Postanalytic and Metacontinental: Crossing Philosophical Divides 9781472547347, 9780826424419, 9781441109996

Citation preview

Contributors

Jack Reynolds is senior lecturer in Philosophy at La Trobe University and author of Merleau-Ponty and Derrida: Intertwining Embodiment and Alterity (Ohio 2004), Understanding Existentialism (Acumen 2006) and co-author of Analytic Versus Continental? (Acumen 2010). James Chase is lecturer in Philosophy at the University of Tasmania and the author of articles in Philosophy and Phenomenological Research, Analysis and the Australasian Journal of Philosophy, and co-author of Analytic Versus Continental? (Acumen 2010). James Williams is Professor of European Philosophy at the University of Dundee. He has written widely on French philosophy, including books on Gilles Deleuze and Jean-François Lyotard. His latest book is Gilles Deleuze’s Logic of Sense: A Critical Introduction and Guide (Edinburgh University Press 2008). Edwin Mares holds a chair in philosophy and is director of the Centre of Logic, Language and Computation at Victoria University of Wellington (New Zealand). He is the author of Relevant Logic: a Philosophical Interpretation (Cambridge 2004) and (with Stuart Brock) Realism and Antirealism (Acumen 2007). Massimiliano Cappuccio is a post-doctoral researcher at the University of Stirling and works in the field of phenomenology applied to cognitive science, publishing on problems concerning mirror neuron theory, motor intentionality and simulation theory applied to social cognition. Pierre Cassou-Noguès is researcher at the CNRS, University of Lille III. His work concerns philosophy of logic with special reference to the epistemological tradition in France and to the role of imagination. He has published De l’expérience mathématique (Vrin 2001) and Les démons de Gödel (Seuil 2007). George Duke has completed a PhD in Philosophy at the University of Melbourne and is a Research Associate at La Trobe University. He is writing a dissertation on Michael Dummett’s theory of abstract entities. He completed a PhD in literature in 1999. Pascale Gillot is Director of program at the Collège International de Philosophie. She has published L’esprit. Figures classiques et contemporaines (CNRS Editions

viii

Contributors

2007) and Althusser et la psychanalyse (PUF 2009). She has co-edited, with Pierre Cassou-Noguès, Le concept, le sujet et la science (Vrin 2009). Simon Glendinning is Reader in European Philosophy at the London School of Economics and Political Science. He is author of The Idea of Continental Philosophy (Edinburgh 2006) and editor of The Edinburgh Encyclopedia of Continental Philosophy (Edinburgh 1999). Dale Jacquette is Senior Professorial Chair in Theoretical Philosophy at the University of Bern, Switzerland. He has recently published Wittgenstein’s Thought in Transition, Ontology, David Hume’s Critique of Infinity and The Philosophy of Schopenhauer. Paul Redding is Professor of Philosophy at the University of Sydney, and is the author of Hegel’s Hermeneutics (Cornell 1996), The Logic of Affect (Cornell 1999), Analytic Philosophy and the Return of Hegelian Thought (Cambridge 2007) and Continental Idealism: Leibniz to Nietzsche (Routledge 2009). Robert Sinnerbrink is lecturer in philosophy at Macquarie University, and Chair of the Australasian Society for Continental Philosophy. He is the author of Understanding Hegelianism (Acumen 2007), and co-editor of Critique Today (Brill 2006) and Work, Recognition, Politics (Brill 2007). Nicholas H. Smith is Associate Professor in Philosophy at Macquarie University. He is the author of Strong Hermeneutics (Routledge 1997) and Charles Taylor (Cambridge 2002). He is also editor of Reading McDowell (Routledge 2002) and a co-editor of Critique Today (Brill 2006) and New Philosophies of Labour (2010). Elena Walsh is an MA candidate in at the University of Melbourne. Her thesis relates Indian Buddhist philosophy and contemporary analytic approaches to relativism and Cartesian scepticism. Michael Wheeler is Professor of Philosophy at the University of Stirling. His primary research interests are in philosophy of science, philosophy of mind and he also works on Heidegger. His book, Reconstructing the Cognitive World: the Next Step, was published by MIT Press in 2005.

Acknowledgements

We’d like to acknowledge the support of our colleagues at La Trobe University, University of Tasmania, University of Dundee and Victoria University at Wellington. We’d also like to recognize the important contributions of: the audience and participants in the ‘Analytic Versus Continental?’ mini-conference (Melbourne, July 2008); those involved in the workshop on transcendental reasoning, including Jeff Malpas and Andrew Benjamin (Melbourne, September 2008); audiences at papers at La Trobe University, Monash University, University of Tasmania, and the 2009 AAP conference. We’d also like to thank the Carnegie Fund for the Universities of Scotland and the Australian Research Council for funding this research project, as well as the associated team: George Duke for so effectively managing to incorporate his research on this project with his various other jobs; Sherah Bloor and Ricky Sebold for many stimulating conversations around these themes; Jo Shiells for editing help, thanks are also due to Ricky and Jon Roffe for compiling the index. On a more personal front, James Chase and Jack Reynolds would like to thank their respective families: Emily, Ada and Max; Jo, Rosa, and a new arrival: Penelope. We are also grateful to all at Continuum for their help in this book.

This page intentionally left blank

Chapter 1

Introduction: Postanalytic and Metacontinental Philosophy Jack Reynolds, James Chase, James Williams and Edwin Mares

Although there is considerable diversity of approach within the fields of what have come to be called analytic and continental philosophy, it is probably fair to say that, since the birth of the analytic movement with Russell and Moore at the start of the twentieth century, they have become increasingly specialized and differentiated fields of endeavour. Anyone who works within academic philosophy is familiar with the distinction between analytic or Anglo-American philosophy and its so-called Continental or European counterpart (overdetermined and vague as the distinction is). Historically the distinction was given weight by a perceived geographical divergence, with the one practised in Anglo-American English-speaking countries and the other in Continental Europe, but these days the geographical differentiation has become far less reliable. It was, of course, problematic from the start. Many of the heroes of early analytic philosophy (most notably Wittgenstein, Frege, Schlick and Carnap) were Austrian or German, and the analytic movement was strongly influenced by early twentieth-century Polish logic and mathematics, for instance; a set of connections that are extremely well known, but that tend to be viewed through the filter of the subsequent migration of many of these philosophers to the Anglophone world. In the other direction, much early twentieth-century philosophy in the United States of America (USA) or United Kingdom (UK) was outside the analytic movement, and straightforwardly Kantian or Hegelian. The analytic dominance in the USA and UK after the Second World War is too easily read back into the earlier twentieth-century philosophical history of both countries, filtering what we notice in the journals, appointments and monographs of the time. Without geography to help us, is the distinction a spurious case of crossclassification, ‘rather as though one divided cars into front-wheel drive and Japanese’.1 Or are there meaningful philosophical differences that can be delineated? The problem is that most attempts to get to the bottom of what it is that separates the analytic and continental traditions of philosophy tend to fall into one of two main camps: an essentialism about the differences between these two traditions that by implication denies the possibility of any meaningful

2

Postanalytic and Metacontinental

rapprochement; and a deflationary response that wants to call into question any value that this distinction, and these terms, might have. It seems to us that both the essentialist and deflationary responses are misleading on their own, and unhelpful for the future of philosophy. Deflationism sheds little light on what is going on in philosophy today under the surface and does not offer an explanation about the very real disdain (which has a philosophical component to it) that many analytic philosophers feel for continental philosophers and vice versa; essentialism ignores the internal diversity of each tradition, the potential for engagement between them, and the places where this is already underway. Innumerable essentialist accounts have been proffered by both analytic and continental philosophers, often dropping out of a general or programmatic conception of philosophy as a whole. If philosophy has to engage in a critique of modernity, or has to embrace its historicity, then it’s tempting to reject analytic philosophy as a faux-neutral and ahistorical game. If philosophy has to adopt norms of argument discourse constrained by formal logical systems or probability theory, or has to be continuous with the natural sciences, then it’s equally tempting to reject continental philosophy as nonsense, mysticism or literature. These sorts of claims are natural enough given the tribal allegiances involved here, and indeed have become something of a rite of passage for some, but they obviously need close examination and equally obviously are at best contestable. Moreover, they are contested, each within the very camp that the essentialist is championing. Analytic philosophers, for instance, vigorously debate the role of logic, the relations between philosophy and science, the worth of standard analytic techniques and so on among themselves – and these are not peripheral debates carried out in minor journals by obscure philosophers, but rather major debates that mark serious divides within analytic philosophy, obvious to participants. So even if one prefers to keep essentialist walls up, they simply don’t quite run where the essentialist wants them to. And, of course, there is evidence that practice undermines such essentialism in any case. After all, there are places (such as ethics and feminist philosophy) where the two traditions come closer and even perhaps sometimes overlap, there are topical areas where significant rapprochement and helpful dialogue has been achieved, and indeed there are examinable methodological preferences bound up with the various different topical and doctrinal commitments made on each side of the divide. Finally, essentialism is to some extent risky. By describing philosophy’s ‘divided house’ in an essentialist manner, one runs the risk of performatively making it more real and oppressive than it is or needs to be – it is in this sense that appeals to the idea of the ‘divide’ can be said to partake in what Simon Glendinning describes as the ‘rotten contemporary scene’ rather than merely to neutrally describe it.2 Deflationary responses assert that we are all simply doing philosophy, with the corollary that we should ignore the so-called analytic and continental distinction. Those who argue that style is the main difference between the

Introduction

3

traditions tend to be deflationary, in that they usually take there to be no significant problem with the translation of the issues and problems that concern the one group to the other. In other words, there is the assumption of a shared referent or explanandum that stylistic differences are thought to merely cover over. Bernard Williams suggests that appeals to the idea of a ‘divide’ involve positing a distinction without a difference, and he intimates that it largely comes down to matters concerning style.3 Graham Priest also suggests that the distinction ultimately comes down to style and consequently advocates a deflationary practice in his comparative work on Hegel, Heidegger, Derrida and others.4 Hilary Putnam asks ‘why can we not just be philosophers without an adjective?’5 Mark Wrathall comments: ‘it strikes me that the best way to overcome the analytic/continental divide is therefore to ignore, at least provisionally, the differences in approach and instead explore the areas of agreement.’6 C. G. Prado, in his ‘Introduction’ to his edited volume on this theme, A House Divided, states: ‘for the most part, this interest has tended to focus on the work of individual philosophers, which certainly is the most productive approach to closing the gap. Generalizations about traditions are less useful than better understanding of the work of particular thinkers.’7 Without wanting to dispute that this may be an effective way to do comparative philosophy between the traditions on particular occasions, there is a risk in such a view as the final judgment on the divide as a whole. Rightly wary of generalizations and caricatures, comparative deflationary accounts might not really plumb the depth of important structural and methodological differences that continue to be part of the contemporary philosophical landscape. Moreover, the risk in over-emphasizing the divide also has an opposite peril in the danger of ignoring it, and in not strategically taking account of it, in academic interactions on social, publishing and career lines. Academia is a powerfully networked form of social organization, where networks are neither homogeneous, hence the divide, nor run according to clear and overtly moral principles, hence the emotional and visceral nature of some engagements around interferences and conflicts between networks. Therefore one of the encouraging and laudable aspects of recent developments across the divide has been the chances taken by many thinkers, notably postgraduate researchers. They have risked careers to investigate connections, yet both the peril and the connectivity are intrinsic values of deep research. However we characterize or dismiss the distinction in theory, in practice it has for many years been very much a feature of the day-to-day activities of contemporary philosophers. Academic philosophers, journals, conferences, publication series and even entire publishing houses, all now often live entirely within one or the other tradition. In some cases, the result is that continental philosophers have effectively been consigned to other disciplines, like comparative literature. More usually, philosophers simply inhabit their own tradition without really attending to the other – perhaps looking at or attending occasional papers from the other side out of collegial politeness or personal loyalty, and

4

Postanalytic and Metacontinental

often regretting it when they do. All the signs are there that many people on each side of the divide simply have nothing to say to each other. The divide has also ramified throughout the social sciences and humanities. It is familiar wherever a confirmation theorist meets a constructivist, or a post-structuralist meets a positivist, in myriad debates about the significance of Foucault or the scope of covering law explanation (or indeed in the careful avoiding of such encounters). Arguably, the analytic/continental divide has underwritten methodological incomprehension and rejection between different camps in sociology, history, anthropology, literary theory, archaeology and many other fields. That is, at least until recently, for some developments in philosophy and associated fields suggest that perhaps the house is no longer quite so divided (or at least that its inhabitants are prepared to sort through the rubble). Essays devoted to the topical, methodological and stylistic differences between analytic and continental philosophy have begun to proliferate. Comparative books on specific representative authors from each tradition abound, and there are recent books examining central continental philosophers and their relationship to, and differences from, analytic philosophy – for instance, Richard Cobb-Stevens’ Husserl and Analytic Philosophy, Robert Hanna’s Kant and the Foundations of Analytic Philosophy, Samuel Wheeler’s Deconstruction as Analytic Philosophy and Paul Redding’s Analytic Philosophy and the Return of Hegelian Thought (see bibliography for details). A number of prominent philosophers trained in the analytic tradition have been perceived within that tradition to have jumped the fence, and have been taken up recently by continental philosophers – for instance, Robert Brandom, Charles Taylor, Donald Davidson and John McDowell, all of whom are given sustained consideration in this book. Metaphilosophical books on each tradition have been recently published, including Hans-Johann Glock’s What is Analytic Philosophy? and Glendinning’s The Idea of Continental Philosophy. Several journals have appeared in the past twenty years that include in their remit the goal of bridging the divide. Signs of the influence of these philosophical developments can be seen in wider fields, such as literature (Deleuze and postanalytic philosophy), politics (Žižek and Badiou), theology (new developments in phenomenology, analytic metaphysics), film (Rancière, Deleuze, Cavell, Mulhall) and aesthetics (analytic work on fiction, continental aesthetic critique and art criticism). Such metaphilosophy perhaps indicates that both traditions are in a period of transformation, no longer taking for granted the methods and conception of the value of philosophy that has been dominant throughout the twentieth century. Perhaps. Quite apart from the fact that many such encounters end in rancour rather than rapprochement, one might also think that this diversity only really occurs at the margins, and that the majority of philosophers ignore or avoid it. Without thinking that we might have turned the page on a distinction that has had such socio-political force within academic philosophy, this edited collection aims to explore this trend. Chapter 2, by George Duke et al., sets the scene for what follows, in that it tests the extent to which the alleged

Introduction

5

‘crossover’ figures of Wittgenstein, Davidson, Rorty and McDowell actually serve to bridge the divide. (Do analytics and continentals read the same Davidson? If they do, do they read each other as well, or is the relevant text merely a common object looked at from different places?) The chapters of Part One then look at the role played by forms of reasoning, or background presuppositions about argument, on one or the other side. Sometimes the result is a justification of practice that has at best an ‘internal validity’, precisely because of differences in presupposition on each side of the divide. James Chase and Jack Reynolds (Chapter 3) examine the differing reception of transcendental argument, linking this to differing background commitments to argument form, on the one side, and interests in ‘situated thought’ and projects of anteriority, on the other. James Chase (Chapter 6) argues for a background epistemic conservatism in analytic philosophy, and suggests that arises as a result of a commitment to dialogic norms concerning inferential connectivity of research. Sometimes, on the other hand, the project is to find a kind of common ground. Edwin Mares (Chapter 4) looks at the relative priority of logic and metaphysics for Dummett and Heidegger and finds a middle way in understanding the development of both a metaphysics and a logical system in reflective equilibrium. Simon Glendinning (Chapter 5) argues against glib appeals to the analyticcontinental distinction and instead makes ‘effective propaganda’ for the view that narrow argument cannot be the exclusive mode of inheritance of philosophy, especially if philosophy aims to cultivate an understanding of ourselves. The chapters of Part Two concern philosophy of mind, and here the emphasis is on convergence. Pascale Gillot (Chapter 7) suggests through a close comparison of John Searle and Descartes that a set of shared and inherited assumptions limit the extent to which Searle (and other contemporary philosophers of mind) have distanced themselves from Cartesianism. Massimiliano Cappuccio and Michael Wheeler (Chapter 8) use Hubert Dreyfus and Heidegger’s philosophy, as well as recent work on the frame problem, to outline the extent to which the cognitive sciences are becoming a meeting place between the traditions, with benefits for both. In Part Three, concerning meaning, expression and aesthetics, the focus is on specific thinkers who have been characterized as postanalytic and metacontinental. Nick Smith (Chapter 9) examines the differing forms of expressivism appearing in work by Robert Brandom and Charles Taylor, two different ways in which the representationalism standard in analytic philosophy has been departed from. Dale Jacquette (Chapter 10) presents a reading of Wittgenstein as a ‘trans-analytic-continental’ philosopher, rejecting the standard understandings of his work in both traditions. Robert Sinnerbrink (Chapter 11) argues that a philosophical disenfranchisement of film has arisen as a result of the ‘cognitivist turn’ away from the continentalinspired film theory of the 1970s and 1980s, and to the work of analytic philosophers like Noël Carroll. Finally, in Part Four, concerning metaphysics and mathematics, Paul Redding (Chapter 12) examines the recent ‘Hegelian turn’ in analytic philosophy, evident in Brandom and John McDowell, both looking

Postanalytic and Metacontinental

6

at how each found his way to Hegel and at the possibilities of an analytic philosophy informed by German idealism. James Williams (Chapter 13) finds similarities in a Deleuzian and a Davidsonian account of events, again exhibiting the possibilities of engagement between two distinct metaphysical traditions while also being attentive to what separates them. And Pierre Cassou-Nogués (Chapter 14) uses a Deleuzian framework to find parallels between the forms of mathematical Platonism adopted by Albert Lautman and Kurt Gödel. It seems to us that substantive engagement in this way is desirable, even if it runs very much against the grain of academic specialization in each tradition. We feel that some kind of conversation – even if antagonistic – is necessary in order for each tradition to avoid some of the weaknesses of complacency that can arise with insulation from engagement with a philosophical ‘other’. While no long-term coming together or union is foreseeable, we can at least hope for ongoing encounters, for not being blinded by the allure of one particular approach, and for the modest pluralist benefits that follow from being forced to look at one’s own presuppositions through sceptical eyes. Let us attempt to leave behind our habitual specializations and confident assertions of knowledge, and revisit the question of philosophy’s heritage and its future.

Notes 1

2

3

4

5 6

7

Bernard Williams, ‘Contemporary Philosophy: A Second Look’, The Blackwell Companion to Philosophy, N. Bunnin and E. P. Tsui-James eds (Oxford: Blackwell, 1996), 25. Simon Glendinning, The Idea of Continental Philosophy: A Philosophical Chronicle (Edinburgh: EUP, 2006). 7. Bernard Williams, ‘Preface’, Ethics and the Limits of Philosophy (Cambridge: Cambridge University Press, 1985). Graham Priest, Beyond the Limits of Thought (Oxford: Oxford University Press, 2003), 237. Hilary Putnam, ‘A Half Century of Philosophy’, Daedalus 12 (1997), 203. Mark Wrathall, ‘The Conditions of Truth in Heidegger and Davidson’, The Monist 82 (1999), 304–323 at 306. C. G. Prado, ‘Introduction’, A House Divided: Comparing Analytic and Continental Philosophy, C. G. Prado ed. (Amherst, NY: Humanity Books, 2003), 9.

Chapter 2

‘Postanalytic’ Philosophy: Overcoming the Divide? George Duke, Elena Walsh, James Chase and Jack Reynolds

1. Introduction The notion of the ‘postanalytic’ has been appealed to as evidence of the increasing irrelevance of the ‘analytic-continental’ divide in contemporary philosophy.1 Such a claim presupposes that there are sufficient thematic and methodological continuities in the work of putatively postanalytic thinkers to endow the notion with a determinate meaning, and that this work has made a significant contribution to the bridging of the ‘analytic-continental’ divide. The purpose of this chapter is to assess these claims through an examination of the extent to which the work of four philosophers who could plausibly be regarded as postanalytic – Wittgenstein, Davidson, Rorty and McDowell – may be characterized as representative of a movement which has promoted increased dialogue between the analytic and continental traditions.2 It will be argued that there are indeed continuities between the relevant thinkers suggestive of a broad notion of the postanalytic, but that it would be an exaggeration to speak of a unified philosophical movement embodying a rapprochement. The term ‘postanalytic’ has been used to characterize the work of thinkers who, having started out in the mainstream analytic tradition, came to place in question some of its central presuppositions.3 Postanalytic philosophy in this overarching sense is often associated with the work of Richard Rorty, a central figure in the collection of essays devoted to the topic edited by John Rajchman and Cornel West.4 Other prominent figures sometimes regarded as postanalytic, in large part due to Rorty’s influence, include Brandom, Davidson, McDowell, Putnam and Wittgenstein. From this perspective, postanalytic philosophy is a ‘deconstructive’ approach defined by what it stands in reaction to, namely a particular characterization of the goals of analytic thought. Rorty has tended to identify the overarching aspiration of analytic philosophy in particular with a ‘transcendental project’5 to provide certain epistemological foundations for science and human practice in general, an aspiration he vehemently rejects in favour of a pragmatic concern with ‘solidarity’ or ‘consensus’. According to Rorty, a postanalytic philosopher works on the assumption that philosophy

8

Postanalytic and Metacontinental

does not have a privileged cognitive role over and against other forms of human knowledge and artistic expression. Of particular concern in this regard to Rorty, and many of the other philosophers in Rajchman and West’s collection, is what they regard as the pernicious scheme-content dualism that is concomitant with the epistemological quest for foundations. Scheme-content dualism is, for Rorty, what allows knowledge to be understood as consisting in a ‘fit’ between the mind’s representations of the world, and the world as it exists in and of itself. According to Rajchman, it is a rejection of this representationalist paradigm, one that has held philosophy ‘captive’ since Kant,6 which unites the postanalytic philosophers.7 While this account of the postanalytic in terms of a pragmatic anti-Kantianism is arguably overly restrictive (it sits uncomfortably, for example, with the case of McDowell), a similar but more inclusive working characterization can be found in Putnam’s sketch of the history of twentieth-century analytical philosophy in his 1985 essay ‘After Empiricism.’8 According to Putnam, the failed attempts of Frege, Russell, Carnap and the early Wittgenstein to give an account of the structure of reality through logico-linguistic analysis led to an increased self-consciousness regarding the difficulties associated with giving an account of the way words hook onto the world. This led in turn to a placing in question of ‘the entire enterprise’ of dividing mundane reality into the Furniture of the World and our projections.9 While such a critical approach may be prosecuted through anti-Kantian pragmatism of the Rortyian variety, it is also compatible with the use of a Kantian framework with therapeutic intent of the kind found in the work of McDowell. In this context, the postanalytic refers to a broad movement of thought which has placed in question some of the guiding presuppositions of analytic philosophy in its ambitious, classical phase of logico-linguistic analysis; in particular representationalism and the ‘third dogma of empiricism’ embodied in scheme-content and subjective-objective dualisms. There are, however, competing assessments of the status of the postanalytic as a coherent philosophical movement. While Rajchman and West see a deep thematic continuity uniting the postanalytic thinkers in their collection, and regard the combined attack of such thinkers against the analytical tradition as paving the way for a new sort of philosophy, Pascal Engel, by contrast, thinks this assessment is premature: [Rorty’s] appropriation of Quinean, Davidsonian, Sellarsian, Goodmanian and Putnamian themes should raise our suspicions, although he is certainly right in believing that these philosophers have undermined a lot of themes in classical analytic philosophy. But there is no reason to think that the various criticisms of their peers that these philosophers have produced transport us into a new scene in which analytic doctrines and styles are abandoned.10 While accepting the thematic continuity of the postanalytic thinkers, and their legitimate criticisms of central aspects of analytical philosophy, Engel denies

Postanalytic Philosophy: Overcoming the Divide?

9 3650

2061

Davidson 1957 to 2008 (Baseline 19, 965)

843

883

McDowell 1970 to 2008 (Baseline 15, 254)

Rorty 1961 to 2008 (Baseline 18, 673)

Wittgenstein 1921 to 2008 (Baseline 27, 352)

Figure 2.1 Citation of postanalytic figures (Articles citing each postanalytic figure in 15 leading journals of analytic philosophy)* *

Journals listed in note 11.

that this is indicative of the creation of a new philosophical movement embodying a wholesale rejection of analytic themes, doctrine and style. This assessment is consonant with the fact that an analysis of citation patterns for Wittgenstein, Davidson, Rorty and McDowell suggests that, despite their alleged polemical attitude towards the analytical tradition, they remain among the most frequently cited philosophers in major analytic journals during the twentieth-century (see Figure 2.1).11 These contrasting positions suggest a weaker and a stronger characterization of postanalytic philosophy. According to the weaker characterization, the postanalytic refers to the work of philosophers who have criticized aspects of mainstream analytical philosophy. According to the stronger characterization, the postanalytic represents a radical form of critique that culminates in the dismissal of the analytic tradition as a failed project12 and hence suggests the need for the creation of a new sort of philosophy. The weaker characterization, it could be argued, is in danger of rendering the notion of the postanalytic anodyne, in that it suggests a form of criticism that is indistinguishable from the general process of questioning and argument that is constitutive of philosophical debate itself. From this perspective, it is difficult to see what would differentiate the postanalytic thinkers in question, insofar as analytic philosophers put forward theses that are then subjected to subsequent criticism by other analytic philosophers, whose intent is to attack the thesis in question, and even some of the presuppositions upon which it is built, but not the coherence and relevance of the discipline itself. Implicit in the stronger characterization, however, is the implausible claim that the postanalytic represents a completely new sort of philosophy entailing a wholesale rejection of broad analytic preoccupations, method and style.

10

Postanalytic and Metacontinental

The first section of this chapter argues for a notion of the postanalytic intermediate between the weaker and stronger characterizations in the context of a brief overview of Wittgenstein, Davidson, Rorty and McDowell’s criticisms of some central assumptions of classical analytical philosophy. Despite significant thematic unity around the need to overcome a dualistic conception of the relationship between the ‘given’ and ‘that which is added by the mind’,13 these philosophers have very different views on the consequences of this critique for analytical philosophy as a discipline, to an extent that could be thought to call into question the viability of the postanalytic label itself. In order to differentiate the weaker characterization of the postanalytic from the general process of critique that takes place within the analytic tradition, it will be necessary to point to the radical nature of the questioning of conceptual presuppositions suggested by the work of postanalytic thinkers; a radical questioning which implies a rejection not only of scheme-content and subjective-objective dualisms but also some of the accompanying methodological and stylistic assumptions of analytical philosophy in its classical phase. The second section of the chapter considers the relationship between postanalytic philosophy and continental thought. The figures here examined – Wittgenstein, Davidson, Rorty and McDowell – have all been identified as potential ‘crossover’ figures: thinkers who have, self-consciously or otherwise, contributed to ‘bridging the divide’ between the analytical and continental traditions. A quantitative analysis of citation patterns of these thinkers in continental journals suggests that they are indeed engaged with by the continental tradition. It will be argued, however, that such engagement is insufficient to establish these thinkers as crossover figures in a strong sense. The fact that analytic and continental scholars both engage with the work of Davidson or Wittgenstein, for example, is inconclusive until it is demonstrated that this secondary work is based on shared thematic concerns and methodological assumptions and that there are significant intersecting bodies of literature which go beyond the mere mention of the relevant figures. A preliminary analysis of cross-citation patterns in the commentarial bibliography of each thinker’s central works suggests that there is little engagement of this nature occurring, despite the fact that each thinker is cited regularly in both traditions. This suggests the need for a cautious assessment of the extent to which the four thinkers in question have contributed towards a genuine rapprochement.

2. The Postanalytic In his What is Analytic Philosophy? Hans-Johann Glock argues persuasively against attempts to provide a one-sentence characterization of analytic philosophy that would definitively capture its essence. This is not because the notion of analytic philosophy lacks a determinate meaning altogether, but rather because it represents a broad and divergent tradition held together by family resemblances

Postanalytic Philosophy: Overcoming the Divide?

11

rather than a specific doctrine.14 An examination of the work of Wittgenstein, Davidson, Rorty and McDowell suggests the need for a similarly nuanced account of the notion of the postanalytic. As is argued below, the postanalytics share a critical attitude towards some of the major conceptual presuppositions of the classical mainstream analytic tradition, including a narrowly ‘representationalist’ account of the way words hook onto the world and scheme-content dualism. Concomitant with this radical questioning, the work of all four thinkers raises broad concerns about the analytical project as a whole, and its predominant style and method, in a way that differentiates their thought from the mainstream analytic tradition, despite the fact that they diverge not only in their own stylistic and methodological assumptions, but also in their accounts of the significance and consequences of these concerns for contemporary philosophy. In his Philosophical Investigations Wittgenstein criticizes what he takes to be St. Augustine’s naïve conception of language, according to which individual words name objects and sentences are combinations of such names.15 It is in reaction to this view of language, attributable to his own picture theory of meaning as outlined in the Tractatus, that much of Wittgenstein’s later work is directed. Here Wittgenstein suggests that, in order to understand meaning, we must primarily look to the way a word is used, rather than towards the object that a word represents.16 This use-based account of language characterizes particular applications of a word in terms of ‘family resemblances’ determined by observation of our linguistic practices. Wittgenstein’s use-based account of meaning, consistent with Rorty’s reading of the notion of the postanalytic, thus rejects a ‘representationalist’ theory of language in favour of a broadly pragmatic and instrumental view.17 In linking meaning to use, Wittgenstein also rejects some of the methodological and stylistic assumptions of his earlier work, which are suggestive of the possibility of an independent objective stance on our linguistic practices. Although some interpreters have taken his position to culminate in a form of epistemological subjectivism,18 in parts of On Certainty it is clear that Wittgenstein thinks of the world as ‘the inherited background against which I distinguish between true and false’.19 As a precondition for intelligible discourse and thus for knowledge itself, the world should not, on this view, be understood on the basis of a dichotomy between the subjective and objective. Like Wittgenstein, Davidson attempts to free philosophy from the view that language is the via media through which the mind comes to represent the world. In his influential paper ‘On the Very Idea of a Conceptual Scheme’, Davidson suggests that such a position implies that there is a sharp distinction between the raw data provided by sensory experience, and a conceptual scheme, seen as either representative of or identical with thought, which provides structure to empirical content. Davidson argues that this dualism is incoherent,20 and holds that, in renouncing the dualism of content and scheme, we ‘do not give up the world’ but instead ‘reestablish unmediated touch with the familiar objects whose antics make our sentences and opinions true or false’.21 Against

12

Postanalytic and Metacontinental

this background, Davidson seeks to develop a systematic and constructive philosophical project, which is to explain how knowledge of the world is possible, via an account of what it is in virtue of which linguistic meaning consists. For Davidson, knowledge is essentially intersubjective, arising not from uninterpreted sense data, but from communicative practices that occur between individuals, and the shared experiences that make acts of interpretation possible.22 In this sense, it is possible to connect Davidson’s work with a broadly pragmatic outlook that replaces representationalism with an holistic and instrumentalist view of knowledge. Both Davidson and Wittgenstein could be labelled postanalytic thinkers insofar as their rejection of the dichotomy between the subjective and the objective places in question, at least implicitly, the very cogency of the classical analytic project. The two stand apart, however, in their broad conception of the purpose of philosophical activity. Wittgenstein’s critique of representationalism is connected with a view of philosophy according to which it may not interfere with the actual use of language, nor provide a foundation for it, but only ‘put everything before us’ in such a way that there is nothing left to explain.23 The philosopher on this view is a kind of therapist who does not answer questions constructively, but rather seeks to uncover the conceptual presuppositions that make such questions intelligible in the first place. Davidson, by contrast has an explicitly constructive project, namely that of explaining how we may have knowledge of our own minds, knowledge of the minds of others, and knowledge of a shared world. For Davidson, it is ‘up to philosophy to talk about the relations between the three kinds of knowledge, how they differ, and how they interdepend’.24 In this context his project is neither quietistic nor does it seek to overcome philosophy as conventionally understood by its analytic practitioners.25 In Philosophy and the Mirror of Nature, Rorty draws the work of both Wittgenstein and Davidson together in a narrative that describes how philosophy has been held captive by the metaphor of the mind as a mirror that reflects reality.26 Rorty supports Wittgenstein’s denial that meaning consists in representation, and follows Davidson’s rejection of a dualism between content and scheme. For Rorty, adherence to this duality is problematic insofar as it demands that the ‘given’ exert a normative constraint upon the ‘space of reasons’. This is untenable from Rorty’s perspective because the world may provide only a causal (and not normative) influence upon the mind.27 Rorty’s claim that it was the task of the traditional philosopher to explain how accurate representation of the world is possible28 suggests, however, without a dualism of content and scheme, the entire legitimacy of epistemology (and, as a consequence, of philosophy) is undermined. Philosophy’s certification as a foundational discipline, one both ‘distinct from and sitting in judgment upon both religion and science’,29 is thus rejected by Rorty, who proposes a redefinition of philosophy as ‘social and cultural criticism’,30 the philosopher acting as a sort of ‘Socratic intermediary between various discourses’, who approaches cultural and social criticism with

Postanalytic Philosophy: Overcoming the Divide?

13

a hermeneutic spirit, rather than seeking to ‘get behind’ our chosen vocabularies to establish the veridicality of our claims to knowledge.31 While he connects Wittgenstein with his own therapeutic pragmatism, by claiming that both think of language as a tool, rather than a mirror that represents reality,32 Rorty’s position is accompanied by a fin de siècle rhetoric and explicitly political focus which differentiates him from the other thinkers here under examination. McDowell approaches the problematic dualism between subjectivity and objectivity by providing an account of how the modern scientific revolution has concealed both our ‘openness’ to the world and ‘second nature’ as rational animals. Rejecting the assumption that we should understand our perceptual experience in terms of modern natural-scientific ‘laws of nature’ or causality, McDowell argues that the perceptual experience of the world by normal mature humans is always already informed by conceptual content. McDowell’s account differs from that of the other postanalytic figures considered in this chapter insofar as he would seem to provide a minimal endorsement of a representationalist framework. Though McDowell defends Kant’s view that empirical knowledge results from an intermingling of receptivity and spontaneity, however, his account marks a departure from Kant insofar as our conceptual capacities are not exercised upon receptivity, but instead drawn on in receptivity itself, as a precondition for the possibility of perceptual experience.33 Thus, McDowell’s analysis of the mind’s active employment of concepts in perceptual experience allows him to argue against external world scepticism, developing a reading of Kant which overcomes the oscillation between ‘the myth of the given’ on the one hand and the ‘frictionless spinning in the void’ of coherentism on the other.34 McDowell’s work has affinities with Wittgenstein and Rorty, insofar as it seeks to overcome the oscillation between coherentism and the myth of the given, and holds that this requires a non-coercive deconstruction of the philosophical anxieties that gave birth to this dualism.35 According to McDowell’s naturalized Platonism, the structure of the space of reasons has a kind of autonomy in the sense that it is ‘not derivative from, or reflective of, truths about human beings that are capturable independently of having that structure in view’, even though it is ‘not constituted in splendid isolation from anything merely human’.36 McDowell’s work also, like that of Wittgenstein and Rorty, attempts to demonstrate that ‘the supposed obligations of traditional philosophy are illusory’.37 McDowell writes: If we could achieve a firm hold on a naturalism of second nature, a hold that could not be shaken by any temptation to lapse back into ordinary philosophical worries about how to place minds in the world, that would not be to have produced a bit of constructive philosophy of the sort Rorty aims to supersede. In Wittgenstein’s poignant phrase, it would be to have achieved ‘the discovery that gives philosophy peace’.38

14

Postanalytic and Metacontinental

Although McDowell here provides explicit endorsement of both Wittgenstein’s quietism and Rorty’s desire to expose the constructivist conceptual presuppositions that gave rise to the fundamental problematic of modern epistemology,39 his advocacy of a form of transcendental empiricism40 and incorporation of the Aristotelian notion of second nature differentiates his thought from that of the other postanalytic philosophers under consideration. This brief assessment of the status of Wittgenstein, Davidson, Rorty and McDowell as postanalytic thinkers suggests that there is indeed a broad critical or ‘deconstructive’ project uniting their work that is deeper and more radical than the standard process of debate and critique that has taken place within the mainstream analytic tradition. In placing in question some of the guiding presuppositions of the analytical tradition, the work of these thinkers engages not only with a broader set of concerns than most analytic thought, but also suggests the inadequacy of some of that tradition’s standard methodological and stylistic resources. Wittgenstein, Davidson, Rorty and McDowell nonetheless differ between themselves not only in terms of method and style, but also in their view of the consequences and significance of the necessary ‘deconstruction’ of contemporary philosophy. These divergences can be seen more clearly through a consideration of the status of these thinkers as ‘crossover’ figures in relation to the continental tradition.

3. Postanalytic philosophy and continental thought In order to assess whether the work of allegedly postanalytic thinkers such as Wittgenstein, Davidson, Rorty and McDowell has genuinely contributed towards an overcoming of the divide between analytical and continental philosophy, it is necessary to consider their status as ‘crossover’ figures. One plausible strong indicator that the four philosophers under examination have contributed towards such a rapprochement would be evidence that their work has led to sustained and informed scholarly debate on shared philosophical concerns across both sides of the divide. This indicator of cross-traditional engagement needs to be distinguished from two related, but weaker indicators: (i) that the work of the postanalytic thinkers in question is ecumenical in the sense that it draws influence from both traditions, and (ii) that the work of the postanalytic thinkers is discussed in depth by both analytic and continental scholars. The following account, which draws on detailed analysis of journal citation patterns,41 suggests that an unqualifiedly positive assessment of the status of Wittgenstein, Davidson, Rorty and McDowell as crossover figures would be premature. Let us take Davidson as a first example. Davidson’s account of knowledge as intersubjective, as something that can only be understood once a dualism of scheme and content has been abandoned, depends, in part, upon his theory of radical interpretation, along with his view that our best understanding of truth is captured via the notion of translation. It is these broad themes – which underlie

Postanalytic Philosophy: Overcoming the Divide?

15

the sense in which Davidson may be thought of as a postanalytic philosopher – that have inspired continental scholars to relate his thought to Heidegger, Gadamer and Derrida.42 Furthermore, articles which focus primarily upon one or more aspects of Davidson’s thought did indeed begin to appear in continental journals from 1987 onwards (see Figure 2.2). This suggests that Davidson is a crossover figure insofar as his work appeals to both analytic and continental scholars (see Figure 2.3). It is, however, less straightforward to determine whether Davidson is a crossover figure in terms of providing a focal point for debate between contemporary analytic and continental scholars.

90 80 70 60 50 40 30 20 10 0 1960s

1970s

Analytic

1980s Crossover

1990s

2000+

Continental

Baseline of 148,570 articles, 233 betray a dominant or predominant focus on Davidson and are categorized above by journal type and date published.

Figure 2.2 Davidson’s reception (Journal articles published on Davidson by decade, 1963 to 2008)

150 67 16

Analytic

Crossover

Continental

Baseline of 148,570 articles, 233 betray a dominant or predominant focus on Davidson and are categorized above.

Figure 2.3 to 2008)

Davidson’s reception (Journal articles published on Davidson, 1963

16

Postanalytic and Metacontinental

A preliminary cross-citation check yields no evidence that analytic philosophers are engaging with scholarly work on Davidson by continental philosophers, and vice-versa.43 Additionally, while Davidson’s articles cited in analytic journals concern his more technical and constructive projects within analytic philosophy of action, mind and language,44 those cited in crossover journals tend to deal with his deconstructive critique of scheme-content dualism, and the account of truth and interpretation upon which it depends.45 This suggests that Davidson’s work has provided only limited impetus for debate between analytic and continental scholars over shared philosophical themes or the nature of his work in general. The status of McDowell as a crossover figure is on the face of it more compelling. McDowell’s early work within analytical philosophy was known only to a small group of professional philosophers working within the Anglo-American tradition.46 The publication of Mind and World, however, introduced McDowell’s thought to a broader philosophical public, leading to increased citation of his work by both analytical and continental scholars (see Figures 2.4 and 2.5).47 McDowell’s work has inspired interest from continental interpreters over the problematic character of naturalism and the status of ‘second nature’,48 the

24

Analytic

2

0

Crossover

Continental

Baseline of 41,810 articles, 26 published before 1994 betray a dominant or predominant focus on McDowell and are categorized above.

Figure 2.4 McDowell’s reception (Journal articles published on McDowell before Mind and World, 1980 to 1993)

72

74

16

Analytic

Crossover

Continental

Baseline of 81,134 articles, 162 published before 1993 betray a dominant or predominant focus on McDowell and are categorized above.

Figure 2.5 McDowell’s reception (Journal articles and book Chapters published on McDowell after Mind and World, 1994 to 2008)

Postanalytic Philosophy: Overcoming the Divide?

17

120 64

Analytic

60

Crossover

Continental

Baseline of 138,257 articles, 244 betray a dominant or predominant focus on Rorty and are categorized above.

Figure 2.6 Rorty’s reception (Journal articles published on Rorty, 1972 to 2008)

status of transcendental arguments,49 and on the proper interpretation of Kant and Hegel.50 Conversely, these aspects of the work of McDowell have alienated some of the more parochial adherents of analytic philosophy. Crispin Wright, for instance, is critical of the ‘therapeutic’ tone of Mind and World, one which he regards as characterized by ‘stylistic extravagance’ and ‘rhetorical metaphysics.’51 Moreover, even though there is evidence that McDowell’s work is discussed by both philosophical traditions, a preliminary cross-citation check once again yields little proof of substantial cross-traditional engagement. This again suggests that while both analytical and continental interpreters deal with McDowell’s thought, in many cases they approach his work from sufficiently different perspectives to render his status as a crossover figure somewhat ambiguous. Rorty’s work, like that of McDowell, has yielded a large number of articles published in both analytic and continental journals (see Figure 2.6). Some of these articles demonstrate that Rorty could legitimately be regarded as a crossover figure, insofar as they highlight affinities between analytic and continental thought on the basis of a pragmatic view of knowledge as resting upon ‘consensus’ rather than representation.52 Continental scholars have also engaged with Rorty’s suggestion that philosophy be replaced with social and cultural critique, and with his readings of continental philosophers like Nietzsche, Heidegger and Derrida.53 While Philosophy and the Mirror of Nature is still heavily cited within analytic journals,54 however, much of this work is concerned with criticizing Rorty’s alleged epistemological relativism and anachronistic or inaccurate understanding of figures in the history of philosophy. This suggests that Rorty has served less as a focus for genuine cross-traditional engagement on philosophical issues, than as a common focal point for critical assessments of his interpretative approach and supposed relativism.55 Wittgenstein presents a more convincing case of an allegedly postanalytic philosopher who has contributed to a genuine rapprochement. As would be expected, Wittgenstein is heavily cited in both analytic and continental traditions (see Figures 2.7 and 2.8),56 and there is evidence that his work deals with

Postanalytic and Metacontinental

18

630 468 182

Analytic

Crossover

Continental

Baseline of 156,843 articles, 1,280 published betray a dominant or predominant focus on Wittgenstein and are categorized above.

Figure 2.7 Wittgenstein’s reception (Journal articles published on Wittgenstein, 1951 to 2008)

341

159

134 63

Analytic

Crossover/Continental

Baseline of 8,905 journal articles.

Baseline of 3,688 journal articles.

TLP

PI

Figure 2.8 Reception of early and late Wittgenstein (TLP and PI citations in 4 leading journals of analytic philosophy, 1926 to 2007 and 5 leading journals defined as ‘crossover’ or continental, 1954 to 2007)* *Journals listed in note 57.

philosophical concerns common to the two traditions.57 Furthermore, a preliminary cross-citation check suggests a small amount of genuine cross-traditional engagement, centring in particular upon the so-called therapeutic reading of Wittgenstein championed by Cora Diamond and James Conant. This interpretation sees Wittgenstein rejecting the notion of an external standpoint on language as entirely nonsensical, while asserting that abandonment of this ideal will not leave philosophy defenceless against the Cartesian sceptic. Genuine engagement between analytic and continental thinkers has been inspired by this unorthodox reading, including investigations of the relationship between Wittgenstein and thinkers like Kant, Levinas and the Frankfurt School, and an analysis of the way in which this therapeutic interpretation relates to more traditional interpretations of Wittgenstein within analytic thought.58

Postanalytic Philosophy: Overcoming the Divide?

19

This brief analysis has demonstrated that there are legitimate grounds for thinking that Wittgenstein, Davidson, Rorty and McDowell are crossover figures insofar as they draw on diverse influences and have attracted significant scholarly attention from both the analytic and continental traditions. In their rejection of some of the major conceptual presuppositions of the mainstream analytic tradition, the figures examined have not only attracted the interest of continental thinkers, but have also provided some common ground for crosstraditional reflection on shared themes and concerns. It would nonetheless be premature to assert on this basis that there is sufficient evidence to prove that Wittgenstein, Davidson, Rorty and McDowell are crossover figures in the strong sense of providing a focal point from which sustained and informed scholarly debate between the analytic and continental traditions has taken place.

Notes 1

2

3

4

5

6 7 8

Ray Monk for example describes postanalytic philosophy as ‘a conception of philosophy that seeks to retain the virtues of the analytic tradition – rigor, clarity, intellectual honesty – while seeking to broaden the scope of the issues and texts customarily dealt with by analytic philosophers, and, in particular, to overcome the disastrous gulf between analytic philosophers and the Continental tradition’. Quoted in Deane-Peter Baker and Patrick Maxwell, eds, Explorations in Contemporary Continental Philosophy of Religion (Amsterdam: Rodopi, 2003), 8. For examples of the claim that the work of these four indicates rapprochement or the redundancy of the divide, see Chris Lawn, Wittgenstein and Gadamer: Towards a Post-Analytic Philosophy of Language (London: Continuum Press, 2005), 3; Jeff Malpas, Davidson and the Mirror of Meaning (Cambridge: Cambridge University Press, 1992), 2–3; James Tartaglia, Routledge Philosophy Guidebook to Philosophy and the Mirror of Nature (London: Routledge, 2007), 14–19; Richard Bernstein, ‘McDowell’s Domesticated Hegelianism’, in Reading McDowell: On Mind and World, ed. Nicholas Smith (London: Routledge, 2002), 9. The term ‘postanalytic’ is not established in the philosophical lexicon. A literature review reveals only 32 sources between 1975 and 2007 making use of it, and many of these discuss the viability of the category. See John Rajchman and Cornel West, eds, Post-Analytic Philosophy (New York: Columbia University Press, 1985). The expression here has a paradigmatically Kantian sense, referring to the alleged analytic attempt to put philosophy ‘on the secure path of a science’ by ‘putting outer space inside inner space (the space of the constituting activity of the transcendental ego) and then claiming Cartesian certainty about the inner for the laws of what had previously been thought to be outer.’ See Richard Rorty, Philosophy and the Mirror of Nature (Princeton, Princeton University Press, 2008), 137. Rorty, Philosophy and the Mirror of Nature, 7–13. Rajchman and West, Post-Analytic Philosophy, xiii. Hilary Putnam, ‘After Empiricism’, in Post-Analytic Philosophy, ed. John Rajchman and Cornel West (New York: Columbia University Press, 1985), 28–29. Cf. Robert

20

9 10

11

12

13

14

15

16 17

18

19 20

21 22

23 24

Postanalytic and Metacontinental Brandom’s 2005–2006 John Locke Lectures on the ‘pragmatist challenge’ to the ‘semantic logicist’ assumptions of analytical philosophy in its ‘classical’ phase. Ibid. Pascal Engel, ‘Is There Any Such Thing as Post-Analytic Philosophy’, in Conférence Clermont Ferrand Société de philosophie d’Auvergne (1995), 18. A search for total citations of each postanalytic figure was conducted via JSTOR between the date of each thinker’s first publication and 2008. The following journals were searched: American Philosophical Quarterly, Analysis, Erkenntnis, The Journal of Philosophy, Mind, Noûs, Philosophical Issues, Philosophical Perspectives, The Philosophical Quarterly, The Philosophical Review, Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, Philosophy, Philosophy and Phenomenological Research, Proceedings of the Aristotelian Society and the Proceedings of the Aristotelian Society Supplementary Volumes. For example, Rorty’s claim that ‘abandoning the scheme-content distinctions and accepting pragmatism does, in a sense, mean abandoning philosophy’. See Richard Rorty, ‘Transcendental Arguments, Self-Reference, and Pragmatism’, in Transcendental Arguments and Science, ed. Peter Bieri Rolf-Peter Horstmann and Lorenz Krüger (Dordrecht: R. Reidel, 1979), 78. Putnam writes in the same context that ‘the enterprises of providing a foundation for Being and Knowledge – a successful description of the Furniture of the World or a successful description of the Canons of Justification – are enterprises that have disastrously failed.’ See Hilary Putnam, Realism with a Human Face (Cambridge MA: Harvard University Press, 1990), 19. On the other hand, Rorty appears occasionally to contradict this stronger characterization, as when he says that ‘I think that analytic philosophy can keep its highly professional methods, the insistence on detail and mechanics, and just drop its transcendental project.’ See Richard Rorty and Eduardo Mendieta, Take Care of Freedom and Truth Will Take Care of Itself: Interviews with Richard Rorty (Stanford: Stanford University Press, 2006), 23. In Rorty’s words, alluding to Sellars’ critique of the ‘Myth of the Given’. See Rorty, Philosophy and the Mirror of Nature, 169. Hans-Johann Glock, What Is Analytic Philosophy? (Cambridge: Cambridge University Press, 2008). See Ludwig Wittgenstein, Philosophical Investigations (Oxford: Blackwell, 1953) §1. Ibid., §43. It is in this sense that Putnam describes Wittgenstein as a postanalytic thinker. See Rajchman and West, Post-Analytic Philosophy, 29. See, for example, Daniel D. Hutto, Wittgenstein and the End of Philosophy: Neither Theory nor Therapy (Hampshire: Palgrave Macmillan, 2006), 174–190. Ludwig Wittgenstein, On Certainty (London: Blackwell, 1969), §94. Donald Davidson, ‘On the Very Idea of a Conceptual Scheme’, Proceedings and Addresses of the American Philosophical Association 47 (1974): 14–16. Ibid., 20. Donald Davidson, ‘Three Varieties of Knowledge’, in Subjective, Intersubjective and Objective (Oxford: Oxford University Press, 2001), 205–220. Wittgenstein, Philosophical Investigations, §126. See the interview with Davidson in Giovanna Borradori and Rosanna Crocitto, The American Philosopher (Chicago: University of Chicago Press, 1994), 54.

Postanalytic Philosophy: Overcoming the Divide? 25

26 27

28

29 30 31 32 33

34 35

36 37 38

39

40

41

21

See Donald Davidson, ‘The Third Man’, Critical Inquiry 19, No. 4 (1993): 607–608: ‘Nothing has surprised me more than to discover myself anthologised in books with titles such as Post-Analytic Philosophy or After Philosophy . . . Is there something sinister, or at least fin de siécle, in my views that I have failed to recognise, something that portends the dissolution not only of the sort of philosophy I do but of philosophy itself?’ He suggests the answer to this question (and the related question why he is associated with figures like Heidegger and Derrida) may turn on his ‘rejection of subjectivist theories of epistemology and meaning, and [his] conviction that thought itself is essentially social.’ Ibid. Rorty, Philosophy and the Mirror of Nature, 12. See Brandom’s summary of Rorty’s critique in Robert Brandom, ed., Rorty and His Critics (Oxford: Wiley-Blackwell, 2001), 160. Rorty, Philosophy and the Mirror of Nature, 375. Also see Donald Davidson, ‘Three Varieties of Knowledge’, in Subjective, Intersubjective and Objective (Oxford, Oxford University Press, 2001). Rorty, Philosophy and the Mirror of Nature, 168–169. Also Brandom, ed., Rorty and His Critics, 10. Rorty, Philosophy and the Mirror of Nature, 131. Rorty and Mendieta, Take Care of Freedom and Truth Will Take Care of Itself, 24. Rorty, Philosophy and the Mirror of Nature, 317. Ibid., 9. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994), 9. Ibid., 67. See Michael Friedman, ‘Exorcising the Philosophical Tradition’, in Reading McDowell: On Mind and World, ed. Nicholas Smith (London: Routledge, 2002), 28–29. ‘McDowell’s work is intended, therefore, to contribute towards a transcendence or overcoming of the philosophical tradition – an exorcism of the dualistic oppositions that have given rise to the traditional ‘“problems of philosophy.”’ McDowell, Mind and World, 92. Ibid., 147. Ibid., 86. Also see 93, where McDowell states that his work is aimed at a Wittgensteinian quietism. McDowell, however, dismisses Rorty’s attempted deconstruction as ‘half-baked’ insofar as it rests upon a problematic dualism of reason and nature. Ibid., 155. See John McDowell, ‘Transcendental Empiricism’. Paper presented at the third Athens-Pittsburgh Symposium, Rethymnon, October 2000; Greek translation in Deuka-lion, vol. 21, (2003). This analysis made use of the wide range of both analytic and continental journals accessible in the Philosopher’s Index to assess the analytic and continental engagement with the work of each postanalytic figure. The study targeted peerreviewed journal articles where each figure’s work predominated, using as a search criterion mention of a figure’s name in either the title or abstract. This method aimed to eliminate incidental citations and utilises the searchable fields in the Philosopher’s Index, but obviously does not capture articles in which a figure’s work is discussed in detail but his or her name is not mentioned in the title or abstract. The results of this preliminary analysis can be seen in Figures 2.1–2.7. Each relevant article has been categorized as either analytic, crossover or continental, determined in accordance with the stated publishing policy of the journal

22

42

43

44

45

Postanalytic and Metacontinental of publication. Overall, such articles were drawn from 55 analytic journals (e.g. Analysis, Erkenntnis, Mind, Noûs, The Philosophical Review and Ratio), 65 ‘crossover’ journals (e.g. European Journal of Philosophy, Southern Journal of Philosophy and International Journal of Philosophical Studies, along with selected interdisciplinary publications like Philosophy and Literature and Critical Inquiry), and 26 continental journals (e.g. Continental Philosophy Review, Philosophy Today, Philosophy and Social Criticism, Research in Phenomenology and The Review of Metaphysics). The relevant date range for each figure begins from publication of the first article with a dominant or predominant focus on that figure. For Davidson’s relation to Heidegger, see Malpas, Donald Davidson and the Mirror of Meaning (Cambridge, Cambridge University Presss, 1992). For his relation to Gadamer, see chapter 4 of Linda Alcoff, Real Knowing: New Versions of the Coherence Theory (Ithaca: Cornell University Press, 1996). For his relation to Derrida, see Giovanna Borradori, ‘Two Versions of Continental Holism: Derrida and Structuralism’, Philosophy & Social Criticism 26, No. 4 (2000), and S. Pradhan, ‘Minimalist Semantics: Davidson and Derrida on Meaning, Use and Convention’, Diacritics 1, No. 1, (1986), 66–77. This cross-referenced a small sample of recent articles published on Davidson’s thought against a compiled list of articles, books and book chapters previously published on Davidson in both analytic and continental traditions, in order to assess how far each sampled article engaged with previously published work on Davidson in each tradition. The intention was to find out whether Davidson has inspired genuinely cross-traditional debate, as measured by citation between contemporary analytic and continental work on Davidson (and not just by citation within each group of Davidson himself). The preliminary search showed no crosscitation of this kind at all. A similar analysis was conducted for each postanalytic thinker, with a similar trend evidenced in the case of both Rorty and McDowell. In each case, the sample set included peer-reviewed journal articles in which the work of the thinker in question was of predominant focus. Articles sampled were published, for the most part, in 2007, though where there were more analytic articles in a given year than ‘crossover’/continental articles, articles published in 2008 or between 2000 and 2006 have been added to ensure the number of analytic and ‘crossover’/continental articles sampled remained roughly even. Given the small sample set under consideration (approximately 10 articles for each postanalytic figure), these results are necessarily preliminary. The three most popular Davidsonian articles appearing in Mind, Noûs, the Journal of Philosophy, and the Philosophical Review address these issues: ‘Mental Events’ (1970), ‘Truth and Meaning’ (1967), and ‘Actions, Reasons and Causes’ (1963), cited 89, 71 and 70 times, respectively. Baseline of 4, 796 articles (1967 to 2008). The three most popular Davidson articles in crossover journals are ‘A Coherence Theory of Truth and Knowledge’ (1983) ‘On the Very Idea of a Conceptual Scheme’ (1973) and ‘A Nice Derangement of Epitaphs’ (1986), cited 13, 12 and 9 times respectively in three major ‘crossover’ journals between 1984 and 2008 (European Journal of Philosophy, Southern Journal of Philosophy and Inquiry). Baseline of 896 journal articles (1983 to 2008). Note that the EJoP and the SJoP are only indexed online between 1998/2003 to 2008, respectively.

Postanalytic Philosophy: Overcoming the Divide? 46

47

48

49

50

51

52

53

54

55

56

23

A search of Philosophers Index reveals that 26 of 28 refereed articles engaging in detail with the work of McDowell up until the publication of Mind and World were published in analytic journals or collections. Subsequent to Mind and World, McDowell became a much-discussed figure in ‘crossover’ and German journals. A handful of articles in continental journals more narrowly defined, such as Philosophy Today, appeared subsequent to 2000, six years after the publication of Mind and World. Special journal editions devoted to the discussion of Mind and World appear in European Journal of Philosophy 14, No. 2 (2006), Teorema 25, No. 1 (2006), and Deutsche Zeitschrift für Philosophie 48, No. 6 (2000). See Ståle Finke, ‘Concepts and Intuitions: Adorno after the Linguistic Turn’, Inquiry: An Interdisciplinary Journal of Philosophy 44, No. 2 (2001): 171–200. Bryan Baird, ‘The Transcendental Nature of Mind and World’, Southern Journal of Philosophy, 44, No. 3 (2006): 381–398. Kenneth R. Westphal, ‘Contemporary Epistemology: Kant, Hegel, McDowell’, European Journal of Philosophy 14, No. 2 (2006): 274–301. Crispin Wright, ‘Human Nature?’, in Reading McDowell: On Mind and World, ed. Nicholas Smith (London: Routledge, 2002), 157–158. Wright implies that McDowell should be banished from the analytical tradition on the basis of Mind and World: ‘if analytical philosophy demands self-consciousness about unexplained or only partially explained terms of art, formality and explicitness in the setting out of argument, and the clearest possible sign-posting and formulation of assumptions, targets, and goals, etc., then this is not a work of analytical philosophy.’ Nonetheless, Wright himself engages with McDowell’s work with a critical rather than dismissive stance, and Mind and World continues to provoke debate in analytic philosophy of mind and language. Examples include Herman J. Saatkamp, that is, Rorty and Pragmatism: The Philosopher Responds to His Critics (Vanderbilt University Press, 1995), and Josef Niznik and John T. Sanders, eds, Debating the State of Philosophy: Habermas, Rorty and Kolakowski (Westport: Praeger Publishers, 1996). Both include contributions from leading analytic and continental philosophers that engage with relativism. Examples include Robert Hollinger, ed., Hermeneutics and Praxis (Notre Dame: University of Notre Dame, 1985); Kai Nelson, After the Demise of the Tradition: Rorty, Critical Theory, and the Fate of Philosophy (Boulder: Westview Press, 1991); and John McCumber, Philosophy and Freedom: Derrida, Rorty, Habermas, Foucault (Bloomington: Indiana University Press, 2000). Philosophy and the Mirror of Nature is cited 175 times in these 15 analytic journals. Baseline of 10,188 journal articles (1979 to 2008). In his annotated bibliography of secondary literature on Rorty, Rumana notes that, of over twelve hundred noted citations, only a small percentage are friendly to Rorty. Richard Rumana, Richard Rorty: An Annotated Bibliography of Secondary Literature (Amsterdam: Rodopi, 2002), ix. Furthermore, a preliminary crosscitation check yields no evidence of substantive cross-traditional engagement (see note 44). Citation patterns also cast doubt on the view that while the early Wittgenstein is an analytic philosopher, analytic scholars ceased to be interested in his later work at the same time as he became more popular in the continental tradition

24

57

58

Postanalytic and Metacontinental (see, for example, Anat Biletzki, ‘Bridging the Analytic-Continental Divide’, International Journal of Philosophical Studies 9, No. 3 (2001): 293–294. (Evidence against this view is provided in Figure 2.8, from which it can be seen that the Philosophical Investigations is cited almost two-thirds as often as the Tractatus in both analytic and continental or ‘crossover’ journals. Analytic journals searched include Mind, Noûs, the Journal of Philosophy and the Philosophical Review. ‘Crossover’ and continental journals searched include the European Journal of Philosophy, Inquiry, Southern Journal of Philosophy, Continental Philosophy Review and the Review of Metaphysics). For example, recent scholarship has displayed affinities between Wittgenstein and Derrida’s explorations of the role of language, grammar and logic in relation to metaphysics. See Garver and Lee, Derrida and Wittgenstein (Philadelphia: Temple University Press, 1995), and Henry Staten, Wittgenstein and Derrida (Lincoln: University of Nebraska Press, 1986). The cross-citation check cross-referenced a small sample set of articles published on Wittgenstein’s work in 2007 against previously published work on Wittgenstein in both analytic and continental traditions. Three of four articles published in continental journals in 2007 drew significantly on interpreters of Wittgenstein frequently published in major journals of analytic philosophy, including P. M. S. Hacker, Cora Diamond and James Conant. Hacker is perhaps most well-known, but Diamond has 10 published articles in the 15 analytic journals listed in note 11, and 51 articles cite her work between 1980 and 2004 (baseline of 10, 564 articles). Conant has two published articles in the same set, and 33 articles cite his work between 1988 and 2004 (baseline of 6,874 articles). Of particular interest are: Paul Livingston, ‘Wittgenstein, Kant and the Critique of Totality’, Philosophy and Social Criticism 33, No. 6 (2007): 691–715; and Søren Overgaard, ‘The Ethical Residue of Language in Levinas and Early Wittgenstein’, Philosophy and Social Criticism 33, No. 2 (2007): 223–249. Though both are published in paradigmatically continental journals, they draw substantially on analytic interpretations of Wittgenstein, and discuss the relationship between Wittgenstein and the analytic tradition.

Chapter 3

The Fate of Transcendental Reasoning in Contemporary Philosophy James Chase and Jack Reynolds1

1. Introduction Analytic and continental philosophers differ on the worth of transcendental reasoning. Analytic concern with transcendental reasoning was evident from the beginning of the movement, and although the analytic literature saw a brief mini-industry on the subject following Peter Strawson’s prominent use of the method, discussion of their acceptability has always been more common than their actual use, and the trend of the discussion has run against the argument form. To the extent that continental philosophy persists in the use of such modes of reasoning, then, it comes under analytic question. By contrast, in the continental traditions (from Kant to the present), it seems to us that transcendental reasoning is close to ubiquitous2 – although what the transcendental involves has been significantly and separately reconfigured by phenomenology and the genealogical turn, as well as by a more constructivist understanding of philosophy emphasizing the transformative potential of the method in concept creation. There are continental concerns about the status of transcendental reasoning, but continued creative use persists and there is no general agreement that transcendental argumentation is especially problematic. In fact, it is sometimes claimed and frequently implied that a transcendental dimension is of the essence of philosophy: any philosophical activity that does not reflect on its own conditions of possibility is naïve or pre-critical; and reflections on the conditions of contemporary philosophical discourse, subjectivity, and cultural life more generally lead to an appreciation of the ‘problem of modernity’. And, of course, this is one way of throwing the analytic project into question. On the other hand, any suggestion that a transcendental dimension is necessary to philosophy need not automatically exclude analytic philosophy. To the extent that analytic philosophers do meta-philosophy of some kind (as very many do), especially in dwelling on the relationship between experience and reflection on that experience (not so much from nowhere but rather from ‘within’), they are also perhaps minimally transcendental philosophers, even if a certain commonsense attitude or commitment to naturalism might quickly short-circuit or limit

28

Postanalytic and Metacontinental

the significance given to such reflections.3 Tom Baldwin, the current editor of Mind, exemplifies the resulting attitude when he suggests, in a paper on Derrida and death, that where transcendental analyses collide with common sense the former should be eschewed.4 But of course, such an attitude, even though not explicitly hostile, radically diverges from the spirit of Kantian ‘critique’ that has continued in much continental philosophy. In this chapter, we trace these divergent attitudes vis-à-vis transcendental reasoning, with the goal of identifying some of the background differences in each tradition that (internally) justify the divergence. We begin with the analytic attitude to transcendental reasoning, which we argue is due in part to the explicit objections to transcendental reasoning absorbed by the analytic community, but also due in part to the methodological role of empiricism, the analytic understanding of transcendental argument as a form, and to a wider analytic attitude to ‘necessity-mongering’ claims. We then look at continental appropriations of such arguments, making use of Mark Sacks’ notion of a situated thought to outline transcendental reasoning involving embodiment and time, and considering the extent to which analytic criticisms apply to such usages.

2. Analytic attitudes to the transcendental In Transcendental Arguments and Scepticism, Robert Stern observes that there is a ‘widespread conviction that there is something vaguely disreputable or even dishonest’ about transcendental arguments.5 This accurately captures the peripheral status of transcendental argument in contemporary analytic philosophy; that is, in the period after Strawson’s influential defence of the argument form and the subsequent objections by Stephen Körner, Barry Stroud and others. For the past thirty years or so, transcendental arguments have neither been used nor analysed extensively in the analytic journal literature, and when such prominent analytically trained philosophers as Putnam, Davidson,6 and McDowell7 do make use of them, it appears to be taken as something of a sign of non-analyticity.8 Beyond the situation of these well-known ‘postanalytic’ philosophers, there are other indications that the transcendental has become problematic for analytic philosophy. For instance, Pascal Engel denies that many of those who are generally taken to be French analytic philosophers should be considered thus, precisely because of their residual allegiance to the transcendental.9 Analytic and continental philosophy of religion is also largely separated on this point,10 and a related methodological division is arguably apparent within pragmatism.11 The analytic and continental understandings of transcendental argument share common ground in Kant’s Critique of Pure Reason.12 It might seem, then, that differing views about the transcendental across philosophy could be understood as differences about how to handle the Kantian heritage – differences about the

The Fate of Transcendental Reasoning

29

dependence of transcendental argument on transcendental philosophy, or about the most fruitful way to generalize on or depart from the model established by Kant’s usage. Given the well-known differences in the interpretation of Kant in the two traditions, this is not an unreasonable view.13 But the move back to Kant brings into prominence a basic problem in discussing differing analytic and continental receptions of transcendental reasoning. There is arguably no common referent here. Kant is indeed read across both traditions, but it does not follow that there is a neutral characterization of the transcendental argument form common to both analytic and continental philosophers as the locus of disagreement. Consider, for instance, the following ‘minimal’ characterization of the transcendental argument form: A transcendental argument is an inference from a state of affairs that indisputably obtains, to the existence of a further, contested, state of affairs that is recognised to be a necessary condition for this obtaining (a ‘condition for its possibility’). The uncontested state of affairs is almost always subject-involving – it might involve such first personal matters as having knowledge, or certain experiences, or beliefs, or conceptual capacities, or it might involve intersubjective relations or practices of some kind. The contested state of affairs may or may not be subject-involving. The structure of such an argument is then as follows: 1. Subject-involving state of affairs p obtains. 2. A necessary condition for p obtaining is that q obtain. 3. So q obtains. If this is all a transcendental argument comes to, it is a special case of reasoning by modus ponens and so not a distinct argument form at all. One might seek a different inferential frame for the argument (interpreting it as a species of inference to the best explanation, say), but again the result is nothing distinctive. More must be said if we are to get at the contested nature of the form. Much of Kant’s own practice is missing in this skeletal characterization – most obviously the intended role of such demonstrations as in some sense securing synthetic a priori principles or concepts – and one can select from the two post-Kantian traditions many different ways of adding flesh to these bones, in the form of further conditions or constraints that attempt to capture the distinctive feature (and so the distinctive possibilities) of such reasoning. For instance, it might be that the notion of necessity in play in (2) is not simply conceptual or logical; or that the condition it expresses is to be thought of as a type of non-logical, non-psychological dependence relation; or that the goal of the transcending manoeuvre is to re-conceive the two states of affairs so as to understand them as distinct though connected. Or the contextual surrounds of a transcendental argument might be critical – it might be that it has an essentially anti-sceptical function; or that it starts from the explanatory question ‘how is p possible?; or that the kinds of conclusion that it can establish are themselves premises for

30

Postanalytic and Metacontinental

a further (transcendentally idealist) project. There is no general agreement across the traditions (or within each) on each of these matters; different ways of understanding transcendental argument have developed in parallel. Indeed, it is not unusual for this state of affairs itself to be celebrated as a mark of the vitality of the method. For instance, Ameriks remarks: Even if Kant’s system is a multilayered complex that has only one root in common sense . . . it is no crime if the other root of his work (the ‘transcendental’ arguments), like all good philosophy, is inextricably involved with some abstract and endlessly disputable concepts. The fertile ambiguity of these concepts . . . has by now surely demonstrated their worth.14 Although one might dispute Ameriks’ contention that there is an inextricable connection between good philosophy and endlessly disputable concepts, historically speaking (and in continental philosophy), the ‘fertile ambiguity’ he celebrates is certainly present here. But, of course, much of the fertile ambiguity of transcendental reasoning is not on show in analytic treatments, even though it is the analytic movement that has been more concerned with the identification and discussion of a transcendental argument form. This needs explanation. Analytic work in other cases vectors in on such ‘fertile ambiguity’ precisely to see whether it is indeed so fertile. Contemporary analytic discussions of the ontological argument form distinguish Anselmian, Cartesian, Kantian, and Gödelian forms, with numerous variations depending on whether conceptual, experiential, modal or other circumstances are used to launch them. Why is there not the same sustained analytic persistence with transcendental modes of reasoning? Why is the project of explication dropped so quickly? So it seems to us that what needs to be explained here is how the analytic interest in transcendental argument as a form arose, without being accompanied by this same exploration of potential variation. In the early analytic context, at least, the answer seems clear, and it involves the way that empiricism functioned as a methodological constraint, rather than merely as a truth claim. Even non-empiricists among the early analytics – such as Russell at some times – took empiricist scruples very seriously as a mark of epistemic respectability (in his last major work, Russell is concerned to show that his deviations have ‘what may be called an “empiricist flavour”’15), and empiricist scruples baulk at transcendental methods. Methodological worries of this kind are never with just the particular issue at hand (a particular use of transcendental reasoning, say); rather, the potential cost is in the confounding of the whole research program. Hostility to transcendental reasoning was thereby part and parcel of Russell and Moore’s formative attempts to distance themselves from British idealism.16 The move within the early analytic period to a more thoroughgoing empiricism made transcendental reasoning even more difficult to take seriously. The logical positivists influentially rejected the synthetic a priori, suggesting that such

The Fate of Transcendental Reasoning

31

knowledge claims amounted to poetry, mysticism or nonsense. Of course, in different ways, Wittgenstein’s Tractarian picture, the verification principle and the pluralist view involved in Carnap’s semantic phase seem to have a synthetic a priori status by their own lights, as has often been noted,17 but this was not taken as any kind of spur to transcendental reasoning.18 Instead, epistemological self-application difficulties were seen as theoretical problems arising within the overall empiricist project, and were hardly to be solved by jettisoning that most basic commitment. Rather, within the empiricist analytic tradition such problems were most obviously to be resolved by adopting a more thoroughgoing coherentism and the radical empiricism of Quine. At this point, then, transcendental reasoning ran squarely against the research program of the emerging analytic movement; to take it seriously was to be some other kind of philosopher. Given this attitude, it is hardly surprising that analytic discussion of transcendental reasoning could only come about with the rise of obviously non-positivistic and even non-empiricist schools, constrained not in the least by methodological empiricism. These conditions obtained for the mid-century ordinary language movement in England. And from the start there is a conception of transcendental argument as a separate argument form with a logical or semi-logical structure, capable of being recognized in the work of other philosophers, past and present, and perhaps available for use in analytic projects. The first clear attempt at this is J. L. Austin’s isolation of transcendental argument as a distinctive (if limited) method of reasoning in his 1939 paper ‘Are there A Priori Concepts?’19 When the argument form makes its appearance in the analytic tradition, it does so in the teeth of a characteristically pointed piece of Austinian rhetoric: People (philosophers) speak of ‘universals’ as though these were entities which they often stumbled across, in some familiar way which needs no explanation. But they are not so. On the contrary, it is not so very long since these alleged entities were calculated into existence by a transcendental argument: and in those days, anyone bold enough to say there were ‘universals’ kept the argument always ready, to produce if challenged.20 Next Austin sets out a particular argument for the existence of universals, and clarifies what he means here by ‘transcendental argument’: This is a transcendental argument: if there were not in existence something other than sensa, we should not be able to do what we are able to do (viz. name things).21 After a little discussion of the very limited knowledge of universals such an argument can yield, Austin presents a further argument for the existence of universals, notes that it is also transcendental in this sense, and then raises what

32

Postanalytic and Metacontinental

one could take to be a difficulty with any general strategy of using such arguments repeatedly (rather than as one-offs, as it were): Now it must be asked: what conceivable ground have we for identifying the ‘universals’ of our original argument with the ‘universals’ of this second argument? Except that both are non-sensible, nothing more is known in which they are alike. Is it not odd to suppose that any two distinct transcendental arguments could possibly be known each to prove the existence of the same kind of thing?22 It is perhaps not surprising that the first analytic discussion of the transcendental argument form turns out to be an obituary, but there are other points worth noting here. First, in treating transcendental argument as a form entirely detachable from its Kantian heritage, Austin is making a characteristically analytic move. The implication is that this is something we can lift out of the general milieu of transcendental idealism or critical philosophy and treat on its own, or recognize at work in other contexts, and thus it is also available for analysis as a logical structure. Second, considered as such a form, in Austin’s hands the transcendental argument becomes much the same sort of device as the standard argument forms of the ordinary language school; say, the paradigm case argument, or the argument from excluded opposites. In each of these three cases, the argument form involves reasoning to a state of affairs from a capacity we have that presupposes it (our wielding of ‘red’ presupposes red things exist; our talk of ‘counterfeit coins’ presupposes there are real ones). One can see Austin, then, as offering a very ‘thin’ understanding of transcendental argument, on which the argument form comes down to a kind of logical or conceptual relationship between capacity and ground. This is surely in part due to the context of ordinary language philosophy itself, inclined to be suspicious of traditional metaphysical (or critical) terminology. In such a context, one can see Austin’s version of the transcendental argument, like Hume’s account of causation, as something of a revisionary clean-up job: the intention is not to engage in Kantian exegesis, let alone to enjoy the ‘fertile ambiguity’ that Ameriks celebrates. Rather it is to identify an argument form that can in theory be given general work to do (because sufficiently purged of links to the transcendental philosophy), but that is perhaps not as useful as its proponents suppose. Austin’s transcendental argument form is (by his lights) a bad argument form, but a clear one. The ordinary language movement also offers the first explicit analytic rejection of this view, and the first sign of a more optimistic role for transcendental argument. Peter Strawson’s 1959 monograph, Individuals, endorses ‘descriptive metaphysics’ as a middle way between revisionary metaphysical activities and the ordinary activities of philosophical analysis (such as the appeal to paradigm

The Fate of Transcendental Reasoning

33

cases). In setting out the case for going beyond mere analysis, Strawson highlights the limits of analysis as a path to understanding: Up to a certain point, the reliance upon a close examination of the actual use of words is the best, and indeed the only sure, way in philosophy. But the discriminations we can make, and the connexions we can establish, in this way, are not general enough and not far-reaching enough to meet the full metaphysical demand for understanding. For when we ask how we use this or that expression, our answers, however revealing at a certain level, are apt to assume, and not to expose, those general elements of structure which the metaphysician wants revealed.23 Descriptive metaphysics is the project of tracing such structures, and it requires arguments that are at the same time explanations and even, perhaps, dramas.24 This is the natural home of transcendental argument; Strawson regards Kant as a descriptive metaphysician, and the ‘reidentification’ argument he puts forward in section 2 of the first chapter of the book is the exemplar of the analytic transcendental argument. Our conceptual scheme of a single unified system of spatio-temporal relations is the framework within which we organize our individuating thoughts about particulars, but a condition of our having such a scheme is our acceptance of the persistence of at least some individual particulars through periods of non-continuous observation; hence we are justified in believing in the persistence of (some) individual particulars.25 Being descriptive, this argument is explicitly modest (as Austin’s was not) – it concerns only the tracing of our conceptual commitments and the structure of our conception of the world. Nonetheless, Strawson holds that it has value in defeating the sceptic, since the argument shows the sceptic ‘pretends to accept a conceptual scheme, but at the same time quietly rejects one of the conditions of its employment’.26 To say the least, this conclusion has been strongly contested in the analytic literature (and Strawson himself came to drop it, and to adopt a more cautious attitude to transcendental argumentation27). Two objections have had especially great influence in the analytic literature. (i) First, as Barry Stroud points out, the sceptic has hardly been defeated.28 For the conclusion of a modest transcendental argument simply tells us a fact about how we think about the world. It is a further step from this claim to the claim that this is how things are. The proponent of the argument can only (immodestly) take this step if the premises of the argument are bolstered with a commitment to a thesis strong enough to bridge the gap (idealism, or verificationism, for instance). Yet this is a premise the sceptic will not accept. Stroud goes further, pointing out that the utility of the transcendental argument is in question even if such a background is assumed, since direct (non-transcendental) argument from idealism or verificationism is now sufficient to establish the conclusion.

34

Postanalytic and Metacontinental

(ii) Second, as Körner and Stroud both point out, the necessary connection identified in Strawson’s arguments (between experiential fact and transcendental ground) is itself difficult to defend against the skeptic.29 For the only necessity in play here is a necessity apparently arising out of our conceptual framework. Hence a suppressed premise of the argument is that no other conceptual framework is available.30 Yet it is difficult to see how this unique applicability is to be established. This is a version of a concern Russell set out in classical form in The Problems of Philosophy – the necessities established by transcendental reasoning must be relative necessities, in that they depend on features of our conceptual framework which themselves might well be contingent. Both of these objections are widely known within the analytic community, having been discussed in the literature particularly in the decade after Stroud’s first paper on the topic. And, of course, this suggests a straightforward internal explanation of the current analytic attitude to transcendental reasoning. According to this explanation, the contemporary attitude is a function of the literature on the Körnerian and Stroudian objections: analytic philosophers now simply have reason to regard transcendental arguments as bad arguments; as a result there has been a decline in interest in the argument form. There is indeed something to this explanation, but it is a little too straightforward as it stands, for two reasons. First, it doesn’t explain a difference in the analytic treatment of what one might call dubious argument forms. Other historically influential argument forms are discussed in the analytic literature, even though they are generally regarded as unsuccessful. The ontological argument, as noted earlier, remains a lively subject of discussion in the analytic journals, and is the subject of several recent monographs.31 True, there exist proponents of the argument who believe it to be sound, but they are very much a minority. One might in the same way regard Putnam and Davidson as part of a minority on the subject of transcendental reasoning, but there is an important disanalogy – in the first case, but not the second, being in the minority is not taken as a sign of non-analyticity. Second, it doesn’t explain the differences in the analytic treatment of these particular objections in this case as opposed to others. Objections much like those of Stroud and Körner can be raised against other argument forms that remain central to the analytic literature. Many argument forms that ostensibly see off the sceptic – argument patterns based on reliabilism, say, or those based on some form of contextualism – fall foul of sceptical replies of a Stroudian flavour. Many argument patterns are only valid within the scope of particular assumptions, or can only lend coherence-building support to a position; they might therefore seem open to Stroud’s follow-up objections about the utility in argument of the transcendental argument form. Körnerian worries about strong claims of necessity arise wherever an argument depends on a claimed necessity or impossibility – for instance, in arguments based on thought experiments of the kind Roy Sorensen calls ‘possibility refuters’.32 These are certainly not the most common use of thought experiment in the analytic tradition, but

The Fate of Transcendental Reasoning

35

they are a respectable minority, and they involve claimed necessities (thinly disguised as claims about impossibility). Each of these situations is one in which a Stroudian or Körnerian move might be made or has been made, pointing to a sceptical countermove or an overly confident claim of necessity. Yet in none of these cases is the fact of objection of this kind taken as debate-closing in the way that it apparently has done for transcendental reasoning. So the internal explanation taken entirely by itself won’t do; at best it is incomplete. The analytic attitude to transcendental reasoning genuinely depends also on particular norms and habits peculiar to the analytic tradition. One of these is simply the focus on argument form in general. Stroud’s sceptical objection can seem puzzling or nit-picking to continental philosophers who are generally unconcerned with the anti-sceptical role of transcendental arguments, but focusing on the sceptic as the opponent can be a little misleading. Arguably Stroud’s objection causes far more general trouble to the argument form as a whole, assuming indeed (as the analytic tradition does) that such forms are in a sense universal, capable of use across differing contexts. Any argument form at all that can be taken without (too much) distortion out of its original context can be re-characterized as having an anti-sceptical function, in that it could in principle be wielded in a different context against someone sceptical about its conclusion. Similarly, any putatively anti-sceptical argument that has the same kind of potential transferability can simply be seen as an argument for its conclusion, no matter what the dialogic context it arises in. By contrast, if we do not think of transcendental reasoning as having such universal form, Stroud’s objection remains localized to sceptical debates, with which continental philosophers are not generally concerned. A second feature of the landscape that comes into play here, we think, is the attitude of the analytic tradition to claims of necessity or necessary connection. The internal force of the Körnerian objection can be misjudged by non-analytics for precisely this reason. On the one hand, naturalized epistemology and radical empiricism have become far more significant in the analytic tradition under the influence of Quine. From this point of view, such exercises as Strawson’s can, at most, show the structure of a conceptual framework that we happen to hold, and no part of which is beyond potential revision. In many dialogues within analytic philosophy, necessity-mongering is regarded sceptically even by those who are not themselves Quineans; the attitude here is captured by Daniel Dennett’s oft-repeated remark that philosophers too often mistake a failure of imagination for an insight into necessity, and by the neo-Humean dictum that necessities are theoretical costs, to be minimized wherever possible.33 On the other hand, the early twentieth century development of modal logic led to an unexpected problem of massive overabundance which the analytic tradition still lives with. It rapidly became clear that formally specifiable alethic modalities are legion, each yielding a distinct sense of necessity, and that temporal, deontic and other necessities could be modelled in the same way, and post-Kripkean diagnoses of recurrent equivocation between types of necessity are habitual.

36

Postanalytic and Metacontinental

As a result, contested philosophical argument forms that depend on alleged but unexamined necessities have to be treated with care even by those who endorse a priorist reasoning; clarity, rather than fertile ambiguity, is at a premium here. Of course, these are general concerns, and much analytic argumentation does involve claims about what is necessary, so again a double standard might seem to be in play. Not so. First, some analytic necessity claims are just ways of pointing out the impossibility of contradiction, or unpacking an explicit definition for analysis and potential refutation. Transcendental arguments do not make similarly secure claims about the necessity of the connection between an agential and a non-agential state of affairs obtaining. Second, there is a difference for the analytic philosopher in what might be called the constructive potential of various premises. Suppose, as many analytic philosophers from Bertrand Russell on have declared, the task of philosophy is to refine and regiment, rather than replace, our common-sense stock of opinions. Such an attitude does not prevent the philosopher arriving at highly unusual places (David Lewis’s modal realism, Peter van Inwagen’s views on persons). But it does place certain limits on the kinds of premises that will be readily accepted; premises that do not receive the immediate backing of intuition and fail to be credentialed by common sense are open to doubt. If such premises are claims about the necessity or impossibility of a certain state of affairs obtaining, and it is not clear that the necessity or impossibility is logical or definitional in nature, then this doubt becomes a form of permanent contestation. No matter how detailed the philosophical work taking one to the claimed necessity, it remains an entirely provisional claim in the sense that its rejection is conceivable.34 Now, this conclusion was reached from a particular analytic conception of the role of common sense, but we submit that there are other analytic paths to the same conclusion. Van Inwagen, for instance, argues that the problem of evil is an unsuccessful argument because it has philosophically controversial premises, and so cannot move a person ideally agnostic on the point at issue.35 At this point we have departed from Körner’s own concern about conceptual frameworks in transcendental reasoning. Körner’s concern is that a claimed necessity may simply reflect the limits of our conceptual framework; the current objection is that the necessitarian premises of transcendental arguments do not reflect the immediately apparent limits of our conceptual framework. As such, a transcendental argument will never or rarely be able to play a constructive role in the fruitful development of an analytic position. At best it will be met with a shrug and a ‘could be’ – and that is just not enough to go on with.

3. Continental attitudes to the transcendental In continental circles, transcendental reasoning is controversial and perhaps even permanently contested terrain. Certainly it is contested by all of the great

The Fate of Transcendental Reasoning

37

philosophers of the tradition, and the risks associated with such forms of reasoning are acknowledged. But it is not controversial enough to induce general abstinence. Instead, the implicit rationale seems to be a bit like Pascal’s Wager – believing in the efficacy of transcendental arguments, if they work, may result in tremendous results (a Copernican revolution); if they do not, some important concepts will have nonetheless been created. Better that, on this view, than disbelieving and being the under-labourer of science. While various continental philosophers have subjected Kant’s conception of the transcendental to critique, few have thought that this signalled the end of the transcendental project. Rather, the aim, scope and structure of transcendental reasoning is instead consistently reinvented by all of the major philosophers associated with this tradition. Indeed, the assumption is more that whatever his failings, Kant’s transcendental project (and the project of philosophical critique more generally) achieved something of the utmost importance. The German Idealists (Fichte, Schelling, Hegel) all described themselves as transcendental philosophers, and the revival of interest in Kant in the late nineteenth and early twentieth century in Neo-Kantianism (Cassirer, Brunschvig, etc.) produced significant revisions in the understanding of transcendental methods and arguments. In the twentieth century, transcendental arguments, proofs, interpretations, etc., have continued to be frequently deployed in continental philosophy in very different ways. The roll call of the famous names includes: Husserl, and the phenomenological intention to use such arguments to reveal the transcendental, a whole ‘region of being’ that was previously unexplored (and so both a method and an object of inquiry, as with Kant); Heidegger, and the association of the transcendental project with both phenomenology and hermeneutics; Sartre, who synthesizes the work of Husserl and Heidegger and gives transcendental philosophy an existential turn, as well as a famous ‘proof’ of the existence of the other; Merleau-Ponty, whose account of omnipresent bodily intentionality seeking equilibrium with the world (through the refinement of our ‘body-schema’ and the acquisition of flexible habits and skills) is argued to be the condition ensuring that sensory experience is more than a fragmented relation to raw sense data; Deleuze, who espouses ‘transcendental empiricism’ and rejects the possible as a category, instead looking to show how particular specific things are actualized, and who deduces the necessity for the ‘virtual’; Derrida, who consistently talks about ‘quasi-transcendentals’ in association with themes like différance despite also deconstructing the transcendental philosophy of Husserl, Heidegger, and others; and Vattimo, whose project has been characterized by one of his commentators as ‘temporalising the a priori’.36 Irigaray discusses the ‘sensible transcendental’, Stiegler and Nancy are both invested in transcendental philosophy,37 as are Marion, Henry and the thinkers of the theological turn in phenomenology. So too, are various contemporary German philosophers of recognition, including Habermas and Apel, whose concern is primarily with performative contradiction and the way they provide transcendental conditions

38

Postanalytic and Metacontinental

for communicative rationality.38 For each of these philosophers, transcendental arguments operate differently, with more and less scope, and more or less claimed universality and more or less able to be formalized in deductive terms. While we cannot offer an adequate survey of these differing continental treatments of transcendental reasoning, nor establish for each whether the analytic criticisms apply, we can usefully abstract from these varied projects to note that all of these philosophers make claims about anteriority.39 We will explore two particular claims of anteriority that we contend are major factors in contemporary continental philosophy: those revolving around body/perception, and those involving time.40 (These particular foci allow us to directly consider examples from some phenomenological and post-structuralist thinkers; moreover, family resemblance versions of one or another of these positions are in evidence in most of the usual suspects.) Each of these anteriorities has a reasonably clear role to play in what Mark Sacks calls situated thought: the thought that one would have from a particular point within a framework, the content of which is informed by it being grasped as if from that perspective. It is not bare propositional content as if from nowhere, but is rather informed by being phenomenologically embedded and directed.41 According to Sacks, much of the difficulty associated with transcendental arguments arises because they are understood as articulating relations between concepts or propositional contents. Construing transcendental arguments as formally valid inferences cannot be adequate, and while they can be ‘modestly’ understood as conceptual claims, for any transcendentally reached conclusion to be a priori true requires some kind of shifting to the level of experience. As a result, the thinker needs to be genuinely co-implicated with what is thought. This phenomenological (or performative) element thus links all transcendental arguments that are not merely analytic to a kind of minimal phenomenology, at least in the relation between the propositional content and the speaker/ experiencer of such content. And while there is certainly a circularity here, it is one that is borne of reflection, and is not something that can be overcome.42 Sacks’ account usefully explains why time, place, space and the body (elements of any perspectival situation) are central to the continental preoccupation with transcendental reasoning. A transcendental argument must have some relationship to experiences that are possible (or actual) in this world, or to concepts that structure our world, as is the case with the ‘historical a priori’ of Foucault’s early work, which is avowedly transcendental.43

3.1 Embodiment, perception and being-in-the-world We start with transcendental arguments about the body, perception, motorintentionality, or being-in-the-world, most of which come from phenomenology. While Heidegger has a complicated relation to transcendental philosophy, there are many chains of priority claims in Being and Time, often posed in the

The Fate of Transcendental Reasoning

39

telltale language of ‘primordiality’ and the ‘always-already’. Famously, the ready-to-hand is claimed to be a condition of possibility of apprehending objects as present-at-hand, or as unready-to-hand. Philosophy has traditionally prioritized the theoretical encounter with things; on Heidegger’s view Dasein associates with things first and foremost on a practical and immediate basis that he calls the ready-to-hand, which refers to the availability of things for our use and deployment in relation to the completion of tasks. The world is not primarily the scientific world, but the practical one of everyday life, and the transcendental claim is that any present-at-hand analysis never leaves behind the practical but presupposes it, in the sense that it is made possible by it. Some closely related claims are put forward by Merleau-Ponty in Phenomenology of Perception. Often associated with the thesis of the ‘primacy of perception’, rather than rejecting scientific and analytic ways of knowing the world MerleauPonty argues that such knowledge is always derivative in relation to the more practical aspects of the body’s exposure to the world, notably our bodily intentionality that seeks equilibria or ‘maximum grip’ with the world through the refinement of our ‘body-schema’ and the acquisition of flexible habits and skills. For him, these aspects of bodily motility and perception are the transcendental conditions that ensure sensory experience has the form of a meaningful field rather than being a fragmented relation to raw sense data, and this kind of know-how (the ‘I can’) is also said to be the condition of possibility of knowledge-that (the ‘I think’). Of course, there is a difficulty here (as there is with Heidegger), since philosophical reflection is required to illuminate this priority of know-how. (Whether or not this is an insuperable problem is debatable, but for most phenomenologists it is, rather, a fertile occasion for thought precisely of the transcendental variety.) A concrete example of Merleau-Ponty’s claim is the argument that, to put it bluntly, grasping is a condition of possibility of pointing; the ability to point to one’s nose (an abstract, reflective activity), depends on one’s ability to grasp one’s nose (a practical response to solicitation from the world; say a mosquito bite, or a need to scratch), but not vice versa. Merleau-Ponty backs the claim by appeal to empirical studies of injuries, phenomenological descriptions of what is involved in these two ways of inhabiting space, the alleged inability of empiricism and intellectualism to adequately describe or explain either the phenomenology or the empirical facts of the case (a kind of inference to a better explanation), and also transcendental reflection about enabling conditions for our experience of the world (motor intentionality).44 This kind of practical non-cognitive grasping is, for Merleau-Ponty (and Dreyfus, Sean Kelly, and others), largely what takes place in learning, and bound up with this privilege given to motor intentionality and skilful coping is a suggestion that such basic activities involve non-inferential perceptual norms. As Dreyfus and Kelly put it: The agent feels immediately drawn to act a certain way. This is different from deciding to perform the activity, since in feeling immediately drawn to do

40

Postanalytic and Metacontinental something the subject experiences no act of the will. Rather, he experiences the environment calling for a certain way of acting, and finds himself responding to the solicitation . . . there is an irreducibly normative component to the experience of perceptual objects . . . it is ‘right’ or ‘appropriate’ to stand a certain distance from a picture.45

While few phenomenological descriptions are uncontested, all phenomenologists will maintain that our perceptual field is normatively structured. Perception has an orientational structure (up/down, figure/ground,46 etc.) that solicits us to optimally come to grips with it, and we see things in terms of actions and in relation to potential uses of them by others. If we grant for the moment that there are non-inferential perceptual norms of this kind, then it is at least open to make a case of the kind that Sacks puts forward – one on which beliefs can be justified without appeal to inductive or other non-deductive evidence, and yet not in virtue of anything like straightforward conceptual analysis. Of course, an analytic philosopher suspicious of transcendental reasoning might well maintain that to show that something is non-inferential is not to show that it is infallible,47 nor that it is prior to inference. Indeed, the more controversial part of any such argument would be to establish in what sense these non-inferential aspects (say Heidegger’s ready-to-hand, or Merleau-Ponty’s bodily intentionality) are the ground for inference. Robert Brandom, for example, refuses to concur with Heidegger on this.48 It need not follow from this that transcendental arguments are intrinsically misleading, however, but rather that they are deployed as part of a package deal. These kind of phenomenological arguments are not straight a priori arguments. While they depend upon experience, and claim that certain enabling conditions make possible this experience, phenomenological reflection also allows us to attend through reflection (to some extent) to these so-called enabling conditions which were previously in the background; they also become perspicuous in cases of breakdown (when things are unready-to-hand). This need not lead to a problem of infinite regress. Bodily intentionality is (claimed to be) primordial, and not everyone must see the claimed necessity, nor indeed need the process of coming to see this putative necessity itself be immediate. On the contrary, it depends on detailed descriptions of phenomena and the structures of experience, and the concepts involved in such descriptions are capable of alternative understandings.49 Empirical data will hence rarely establish that a transcendental argument is wrong tout court, since the difficult task is always to describe each of the premises/concepts in detail, and there is often sufficient ambiguity in the situation to allow for the reconstruction of a transcendental argument that fits with empirical findings. However, not any reconstruction will be possible and the empirical hence remains an important constraint. The important question for the continental philosopher here is just when such reconstructions become a pathological attempt to immunize one’s theory against any possibility of error. This is always a judgment call, depending

The Fate of Transcendental Reasoning

41

on an array of background norms regarding justification and explanatory comprehensiveness. It does suggest, however, that cognitive and empirical science cannot be ignored by any phenomenologist working in this particular embodied transcendental tradition, and that empirical discoveries can and should (potentially) be able to cast doubts on the claims of transcendental philosophy. There is, however, a strong and a weak (immodest and modest) way to interpret Heidegger and Merleau-Ponty’s claims to a primacy of the ready-to-hand and embodied perception. They might just be a timely reminder to any cognitivist philosophy that something is left out when structures of cognition are posited everywhere. The stakes are raised, however, when it is declared that the non-inferential coping of the lived-body necessarily could not be explained or understood within the mechanistic and inferential terms of science, psychology, biology, etc.50 This is where the ‘anteriority complex’ becomes, perhaps, a more apt term. But if we do not take this more immodest step, how does what we have described thus far fare in relation to the standard analytic objections to transcendental reasoning? The problem is that nothing seems to bridge the Stroudian/Kantian gap between appearance and reality, although from the phenomenologist’s perspective that very gap itself presupposes the ready-to-hand, and bodily coping, as its condition of enunciation. From the analytic point of view, therefore, at best these kinds of thoughts are an appeal to some kind of coherence condition; for instance, a kind of reflective equilibrium concerning our judgments and pre-judgments. The ‘transcendental’ move is no doubt thought of by its proponents as involving more than just the transformative potential of concept creation, but (on this view) it then becomes rather unclear why the continental philosopher does not make out that part of the case that can be made in such theoretically uncontentious terms. In other words, what’s wrong with charitably interpreting Heidegger or Merleau-Ponty in such a way, and so protecting them from Stroudian critique? A residual anti-scientism and the preference of many continental philosophers for anti-mechanistic explanation is certainly doing at least some work here.

3.2 Time We have suggested that one of the other key trajectories vis-à-vis transcendental reasoning concerns time, a link that was first noted in Kant.51 In his reflections on the phenomenology of internal time-consciousness, Husserl famously suggests that our integrated experience of a melody – even on first listening – implies that any so-called ‘now’ must have a retentive element that retains the past notes, and a protentive moment that anticipates future elaborations. No doubt there are other possible explanations one might give here, but these, Husserl would claim, presuppose these same aspects of time-consciousness. We have here a transcendental claim, and one that is developed by the existential

42

Postanalytic and Metacontinental

phenomenologists, perhaps especially by Heidegger in Being and Time, where temporality is claimed to be a necessary condition for Dasein.52 But if philosophy of time is important to phenomenology it is an even more central feature of post-structuralism, where the conception of the transcendental is rendered more modest, localized to times and places, historicized in an effort to avoid some of the problems associated with the Kantian and Husserlian conceptions, and so often without the strong claims to synthetic a priori knowledge. Transcendental arguments have increasingly been buttressed by phenomenology, hermeneutics and genealogy; extra forces are marshalled, just as Stroud says they must be, to avoid begging the question. But for both genealogy and phenomenology the transcendental dimension remains crucial rather than an unnecessary add-on. Without it, the one arguably commits the genetic fallacy, and the other remains enthralled in a philosophically uninteresting form of subjectivism. That said, there is an important sense in which post-structuralism’s genealogical or deconstructive analyses take the critical dimension of Kant’s Copernican revolution so seriously that they frequently put the lie to the necessitarian ambitions of transcendental philosophy. In archaeological, genealogical and deconstructive analyses, any claims to transcendental neutrality or to having established a uniquely applicable condition (or set of conditions) for experience that holds for all times and places is radically relativized. While this is often done immanently, by analysis of textual inconsistencies, or by contextualizing these claims historically (rather than by deductively formalizing the argument), the ultimate result is that many philosophers in this tradition can actually concur with the force of Stroud and Körner’s objections – although the pressing question then concerns the way their work is proffered in the name of transcendental philosophy itself. Deleuze, for example, criticizes Kant for tracing the transcendental from the empirical and reinvents transcendental philosophy so as to avoid this mistake,53 and Derrida’s deconstruction of logocentrism and the metaphysics of presence includes within its purview the problematization of transcendental projects like those found in Kant, Husserl and Heidegger, but doesn’t abandon transcendental philosophy. In Ashley Woodward’s terms, the minimal transcendental significance of this ‘temporal turn’ consists in the idea that ‘beings are disclosed within changeable cultural horizons of interpretations, and these cultural horizons are the transcendental conditions (the temporalized a priori) for the meaning and intelligibility of things in general’.54 Perhaps this historicizing doesn’t really avert analytic concerns, in that it continues to involve a question-begging kind of anti-realism. Or, if it does avert them, it might be that it brings with it other commitments that are equally anathema to many analytic philosophers (relativism, say). Transcendental reasoning in this guise is both modest (in the sense of not universalist), and, prima facie, immodest, in arguing for the existence of discernible epochs in a manner that most historians would be wary of.

The Fate of Transcendental Reasoning

43

The key transcendental arguments in post-structuralism are manifold, emerging around questions about the condition of possibility for change, novelty and difference. Where there is difference, the question then concerns the genetic conditions of specific changes, and what might be deduced to hold more generally that explains these specific changes. Deleuze’s project in Difference and Repetition is a straightforward example. He attempts to establish that repetition is, and never could be, the simple repetition of the same, and a key part of this project is his demarcation of three different approaches to time – habitual time, memorial time and futural time – all of which, he contends, involve repetition, as well as difference in repetition. The first synthesis of time, for Deleuze, is that of habit, which gives us the phenomenological experience of the ‘living-present’. On this view, time is constituted by an originary synthesis that operates on the repetition of instants; it contracts the independent instants into one another to constitute the living present. The past and the future then become but aspects or dimensions of this living present (which is not itself an instant): the future is that which is anticipated to occur, whereas the past is the preceding instants and background conditions that are retained in the contraction that makes up the present. This living-present also sets up a directionality or arrow of time in that it goes from the past to the future, and from the particular to the general. On Deleuze’s partly Humean view, then, habit is the condition of the self or ego that accompanies the contractions of the living present; the ‘I’ is produced by myriad habitual syntheses of time. It is clear enough that habit involves some kind of repetition, but for Deleuze, habit also involves difference primarily because ‘habit draws something new from repetition – namely difference (in the first instance understood as generality)’ (p. 73). Habit is hence not simply a mechanical repetition. Rather, it also involves a pre-reflective recognition (based on the passive synthesis) that the activity that is being engaged in is something that has been done before. For example, in the Humean series AB AB AB AB, it is habit that introduces a difference between one set of the series and the next, leading us to expect a B whenever we encounter an A. As Deleuze suggests, ‘when A appears, we expect B with a force corresponding to the qualitative impressions of all the contracted ABs. This is by no means a memory, nor . . . a matter of reflection’ (p. 70). His point is that it is not that we reflect upon the past, or even consciously remember the past, but that we simply know how to go on in a non-reflective way. While this process partially depends upon the past experiences that are involved in the synthesis, at least according to Deleuze, there is no memory (except what is sometimes called procedural memory) involved in the living present of habit (p. 70). This passive synthesis of time occurs in the mind – it is not carried out by the mind. Deleuze is not entirely satisfied with this habitual explanation of time, in which the chain of events, or passing present moments, constitute time. While it seems to offer an explanation of the constitution of the ‘living-present’,

44

Postanalytic and Metacontinental

the present also passes in the time that is thus constituted (it can be exhausted and is hence not co-extensive with time per se), and his basic question is hence something like, ‘why is it that a habitual present, or temporal “now” moment, can pass’, or, ‘why is it that the present is not totally co-extensive with time?’ Deleuze suggests that this necessarily refers us to a virtual or transcendental condition for the living present. To put this another way, there needs to be a second synthesis of time that causes the present to pass, and this, he argues, is the time of the past, or memorial time (p. 79). The fundamental idea is that we cannot represent a former present (i.e. the past) without also making the present itself represented in that very representation. So, if we think about our past, we also in some sense bracket away the present, or cause the present to cease to be. This means that whenever we remember, there will be two main aspects to this: firstly, the ‘actual’ memory of that past; but also a representation of the present (or the self) as itself being engaged in remembering. Deleuze describes these two aspects as memory and understanding (p. 80). We can, perhaps, schematize his argument as follows: (1) the present passes; (2) ‘no present would ever pass were it not past “at the same time” as it is present’ (p. 81); (3) hence to explain this we need to posit a ‘pure past’ (a virtual past) that has never yet been present. And, as Körner suggests, the perhaps suppressed premise here is that no other explanation will suffice. According to Deleuze, then, this second synthesis of time, the past, is the ground that means that any ‘present’ always necessarily passes (as it is bracketed away in the attempt to remember), and it hence allows for the arrival of another ‘present’ (p. 81–82). Deleuze ultimately contends, however, that neither of these modes of time are sufficient, since they do not properly institute time in thought. Although we no doubt can think habitually, and although there is ‘understanding’ involved in memory, this is not genuine thought. Deleuze argues that truly philosophical thought involves a futural form that breaks open time (p. 88), and interrupts time, even if it always also pertains to time. Deleuze associates this affirmation of pure difference, of the future, perhaps surprisingly, with Nietzsche’s famous thought of the eternal return of the same. Without dwelling on this, we can schematize Deleuze’s argument as follows: (1) there is actual change/ difference (the new sometimes happens; creativity occurs): (2) neither of the above models of time adequately explain this, although they both presuppose change/difference: (3) there must, therefore, be a third temporal synthesis oriented to chance and difference. The synthesis that is involved in futural time is not conjunctive, but instead must be conceived of as disjunctive, in that the only unity of the eternal return is the negative unity that difference does, in fact, return (p. 126). There are intuitive problems with the notion of difference returning, but in this brief account of Difference and Repetition we can see the manner in which the unknown future serves as a condition for explaining genetic change/ difference, something that is also the case, albeit in different ways, with the

The Fate of Transcendental Reasoning

45

other post-structuralist thinkers. Of course, these arguments will not convince everybody, either in summary form or in Deleuze’s own (far more nuanced) account. To those sceptical, this might all seem a bit murky, and murkiness is indeed one of the potential risks associated with transcendental arguments, both in Deleuze but also in other philosophers. While Derrida, for example, is one of the great (internal) critics of transcendental philosophy, there are arguably also some important slides in his work – from questions of transcendental priority to ethical priority, for example – that are made possible by the ambiguities that are part of transcendental philosophy. For example, it is not always clear how to understand statements like the following: There where the possible is all that happens, nothing happens, nothing that is not the impoverished unfurling of the predictable predicate of what finds itself already there, potentially, and thus produces nothing new.55 Derrida’s language here is strong, but is he making simply a conceptual point about the limitations of our concept of possibility, something that could be by definition true (or at least true-given-our-genealogical-history), or is it a claim that bears on reality, however loaded that term may be? There is a difference between what Stern refers to as concept-directed and truth-directed transcendental arguments,56 but Derrida’s quasi-transcendental claims can almost always be read in both of these ways, and the problem (within which deconstruction avowedly situates itself) revolves around whether we understand Derrida primarily as a genealogist of concepts, as a transcendental philosopher,57 or as an ethicist, albeit in a highly restricted sense. Derrida might contest Stern’s distinction, suggesting that concepts and metaphysics are inextricably intertwined, but is this opposition between the event and the predictable predicate a necessary one? Derrida claims it to be a condition of thinking the event, but is every philosophy of mediation, of continuums, necessarily condemned to be unable to think, or account for, the event? It is not clear that this is so. Presumably ‘tipping points’ of various kinds can still be theorized, and the risk is that transcendental reasoning of this sort depends upon a contrast that excludes other possibilities.

4. Conclusion The key explicit objections that analytic philosophers raise – the general objection from empiricism, Stroud’s ‘idealist/verificationist’ objection, and Körner’s worries about the uniqueness of the condition – are sufficient to make one at least appreciate the potential difficulties of transcendental arguments. From two different sides we have developed the point that the explanatory background genuinely differs across the divide. The suggestion is that we are in a situation akin to what Lyotard calls a differend: a dispute in which two parties

46

Postanalytic and Metacontinental

cannot agree on a common rule of judgement, or a meta-narrative which would act as a tribunal of arbitration. In conclusion, let us consider again how this might be so. The empiricist objection has been in the background during much of the above discussion, but it is fundamentally a worry about the ampliative ambitions of the transcendental move, and the methodological disaster that unconstrained transcendental theorizing threatens to visit on philosophy. Moreover, the worry often concerns the naturalistic ambitions of much analytic philosophy quite as much as empiricism itself. How could such reasoning be sensibly incorporated within a framework that often regards the finding of the sciences as epistemically privileged and that seeks coherence between its disparate knowledge claims? Prima facie, not easily, although there are a couple of important points to note here. First, one would need to fill out just what the commitment to methodological empiricism or naturalism here means, since there are many different ways of understanding this, and perhaps the particular version of naturalism being endorsed or presupposed is itself dogmatic or problematic (partes extra partes). That is certainly what Merleau-Ponty would maintain. Second, a wide range of non-positivist variants on empiricism and naturalism have now been developed within analytic philosophy, and so the question of the compatibility of transcendental arguments with such views has to be rethought in this context. John McDowell engages in this kind of project, and it is not a coincidence that most of the major continental philosophers associated with transcendental arguments (including Husserl, Bergson, Heidegger, Deleuze) are either anti-naturalists, or involve a critical and transformative relationship to both nature and the natural sciences. Certainly elements of vitalism can be discerned in the work of the last three philosophers mentioned. In addition, various philosophers working with transcendental arguments in phenomenology and philosophy of mind are engaged in projects of naturalizing phenomenology, or perhaps more aptly ‘phenomenologising nature’ as David Morris puts it.58 On such views, empirical analyses must be relevant, and in a relationship of mutual constraint with any transcendental claim, and transcendental claims must be potentially refutable by empirical analysis, noting that empirical analysis is never completed once and for all, and that the negative heuristics associated with naturalism would be somewhat different from those that typically guide most analytic philosophers. Still, transcendental reasoning on this view earns its keep much like any other theoretical approach with predictive implications. Recall, for a moment, Merleau-Ponty’s analyses of the body. In Phenomenology of Perception, written in 1945, rich phenomenological analyses are coupled with transcendental claims about the primacy of bodily intentionality, along with other arguments that also rely on inference to the best explanation and engagement with the psychology, biology and social sciences of his day. Since that time, detailed analyses of proprioception, mirrorneurons in monkeys and other relevant studies have seemed to confirm some of his claims, and to challenge others. But his analyses have also made possible

The Fate of Transcendental Reasoning

47

some of the subsequent empirical work, both in inspiring it, and framing it; we might, for example, trace a lineage from Merleau-Ponty, to Hubert Dreyfus’s What Computers Can’t Do, to Rodney Brooks’s embodied robotics, and to accounts of the embodied mind that are increasingly influential in cognitive science. The result is a conception of the relative roles of philosopher and scientist that is not too dissimilar from that of naturalizing analytic philosophers (Daniel Dennett, say) – philosophical theorizing is broadly constrained by the deliverances of the sciences, but philosophical accounts of subject matters relevant to the sciences can guide inquiry or play an inspirational or transformative role in theory development. The Stroudian objection lends itself less to such middle-way approaches. There does appear to be a question-begging assumption of something like anti-realism in virtually all non-externalist uses of transcendental arguments (we say ‘something like’ because, while to an analytic philosopher the dictum ‘being is phenomenon’ has to be idealistic, for many phenomenologists it is a middle-way between idealism and realism). Two rejoinders might be offered. One might concede the point, as Charles Taylor does, but nonetheless maintain that transcendental analyses still establish necessary conditions for our own self-relation – for ourselves we are, for instance, necessarily embodied subjects – and insist that this has significance for the humanities and social sciences writ large, even if we should be metaphysically modest regarding what it means about reality per se.59 Second, one might point out that Stroud’s influential criticisms of transcendental argumentation are not premise free. As Anita Avramides notes, the Stroudian position supposes a logical and metaphysical gap between me and everything else, and further presupposes that knowledge has primacy over action.60 From a pragmatic or action-oriented picture, however, this idea of a separated Cartesian subject attempting to know things seems peculiar, as does the co-implicated idea of an objective reality that is independent of subjectivity. It is certainly very different from the dominant holism in continental philosophy in which the subject is co-implicated with the real. This charge of partiality could be extended to the analytic project of classifying and formalizing different argument types. While undoubtedly useful in many contexts, this can also create problems of its own. It seems to one of us that the risk is that a characteristically analytic or decompositional approach obscures the manner in which thought itself is (at least sometimes) synthetic. Effective reasoning is arguably more complicated than this, and for the continental philosopher to think otherwise would be to mistake the means of reasoning for the end; that is, to confuse particular useful tools of thinking with the task itself. In that respect, as we have seen, most continental philosophers will maintain that there is inductive support (or support by inference to the best explanation) for the ostensibly non-inductive method of transcendental argumentation. Even for those not interested in the sciences, the history of Western philosophy would be another such ground, including both its problems and its successes.

Postanalytic and Metacontinental

48

We have seen that transcendental arguments also depend on a certain active involvement. In that sense, they do beg the question against the sceptic, and they do call for a certain abeyance of incredulity (a certain trust) in reading and reflection. Philosophers have to be critical, but none are critical all of the time, and for a transcendental argument to work, one has to be taken along by a story of sorts (a description, a genealogy, etc.), one has to imagine and reconstruct experiences (from an embodied situation), and one has to critically and sceptically reflect in another moment, using various other devices of argumentation. One can, of course, err in navigating these demands. We can be bewitched by sophistry. But this also means that, without some trust, transcendental reasoning will never work, in the sense of being useful or helping to induce a perspectival shift. The analytic community has, by and large, decided not to trust such arguments; the continental community, by and large, has put a significant degree of faith in them. As William James made clear in a more general epistemic context, the precise trade-off between the goal of avoiding false beliefs and the goal of seeking true beliefs can differ from agent to agent, and that is the situation we have here between the traditions.61 The analytic decision is arguably reasonable given the difficulties there are in finding a constructive role that transcendental reasoning can play within the norms of the analytic community. The reasonableness of the continental decision requires consideration of the value of the activity of engaging in this Wager while also critically reflecting, and the value of the perspective that dual combination affords. Jeff Malpas has observed that the comparison of analytic and continental approaches to transcendental arguments might help to bridge the divide or to deepen it.62 Our feeling is that the latter is rather more likely in this case. Perhaps, however, these rather stark differences need to be made perspicuous in order for a genuine conversation to be possible.

Notes 1

2 3 4

5

6

We are indebted to the support of the Australian Research Council, and to James Williams, Andrew Benjamin, Jeff Malpas, Sherah Bloor, Ricky Sebold, Ashley Woodward, Chris Cordner, Jon Roffe, Paul Patton and Robert Sinnerbrink for engaging with some of these views. Some exceptions might include the structuralists and Quentin Meillassoux. Thanks to Sherah Bloor for this idea, which she is developing in her MA thesis. Thomas Baldwin, ‘Death and Meaning’, Arguing with Derrida, ed. Simon Glendinning (London: Blackwell, 2001), 89–102. Robert Stern, Transcendental Arguments and Scepticism (Oxford: Oxford University Press, 2004), 1. Davidson’s reflections on triangulation and his later formulations of the principle of charity are often thought to involve them, and it is perhaps not a coincidence that discussions of his work in central journals of analytic philosophy have dwindled from this date. In Mind and the Journal of Philosophy, those of Davidson’s papers

The Fate of Transcendental Reasoning

7

8

9

10 11 12

13

14

15

16

17

18

19

20 21

22 23 24 25 26

49

that are most cited are: ‘Actions, Reasons and Cause’, ‘Truth and Meaning’ and ‘Mental Events’. In more ‘crossover’ journals, like European Journal of Philosophy and Southern Journal of Philosophy, it is essays after ‘On the Very Idea of a Conceptual Scheme’ that are more frequently cited. Thanks to George Duke for his citational analyses of this. Crispin Wright questions McDowell’s analytic status because of his lack of clarity, rather than because of a use of transcendental arguments, but the issues seem to us interwoven. See Chapter 2, endnote 51. Sometimes this questioning is explicit; more often it is implied by the comparative lack of attention that they (or their late ‘transcendental’ work) receive in the standard-bearing analytic journals. Pascal Engel, ‘Continental Insularity: Contemporary French Analytical Philosophy’. Contemporary French Philosophy. ed. A. Griffiths (Cambridge: Cambridge University Press, 1989), 1–20. See Nick Trakakis, The End of Philosophy (London: Continuum, 2008). See Sami Pihlstrom, Naturalising the Transcendental (Prometheus Books, 2003). What we call a ‘transcendental argument’ is something of an abstraction from Kant’s practices of transcendental deduction and transcendental exposition; Kant does use the term (‘transzendentalen Argumente’) in the Critique of Pure Reason at A627/B655, but in a different sense. The connection between transcendental exposition and transcendental argument is controversial. See Henry E. Allison, Kant’s Transcendental Idealism (Newhaven: Yale University Press, 1986). Karl Ameriks, Kant and the Historical Turn (New York: Oxford University Press, 2006), 38. Bertrand Russell, Human Knowledge: Its Scope and Limits (London: Routledge, 2009), 446. While Russell on occasion maintained that mathematics involved synthetic a priori knowledge, transcendental justifications for such claims are always avoided. See Laurence BonJour, In Defense of Pure Reason, Cambridge: Cambridge University Press, 1998, chapter 2, and Mark Sacks, ‘Naturalism and the Transcendental Turn’, Ratio 19 (2006): 92–106. Such self-application problems could also be viewed as inherent in epistemology itself, whether empiricist or not; consider the diallelus (the problem of the criterion), for instance. One can find a number of earlier uses stretching back further; for instance, see A. J. Balfour’s ‘Transcendentalism’. Mind 3 (1878): 480–505. As far as we can tell, Austin is the first to explicitly generalize the argument form in the modern analytic way. J. L. Austin. Philosophical Papers (Oxford: Oxford University Press, 1961), 33. Austin. 34. In a footnote, Austin makes it clear he is referring to (or generalizing from) the Kantian argument form. Austin, Philosophical Papers, 35–36. Peter Strawson, Individuals (London: Methuen, 1959), 9–10. Peter Strawson, The Bounds of Sense (London: Methuen, 1966). Strawson, Individuals, 25–40. Strawson, Individuals, 35.

50 27

28

29

30

31

32

33

34

35

36 37

38

39

40

41

42 43

Postanalytic and Metacontinental Peter Strawson, Skepticism and Naturalism: Some Varieties (New York: Columbia University Press, 1985), 20–23. Barry Stroud, ‘Transcendental Arguments’, Journal of Philosophy 65 (1967): 241–256. Stephan Körner, ‘Transcendental Tendencies in Recent Philosophy’, The Journal of Philosophy, 63 (1966): 551–561, and S. Körner, ‘The Impossibility of Transcendental Deductions’, The Monist, 51, no. 3 (1967): 317–331. Actually, the required premise is even stronger. Stroud points to the further need to distinguish between all such conditional necessities and necessities simpliciter; even if there could be no other conceptual schemes, it can still be the case that the existence of subjects is contingent, and in this case the inference will still only provide a contingent conclusion. Graham Oppy, Ontological Arguments and Belief in God (New York: Cambridge University Press, 1995), N. Everitt, The Non-Existence of God (London: Blackwell, 2004), J. Sobel, Logic and Theism (New York: Cambridge University Press, 2004), D. Dombrowski, Rethinking the Ontological Argument: A Neoclassical Theistic Response (Cambridge: Cambridge University Press, 2006). Roy Sorensen, Thought Experiments (Oxford: Oxford University Press, 1992), 153–157. In the case of Kant, such warnings are backed by several awful historical examples – most egregiously, for the analytic tradition, his confidence in the finality of the logic of his day and his sandbagging by the rise of non-Euclidean geometries. The attitude expressed here is set out and defended in W. Lycan, ‘Moore against the New Skeptics’, Philosophical Studies 103 (2001): 35–53. P. Van Inwagen, The Problem of Evil (Oxford: Oxford University Press, 2006), 37–55. D. Antiseri, The Weak Thought and its Strength (Ashgate: Aldershot, 1996). Stiegler says ‘I think one must pass through the transcendental in order to get beyond the transcendental. The misunderstanding between continental philosophy and Anglo-Saxon philosophy relates to this point, for one cannot short-circuit transcendental experience; it is impossible.’ See Bernard Stiegler, ‘Technics, Media, Teleology’, Theory, Culture and Society, 24, No. 7–8 (2007): 340. See Karl Apel, ‘The Problem of Philosophical Foundations in Light of a Transcendental Pragmatics of Language’, After Philosophy: End or Transformation, ed. Kenneth Baynes, James Bohman and Thomas A.McCarthy (Cambridge: MIT Press, 1987). He sees his work as ‘mediating between the philosophy of consciousness and analytic philosophy of language’ (p. 271). Alan Murray, ‘Philosophy and the Anteriority Complex’, Phenomenology and the Cognitive Sciences, 1 (2002): 27–47. Some would claim that place subtends this opposition. See, for example, the work of Jeff Malpas and Edward Casey. Mark Sacks, ‘The Nature of Transcendental Arguments’, International Journal of Philosophical Studies, 13, No. 4 (2005): 444. Ibid., 446. Michel Foucault, The Order of Things: Archaeology of the Human Sciences (New York: Vintage Books, 1970), and Michel Foucault, Archaeology of Knowledge, trans. A. Sheridan (London: Routledge, 2002). Indeed, even in Foucault’s middle genealogical period, which appears to be resolutely anti-transcendental, it might

The Fate of Transcendental Reasoning

44

45

46

47

48

49

50

51

52

53

54 55

56 57

51

still be maintained that his genealogies derive their force from reflexion between the performative act of thinking about the conditions of that and other thoughts/ experiences. Recent cases of people without proprioception complicate this view, but do not necessarily refute it. See Shaun Gallagher, How The Body Shapes the Mind (Oxford: Oxford University Press, 2006), and Jose Bermudez, ‘Transcendental Arguments and Psychology: The Example of O’Shaughnessy on Intentional Action’, Metaphilosophy, 26, No. 4 (1995): 379–401. Hubert Dreyfus, and Sean Kelly, ‘Heterophenomenology: Heavy-handed Sleightof-hand’, Phenomenology and Cognitive Sciences, 6 (2007): 45–55. Every figure has an optimal context in which to view it, for example. See Sean Kelly, ‘Seeing Things in Merleau-Ponty’, Cambridge Companion to Merleau-Ponty, Taylor Carman, and Mark Hansen eds (Cambridge: Cambridge University Press, 2005): 74–110. The suggestion that phenomenological arguments are committed to infallibility or incorrigibility is convincingly disputed in Taylor Carman’s ‘On the Inescapability of Phenomenology’, Phenomenology and Philosophy of Mind, D. Smith and A. Thomasson eds (Oxford: Oxford University Press, 2006). See Robert Brandom, Tales of the Mighty Dead (Cambridge, MA: Harvard University Press, 2002). Charles Taylor makes this point well. See ‘The Validity of Transcendental Arguments’, Philosophical Arguments (Cambridge, MA: Harvard University Press, 1995). See Jose Bermudez, ‘The Phenomenology of Bodily Awareness’, Phenomenology and Philosophy of Mind, D. Smith, and A. Thomasson eds (Oxford: Oxford University Press, 2006): 295–316. One of us elsewhere argues that a ‘temporal turn’ gives continental philosophy in the twentieth century a minimal philosophical unity. See Jack Reynolds, ‘Continental Philosophy and Chickening Out’, and ‘Reply to Glendinning’, International Journal of Philosophical Studies, 17, No. 2 (2009): 255–272, 281–287. That is, it is the ground without which there is no Dasein at all. Heidegger argues that what he calls ‘understanding’ depends upon the futural ecstasis of time; that ‘attunement’ (or mood) is structured by the past ecstasis of time (throwness, having-been); and that ‘fallenness’ is an attachment to present things that denies these other escstases. See Gilles Deleuze, Difference and Repetition, trans. Paul Patton (New York: Columbia University Press, 1994) – all of the following citations of this book come from chapter 2 – and Gilles Deleuze, Kant’s Critical Philosophy (Minneapolis: University of Minnesota Press, 1985). Ashley Woodward, unpublished paper. Jacques Derrida, On Touching – Jean Luc Nancy, trans. C. Irizarry (Palo Alto: Stanford: Stanford University Press, 2005), 57. See Robert Stern, Transcendental Arguments and Scepticism, 10–11. Derrida notes that he is an ‘ultra-transcendentalist or quasi-transcendentalist’ and explains this in terms of his concern to avoid empiricism or at least certain forms of it. See Arguing With Derrida, ed. Simon Glendinning (Oxford: Blackwell, 2001), 107.

52 58

59 60 61

62

Postanalytic and Metacontinental Morris suggests this may herald a new conception of nature and hence naturalism: ‘nature is not space-time-matter unfolding according to laws, nature is moving being organising.’ We cannot address this here, except to say there is some support for this kind of view within continental-inspired philosophy of mind. See David Morris, ‘Continental Philosophy of Mind’, Columbia Companion to Twentieth Century Philosophies, ed. C. Boundas (New York: Columbia University Press, 2007), 535, 541. Charles Taylor, ‘The Validity of Transcendental Arguments’, 25. Anita Avramides, Other Minds (London: Routledge, 2001). William James, The Will to Believe and Other Essays in Popular Philosophy (New York: Dover Publications, 1956). Thanks to David Coady for the analogy. Jeff Malpas, ‘Introduction’, From Kant to Davidson: Philosophy and the Idea of the Transcendental (Cambridge: Cambridge University Press, 2005), 5–6.

Chapter 4

Logic and Metaphysics: Dummett Meets Heidegger Edwin Mares1

1. Introduction This book is supposed to be a rapprochement between analytic and continental philosophy. As such, it is fitting that this be a chapter about logic but take its theme from two books, one by Dummett and the other by Heidegger. The titles of the two books are rather similar: Dummett’s is The Logical Basis of Metaphysics (henceforth ‘LBM’) and Heidegger’s is The Metaphysical Foundations of Logic (henceforth ‘MFL’).2 The background to this chapter is the problem of choosing a logical system among all the alternatives available to us. The specific question that we will deal with here is whether a logical system is required to construct and justify a metaphysical view or whether a metaphysical theory is required to justify a choice of a logical system. Dummett is an advocate of the former view, which we will call the ‘logic first view’. Heidegger is an advocate of the latter view, which we call the ‘metaphysics first view’. I argue that neither view is satisfactory. I suggest instead that the reasoning process that we should use to choose both a logic and a metaphysics is a form of reflective equilibrium. The outline of the chapter is as follows. In the first few sections, we see the logic first view. I give an ad hominem argument against it – that even Dummett could not maintain his belief in the system (I think for good reason). We then look at Heidegger and the metaphysics first view. Although this view seems in some ways more intuitively satisfying, it does not accord with much of what I think is good practice in philosophy. In the last few sections, I propose a process of finding reflective equilibrium that philosophers should (and, I think do) use to bounce views about logic and metaphysics off one another. There are worries about both the terms ‘logic’ and ‘metaphysics’. Heidegger regularly uses ‘metaphysics’ in a disparaging way. Metaphysicians have, on his view, avoided the fundamental question of philosophy by creating theories about what there is. They have replaced the question of the nature of being by examining the nature of particular things (beings) instead. Luckily for me, I do not need to examine Heidegger’s notion of fundamental ontology. Rather, we only have to examine

54

Postanalytic and Metacontinental

the relationship of logic to a very general notion of metaphysics, as the theory of what there is, regardless of whether it is mistaken (as Heidegger thinks is the case in most post-Platonic philosophy) and so we can take this term as covering both Heidegger and the philosophers of the analytic tradition. The use of the term ‘logic’ is similarly problematic. Even among logicians the word ‘logic’ means at least two different things. Some logicians take ‘logic’ to be synonymous with a proof theory. A proof theory is, roughly, a set of rules and/or statements (the theorems of the logic). Most philosophical logicians, however, take a logic to be a proof theory together with an interpretation. I will mostly use ‘logic’ in this second sense, but I have found it unavoidable sometimes to distinguish between a logic and its interpretation. I hope no confusion will result because of this. An interpretation is, of course, a way of understanding the meanings of the symbols in the logical language. Some interpretations seem already laden with metaphysics – such as possible world semantics for modal or counterfactual calculi – but, as we shall see, even these can be drained of their ontological commitments. Other interpretations, such as the interpretation of the intuitionist calculus that we shall meet in our discussion of Dummett claim from the outset to be devoid of metaphysics. Some proof theories have more than one interpretation. The intuitionist calculus, for example, has several.

2. Let us calculate There is an ancient tradition of taking deductive logic to be a neutral form of reasoning and debate. It has been used to support or refute metaphysical views at least since Zeno’s use of dialectic to deny the reality of motion.3 The late mediaeval philosophers carefully put every inference they discussed into syllogistic form, discussing first the major premise and then the minor, to defend their own views and attack those of their opponents. This view of logic receives its strongest expression in Leibniz’s view that his logical system – his Characteristica Universalis – is to provide a tool with which philosophers can settle all disputes. Here is a famous passage from Leibniz, as translated by Bertrand Russell: If controversies were to arise, there would be no more need of disputation between two philosophers than between two accountants. For it would suffice to take their pencils in their hands, to sit down with their slates and say to each other (with a friend as witness, if they liked): Let us calculate.4 Logic was supposed to be a tool with which one could arbitrate in metaphysical disputes. As such, one logical system would have to be universally accepted among philosophers. Unfortunately, this is not the case. In the following section, we look at an attempt to justify the use of a single logical system in a

Logic and Metaphysics: Dummett Meets Heidegger

55

non-metaphysical way and then apply it more or less as this tradition would have us do.

3. Logic first: Dummett Perhaps the most forthright contemporary advocate of the view that logic precedes metaphysics is Dummett. Here we will look at his position in LBM. There Dummett sets out a programme of settling metaphysical questions through the logical analysis of language. At first, this sounds like a revival of a method that was used by the logical positivists. But the positivists wanted to eliminate metaphysics through the logical analysis of language. Dummett does not treat metaphysical questions as pseudo-questions.5 He just thinks that we can solve metaphysical questions through the proper use of logic. In this, he is a modern successor to Leibniz. Dummett approaches logic through a theory of meaning. A theory of meaning, moreover, can only be justified to the extent that it predicts and explains human linguistic behaviour. To determine whether a theory of meaning fits with our linguistic behaviour is an empirical matter, not an enterprise in metaphysics: The task of constructing a meaning-theory can, in principle, be approached without metaphysical presuppositions or arrière-pensées: success is to be estimated according as the theory does not provide a workable account of the practice that agrees with that which we in fact observe.6 This theory of meaning is then used to justify the adoption of a logical system. The theory of meaning will tell us what the meanings of the logical constants are.7 On Dummett’s view, the meaning of an expression is its use. When he applies this Wittgensteinian theory of meaning to the logical constants, he claims that the meaning of a constant is its behaviour in a system of proof. Thus, our observations of individuals’ inferential behaviour (and our own logical intuitions) are used to support the adoption of a logical theory (in Dummett’s case, intuitionist logic). Using this method, Dummett thinks that we can justify the acceptance of a particular logical system without appeal to any metaphysical principles. Intuitionist logic as a theory of deductive inference is part of Dummett’s general theory of meaning – his justificationism. Justificationism is a generalization of the intuitionists’ view of mathematics. Intuitionism takes the meaning of a mathematical statement to be the conditions under which that statement is proven. Justificationism holds that the meaning of a statement is the set of conditions under which there are ‘grounds sufficiently compelling to warrant asserting the statement’.8 In fact, the meaning of a statement is given only by its canonical justification conditions. Consider, for example, the statement ‘either

56

Postanalytic and Metacontinental

there is some leftover spaghetti or pizza in the refrigerator’. The canonical method for justifying this statement is to look and verify that at least one of the disjuncts obtains. There are other non-canonical means such as recalling what one had for dinner last night and reasoning from there. Restricting meaning to the set of canonical methods of verification allows Dummett to claim that the meaning of a statement is a finite set and can be learned. We will not go through Dummett’s arguments for that position here. It is not the theory of meaning itself or its support about which we are concerned. Rather, it is the link between meaning and logic and the relationship between logic and metaphysics that interests us. As we have said, justificationism is a generalization of the intuitionist treatment of mathematical statements. Consider for example, the statement ‘2 + 2 = 4 and 3 + 5 = 8’. According to intuitionism, the meaning of that statement is given by the (canonical) method of proving both of its conjuncts. More controversially, intuitionists hold that the canonical means for justifying the assertion of a negated sentence, not-A, is to show that from the assumption of A a contradiction can be derived. One controversial aspect of this interpretation of negation is that it requires that we have an idea of a contradiction that precedes our understanding of negation. Another controversial aspect of this treatment of negation is that it leads us to deny the principle of bivalence, viz., that for any sentence either that sentence or its negation is true (or, for justificationism, that it can be justified). In ordinary discourse this is even clearer than it is in mathematics. There are always statements that are neither provable nor entail a contradiction (given our present information). For example, we currently can neither prove nor disprove that there is life on other planets. So, for the statement ‘there is life on other planets’ Dummett claims that the law of bivalence fails. If we have a collection of sentences among which are some that fail this test of bivalence, then we can be said to be anti-realistic about that set of sentences. For example, if there are some sentences about quantum mechanics which are neither verifiable nor refutable (i.e. the assumption of which allows us to derive a contradiction), then we must accept anti-realism about quantum mechanics. Thus, on Dummett’s view, the theory of meaning for the logical connectives gives us a test for realism about any given realm of discourse. Dummett says: Will it also settle the metaphysical controversies themselves? It is my contention that it will. In the process of constructing a general model of the meaning of a sentence belonging to each sector of the language, the theory will elucidate the concept of truth, as applied to statements belonging to that sector – statements about physical reality, mathematical statements, statements in the past tense, or the like – by setting that concept in its proper place in the characterization of the meanings of those sentences. It will thus adjudicate between the rival conceptions of truth advocated by realists and anti-realists.9

Logic and Metaphysics: Dummett Meets Heidegger

57

What is of interest to us here are two points: (1) Dummett thinks that we can justify the acceptance of a particular logic and accompanying theory of meaning without any metaphysical presuppositions; (2) the logic and theory of meaning themselves will settle every question of the reality of any disputed realm and there will be no further metaphysical questions to be asked. On this last point, Dummett is extremely explicit: It will resolve those questions without residue: there will be no further, properly metaphysical, question to be determined.10 Thus, Dummett attempts to make Leibniz’s dream come true of having a logical means of deciding metaphysical disputes. Should we accept the claim then that logic can be used in this way? The case of Dummett is instructive in this regard. Although he constructed this logical method, there were times at which he could not bring himself to accept the results of applying it. In particular, applying the method to the problem of the reality of the past gave Dummett the wrong answer: I confess to being in just the position of the thinker inclined to a justificationist theory of meaning but troubled by the reality of the past. I am indeed attracted by a justificationist theory of meaning; at the same time, I have long been worried about reconciling the reality of the past with that account. Many years ago, I wrote and published an article to explore the issue. When I began to write it, I had hoped to arrive at one or other conclusion: that antirealism about the past was a benign and acceptable view; or that it was incoherent, and that its incoherence would expose a fallacy in the argument for a justificationist theory of meaning. In the result, the conclusion that I reached was the most disappointing possible. Antirealism about the past was not incoherent; but it was not believable, either.11 I think this shows that there is a problem with the logic first point of view. Looking at Dummett’s response to his own method is not merely to give an ad hominem argument. Rather it demonstrates that our metaphysical intuitions are often stronger than our intuitions about semantics and logic. What Dummett does in Truth and the Past in response to his qualms about anti-realism about the past is to move his theory of meaning more towards a truth conditional theory. Thus, Dummett in this case allows his semantical (and logical) ideas to be influenced by his metaphysical intuitions. And he seems perfectly justified in doing so. Let’s turn now to the metaphysics first point of view.

4. Metaphysics first: Heidegger Heidegger is often very disparaging about logic. Logic, on Heidegger’s view, has become ‘the discipline that has most of all succumbed to induration and

58

Postanalytic and Metacontinental

separation from the main problems of philosophy’.12 He thinks logic is often treated (by people like me) as an overly abstract field that has little to do with the rest of philosophy. This abstractness is not just a problem for logic, but rather it is a problem for philosophy as a whole.13 The problem is that some issues get ‘set aside’ into logic – they are treated as purely logical problems and then their metaphysical core is ignored because of the abstract nature of logic as logicians treat it. In The Basic Problems in Phenomenology, for example, he says that the issue of the nature of the copula (‘the question of being in the sense of the copula’) has been treated as a properly logical problem. But then logic, because it claims to abstract away from particular philosophical content, does not treat the problem of the real meaning of the copula seriously. In becoming separate from the rest of philosophy, logic took the problem of the copula with it and eliminated it as a philosophical problem. In this way Heidegger thinks that the treatment of logic as an abstract discipline has been destructive to philosophy. But we will set aside Heidegger’s value judgments about logic. Rather, we want to see how he thinks that logic and metaphysics are linked. There are two senses in which Heidegger thinks that logic presupposes metaphysics. First, metaphysics is needed to show that logical thought is possible. Second, logic requires a grounding in metaphysics. The second is the main theme of his book, and it will take a bit of setting up to explain. The first, however, is not that difficult. Logic, as Heidegger recognizes, is a normative discipline. In order to think logically, one needs to make his or her thoughts conform to the rules of logic. To be deemed to be obeying these rules, Heidegger thinks that one needs to do so freely: Only what exists as a free being could be at all bound by an obligatory lawfulness. Freedom alone can be the source of obligation. A basic problem of logic, the lawfulness of thinking, reveals itself to be a problem of human existence in its ground, the problem of freedom.14 This is an extremely Kantian position. For Kant, only a free being can be said to do his or her duty.15 Kant goes further than Heidegger, however, and claims that only a person who is doing his or her duty is actually free.16 This is because Kant has a different view of freedom than Heidegger. For Heidegger, freedom is not something that humans can fail to have. Human beings are themselves ‘the possibility of freedom’.17 Let’s set aside the content of Heidegger’s claim – this chapter is not an investigation into the real nature of logic – to see what exactly is going on here. What we have is a passage that is to a large extent Kantian, not only in content, but also in method. We have, in effect, a transcendental argument for free will. Free will is a presupposition of our being able to obey the dictates of logic. Thus, if we are to hold that we can obey the dictates of logic, we must be free.

Logic and Metaphysics: Dummett Meets Heidegger

59

The argument may not convince many of us, but it does usefully point out one way in which logic and metaphysics have been bound together. There are various transcendental arguments based on the possibility of logic in the literature. For example, the way in which Frege postulates his ‘third realm’ can easily be recast as a transcendental argument. Frege’s central premise is that the normative force of logic presupposes that logic be about entities that are extramental. The second premise is that, if the truths of logic are to be eternal, these objects cannot be physical. Thus, Frege concludes that there is a third realm; that there are objects that are neither mental nor physical, that are abstract. This too is an argument that the possibility of logic requires a metaphysical posit. Heidegger’s main aim in MFL is to show that the specific content of certain logical systems requires a metaphysics for its ground. Before we can discuss his argument for this we need to understand some of what Heidegger means by ‘ground’. I shall not plumb the full depths of Heidegger’s notion of a ground – that is not needed for our purposes here. Rather, we can think of ‘ground’ as having two central elements to its meaning. First, there is the notion of a source (what something is at bottom – from its ground). Second, there is the reason or justification for that thing.18 The second meaning will have the most importance for us here. We use a metaphysical view as the justification for a logical principle, say, when we infer that if something occurs then it will always be the case that it has occurred. This inference presupposes the fixity of the past and it can be said to be justified by the view that the past is in fact fixed. Heidegger’s main argument for his thesis is a case study. He looks at Leibniz’s logic and claims that his view of the logical analysis of subject-predicate expressions presupposes particular features of Leibniz’s metaphysics.19 One again, our purpose in examining Heidegger’s argument is metaphilosophical. We are not concerned with the soundness of his argument, nor are we concerned with the correctness of his interpretation of Leibniz. For Heidegger, as for many philosophers, the central notion of logic is that of truth. So, in order to study Leibniz’s logic, he examines Leibniz’s theory of truth. The truth or falsity of statements for Leibniz is rooted in predication. A true subject-predicate statement is true because the predicate is appropriately connected to the subject.20 In an identity statement, this connection is of course one of identity. All other subject-predicate statements are really hidden identity statements – they are to be analysed to make explicit the identities that they claim to hold.21 A contingent truth is one for which we cannot do the analysis; the analysis requires an infinite number of steps. But for the mind of God, all truths are equal – they are determined by identities that follow from the essences of the individuals in the world.22 The complete concept of an object contains all its predicates.23 A true predication of an individual tells us, in part, what the individual is. It is thus an identity statement of a sort. Thus, the claim that we can reduce all true predications to identity statements rests on (or is grounded in) Leibniz’s view that the complete concept of an

60

Postanalytic and Metacontinental

individual substance contains all its predicates. This view of concepts is based on Leibniz’s view of substances as self-contained monads. The world is made out of monads. Thus, Heidegger characterizes Leibniz’s view as ‘spiritualism’: he takes the world to be constructed from spirits or ‘little gods’.24 Each monad, moreover, represents the entire world. Leibniz’s doctrine of pre-established harmony is supposed to solve the mind-body problem by having the universe made up of monads which represent one another and act in parallel in a manner consistent with one another. When God creates the world he chooses which monads to create. In each of them is contained a total representation of the world – including its relations to other monads. Thus, all true predications are determined by the concepts (God’s concepts) of the individual substances. What is interesting for us (and for Heidegger) in all of this, is that Leibniz’s notion of logical analysis is grounded in (i.e. justified by) a particular metaphysical view, that is, his theory of substance. There is no metaphysical neutrality here. This view of analysis would not (and has not) transferred to other philosophers who hold different ideas of substance. Heidegger also holds that metaphysical theorizing does not presuppose any particular logical views. I am not going to argue that this is false throughout the history of philosophy (although I believe it is).25 But the sort of metaphysics that I am interested in, and discuss in the following section, provides us with a foundation for particular logical systems (such as modal and counterfactual logics) and does presuppose specific logical doctrines.

5. Logic and metaphysics in contemporary philosophy Does logic need a metaphysical justification? In some contexts it does. If the logical validity of inferences is supposed to be analysed as truth preservation,26 then we need something to ground the truth of premises and conclusions of inferences. There are some clear cases in which philosophers have allowed their metaphysics to shape their logical views. Quine, at various times in his career would refuse to use second-order logic because having quantifiers binding predicate variables would commit him to accepting universals or classes in his ontology.27 Another example is David Lewis’ abandonment of modal logic in favour of a first order logical theory that directly represents his counterpart theory. Although he originally put forward his counterpart theory as a semantics for modal logic, he also recommended that we talk directly in terms of counterparts in a first order logic in which the modal operators were replaced by analyses in terms of relations. Later it turned out that various standard theses of modal logic were not valid in his semantics, and his reply was to completely abandon standard modal logic.28 So, it is clear that Lewis put his metaphysical views ahead of well-accepted logical principles.

Logic and Metaphysics: Dummett Meets Heidegger

61

But what may be even more common than the reflection of a philosopher’s metaphysics in her logic is the postulation of a metaphysical structure in order to justify one’s logical views. For example, Frege wanted to reduce arithmetic to logic. In order to do so, he had to find logical derivations of statements such as ‘the addition of any two numbers results in a number’. In order to prove this, he allowed expressions that referred to numbers as if they are objects. Given his views about logic and reference, he had to postulate an infinite number of ‘logical objects’ (his courses of values). An example that comes to mind almost immediately in thinking about philosophers’ postulating entities to justify specific logical laws is the postulation of possible worlds to justify the laws of the various modal logics. Modal logic – as a branch of modern symbolic logic – dates back to at least the 1910s. Possible world semantics, as a semantics for modal logic, dates back only as far as the 1940s. And the founder of modern modal logic, C.I. Lewis, was never an advocate of possible worlds. C.I. Lewis offered another interpretation of logical necessity. On his view, a statement is necessary if and only if it is analytically true, and for him analyticity was not a metaphysical matter.29 After Quine’s attack,30 the notion of analyticity fell out of favour with logicians and with it Lewis’ interpretation faded in popularity. Another interpretation of modal logic was needed, and that came in the form of possible worlds. What is happening in these last two cases is that philosophers have a logical intuition and this shapes their metaphysics.31 This is different than what happens according to the logic-first view. According to the logic-first view, we should have a fully interpreted logical system and use it to determine our metaphysical theory. Similarly, the metaphysics first view says that the logic should be a reflection of one’s metaphysical view. Here instead logical intuitions are used to shape a metaphysical theory, but the metaphysical view is used afterwards to interpret the logical ideas. In recent logical theory there are, however, some philosophers who have given decidedly non-metaphysical treatments of (fragments of) logic. For example, according to Ernest Adams, Dorothy Edgington, and others, conditionals play an epistemological role in our thought and talk, but do not describe the world. As such, they are given an epistemological foundation rather than a metaphysical one. Perhaps a more extreme example of an anti-metaphysical treatment of logic is the fictionalist treatment of possible world semantics. Gideon Rosen and others have proposed that the theory of possible worlds is a mere fiction that is appropriate in some way to provide truth conditions for modal statements; they help themselves to the logical power of possible world semantics without buying into its metaphysical commitments. Fictionalists, however, are committed to providing a reason (a ground) for the acceptance of this fiction – why it rather than some other fiction is appropriate. So, even in these cases, we can agree with Heidegger when he says that logic is not ‘free floating, something ultimate’.32

62

Postanalytic and Metacontinental

Where we do disagree with Heidegger is in his claim that metaphysics is prior to logic. As we have seen, logicians can begin with logical views and then proceed to develop metaphysical views just as they may do the reverse. Logical reasons can be good reasons also to abandon metaphysical views. Thus, we need a more complicated view of the relationship between metaphysical theories (and epistemological theories, pragmatic theories, and so on).

6. Reflective equilibrium The obvious way to understand the justificatory relationship between metaphysics and logic is, in my opinion, through a form of reflective equilibrium. Philosophers try to construct a metaphysics to satisfy their logical beliefs and their logic to satisfy their metaphysical beliefs. Sometimes (often in fact) there is no perfect fit. In order to maintain coherence some beliefs have to be modified or erased. And others may have to be added (such as, for example, the belief in possible worlds when one comes to accept modal logic). The notion of coherence involved is some sort of explanatory coherence. Since the aim is to get a justification of logical and metaphysical statements by one another, some sort(s) of explanatory connections must hold between them. Thus, one goal of this process of reflective equilibrium is to attain as coherent a set of judgments as possible. Dummett’s move away from justificationism can be interpreted in terms of this sort of process. He has a well-entrenched belief in the fixity of the past. That is, he believes that statements about the past are bivalent. But there are statements about the past that can neither be justified nor falsified. So, his response was to abandon pure justificationism. In Dummett’s case, we can think of his move away from justificationism to a large extent as a way of maintaining the consistency of his beliefs. But not all philosophers of logic believe that we should have consistent beliefs. Some of them even believe that there are true contradictions. So, we cannot impose the condition that our beliefs should be consistent. To do so would be question begging. Instead, I propose that we use a concept that should be amenable to a very wide range of philosophers. This is the notion of pragmatic consistency.33 To explain this notion, we start by understanding the distinction between accepting and rejecting a statement. The use of this distinction in a theory of belief revision is given in my paper ‘A Paraconsistent Theory of Belief Revision’.34 The theory of that paper is quite technical and it differs slightly from the view that we give here, so I will not go into any detail about it.35 Assertion and denial, or acceptance and rejection, are attitudes towards statements. There are certain statements that one believes or accepts. There are also certain statements that she denies or rejects. And there are others that she refrains from judging about. We do not want to reduce the rejection of a statement to the acceptance of the negation of that statement, since (given the constraint of pragmatic consistency that we will state below) this will force the issue towards classical logic.

Logic and Metaphysics: Dummett Meets Heidegger

63

Most non-classical logicians have to accept the distinction between rejecting a statement and accepting its negation. For example, in intuitionist logic, the law of excluded middle – for all statements, either it or its negation obtains – is not a theorem. Dummett, for example, rejects excluded middle. But he does not assert its negation. Rather, its negation is a contradiction according to intuitionist logic. A person is pragmatically consistent if and only if his or her beliefs do not entail any statements that he or she rejects. This definition uses a logical notion, the concept of entailment.36 Thus, it presupposes that we have some logic. Now we have what looks like a thorny problem: how can we justify using a logical system when we are trying to determine the right logic? There are two answers to this question. First, what is needed for the justification of this logic is not a metaphysical ground (if it were, then we would be stuck in a nasty circle). Rather, it is an epistemological justification. Second, the logic required can be very weak. This means that we can use a logical system that has very few logical rules and theorems.37 In fact the logic is so weak that few philosophers would reject any of its rules or theorems. Thus, whereas I do not claim (with Heidegger) that metaphysical theorizing does not require any specific logical views, I do hold that metaphysical and logical theorizing can be done together with the aid of a logic that requires only certain logical views that almost every contemporary philosopher or logician would accept. (But just because we all accept the logical theses required does not mean that they are philosophically neutral. They may well have important metaphysical baggage.) I have gone back and forth about the status of pragmatic consistency. It may just be a value – a normative constraint on people’s judgments. But it also might be a condition of rejecting a statement: if a person accepts a statement then any attempted rejection of it fails (at that time). I’m not sure which to believe, but luckily it will not matter in this context. Although if I do opt for the merely normative account I will have to give a ground (to use Heidegger’s terminology) for that norm. In the remaining sections of this chapter, I examine other features of this form of reflective equilibrium. We will look at the role of interpretations of logical systems in the inferential process, and we will discuss a particular logicalmetaphysical debate to give the theory more detail and to show how the theory can be applied.

7. Interpretations as connecting principles In order to be able to bounce logical and metaphysical views off one another in the process described above, we need to be able to tell when they conflict or entail each other. To do so, we need some connecting principles. Interpretations of logical systems can do this, at least to some extent.

64

Postanalytic and Metacontinental

Consider, for example, Dummett’s interpretation of intuitionist logic. On that interpretation, the meaning of a statement is the set of its canonical justification conditions. In order to formalize this, we could add singular terms that denote justification conditions to our language. This might seem like an odd thing to do, but Per Martin-Löf does it in his influential interpretation of intuitionist logic.38 Then we can talk about purported entities that can be connected with statements about the nature of the world.39 Another sort of interpretation is a model theoretic interpretation of logic. In 1930 Alfred Tarski introduced this sort of interpretation into logic (and philosophy).40 Tarski sets out the notion of truth in a structure. In his original work, this structure is supposed to be the real world, but in subsequent work by him and others, a variety of structures – models – are used. What was most important about Tarski’s semantics is that it gives an inductive definition of the truth of sentences. Given some basic clauses that state to which objects names refer and which sets of entities are in the extensions of predicates, Tarski’s theory tells us the truth values of every statement in a (logical) language. Out of Tarski’s theory grew the truth conditional theory of meaning. On this theory of meaning, we understand declarative sentences by knowing their truth conditions. Now, there is a lot of controversy about this view of meaning, but we will assume it. What is wonderful about the truth conditional theory of meaning, when combined with a model theory view of truth, is that it tells us how we understand the meanings of complex statements given an understanding of the meanings of simple statements (and how we understand those if we understand the basic notion that a simple subject-predicate statement is true if the subject refers to a thing that is in the extension of the predicate). In other words, on this view of meaning, meaning is compositional – the meaning of a sentence is determined by the meanings of its parts.41 In the following section we look at an example of a debate among logicians. To set up that debate, we need to know a little bit about possible world semantics. Here is a very simple definition of a possible world model for the logical system S5. A possible world model consists in a set of entities (which we will call ‘worlds’) together with a function known as a value assignment. We are not going to consider predicate logics (logics with quantifiers) – just propositional logics. Our value assignment tells us which simple sentences (atomic formulas) are true at which worlds. A sentence A and B is true at a world if and only if both A and B are true at that world. A sentence not-A is true at a world if and only if A fails to be true at that world. A sentence A or B is true at a world if and only if at least one of A or B is true at that world. And a sentence it is necessary that A is true at a world if and only if in every world in the model A is true. A sentence is said to be valid on the model if and only if it is true in every world in that model. A set of sentences is said to semantically entail a sentence A if and only if in every world in which every sentence of that set is true A is also true. A rule is said to be valid in a model if every instance of it is such that its premises semantically entail its conclusion in that model.

Logic and Metaphysics: Dummett Meets Heidegger

65

A logical system is sound and complete over a class of models if and only if all of its rules and theorems are valid in every model.42 Proofs of soundness and completeness – the staples of a philosophical logician’s career – show that a class of models meet the basic formal requirements for being an interpretation of a logic. There are, of course, philosophical requirements for a model theory’s acting as an interpretation. The foremost of these is that it leads people better to understand the logical system. Consider the advent of possible worlds semantics for modal logic. The notion of necessity was previously often despised (especially by Russell and Quine) for being unintelligible. Then, possible world semantics gave us at the very least a heuristic for understanding that notion. In imagining whether something could happen in a possible universe, we have a philosophical toe-hold on necessity, and one that is used in a huge number of arguments and explanations in our field. Model theory is particularly of interest to us because it seems to wear its ontological commitments on its sleeve. Possible world semantics, for example, seems committed to the existence of possible worlds. As we shall see, some philosophers try to accept this theory without accepting the existence of worlds. Note that all model theoretic semantics presupposes the existence of sets. The theory is formulated in the language of set theory, which is itself formulated in terms of some logic. We won’t be looking at the debate over the existence of sets here.

8. Case study: ex falso and the problem of impossible worlds One tradition uses logic as a way of reasoning, not just about what is actually true, but about the contents of stories, the contents of databases, the contents of people’s beliefs, and so on. Beliefs, stories, databases, and the like can contain contradictions. Classical logic, however, contains the rule that any statement can be derived from a contradiction. Schematically, this can be represented as A Not-A ∴B Here B is arbitrary – it can be any statement, even one that has no connection with A. This rule is often called ex falso quodlibet. Most philosophers do not want to say that inconsistent stories contain every statement nor do they want to say that a person who believes a contradiction in effect believes every statement. On the simple version of the model theoretic conception of validity, in order to make a rule invalid, we need to accept a model in which there is at least one

66

Postanalytic and Metacontinental

world in which at least one instance of inference does not preserve truth. This means that we need a model in which there is a world in which for some statement A both A and not-A are true and in which there is some statement B that fails to be true. Thus, in order to invalidate ex falso on the simple version, we have to have a world in which there is a true contradiction. In the literature, worlds in which there are true contradictions are known as impossible worlds, but in a sense this is a misnomer. For some philosophers – like Graham Priest and Richard Sylvan – think there are contradictions true in the actual world. But we will follow the tradition here and continue to refer to them as impossible worlds. In order to include impossible worlds in our semantics, we have to reject the truth condition for negation given in the previous section. If we are to have both A and not-A true in the same world, then we cannot have not-A true just in case A fails to be true. There are various ways to deal with this, but we do not need to list them here. The issue that we are to deal with is whether there are impossible worlds. Now, it might seem that this isn’t much of a problem at all. For we have only claimed that we need a model in which there is an impossible world. We didn’t say that this is the intended model. But for some philosophers, such as David Lewis, the only model that can determine which rules are valid is the intended model. He thinks that we should model our beliefs, stories, and so on, using only the ‘real’ worlds. And there are other philosophers, such as Peter Schotch and Ray Jennings, who have semantical reasons for rejecting impossible worlds. They think that semantics with impossible worlds get the truth conditions for the logical connectives wrong. There are then two ways of dealing with the problem of ex falso. Either we allow impossible worlds in our models (and alter the truth conditions for the connectives) or we modify the concept of logical validity. The second approach is taken by Peter Schotch, Ray Jennings and David Lewis (among others). Let’s consider the case of an inconsistent story to explain the view. On their view, a story represents not a single set of worlds, but a set of sets of worlds. Suppose that the statement that A is true in the story and so is the statement not-A. On this view, this means that there is one of these sets in which A holds in every world and one set in which not-A holds at every world. But they do not allow that there is any world in which both sentences are true. Then, on their view, ex falso is blocked because we can construct such a model in which there is some statement that is not true throughout any of these sets. On the downside, this theory denies that it is a logical law that we can derive a conjunction from its two conjuncts. For we can have a story (say) represent a set that contains two sets of worlds. On one of these sets, A is true in every world and on the other B is true. But we cannot conclude from this that A and B is true in any world. So, this theory rejects what is usually thought to be a central rule of reasoning.43 Our purpose here is not actually to support one side or another in this debate. Rather, it is to do some metaphilosophy and show how someone’s reasoning about this issue could work.

Logic and Metaphysics: Dummett Meets Heidegger

67

One assumption we made is that one would want to reject ex falso. This, however, is a rule and not a statement, when phrased in the proof theory. But, in a metalanguage, it can be phrased as a statement (the metalanguage contains as a relation, for example, ‘entails that’). Suppose our reasoner begins by believing in possible world semantics. Then, her beliefs entail one of her rejections. If she wants to retain that rejection, she can modify her semantical theory to include impossible worlds or she can revise the relationship between rules and the models that she accepts. How is she to choose? There is no other option but to say that the overall result has to justify the choice. It is the overall coherence of the result that determines whether the choice was a good one. Suppose that our reasoner accepts impossible worlds but claims that there are no contradictions true at the actual world. As Graham Priest argues in ‘Motivations for Paraconsistency’ this position is unstable. Once impossible worlds have been put into her ontology, she needs to explain why this world is not among them. The danger that has appeared in her view is what we might call the dialethic peril. Clearly, since the mutual justification (and explanation) of metaphysics and logic is the aim of the reasoning process, the form of justification to be used would naturally be some form of explanatory coherence theory. The rejection of true contradictions needs some explanation. One way of avoiding the dialethic peril is to give the actual world a different metaphysical status from other worlds. This is what so-called ersatz theories of worlds do. They construct worlds from abstract objects and then claim that only the actual world is instantiated. If our reasoner were to take this tack, she could give some metaphysical reasons why only a consistent world can be instantiated. Of course there are other options open for our reasoner. For example, she could become a fictionalist about her semantics and claim that it is not true, but a story that makes decides whether a rule of inference is a true law of logic. Fictionalism about one’s model theory destroys the intermediary’s role of model theory. But if she makes this choice she needs to explain why this story (rather than any other) has this central normative role in deciding how we should reason. On the other hand, she could almost give in to the dialetheist and accept the belief that it is possible that there be a true contradiction (and eliminate the rejection of the statement that there is a true contradiction). This case study, then, shows how one should use the process of finding reflective equilibrium regarding his or her views about logic and metaphysics.

Coda We have seen that the relationship between logic and metaphysics is complicated. As Heidegger says logic is not ‘free floating, something ultimate’.44 Our logical intuitions, as we saw in our discussion of Dummett, may be overridden by our intuitions about metaphysics. But our metaphysical intuitions fare no better. Often our reasons for postulating a metaphysics are logical, as in the case of the

Postanalytic and Metacontinental

68

postulation of other possible (or even impossible) from logical need. Heidegger may be right that ultimately we need metaphysics, or at least some extralogical theory, to justify logical statements and rules, but in our reasoning metaphysics beliefs need not precede logical beliefs. We thus need a way of reasoning about logic, metaphysics, epistemology, semantics and pragmatics all at once. For this form of ‘all-in reasoning’ I have suggested a form of reflective equilibrium.

Notes 1

2

3

4

5

6

7 8

9

I would like to thank Jack Reynolds, James Williams, James Chase, and Frank Jackson for comments on earlier versions of this chapter. This chapter was written with financial assistance from a Discovery Grant from the Australian Research Council. Mark Wrathall thinks that the contrast between analytic and continental philosophy in general can be illustrated by examining the differences between these two texts: Wrathall claims that whereas analytic philosophy attempts to discover metaphysical truths by ‘clarifying the logical structure of language and mind’, continental philosophers reflect on the structure of language and mind to discover the ‘historical, existential dimension of our language and thought.’ See Mark A. Wrathall, ‘The Conditions of Truth in Heidegger and Davidson’, The Monist 82 (1999): 305. See Plato, Parmenides, trans. A. Whitaker (Newburyport, MA: Focus Publishing), 1996, 128D, and William Kneale and Martha Kneale, The Development of Logic (Oxford: Oxford University Press, 1984), §3. Bertrand Russell, The Philosophy of Leibniz (London: George Allen and Unwin, 1958), 170. Actually, it is rather difficult to characterize exactly Dummett’s view of metaphysical disputes. They are really disputes about whether a set of sentences (characterizing some subject matter) are bivalent (see below). This is Dummett’s way of making disputes about, say, the existence of a supersensible realm of mathematical objects intelligible to him. He says ‘to affirm, or deny, the existence of an objective reality described by our statements, and rendering them determinately true or false, is to adopt a picture which accords with one or other conception of the kind of meaning which those statements have, but a picture which has in itself no substance other than as a representation of the given conception of meaning.’ See Michael Dummett, Elements of Intuitionism (Oxford: Oxford University Press, 1977), 382–383. Dummett always approaches the question of the existence of mathematical (or historical, or subatomic, . . .) objects as questions about the determinate truth value of sentences about those objects. Michael Dummett, The Logical Basis of Metaphysics (Cambridge, MA: Harvard University Press, 1991), 13–14. Dummett, The Logical Basis of Metaphysics, 13f. Michael Dummett, Truth and the Past (New York: Columbia University Press, 2004), 45. Michael Dummett, The Logical Basis of Metaphysics, 14.

Logic and Metaphysics: Dummett Meets Heidegger 10 11 12

13

14 15

16

17

18

19

20 21 22 23 24 25

26

27

28

69

Ibid. Dummett, Truth and the Past, 45. Martin Heidegger, The Basic Problems of Phenomenology (Bloomington: Indiana University Press, 1982), 177. It seems that Heidegger thinks that mathematical logic is even worse in this regard: ‘Contemporary logic shows a new distortion of the problem. Not only is metaphysics reduced to logic, but logic itself is reduced to mathematics. Contemporary logic is symbolic, mathematical logic, and thus a logic which follows the mathematical method.’ See Martin Heidegger, The Metaphysical Foundations of Logic (Bloomington: Indiana University Press, 1984), 106. Ibid., 176. Immanuel Kant, Critique of Practical Reason (Indianapolis: Bobbs-Merrill, 1956), part I, chapter 1. ‘[F]reedom and unconditional practical law reciprocally imply each other’ (Kant, Critique of Practical Reason, 50). This is Kant’s famous ‘reciprocity thesis’. For an in-depth analysis of the reciprocity thesis, see Henry E. Allison, Kant’s Theory of Freedom (Cambridge: Cambridge University Press, 1990). For Heidegger’s criticism of the reciprocity thesis, see The Basic Problems of Phenomenology, 135. Martin Heidegger, The Essence of Human Freedom: An Introduction to Philosophy (London: Continuum, 2002), 93. Further, ‘Human freedom is no longer means freedom as a property of man, but man as a possibility of freedom. Human freedom is the freedom that breaks through in man and takes him up unto himself, thus making man possible’. Martin Heidegger, Being and Time, trans. J. Macquarrie and E. Robinson (Oxford: Basil Blackwell, 1927), 34. Heidegger wasn’t the first to claim that Leibniz’s logic is based on his metaphysics, Russell did so in 1900. On Benson Mates’ view, however, any claim about the priority of metaphysics over logic or logic over metaphysics in Leibniz is wrongheaded. He says that it is ‘futile to argue about whether Leibniz’s metaphysics is derived from his logic or his logic from his metaphysics, an issue that seemed of major importance to a number of commentators, including, preeminently, Russell and Couturat. For not to belabor the point, in Leibniz’s philosophy “everything is connected to everything”’. See Benson Mates, The Philosophy of Leibniz: Metaphysics and Language (Oxford: Oxford University Press, 1986), 5. Martin Heidegger, The Metaphysical Foundations of Logic, 37. Ibid., 39. Ibid., 48. Ibid., 70. Ibid., 76. For example, I think that Aristotle’s substance philosophy comes as much out of his syllogistic as the syllogistic is a reflection of his views on substance, Plato’s dialectic is a source for the theory of forms, and so on. Actually, I think that logical validity should be interpreted as information preservation, but this view presupposes a metaphysics of information. W. V. O. Quine, From a Logical Point of View (New York: Harper and Row, 1961), essay 6. Whenever he was asked about these failures he would say ‘so much the worse for modal logic’. He never considered them a problem for his metaphysical views.

70 29

30 31

32 33 34

35

36

37

38 39

40

41

42

43

44

Postanalytic and Metacontinental C. I. Lewis, An Analysis of Knowledge and Valuation (LaSalle, IL: Open Court, 1946). W. V. O. Quine, From a Logical Point of View, essay 2. Clearly, the choice of a metatheory (a logic in which to frame one’s semantics) may determine what one says about the metaphysics. Martin Heidegger, The Metaphysical Foundations of Logic, 103. Unfortunately, I used to call ‘coherence’ what I now call ‘pragmatic consistency’. Edwin Mares, ‘A Paraconsistent Theory of Belief Revision’, Erkenntnis 56 (2002): 229–246. The variation is indeed slight. That view uses a ‘partial meet’ theory of contraction. When contracting, it chooses a set of reasonably maximal subtheories and then takes the intersection of them. On the present view, it is more appropriate just to have a selection function that chooses one reasonably maximal theory – the one that exhibits the most explanatory coherence. Actually it appeals to several notions. A set of beliefs is said to entail a set of rejections if and only if (according to some logic that is chosen) it is a theorem that a conjunction of the beliefs implies a disjunction of rejections (see Mares, ‘A Paraconsistent Theory of Belief Revision’). The logic can be clearly as weak as the basic relevant logic B. I conjecture that we can make the logic even weaker than this by replacing the notions of implication, conjunction and disjunction with structural connectives in a sequent calculus. This will allow us to reconstruct the theory without assuming, say, anything about formulas with nested implications. Per Martin-Löf, Intuitionistic Type Theory (Naples: Bibliopolis, 1984). This would of course contradict Dummett’s claim that his theory of meaning is completely neutral with regard to metaphysics. Given my theory, and the failure of his programme, I do not think that this is such a bad thing. Alfred Tarski ‘The Concept of Truth in Formalized Languages’, in Tarski, Logic, Semantics, Metamathematics (Indianapolis: Hackett, 1984), 152–278. Some philosophers call this a ‘decompositional’ semantics. Those philosophers usually hold some form of Frege’s context principle – that words only have meaning in the context of a sentence. So we need the sentence before we can decompose it to find the meanings of its parts. I am running roughshod over an important distinction here. Not all rules are valid in the sense described. Some rules, such as the rule of necessitation (that from A’s being a theorem we can derive that necessarily A is also a theorem) are valid only in the sense that in any model in which the premise is valid so is the conclusion. Moreover, there are rules that are such that if their premises are all valid in all models, then their conclusion is valid in all models, but are such that individual models need not be closed under these rules (the rule of uniform substitution of formulae for propositional variables is of this kind). But for our purposes, we do not need to discuss such rules. This isn’t entirely fair to Schotch and Jennings, who have a very complicated view of when conjoining is allowed and when it is not. Martin Heidegger, The Metaphysical Foundations of Logic, 103.

Chapter 5

Argument All the Way Down: The Demanding Discipline of Non-Argumento-Centric Modes of Philosophy Simon Glendinning

One may say in such a case that the term doesn’t have one legitimate heir only . . . Consider the question: Why should what we do here be called ‘philosophy’? Why should it be regarded as the only legitimate heir of the different activities which had this name in former times? (Ludwig Wittgenstein)

1. Situations of war Things are still bad round here, sometimes still really bad.1 This is how we tend to see it: a war has been going on in the philosophical culture, and it is not over. Geoffrey Bennington sketches the scene in the following vivid description: As is not unusual in situations of warfare, the sides will often find it to their advantage to adopt strategic means to achieve certain ends – so we might expect propaganda of various sorts to be used at various levels of the conflict, from published polemics and denunciations and even smears, to departmental manoeuvrings and even the way academic appointments are made.2 It is at this level of activity and activism that the idea of a distinction or opposition or division between analytic and continental philosophy is at its most powerful, most disturbing, and most persuasive. And yet I am not persuaded. In my view, the construal of the situation which elaborates itself through the prosecution of this war is less a truthful presentation of something rotten in the state of our philosophical culture than the most significant expression of its rotten condition; less a response to a fault within the culture than the expression of a cultural fault. As I see it, most talk informed by the idea of the analytic/continental division belongs to a dimension of our inhabitation of our philosophical culture that is fundamentally insensitive to the questionable character of its own terms of trade.

72

Postanalytic and Metacontinental

People seem happier to render inaccessible to themselves whatever they are (for some reason) interested in underestimating. And that, I think, is one of the main functions of the idea of the analytic/continental division, especially within analytic philosophy where the idea first developed. It has rationalized a willingness not to read, a willingness not to discover oneself capable of reading well. Of course one can’t read everything, let alone read everything well. And we know too that writings that stake a claim to the title ‘philosophy’ but which do not belong to the mainstream of analytic philosophy will typically be experienced as distinctively difficult to read by people whose studies are centred on that mainstream. However, I want our thinking about this reading difficulty to be more cautious and more refined than one generally finds. Saying simply ‘I can’t make head nor tail of this: it is continental philosophy’ is not even the beginning of an explanation of this difficulty. On the contrary, it is a major symptom of a distinctive and also rather shocking way in which (certain) philosophers have treated a question.3 Needless to say, left untreated such a treatment produces its own reality. And today it is still dominant. Nevertheless, it does not altogether exclude something else making its way. The situation remains open to another, a different, more differentiated, more constructive and better informed response. Bennington’s description of the cultural war belongs to just such an effort, cultivating a construal of the situation beyond the take of warring sides: There are two broad ways of presenting this situation, I think. One way thinks of it as a contest or war: on this view there is logocentrism (or argumentocentrism) on the one side and something else (style, literature, Heidegger, ‘the continent’ . . . ) on the other, and these two sides really have to fight it out. I believe that this has been the dominant way in which these matters have been presented . . . The other way of construing the situation is to cast doubt on the very presentation of it as a conflict . . . This is not to suggest that there are no differences between what are no longer quite two sides, but that . . . the multifarious differences between and within each side are being grossly (violently) simplified when they are presented as always finding their truth in opposition and confrontation.4 In this chapter I attempt to forge a path through and beyond the first construal by developing an argument for a commitment to the second. I will begin with an attempt by Bernard Williams to restyle the first.

2. Le style c’est l’homme même The proposal from Bernard Williams that I want to launch from is to read the distinction between analytic and continental philosophy as a ham-fisted way of responding to differences in the style of philosophical writing. Labelling a work

Non-Argumento-Centric Modes of Philosophy

73

by one or other of these ‘obscure’ and ‘misleading’ titles, and the absurd ‘crossclassification’ that results,5 is usually thought to have a reasonable basis in the distinctive differences between the methods and problematic fields of two philosophical traditions. According to Williams, however, what is really at issue is not primarily a matter of content at all but, rather, of style. Without trying to develop a rigorous taxonomy of philosophical styles, Williams suggests that we can distinguish the style of analytic philosophy from other styles of philosophical writing with reference to the fact that it (analytic philosophy) typically employs ‘moderately plain speech’.6 This plain style is, he insists, still a style – and hence ‘must to some extent determine subject matter’ – but whatever shortcomings or limitations accrue as a result it has the distinctive virtue of achieving the kind of clarity and accuracy possessed by and aimed at in the natural sciences. The concept of ‘style’ is standard fare in the dominant construal. However, Williams’ reference to the style of analytic philosophy serves to remind analytic warmongers that style is not an optional (and best avoided) extra for philosophical writing. What one regards as ‘the responsible way of going on, the most convincing expression of a philosopher’s claim on people’s attention’7 always includes a view concerning what it is for philosophical writing ‘to be well shaped and disciplined’.8 It is, in this sense, unavoidably, a question of style all the way down. On this conception, what is regarded as the responsible way of going on in philosophy is always caught up in a distinctive stylistic choice or decision, in play from the outset, concerning how to make moves within a language that will – by the lights of the philosopher – best cultivate its expressive power with respect to a particular question. In our time, many philosophers, especially (but neither all nor only) analytic philosophers, construe the most responsible way of going on as proceeding in a manner that attempts ‘to make language as unambiguous, clear and as ‘scientific’ as possible’.9 As we shall see, the style aspired to here is closely connected to a desire to proceed though the development of a certain kind of argument. Geoffrey Bennington specifies this idea of argument in terms of ‘something like a sequence of related propositions tending to establish the validity of a position or a thesis of some sort’.10 We should probably stress the ‘something like’ rider in this formulation,11 but Bennington is sure that Jacques Derrida for one really does not proceed in anything like this sort of way, at least not most of the time. In general, Derrida’s texts don’t look at all like the elaboration of an argument in this narrow sense. So what do his texts look like the elaboration of then? ‘Well’, says Bennington, ‘an argument’.12 He continues: ‘Derrida has an argument with argument, if you like, but that argument does not involve simply denying, criticising or even ignoring the claims of argument, but rather loving argument.’13 This is such an under-appreciated point in relation to Derrida (for one) that it is worth underlining (if not automatically multiplying). Distancing himself from a certain reception of deconstruction as a profoundly negative or

74

Postanalytic and Metacontinental

destructive engagement with the philosophical heritage, Derrida relates his approach as follows: I love very much everything that I deconstruct in my own manner; the texts I want to read from a deconstructive point of view are texts I love, with that impulse of identification which is indispensable for reading. They are texts whose future, I think, will not be exhausted for a long time . . . Plato’s signature is not yet finished . . . – nor is Nietzsche’s, nor is St Augustine’s.14 So Derrida does not write in a ‘critical fury’ against the great classic texts of the philosophical heritage, but engages with them out of love for them, attempting to show that these texts still have something yet to say to us, remain ahead of us and still to come. Analytic philosophy is dominated by an inheritance that Bennington aptly calls ‘argumento-centric’. And no doubt cultivating this commitment is also its distinctive act of love, its effort at fidelity to the philosophical heritage. But between such acts of love (and must we count the legitimate ways here?) is there, must there be war? Must the situation be construed in terms of a confrontation between two opposing sides? Two styles? Two fronts? A divided inheritance?15 We are concerned with texts which attempt faithfully to inherit the different activities which had the name ‘philosophy’ in former times by cultivating what is, by the lights of the philosopher writing, the greatest expressive power of language with respect to some question. This certainly is a matter of taking a stand: it is a way of respecting a commitment to what one regards as the most responsible way of going on. However, on the site of an asymmetrical conflict that is skewed ‘rather massively on the “analytic” side’,16 I want to challenge the long-cherished idea that taking responsibility here must involve finding fault in anything but argument in the narrow sense. In what follows, I intend, therefore, to make effective propaganda for a change in our way of thinking.

3. Anything but argument Philosophy (in general) has a constitutive relation to the distinction between legitimate and illegitimate forms of persuasion; between being, say, rationally convinced (the aim of philosophy proper) and being merely persuaded (the upshot of being taken in by, say, sophistry and illusion). Within analytic philosophy today the quasi-official understanding of this distinction is most often cast in terms of the presence or absence in a text of a clearly stated argument.17 Such an argument would present something like a series of valid inferential steps from premises to conclusions. In the absence of such an argument – henceforth referred to as ‘narrow argument’ – what one does cannot be a rationally ‘convincing expression of a philosopher’s claim on people’s attention’. Now, among the complaints one often hears directed against so-called ‘continental philosophy’ is the perceived lack in it of clearly stated narrow arguments.

Non-Argumento-Centric Modes of Philosophy

75

I think it is probably a mistake to suppose that any of the major authors who are usually captured or corralled into the continental collection pursue philosophy utterly without such arguments. However, it is true that there is typically a lot going on in their writings that is not just ‘developing an argument’ in this narrow sense. Nevertheless, rather than seeing this as a problematic dearth of argument, I want to argue that such writings should be conceived, in another sense, as argument all the way down: for they are all written with the aim of convincing other people. On the other hand, yes, it is not only or primarily by way of narrow argument that they proceed. The worry, of course, is that in the absence of clear and explicitly developed narrow arguments their contribution is of limited philosophical interest. Even if their writings are found persuasive they cannot be said to offer the chance of bringing reasonable conviction. Consequently, many reckon such writings philosophically weak or questionable, or worse. Socrates is often cited as a model of philosophical objectivity and integrity, famed for his willingness to ‘follow the argument where it leads’.18 Inheritances of philosophy that do not take the narrowly argumento-centric path are often thought to have departed from that line of ‘the dialogue of reason’ which takes Socrates as its model.19 On the other hand, however, Socrates is equally famed as the gadfly who addressed himself to others, the one who talked philosophically to other people. And, whether or not one explicitly attends to the point in one’s own writing it is clear that who one is addressing makes a huge difference to one’s prospects of convincing by means of narrow argument. As Cora Diamond observes in the following passage, this issue is especially acute in relation to our moral thinking: When we engage in philosophical discussion about such a subject as abortion, or the moral status of animals, whom should we think of ourselves as trying to convince? For if we proceed by giving arguments, we presumably do not expect to be able to convince anyone who is incapable of following our arguments, or who is too prejudiced to consider them. And if we are talking about convincing human beings, surely it is a fact about many of them that one certain way of not convincing them is to try arguing the case . . . No one who urges another philosopher to give arguments thinks of arguments as capable of convincing everybody. When we put forward arguments, or urge someone else to do so, we have a conception of what it would be to succeed in giving genuinely convincing arguments, and also of those who would nevertheless not be convinced, even should they attend to the arguments. Now, argument is simply one way people approach moral questions, and there are other ways of trying to convince someone of one’s view of animals or foetuses or slaves or children or whatever it may be.20 This is not an argument utterly against narrow argument. Diamond has no intention of suggesting any impropriety in the thought that developing such arguments is what ‘all [philosophers] do some or most of the time’.21 Indeed,

76

Postanalytic and Metacontinental

as I have suggested, one is likely to find narrow arguments in all the writings of the major so-called ‘continental philosophers’ too. However, what Diamond wants to encourage her readers to acknowledge is that it is actually quite perverse to think that this is all that any philosophers do, or that this is the only thing that any philosophers do as proper philosophers that leads others to be reasonably convinced of something, as if giving an argument in the narrow sense was the only thing which would represent a genuinely convincing expression of a philosopher’s claim on people’s attention. When he reflects on the construal of that expression that dominates in analytic philosophy today, Bernard Williams is struck by how little attention is given to the fact that the recommended way of going on in philosophy would not actually convince many of those whose attention it claims. The idea is rather that it is being held up as exemplary, and hence is worth shouting about even to people who, in fact, lack the interest or capacity to attend to its procedures themselves. Diamond puts the point as follows: ‘If we think that philosophers should try to work out moral views [in something like this sort of way], it is because we think that people should aim to order their own moral thought in something like this sort of way.’22 Diamond, however, is not convinced that this sort of way of thinking is genuinely exemplary, for she is not convinced that it correctly identifies ‘which human capacities are characteristically exercised in the development of someone’s moral life, and more specifically of what it is for someone to exercise his capacities as a thinking being in that development’.23 The quasi-official understanding of argumento-centric philosophy conceives ratiocinative capacities as of first importance in our moral thinking, and thus thinks of narrow argument as the most (perhaps only) legitimate way of reasonably convincing others. Diamond (and, in fact, Williams himself to some extent) disagrees: this construal massively underestimates the importance of imaginative capacities in our moral thinking. Now, while there may be a certain imaginative poverty to the quasi-official selfunderstanding of argumento-centric philosophy and to the examples used in the development of narrow arguments in moral philosophy, Diamond is not calling for an imaginative supplement to standard discussions. For Diamond, the imagination has a more ubiquitous and fundamental role to play because what it means to be a human being is something that is in itself imaginatively shaped and reshaped by us, it is in itself something we have creatively made something of in our thought and talk.24 On this view, unless we pursue moral philosophy in ways that involve our capacity to ‘bring imagination to bear on observation’ or to ‘recognise that that has been done’,25 the development of narrow arguments will prove ‘in a sense quite useless’:26 what we do in moral philosophy will have foregone the kind of thinking that can actually touch us or turn us round. So getting our imaginative capacities in play in moral thinking is not simply of supplementary assistance to achieving some kind of purely rational clarity about the evaluation of actions, or the solution of practical problems but is the sine qua non of bringing about the kind of lucid ‘seeing’ that goes with an alteration in a person’s moral view: of bringing a person to see something they had (by one’s own lights) hitherto failed

Non-Argumento-Centric Modes of Philosophy

77

to see. It should come as no surprise then that for Diamond the best model for such an effort is not narrow argument but imaginative literature. On this view, what is at issue for a moral philosopher is not especially a concern with the evaluation of actions or the solution of practical problems, but is a matter of addressing others with the aim of bringing about changes to what Iris Murdoch calling ‘the texture’ of someone’s being: that ‘vision of life’ which shows up in all someone’s ‘reactions and conversation’ – and not just in their ‘ethical statements’.27 For example, of making them into more sensitive and more refined readers of what legitimately goes on in philosophy that is not reducible to science-inspired plain-speaking and narrow argument.

4. A turn in our time The account of moral development outlined here is clearly controversial, and I cannot hope to do much more than introduce it at this point. However, there are some objections which should be dealt with. First, there is the question of what is to constitute an exercise of imagination. In a thought that is related to Diamond’s resistance to construals that restrict the concerns of the moral philosopher to one (supposedly isolable) domain of our lives, Kant (and then Strawson via Wittgenstein) argues that something like imagination is drawn on everywhere in experience; simply in virtue of the fact that judging any particular to be a such-and-such brings the present experience into a relation with nonpresent, non-actual perceptions.28 But Strawson (and Kant) acknowledges this to be a somewhat technical sense of imagination, and what is at issue in the present discussion is a more hum-drum sense: imagination as the capacity to engage in a creative elaboration of an idea, thought, or vision of a thing. This is more hum-drum but may still be deep-rooted and wide-ranging in our lives. For example, according to Diamond there is a work of imagination in the historical and cultural development of the idea of human beings as, for example, ‘rational animals’, it belongs, that is, to an imaginative (and not merely factual or scientific) elaboration of an idea of the difference (not the factual differences) between ourselves and other animals.29 To explore and test this claim concerning the general importance of the imagination in human life a little further, I want briefly to consider a plausible counter-example to Diamond’s view about changes in moral outlook: a case in which it might be claimed that it really was narrow argument which played the crucial role in some stretch of human moral development. The case concerns the campaign for female suffrage in America and Europe. We know that for thousands of years in post-Neolithic human culture women have been regarded by the dominant and dominating culture as ‘the second sex’ (in Simone de Beauvoir’s striking expression). It has also been suggested that in early Neolithic times human beings lived in a rather more matriarchal culture. Now, it is natural to think that the kind of change-over in human culture at issue here must be accompanied by a major alteration in communal psychology, and in

78

Postanalytic and Metacontinental

this case a major alteration in the cultural significance attached to the difference (not the factual differences) between men and women. But what might drive such a change? I am not about to speculate on what took place in Neolithic times, but at least as far as modern humanity is concerned, an explanation for a change in opinions on voting rights can certainly be offered in the terms favoured by a defender of the primacy of narrow argument; an explanation in terms of the arguments that provided the impulse to back proposals for change. An argument of the favoured sort might be rendered as follows: If at least some women are smarter and more public spirited than some men, there is no decent justification for giving men but not women the right to vote. At least some women are smarter and more public spirited than some men. Thus, there is no decent justification for giving men but not women the right to vote.30 The thought here is that narrow arguments of this sort are, in virtue of their plain unanswerability, capable of ‘turning people round’ and changing minds. However, in defence of the emphasis on the role of imagination, it seems clear that if this sort of narrow argument really was effective in leading to female suffrage (and we shall see shortly that this particular one was not), then that was likely to be because it was advanced at a time when there was already a shifting cultural perception of the difference between men and women. A hundred years ago narrow arguments for female suffrage began to present themselves as convincing in America and Europe; a hundred years earlier than that hardly at all. But it is not as if the narrow arguments were defective earlier. So while it seems right to say that an appreciation of narrow arguments for suffrage does not involve much imagination (which is my point too of course) it is still wrong to think that they were effective independently of some considerable work of the imagination. The cultural conception of the difference between men and women was already mutating and narrow arguments could begin to make a purchase in this changing space of what it does and what does not make sense to say about women. In particular, the change is one in which it is no longer so matter of course to regard engagement in public affairs as utterly unfeminine or to suppose that the proper ‘sphere of women’ is not politics or publicness, but domestic and private life. The argument against the defender of narrow argument then is this. It might be thought that narrow arguments for universal suffrage lead to widespread moral improvement in Western societies about one hundred years ago. However, it looks as if the crucial improvement was already underway in American and European culture, and that this centrally involved a new imaginative elaboration of what it means to be a woman. Moreover, as Harriet Taylor Mill noted, most of the really effective narrow arguments were formulated in terms which focused on what is right for human beings in general, not on what is right for men or right for women. These arguments were put forward in a way which, as she says,

Non-Argumento-Centric Modes of Philosophy

79

‘refused to entertain the question of the particular aptitudes either of women or men’.31 In other words, the proponents of universal suffrage strategically avoided engaging in narrow argument on just the issue which mattered most to their opponents: the difference between men and women. So, strategically speaking, arguments like the one spelled out above were not the ones worth making. On the other hand, insofar as they are conceived as aiming to speak to our ratiocinative capacities, even those which were worth making were, to go back to Diamond’s phrase, ‘in a sense quite useless’. One of the great virtues of a classic deductive argument is the way it can be used to make perspicuous the internal coherence of a conception, and this might be a very useful contribution. What lacks coherence is the idea that one could create the impetus for change exclusively from a raft of arguments narrowly conceived.32

5. The demanding discipline The argument is that the excision of imagination from philosophical activity is not just an annoying shortcoming in a moral philosopher, but a radical ‘closing off’ of oneself to oneself as a thinking, feeling being, a blindness to one’s own appreciation of and imaginative involvement in what it is to be human.33 Conversely, careful and critical attention to the sometimes simply astonishingly disciplined use of language that one finds in imaginative literature need no longer be regarded as a second best means of developing one’s vision of life, or as a problematically questionable way of bringing about the kind of gestalt shift which might transform the texture of one’s being: it does not fall short of a gold-standard of ‘convincing expression’ that can be got from science-inspired plain speech and narrow argument. Indeed, it is not obvious that a change in one’s moral vision might actually be brought about at all in this area unless our imaginative capacities are fairly intensely brought to bear in our thinking. To suppose that philosophical writing about moral matters is properly disciplined only when it occupies itself with narrow argument is first of all an implicit commitment to a certain conception of the proper formation and moral development of human beings. Indeed, it belongs profoundly to what one might call the modern imagination’s development or elaboration of what it means to be a human being, although it does not recognize itself as such.34 And, more to the point here, it is one that does not recognize that one can be reasonably convinced by something other than narrow argument; for example, by the work of words one can find in a novel, say, or a poem, or some other kind of in its way (judged to be) convincing appeal to our attention. None of this is an invitation to abandon narrow argument. But the idea that something that is judged to be convincing must, if one is to be reasonably convinced by it, be capable of being presented in a narrow argument seems, well, hopelessly unreasonable. This is perhaps obviously the case when moral matters are at stake. Does it need arguing that our moral sensibility can be given

80

Postanalytic and Metacontinental

an unparalleled working over by works of imaginative literature? However, if we accept that the exercise of imagination that is involved in our moral life is involved in shaping our life simpliciter,35 there is an emerging plausibility to the thought of extending something of the claim more widely in philosophy: to see that what is often regarded as a problematic meagreness of narrow argument in a work of (not-specifically-or-exclusively-moral) philosophy may, in fact, still be the most philosophically convincing expression one could wish for. The aim in philosophical writing is to find ways of making moves within language which, by the lights of the philosopher, best cultivate its expressive power with respect to a question. Adherence to the plain style is one way in which this is attempted. For its adherents it offers a chance of eliminating unnecessary ambiguities and unclarities that cloud our everyday understanding; a way of making what we are saying as clear and precise as possible and so to avoid confusions or superstitions or prejudices or distortions. It aims to cultivate a style of philosophical language that attains a status that is as ‘scientific’ as possible. For others, however, this ambition simply expresses distinctively modern prejudices concerning what it is for philosophical writing ‘to be well shaped and disciplined’.36 Of course, bringing it about that others see this too is also part of the philosopher’s task. What, in this case, will be the most convincing expression one could wish for? Can the kind of lucid ‘seeing’ that is required to release us from the felt need for narrow arguments be achieved only or even primarily by means of narrow arguments? One of the best lessons of non-argumento-centric modes of going on in philosophy may be that the intellectual motivations of modernity one must combat here run too deep for that. Or better: these motivations are not simply intellectual but belong to the distinctively imaginative elaboration of the life of human beings living in communities with a history that weaves the fabric of modern times. Hence, just as it is not clear that, when our moral convictions are at issue, narrow argument alone can be anything but ‘in a sense quite useless’, so also it is not clear that narrow argument alone can hope to answer to what will provide the in its way (judged to be) convincing appeal to our attention sought by writings of philosophy that take seriously the threat posed to thinking in a time when ‘science [has] become our passion’.37 Narrow argument cannot be the exclusive mode of that inheritance of philosophy which aims to cultivate an understanding of ourselves and the world denied to us by (judged to be) inhospitable modern standpoints. How non-argumento-centric modes of philosophy attempt to loosen the grip of modern science on our thinking in philosophy resides in the particular and sometimes strikingly novel ways that words are put to work there. And for that reason we should not expect that a talent for writing such philosophy will belong to everyone equally, or that all who attempt it will respond to questions concerning ‘the proper place and manner of its own commencement’38 in the same way.39 Nevertheless, it seems to me that it cannot take place at all without an awareness of the need to get rid of distorting presuppositions and assumptions not simply by narrow argument – and not by writing a novel or a poem either – but by way of a composed work of words which offers some other

Non-Argumento-Centric Modes of Philosophy

81

kind of in its way (aiming to be) convincing appeal to people’s attention; through writing whose demanding discipline resides in its capacity to enable us to retrace our steps, to turn people round so that they can see clearly what (by the philosopher’s lights) in modern times we find it hard to see. As we saw at the start, it is not news that writings outside mainstream analytic philosophy typically pose a distinctive problem for readers whose studies are centred on that mainstream: the texts are distinctively demanding and difficult to read. It should now be clearer why this is so, and why the treatment of the question of this reading difficulty that consists in consigning it to (already understood) ‘continental’ obscurity closes one off to investigations in which the way the philosopher’s words are chosen, ordered and fitted together is an integral part of its real business, not something supplementary to its work of argument and demonstration. With such texts, coming to terms with a philosopher’s treatment of a question and coming to terms with the text in which the question is treated are not even notionally separable tasks; even punctuation matters. On the other hand, where this idea is taken to heart we find something like the affirmation, explicit in Wittgenstein, that writing philosophy should be as ‘nur dichten’ (as a poetic composition). This is not to suppose that the way philosophy should be written is as ‘nur Gedicht’ (as a poem).40 If someone finds this distinctively dense (dicht) way of writing irreplaceably fitting in philosophy that does not mean that they are really writing poems, and by the same token it is not as if a proper concern with a philosophical text that is composed in this way demands that one should pay special (still less exclusive) attention to the images and metaphors that it makes use of. On the other hand, it does call for readers who are prepared not to assume that they know what it means to be an appropriately sensitive reader of every text that might, with justice, be called philosophical. When you come to see that the writing of philosophy should be as ‘nur dichten’, your own reading will be marked by a distinctive openness to composed works of words which, in an age dominated by science, see their own task as one of cultivating and developing your capacity to elaborate for yourself a radically re-vis(ion)ed understanding of yourself and your place in the world and with others. As a matter of fact, such an understanding of the task of philosophical writing does not belong to just any old inheritance of philosophy, but distinctively to texts that have presented themselves during the last one hundred years or so in the name of phenomenology. (This would include the work of Ludwig Wittgenstein,41 J. L Austin42 and Gilbert Ryle.43) Here more than anywhere, the ambition has been to treat questions in such a way that, in the face of the modern obsession with scientific method and scientific conceptions of ourselves and the world, we can be freed to ‘see’ what is right before our eyes.44 So is phenomenology the right way, the most responsible way, to go on with philosophy? There is and can be no radical justification for phenomenology that is not already phenomenological in character. Merleau-Ponty says: phenomenology ‘rests on itself’.45 Austin says: ‘there is gold in them thar hills.’46 One can try to say more but, really, one should simply ‘cut the cackle’, as Austin

Postanalytic and Metacontinental

82

insists, because the phenomenological work of words is already, and in itself, argument all the way down.

Notes 1

2

3

4 5

6

7 8

9

10 11

12 13 14

15

This text continues the assessment of the contemporary philosophical culture developed in Simon Glendinning, The Idea of Continental Philosophy: A Philosophical Chronicle (Edinburgh: Edinburgh University Press, 2006). Geoffrey Bennington, ‘For the Sake of Argument (Up to a Point)’, in Arguing with Derrida, ed. Simon Glendinning (Oxford: Blackwell, 2001), 40. This conception of the situation is informed by Stephen Mulhall’s reading, in Wittgenstein’s Private Language (Oxford: Oxford University Press, 2007, chap. 6), of the ambiguities in Wittgenstein’s famous (but hitherto questionably translated) aphorism, with its enigmatic semi-colon, ‘The philosopher treats a question; like an illness’ (Wittgenstein, 1958, §255, Mulhall’s translation). Bennington, ‘For the Sake of Argument (Up to a Point)’, 40–42. As Bernard Williams suggests, the cross-classification at issue is ‘rather as though one divided cars into front-wheel drive and Japanese’. See ‘Contemporary Philosophy: A Second Look’, in The Blackwell Companion to Philosophy, ed. N. Bunnin and E. P. Tsui-James (Oxford: Blackwell, 1996), 25. Bernard Williams, Ethics and the Limits of Philosophy (Cambridge: Cambridge University Press, 1985), ‘Preface’. Williams, ‘Contemporary Philosophy’, 27. Stephen Mulhall, Inheritance and Originality (Oxford: Oxford University Press, 2001), 1. Herman Philipse, Heidegger’s Philosophy of Being (Princeton: Princeton University Press, 1998), 230. Bennington, ‘For the Sake of Argument (Up to a Point)’, 39. Some analytic philosophers collect unusual argument forms as other people might collect unusual smoker’s pipes or egg-cups. And there is wide interest in, for example, inference to the best explanation, Bayesian reasoning, probabilistic argument and so on. So it would be a mistake to claim that analytic philosophers restrict argument to deductive forms. However, Bennington’s ‘something like’ rider allows for this variation, and the basic claim need only be that, for analytic philosophers, deductive argument has a normative centrality however much their arguments might also roam beyond this, and indeed might roam willy-nilly into what I will call the demanding discipline that it might officially prefer to avoid altogether. I am grateful to James Chase for drawing my attention to these points. Bennington, ‘For the Sake of Argument (Up to a Point)’, 46. Ibid., 47. Jacques Derrida, The Ear of the Other, trans. by Peggy Kamuf (Lincoln: University of Nebraska Press, 1988), 87. Within this scene one might see the work of an ambiguity inside the very name ‘philosophy’, which might refer us (in Greek style) to the love of wisdom or (in Judaic style) to the wisdom of love. The mistake would be to suppose that one style was finally or analytically separable from the other. As Molly Bloom famously

Non-Argumento-Centric Modes of Philosophy

16 17

18

19

20

21 22 23 24

25 26 27 28

29 30

31 32

83

says in her soliloquy in Joyce’s Ulysses ‘Jewgreek is Greekjew’, an idea that is cited and discussed in Jacques Derrida, Writing and Difference, trans. Alan Bass (Chicago: University of Chicago Press, 1981), 152. I am grateful to Jack Reynolds for reminding me of this ambiguity within every inheritance of the name of philosophy. Bennington, ‘For the Sake of Argument (Up to a Point)’, 41. By quasi-official I mean the understanding one typically finds in student guidelines for writing a philosophy essay, and which, as indicated in note 10 above, has a normative centrality for the analytic self-conception. An excellent example of this sort of thing is James Pryor’s online resource for students at http://www. jimpryor.net/teaching/guidelines/writing.html. As I have indicated, the philosophically significant question is whether this conception ever exhausts what is, willy-nilly, going on in philosophical texts written under its aegis, and hence how far such writings are always already not (in this sense) properly philosophical through and through, or indeed how far anything aiming at bringing reasonable conviction could ever be. On this theme see Darren Sheppard, ‘The Reading Affair: On Why Philosophy Is Not “Philosophical” Through and Through’, in Arguing with Derrida, ed. Simon Glendinning, 121–134 (Oxford: Blackwell, 2001). Geoffrey Bennington’s essay in the same volume, referred to already, raises similar questions, noting various argumentatively pivotal moments in texts elaborated in the name of philosophy which are, by their own lights, ‘perhaps not very philosophical’ (cf. 34). He undertakes a detailed case study of this tendency in Geoffrey Bennington, Other Analyses: Reading Philosophy. CreateSpace, 2008. John Cottingham cites this central feature of Socrates’ way of thinking, and helpfully stresses, as many forget to do, that this is not incompatible with his also avoiding a certain kind of conclusion or conclusiveness. In response Geoffrey Bennington wondered whether Socrates is really following the argument wherever it went or only ‘where he wants [it] to go’ – see Simon Glendinning, ed. Arguing With Derrida (Oxford: Blackwell, 2001), 52. This argumento-centric conception of philosophy is to the fore in J. L. Cohen’s The Dialogue of Reason: an analysis of analytic philosophy (Oxford: Clarendon: 1986). Diamond, Cora. The Realistic Spirit: Wittgenstein, Philosophy, and the Mind (Cambridge: The MIT Press, 1996), 292. Ibid., 293. Ibid., 305. Ibid. Cora Diamond, ‘The Importance of Being Human’, in Human Beings, ed. D. Cockburn (Cambridge: Cambridge University Press, 1991), 47–48. Diamond, The Realistic Spirit, 305. Ibid., 306. Ibid., 374. Peter Strawson, ‘Imagination and Perception’. Kant on Pure Reason, ed. R. C. S. Walker (Oxford: Oxford University Press, 1982), 90–104. Diamond, ‘The Importance of Being Human’, 39. I am grateful to the generosity of Brad Hooker for engaging with my view and suggesting this interesting counter-example. Harriet Taylor Mill, The Enfranchisement of Women (London: Virago Press, 1983), 14. I am not suggesting that such an impetus comes from outside thinking but that the kind of thinking that can be an impetus for change, for example the careful

84

33 34

35 36 37

38 39

40

41

42

43

44

45

46

Postanalytic and Metacontinental consideration of reasons, should not be conceived as the actualisation of exclusively ratiocinative capacities. I am grateful to James Chase for pressing this issue on me. Diamond, ‘The Importance of Being Human’, 45, fn. 18. The modern outlook at issue here can be defined as one in the grip of the thought that an intellectually satisfactory conception of ourselves should be as scientific as possible, and should be drawn in terms that belong to a scientific depiction of nature. Philosophy in this mode is thus the kind of thinking which looks for its inspiration in the style, method and achievements of the natural sciences. Diamond, The Realistic Spirit, 312. Mulhall, Inheritance and Originality, 1. Martin Heidegger, Basic Writings, ed. D. D. Farrell Krell (London: Routledge, 1993), 94. Mulhall, Inheritance and Originality, 2. Regarding the distribution of talent here as uneven is, in a certain way, to depart from the classical ideal in which philosophical competence belongs to us all in virtue of the fact that we are all rational creatures. Argumento-centric philosophy thus appears as normative for all, even if only convincing for a small few. See Diamond, The Realistic Spirit, 26. Wittgenstein’s remark is ‘Philosophie dürfte man eigentlich nur dichten’, which Winch (I think fairly) rendered as ‘philosophy ought really to be written only as a poetic composition’. See Ludwig Wittgenstein, Culture and Value, trans. Peter Winch (Chicago: University of Chicago Press, 1984), 24. ‘You could say of my work that it is “phenomenology”‘, (Wittgenstein, Culture and Value, 116). ‘I think it might be better to use, for this way of doing philosophy, some less misleading name than those given above – for instance ‘linguistic phenomenology’, only that is rather a mouthful’ – J. L. Austin, Philosophical Papers (Oxford: Oxford University Press, 1961), 182. ‘Though it is entitled The Concept of Mind, it is actually an examination of multifarious specific mental concepts . . . The book could be described as a sustained essay in phenomenology, if you are at home with that label.’ See Gilbert Ryle, Collected Papers (London: Hutchinson, 1971), 188. This conception of writings pursued in the name of phenomenology is the subject of Simon Glendinning, In the Name of Phenomenology. London: Routledge, 2007. Some of this essay develops material from a section on phenomenological arguments in the opening chapter. Maurice Merleau-Ponty, Phenomenology of Perception, trans. C. Smith (London: Routledge, 1962), xx. Austin, Philosophical Papers, 181.

Chapter 6

Analytic Philosophy and Dialogic Conservatism James Chase1

1. Introduction Those in the continental philosophical tradition have occasionally criticized analytic philosophy on the grounds of its alleged political or institutional conservatism. The charge sheet here is long. Horkheimer notoriously held that logical positivism was in some sense dependent on the ‘existence of totalitarian states’, and the positivist ascendancy in post-war America has been unsympathetically tagged a form of intellectual McCarthyism. Ordinary language philosophy and the linguistic turn have been seen, particularly in ethics and political philosophy, as involving a retreat from engagement and issues of significance in favour of wordplay. Analytic philosophers don’t in general regard the critique of modernity as core business. The analytic is seen as aping Kuhnian ‘normal science’ and so incurring all the ills of professionalism, over-specialization, and technocracy, while at the same time exhibiting a cringe-worthy obeisance to science itself. And so on. Much of this comes across as exaggerated, unconvincing, selective or ignorant to analytic ears.2 But there are several residual issues that an analytic philosopher would be wise to concede. Many of these are the kinds of institutional issues faced by most academic disciplines with much history behind them – recurrent sexism, inertial resistance to new ideas, and so forth. But at least one of them is peculiarly philosophical. The enterprise of analytic philosophy, both as currently practised and for much of its history, exhibits in many ways a distinct epistemic conservatism, as I shall argue below. True, this is directly a matter of the epistemic norms presupposed by the methods of analytic philosophy, rather than any of the socio-political issues brought up in the continental tradition, but it’s not hard to see how this theoretical conservatism in turn furthers some of the features of institutional conservatism (most notably inertia). Indeed, it seems to me that epistemic conservatism is also one of the points at issue in the debate between traditional analytic philosophy and those who wish to further its engagement with the sciences, such as the experimental philosophy movement. But, quite apart from these connections to institutional or political forms of conservatism, the ways in which analytic philosophy exhibits epistemic

86

Postanalytic and Metacontinental

conservatism are of interest in themselves. Moreover, they stand in need of something of a justification, as many of the standard motivations for epistemic conservatism don’t easily apply to this case (or so I shall argue below). In section 2, I argue that there are many ways in which the practices of analytic philosophy can exhibit epistemic conservatism. Broadly, conservatism is inherent in some analytic credos about the nature of philosophy or epistemology, and also in the many analytic practices involving some relationship to common sense, local coherence-building or appeal to intuition. In section 3, I then argue this situation is not completely justifiable by the standard motivations for conservatism. Rather, some instances of conservatism are induced by the nature of contemporary analytic philosophy itself as a dialogic enterprise, and I present an argument to this effect. The argument does not cover the field entirely, in that some particular analytic commitments to epistemic conservatism have other sources. However, it does make clear the extent to which structural factors might be playing a role in analytic philosophical practice.

2. Analytic conservatism In an influential paper, Richard Foley characterizes epistemic conservatism (henceforth ‘conservatism’) as the thesis that ‘a proposition acquires a favorable epistemic status for a person simply by being believed by him.’3 Since the 1970s, analytic epistemologists have considered the plausibility or implausibility of claims of this kind, both with respect to the beliefs of individual agents and with respect to the commitments of epistemic institutions, and in fact many analytic epistemologists have explicitly endorsed such claims. The relevant conservative claims can be sorted into three lineages:4 i. Generator conservatism: the thesis that belief itself has epistemic weight, whether or not that epistemic weight is ultimately enough to justify its continuance. Roderick Chisholm is perhaps the most well-known proponent of generator conservatism, and although the fact that his foundationalist commitments entailed such a thesis was less clear in his early work, Chisholm’s late work contains such explicitly conservative claims as the following: Anything we find ourselves believing may be said to have some presumption in its favour – provided it is not explicitly contradicted by the set of other things we believe.5 ii. Perseverance conservatism: the thesis that belief revision requires positive grounds. That this claim has to be assessed quite distinctly from generator conservatism was first made clear by Gilbert Harman. His preferred formulation is as follows: One should stop believing p only when one positively believes one’s reasons for believing p are no good.6

Analytic Philosophy and Dialogic Conservatism

87

iii. Comparative conservatism: the thesis that choice among theories with similar epistemic status should be governed in some way by closeness to current belief. Quinean epistemology involves a conservatism principle of this kind, although (as in Chisholm’s case) the claim was latent rather than explicit in Quine’s early work. Most of the analytic literature on this form of conservatism has instead focused on a related claim advanced by Lawrence Sklar in one of the earliest explicit discussions of conservatism in the analytic literature. In Sklar’s formulation, the thesis is as follows: If you believe some proposition, on the basis of whatever positive warrant may accrue to it from the evidence, a priori plausibility, and so forth, it is unreasonable to cease to believe the proposition to be true merely because of the existence of, or knowledge of the existence of, alternative incompatible hypotheses whose positive warrant is no greater than that of the proposition already believed.7 There are many variants on each of these varieties of conservatism that one might endorse, involving different epistemic properties, psychological states (instead of belief, one might consider acceptance, judgment, hypothesis, or some other ‘fit to world’ representational state) or deontic status (is the principle obligatory, defeasible or permissive?). Some of these variations have been explored in the literature, as part of the general assessment of conservatism claims, and a number of other analytic epistemologists have, in setting out a general epistemology, made explicitly conservative claims of these kinds. Presumably in all of these cases, as in the cases of Chisholm, Harman and Sklar, something like an entirely general commitment to conservatism is being made. Of course, most analytic philosophers do not present explicitly conservative epistemologies (or any other kind of epistemology), and many of the participants in the literature on conservatism within analytic epistemology are of course concerned to argue against such general conservatism principles. But such explicit statements of conservatism articulate principles that, it turns out, many analytic philosophers implicitly presuppose in their daily practice of philosophy. A number of common analytic philosophical practices are only justifiable if some kind of conservatism principle obtains locally. Moreover, some ways in which analytic philosophers explicitly conceive of the overall goal or nature of philosophy are markedly conservative, and these conceptions of philosophy remain very influential within the analytic tradition. I’ll briefly consider these points in turn.

2.1 Conservatism of technique Contemporary analytic philosophy makes use of a range of consciously adopted and studied techniques of argumentation or analysis. Many of these presuppose some prior confidence in the worth or applicability of the technique, but prior

88

Postanalytic and Metacontinental

confidence need not involve conservatism in the sense outlined above. For instance, setting arguments out in premise and conclusion form (and bringing formal logical systems, probability theory or the like to bear on them) is not a practice that presupposes epistemic conservatism. It simply relies on the particular evidence one might have as to the suitability of this way of going about things. For a technique to presuppose conservatism it must only make sense if the prior beliefs of the subject on the topic are to be accorded some kind of epistemic weight merely in virtue of their being believed. There are three ways that this obtains in the analytic literature: in techniques involving common sense as a form of input, in local coherence-building devices, and in appeals to intuition. i. Common sense Pre-reflective opinions, or elements of them, are often taken to have epistemic weight independent of our being able to articulate reasons in favour of them – occasionally ‘inarticulate’ common-sense is given the power to defeat philosophical argument. Consider William Lycan’s reading of G. E. Moore’s two-hands refutation of scepticism.8 Moore, according to Lycan, is engaging in a method of differential certainty, and a perfectly rational one at that, since one philosopher’s ponens is another’s tollens. Any argument from premises P1, . . ., Pn to conclusion C, if valid, leaves us comparing the relative plausibility of the collection (P1, . . ., Pn) and the negation ~C. If it turns out that the former are relatively more plausible than the latter, we treat the argument as evidence for the conclusion; if the latter is more relatively plausible than the former, we treat the implausibility of the conclusion as evidence for a defect in the premise set. What makes Moore’s invocation of this familiar point distinctive is that the judgment of relative plausibility is distinctly tilted towards common-sense belief of a certain kind (a category that for Lycan includes not just common-sense claims about the existence of hands, but also knowledge claims about such matters).9 Lycan accepts that common sense is not irrefutable, but its status in possession is not trivial. It can be overturned by evidence (‘careful empirical investigation and scientific theorizing’), but: No purely philosophical premise can ever (legitimately) have as strong a claim to our allegiance as can a humble common-sense proposition such as Moore’s autobiographical ones. Science can correct common sense; metaphysics and philosophical ‘intuition’ can only throw spitballs.10 In the jargon introduced earlier, Lycan’s Moore accepts a fairly strong kind of perseverance conservatism with respect to common sense. If Lycan is right, Moore sets up a conservative background against which a philosophical theory is to be toppled by common opinion. But it doesn’t always have to be so convenient. Analytic philosophers who defend conclusions running in the face of common sense nonetheless can acknowledge a kind of

Analytic Philosophy and Dialogic Conservatism

89

conservatism (which, on this occasion, is being trumped by other considerations). In defending modal realism against all comers, David Lewis unsurprisingly faces a challenge from common sense, and his response is whole-heartedly conservative: Common sense has no absolute authority in philosophy. It’s not that the folk know in their blood what the highfalutin’ philosophers may forget. And it’s not that common sense speaks with the voice of some infallible faculty of ‘intuition’. It’s just that theoretical conservatism is the only sensible policy for theorists of limited powers, who are duly modest about what they could accomplish after a fresh start. Part of this conservatism is reluctance to accept theories that fly in the face of common sense, but it’s a matter of balance and judgement . . . The proper test, I suggest, is a simple maxim of honesty: never put forward a philosophical theory that you yourself cannot believe in your least philosophical and most commonsensical moments.11 Lewis’s general view here underwrites a further philosophical technique. Many analytic philosophers engage in regimentation practices, where common sense is taken in as a kind of input, and the philosophical job involves revising it as needed on grounds of overall coherence. Such regimentation can involve potential reduction or elimination, so to speak (what is sometimes called ‘different levels analysis’). It is not always possible to revise in this way without discarding an essential feature of the subject matter, in which case some kind of eliminativism is likely (for instance, Mackie’s error theory,12 Churchland’s eliminativism about propositional attitudes13); more usually the end result is to accept some form of reductive realism about the domain (for instance, ethical naturalism, or empirical functionalism about the mind); there are other alternatives here as well. Such methods can be local to a particular subject matter, or they can be generalized across the board as a philosophical program (such as the Canberra plan approach to conceptual analysis14) or a generally available technique (such as the method of Ramsification, a device for making network definitions manifest). Alternatively, the regimentation process can be a progressive refinement of the common-sense materials on hand, as in the way that reflective equilibrium is actually put into practice in much moral epistemology, or as in the approach to philosophy that Roderick Chisholm tags ‘particularism’.15 In all of these cases, the point behind regimentation is conservative – that the common-sense starting place for reflection is not at all worthless or trivial, even if the result of the process is potentially rejection of the claims of common sense. ii. Local coherence-building Coherence-building methods are legion in analytic philosophy. Consider argument within analytic philosophy by inference to the best explanation, an argument form in which we accept that claim which serves as the best explanation of the relevant evidence among those explanations we can generate.16 A well-known

90

Postanalytic and Metacontinental

example is Jerry Fodor’s original case for the language of thought hypothesis.17 Again, coherence-building lies behind Quinean appeals to theoretical virtue; preferring, ceteris paribus, those theories that are close to current belief, modest, simple, general and/or refutable. This is Quine’s own list of virtues,18 and much of his work shows him navigating by these lights (for instance, the recurrent wielding of Occam’s razor in Word and Object19). (Other philosophers influenced by Quine may emphasize other virtues, such as fruitfulness, but these are additional, rather than competitors to Quine’s list.20) As a third example, the practice of narrow reflective equilibrium involves working back and forth between a set of norms and a set of judgements about particular instances of application of those norms, until coherence is achieved. This is near-ubiquitous in analytic ethics; for instance, it’s the method at work in Judith Thomson’s violinist argument.21 (I might also mention wide reflective equilibrium, in which the coherence building extends out to one’s general beliefs, but that seems less common in practice.22) Each of these is a method that recommends theory choice be weighted by closeness to current opinion, whatever that turns out to be (and it need not be common-sense – it can be highly particular, reflective, opinion). Fodor recommends the language of thought hypothesis in part because it fits in better with contemporary (for the time) theories of concept learning; Quine recommends the use of classical logic, say, in part because it allows us to retain more of our current beliefs about inference; Thomson’s case is in effect that a particular account of the rightness of abortion best allows us to hang on to those particular judgments we are disposed to make about comparable cases. Each is a type of comparative conservatism. iii. Appeals to intuition Finally, and most prominently, many analytic philosophers make explicit or implicit appeal to intuition, a trend that has burgeoned since the 1970s. An intuition, in this sense, is to be thought of as a fallible intellectual seeming that is (apparently) effortless to produce, immediately given (upon understanding of the relevant situation), reasonably stable from time to time, and very often able to be given a modal inflection (we can intuit that p is possible, or that it is necessary, or that it is impossible, etc).23 Quite apart from the way intuition is directly appealed to as backing for the premises of particular arguments, appeal to ‘grammatical’ intuition is the basic engine of ordinary language philosophy, intuition elicitation is the job of thought experiments (‘intuition pumps’, as Dennett calls them), and the particular judgments that reflective equilibrium requires at each stage are intended to be in some sense deliverances of intuition. Now, there is certainly an analytic tradition that conceives of intuition as an independent, primary source of evidence, more or less akin to perception. This can be understood as some kind of rationalism (as George Bealer argues), with intuitions grounding a priori knowledge,24 or the intuitions in question might be ‘conscious, spontaneous judgments’ (where the instantaneous judging process somehow brings into play further evidence of some kind that might

Analytic Philosophy and Dialogic Conservatism

91

not be easy to investigate otherwise).25 If this is right, there is nothing especially conservative about intuition. But the analytic tradition also includes many who do not regard philosophical intuition in this way, while still making use of intuition-dependent methods such as thought experimentation. On this view, intuitions are simple beliefs, ordinary counterfactual judgments or dispositions to believe, given a light paint job or elicited by an unusual pathway.26 I find this view plausible myself. But if anything like this account of intuition is right, the use of intuition as a starting place in analytic techniques looks to be a further type of conservatism, since prior beliefs, counterfactual judgments, and dispositions to believe are being given epistemic weight to some extent simply because they are there. Surprisingly, the intuition-conservatism link in contemporary analytic practice is not much discussed by those who appeal to intuition.27 Richard Fumerton has recently flagged the connection, and expresses the same puzzlement about the situation: . . . a great many philosophers who would, I suspect, distance themselves from epistemic conservatism, do talk a great deal about the importance of relying on intuition. Or if they don’t talk about its importance, they are not shy about appealing constantly to intuitions in the course of presenting arguments . . . These appeals to intuition are often frustratingly unaccompanied by any account of what intuition is supposed to be. At least sometimes, however, one suspects that the appeal to intuition is just appeal to how things seem, and if intuition does degenerate into such a fallible intentional state, the intuitionism in question can look again very much like a version of epistemic conservatism.28 More is made of the connection by explicit conservatives; conservatism principles in analytic epistemology are sometimes framed to explicitly catch intellectual seemings. Two fairly well-known examples are William Lycan’s Principle of Credulity (‘Accept at the outset each of those things that seem to be true.’29) and Michael Huemer’s Phenomenal Conservatism thesis (‘If it seems to S as if p, then S thereby has at least prima facie justification for believing that p.’30)

2.2 Conservative credos: Russell and Quine Common sense and coherence don’t just underwrite particular conservative methods. Each is also the basis of a discernable analytic philosophical movement with an explicitly conservative conception of the nature of philosophy. The conservative attitude to common sense was first explicitly set out by Russell,31 and it has contemporary analytic influence. For instance, here is David Lewis on the goal of philosophy: One comes to philosophy already endowed with a stock of opinions. It is not the business of philosophy either to undermine or to justify these pre-existing

92

Postanalytic and Metacontinental opinions, to any great extent, but only to try to discover ways of expanding them into an orderly system.32

The claim is not that all philosophical conclusions must confirm the opinions brought to the enterprise; rather, although such opinions are defeasible, they are not to be brought into play as already suspect, so to speak, or ripe for revision – the job of the philosopher is not to begin with the aim of getting somewhere else. This attitude towards common sense entails a very clear conservatism on the part of the philosophers carrying out such projects; indeed, in Lewis’s case in particular, notwithstanding the notorious conclusions he reaches in On the Plurality of Worlds, conservatism of all three kinds outlined above is fairly easy to find acting as a methodological constraint. The other general movement with epistemically conservative implications is naturalizing philosophy, as underwritten by the Quinean revolution in analytic epistemology and philosophy of language, and the associated rise of coherentism as a leading theory of epistemic justification. The Quinean position is readily seen to be conservative, as are forms of naturalized epistemology deriving from it. If our beliefs form a web, and are to be managed by norms requiring the preservation of central (highly connected) beliefs, a range of conservative phenomena arise. The order in which we receive potentially conflicting information may become important; those beliefs we acquire earlier will receive some epistemic weight simply for getting into the system first. Close jumps in theory space will be preferred to long jumps, on pain of rending coherence. Theories that preserve as many of our prior beliefs as possible will ceteris paribus be preferred to those that do not, for this is an aspect of coherence. An influential way of elaborating the link (due to both Sklar and Quine33) appeals to the properties of Neurath’s raft. All epistemic evaluation – whether it be justification or challenge – is essentially local, and needs to be measured against stable background beliefs, which therefore can’t be thrown into question all at once. Other forms of coherentism are less often explicitly identified as conservative, although the conservative implications of coherentism occasionally appear in discussions of the isolation problem for coherentism (the observation that, if coherence both makes for justification and is a function of the internal relations of a set of beliefs, justification does not require contact with the world).34 Recently, in Bayesian mode, conservatism has been explicitly derived from the requirement of diachronic coherence by itself (rather than from overall coherentism).35 Diachronic coherence requires agents to exhibit coherence in their beliefs from one time to another (or the next) time, and is arguably yielded by some Bayesian variants on ordinary conditionalization. But if diachronic coherence is required, doxastic stability is itself a virtue, and conservatism quickly follows. So there exist prominent analytic conceptions of the nature of philosophy that have conservative implications. It’s only fair to flag other conceptions

Analytic Philosophy and Dialogic Conservatism

93

without such implications. Many characterizations of philosophy regard it as essentially concerned with exhibiting meaning relations, from Ryle’s declaration in ‘Systematically Misleading Expressions’ in 1931 on to fairly recent times. Again, in the spirit of the late Wittgenstein, analytic philosophy has been seen as therapeutic in nature. Again, some analytic philosophers (such as Chisholm or Sellars) regard their practice as bringing appropriately analytic methods of ‘honest toil’ to bear on traditional pre-analytic topics (for Chisholm the continuity is traced back to Brentano, for Sellars to Kant). And most famously, the logical positivists present a highly programmatic (and restrictive) vision of the role of philosophy. None of these conceptions of the analytic tradition is conservative in itself, even if practice of it might involve the occasional use of conservative methods (such as appeal to grammatical intuition). The logical positivists, for example, certainly come to their task according positive epistemic weight to some matters and not others, as Carnap makes clear in a late reminiscence of the Vienna Circle: The experience in our investigations and discussions led us to the following practical attitude. We regarded terms of the traditional philosophical language with suspicion . . . and accepted them only when they have passed a careful examination; in contrast, we regarded terms of mathematics and physics as innocent . . . unless cogent reasons had shown them to be untenable.36 But in such cases, the positivist will point to the evidence behind contemporary mathematics and physics as reason for this assignment of weight; the resulting constraint of philosophy by science, and deference to science, thereby rule out certain revolutionary philosophical transitions, but they do not do so in virtue of a species of conservatism obtaining. I’m not sure that this concession reduces the force of the preceding examples very much, however, when it comes to contemporary rather than historical forms of analytic philosophy. Very few analytics today raise the flag of ordinary language philosophy, or philosophy as therapy, or logical positivism. Many take seriously appeals to common-sense or intuition, and will not be unsympathetic with Lewis’s credo; many others are naturalizing philosophers in the Quinean tradition, and so inclined to some kind of comparative conservatism. And still more simply make use of appeals to intuition, reflective equilibrium, thought experiment, regimentation, inference to the best explanation and so forth in their work.

3. Dialogic norms and background conservatism Supposing that much analytic philosophy is conservative in practice, what kind of justification can be given for this? A range of off-the-shelf motivations for conservatism have been discussed in the literature, and perhaps one or more of

94

Postanalytic and Metacontinental

these fit the bill. Most of the motivations people have articulated for conservatism involve coherence considerations (discussed below), or some kind of pragmatic feature of the situation. Lycan presents an argument for conservatism based on considerations of cognitive efficiency – belief revision is costly, so revision for no epistemic benefit is senseless.37 Another case, put forward by Harman, involves the phenomenon of lost evidence.38 I cannot now recall the evidence I once had for very many of my beliefs (those formed in childhood, say, on the basis of testimony), yet intuitively it would be unreasonable to now suspend judgment about so many matters; hence I must give my bare beliefs epistemic weight as they are. Again, one might appeal to the committal force of past decisions, as Jonathan Adler does – beliefs reflect a past epistemic event, in virtue of having action-guiding force (the decision to cease inquiry and the act of accepting p); however, decisions, like intentions, are (second-order) reasons for action.39 Finally, one can bluntly invoke the epistemic predicament of the agent, as Timothy Williamson does in arguing for a limited form of conservatism.40 Agents have to start somewhere, and there is a practical necessity to start where one is. Most of these considerations are obviously not applicable to analytic philosophical practice, if only because the context is one in which many pragmatic barriers to belief assessment are down – it’s not as though there’s any especially urgent need to decide matters once and for all, or as though the evidence for common-sense beliefs or behind common intuitions cannot be sifted for or discussed. Bluntly, cognitive efficiency, lost evidence and the committal force of past decisions all seem to be simply irrelevant in philosophical investigation. Williamson’s point is more effective, since it applies in philosophy as well as it applies anywhere else. Nonetheless, it also does not quite cover the field (indeed, in presenting that account of conservatism Williamson argues that it cannot be used to shore up the practice of appealing to intuition). A second possibility is that analytic philosophers who use such techniques generally endorse the self-consciously common-sensical credo of Russell or Lewis, or accept some form of coherentism as a philosophical epistemology. This is not impossible, I suppose, but given the controversial status of each within the analytic tradition as an actively contested conception of philosophy, it hardly seems likely that such commitments can account for all of the conservatism norms that analytic philosophers abide by in practice. In particular, the role that appeal to intuition plays in analytic philosophy is left more or less unexplained on this account.

3.1 The analytic dialogue (and what’s unusual about it) I think a further explanation of the conservatism of analytic methods comes to light if we consider the nature of analytic philosophy itself. At different times, it has appeared to be concerned with clarity above all (Moore’s ‘work of analysis

Analytic Philosophy and Dialogic Conservatism

95

and distinction’), the use of the tools of logical analysis and an associated interest in argumentation, a certain liaison with (and deference to) the physical sciences, or with linguistic analysis (on either side of the ‘Homeric struggle’ between use theories and truth conditional theories of meaning). But although such features are recognizably present, it’s equally clear that none give the whole story about contemporary analytic philosophy, let alone the tradition as a whole. What is the point of the contemporary analytic enterprise? It seems unlikely that the internal point of view among analytic philosophers is that it is carried out only to achieve consensus, or purely to further our pragmatic interests. Again, it is unlikely that the point of the thing is purely in the procedures it exemplifies (a degenerate form of a Rortyan conversation, say, or an attempt at a Habermasian ideal speech situation – unsurprisingly, these kinds of suggestions are from outside the analytic tradition). Analytic philosophers by and large take themselves to be in either a truth-directed business, or a reasondirected business. A fairly natural conception, then, is that of the ‘dialogue of reason’ that Jonathan Cohen puts forward as the characteristic of analytic philosophizing: If we examine seriatim the problems that actually puzzle analytical philosophers we shall find that the problems of analytical philosophy are all normative problems connected in various ways with rationality of judgement, rationality of attitude, rationality of procedure, or rationality of action. Analytical philosophy seeks a reasoned resolution of such problems . . . Not that it ever hypostatizes rationality in the high, Hegelian manner. No analytical philosopher would proclaim, for example, that ‘Reason is both substance and infinite power, in itself the infinite material of all natural and spiritual life as well as the infinite form, the actualization of itself as content.’ Analytical philosophy occupies itself instead, implicitly or explicitly, with the humbler task of reasoning, at an appropriate level of generality, about how good a reason what can be for what.41 This is unlikely to convince a non-analytic philosopher, who can correctly enough point to other reason-giving activities in other philosophical traditions. Yet we can stress the ‘dialogue’ side of things, rather more than the ‘reason’ (although, like many analytic philosophers it seems right to me to characterize the analytic tradition in terms of a particularly careful attitude to argumentation as well). Contemporary analytic philosophy is a common dialogic enterprise of some kind – an institution, of the kind studied in social epistemology – and is characterized by a fair degree of internal interactivity and responsiveness. Even if other philosophical enterprises also exist which could also be characterized as concerned with normative questions of the kind Cohen raises, they are different dialogues (or monologues), simply in virtue of lacking the appropriate connections to this one. A similar point is made by Hans-Joachim Glock, who

96

Postanalytic and Metacontinental

suggests that analytic philosophy is best understood when characterized in terms of the ties of mutual influence between particular philosophers (both diachronically and synchronically), and some kind of family resemblance relation;42 the synchronic relation of mutual influence is, of course, the fact of dialogue I am stressing now. Not every dialogue shares the same features, and the analytic dialogue is no more constructed on the natural scientific model, say, than most other dialogues within an academic discipline. Four features of the analytic dialogue seem to me relevant to the conservatism of contemporary analytic philosophy, and some of them are distinctive in the sense that not all academic fields exhibit them. First, inferential links are regularly claimed to hold between different pieces of work in analytic philosophy, often on very different topics. You have presented a case for p; I claim to show that if p then q, and I join my result to your result to conclude q. Much work in analytic philosophy is explicitly conditional on other analytic philosophical work in this way, whether or not the connection is meant to extend our knowledge of the antecedent domain or act as a reductio, to shore up the consequent claim or simply to make clear an interesting link. Which of these jobs the conditional connection does depends, of course, on one’s point of view. (For instance, Nick Bostrom’s simulation argument is conditional on a functionalist account of mind, and while he views the simulation conclusion as the point of interest, it’s not hard to reshape this into an objection to functionalism.43) Some analytic philosophers go further, and draw inferential connections between analytic philosophical work and claims in other fields, but these conditional connections have quite different properties, as I shall outline below. Second, as in many research enterprises the dialogue is not essentially hierarchical – no participant’s views are to be discounted or accepted because of their identity. Of course, more and less influential philosophers have emerged, and on occasion authority can be misused to stare down or silence, as in any other enterprise. Allocations of influence can arise in a perfectly respectable manner from experience (everything of X’s that I’ve read has been excellent, so I expect much from this article of hers), but silencing or trumping activities of this kind are in effect breaches of the norms of inquiry and dialogue in play, which are meant to be underlined by standard academic practices of double-blind refereeing and the like. The modest claim here is that participants in the dialogue generally abiding by the constitutive norms of the dialogue are entitled to take themselves to have the same status as potential contributors, as others. Third, the analytic dialogue has no explicit research agenda, setting out an order in which topics are to be treated (although fashion can come and go as in any literature). In this way it diverges from most scientific enterprises, which often have a much more unified sense of the key problems to consider, and what it would be interesting to get ahead on. Philosophy sprawls by comparison. As a result, philosophical researchers, in choosing a starting place for their work, simply have to jump in somewhere and hope that upstream debates are not to

Analytic Philosophy and Dialogic Conservatism

97

be settled entirely against the views they thereby conditionalize on. If my chosen research topic involves refining defeasibility analyses of knowledge to avoid Gettier problems, for instance, then everything I do might well depend on the philosophy of mind yielding an appropriate account of belief, or at least not yielding the reverse. Finally, analytic work very often makes use of devices of common reference. Analytic literatures involve standard cases or patterns of argument; well-known thought experiments, classical ways of putting an objection, and so on. Again, there is a culture, often parodied, of common ostension when it seems at first that nothing would be lost by variation. A set of standard examples (billiard balls hitting one another, elms and beeches) and display sentences (‘The present king of France is bald’, ‘Water is H2O’, etc) do service in the literature, and their appearance usually invokes the problem domain that will be discussed. Again, discussions of problems or phenomena that potentially impact on a range of positions will often use a standard off-the-shelf account as exemplar of the category as a whole; vanilla Bayesianism with ordinary conditionalization might not be widely believed (since many subjective probabilists add their own tweaks to the view), but it gets a regular airing whenever an account of this type is needed to make some point. Finally, analytic philosophers regularly make use of ‘placeholder’ concepts; common concept sketches that are intended to fill a functional role in an argument or position, and for which the right details are something to be filled in later. For instance, a common practice in discussing theories of knowledge has been to treat the belief condition as a proxy, black-boxing the issues it gives rise to, in order to get on to questions of justification (say). I suggest that analytic concern with common reference is entirely rational, given the focus on inferential connectivity the analytic dialogue instances. Such common reference makes for a defeasible way to get the most out of the conditionalizing connection-making practices of the analytic tradition. Real debates can thereby occur (since there is implicit agreement on starting places). Moreover, the inferential connections between different pieces of work can thereby be maximized. If we both take as read the complex background facts p, and you conclude within this that if q then r and I conclude that not-r, it follows from our work – if sound – that within the scope of the assumption p, q is false. Our concern for keeping the background facts straight has yielded extra inferential possibilities. By contrast, if we regularly conditionalize on very different background assumptions, which cut across one another and are not easily brought into relation, your finding that (given p), if q then r, and my finding that (given w) not-r, are not going to get us much further. Inferential overlaps will be impossible or very difficult to establish unless we change our ways. Of course, such common reference does not (or should not) ride over every other consideration. Philosophers who have proved unable to maintain this practice have often moved ‘upstream’ to consideration of the relevant background issue or domain instead.44 And this is obviously a practice that can

98

Postanalytic and Metacontinental

become somewhat pathological. David Lewis complains that the siren lure of the paradox of deterrence has led analytic philosophers to value inferential suggestiveness over practical utility;45 the early Gettier industry arguably accepted a fairly narrow set of terms of the debate about knowledge a little too enthusiastically; and so forth.

3.2 Common assumption, non-deference, and conservatism These features of the analytic dialogue provide the makings for an explanation of at least some of the conservatism of analytic practices. To begin with, analytic conditionalization practices indicate that there is a defeasible norm of the following kind at play here: Common Assumption: Ceteris paribus, maximize overlap between the premises of your work and that of other participants on the same topic within the tradition. Such a norm does not embody any kind of deference to upstream issues within philosophy; modus tollens requires connectivity just as much as modus ponens. Nor does it preclude the challenging of background assumptions where things seem to have gone wrong as the result of trouble upstream. Rather, a norm of common assumption in effect requires the suppression of differences of position where this can be done without much trouble. The agent is giving a kind of default weight to the background assumptions made within discussion of a particular topic in the tradition, in order to allow his or her work to be inferentially connected to as much of the literature as possible. Moreover, this is a social norm, in the sense that whatever justification it has is in terms of its effects on the enterprise as a whole (promoting inferential connectivity), and not merely in virtue of being independent good advice for analytic philosophers as individuals. If participants abide by this norm on the whole, an immediate consequence is that a weak kind of Reidian credulity then obtains within the analytic community. To learn that another participant sets out from a particular set of premises or assumptions in examining the topic you both work on is to have at least some reason to do the same. So, the norm of common assumption will motivate the following claims: (RC1) If S learns that there exists another participant in the dialogue for whom it seems that p, then S should give some positive weight to p. (RC2) If S is deliberating between evidentially equivalent hypotheses p and q and p is closer to the background beliefs of another participant in the dialogue than q, ceteris paribus S should prefer p to q.

Analytic Philosophy and Dialogic Conservatism

99

These are very weak (in the sense ‘not committal’) conditions. Giving some positive weight to a claim, or having a ceteris paribus preference for it, is entirely consistent with deciding against it or suspending judgment on the issue. For a norm of common assumption only asks of the philosopher that he or she suppress differences where this can be done without much trouble, not that differences be suppressed come what may. Moreover, on many philosophically interesting issues, these conditions will effectively be idle, because other participants in the debate will have reported differing intuitions or background beliefs. (RC1) and (RC2) in effect embody a kind of social conservatism, which requires participants to take seriously the intuitions and beliefs reported beforehand and by others in the dialogue. Social conservatism is, however, not conservatism of the kind on display in the philosophical practices considered earlier. Those involved philosophers setting off from their own intuitions, their own judgments about common sense, and so on. Putting things another way, one might imagine a highly precedentbound research community in which (RC1) and (RC2) are met, but in which they entirely trump the intuitions and beliefs of those who come later. Getting from this social conservatism to the kinds involved in analytic practice involves a further norm, which I claim is also indicated by the features of analytic philosophy that I have outlined above. For the lack of an explicit research agenda and the non-hierarchical nature of the dialogue mean that deference on the basis of the relative importance of different pieces of philosophy, or deference on the basis of the mere identity and institutional importance of another, are not in play. Epistemology is not to defer to metaphysics on a point of common interest; humbler epistemologists are not to defer to the more prominent in virtue of their prominence. I suggest that these matters indicate that participants in the analytic dialogue abide by a further defeasible norm: Non-deference: Ceteris paribus, don’t take yourself to be a worse epistemic authority than other participants on the same topic in the dialogue. A clarification: I am not claiming that there is a norm forbidding the recognition of superior merit and general excellence in another philosopher, or preventing one from being disposed to take a view more seriously (on learning that the Great One has argued for it) before coming across the details of the case for it. Rather, the intended limitation is to cases where the reasons at play in the dialogue are available and have been absorbed. One may have appropriately great expectations for the Great One’s work, but upon actually absorbing it, it is the details of the case presented that matter, rather than the fact it has that particular author. The norm is, nonetheless, defeasible because it’s entirely possible for a philosopher to acquire evidence that there are parts of philosophy where he or she is not strong, easily muddled, or simply doesn’t get the point at issue. The norm could also conceivably be defeated if most philosophers regularly defer to a minority on a particular kind of issue, but this appears to be

100

Postanalytic and Metacontinental

genuinely rare; as far as I can tell, regular exceptions of this kind only arise where philosophers defer to those with technical skills when there is a technical aspect to a discussion (a good example being the not uncommon deference to George Bealer and George Boolos as second-order logic whisperers). These cases seem in some ways to resemble the not unusual analytic deference to the deliverances of the sciences and mathematics. The non-deference that I have in mind here, and the way it interacts with the analytic norm of inferential connectivity, comes out more clearly if it’s compared to the attitude that naturalizing philosophers in the Quinean tradition take to the natural sciences and mathematics. Naturalizing philosophers regard philosophy as continuous with, but constrained by, the deliverances of the relevant natural sciences and mathematics. Such continuity and constraint imply inferential connections between the two, and in fact this is a common naturalistic move. Quine’s ruling scepticism out of court in naturalized epistemology, for instance, amounts to making his philosophical conclusions conditional on the sciences he trusts being right. Now if a conditional connection is made out between a scientific matter and a piece of philosophy in this way, one might think both sides are hostage to it. If the scientific claim turns out to be true, the philosophical claim follows by modus ponens. But if the debate in philosophy goes against that philosophical claim, modus tollens tells us that the scientific claim is false as well. Of course, naturalizing philosophers are not at all inclined to draw that conclusion; rather, they will diagnose philosophical pathology. The conditional here is a one-way conditional, so to speak. By contrast, within the analytic literature no such one-way conditionals can be readily discerned – where inferential connections between different philosophical positions are accepted, philosophers are apt to employ both ponens and tollens.46 The argument now should be clear. Suppose I generally abide by (RC1) and (RC2). If so, I am taking the intuitions and background beliefs of other participants in the dialogue to have some positive epistemic weight, and so I am allocating those participants at least some status as epistemic authorities. But the norm of non-deference now extends this consideration to my own judgments of this kind, except in those cases where I am aware that I have a special weakness. As a result, more familiar types of conservatism can now be derived: (C1) If it seems to S that p, then S should give some positive weight to p. (C2) If S is deliberating between evidentially equivalent hypotheses p and q and p is closer to S’s beliefs than q, ceteris paribus S should prefer p to q. The upshot is that appeals to intuition and common sense have at least some weight as starting places (by (C1)), and coherence-building methods have at least some reason to stay local (by (C2)), simply because the analytic dialogue aims to maximize a certain kind of inferential connectivity without employing such structuring devices as hierarchical authority or explicit research agendas. Of course, this is a background conservatism, so to speak; philosophers may

Analytic Philosophy and Dialogic Conservatism

101

well have further grounds in their general epistemic views or conception of philosophy for being conservative. The features and norms I have ascribed to analytic philosophy do not presuppose an implausible analytic saintliness or self-effacement. For instance, (C1) and (C2) are consistent with the participant giving his or her own intuitions and beliefs much greater weight than those of other participants; this is not necessarily an epistemically impartial position. But the fact that (RC1) and (RC2) also hold means that participants can all accept (C1) and (C2) – and so appeal to intuition, common sense, and the rest – while varying in their degree of epistemic impartiality. I suggest that this is also an observable feature of the analytic dialogue; people stick to their guns in the face of contrary intuitions to varying degrees. A persistent minority of philosophers don’t share the prevailing intuition about Twin Earth, for instance; some of the majority are happy to ignore this and some are not.47 Again, my claim that the analytic dialogue is essentially non-hierarchical is likely to excite scepticism (or derision) from some non-analytics. The claim is, of course, about the entitlements or expectations of participants, a concept I am leaning on quite strongly. It might be that not everyone counts as a participant who wishes to, simply because the relation of mutual influence that Glock mentions does not obtain (and of course in some cases this might be the result of bigotry, elitism or the like). Many analytic journals are widely respected within the analytic community, and I’d suggest that having one’s work appear in such journals is generally sufficient for dialogic participant status (whether or not it is necessary). My case is consistent with participation being a matter of degree, and I don’t want to aim for false precision on this point. A second objection on the non-hierarchical status of analytic philosophy might simply begin from the charges of institutional and political conservatism I left in my first paragraph. But I doubt this has much bearing on the point at issue here either, because the conservatism claims I reach depend on analytic philosophers regarding themselves as non-hierarchical, rather than on this attitude actually reflecting reality. A final matter. I claim that the norms of common assumption and nondeference obtain in contemporary analytic philosophy as a matter of descriptive fact. And the explanatory, as opposed to justificatory, function of the argument I have given doesn’t depend on anything beyond this bare claim. But of course one can go on to ask whether common assumption and non-deference are justifiable norms. They are, I think, if inferential connectivity and a non-hierarchical structure are themselves reason- or truth-promoting practices in philosophy. I suspect most analytic philosophers would agree, because this view coheres very strongly with the features of analytic philosophy we tend to cite in our defence or favour – the claims that careful argumentation, clarity, formal methods and so on genuinely further reason-giving or truth-seeking. As a result, my conditional justification of conservative analytic practice (given the features and norms of the field) is also potentially part of a wider justification of such

102

Postanalytic and Metacontinental

conservatism from an analytic point of view. Such a wider justification need not ignore the ways in which dialogic conservatism might well be unsatisfactory; to say that the current analytic enterprise is broadly reason- or truth-promoting is not to say that it could not be improved, and an improved model might well be less conservative. But it might well explain why charges of (epistemic) conservatism are not likely to sway the analytic.

Notes 1

2

3 4

5

6 7

8

9

10 11 12 13

14

15

16

I would like to thank Jack Reynolds, James Williams, Ed Mares, Mark Hodges and audiences at La Trobe University and the University of Tasmania. This chapter was written with financial assistance from the Australian Research Council. See, for instance Glock’s discussion of similar claims in What is Analytic Philosophy? (New York: Cambridge University Press, 2008), 180–189. Richard Foley, ‘Epistemic conservatism’, Philosophical Studies 43 (1983): 165–182. The distinction, and some of the terminology here, is due to Hamil Vahid, ‘Varieties of Epistemic Conservatism’, Synthese 141 (2004): 97–122. Roderick Chisholm, ‘A Version of Foundationalism’, Midwest Studies in Philosophy 5 (1980): 551–552. Gilbert Harman, Change in View (Cambridge, MA: MIT Press, 1986), 38. Lawrence Sklar, ‘Methodological Conservatism’, Philosophical Review 84 (1975): 374–400 at 378. W. Lycan, ‘Moore against the New Skeptics’, Philosophical Studies 103 (2001): 35–53. These are, roughly, beliefs that are about matters noted at a place and time but that are modestly generalizable to other places and times, that don’t require technical language to express, that are such that their modest generalizations are accepted by nearly everyone regardless of theory, and are too obvious to ordinarily mention, and that we are inclined to go on believing regardless of philosophical argument. See Lycan ‘Moore against the New Skeptics’, at 48–49 (and note that there are further conditions of this kind). Lycan treads with some care around the simpler claim that, for Moore, common-sense claims are to be invested with positive epistemic weight in virtue of being common sense. Lycan, ‘Moore against the New Skeptics’, 41. D. Lewis, On the Plurality of Worlds (Oxford: Blackwell, 1986), 134–135. J. Mackie, Ethics: Inventing Right and Wrong (Harmondsworth: Penguin, 1977), ch.1. P. Churchland, ‘Eliminative Materialism and the Propositional Attitudes’, Journal of Philosophy 78 (1981): 67–90. See, for instance, Frank Jackson, From Metaphysics to Ethics: A Defence of Conceptual Analysis (Oxford: Oxford University Press, 2000), ch.2. R. Chisholm, The Problem of the Criterion (Milwaukee: Marquette University Press, 1973), 12. G. Harman, ‘The Inference to the Best Explanation’, Philosophical Review 74 (1965): 88–95; ‘Knowledge, Inference, and Explanation’, American Philosophical Quarterly 5 (1968): 164–173.

Analytic Philosophy and Dialogic Conservatism 17

18

19 20

21

22

23

24

25

26

27

28

29

30

31

32 33

34

103

Jerry Fodor, The Language of Thought (Cambridge, MA: Harvard University Press, 1975). W. V. Quine and J. Ullian, The Web of Belief (New York: Random House, 2nd ed. 1978), ch. 6. W. V. Quine, Word and Object (Cambridge, MA: MIT Press, 1960). Using David Lewis as an example again, although his concession is toward common sense, his guiding argument for modal realism is by theoretical virtue: ‘Modal realism is fruitful; that gives us good reason to believe that it is true.’ (On the Plurality of Worlds, 4). J. Thomson, ‘A Defence of Abortion’, Philosophy and Public Affairs 1 (1971): 47–66. But see N. Daniels ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’, Journal of Philosophy 76 (1979): 256–282 for an influential argument that it too should be standard in analytic ethics. This is a standard characterization of intuition in the literature, although not totally uncontroversial. See J. Weinberg, ‘How to Challenge Intuitions Empirically Without Risking Skepticism’, Midwest Studies in Philosophy 31 (2007): 318–343. The classic defence of this view in contemporary analytic epistemology is George Bealer and Peter Strawson, ‘The Incoherence of Empiricism’, Proceedings of the Aristotelian Society, Supplementary Volume 66 (1992): 99–143. The quote is from Alvin I. Goldman, ‘A Priori Warrant and Naturalistic Epistemology’, in his Pathways to Knowledge: Private and Public, (New York: Oxford University Press, 2002), 24–50 at 45. I’m not asserting Goldman’s own account is of this kind; he has a slightly more complicated view because he brings a reliabilist account of justification to the issue, and so, for instance, intuitions might well justify beliefs not because of their origin but merely because the intuition-to-belief transition has a high truth ratio. See Timothy Williamson, ‘Philosophical “Intuitions” and Scepticism about Judgement’, Dialectica 58 (2004): 109–153; Jaakko Hintikka, ‘The Emperor’s New Intuitions’, Journal of Philosophy 96 (1999): 127–147. For an overview of analytic use of intuition since the 1970s that does flag the link to conservatism, see John Symons, ‘Intuition and Philosophical Methodology’, Axiomathes 18 (2008): 67–89. R. Fumerton, ‘Epistemic Conservatism: Theft or Honest Toil?’, Oxford Studies in Epistemology 2 (2008): 63–86 at 71. William Lycan, Judgement and Justification (Cambridge: Cambridge University Press, 1988), 165. Michael Huemer, Skepticism and the Veil of Perception (Oxford: Rowman and Littlefield, 2001), 99. Bertrand Russell, The Problems of Philosophy (London: Oxford University Press, 1959), 25. David Lewis, Counterfactuals (Cambridge, MA: Harvard University Press, 1973), 88. See Lawrence Sklar, ‘Methodological Conservatism’; Quine and Ullian, The Web of Belief. For instance, see Laurence. BonJour, The Structure of Empirical Knowledge (Cambridge, MA: Harvard University Press, 1985), 108. (In passing, BonJour’s discussion

104

35

36

37

38 39

40

41 42

43

44

45

46

47

Postanalytic and Metacontinental

of Brand Blanshard in appendix B involves a quite different sense of conservatism.) See W. J. Talbott, ‘Two Principles of Bayesian Epistemology,’ Philosophical Studies 62 (1991): 135–150 and David Christensen, ‘Clever Bookies and Coherent Beliefs’, Philosophical Review 100 (1991): 229–247. Christensen presents the case against diachronic coherence and associated Bayesian principles in ‘Diachronic Coherence vs. Epistemic Impartiality’, Philosophical Review 109 (2000): 349–371. Rudolf Carnap in Paul Schilpp (ed), The Philosophy of Rudolf Carnap (La Salle, IL: Open Court, 1963), 65–66. Lycan, Judgement and Justification, 174–175. Lycan is developing a theme from Lawrence Sklar’s ‘Methodological Conservatism’. G. Harman, Change in View, ch.4 and ch.5. Jonathon Adler, ‘An Overlooked Argument for Epistemic Conservatism’, Analysis 56 (1996): 80–84. T. Williamson, ‘Philosophical “Intuitions” and Scepticism about Judgement’, 125–131, particularly 126 (though note the qualifications given thereafter). J. L. Cohen, The Dialogue of Reason (Oxford: Clarendon Press, 1986), 49–50. H-J. Glock, What is Analytic Philosophy? (New York: Cambridge University Press, 2008), 222–223. Nick Bostrom, ‘Are You Living in a Computer Simulation?’, Philosophical Quarterly 53 (2003) 243–255. For instance, in explaining his philosophical career, Fred Dretske notes ‘It was a growing sense of these lacunae [unsatisfactory accounts of belief and experience] in my epistemology that drove me into philosophy of mind).’ Fred Dretske, Perception, Knowledge and Belief (Cambridge: Cambridge University Press, 2000), x. ‘I am afraid that because paradoxical deterrence is philosophically fascinating, it will be much discussed; and because it is much discussed, it will be mistaken for reality.’ D. Lewis, ‘Devil’s Bargains and the Real World’, Papers in Ethics and Social Philosophy (Cambridge: Cambridge University Press, 2000), 211. This difference also explains why the norm of common assumption doesn’t overgenerate to any field that can be inferentially connected to the analytic literature. The way analytic philosophers are wont to react to divergent intuition has become an issue in general, as the result of well-known work in ‘experimental philosophy’ by Stephen Stich, Jonathan Weinberg and Shaun Nichols.

Chapter 7

Cartesian Echoes in the Philosophy of Mind: The Case of John Searle Pascale Gillot

Recent studies, such as Hilary Putnam’s, have shown the underestimated importance of the ‘Cartesian picture’ of the mental in contemporary philosophy of mind which would remain widely dominated by ‘Cartesianism cum materialism’.1 Even though most contemporary approaches to mind have long abandoned the classical dualist claim of an ontological distinction between mind and body, they appear to be deeply influenced still by Cartesian vocabulary and categories (categories such as: ‘mental’ and ‘physical’, ‘inner’ and ‘outer’, ‘subjective’ and ‘objective’). John Searle’s program seems to be based, to the contrary, on an explicit and recurring refusal of Cartesianism in the philosophical analysis of mind; this would mean a rejection of concept dualism, as well as of property dualism and substance dualism. The reasons for the failures or stagnation in philosophy of mind in the last decades, according to Searle, lie in the amazing persistence of a terminology and of a network of concepts inherited from early modern philosophy, especially from Descartes. Searle’s solution to the Mind-Body Problem, therefore, presents itself as an attempt to reject the theoretical assumptions implied by the ‘traditional categories’ which are shared by monists and dualists, idealists (even if there are very few of them) as well as materialists (the immense majority). Such a solution, identified with ‘biological naturalism’, would then be emancipated from classical and fruitless oppositions (monism/ dualism, idealism/materialism), and would eventually offer a genuine non Cartesian picture of the mind. Nevertheless, the Searlian conception of the mind, through its central use of the term ‘consciousness’, might not have ‘escaped’ the traditional categories as much as it claims it has. In this chapter, I interrogate the paradoxical relationship that links Searle’s approach to Descartes’ original theory, since his approach represents one of the most direct attempts to reject Cartesian philosophy. First, I will sketch out some of the main lines of Searle’s general theory of mind, insofar as it criticizes the traditional alternative between dualism and materialism, and also aims at a fundamental subversion of the computational theory of mind upon which cognitive science was built. Second, I examine more precisely the

108

Postanalytic and Metacontinental

Searlian critique of Cartesianism, which is not only about substance dualism, but also regards a whole ‘grammar’, the whole set of Cartesian terms and concepts. Finally, I will analyse the limits of such a critique, insofar as internalism and mentalism, among others, are concerned.

1. Searle’s Program: Beyond the choice between dualism and materialism As is well-known through the wide diffusion of the ‘Chinese room argument2’, John Searle aims at the disqualification of the mind-computer theoretical model, which has been dominant since the ‘cognitivist turn’ of the 1960s. Such a computational model, reducing the mind to a logical automaton, is particularly at stake in the artificial intelligence program; it is also generally implied in cognitive science, at the intersection of neurosciences, psychology and philosophy. Its philosophical background lies in functionalism, a theory that was developed particularly by Hilary Putnam with the identification of the human mind with a Turing Machine.3 John Searle, who was brought to philosophy of mind on the basis of a prior interest in philosophy of language, and the need for an account of the intentionality of speech acts, claims that functionalist premises involve a fallacy. In The Rediscovery of the Mind, he underlines the limits and contradictions inherent to the computational model and to the ‘strong artificial intelligence’ view. The human mind is not a computer program, nor a Turing Machine. The very existence, in the mind, of mental contents, meanings and conscious states, refers to a semantic field which is radically absent from the syntactical operations at work in any computational device. If one postulates, as Searle does explicitly, that semantics, which envelops consciousness and intentionality, cannot be reduced to syntax, then functionalism, namely computer functionalism, is false. It would not resist, Searle argues, the ‘commonsense objection’ according to which ‘strong AI (. . .) [leaves] out the crucial things about the mind such as consciousness and intentionality’.4 Through this fundamental differentiation between syntax and semantics, it appears that standard cognitive science, insofar as it is linked with strong AI, does not constitute an adequate account of mental phenomena: it would necessarily fail in the attempt to explain these phenomena, because it does not take into consideration essential features of the mind like consciousness, qualitativeness, intentionality and the first-person point of view. In this respect, it shares common postulates with methodological behaviourism and its singular claim that inner conscious states are theoretically irrelevant. Searle’s ‘critique of cognitive reason’, therefore, must be set within a larger frame defined by the critique of materialism in all the forms it took during the twentieth century. Behaviourism and functionalism undoubtedly constitute the main targets of the Searlian critique. But physicalism and the many versions of

Cartesian Echoes in the Philosophy of Mind

109

the identity theory, which was first formulated by Place and Smart in the 1950s, also get disqualified, for the main reason that they deny the irreducibility of the first-person point of view and of phenomenal, subjective experience. Identity theory infers that mental, conscious states are just brain states, and may be reduced to their neurophysiological correlates, from the premise that brain states cause mental states. If pain is caused by the activation of C-fibres, then one should admit that it is nothing else than a certain activation of C-fibres, that is some objective phenomenon, explicable entirely from the third-person point of view. But such an inference would be mistaken. Among Searle’s main claims is the point that consciousness, and mental phenomena in general, are indeed caused by lower-level brain states, and yet are ontologically irreducible to these; causal reduction is perfectly acceptable, but ontological reduction is not. The causal emergence theory, sustained in The Rediscovery of the Mind and other texts,5 is then explicitly combined with the non-physicalist view that the firstperson ontology remains irreducible. This original combination takes the name of ‘biological naturalism’, and represents Searle’s solution to the Mind-Body Problem: ‘Mental phenomena are caused by neurophysiological processes in the brain and are themselves features of the brain.’6 In other terms, ‘consciousness is caused by microlevel processes in the brain and realized in the brain as higher-level or system feature.’7 From such a perspective, Searle’s position assumes that the mental, insofar as consciousness constitutes its principal attribute, must be understood as a biological phenomenon, just like photosynthesis or digestion.8 But if consciousness, and all mental capacities without exception are just biological features of the human brain, the result of biological evolution, then the dualistic claim, according to which reality divides into two independent realms, the mental and the physical, has to be rejected. This characterization of consciousness as part of the biological, natural world, leads Searle, like so many of his contemporaries in the field of the philosophy of mind, to the explicit refusal of Descartes’ doctrine, insofar as it would represent the most entrenched dualistic system, with the metaphysical view of substance dualism. Indeed, Descartes asserts very clearly, in particular in the Meditations on First Philosophy, the existence of a real distinction between the res cogitans and the res extensa, whose respective attributes, thought and extension, are conceived independently one from another.9 In the Second Set of Replies to the Objections, he offers a systematic demonstration of the real, substantial distinction between mind and body.10 According to such a demonstration, Descartes’ dualism (to use this rather problematic, controversial term) would principally consist in the inference of a substantial, ontological distinction (the distinction between thinking substance and extended substance) from a conceptual one: mind and body must be seen as two different substances because their concepts, thought and extension, differ entirely one from another. In that respect, ontological dualism would also represent a consequence of the modern representation of the universe implied by mechanical philosophy and the Galilean science

110

Postanalytic and Metacontinental

of nature, which stand at the core of Descartes’ philosophical and epistemological doctrine. In some way, the real distinction claim is read through its relationship to the modern science of nature which led to ‘the exclusion of consciousness from the subject matter of science by Descartes, Galileo and others in the seventeenth century’. It is even described, in The Rediscovery of the Mind, as ‘a useful heuristic tool in the seventeenth century, a tool that facilitated a great deal of the progress that took place in the sciences’: the epistemological autonomy for the new science of nature, characterized by objectivity and quantitativeness, was also at stake in ‘the separation between mind and matter’.11 Yet at the same time this very representation of a separation between two different types of reality, a mental one and a physical one, is recurrently denounced by Searle as some kind of metaphysical aberration, a philosophical confusion; it necessarily fails in the task of providing ‘an adequate or even coherent account of the relationship between the mind and the body’, and ‘just seems out of the question as a scientific hypothesis’.12 The Rediscovery of the Mind, for example, describes ‘Cartesian dualism’ as an absurd view,13 and proposes to substitute for the famous Cartesian inference in the Discours de la méthode, ‘I think, therefore I am’, the following, unorthodox proposition: ‘I am a thinking being, therefore I am a physical being.’14 Searle’s general program appears then to be deeply antiCartesian, for it tends to the removing of an ‘obstacle’ that would prevent us from understanding the place of consciousness in nature, namely the fact that it is a biological feature; the obstacle is nothing but the Cartesian-style dualism and the classical oppositions between the mental and the physical, the subjective and the objective, the unextended and the extended, etc., that lead to inextricable problems and persistent dead ends. In The Rediscovery of the Mind, one can read a genealogy of the main and still operating problems in philosophy of mind. In a section significantly called ‘Descartes and other disasters’, ambivalence in the Searlian reading of Cartesian philosophy can be noticed.15 On the one hand, like other theoreticians of the mind from very different schools, Searle sees in Descartes’ work the very origin of modern philosophy of mind;16 as such, Descartes’ most influential conceptions and axioms, like the identification of consciousness to the essence of mind, cannot be simply ignored. On the other hand, Cartesian philosophy is seen as the main source of endless and sterile debates around the mind-body problem, the problem of other minds, the problem of personal identity . . . and even, more indirectly, the problem of mental causation in a physical, causally closed world. Descartes’ influence would then consist less in a defined doctrine or system (if it exists) than in some widely diffused network of concepts, issues and topics that survived substance dualism: as a matter of fact, ‘many of his [Descartes’] views are routinely expounded, and uncritically accepted today by people who cannot even pronounce his name.’17 The real solution to those persistent problems, responsible for the failures of the philosophy of mind in the last decades, would then imply not only the rejection of Descartes’ ontology

Cartesian Echoes in the Philosophy of Mind

111

(substance dualism), but also and above all the dissolution of the Cartesian conceptual and verbal apparatus.

2 ‘The inadequacy of the traditional vocabulary’:18 The attempt to reject Cartesian terminology One of Searle’s main points is the strongly asserted disqualification of classical and lasting alternatives, such as materialism vs. idealism, monism vs. dualism, reductionism vs. anti-reductionism, etc. Such a disqualification, implied by the ‘biological naturalism’ assumed in The Rediscovery of the Mind, is built upon the rejection of conceptual dualism, and has great critical value. It implies an opposition to the hegemonic trend in philosophy of mind for many decades, materialism, considered in all its various modalities: behaviourism, identity theory (type-type identity or token-token identity), functionalism, eliminativism. But it also leads to the critique of recent attempts to abandon strong reductionism, such as property dualism, dual aspect theory and anomalous monism. Insofar as the ‘standard’ materialistic view is concerned, Searle’s particular statement is the following: ‘Materialism is (. . .) the finest flower of dualism’.19 As a matter of fact, the main claims of materialism, such as the reducibility of mental phenomena to physical or objective processes (computational or neurophysiological), the exclusive existence of public observable phenomena, the negation of independent subjective processes, like consciousness or qualia, are based upon a particular reading of the relation between the mental and the physical. Materialism adheres to the scientific representation of reality, and it does not admit the existence of a mental, that is non-physical, entity or cause in the universe. Yet this exclusion derives from the definition of the relation between the mental and the physical in the terms of an exclusive disjunction: the mental qua mental excludes the physical qua physical, as the physical qua physical excludes the mental qua mental. But such an exclusive disjunction is inherited from Descartes’ philosophy, which precisely took the conceptual distinction of mind and body, that is their supposed logical independence as concepts, as the premise of substance dualism. According to this Searlian analysis, the materialistic tradition is still deeply, ‘inadvertently’ informed by Cartesian categories and conceptual distinctions. Thus, Searle writes, ‘in denying the dualist’s claim that there are two kinds of substances in the world or in denying the property dualist’s claim that there are two kinds of properties in the world, materialism inadvertently accepts the categories and the vocabulary of dualism. It accepts the terms in which Descartes set the debate. It accepts, in short, the idea that the vocabulary of the mental and the physical, of material and immaterial, of mind and body, is perfectly adequate as it stands. It accepts the idea that if we think consciousness exists we are accepting dualism’; yet, Searle specifies, ‘what I believe (. . .) is that the vocabulary, and the accompanying categories, are the source of our deepest philosophical difficulties.’20 In that sense, cognitive

112

Postanalytic and Metacontinental

science, just like behaviourism through its very rejection of the ontology of mental subjective states, fails in its fundamental program; that is, the dissolution of the dualistic view. Nevertheless, Searle’s critique of the main and most famous theories in contemporary philosophy is not only addressed to the materialistic, reductionist tradition. The current non-physicalist view, as it is represented for example in Thomas Nagel’s dual aspect theory21 or in Donald Davidson’s anomalous monism,22 is also criticized insofar as it seems to postulate (especially in Nagel’s view) some other form of dualism, property dualism. Property dualism, as a matter of fact, remains prisoner of Cartesian categories, even though it rejects the ontological distinction between mind and body: ‘it still forces us to suppose that there are properties of the body, presumably properties of the brain that are not ordinary physical properties like the rest of our biological makeup. The problem with this is that we do not see how to fit an account of these properties into our overall conception of the universe and of how it works. We really do not get out of the postulation of mental entities by calling them properties. We are still postulating nonmaterial mental things. It does not matter whether we say that my conscious pain is a mental property of my brain or that it is an event in my brain. Either way, we are stuck with the traditional difficulties of dualism.’23 Insofar as property dualism would imply the causal irreducibility of the mental, Searle is then led to reject such a philosophical view.24 One can then measure the specificity of the ‘biological naturalism’ program, considered through its attempt to constitute a genuine non-Cartesian picture of the mind, and above all through its refusal of conceptual dualism as the very principle of property dualism and substance dualism. Searle’s diagnosis of theoretical dead ends in philosophy of mind emphasizes ‘the inadequacy of the traditional vocabulary’,25 that is to say the inadequacy of the Cartesian vocabulary; it may sound curiously like an echo of externalist, anti-mentalistic readings of the ‘category mistake’, or ‘grammatical error’, at stake in the traditional philosophical understanding (materialist or not) of the relation between mind and body. First of all, such a diagnosis reminds us of the genealogy of the ‘official doctrine’ associated to ‘Descartes’ myth’ about the mental, established by Gilbert Ryle in The Concept of Mind.26 The hypothesis of a ‘double life’ in each individual, a mental life and a physical life that would constitute two independent processes, is based upon this ‘category mistake’ through which the mind is identified to a category comparable with the category of body. The concept of mind in its Cartesian acceptance is then constructed on the basis of a systematic negation of the classical attributes that belong to the body in the mechanistic conception of the universe – mind is the obscure double of the body: it is unextended, unobservable, etc. The official doctrine, in the current dualistic perspective, lies upon a fundamental conceptual and verbal antithesis: the antithesis of inner and outer, which gets contested by Ryle in his analysis of mental terms. Searle, when he proposes to revise the ‘list of the traditional features of

Cartesian Echoes in the Philosophy of Mind

113

the mental and the physical that are supposed to render them mutually exclusive’ (subjective/objective, qualitative/quantitative, intentional/nonintentional, not spatially located/spatially located . . .),27 appears to be strikingly close to the approach developed in The Concept of Mind, which consisted in trying to dissolve the elementary conceptual distinctions and categorizations that ground the Cartesian ‘myth’ about individual human identity. More generally indeed, it also seems to renew certain topics and analyses inherited from ordinary language philosophy, in the Wittgensteinian style. Strange and paradoxical as it may be, Searle’s critique of dualism in all its forms is not so far from the way in which Wittgenstein, in The Blue Book, resets the classical distinction between mind and body in the context of a philosophical confusion, which makes the mind the ghostly double of the body, a thing of a particular type, and then does not take into account the specific grammar of mental terms. Furthermore, the Searlian statement, ‘(. . .) we lack a neutral vocabulary in which to describe mental phenomena’,28 could also be read in reference to William James’ philosophy of pure experience, in the Essays in Radical Empiricism. Indeed, if Searle does not share the philosophical postulates of radical empiricism and ‘neutral monism’, he nonetheless renews the Jamesian project to dissolve the classical verbal and conceptual oppositions between subject and object, thought and matter.29 Among the latent theoretical assumptions behind the traditional vocabulary, Searle rejects what he calls ‘the basic assumption’ (still silently at work in the reductionist claim), that is “the distinction between the mental and the physical”, which entails that ‘“mental” and “physical” name mutually exclusive ontological categories.’ The rejection of such a basic distinction is part of the general program propounded by Searle in philosophy of mind: ‘we should not accept the traditional terminology and the assumptions that go with the terminology’, for ‘expressions like “mind” and “body”, “mental” and “material” or “physical” . . ., as they are used in discussions of the mind-body problem are the source of our difficulty and not tools for its resolution.’30 Also quite significant, in that respect, is Searle’s refutation of the traditional ‘representative theory of perception’, supposedly at work in Cartesian philosophy, which claims that ‘we cannot directly perceive tables, chairs, mountains, etc. but can only perceive our ideas of these things.’31 Rejecting such a theory, ‘the greatest single disaster in the history of philosophy’, Searle defends what he calls a ‘direct realism’.32 Such a realism offers remarkable concordances with the dissolution of the partition between things and sense-data, between objects of the world and what would be their mental duplicates, engaged by James in particular on the occasion of the 1905 paper ‘La notion de conscience’. In that respect, it might also be compared with Putnam’s defence of direct realism, and rejection of what he calls the ‘interface theory’ involved in ‘Descartes’ own conception of the mental as kind of inner theatre’.33 Finally, to conclude this bringing together of an externalist (yet not physicalist) tradition that would go from William James to contemporary theories, it is

114

Postanalytic and Metacontinental

tempting to draw an analogy between the general position sustained in The Rediscovery of the Mind about materialism being ‘the finest flower’ of dualism, and Putnam’s verdict concerning the existence of a Cartesianism cum materialism that would have become the hegemonical view in the philosophy of mind today. Yet, of course, such a bringing together admits of obvious limits, dealing with claims about the mental that cannot be reconciled. Gilbert Ryle, for instance, defends a kind of ‘logical behaviorism’ which is refused by Searle because it denies the theoretical relevance of interiority and privileged access in the analysis of mental terms. Wittgenstein’s ‘philosophy of psychology’, since it is opposed to any kind of referentialist theory of the first person, would be no more acceptable. As for the Jamesian philosophy of pure experience, aiming at the dissolution of the very ‘notion of consciousness’, it obviously cannot fit the Searlian approach to the mind. Putnam’s externalist theory of reference, according to which concepts, intentional contents and meanings are just ‘not in the head’, holds to a fallacious conception of intentionality, insofar as it makes a distinction between intentional phenomena and conscious, inner states.34 Indeed, this complex and paradoxical relation to the externalist, anti-mentalistic tradition may also be seen as the sign of some kind of ambivalence in the Searlian analysis of Cartesian dualism. On the one hand, Cartesian dualism is clearly and thoroughly disqualified. On the other hand, Searle claims that dualism is no more false a doctrine than materialism, and that it will not disappear, since it is based upon the central, ‘obvious’ view that we have conscious experiences and that consciousness is irreducible to the third-person point of view; in that respect, he even seems to recognize a major theoretical relevance for dualism rather than materialism. Thus, he writes about the recent resurgence of concept dualism, for example in Nagel’s anti-reductionist claim and rehabilitation of phenomenal experience: ‘My guess is that dualism, in spite of being out of fashion, will not go away. Indeed in recent years dualism, at least property dualism, has been making something of a comeback, partly due to a renaissance of interest in consciousness. The insight that drives dualism is powerful. Here is the insight, at its most primitive: we all have real conscious experiences and we know that they are not the same sort of things as the physical objects around us.’35 If Searle does not accept the existence of a non-biological reality, the existence of a mental reality separated from the physical one, his entire approach to mind is nevertheless constructed on the postulate of the primitive, phenomenal existence of consciousness, together with the claim that the ontology of mental states is a first-person ontology. He therefore rejects, for example, Daniel Dennett’s attempt to ‘explain consciousness’ in a functionalist perspective through the method of heterophenomenology.36 This method does not take into account the first-person point of view, and would eventually lead, according to Searle, to the denial of consciousness itself as inner, subjective experience. He also refuses, as we have noticed, externalism in the philosophy of language as well as in the

Cartesian Echoes in the Philosophy of Mind

115

philosophy of mind, and does not admit Putnam’s argument against internalism and mentalism (namely the Twin-Earth argument presented in ‘The Meaning of “Meaning”’). Indeed, the Searlian refutation of Putnam’s Twin-Earth argument, more particularly developed in Intentionality, goes with an overt defence of internalism: intentional contents, ‘in the head’ of the speaker or hearer, are considered sufficient to determine reference.37 Moreover, by asserting that consciousness is ‘the central mental notion’,38 so that any philosophical account of the mind would necessarily be an account of consciousness, and consequently by identifying mental states with inner, private states, by separating mind and behaviour (as is shown in the use of the zombie hypothesis), Searle seems to reactivate the classical ‘antithesis of outer and inner’. He would then renew the fundamental theoretical distinction at the origin of the Cartesian ‘official doctrine’ (in Ryle’s terminology) about the mental and its relationship with public events occurring in the objective, observable world. In that particular respect, Searle’s attempt to escape from the Cartesian vocabulary and categories does not appear to be fully achieved. One may even state the hypothesis that Searle’s approach to the mind is de facto built upon Cartesian concepts, such as first-person mental ontology, the interiority of mental phenomena, and the cerebral localization of the mind.

3. Searle’s Implicit Cartesianism: Internalism, mentalism, neuropsychology Searle’s specific perspective, and his criticism of materialism and cognitive science, involves the definition of the mind through the categories of consciousness, subjectivity and first-person ontology. Those are supposed to constitute, together with qualitativeness, essential features of the mind. They could not be denied or neglected as such, except by philosophers whose elaborate far-fetched theories would obliterate the common sense, ‘obvious’ and fundamental experience that we have qualitative, inner mental states. For example, one may read in The Rediscovery of the Mind that ‘mental states are always somebody’s mental states’, so that one must admit that ‘there is always a “first-person”, an “I”, that has these mental states’.39 But such an axiom, as a matter of fact, is fully concordant with the Cartesian conceptualization of the cogito, and the assignation of thought to a certain subject in which mental processes occur. The Cartesian ‘discovery’, in the thought experiment of radical doubt set up in the Méditations is precisely bound to the view that all the objects of thought without exception can be considered as false (the hypothesis of the ‘Malin génie’, or evil demon). It consequently entails that every kind of mental activity, even though it would represent to us fictitious objects, necessarily requires an I as its condition and principle, an I whose existence escapes any doubt, but reveals itself in the very experience of hyperbolical doubt. The first-person ontology is undoubtedly involved in the Cartesian definition of the mind as a res cogitans, a thinking ‘thing’.

116

Postanalytic and Metacontinental

Moreover, Searle seems almost paradoxically to acknowledge some debt to Descartes’ conceptualization of the mind when he writes that according to Descartes ‘the essence of mind is consciousness, or as he called it “thinking”’.40 Indeed Descartes, at the heart of the seventeenth century, propounds to subsume all the various modalities of thought (rational thought, desire, will, sensation, etc.) under the general category of immediate inner knowledge (conscientia), by the I, of everything that occurs inside of the I. One can read, in the Second Set of Replies: ‘I use this term [Thought] to include everything that is within us in such a way that we are immediately aware of it (immediate conscii)’;41 or even, in the Principles of Philosophy: ‘By this term “thought”, I understand everything which we are aware of as happening within us, insofar as we have awareness of it (in nobis conscientia est)’.42 Such a conscientia, which literally defines in the Cartesian view the essence of all forms of mental activity, is remarkably close to the contemporary use of the term consciousness, even though Descartes does not use the term ‘conscience’ in the French versions of his writings. Be that as it may, Searle’s definition of thought in general sounds deeply marked by the original Cartesian definition just mentioned: ‘In one way or another, all other central notions – such as intentionality, subjectivity, mental causation, intelligence, etc. – can only be fully understood as mental by way of their relations to consciousness.’43 One may even consider that such a postulate about the essence of thought is implied in Searle’s critique of the traditional distinction, concerning mental states, between qualia and propositional attitudes: the limitation of qualia to a determined category of mental states would entail the fallacious claim, Searle argues, that all mental states are not necessarily qualitative and conscious ones. Searle’s demonstration of the conceptual equivalence between consciousness and mental activity in general, in relation to the obviousness of the existence of the conscious I, sounds identical to the Cartesian argument, that might be called the ‘it seems to me that’ (‘il me semble que’) argument. Descartes’ general conceptualization of the nature of mental acts, establishes the essential features of thinking, subjectivity and qualitativeness by mentioning a specific nondistinction between appearance and reality in the case of consciousness. This conceptualization can be found in the Principles of Philosophy, when Descartes claims that the subjective experience of seeing or walking, insofar as seeing or walking consist in ‘bodily activities’, cannot prove the existence of the I; but that they can certainly do so if we take ‘seeing’ or ‘walking’ to be the actual sense or awareness of seeing or walking, ‘since it relates to the mind, which alone has the sensation or thought that it is seeing or walking’.44 The ‘it seems to me that’ argument was already presented in the second Meditation, and responded to the need for a demonstration of the necessity to include sensation within the general frame of mental activity. Thus, according to Descartes, sensory perceptions like ‘seeing light’, ‘hearing a noise’, or ‘feeling heat’ might be reputed false when related to the ‘bodily things’ they express, as shown by the hypothesis of a continuous sleep; yet ‘I certainly seem to see, to hear, and

Cartesian Echoes in the Philosophy of Mind

117

to be warmed. This cannot be false; what is called “having a sensory perception” is strictly just this, and in this restricted sense of the term it is simply thinking.’45 Searle’s argument in favour of the irreducibility of consciousness is also grounded, in turn, on this claim for a non-distinction between appearance and reality and literally involves the ‘it seems to me that’ formula: Searle writes ‘where consciousness is concerned . . . the appearance is the reality. (. . .) if it consciously seems to me that I am conscious, then I am conscious. I can make all sorts of mistakes about the contents of my conscious states, but not in the way about their existence.’46 Of course, a difference would nevertheless remain: Searle’s analysis is opposed to the identification of consciousness and subjectivity to any kind of ‘inner introspection’. But, following Georges Canguilhem’s famous analysis about the historical misinterpretation concerning the original signification of the cogito, which would not be a psychological one, it is very doubtful even that the ‘inner introspection’ model should be imputed to Descartes himself.47 In the end, even though Searle claims a direct realism, against the classical (Cartesian) theory of perception, that is the ‘representative theory of perception’, his general position remains fundamentally internalist and mentalist. Interiority, as well as consciousness, is an essential feature of the mind, according to Searle, who constantly maintains the conceptual equivalence between mental states and inner states. Such an internalist notion of the mind, as we noticed, is at stake in Searle’s critique of cognitive science; it is implied by the distinction between semantics and syntax in the Chinese room argument. It is also postulated in the very use of the ‘Zombie argument’, which supposes that an individual might present all the signs of a human, normal, observable behaviour, and yet at the same time be deprived of any mental states. This contemporary renewal of the Cartesian hypothesis of the ‘hommes-machines’, that one may already read at the beginning of the Treatise on Man, makes sense only if mental states are identified with private, unobservable states, and thus supposes the privileged access epistemology. Indeed, one of Searle’s thought experiments, the ‘Silicon Brains’, in The Rediscovery of the Mind, entails that ‘it is perfectly conceivable for you to imagine that your external behaviour remains the same, but that your internal conscious thought processes gradually shrink to zero. From the outside, it seems to observers that you are just fine, but from the inside, you are gradually dying. In this case, we are imagining a situation where you are eventually mentally dead, where you have no conscious mental life whatever, but your externally observable behaviour remains the same.’48 Such a thought experiment seems widely indebted to Descartes’ supposition in the second Meditation, that he might see, when looking out of the window, not real human individuals walking in the street, but android automata: ‘Yet do I see any more than hats and coats which could conceal automatons?’.49 This hypothesis of ‘concealed automatons’ presupposes the general conceptual distinction between interiority and exteriority: it is noticeable for example in the Cartesian differentiation between interior actions, associated with mental events,

118

Postanalytic and Metacontinental

thus identified with specific human acts, and exterior actions, occurring in the observable, physical world and accomplished by machines and animals as well.50 In that respect, the constant use of the ‘internal-external’ vocabulary in Searle’s writings about the mind is quite significant: it indicates the silent persistence of the modern notion of mental interiority. The Cartesian apparatus of internalism, consequently, appears to be still at work in Searle’s perspective. As a matter of fact, the emphasis put on ‘the difficulties in the currently orthodox externalist accounts of intentional content’51 leads Searle to reject, for example, Putnam’s arguments against the current mentalistic view according to which meanings are just ‘in the head’.52 His criticisms of Putnam’s ‘Twin-Earth’ thought experiment entails disagreement with the externalist theory of signification, and with the assertion that ‘the extension of a term is not fixed by a concept that the individual speaker has in the head’, but that it is determined ‘socially’ (‘the linguistic division of labor’) and by the causal relations between the features of the world and the utterances about the world.53 But, if that is so then internalism and mentalism in philosophy of language and in philosophy of mind, appear to be more coherent than the externalist account. Nothing shows, Searle writes, that ‘meanings are not in the head’.54 The refusal of the causal theory of reference was already at work in Intentionality, through the critical examination of Putnam’s claim that ‘the speaker’s internal Intentional content is insufficient to determine what he is referring to, either in his thoughts or in his utterances’.55 Searle’s general theory of intentionality clearly claims that the contribution of society and the contribution of the real world are not sufficient to determine reference. But if it is not true that ‘meanings are not in the head’, it seems then difficult not to adopt a mentalistic view, in the theory of language and in philosophy of mind. Such a mentalistic claim, associated with an internalist approach of mental phenomena, appears also to stand at the core of Searle’s critique of Daniel Dennett’s program of explaining consciousness from the third-person point of view, and thus disqualifying qualia in consequence.56 Of course, internalism should not be separated, in Searle’s view, from the refusal of any kind of representative theory of perception: this refusal is at stake in the ‘direct’ or ‘naïve’ realism propounded in his recent writings. Direct realism may be understood as a consequence of the non-metaphysical approach to the mind through which Searle, since Intentionality, intends to escape from ‘traditional philosophy’ (a tradition that would be represented by Descartes, Locke, Hume, Berkeley, Kant and Hegel). Yet, paradoxically, intentionality is still defined under the general category of representation, and intentional, or representative contents; though they do not constitute shadowy intermediary entities between the mind and the actual objects of the world, they are still identified with inner states, ‘inside our heads’. But if intentional contents are sufficient to determine reference, insofar as the conditions of satisfaction of the intentional state are seen as internal to the intentional state, and if meanings just are in the head, then the internalist, intentionalistic tradition renewed by Searle appears to be

Cartesian Echoes in the Philosophy of Mind

119

strikingly close to the mentalistic one, which Putnam’s externalism and causal theory of reference precisely intended to dissolve. What we might call Searle’s hidden Cartesianism is not limited, nevertheless, to the categories inherited from the ‘res cogitans’ notion, in relation to the conception of an immaterial mind. It is even more striking (and more unexpected) insofar as ‘neuropsychology’ is concerned. By the term neuropsychology, I do not mean any kind of reductionist theory that would assert that mental states are nothing but brain states. Rather, in the wide meaning of this term, I mean the view that the human brain is the seat of the soul, and that there exist, at least in the case of sensation and perception, some systematic correlations between mental processes and the corresponding neurophysiological processes. Such a view is entirely compatible with Searle’s ‘biological naturalism’. But in a very strange way, he seems to deny to Descartes himself any insight into such a correspondence between events in the mind and events in the brain. For example, in The Rediscovery of the Mind, he asserts that the original Cartesian view simply ignores the role of cerebral processes for mental representations: criticizing the mind-computer model and the functionalist multiple realizability thesis, he writes that ‘Artificial intelligence and functionalism coalesced, and one of the most stunning aspects of this union was that it turned out that one can be a thoroughgoing materialist about the mind and still believe, with Descartes, that the brain does not really matter to the mind’.57 But is it that obvious that in the Cartesian approach to mental phenomena, ‘the brain does not really matter to the mind’? On a more precise examination, such an interpretation appears contradictory with Descartes’ conceptualization of mental processes, and more particularly of sensation and sensory perception, from the Treatise on Man to The Passions of the Soul. One can find, in the Treatise on Man, a first explicit theorization of the cerebral localization of the soul: ‘when God unites a rational soul to this machine [the human body] he will place its principal seat in the brain’.58. Indeed the Cartesian ‘solution’ to the mind-body problem as shown in the sixth Meditation, that is the understanding of psychophysical unity, requires such a ‘neuropsychological’ postulate. This postulate is also required, from a scientific point of view, for the Cartesian account of visual perception, based upon the modern mechanical philosophy. Thus, according to the new science of vision propounded in the Optics (1637): ‘We know for certain that it is the soul which has sensory perceptions, and not the body (. . .). And we know that it is not, properly speaking, because of its presence in the parts of the body which function as organs of the external senses that the soul has sensory perceptions, but because of its presence in the brain, where it exercises the faculty called the “common” sense’.59 It is quite remarkable that in Descartes’ view, the refusal of a bodily subject of perception and representation, which entails that the soul itself, as a thinking thing, is the real subject of sensation, does not exclude, but on the contrary implies, the assignation of the soul to the cerebral sphere, together with the general representation of psychophysical interaction (sensation

120

Postanalytic and Metacontinental

and voluntary movement). If one now considers The Passions of the Soul, in which the conceptual apparatus of Cartesian neuropsychology is reinforced, with, among others, the famous notion of the ‘pineal gland’ (the most inner part in the brain identified with the seat of the soul), one is led to the same conclusion: ‘Let us therefore take it that the soul has its principal seat in the small gland located in the middle of the brain.’60 Strange as it may seem, the modern characterization of mind as the inner principle of thought, at stake in the Cartesian philosophy, cannot be separated from the modern neuropsychological view.61 In that respect, it is not so surprising that by reactivating Cartesian internalism, Searle should ipso facto (without admitting it in any circumstance) reassume, in a non-materialist and non-reductionist perspective, the main theoretical principles of the Cartesian claim about the brain being the seat of the soul. Indeed, the view in The Rediscovery of the Mind that inner introspection is not the ‘object’ of consciousness, yields the following statement: ‘it is easy to see that it [consciousness] is spatial, because it is located in the brain.’62 Such a statement does not sound ultimately anti-Cartesian, since in Descartes’ system, the hypothesis of an immaterial, unextended mind coexisted with the representation of the localization of the soul in the brain, namely in the innermost part of the brain; this representation, as we have seen, appeared necessary for the account of the causal action of the mind upon the body (voluntary movement) and the causal action of the body upon the mind (sensation, sensory experience). In The Rediscovery of the Mind, it is again asserted that ‘qualitativeness, subjectivity, and intentionality’, as features of the mental, ‘are located in the space of the brain at certain periods of time, causally explicable by lower-level processes and capable of acting causally’. In other words, according to Searle, ‘all my mental life occurs in the space of my brain, is caused by microprocesses there, and acts causally from there.’63 But would Descartes have refused such a cerebral ‘space’ of mental events, in the name of his definition of the mind as a non-material, that is non-extended substance? On the contrary, as the correspondence with Elisabeth shows (1643), he might have accepted such a paradoxical ‘spatialization’ of the soul. When Elisabeth asks him how it was conceivable that the soul, an immaterial thing, could yet determinate the movement of the body, in the case of voluntary actions, Descartes gives this astonishing answer: ‘Your Highness observes that it is easier to attribute matter and extension to the soul than to attribute to it the capacity to move and be moved by the body without having such matter and extension. I beg her to feel free to attribute this matter and extension to the soul because it is simply to conceive it as united to the body.’64 The argumentative context of such a remarkable attribution, by Descartes, of an ‘extension’ to the soul is precisely defined by interactionism, and the need for an account for mental causation, as well as for sensory representations. It must be noticed that it is a similar argumentative context, and the will to refute epiphenomenalism, among other things, that lead Searle to the claim that mind should be regarded as spatial, that is located in the brain. Searle’s subterranean revival of Descartes’ representation of the ‘seat of the soul’,

Cartesian Echoes in the Philosophy of Mind

121

consequently, goes with the revival of Cartesian interactionism. As a matter of fact, in his examination of the Cartesian hypothesis of the ‘pineal gland’, identified with ‘the point of contact between the mental and the physical’, Searle even recognizes, against the current views on this point, that such an hypothesis ‘is not as crazy as it sounds’.65 To conclude, I would like to focus very briefly on the critique of the notion of unconscious which occupies a whole chapter of The Rediscovery of the Mind. In chapter 7, Searle rejects two definitions of the unconscious, the functionalist one, and the Freudian one, to retain only the weak version of the unconscious, that is an unconscious reduced to some potential consciousness. First of all, Searle rejects the idea that there could be, as Chomsky asserted for example, deeply unconscious rules of syntax, inaccessible to consciousness. His main argument, linked to his theory according to which ‘only intrinsic intentionality is genuinely mental ’, is called the ‘connection principle’: it is the idea that all unconscious states, being intrinsically intentional, ‘are in principle accessible to consciousness’.66 But such a view does not entail only the invalidation of the computational syntactical definition of mental activity, which separates intentionality from consciousness. It is also and above all, one might say, directed against the Freudian claim that the unconscious, in its dynamic sense, is the essence of psychical reality. The first topic, as it is well-known, distinguished between two meanings of the term ‘unconscious’, that is the preconscious, which involved the representations that could become conscious, and the unconscious as system, which involved representations that could not, as such, have access to consciousness. The latter version of the unconscious, the dynamic unconscious, led Freud, in the Traumdeutung, to contest the classical hypothesis, both philosophical and psychological, of an essential equivalence between consciousness and psychical life. Searle, through his refutation of the Freudian perspective, seems in his turn to reactivate, inadvertently or not, such a classical hypothesis about the mind. ‘If Descartes had not already destroyed the meaning of the sentence, we could say “the essence of mind is consciousness”’.67 Indeed, it is exactly this supposed essence that Searle asserts, following the original view propounded in the Meditations on First Philosophy. The identity statement between mental activity and consciousness, at the very heart of Searle’s ‘rediscovery of the mind’ and critique of cognitive reason, appears to be inseparable from the Cartesian move in the history of philosophy, namely the modern conceptualization of the mind as thinking subject.

Notes 1

2

Hilary Putnam, The Threefold Cord. Mind, Body, and World (New York: Columbia University Press, 1999). The Chinese room argument was first presented by Searle in a paper called ‘Minds, Brains, and Programs’, Behavioral and Brain Sciences 3 (1980), 417–424. Also see John Searle, The Rediscovery of the Mind (Cambridge, MA: MIT Press, 1992), 45.

122 3

4 5

6 7 8

9

10

11 12 13

14 15 16

17 18 19 20 21

22

Postanalytic and Metacontinental

See in particular Hilary Putnam, ‘Minds and machines’, in Mind, Language and Reality. Philosophical Papers, vol. 2 (Cambridge: Cambridge University Press, 1975), 362–385. Searle, The Rediscovery of the Mind, 45. Cf. in particular Ibid., 111–126. See also John Searle, Mind. A Brief Introduction (Oxford: Oxford University Press, 2004), 113–115; John Searle, The Mystery of Consciousness (New York: New York Review of Books, 1997). Searle, The Rediscovery of the Mind, 1. Searle, The Rediscovery of the Mind, 158. ‘Consciousness, in short, is a biological feature of human and certain animal brains. It is caused by neurobiological processes and is as much part of the natural biological order as any other biological features such as photosynthesis, digestion, or mitosis’ (Searle, The Rediscovery of the Mind, 90). The very title of the Méditations métaphysiques, in the french version of 1647, is particularly explicit in that respect: ‘Méditations métaphysiques (. . .) dans lesquelles l’existence de Dieu, et la distinction réelle entre l’âme et le corps de l’homme, sont démontrées’, (René Descartes, Oeuvres de Descartes, Charles Adam and Paul Tannery eds (Paris: Vrin, 11 volumes, 1996), vol. 9, 13; René Descartes, The Philosophical Writings of Descartes, trans. by John Cottingham, Robert Stoothoff, Dugald Murdoch, Anthony Kenny (New York: Cambridge University Press, 3 volumes, 1991), vol. 2, 12. Descartes, Second Set of Replies, in René Descartes, Oeuvres de Descartes, vol. 2, 119–120. Searle, The Rediscovery of the Mind, 85. Searle, Mind. A Brief Introduction, 41–43. Searle mentions ‘the bankruptcy of the Cartesian tradition, and the absurdity of supposing that there are two kinds of substances or properties in the world, “mental” and “physical”‘ (Searle, The Rediscovery of the Mind, 13). Searle, The Rediscovery of the Mind, 15. Searle, Mind. A Brief Introduction, 13–27. In that respect, Searle seems to share the point of view expressed by one of his theoretical adversaries, Daniel Dennett, who explains that he first took interest in the issue of the relation between mind and body when he read Descartes’ Meditations (Daniel Dennett, preface to Consciousness Explained (Boston: Little, Brown and Company, 1991). Searle, Mind. A Brief Introduction, 13. Searle, The Rediscovery of the Mind, 15. Ibid., 26. Ibid., 54. See in particular Thomas Nagel, The View From Nowhere (Oxford: Oxford University Press, 1986). In this book, Nagel reactivates explicitly the Spinozistic thesis according to which the mental and the physical do not constitute two different substances, but are the distinct attributes of a unique substance that would be per se neither mental nor physical. Cf. Donald Davidson, Essays on Actions and Events (Oxford: Clarendon, 2001). The denial of any strict psychophysical laws is inferred by Davidson from the non-nomological character of the mental, its holistic nature. In that respect, one has to postulate the explanatory irreducibility of mental events.

Cartesian Echoes in the Philosophy of Mind 23 24

25

26 27 28 29

30 31 32 33 34 35 36 37

38 39 40 41

42

43 44 45

46

47

48 49

50

51

123

Searle, Mind. A Brief Introduction, 45–46. Cf. in particular ‘Why I am not a property dualist’, in John Searle, Philosophy in a New Century: Selected Essays (Cambridge: Cambridge University Press, 2008), 152–160. ‘I am now convinced, after several years of discussing these issues, that with very few exceptions all of the parties to the disputes in the current issues in the philosophy of mind are captives of a certain set of verbal categories. They are the prisoners of a certain terminology, a terminology that goes back at least to Descartes if not before’ (Searle, The Rediscovery of the Mind, 31). Gilbert Ryle, The Concept of Mind (London: Hutchinson, 1949). Searle, Mind. A Brief Introduction, 116. Ibid., 2. Cf. more particularly ‘La notion de conscience’ and ‘Does “Consciousness” Exist?’, in William James, Essays in Radical Empiricism (London: Longmans, Green, and Co., 1912). Searle, Mind. A Brief Introduction, 108–109. Ibid., 22–23. Ibid., 273–277. Putnam, The Threefold Cord, 101. Searle, Mind: A Brief Introduction, 178–192. Ibid., 46. See Searle, The Mystery of Consciousness, ch. 5. John Searle, Intentionality. An Essay in the Philosophy of Mind (Cambridge: Cambridge University Press, 1983), 197–208. Searle, The Rediscovery of the Mind, 84. Ibid., 20. Searle, Mind: A Brief Introduction, 13. René Descartes, Second Set of Replies, in The Philosophical Writings of Descartes, vol. 2, 113 (Oeuvres de Descartes, vol. 9, 124 for the text in French; vol. 7, 160 for the text in Latin). Descartes, Principles of Philosophy, I, 9, in The Philosophical Writings of Descartes, vol. 1, 195. (Oeuvres de Descartes, vol. 9, 28 for the text in French; vol. 8, 7 for the text in Latin). Searle, The Rediscovery of the Mind, 84. The French and Latin versions are even more eloquent in that respect. Descartes, Meditations on First Philosophy, II, in The Philosophical Writings of Descartes, vol. 2, 19 (Oeuvres de Descartes, vol. 9, 23 for the text in French). Searle, Mind: A Brief Introduction, 122. See also Searle, The Rediscovery of the Mind, 121–122. Cf. Georges Canguilhem, ‘Qu’est-ce que la psychologie?’, Etudes d’histoire et de philosophie des sciences concernant les vivants et la vie [1968]. (Paris: Vrin, 1994), 365–381. Searle, The Rediscovery of the Mind, 66–67. Descartes, Meditations on First Philosophy, II, in The Philosophical Writings of Descartes, vol. 2, 21. See in particular The Correspondence, To Reneri for Pollot, April or May 1638, in The Philosophical Writings of Descartes, vol. 3, 99–100. See also To the Marquess of Newcastle, 23 November 1646, 303. Searle, Mind: A Brief Introduction, 192.

124 52

53 54 55 56 57 58 59

60

61

62 63 64

65 66 67

Postanalytic and Metacontinental

Cf. Hilary Putnam, ‘The Meaning of “Meaning”’, Mind, Language and Reality: Philosophical Papers, vol. 2, 215–271. Putnam, ‘The Meaning of “Meaning”’, 245. Searle, Mind: A Brief Introduction, 187. Searle, Intentionality, 199. Cf. Searle, The Mystery of Consciousness, ch. 5. Searle, The Rediscovery of the Mind, 44. Descartes, Treatise on Man, in The Philosophical Writings of Descartes, vol. 1, 102. Descartes, Optics, Discourse IV, in The Philosophical Writings of Descartes, vol. 1, 164. Descartes, The Passions of the Soul, 34, in The Philosophical Writings of Descartes, vol. 1, 341. Concerning Descartes’ ‘neuropsychology’, cf. in particular Pascale Gillot, L’esprit. Figures classiques et contemporaines (Paris: CNRS Editions, 2007), 40–50. Searle, The Rediscovery of the Mind, 105. Searle, Mind: A Brief Introduction, 117–118. Descartes, The Correspondence, To Princess Elisabeth, 28 June 1643, in The Philosophical Writings of Descartes, vol. 3, 228. Searle, Mind: A Brief Introduction, 33–34. Searle, The Rediscovery of the Mind, 153–156. Searle, Mind: A Brief Introduction, 158.

Chapter 8

When the Twain Meet: Could the Study of Mind be a Meeting of Minds? Massimiliano Cappuccio and Michael Wheeler1

1. Route 666 What happens when (so-called) analytic philosophy is brought face to face with (so-called) continental philosophy? In its present form, this question is too nebulous to deserve serious attention, since any answer that achieves more than unhelpful over-generalization will depend on which particular lines of thought within these two great philosophical traditions are selected to represent them. So we need to replace our opening question with one that limits the range of views under consideration, and is correspondingly more tractable. Here goes. What happens when scientifically oriented philosophy of mind in the analytic tradition is brought face to face with phenomenological thought in the continental tradition? Given that both these modes of inquiry seek to investigate human experience, one might reasonably think that there must be some sort of overlap between them. So what happens when they are brought together? Do we end up with a meeting of minds or a clashing of heads? Is it a case of friends reunited or of mortal combat? Is there any prospect that the interactions in question will provide the platform for a productive overcoming of the apparent divisions that are standardly thought to separate the two traditions? These are the programmatic questions that ultimately concern us in this chapter. To anchor our discussion in a concrete issue, we propose to address them by way of a significant challenge that confronts research in artificial intelligence (AI) and cognitive science. The uninitiated might think of such research as paradigmatically relevant to scientifically oriented philosophy of mind in the analytic tradition, but of little interest to continental phenomenology – except perhaps as an example of the sort of dehumanizing technological thinking that ought to be criticized. The uninitiated are in for a shock. To bring our concrete issue into view, consider the following questions. In the sort of dynamically changing and open-ended world in which, most of the time, human beings think and act, how is it that we are able to home in on just those aspects of all the things we sense, know or believe are relevant in some current context of activity, while ignoring everything that is contextually irrelevant? And how are we able to revise our beliefs and/or perform actions in

126

Postanalytic and Metacontinental

a contextually appropriate manner? Philosophers of cognitive science will recognize these questions as placing us on the road to what is known in the business as the frame problem, the problem of explaining how human beings think and act in ways that are fluidly and flexibly sensitive to context-dependent relevance.2 It has been argued by more than one notable authority that the frame problem is a sure sign that AI and cognitive science are bound to fail. If this is right, then the route to the frame problem is not just any old road. For researchers in AI and cognitive science, and for any philosopher who rides shotgun with these sciences, it’s the highway to hell. As we shall see, the harbingers of doom may be overly pessimistic. Nevertheless, it remains fair (in our view) to say that the frame problem constitutes a significant barrier to the more ambitious research programs in AI and cognitive science. At the same time, and highly germane to our present concerns, it has emerged as an unusually fertile locus for cross-tradition philosophical interaction. Our first task, then, is to understand the nature of the frame problem and its relationship with the two philosophical traditions.

2. Representational machines But why, we hear you wondering, is the phenomenon of context-dependent relevance-sensitivity – henceforth CDRS – such a problem for cognitive science and its philosophical bedfellows? Here it is useful to begin by saying something about how a scientifically oriented analytic philosopher of mind under the influence of cognitive science and AI will be disposed to think about human beings. The first crucial idea, which may be conceived as a methodological assumption adopted in order to provide an account of human intelligence consistent with the epistemological framework of cognitive science, is that human intelligence can be analysed into a set of functions that, ultimately, is governed by nothing more than a system of physical machines.3 Philosophically speaking, one might think of this idea as resulting from a rejection of Cartesian substance dualism combined at the same time with a preservation of Descartes’ understanding of the human body, including its brain, as a complex physical machine. If, in some sense, the mind is to be identified with the body, or a certain special part of it, and if the body is functionally equivalent to a complex physical machine, then the mind too, it seems, will have a wholly mechanistic explanation. Now, of course, mechanistic phenomena in general don’t automatically realize CDRS. So if the human mind can be explained mechanisitically, its capacity for CDRS will be explained by the fact that it functions like a machine of a certain sort, a machine with a certain organization. This ushers in a second and equally crucial idea. There is a view, awash with philosophical and scientific respectability, according to which, whatever else may be true of human beings qua complex physical machines, it simply must be

Could the Study of Mind be a Meeting of Minds?

127

true that we work by building, storing and manipulating inner representations.4 One generic argument goes something like this. Intelligence (including CDRS) is fundamentally a matter of sensitivity to information, and what explains how certain physical systems may realize such informational sensitivity is the capacity of those systems to build, store and manipulate physically realized (for which read neurally realized) internal representations. One might justifiably think of this as another way in which contemporary thinking about the mind continues to be haunted by Descartes’ ghost. At the moment when the hyperbolic doubt runs aground on the Cogito, the Cartesian mind is temporarily abandoned without a world. Certain cognitive structures – Descartes calls them ideas, we can call them representations – serve as the vehicles by which reliable epistemic contact between mind and world is re-established. Thus God’s role, as Descartes conceives it, is to ensure that, for the most part, our inner representations line up accurately enough with the world. With the subsequent rejection of substance dualism by naturalistic analytic philosophy, the immaterial Cartesian mind was given physical form, but the method by which it gained proper and useful epistemic contact with its world (by which it achieved informational sensitivity) remained unaffected. Mind is, and perhaps must be, a representational system. At least that (usually minus the history lesson) is how it has seemed to most scientifically oriented analytic philosophers of mind. With representationalism on the table, the problem of accounting mechanistically for CDRS now looks like this: how is a representational machine able to find, interpret, deploy and update appropriately just those representations that are relevant right now, in the present situation? It seems clear that any such machine capable of achieving human-level performance would need to contain a vast sea of information stored in a huge number of inner representational structures; so, in a dynamic and changing environment, a sequence of unguided global searches of its entire internal informational storehouse would surely fail in general to produce the sorts of timely and appropriate responses that CDRS requires. The obvious move here is to equip the machine with internally stored relevancy heuristics (represented rules of thumb) or second-order representations of context that determine which of its first-order representations are relevant, under particular interpretations, in a particular context. But although this might seem like a promising strategy for the representationalist to pursue, all it does is push the real problem one stage back. For how does the system decide which of its stored heuristics or potentially context-specifying representations are relevant? Another, higher-order set of heuristics or representations would seem to be required. But of course the same issue will re-emerge at that higher level. As Hubert Dreyfus once concluded, ‘if each context can be recognized only in terms of features selected as relevant and interpreted in terms of a broader context, the AI worker is faced with a regress of contexts.’5 This regress is one signature of the frame problem.

128

Postanalytic and Metacontinental

3. Phenomenology and the frame problem It is with Dreyfus that continental phenomenology enters the debate over the frame problem. Drawing on Heidegger and Merleau-Ponty, Dreyfus has argued that one source of the frame problem is the assumption that human intelligence is representation-driven. As he put it recently, ‘for Heidegger, all representational accounts are part of the problem.’6 Dreyfus’s account of the link between representation and the frame problem has three strands.7 1. Phenomenological analysis reveals CDRS to involve sensitivity within and between large-scale networks of semantic connections. Here Dreyfus is guided by Heidegger’s phenomenological analysis of the contexts that characterize everyday human activity. Heidegger argues that we ordinarily encounter entities as equipment, that is, as being for certain sorts of tasks (cooking, text-editing, baby bouncing and so on). He then introduces the concept of involvements to capture the significance of equipmental entities (the ways in which they are involved) in our everyday projects, activities and tasks. Each involvement is a relationally defined link in a web of significance, a totality of involvements. Thus someone might currently be working with a computer (an involvement that Heidegger calls a with-which), in the practical context of a study (an in-which), in order to write a letter of explanation (an in-order-to), which is aimed towards explaining the validity of her recent expenses claims (a towards-this), for the sake of safeguarding her political future, that is, for the sake of her being a politician (a for-the-sake-of-which). Crucially, such networks of equipmental significance typically spread out over vast regions of involvement-space through multiple and overlapping projects and activities. Contextual significance is thus massively holistic. According to Dreyfus, this is one reason why representational accounts of CDRS founder, because such massive holism renders the contextual structures in play resistant to specification in terms of determinate representational content. 2. Representations paradigmatically realize a form of knowledge-thatsomething-is-the-case. CDRS is a kind of knowing-how-to-navigate-one’s-world. Knowledge-how cannot be reduced to knowledge-that, so any attempt to representationally encode the knowledge that enables CDRS is bound to generate difficulties. Here Dreyfus is influenced by Heidegger’s notion of readiness-to-hand, and the associated context-sensitive practical engagement with equipmental entities that Heidegger characterizes as a skilful coping (more on which later), a knowing of one’s way around one’s world which cannot be reduced to a thematic knowledge-that things are thus and so. 3. Representations are intrinsically context-independent structures to which context-dependent significance must somehow be added. Because of this, any attempt to add significance by adding representations must result in a regress (see above), since those additional representations will themselves need to have significance added by yet further representations, and so on. Here Dreyfus makes use of Heidegger’s claim that adding value-predicates (representational

Could the Study of Mind be a Meeting of Minds?

129

structures specifying that an entity has a certain context-dependent property) to context-independent primitives (e.g. raw sense data or, to give the argument a more contemporary tone, light-intensity gradients at the retina) can never be the ultimate source of CDRS, since any further representational structures of this sort (at whatever order of application) will themselves stand in need of context-dependent interpretation. As Heidegger himself puts it: ‘Adding on value-predicates cannot tell us anything at all new about the Being of goods, but would merely presuppose again that goods have pure presence-at-hand [context-independence] as their kind of Being. Values would then be determinate characteristics which a thing possesses, and they would be present-at-hand.’8 If Dreyfus is right, the frame problem is an artefact of representationalism. This invites the following question: can we develop a nonrepresentational account of CDRS? Here Dreyfus draws on Merleau-Ponty’s (1962) notion of the intentional arc. According to Merleau-Ponty, as an agent acquires skills, those skills are ‘stored’, not as representations in the agent’s mind, but as the solicitations of situations in the world. What the learner acquires through experience is not represented at all but is presented to the learner as more and more finely discriminated situations. If the situation does not clearly solicit a single response or if the response does not produce a satisfactory result, the learner is led to further refine his discriminations, which, in turn, solicit ever more refined responses. For example, what we have learned from our experience of finding our way around in a city is ‘sedimented’ in how that city looks to us. Merleau-Ponty calls this feedback loop between the embodied coper and the perceptual world the intentional arc.9 To illustrate this idea, we can adapt an example from Shaun Gallagher.10 When poised to engage in the action of climbing a mountain, I do not build a representation of the mountain and infer from that plus additionally represented knowledge of my own abilities that it is climbable by me. Rather, from a certain distance, in particular visual conditions, the mountain ‘simply’ looks climbable to me. My climbing know-how is ‘sedimented’ in how the mountain looks to me. This dovetails with Heidegger’s claim that human being is characterized by thrownness. Human being is thrown in that each of us is in primary epistemic contact not with bare context-independent elements, but rather with equipment, the kind of entity that, because of the way its meaning is fixed by its relational position in networks of contextual significance, comes already laden with such significance (see above). Since human being is characterized by thrownness, each of us is always already embedded in some meaningful context, so we are never in the position of having to add contextual significance to context-independent primitives. So once we recognize that human agents are not representational machines, but the kind of nonrepresentational thrown beings revealed by phenomenological analysis, we discover not a solution to the

130

Postanalytic and Metacontinental

frame problem, but a dissolution of it. In short, for phenomenologists in the Heideggerian-Merleau-Pontian tradition, there is no frame problem.

4. Naturalism and beyond From what we have seen so far, phenomenological insights can provide us with a diagnosis of why the frame problem arises for a certain conceptualization of human intelligence, a conceptualization favoured by a scientifically oriented analytic philosophy of mind influenced by AI and cognitive science. Moreover, they can provide an alternative conceptualization of human intelligence given which the frame problem arguably doesn’t arise. So far, then, the intellectual traffic in this chapter has been pretty much one-way. In this section, however, we shall argue that it is by viewing phenomenological analysis through the lens of a certain species of analytic theorizing about cognitive science that continental phenomenology is brought into the most productive relationship with that science. Let’s begin by considering the relationship between scientifically oriented analytic philosophy of mind and empirical cognitive science. Such philosophy is standardly conceived as naturalistic, in the sense that it deliberately constrains its theorizing to be continuous with empirical science. So the strength of the naturalism here will be determined, in part, by how the term ‘continuous’ is interpreted. Perhaps the weakest possible reading is something like consistency with empirical science.11 But it might seem that even if we endow scientifically oriented analytic philosophy of mind with this weakest possible kind of naturalism, the result will still be a conflict with continental phenomenology. For in order to remain a form of naturalism, the analytic view in question must be committed to the principle that if and when there is an inconsistency between philosophical analysis and empirical science, it’s philosophy that must give way. Continental phenomenology is a theoretical enterprise that, through an attentive and sensitive examination of ordinary experience, aims to reveal the transcendental yet historical conditions that shape and structure that experience. That doesn’t sound much like a discipline in any mood to give way to what empirical science might have to say about its subject matter. For the purposes of this chapter, we want to remain neutral on whether or not there is room for phenomenological thought to adopt the weak form of naturalism just scouted. Our point, rather, is that one mark of a kind of postanalytic and metacontinental program in this area will be that it takes seriously both the disciplined phenomenological analysis of human experience and the rigorous science of human cognition, where to ‘take both seriously’ means to show how they may be integrated into an overall package that preserves the explanatory credentials of both modes of inquiry while allowing a critical interplay between them. One model for such a package may be developed out of John McDowell’s distinction between explanations at the personal level, and those that proceed

Could the Study of Mind be a Meeting of Minds?

131

subpersonally. According to McDowell, personal-level explanations are concerned with the identification and clarification of the constitutive character of agency (roughly, what it is to competently inhabit a world), whereas subpersonal explanations are concerned with mapping out the inner states and mechanisms (the parts of agents, as it were) that causally underpin personal-level phenomena. Continental phenomenology is a species of personal-level explanation. Cognitive science, by contrast, is a subpersonal enterprise. The two explanatory levels, and so the disciplines that explore them, are not reductively related (in particular, the personal level is not reducible to the subpersonal level), but neither are they wholly independent of each other. The process, rather, is one that McDowell describes as a ‘perfectly intelligible interplay’ between them.12 On the one hand, the ultimate goal of cognitive science is to map out the subpersonal elements whose organization, operation and interaction make it intelligible to us how it is that personal-level phenomena are modified or disrupted when mechanistic causal processes (such as those realized in brains) are similarly modified or disrupted. Given that what those personal-level phenomena are (their constitutive character) is a matter that may be pursued by philosophical reflection of one kind or another (including phenomenological analysis), there is a sense in which personal-level explanation (and thus phenomenology) isolates phenomena for which subpersonal science then tries to identify underlying causal mechanisms. On the other hand, what cognitive science discovers at the subpersonal level may sometimes lead us to revise our conception of what the personal-level phenomena are. For example, studies on the premotor cortex carried out by Rizzolatti’s group have strongly stimulated in an original way the phenomenological debate over motor intentionality and intersubjectivity.13 Experiments on mirror neurons suggest that a subject’s motor competences enrich and actively modulate her experience of perception, recognition and categorization of intentional actions, eventually providing the embodied experiential base for the pre-reflective understanding of the intentions of other agents. There are, of course, many details here to be settled, and many devils may await therein.14 What seems clear, however, is that our phenomenologically driven understanding of CDRS will be incomplete (it will be mere armchair imperialism) until we have a cognitive-scientific model of the causal mechanisms that render that phenomenon an intelligible aspect of the physical world. The McDowellian personal-subpersonal distinction gives us a powerful way of making space for that completion without risking reductive scientism. Interestingly, Dreyfus himself accepts the challenge to describe the underlying mechanisms that, in the wake of his phenomenological treatment of the frame problem, might causally account for CDRS.15 Here he appeals to the work of neuroscientist Walter Freeman. What emerges is a vision of the brain as a nonrepresentational dynamical system primed by past experience to actively pick up and enrich significance, a system whose constantly shifting attractor landscape is identified as physically grounding Merleau-Ponty’s intentional arc

132

Postanalytic and Metacontinental

by causally explaining how newly encountered significances change the whole perceptual world of the agent. The fine-grained details of this vision need not concern us. What is crucial in the present context is the fact that the sweeping anti-representationalism that we earlier attributed to Dreyfus at (what we can now call) the personal level is ultimately made intelligible (mechanistically speaking) by an equally sweeping anti-representationalism at the subpersonal level. As we are about to see, however, Dreyfus’s dual-level anti-representationalism needs to be qualified in an important respect. At the heart of things here is Dreyfus’s account of coping – the basic modality of intelligent interaction between agent and world that instantiates Merleau-Ponty’s intentional arc and thus explains CDRS. Crucially, Dreyfus distinguishes between two kinds of coping – skilful coping and background coping. In what remains of this chapter we shall interrogate Dreyfus’s phenomenologically driven anti-representationalism and the related distinction between skilful and background coping, and we shall do so within the theoretical space that we have endeavoured to open up via our McDowellian detour. In short, it is time for a brief exercise in postanalytic, metacontinental philosophy, as we understand it.

5. The idea that lived Heidegger distinguishes between (i) readiness-to-hand, the mode of being exhibited by manipulable entities during hitch-free skilled practical activity, in which agent and equipment become transparent within the phenomenological dynamics of the ongoing experience, and (ii) un-readiness-to-hand, the mode of being exhibited by entities when hitch-free skilled practical activity is disturbed, for example by broken or malfunctioning equipment. In his Heideggerian analysis of intelligence, Dreyfus treats both these modes of engagement as examples of skilful coping. It is un-ready-to-hand skilful coping that will largely concern us here. It is important to stress that although un-readiness-to-hand is standardly characterized as resulting from a disturbance to hitch-free activity, such disturbance need not result from an actual breakdown of some pre-established pattern of such activity. The key point is that the world announces itself to the agent by presenting itself as a practical problem to be solved by the agent. Equally important is that such practical problem solving need not involve anything like theoretical reflection. Indeed, when confronted with such disturbances, human agents typically exploit the same kind of practical pre-reflective involvement with entities as characterizes the domain of readiness-to-hand, and so, for the most part, continue to produce richly adaptive, fluid and flexible, real-time, context-sensitive embodied responses. Despite the continuity between hitch-free activity and practical problem-solving, there is a key difference: unlike hitch-free activity which is nonrepresentational in form, practical problem solving is plausibly governed by a minimally representational process. There has emerged a kind of cognitive distance between agent and

Could the Study of Mind be a Meeting of Minds?

133

environment. So, in order to interact appropriately with the context, the agent must produce, by means of its own actions, some sort of temporary ‘description’ of the environmental situation. Given the aforementioned continuity between un-readiness-to-hand and readiness-to-hand, however, such ‘descriptions’ will take on a particular form. Entities are not here disclosed as context-independent objects with activity-neutral properties (the mode of being that Heidegger called presence-at-hand). Because of this, the only environmental aspects encoded in our temporary ‘descriptions’ will be keyed to possibilities of actions, so the representations concerned will map the world in an operative format. They will also be ego-centred (linked to a situated perspective) and intrinsically context-dependent (essentially embedded in specific contingencies). Following previous usage, we shall call such structures action-oriented representations.16 To illustrate the properties that we have attributed to action-oriented representation, it is useful to ‘descend’ to the subpersonal level and look at the way such properties might be reflected in the mechanisms that enable un-ready-tohand problem solving. This also respects the logic of explanatory completion sketched above. Here, then, is an example that one of us has used a number of times before. Franceschini et al. set themselves the task of building a robot that navigates its way to a light source while avoiding obstacles.17 The resulting system achieves this goal by executing a sequence of movements, each of which is generated in the following way. A primary visual system, inspired, in part, by the compound eye of the fly, features a layer of elementary motion detectors (EMDs). Since these components are sensitive only to movement, the primary visual system is blind at rest. What happens, however, is that the EMD layer uses relative motion information, generated by the robot’s own bodily motion during the immediately preceding movement in the sequence, to build a temporary representation (a snap map) of detected obstacles, constructed using an egocentric coordinate system. Then, in an equally temporary second representation (a motor map), information concerning the angular bearings of those detected obstacles is fused with information concerning the angular bearing of the light source (supplied by a supplementary visual system) and a direction-heading for the next movement is generated. This heading is as close as possible to the one that would take the robot straight towards the light source, adjusted so that the robot avoids all detected obstacles. The representations used by this robot are thus keyed to possibilities of actions, in that objects (other than lights) are represented as avoidance-regions or motion-barriers; they are ego-centred, in that the snap map of detected obstacles features an agent-centred coordinate system, and the motor map exploits agent-based angular bearings; and they are intrinsically context-dependent, in that the explicit representation of context is eschewed in favour of special-purpose adaptive couplings that implicitly define the context of activity in their basic operating principles.18 Now for what, in the wake of Dreyfusian anti-representationalism, looks to be a controversial claim. In our view, the frame problem – or rather, as we shall see in a moment, one important dimension of it – may be neutralized by processes

134

Postanalytic and Metacontinental

that involve action-oriented representation. To bring this claim into proper view, we need to adopt a more nuanced understanding of the frame problem, one that depicts it as having the two-dimensional structure implied by our previous description of CDRS as involving sensitivity within and between large-scale holistic semantic networks. The first dimension of the frame problem, call it the intra-context dimension, concerns how AI/cognitive science might explain appropriate, flexible and fluid action within a context, action that is oriented towards the achievement of a specific goal. The second dimension, call it the inter-context dimension, concerns how AI/cognitive science might explain our ability to switch between contexts, flexibly and fluidly, in a relevance-sensitive manner, or to maintain the significance and adaptiveness of an action even though it is performed in a situation in which the boundaries of the context are not clearly determined.19 Given this distinction, we can restate our key claim as follows: although any processing strategy capable of disabling the inter-context dimension of the frame problem may well engage, or even require, nonrepresentational dynamical processes akin to those discussed by Dreyfus,20 the intracontext dimension may be neutralized by processes that involve action-oriented representation. To clear the way for this claim, here is our diagnosis of how the intra-context frame problem comes about. The intra-context frame problem arises when a goal-oriented action is modelled in the wrong way, when, that is, the action is conceived as a set of sensorimotor variations governed by a set of pre-programmed instructions defining (representing) the movements that enable the achievement of a certain goal. So modelled, the action (the organized bodily movement) is conceived as having a merely executing function, a consequence of a prior cognitive process of decision making. The properly cognitive process is the selection of the relevant movements. The purposeful meaning of the action is thus compressed into a specified set of attributes, and trapped within an abstract moment of cognitive elaboration. For this reason the intentional content motivating the action is detached from the execution of the movement and thus decontextualized from the ongoing practical activity. The intra-context frame problem then arises because it is impossible to specify in advance the criteria for selecting the right goal-oriented movements in accordance with the actual context. The intra-context frame problem may be dissolved if, in contrast to the model just described, the purposeful meaning of the movements is embodied in the sensorimotor process itself. This is possible if the motor function is understood as an active element integrated into the cognitive process, that is, if the cognitive process is considered to be (at least partially) carried out by means of an embodied sensorimotor function. On the model we propose, and as realized subpersonally by Franceschini et al.’s robot, skilful coping is embodied in the cognitive process employing action-oriented representations, and action-oriented representations are embodied in the action itself; they are generated within the sensorimotor process and are dynamically up-dated while the action takes place. In this way, action-oriented representations mediate a minimal process of practical problem

Could the Study of Mind be a Meeting of Minds?

135

solving, a basic form of decision making that is embodied in the action itself. In accordance with the continuity between readiness-to-hand and un-readinessto-hand, the purposeful meaning of the action derives from a close dynamical coupling between the embodied agent and its environment. The goal of the action should be understood as embedded in a contextual background that defines dynamically the relationship between body and contingencies. The special-purpose, direct coupling between agent and world determines a reciprocal modification of intelligence (of the actions) and intelligibility (of the objects as affordable). More on this reciprocity later. For the moment the key point is this: action-oriented representations are never decontextualized structures requiring contextual interpretation in order to be deployed in the generation of behaviour. Thus, and pace Dreyfus, the very fact that representations figure in a proposed explanation of CDRS does not usher in the frame problem. The representationalist lives to fight another day.

6. Another day As it happens, Dreyfus is happy to concede that un-ready-to-hand skilful coping involves action-oriented representations.21 This apparent concession to representationalism is not significant, however, since his view is that action-oriented representations are not themselves explanatorily efficacious with respect to CDRS. It is at this point that the second kind of coping mentioned above becomes relevant. For Dreyfus, background coping is a nonrepresentational knowledge of how to get around one’s world that is ontologically more basic than nonrepresentational ready-to-hand coping or minimally representational practical problem solving. It is at the level of background coping that, for example, the skilled climber’s know-how first opens up the world as a familiar place of climbable mountains. Here is Dreyfus himself, drawing once again on Heidegger during a critical response to Wheeler’s advocation of minimal representationalism22: . . . Heidegger says explicitly that our background being-in-the-world, which he also calls transcendence, does not involve representational intentionality, but, rather, makes intentionality possible . . . To be more exact, background coping is not a traditional kind of intentionality. Whereas the ready-to-hand has conditions of satisfaction, like hammering in the nail, background coping does not have conditions of satisfaction. What would it be to succeed or fail in finding one’s way around in the familiar world? The important point for Heidegger, but not for Wheeler, is that all coping, including unready-to-hand coping, takes place on the background of this basic nonrepresentational, holistic, absorbed, kind of intentionality, which Heidegger calls being-in-the-world.23 According to Dreyfus’s theory of motor intentionality, skilful coping involves goal-oriented actions, processes aimed at specific situated interactions between agents and entities. The purpose of skilful coping is to eliminate a pre-reflective

136

Postanalytic and Metacontinental

sense of distress that is established when environmental contingencies result in a non-perfect attunement between agent and environment. By contrast, background coping implicitly fixes the conditions that define the situatedness of the agent as being-in-the-world, that is the general structure of its way of belonging to a context by being situated in a network of equipmental relationships. The background thus defines implicitly the general preconditions that make coping possible, the general preconditions that, in other words, make it possible for an agent to maintain a relationship of dynamical coupling with its world. Such preconditions are the possibilities of reciprocal modification between environment and agent through perception and action, of how the agent can employ specific strategies for acting intelligently in a certain environment, and of how the environment can become intelligible for certain agents. In this way they determine the Merleau-Pontian intentional arc. Moreover, every background precondition of the sort just described exists only as a consequence of other previous preconditions, in such a way that the preconditions that comprise the background are infinite in number and unpredictable in advance. The background is thus an open-ended horizon which can appear only partially and indirectly.24 As just described, the background hosts the primary manifestation of the coping process by which the body responds to its environmental situation, finding a silent dynamical equilibrium with it. It is in the light of this recognition of primacy that Dreyfus’s objection to action-oriented representation now emerges. Even if minimally representational activity is part of skilful coping, the very fact that this is so presupposes the phenomenon of background coping as ‘an even more basic nonrepresentational holistic coping that allows copers to orient themselves in the world’.25 The practical problem-solving process distinctive of un-readiness-to-hand cannot stand alone, separated from the process of background coping out of which it emerges. This is a compelling picture. So what is wrong with it? As we see things, Dreyfus’s mistake is subtle but profound. In effect, he treats skilful coping and background coping as if they are two separate processes, in the sense of being distinct from each other in their patterns of intelligibility. Although such separability is consistent with the idea that background coping has absolute priority over skilful coping as a fundamental platform for the latter (dependency does not require inseparability), it fails to capture the way in which the phenomena in question show up as intelligible to us. In truth, background coping and skilful coping are (in Heidegger’s terms) equiprimordial, because the background and the structure of the action (with its target-object) are mutually defined and always appear together. This claim resonates with Merleau-Ponty’s analysis of motor intentionality (an analysis that, as we have seen, has been used by Dreyfus himself to ground his model of nonrepresentational intelligent bodily action). Merleau-Ponty stresses the fact that background and action are co-determined and co-evolving in the phenomenology of the living body. As he puts it, ‘movement and background are, in fact, only artificially separated stages of a unique totality.’26 Because of the reciprocal determination of background and action, the relationship between background coping and skilful coping is one of ontological

Could the Study of Mind be a Meeting of Minds?

137

co-dependence. The background is made intelligible as such only when it appears in contrast with a well-shaped intentional act, and is not otherwise recognizable in itself. In this sense the background of an action always carries the mark of the actions that make it intelligible. On the other hand, the action is already shaped by the preconditions defined by the background, and in this sense it always brings with it the mark of the specific background out of which it emerged. So although it is true that being-in-the-world (transcendence) is a background phenomenon that can’t be mapped in terms of representational processes, and although it is accepted here that representational intentionality is grounded in transcendence, while transcendence can’t be a matter of representational intentionality, we must nevertheless recognize that background and action can never be conceived as autonomous decoupled elements, and that neither of them is intelligible if it is considered alone. One graphic way to appreciate the intimate dynamical relationship between background and action that is evident in ordinary skilful coping is to see what happens when such intimacy is not in place. From the analyses of motor intentionality carried out by Merleau-Ponty in Phenomenology of Perception, it is evident that, in apraxic patients, the agent’s movement and environment are fundamentally decoupled (not dynamically related to each other): What [the patient] lacks is neither motility nor thought and we are brought to the recognition of something between movement as a third-person process and thought as a representation of movement – something which is an anticipation of, or arrival at, the objective and is ensured by the body itself as a motor power, a ‘motor project’, (Bewegungsentwurf), a ‘motor intentionality’ in the absence of which the order remains a dead letter. The patient either conceives the ideal formula for the movement, or else he launches his body into blind attempts to perform it, whereas for the normal person every movement is, indissolubly, movement and consciousness of movement.27 The analogy between apraxic patients and the kind of AI systems that confront the frame problem is striking. In both cases, the links between movement and environment are stereotyped, that is reduced to a fixed, finite objective frame. When this happens, the link between goal-oriented action and the background must be artificially reconstructed by the agent by means of sensorimotor spatial exploration, conceptual inferences or objective representations of an intellectualized space. The frame problem confronts both human apraxic patients and certain AI systems because, in both these cases, the dynamic, holistic coupling between movement and background has been lost. At this point Dreyfus would surely remind us that Merleau-Ponty’s account of motor intentionality, with its associated claim of background-action co-dependency, has an important nonrepresentational element. As Merleau-Ponty himself puts it, ‘[m]otility . . . is not, as it were, a handmaid of consciousness, transporting the body to that point in space of which we have formed a representation before-hand.28 But here it is important to note that one of the key features of

138

Postanalytic and Metacontinental

action-oriented representations, as we conceive them, is precisely that they do not comprise a set of movement instructions fixed in advance of the overall pattern of sensorimotor activity in which they are implicated, but rather are generated within that activity and are dynamically up-dated while it takes place. Thus they do not pre-select, and then transport the body to, a fixed point in space, in the way that Merleau-Ponty rightly finds objectionable. Indeed, what actionoriented representations support is precisely a productive failure to close down possibilities. This explains why, for example, within limits set by certain relative speeds of movement, the action-oriented representational target-tracking strategy adopted by Franceschini et al.’s robot (discussed earlier) generalizes from a static to a moving light source. On the view we endorse, then, the possibilities that define the background, all the while they remain ‘merely’ part of that background, are not themselves represented, either at the personal (phenomenological) or the subpersonal (mechanistic) level. Indeed, it is not even guaranteed that such possibilities will be represented at the point where they are plausibly thought of as being brought forth to govern behaviour in a concrete manner. That depends on the dynamics of the situation. As we have seen, whereas un-ready-to-hand skilful coping is representational in character, ready-to-hand skilful coping is not. Crucially, the critical properties that explain intra-context CDRS in the case of ready-to-hand coping (context-dependent dynamic structures, close agent-environment coupling) also explain it in the case of un-ready-to-hand coping, because the minimal representations implicated in the latter bear these very properties.29 The point we wish to make is that representations may be an important part of the process by which human beings achieve CDRS, not that they always must be. That said, it seems to us that the situation of complete, transparent, readiness-to-hand, in which the subject-object dichotomy disappears and representationalism has no grip, is something of an ideal state. Coping is never (or very rarely) perfectly smooth. Subjective activity produces a background noise which never really disappears. This keeps intact, to some extent, the distinction between subject and environment. So although background coping is nonrepresentational in nature (the possibilities that define the structure of coping are not represented), actual coping activity is always (or almost always) mixed to some degree with (partly, minimally) representational processes. We can find some clues to the structure of this phenomenon if we look at Heidegger’s discussion of skilful coping mediated by signs. The representations here (the signs) are of course external not internal in character, but the logic, we suggest, carries over. According to Heidegger, signs have the function of indicating; indicating is a particular form of referring; and referring, schematically, is a way of creating an operative relation. In this way, a sign is a representational entity devoted to modulating the practical interaction of an agent with its world-environment. One of Heidegger’s examples is the south wind.30 Considered as a sign, the south wind is a ready-to-hand entity that reveals a perspicuous relationship between the actual environmental situation and the phenomenon (foreseen)

Could the Study of Mind be a Meeting of Minds?

139

of its possible and incipient modification (the rain). When the sign pops up from the background, it represents a potential aspect or modification presupposed by the background itself, revealing that possibility in the form of an affordance available to the agent. What precisely is represented by the sign, and how? In advance of the agent approaching the south wind with a specific practical, operative or interpretative attitude, the affordance in question remains an element of the background. At this time the south wind is not intelligible as a free-standing (present-at-hand) entity. Moreover, it remains at the periphery of any process of skilful coping. As a potentiality, held by the background, for indicating a particular, available practical intervention, it awaits discovery. However, when the agent approaches the south wind with a certain practical perspective, the south wind becomes an object of interaction placed before her. The south wind then emerges as a very specific kind of equipment, one having the function of indicating (representing) the possibility of rain. But, as Heidegger puts it, as a sign it is also ‘an item of equipment which explicitly raises a totality of equipment into our circumspection so that together with it the worldly character of the ready-to-hand announces itself’.31 In other words, the sign represents not only the entity indicated by it, but also the contextual background that will eventually make a new practical interaction possible as skilful coping: standing there, before the farmer, the south wind doesn’t only remind him that it will rain soon; it also makes him aware that he must actively use his meteorological competences in order to cope with that change in the weather. Background coping grounds, prepares and actively shapes the representational process of skilful coping that will accompany it. Moreover, the modalities of intelligibility and of practical operability of the background are actively modulated by the minimally representational process introduced by the appearance of the sign. Incorporated into the background, the south wind is a nonrepresentational element of the holistic process of background coping that defines the situation of the farmer as being-in-the-world; but when it appears as a sign, it breaks the normal flow of absorbed coping and stands in front of the situated agent as a problem, a question mark requiring his attention and his capacity to recollect and employ his practical farming abilities. It is this specific transformation in the attitude of the agent that finally makes possible the appearance of the entity indicated by the sign as an object standing by itself as present-athand. That’s why, during skilful coping, a sign is capable of mediating between the background and the practical capabilities of the agent, establishing the possibility of a continuous reciprocal reversibility between background coping and skilful coping, by means of minimally representational activity. A sign is thus revealed as itself a kind of action-oriented-representation, providing a model for how background and representation are related during the sort of action-oriented internal representational activity that we discussed earlier. In view of all this, how should we understand the mechanistic subpersonal basis of background coping? What is crucial here is to resist any temptation to conceptualize background coping as requiring a distinct set of causal mechanisms,

Postanalytic and Metacontinental

140

separate from (a) the special-purpose adaptive couplings and action-oriented representational processes that (we have suggested) provide the causal-structural basis of intra-context sensitivity to relevance, and (b) the dynamical processes (whatever they may be) that underlie the kind of fluid and flexible contextswitching highlighted by the inter-context frame problem. Rather, background coping identifies sets of implicitly defined possibilities that are presupposed (transcendentally, one might say), by such mechanisms, if those mechanisms are to account adequately for CDRS and thus neutralize the frame problem. Indeed, to ignore this structural feature of any genuinely cognitive machine is to issue an invitation to the frame problem.

7. Towards a postanalytic, metacontinental philosophy To bring things to a close, we need to refocus our attention on programmatic issues. Perhaps surprisingly, scientifically oriented philosophy of mind has recently become the fertile interface for a productive dialogue between the analytic and continental traditions in philosophy. As we have shown via our consideration of CDRS and the frame problem, ideas from thinkers such as Heidegger and Merleau-Ponty may be used positively by cognitive scientists and their philosophical bedfellows. However, this engagement with continental thought is not mere uncritical appropriation. Rather, just as continental ideas may help scientifically oriented analytic philosophy of mind to think anew about its problems, so there is an intellectual current in the other direction, one that enables continental thought to get into a productive relationship with cognitive science. Perhaps the best way to describe this philosophical development is as a productive overcoming of what is often judged to be an irreconcilable opposition between two allegedly separate traditions of thought. If so, this opens the door to a genuinely new, postanalytic, metacontinental philosophy, a philosophy that, as we have encountered it here, and in just one of its potentially many forms, creatively combines a critical respect for human science with a disciplined analysis of human experience.

Notes 1

2

This chapter was prepared thanks to the support of a Caledonian Research Foundation/Royal Society of Edinburgh European Visiting Research Fellowship. Many thanks to the editors of this book for their useful comments on a previous version. The frame problem first emerged in AI as the problem of characterizing, using formal logic, those aspects of a state that are not changed by an action – see M. R. Genesereth, and N. J. Nilsson, The Logical Foundations of Artificial Intelligence (Los Altos: Morgan Kauffman, 1987). It soon turned out, however, that this original frame problem was just one dimension of a more general and multifaceted

Could the Study of Mind be a Meeting of Minds?

3

4

5

6

7

8

9 10

11 12

13

141

difficulty. As a result, in philosophical circles at least, the term ‘frame problem’ has taken on a wider meaning. It is that wider meaning to which we appeal here. The appropriateness of the term ‘frame problem’ is perhaps best illuminated by Fodor’s remark that what we confront is ‘the problem of putting a “frame” around the set of beliefs that may need to be revised in the light of specified newly available information’. See Jerry Fodor, The Modularity of Mind (Cambridge, MA: MIT Press, 1983), 112–113. We describe this appeal to functional explanation as a methodological commitment adopted to achieve consistency with the epistemological framework of cognitive science, in order to respect the fact that empirical cognitive science is not committed to (what we might call) metaphysical or constitutive functionalism, the view that what makes something a mental state of a particular type is the function which that state performs within the system of which it is a part. Thus, to put the point using language that we introduce later in this chapter, one might be an anti-functionalist at the personal level while remaining a functionalist at the subpersonal level. For the purposes of this chapter we shall adopt a very general understanding of the concept of representation. To engage in representational explanation is thus to explain thought and behaviour by invoking internal states with content (i.e. that encode meaning or bear information), states whose role is to act as inner ‘stand-ins’ for (usually external) objects and situations. Thus we do not assume that representations must, or even typically, possess certain properties that are sometimes (especially in cognitive science) linked with the concept (e.g. passivity, duration, decoupleability from the object or state of affairs that is being represented, combinatorial structure, etc.). Hubert Dreyfus, What Computers Still Can’t Do: A Critique of Artificial Reason. (Cambridge, MA: MIT Press, 1992), 288–289. Hubert Dreyfus, ‘Why Heideggerian AI failed and how fixing it would require making it more Heideggerian’, in The Mechanical Mind in History, P. Husbands, O. Holland, and M. Wheeler eds (Cambridge, MA: MIT Press, 2008), 331–371. Hubert Dreyfus, Being-in-the-World: A Commentary on Heidegger’s Being and Time, Division I (Cambridge, MA: MIT Press, 1990). For analysis and discussion, see Michael Wheeler, ‘Cognition in context: Phenomenology, Situated Robotics and the Frame Problem’. International Journal of Philosophical Studies, 16 No. 3 (2008): 323–349, and Michael Wheeler, ‘The Problem of Representation’, in The Handbook of Phenomenology and Cognitive Science, S. Gallagher and D. Schmicking eds (Berlin: Springer, 2009), chapter 18. Martin Heidegger, Being and Time, trans. J. Macquarrie and E. Robinson (Oxford: Basil Blackwell, 1927), 111–112. For discussion of this argument, see Dreyfus, Being-in-the-World, and Michael Wheeler, Reconstructing The Cognitive World: The Next Step (Cambridge, MA: MIT Press, 2005). Hubert Dreyfus, ‘Why Heideggerian AI failed’, 340. Shaun Gallagher, ‘Are minimal representations still representations?’ International Journal of Philosophical Studies, 16 No. 3 (2008): 351–369. Michael Wheeler, Reconstructing the Cognitive World. John McDowell, ‘The content of perceptual experience’. The Philosophical Quarterly 44 No. 175 (1994): 197. G. Rizzolatti, and C. Sinigaglia, ‘Mirror Neurons and Motor Intentionality’. Functional Neurology, 22 No. 4 (2007): 205–210. For a development in the light of

142

14

15 16

17

18

19

20 21 22 23 24

25 26

27 28 29

30 31

Postanalytic and Metacontinental

Husserlian and Merleau-Pontian phenomenologies, see H. De Preester, ‘From Ego to Alter Ego: Husserl, Merleau-Ponty and a layered approach to intersubjectivity’. Phenomenology and the Cognitive Sciences, 7 No. 1 (2008): 133–142. For example, it might be thought that any sort of non-reductive dual level approach falls prey to some sort of overdetermination problem. Thus if both levels are thought of as providing sufficient causal explanations of the same explanandum (say, an action), then that explanandum will be causally overdetermined. However, this problem is not inevitable. For example, and most radically, there will be no overdetermination if the relevant explanations at the personal level are conceived non-causally. One might think that phenomenological explanations designed to reveal the transcendental yet historical conditions for experience will fall into such a category. Thanks to James Chase for raising this issue and to Adrian Haddock for discussion. Hubert Dreyfus, ‘Why Heideggerian AI failed’, 347–357. See Michael Wheeler, Reconstructing The Cognitive World. For a prior use of the term that is closely related although it is not derived from phenomenological considerations, see Andy Clark, Being There: Putting Brain, Body, And World Together Again (Cambridge, Mass.: MIT Press. 1997). N. Franceschini, J. M. Pichon, and C. Blanes, ‘From insect vision to robot vision’. Philosophical Transactions of the Royal Society, series B. 337 (1992): 283–294. To be clear, our claim here is not that Franceschini et al.’s robot has conscious experience. The sorts of subpersonal structures it instantiates may be necessary for there to be un-readiness-to-hand, but they are not sufficient. For this analysis of the frame problem, see Michael Wheeler, ‘Cognition in context’. For some worries, see E. Rietveld, ‘Context-switching and responsiveness to real relevance’, in Heidegger and Cognitive Science, J. Kiverstein and M. Wheeler, eds (Basingstoke: Palgrave-Macmillan, forthcoming). See the works of Michael Wheeler cited above. Hubert Dreyfus, ‘Why Heideggerian AI failed’. Michael Wheeler, Reconstructing the Cognitive World. Hubert Dreyfus, ‘Why Heideggerian AI failed’, 345–346. Although we do not have the space to describe the full phenomenological picture here, it is worth noting that an object will become present as such if its affordances appear to the subject as plainly separated and virtually independent from the background. In this case the object is offering itself as (to use Heidegger’s terminology) a present-at-hand entity. Hubert Dreyfus, ‘Why Heideggerian AI failed’, 345. Maurice Merleau-Ponty, Phenomenology of Perception, trans. C. Smith (London: Routledge, 1962), 159. Ibid., 126–127. Ibid., 161. For much more on this continuity at the subpersonal level, see Michael Wheeler, ‘Cognition in context’, and ‘The Problem of Representation’. Heidegger, Being and Time, 111–112. Heidegger, Being and Time, 110.

Chapter 9

Expressivism in Brandom and Taylor Nicholas H. Smith

There is currently a lot of interest among philosophers in the idea of philosophical ‘expressivism’. This is in so small measure due to the pioneering work of Robert Brandom, who uses the term ‘expressivism’ as a label not just for his own project, but for a whole philosophical tradition that encompasses thinkers as diverse as Kant, Hegel, the American pragmatists, Heidegger and Wittgenstein.1 According to Brandom, this tradition offers a source of insights which, when systematized, can mount a serious challenge to the long-standing dominant paradigm in philosophy, especially analytical philosophy: ‘representationalism’. The representationalist paradigm presumes that the representational relation – say between thought and object, word and thing, language and the world – is primitive, so that representations figure as first in the order of explanation of what we are able to think, know and say. Expressivism denies this and maintains that the ‘expressive’ relation is basic. Brandom’s systematic exposition of the expressivist alternative to representationalism has an impressive ‘staying power’, to use Habermas’s phrase, and it has rightly provided the focus for recent debates about the nature of expressivism.2 It is surprising, though, that Charles Taylor’s work on expressivism, for which he was famous in the 1970s and 80s, has hardly featured in these debates.3 After all, it is fair to say that prior to the recent interest in expressivism sparked by Brandom’s work, the term ‘expressivism’ owed its currency in Anglophone philosophical circles chiefly to two sources: its use as a label for a position in meta-ethics also commonly called ‘emotivism’; and Taylor’s very different use of it (indebted, in turn, to Isaiah Berlin) to designate a conception of human agency and its relation to nature that emerged historically out of Romantic thought.4 It is also unfortunate that Taylor’s work in this area has been neglected because it deals with an aspect of what it means to ‘express’ something which is of great philosophical importance, but which tends to remain in the background – hidden and perhaps even repressed – in the contemporary debate about expressivism.5 At least this is the claim I want to put forward here. I begin by picking up on Brandom’s suggestion that expressivism follows American pragmatism in seeking to advance the cause of the Enlightenment.

146

Postanalytic and Metacontinental

This provides us with a first point of contrast with Taylor’s understanding of expressivism, since Taylor takes expressivism to be inseparably bound up with the Romantic critique of the Enlightenment and as fundamentally opposed to Enlightenment naturalism. I then distinguish two features of what we ordinarily mean by the term ‘expression’, one of which provides an intuitive basis for understanding Brandom’s expressivist program, the other of which provides an interpretive key for understanding Taylor’s version of expressivism. After looking briefly at the main tenets of Taylor’s expressivism, I conclude by considering its relation to Romanticism on the one hand, and to Brandom’s expressivist renewal of the Enlightenment project on the other.

1. Second Enlightenment expressivism In an essay entitled ‘The Pragmatist Enlightenment (and its Problematic Semantics)’, Brandom presents an intriguing account of classical American Pragmatism as ‘the announcement, commencement, and first formulation of the fighting faith of a second Enlightenment.’6 Like the first Enlightenment, the pragmatist second Enlightenment took reason to be ‘the sovereign force in human life’, where reason was understood ‘on the model provided by the forms of understanding distinctive of the natural sciences’.7 Those forms of understanding underwent a shift in the nineteenth century – Darwinian biology took over from Newtonian physics as the paradigm – and the pragmatists sought to bring the Enlightenment project up to date in light of this development. But there were other features of the first Enlightenment the classical American pragmatists wanted to correct. Some of these, Brandom remarks, had already been noticed by the Romantics, and rejection of them gave the pragmatists and the Romantics common cause. Both rejected the ‘spectator theories of knowledge’ favoured by the first Enlightenment, both drew attention to the genesis of theoretical knowledge in practical involvement in the world, and both turned to a ‘concept of the organic’ for healing – as Brandom puts it – ‘the dualistic wound inflicted by the heedless use of an over-sharp distinction between mind and world’.8 But for all their attachment to Romanticism, the pragmatists remained fundamentally committed to the rational principles of the Enlightenment. Whereas, according to Brandom’s story, the Romantics countered the first Enlightenment with an irrationalist ‘privileging of feeling over thought, intuition over experience, [and] of art over science’, the pragmatists held their nerve and sought (to borrow another phrase from Habermas) to enlighten the Enlightenment about itself.9 And although the pragmatists were seduced for a time by the ‘vitalistic biology’ and ‘organic metaphors’ of Romanticism, it soon managed to shake these off. While its semantics took longer to mature (due in part to its enduring weakness for organicism), pragmatism is now in a position

Expressivism in Brandom and Taylor

147

not just to announce, but to defend in systematic fashion the cause of a second Enlightenment. It is able to do this, Brandom argues, because of the powerful theoretical resources that became available to philosophers in the twentieth century who focused on the ‘discontinuities with nature’ that language both ‘establishes’ and ‘enforces’.10 The emphasis placed on the ‘continuities between concept users and organic nature’ prevented the classical American pragmatists from developing an adequate theory of the conceptual. But Brandom acknowledges that the emphasis on continuity was not without its own warrant, and he concedes that ‘we have not yet sorted out the tensions between naturalistic assimilationism and normative exceptionalism about the discursive practices most distinctive of us’.11 This is a nice formulation of one of the most fundamental problems facing contemporary philosophy. But one also cannot help but be struck by how closely it defines the central problem that the Romantic movement (more so than the first Enlightenment) understood itself as facing. From its beginnings with Herder, Hölderlin, Schleiermacher and others, the Romantic movement was concerned precisely with reconciling the autonomy (normative exceptionalism) and creaturely nature (naturalistic assimilationism) of the being marked above all by its language capacities (which is on this account the rational animal). And it was with a view to solving this problem that the Romantics first formulated the outlook that, thanks to the interpretations of Berlin and Taylor, has become widely known (especially among historians of ideas) as ‘expressivism’. Thus, while Taylor and Brandom both take expressivism to rest on the thesis of the semantic priority of expression (that expression not representation comes first in the order of explanation of meaning), and while they substantially agree on who belongs to the expressivist canon (Hegel, Heidegger and Wittgenstein in particular), the Romantics also occupy a prominent place in Taylor’s (but not Brandom’s) pantheon of expressivists, and they do so because of (not despite) their concern to bring out the continuity that exists between the rational human animal and the rest of organic nature. The twentieth-century philosopher arguably most taken up by this concern, Merleau-Ponty, also counts as a key expressivist for Taylor, but not for Brandom.12 In order to understand better where Taylor’s roughly speaking ‘Romantic’ paradigm of expressivism parts company with Brandom’s expressivism of ‘second Enlightenment’, we need to have a clearer idea of what it is about ‘expression’ that makes it fundamental from a semantic point of view. I will consider Taylor’s own views on this in a moment. But we find a clue to Brandom’s and Taylor’s different philosophical conceptions of expression in distinct features of the ordinary concept of expression, that is, the concept as it is used in everyday language. So let us briefly attend to these first.13

148

Postanalytic and Metacontinental

2. Two Features of expression First, consider what we ordinarily mean by a ‘facial expression’. One salient and obvious feature of a facial expression is that it involves activity of some kind. Smiling, grimacing, pouting, are things that we do, more or less voluntarily. I can smile in front of a camera if I decide to, or do it spontaneously and without forethought to acknowledge an approaching friend. As well as being a kind of act, facial expressions reveal something about our inner state. When my face drops at hearing some bad news I convey something about how I feel. Communication of course hinges on such expressions. But facial expressions don’t just seem to reflect how the subject feels. They also actually seem to help constitute the feeling. It may take more than a smile to make me feel happy, but expressing a feeling can certainly make a difference to what the feeling is like. Think now of certain contexts in which the phrase ‘express yourself’ might be used. We can imagine it being used by the coach of a team of skilled players who, having gone through all the training, the tactics and so forth, are in the end told not to worry about obeying this or that instruction, and just play their own game. Of course, here too self-expression involves activity, a doing of something. But it also carries the implication of doing something new, something not done (or perhaps even conceived of) before. For the skilled player to express herself is not simply to repeat what she has done before, though the skill may have been acquired by constant repetition. Rather, it is to create a play, without necessarily meaning (in the sense of consciously intending) to do it, or without knowing in advance what will be done. This is the difference between expressing oneself and, say, asserting oneself. Both involve free action of sorts, but the mode of freedom involved in expression has an openness, contingency and room for creativity lacking in free self-assertion. I mention these examples simply to draw attention to two different features of what in ordinary use ‘expressing’ something means. One is that expression is a matter of doing. The other relates rather to a bringing about, the creation of something that didn’t exist (or that only had the potential to exist) prior to the expression. To the extent that representation involves a relation of correspondence between two distinct items, or a passive reflection of one thing in another, we can readily see how expression (both as a ‘doing’ and a ‘constituting’) contrasts with it. We can also readily see how expressing something, at least in the sense of a doing, is subject to norms. If through an expression one is doing something with something, we can straight away see that the correctness or incorrectness of the expression is a matter of the propriety or impropriety of the deed. The normativity of expression would then be no more mysterious than that of any action and could be reconstructed from the basic principles of practical reason. It is this feature of expression that the expressivism of the second Enlightenment draws on. As Brandom puts it, the pragmatists ‘tried to figure out what it

Expressivism in Brandom and Taylor

149

is we do . . . that adds up to thinking or knowing something, even unsuccessfully’.14 Admittedly, some of the classical formulations of pragmatism suffered from a narrow, instrumentalist understanding of what the basic principles of practical reason were. They seemed to suggest that practical reason was simply a matter of determining the most efficient means to a given end. By rashly proposing that the truth of a theory, for example, consisted in its utility, or that instrumental control over an environment could provide the criterion of genuine knowledge, certain pragmatists repeated a mistake of the first Enlightenment. The instrumentalist error is also evident in the classical statements of ‘expressivist’ meta-ethics.15 This theory correctly analyses the meaning of moral judgements in terms of their expressive use rather than their putative representational function (which in any case they do not have: there is no independent moral reality to represent on this view). But it incorrectly reduces the norms implicit in that expression to those of instrumental rationality. For example, the meaning of ‘X is wrong’, on the standard ‘expressivist’ analysis of the emotivists, merely gives expression to a feeling (of disapproval) which has the practical purpose of eliciting disapproval in the addressee. But while this theory looks in the right place for the norms implicit in making moral judgements (namely, social practice), it reconstructs those practical norms in the wrong way. It is rightly ‘anti-representationalist’ and expressivist, but wrongly instrumentalist. Taylor is also highly critical of emotivism, but even he would acknowledge that it touches on something important. This is the insight that moral judgements, understood as expressions, ‘bring something about’. Taylor would agree that the utterance of moral judgements brings about a certain social relation between the addresser and the addressee, and this is an important aspect of their meaning. In principle such speech acts could serve to establish or support instrumental relations between people, but that will depend on the cultural context.16 A more serious objection Taylor has to emotivist expressivism, however, is that it assumes an ontology of natural facts onto which humans merely project their values. His fundamental objection is to the anti-realism of this kind of analysis, that is to say, its presumption that values are no more than a subjective matter, and that the only objective states of affairs are those described in a value-neutral manner by the modern natural sciences. Taylor thus rejects the subjectivist construal of what is ‘brought about’ by expressions of this kind, and by whom it is brought about. In Taylor’s view, the instrumentalist interpretation of what is done in making moral judgements is one reason for rejecting emotivist expressivism, the subjectivist interpretation of what is brought about by the use of moral language is another. Taylor agrees with Brandom and the pragmatists that expression is a form of human activity. It is a kind of doing. As such, expressions are subject to practical norms. But the semantic significance of expression, in Taylor’s view, arises less from this feature than the second one I identified above, namely their ability to

150

Postanalytic and Metacontinental

bring something about. This is also the feature of expression that MerleauPonty and the Romantic tradition – two key sources of Taylor’s but not Brandom’s expressivism – focused on. While not denying the pragmatist point that as a form of human activity expression is subject to practical norms, the decisive point for the expressivists that inform Taylor’s understanding of expressivism is the role of expression in making something manifest. Let me now say a little bit more about where Taylor takes this idea.

3. Taylor’s ‘Romantic’ expressivism Taylor’s explicit point of departure for thinking about expression is that ‘an expression makes something manifest in an embodiment’.17 This simple formula already reveals much about Taylor’s expressivist philosophical orientation. If this is what, for philosophical purposes, we mean by expression, then the idea that expression is first in the order of explanation of meaning implies that embodiment is an essential feature (or presupposition) of the things that carry meaning. Obviously, one thing we think of as bearing meaning is human experience. If that is the case, then the ‘meaning-content’ of experience cannot, at a fundamental level, be separated from its manifestation in the human body. It would be wrong to suppose, for instance, that experience (as a bearer of meaning) is only contingently connected to the body in which it is made manifest. This idea is encouraged, though, by representationalist theories of experience that oppose an inner realm of representations to an outer realm of represented objects. Taylor’s expressivism is from the start opposed to the ‘inner-outer’ sorting that accompanies standard representationalist theories of the mind.18 Indeed, the whole point of expressivism, what is so attractive about it from Taylor’s point of view, is that it avoids this kind of dualism. Instead, it conceives the mind as inseparable from its incarnation in matter. This is a fundamental axiom of Taylor’s expressivism which he is convinced of early on by Merleau-Ponty, and which his studies of Romantic thought and its legacy (especially Hegel) impressed on him further.19 According to this kind of expressivism, human beings share with other animals a natural capacity to express feelings, desires and purposes in action, and speech is a natural development of this capacity for action. Following Herder, Taylor takes a genetic approach to language that traces it back to our engaged, embodied stance in the world.20 Our original condition is not that of minds with private thoughts and feelings which are subsequently communicated in speech (though this is an ability we later acquire as a refinement of our capacity for linguistic expression). Rather, the thoughts and feelings we recognize in ourselves and others come to be as we recognize them through the way we express and articulate those thoughts and feelings. To say that language expresses meaning is to say that the meaning language

Expressivism in Brandom and Taylor

151

conveys is bound up with its linguistic embodiment: the meaning does not pre-exist its linguistic form in the mind of the language user. In this way, meaning is realized and not just communicated in what Taylor calls ‘the semantic dimension’.21 According to Taylor’s Herder-inspired view, the primary function of language is not to describe something already there, something that would be there even without language; though again this is not to say that language is not often descriptive. Rather language originally expresses things that can only be made manifest through the expression, that is, internally to the semantic dimension. But if meanings do not pre-exist expression, in an important sense the expression constitutes the meaning. This meaning-productive, constitutive role of expression is never far from the focus of Taylor’s writings on language, and it would only be a small exaggeration to say that his whole expressivist philosophy of language is geared towards recovering a sense of its significance.22 The three main ways in which Taylor takes meanings to be constituted internally to language are as follows. First, and as mentioned before, there are certain interpersonal relations that only come about insofar as they issue from an expressive act. All sorts of social relations are only possible on account of the right word being said, at the right time, in the right way. Just by saying ‘sorry’ for instance, a whole new social space can be opened up. Second, there is a range of emotions whose very content depends on the words used to express them. By finding just the right expression, for instance, a confusing, troubling emotion can become clear, and with the clarity a new feeling manifests itself. And third, Taylor draws attention to moral standards whose ability to convince and move us is conditioned by the expression we give to them. In all three cases, linguistic expression brings something about which cannot exist without language or outside the semantic dimension. By expressing thoughts, feelings, desires and purposes in language we can transform them. By giving them a reflective dimension we can, in effect, create new ones. In this sense language is productive of meaning, and the capacity to generate or create meaning is intrinsic to it. Language thereby creates possibilities that don’t exist for non-linguistic beings. Now the normativity of language, on this view, is set by the multiple standards of correctness and incorrectness that hold sway across the semantic dimension, not just truth and falsity. Rightness is not just a question of the fit between the expression and what is expressed, or between a proposition and its object. It is not just descriptive or assertive expressions that are subject to norms. Standards of correctness and incorrectness apply to other kinds of expression too. For example, a word or articulation can be the proper one for establishing a desired interpersonal relation (‘sorry’ bringing about reconciliation); it can be right for clarifying an indeterminate feeling (say as indignation rather than anger); or it can guide an agent to act in the way called for by the situation.

152

Postanalytic and Metacontinental

According to Taylor’s expressivism, therefore, there are normative constraints on non-designative modes of language use, usages which do not involve talking about an independently existing object. We have to look for the right word, or the right articulation, in a plurality of contexts. And in some contexts, being true to the phenomena, or expressing them rightly, may require radical linguistic innovation. Linguistic or semantic innovation can itself constitute the meaning. Not just descriptive prose, but broadly speaking ‘poetic’ modes of expression, may be required for getting (in a manner of speaking) at the truth. This suggests that in doing things such as evoking the right mood, or articulating a feeling with the right nuance, we are just as accountable (or subject to norms) as we are when making literal assertions about some thing. Descriptive prose then loses its status as the paradigm locus of meaning. And it was this productive power of language, Taylor reminds us, that the Romantics took as their point of departure. In looking to the work of art as the paradigm locus of meaning the Romantics were not so much rejecting reason, science and rationality as responding to the distinctive capacity of expressive language to ‘make possible its own content’. Taylor is sympathetic to this move, and his reconstructions of the rationale behind it have done much to bring out the contemporary relevance of Romantic expressivism.

4. The Post-Romantic agenda At the same time, it is misleading to say that Taylor himself is a ‘Romantic’ expressivist. Certainly, it is not a consequence of his expressivism to privilege feeling over thought or intuition over experience. Nor does Taylor show any proclivity for the ‘concept of the organic’ or ‘vitalistic biology’ with which pragmatism briefly flirted under the spell of Romanticism. He agrees with Brandom that the concept of nature has been so thoroughly disenchanted by the natural sciences that the project of recovering meaning in nature at the ‘ontic’ level – that is, at the levels at which objective scientific descriptions work and the metaphysical discourses that ground them – has had its day. But in Taylor’s view this does not mean that the problem of reconciling mind and nature, with which the Romantics grappled with only partial success, has gone away. Nor, in Taylor’s view, can this problem simply be attributed to ‘the heedless use of an over-sharp distinction between mind and world’ as Brandom suggests. This is because, for all the progress that the theory of concepts has been able to make by focusing on the discontinuity between concept users and organic nature, concept users are still always embodied in nature and their understanding develops out of genetically more basic pre-conceptual expressive capacities. Failure to keep in view how meaning finds expression at this level would undermine the whole expressivist project, at least as Taylor understands it. Conversely, retrieval of the pre-conceptual content of lived experience suggests itself as a more promising (and philosophically

Expressivism in Brandom and Taylor

153

less demanding) way of reconciling mind and world than full-blooded Romantic organicism. Taylor’s expressivism also cannot be considered ‘Romantic’ if that means it privileges art over science as a medium for expressing truth. On Taylor’s account, as I have already suggested, objective scientific description realizes a potential for expressive ‘correctness’ that non-scientific discourses cannot match. Alongside science, art realizes other kinds of expressive potential, and has its own standards of appropriateness, much as the Romantics declared. But even here Taylor does not think that art realizes powers of expression in quite the way the Romantics supposed. There are several reasons for this. One has to do with what Taylor calls the ‘fracturing’ of ‘the expressive turn’ initiated by the Romantics, whereby expressive practices that make manifest an order of significance in which the subject is set are diversified and brought into conflict with each other.23 A related reason has to do with the collapse of what Taylor calls a ‘publicly established order of references’ – which the Romantics could presume but which no longer obtains in the post-Romantic world – in relation to which expression through art could claim intersubjective validity.24 A third reason concerns the contestability of the Romantic assumption that art could reconcile mind and nature by disclosing nature as what Taylor calls a ‘constitutive good’, that is, a reality worthy of unambiguous affirmation.25 It fell to Romantic art to reveal, and to make manifest, the course of life (élan vital) running through nature and the heightening of life that came about when nature expressed itself as art – all, of course, in a context of bourgeoning industrialization. But in our own ‘Post-Romantic’ historical context, the goodness of nature, and the worth of a life led in ‘contact’ with it, cannot so readily be taken for granted. Nonetheless, the idea that the work of art is a paradigm case of expression, precisely in the sense that it ‘makes something manifest in an embodiment’, remains fundamental to Taylor’s expressivism. Taylor owes this idea to the Romantics, but in retrospect he can say that the Romantics had a restricted and in some respects naïve understanding of what such expression involved. They rightly saw that in the modern world the creative imagination has an indispensable role to play in the articulation (or expression) of our intuitive sense of belonging to a spiritually significant reality. The work of art makes a ‘moral reality’ manifest that would otherwise be inaccessible; it ‘discloses a world’. But from Taylor’s post-Romantic perspective, the understanding reached through such expressions is ‘indexed to a personal vision’ in a more radical manner than the Romantics could have seen.26 This is not to say that the ‘moral reality’ disclosed in a work of art is just an invention of the subject, a mere product of the imagination with no external grounding. But it is to acknowledge the unavoidably subjective manner in which that ground comes to expression. At this point Taylor parts company with Hegel’s rationalist expressivism, and by implication with Brandom’s. On the one hand, Taylor claims that the work

154

Postanalytic and Metacontinental

of art can give legitimate, non-conceptual and non-substitutable expression to a subject’s intuitive sense of reality. Art is not just the self-awareness of Spirit ‘in default of concepts’, as Hegel supposed, but a genuine source of understanding that at least in some cases lies beyond the reach of the conceptual.27 Taylor thus rejects the Hegelian idea of Absolute Knowledge, of an Aufhebung of art by philosophy. On this account his expressivism amounts to a hermeneutics of finitude that contrasts with the Absolute Idealism Brandom finds congenial.28 On the other hand, Taylor maintains that it is an open question whether the norms to which expressions are subject have a human or nonhuman source. Brandom’s expressivism settles this matter decisively in favour of the secular option. Indeed grasping this is precisely what Absolute Knowledge is all about: taking responsibility for the discursive practices that characterize us. It is also the founding principle of pragmatism, and of the ‘Enlightenment project’ pragmatism advanced. Richard Rorty has remarked that Brandom has managed better than anyone else to vindicate this core principle of Enlightenment thought philosophically, and it is hard to disagree with him.29 Brandom shows how authority by way of expression rests on the freedom of those who participate in expressive practices themselves. The idea that the subject of expression is accountable to some non-human source of norms is anathema to this version of exressivism. Taylor’s expressivist might reply that while the authority lent to expression is just a human, social matter, excellence by way of expression may require the acknowledgement of non-human sources of significance (such as the natural environment or other species). Such acknowledgement would not have to involve obedience to something non-human, but it would seem to require openness to possibilities that the Enlightenment view rules out a priori. The stakes of the debate between Taylor’s and Brandom’s expressivism are thus high. It is a debate well worth taking further.

Notes 1

2

3

Robert B. Brandom, Making it Explicit (Cambridge MA: Harvard University Press, 1994); and Articulating Reasons (Cambridge MA: Harvard University Press, 2000). Jürgen Habermas, ‘From Kant to Hegel: On Robert Brandom’s Pragmatic Philosophy of Language’, trans. by Maeve Cooke. European Journal of Philosophy, 8 No. 3 (2000): 322. See in particular the following works by Taylor: Hegel (Cambridge: Cambridge University Press, 1975); ‘Action as Expression’, in Cora Diamond and Jenny Teichmann eds, Intention and Intentionality (Ithaca: Cornell University Press, 1979); ‘Hegel’s Philosophy of Mind’, in Taylor, Human Agency and Language (Cambridge: Cambridge University Press, 1985); and Sources of the Self (Cambridge: Cambridge University Press, 1989), ch. 21.

Expressivism in Brandom and Taylor 4

5

6

7 8 9

10 11 12

13

14 15

16

17 18

19

20

21

22

23 24

155

Not that Berlin and Romanticism are the only sources of Taylor’s understanding of expressivism. As I have shown elsewhere (Charles Taylor: Meaning, Morals and Modernity [Cambridge, Polity, 2002] chapter 1), Taylor is just as indebted to the ‘expressivist’ account of perception presented by Merleau-Ponty. See also note 12. I leave to one side discussions of expressivism triggered by Deleuze’s interpretation of Spinoza – see Gilles Deleuze, Expressionism in Philosophy: Spinoza, trans. Martin Joughin (London: Zone Books, 1990). Another influential discussion of expression in the continental tradition that I won’t have space to consider is Emmanuel Levinas, Totality and Infinity, trans. A. Lingis (Pittsburgh: Duquesne University Press 1969). Robert Brandom, ‘The Pragmatist Enlightenment (and its Problematic Semantics)’, European Journal of Philosophy (2004): 2. Ibid. Ibid. Ibid., 3, and for Habermas’s case for an ‘enlightened enlightenment’, see Jürgen Habermas, The Philosophical Discourse of Modernity, trans. F. Lawrence (Cambridge, MA: MIT Press, 1987). Brandom, ‘The Pragmatist Enlightenment’, 15. Ibid. Taylor refers to Merleau-Ponty’s expressivism as far back as 1958, in his first published academic article (co-written with Michael Kullman) ‘The Pre-Objective World’, The Review of Metaphysics 12 (1958): 127. For Brandom’s and Taylor’s own reflections on how their philosophical conceptions of expression relate to the concept as it is used in ordinary language, see Brandom, Articulating Reasons, 7f, and Taylor, ‘Action as Expression’. Brandom, ‘The Pragmatist Enlightenment’, 15. See, for example, A. J. Ayer, Language, Truth and Logic (London: Gallancz, 1946). While broadly sympathetic to MacIntyre’s meta-ethics, Taylor does not go as far as to endorse MacIntyre’s thesis in After Virtue that emotivism merely reflects the degenerate state of morality in modern society. See Alisdair MacIntyre, After Virtue (London: Duckworth, 1984). Taylor, ‘Action as Expression’, 73. See for example Taylor, ‘Foundationalism and the Inner-Outer Distinction’, in Nicholas H. Smith ed., Reading McDowell: On Mind and World (London: Routledge, 2002). See Charles Taylor, ‘Genesis’, review of The Structure of Behaviour by Maurice Merleau-Ponty, New Statesman, 70 No. 3 (1965): 326–327. See Taylor, ‘The Importance of Herder’, in Edna Margalit and Avishai Margalit eds, Isaiah Berlin: A Celebration (Chicago: University of Chicago Press, 1991). See Taylor, Philosophical Arguments (Cambridge MA: Harvard University Press, 1995), 103. See Charles Taylor, ‘Theories of Meaning’, Human Agency and Language (Cambridge: Cambridge University Press, 1985). See Taylor, Sources of the Self, 390. Ibid., 491.

156 25 26 27 28

29

Postanalytic and Metacontinental

Ibid., 93. Ibid., 510. Taylor, Hegel, 467. See Robert Brandom, ‘Some Pragmatist Themes in Hegel’s Idealism: Negotiation and Administration in Hegel’s Account of the Structure and Content of Conceptual Norms’, European Journal of Philosophy, 7 No. 2 (2000): 164–189. See Richard Rorty, ‘Cultural Politics and the Question of the Existence of God’, in Nancy Frankenberry ed., Radical Interpretation in Religion (Cambridge: Cambridge University Press, 2002), 53–77.

Chapter 10

Wittgenstein as Trans-Analytic-Continental Philosopher Dale Jacquette

1. Wittgenstein’s thorny legacy Wittgenstein’s Tractatus Logico-Philosophicus fostered and served even if only inadvertently as an iconic touchstone for the analytic movement to which, despite itself, it helped give direction and shape. Wittgenstein’s later ideas after 1930 were deceptively more down-to-earth, but what persisted when in Wittgenstein’s mind the early philosophy was defeated were the questions that had inspired the Tractatus in the first place. It was the need to find new answers to the problem of meaning and its implications for the practice and prospects of philosophy, once the Tractatus was rejected, that was eventually to result in the later Wittgenstein’s remarkable innovations.1 The later Wittgenstein’s difficult thought, more remote from the friendly field of exact logical investigations, were increasingly discussed; although, like much of the early philosophy, they were also frequently misunderstood. It is no secret that many analytic philosophers prefer the early to the later Wittgenstein.2 How, then, should Wittgenstein’s early and later philosophy be properly understood, and how does his work relate to the two major trends in contemporary philosophy, the so-called analytic and continental schools? Is Wittgenstein in some sense a trans-analytic-continental thinker?

2. Across the analytic-continental divide It might appear essential to begin an inquiry into Wittgenstein as a trans-analyticcontinental philosopher with a precise definition of analytic philosophy, continental philosophy and the distinction between the two. Otherwise, what are we talking about? Wittgenstein can nevertheless be interpreted as trans-analyticcontinental in a sense that does not at all depend on the exact division of these categories in theory or practice. Fortunately so, because it turns out to be very difficult to accurately characterize and draw a firm line between analytic and continental philosophy. It will nevertheless be useful to say something about

158

Postanalytic and Metacontinental

the distinction between analytic philosophy and continental philosophy as it is usually intended, if only for the sake of explaining how Wittgenstein outflanks the conventional dichotomy. Roughly, analytic philosophy is that mode of philosophical inquiry in which the clarification of concepts through efforts to provide exact definitions and sound arguments is made an adjunct to, if not identified with, the purposes and methods of logic, mathematics and natural science. Analytic philosophy is often described as the pursuit of a special area of knowledge, with an emphasis on the problem of understanding the meaning of language and the application of semantic theory to the solution, resolution or dissolution of philosophical problems. All this, of course, Wittgenstein, early and late, flatly denies. Analytic philosophy is historically associated especially with mainstream philosophical thought in Great Britain, the USA, Australia, Canada and Austria, although more recently analytic philosophy, in step with modern science, has spread to almost every corner of the globe. It is further linked with the names, among many others, of Gottlob Frege, Bertrand Russell, G. E. Moore, Wittgenstein, Rudolf Carnap, Gilbert Ryle and W. V. O. Quine. Continental philosophy in contrast is comparatively less monolithic, and accordingly more difficult to characterize. It is frequently remarked that continental philosophy is more of a literary than scientific enterprise. Where the concept of what is to count as scientific, let alone literary, is itself philosophically up for grabs, however, the phenomenology of Edmund Husserl and even of Maurice Merleau-Ponty, again, among numerous others, quickly present counterexamples. Other continentalists more naturally fit the literature-over-science stereotype. One thinks here naturally of Friedrich Nietzsche, Miguel de Unamuno, Georges Bataille and Jean-Paul Sartre. An emphasis on the subjective and the problem of coming to terms with (what is often referred to without much effort at explanation as) the human condition are also characteristic of much of the philosophy categorized as continental. It is nevertheless inaccurate to maintain that continental philosophers do not try to define concepts, contribute to the clarity of ideas or use arguments. What it is probably fair to say of many and even most cases is that continental philosophers typically do so with less emphasis and rigour, though sometimes with more flair, than their analytic colleagues, Husserl and others outstanding once again as exceptions. It is also correct to observe that some continental thinkers are indifferent and occasionally even hostile to the possibility of truth, and sometimes contemptuous of the aims of science. Such conclusions, it must now be emphasized, also remain open to and have historically been defended by some analytic philosophers as consequences of their investigations. The thinkers of principal interest to continental philosophy, merely to scratch the surface, include Immanuel Kant, at the apex of the great divide, G. W. F. Hegel, Johann Gottlieb Fichte, F. W. J. Schelling, Arthur Schopenhauer, Nietzsche, Karl Marx, Husserl, Martin Heidegger, Sartre, Merleau-Ponty, Jacques Derrida, Jean-François Lyotard, Michel Foucault and Hans-Georg Gadamer.

Wittgenstein: Trans-Analytic-Continental

159

There is authoritative documentation for a connection between the early Wittgenstein and Kant and Schopenhauer, and between the later Wittgenstein and such stock continental thinkers as Søren Kierkegaard. We know Wittgenstein read Kant and heavily annotated a copy of the Kritik der reinen Vernunft while a prisoner at Monte Cassino near the end of the First World War.3 The influence of Schopenhauer on the early Wittgenstein is manifest, not only in Wittgenstein’s sign-symbol distinction, in which the transcendence of the symbol-aspect of language is made indispensable to the picture theory of meaning, but also, among other key places, in the image of the relation between the eye and the visual field in Tractatus 5.6331, in the compression of ethics and aesthetics as one in 6.421, and in the ladder metaphor of 6.54.4 The later Wittgenstein, with his long-standing religious interests, seems to have found Kierkegaard as rewarding in some ways as Fyodor Dostoyevsky.5 While the analytic versus continental distinction is problematic, we will follow established custom and continue to speak of the distinction crudely distinguished in part by exemplars and historical figures as benchmarks rather than by means of necessary and sufficient conditions for the definition of these dissimilar ways of exploring what each side identifies as the challenges of philosophy. When we examine even superficially a page from the writings of Frege and Hegel, or of Carnap and Emmanuel Levinas, it is clear at once that there is something very different going on philosophically. A sense of this difference is all that we need in order to consider Wittgenstein’s place somewhere along or hovering above this spectrum. How, then, does Wittgenstein fit into this picture of the development of philosophy, split, since the time of Kant, into two opposing branches? The view I shall advance is that Wittgenstein is a trans-analytic-continental thinker in the sense that his work in philosophy does not fit into either major movement regardless of how they, and the differences between them, are characterized. We can interpret Wittgenstein not as bridging but rather as transcending both analytic and continental philosophy, along with all other philosophical traditions, by virtue of belonging to neither and hence legitimately inappropriable by any non-Wittgensteinian philosophical tradition or combination of traditions. Wittgenstein, on the interpretation to be advanced in what follows, belongs to neither one philosophical school as opposed to the other; he is neither analytic nor continental, nor is he in any sense an amalgam of the two.

3. Three strategies for interpreting Wittgenstein If we want to understand Wittgenstein’s relationship to analytic versus continental philosophical traditions, then we need to step back from the stereotypes both of his philosophy and of the attractions that analytic and continental philosophers have sometimes believed themselves to find in his work. Let us assume, at least for the sake of argument, that Wittgenstein is rightly described

160

Postanalytic and Metacontinental

as a trans-analytic-continental philosopher, and ask what such a pronouncement could possibly mean. There are several imaginable interpretations. Here are perhaps the most obvious possibilities to be considered: (A) Wittgenstein started out as an analytic philosopher, and in some sense, together with Frege, Russell, Moore and others, invented this now thriving branch of philosophy. When he came to see that the Tractatus Logico-Philosophicus had failed, or perhaps as a result of other events in his personal life, he turned away from analytic philosophy in the later period, marked most notably by the Philosophical Investigations and the great body of posthumous writings, and embraced themes and methods that are more characteristic of continental thought. Thus, speaking of Wittgenstein as a thinker spanning his entire career from the early, transitional, to later periods, Wittgenstein bridges analytic and continental traditions by virtue of participating in and contributing substantially to both. It is probably fair to say that this interpretation of Wittgenstein’s philosophy as trans-temporally bridging analytic and continental methodologies and ideologies, if it ever had any credence, has by now been thoroughly discredited. The chief objection is that, properly understood, Wittgenstein cannot accurately be classified as an analytic philosopher in the usual sense of the word even in the Tractatus period. The early Wittgenstein, despite being a mascot and unintentional ambassador for the analytic movement, uses and discusses its ideas in order to draw very different kinds of conclusions that look beyond the theory-building of conventional analytic philosophy; he is doing something categorically different. Wittgenstein in the Tractatus may offer a schema for reducing meaning in a logical analysis to simplest elements, but he also tells us in the end that it is all nonsense at least to try to say so, and this bears a marked contrast with what one finds in the work of any orthodox analytic thinker. Moreover, Wittgenstein makes room in the early philosophy for the world-transcendence of the Metaphysical Subject or Philosophical I, logical form, pictorial form, form of representation, unified ethical and aesthetic value, and other aspects of the picture theory of meaning that are generally eschewed by analytic philosophy both before and after Wittgenstein. Nor does Wittgenstein significantly resemble even the crudest description of continental philosophy in the later period, when compared with those of his contemporaries representing continental thought in England, such as R. G. Collingwood, Samuel Alexander or F. H. Bradley, let alone when compared with those carrying on the tradition of philosophy firmly established on the European continent. (B) Wittgenstein’s philosophy in both periods combines discernible elements of both analytic and continental philosophy. He is throughout his career a thinker who bridges the analytic and continental philosophy divide by blending

Wittgenstein: Trans-Analytic-Continental

161

together the ideas and methods of both sides in at least a substantial part of all his work, without self-consciously trying to be loyal or disloyal to either side. Of the two alternatives considered thus far, this is undoubtedly the more reasonable and responsible effort to explain Wittgenstein’s philosophy as trans-analytic-continental. Here we have the makings of an interesting account of why it is that conspicuous aspects of Wittgenstein’s early philosophy, such as his emphasis on the transcendence of logical form, among other things, the sense of existential mystery in a grasp of the world considered as a whole sub specie aeterni (aeternitatus), and, more generally, everything that according to the picture theory of meaning can only be shown rather than said, has sometimes fallen obliquely off the analytic historian’s radar screen. Similarly, we thereby have in hand the basis for an explanation of why strong analytic elements in Wittgenstein’s later philosophy, such as the private language argument, the detailed remarks on the foundations of mathematics, the discussion of Moore’s paradox and defence of common sense, and rule-following, among other more analytically-oriented comments, have not drawn sufficient attention from continental writers. (C) Wittgenstein transcends rather than bridges the analytic-continental philosophy distinction by virtue of being neither one nor the other; instead he is unclassifiable as belonging to either of these or any other schools. On this account, Wittgenstein’s philosophy, intentionally or unintentionally, is unpigeonholeable. That is the right word, I believe, even though my spellchecker has revolted in the strongest possible terms. This is the view I shall develop and defend in the discussion to follow.

4. History of misunderstandings in the analytic tradition It is not astonishing that many of the early analytic interpretations of Wittgenstein’s philosophy should come to be replaced by more continental-leaning readings of his work. Some of the narrowly analytic accounts of Wittgenstein among logicians and philosophers of the recent past are imbalanced and misleading, and in a way begging to be supplanted by a more accurate, pluralistic account of Wittgenstein’s thought. The problem is that for all their show of rigour, and no-nonsense attitude toward what Wittgenstein did and did not say, the classic analytical interpretations in many instances simply have not done justice to the texts and relevant historical data surrounding Wittgenstein’s teaching and life. How can it be maintained, for example, that Wittgenstein’s early philosophy is primarily concerned with logic and theory of meaning in a pedestrian sense, when he tells the reader in the book itself that the sentences it contains are nonsensical?6 Is there not a confusion of means and ends in interpreting Wittgenstein’s Tractatus as a work predominantly about symbolic logic, of an admittedly idiosyncratic sort, beyond its consideration of elementary truth

162

Postanalytic and Metacontinental

tables? We ask this question in particular after noting that Wittgenstein confides in a letter about the Tractatus to Ludwig von Ficker that: ‘The sense of the book is an ethical one. I once wanted to include in the preface a sentence which actually is not now in it but which I will write out for you here since it will perhaps be a key (to the book) for you. I wanted, then, to write: my work consists of two parts: of that which is under consideration here and of all that I have not written. And it is precisely this second part that is the important one.’7 There may be an excuse for omitting reference to this kind of background information before it comes to light. Once sources like Wittgenstein’s letter to von Ficker have surfaced, however, it is hard to justify the old-fashioned readings of Wittgenstein’s Tractatus as essentially a manual of logic and semantic philosophy with an attached logically atomistic metaphysics, the real beating heart and soul of the early philosophy, the good stuff, with some adolescent Schopenhauerian conceits and Spinozaesque obsessions tacked on that we are free to disregard. Wittgenstein in numerous places demonstrates by word and deed that this is not the right way to understand his thought, and that he is not committed to the truth or even the intelligibility of the sentences in his treatise in anything like the way Frege and Russell accept the principles of their respective substantive philosophies, more or less on a par with the propositions of natural science applied to a uniquely philosophical subject matter. This is arguably the main point of the ladder metaphor in Tractatus 6.54. Similarly, in Philosophical Investigations we find Wittgenstein apparently repudiating certain propositions highly cherished by analytic philosophy, such as the possibility of essentialist definitions of certain general concepts. Or, to consider another example, many analytic expositions of the later philosophy maintain that Wittgenstein is some kind of behaviourist (as we find in U.T. Place and D.M. Armstrong8). Such attributions are sometimes breezily assumed without contextually examining Wittgenstein’s remarks in any detail. They are made despite the fact that Wittgenstein plainly states that he is not advancing any theses or arguments, and that his purpose is rather to provide a therapy for philosophical anxieties that will displace philosophical inquiry with an appreciation for those abuses of philosophical grammar that create an illusion of philosophical problems. If we are to take seriously Wittgenstein’s own testimony on the matter, then he makes a point of not embracing behaviourism in Philosophical Investigations §307, when he first asks: ‘Are you not really a behaviourist in disguise? Aren’t you at bottom really saying that everything except human behaviour is a fiction?’ And then replies: ‘If I do speak of a fiction, then it is of a grammatical fiction.” That Wittgenstein flatly denies being a reductivist or eliminativist concerning mental states, whether of a behaviourist or some other type, is evident in the surrounding passages. Wittgenstein begins in §306: ‘Why should I deny that there is a mental process? But “There has just taken place in me the mental process of remembering . . .” means nothing more than: “I have just remembered . . .”. To deny the mental process would mean to deny the

Wittgenstein: Trans-Analytic-Continental

163

remembering; to deny that anyone ever remembers anything.’ And in §308: ‘How does the philosophical problem about mental processes and states and about behaviourism arise? . . . And now it looks as if we had denied mental processes. And naturally we don’t want to deny them.’ We have only touched on a few points of common misinterpretation of Wittgenstein’s early and later philosophy in the analytic philosophical community. All are perfectly understandable and forgivable. Yet if Wittgenstein is a card-carrying analytic philosopher, the misinterpretations of some of his most important ideas by some analytic commentators suggest that there are concepts at work in Wittgenstein’s early and later work that do not really belong to any professional analytic philosophical theories and methodologies. Does this mean then that Wittgenstein, early or late, develops a philosophy that is more akin to the continental philosophy Tower of Babylon, that he is actually a continental philosopher and that this explains why his thought has been so dramatically misunderstood by analytic philosophers?

5. Wittgenstein’s reception among continental philosophers Wittgenstein’s later philosophy, marked especially by the posthumous publication of Philosophical Investigations, eventually caught the attention of continental thinkers. They found more of interest in the later Wittgenstein in contrast with what was all too frequently and erroneously dismissed as the logical positivism of the Tractatus. They could rejoice, contrary to the lamenting of proponents of the analytic movement, that Wittgenstein after 1930 had outgrown his early fixations on logic and conceptual analysis, and had opened his eyes to subjective phenomena, even to phenomenological studies, bereft of that prejudicial label, and to the human context even of semantic theory and mathematics. Wittgenstein’s later concept of a form of life (Lebensformen) ‘resonates’, so continental thinkers sometimes like to say, with that of Friedrich Schleiermacher and Wilhelm Dilthey. Many continental commentators also seem to have found Wittgenstein’s thematization of language games in the later philosophy congenial to their sense of encultured rather than abstract, purely logical, meaning. They appreciate the later Wittgenstein’s loosening up of the rigid uncompromising categories of analytic philosophy, steeped in, and often construing itself merely as offering custodial service for, an absolute logic, mathematics and natural science, and with no legitimate place for art, music, literature, religious experience and the fascinating challenges of understanding these supposedly soft topics. Continental thinkers may have been relieved to discover what they perceived as the later Wittgenstein’s moving sharply away from Russell’s conception of philosophy as identical with logic. Perhaps they cheered, reading Philosophical Investigations §11, and the claim that there are many different ways in which terms and sentences can have meaning; that the distinct parts of

164

Postanalytic and Metacontinental

language are as varied as the tools in a tool box containing hammer, nails, measuring stick and gluepot. We find the later Wittgenstein sensitively attuned to multiple uses of language, not merely to describe the contingent facts of the world, but serving a large and open-ended variety of purposes in human life (§23). Significantly, Wittgenstein takes the opportunity there also to criticize the Tractatus theory of logical atomism. The misunderstandings of early and later Wittgenstein from within the analytic philosophical community are only compounded when we turn to consider those continental writers who have commented on Wittgenstein’s philosophy. A stunning example is found in Theodor W. Adorno’s Against Epistemology. Here he writes, referring to the Tractatus: ‘As long as philosophy is no more than the cult of what “is the case,” in Wittgenstein’s formula, it enters into competition with the sciences to which in delusion it assimilates itself – and loses.’9 A pity, too, and a great irony, because of course Wittgenstein’s position is just the opposite of the view Adorno attributes to him, and much more in line with the distinction Adorno himself wants to emphasize. Stanley Rosen similarly demonstrates misunderstandings of Wittgenstein and lack of exegetical care when he maintains in The Limits of Analysis that: ‘The Absolute is the activity by which subjects and objects are produced. We can, by a kind of negative dialectic, return to the intuition of the Absolute, but we cannot discuss it analytically. Fichte’s Absolute is thus to all intents and purposes Wittgenstein’s unspeakable form of life and language. Wittgenstein is then both in his early and late manifestations a kind of Fichtean.’10 The idea seems to be that for Fichte the ‘Absolute’ is unspeakable, that Wittgenstein thought that some aspects of logic, language and life were unspeakable, and that therefore (preposterously) Wittgenstein is a kind of Fichtean. Rosen tries to support his claim that the later Wittgenstein recommended silence much as the earlier Wittgenstein did. The problem nevertheless remains that if we lapse into silence concerning the propositions in which bothersome philosophical problems are expressed, we can never hope to recover the principles of the appropriate philosophical grammar in which the distinct independent keys to ridding ourselves of the nagging burden of philosophical questioning are finally to be discovered. As a spokesperson for more contemporary postmodern responses to the later Wittgenstein that perhaps began with Lyotard, we might consider Marjorie Perloff. Perloff is a philosophically trained specialist in comparative literature, who has been influenced in interesting ways especially by the later Wittgenstein. She writes, in Wittgenstein’s Ladder: Not an overarching theory, then, but a method for ‘going on’ – this is what makes the Investigations, along with Culture and Value, The Blue and Brown Books, and the various lectures and notebooks published posthumously, so interactive a text. For here, in language always to be rephrased, recomposed, or charged with a new metaphor or example, is the deconstruction of

Wittgenstein: Trans-Analytic-Continental

165

our own language usage at its most insidious, baffling – and yet its most ordinary . . . the Investigations cannot possibly constitute any sort of whole; its negative seriality is an index to the inherent provisionality of the text.11 I find it problematic in the first place to attribute to Wittgenstein ‘the deconstruction of our own language usage at its most insidious, baffling – and yet its most ordinary’. I never find Wittgenstein doing anything that I would be content to describe as the ‘deconstruction’ of language use. Wittgenstein discusses an analysis of language in the Tractatus, but what he says about the meaning of language in both the Tractatus and the later texts Perloff mentions does not seem to involve what has come to be meant by deconstruction. Nor, although I shall not comment further, do I understand what Perloff means when she says of Wittgenstein’s Investigations: ‘its negative seriality is an index to the inherent provisionality of the text.’ Such misunderstanding and expository obscurity in Adorno, Rosen and Perloff only contributes to the stereotype of continental philosophy as slack and confused. Needless to say, not all commentary on Wittgenstein’s possible continental influences and impact has been incompetent. An interesting book was written about Derrida & Wittgenstein by Newton Garver and Seung-Chong Lee, though much to the disadvantage of Jacques Derrida, in the ongoing comparison of the respective merits of the two philosophers.12 Nicholas F. Gier, in Wittgenstein and Phenomenology, makes careful authoritative work of understanding certain aspects of Wittgenstein’s thought in relation to several of the great figures in classical phenomenology.13 To the extent that there is an intelligible succession of ideas interrelating Wittgenstein and continental thinkers, to that extent we can better understand both what Wittgenstein has to offer in the early and later philosophy, and its situation within the larger historical context of extraanalytic European thought.14 Can we go further in this direction, now that the first generation of continental-minded readers of the later Wittgenstein have had their chance, and turned over the helm to more recent thinkers? To answer this question, we turn to two relatively recent commentators on Wittgenstein’s later philosophy from a continental-philosophy-friendly perspective, who offer more sophisticated and textually better-grounded interpretations of the later Wittgenstein than we have seen so far. Stanley Cavell and Richard Rorty have recognized that there is something unique about Wittgenstein’s later philosophy, which they try to explicate and compare with more run-of-the-mill philosophy, both before and after Wittgenstein. I conclude that despite their initial promise neither rightly understands the later Wittgenstein. Moreover, I believe that they could not possibly hope to do so as they propose, detaching the later Wittgenstein from the early, and independently of a more penetrating study of Wittgenstein’s Tractatus. If we anticipate finding in Cavell or Rorty a way of understanding Wittgenstein as a trans-analytic-continental philosopher, acknowledging from the outset that it is not their purpose so to characterize him, then we are bound to be disappointed. Cavell and Rorty,

166

Postanalytic and Metacontinental

as we will see, fail to appreciate the deliberate and methodical way in which Wittgenstein’s early and later work constitutes an anti-philosophy rather than any recognizable type of philosophy. And it is precisely as an anti-philosophy that Wittgenstein’s thought transcends analytic, continental and all other philosophical movements.

6. Cavell on Wittgenstein and the search for knowledge It is impossible in this space to do justice to Cavell’s richly rewarding study, The Claim of Reason. I shall touch instead only a single theme in Cavell that is related in an interesting way to my interpretation of Wittgenstein in both the early and later periods as a trans-analytic-continental philosopher (or anti-philosopher). In the ‘Foreword’ to the new edition of his book, Cavell explains: I have made no effort to sophisticate my early, tentative, efforts to link the English and Continental traditions, because I want them to show that to realign these traditions, after their long mutual shunning, at any rate to write witnessing the loss in that separation, has been a formative aspiration of mine from the earliest of the work I refer to here.15 Thus, Cavell acknowledges an aspiration to write with a heightened awareness of what he thinks has been sacrificed historically in the division between English, read analytic, and continental philosophy. He interprets Wittgenstein’s discussions of criteria primarily as criteria of judgment and knowledge, rather than as criteria of meaning and meaningfulness. It is not until part II, chapter VII that Cavell ventures an ‘Excursus on Wittgenstein’s Vision of Language’, more than one-third of the way through the book, and only as an ‘excursus’ instead of being the main event. Cavell senses on every page that there is something extraordinary about Wittgenstein’s Philosophical Investigations. Yet he never seems to put his finger on exactly what it is that makes Wittgenstein’s ‘philosophy’ so unique. He appears to think of whatever Wittgenstein is up to merely as an innovative theory of knowledge, but a theory of knowledge, despite its remarkable idiosyncracies, on all fours with other great epistemologies. Cavell offers many invaluable insights into the details of Wittgenstein’s thought. He uncovers a great deal of the pragmatism latent in Wittgenstein’s later understanding of the meaning of language, which was inadequately appreciated in the philosophical world, especially at the time of Cavell’s first edition in 1979. Cavell never, however, in my opinion, fully grasps the key purpose behind Wittgenstein’s album of language games, or his emphasis on forms of life, family resemblance predicates, criteria of correctness in naming psychological events as individual entities, the learning of language by children, the phenomenon of reading, and all the other topics of Wittgenstein’s later philosophy.

Wittgenstein: Trans-Analytic-Continental

167

He does not generally see the Wittgensteinian forest for the Wittgensteinian trees. Cavell writes in an extended passage worth reproducing in its entirety: Here I think of Wittgenstein’s remark about philosophical theses . . . What I have heard people say about this [§128] leaves out the ideas of ‘trying to advance’ and ‘not being possible’. On the surface, the remark about theses suggests that if I say, for example, ‘The world exists’, this would not be a (what we call a) thesis, because there’s no two ways about it. But not every philosopher should be expected to see this. And then, if I say ‘God exists’, I am prepared to learn that there’s two ways about that, but it doesn’t follow that what I announced was a thesis. The German runs: ‘Wollte man Thesen in der Philosophie ausstellen, es könnte nie über sie zur Diskussion kommen, weil Alle mit ihnen einverstanden wären.’ I get the following sense from those words: ‘Whatever knowledge philosophizing brings forth cannot be brought out by setting forth a claim; because such a claim could never come into question, i.e., it could never arrive at words: because it is something upon which we have an understanding, i.e., it goes without saying, and only without saying, because we could not understand anyone who claimed it and thereby held it as possibly open to question.’ That will doubtless not express everyone’s view either of the German passage or of philosophizing . . . Such a sense of Wittgenstein’s philosophizing has the virtue of heading off the idea that what he would really like, would he but tell, is to support ordinary or common sense beliefs. To confront beliefs, common or otherwise, with the human agreement in terms of which those beliefs propose to make sense – to bring anything that is said back into the basis upon which we have anything to say – is not a practice of common sense. (p. 33–34) I admit, as Cavell anticipates, that I am among those who do not at all agree with his handling of the German. I have immediate problems in particular with Cavell’s unexplained insertion of the word ‘knowledge’ in the first sentence of the translation, for which there is no justification in the original passage. One cannot but understand this as another reflection of Cavell’s reading of the later Wittgenstein through epistemic goggles. I interpret Wittgenstein rather differently, as concerned with meaning and the rules of philosophical grammar established by the point and purpose of specific culturally ingrained and pragmatically justified language games. I nevertheless agree with Cavell in his complaints about the sorts of things one frequently hears about Wittgenstein’s later renunciation of philosophical theses. Wittgenstein writes in Philosophical Investigations §128: ‘If one tried to advance theses in philosophy, it would never be possible to debate them, because everyone would agree to them.’ Wittgenstein’s meaning at first is obscure, and has remained obscure to generations of commentators. We can assume that,

168

Postanalytic and Metacontinental

whatever he means by a ‘thesis’ ‘in philosophy’, it cannot be disagreed with once articulated. What kinds of theses would these be? Unlike Cavell, I read Wittgenstein on this point as saying that a philosophical thesis would not be like the kinds of propositions one finds in a philosophy that tries to identify, describe, and systematize the truth about some choice of interesting concepts. A philosophical thesis for the later Wittgenstein could only perspicuously explicate the rules of philosophical grammar in the context of particular applications. Wittgenstein is also telling us that such perspicuous explications of the rules of philosophical grammar, by virtue of being rules of grammar, governing the ways in which we can and cannot intelligibly truly or falsely speak about relevant matters, would need to be assented to by every reflective person who hopes to say anything meaningful about the rules of philosophical grammar. This makes it pointless to offer the rules of philosophical grammar as philosophical theses, in the way that philosophical theses are usually thought of as disputable and debatable. The conclusion gives us our first glimpse of why it is pointless to try to categorize or pigeonhole Wittgenstein as a positivist, behaviourist, realist, nominalist, existentialist, modernist, postmodernist or any other kind of -ist. As Wittgenstein tells us, and reminds himself in his most anti-philosophical soliloquies, he is not advancing theses. In philosophy there is no progress to be made. Philosophy is not like the sciences in which a problem is set and solutions are developed according to an appropriate choice of methods. It is something above or below, but not on equal footing with the natural sciences, as he charges in Tractatus 4.111. This is also why efforts to link Wittgenstein with the great analytic and continental traditions inevitably fail, and why it is implausible and too simple-minded to interpret Wittgenstein as a twentieth-century Plato, to draw connections between Wittgenstein and Fichte or Heidegger, or to see Wittgenstein as continuing the projects of Frege and Russell, among others, despite points of overlap with all of these thinkers. The comparisons fail because they do not take account of the means-end distinction in Wittgenstein’s work in philosophy, and because all such interpretations try to make Wittgenstein into something he is not. What Wittgenstein most definitely is not, is a conventional philosopher in any of the recognizable conventional philosophical traditions or their rivals, or some hybrid blend of their distinctive attributes – say, a little bit of mathematical logic with some religious and ethical = aesthetic phenomenological accessories.

7. Rorty’s categories of systematic versus edifying metaphilosophy Among postmodernists, Richard Rorty in Philosophy and the Mirror of Nature arguably has the best grasp of what is unique about Wittgenstein’s thought and its relation to traditional philosophy. Rorty draws an interesting distinction

Wittgenstein: Trans-Analytic-Continental

169

between what he calls systematic (including both normal and revolutionary) thinkers, and edifying philosophers. Eventually he lumps all of conventional philosophy under the systematic head, and singles out Wittgenstein and a few other thinkers by virtue of their extraordinary metaphilosophical attitudes as edifying. Near the very end of the book, Rorty marks the distinction between systematic and edifying philosophy and philosophers in this way: On the periphery of the history of modern philosophy, one finds figures who, without forming a ‘tradition,’ resemble each other in their distrust of the notion that man’s essence is to be a knower of essences. Goethe, Kierkegaard, Santayana, William James, Dewey, the later Wittgenstein, the later Heidegger, are figures of this sort . . . The mainstream philosophers are the philosophers I shall call ‘systematic,’ and the peripheral ones are those I shall call ‘edifying.’ These peripheral, pragmatic philosophers are skeptical primarily about systematic philosophy, about the whole project of universal commensuration. In our time, Dewey, Wittgenstein, and Heidegger are the great edifying, peripheral thinkers.16 Where Rorty appears to go astray is in his choice of terminology for the distinction and in grouping Heidegger’s existential phenomenology and a select group of other philosophers, including the humdrum John Dewey, together with Wittgenstein, as ‘edifying’ philosophers. That Wittgenstein is not a systematic philosopher in Rorty’s sense is true enough, although it must be wondered why Rorty singles out only the later Wittgenstein as edifying, since the early Wittgenstein would seem to fit Rorty’s description of an anti-systematic thinker just as comfortably as the later. To designate Wittgenstein as an edifying philosopher, say, in the manner of Kierkegaard’s religious-existential philosophy misleadingly suggests that Wittgenstein regards philosophy itself as lifting thought out of its mundane surroundings to something higher and improved. As I understand Wittgenstein in both the early and later periods of his thought, this is not at all his purpose. Wittgenstein’s approach is rather not to lift philosophy up, but rather for philosophy to disappear step by step as its problems are exposed as pseudo-problems in the early period, or, in the Investigations, laid to rest through a course of therapy in which an understanding of philosophical grammar enables us to dispel the grip of improperly conceived philosophical questions on our imaginations, like a clinical psychoanalytic treatment of a paralysing intellectual syndrome (§133).

8. Wittgenstein’s anti-philosophical transcendence of analytic and continental philosophy The selective survey of Wittgenstein commentators we have considered indicates that many analytic and continental responses to Wittgenstein’s early and later

170

Postanalytic and Metacontinental

philosophy are dramatically out of sync with each other, and with what we find Wittgenstein saying candidly about what he is trying to do. At the same time, they evidently leave what appear to be key elements of Wittgenstein’s thought in the early and later period entirely out of account. These facts suggest that neither side has rightly understood the unique quality of Wittgenstein’s contributions to philosophy. With scant guidance from such recognized commentators as Cavell, Rorty or Perloff, we stand at the threshold of a need to reinterpret Wittgenstein’s early and later anti-philosophy. If Wittgenstein is finally a trans-analytic-continental philosopher, it is unlikely to be because he borrows topics, principles, concepts, distinctions or theses from both camps and then welds them into something that is simultaneously both analytic and continental. Wittgenstein instead carves out a new and individual style of engaging with the putative problems of philosophy. It seems from internal and external evidence that what has been described as Wittgenstein’s early and later philosophy is really an anti-philosophy that is ultimately means-ends directed to a larger spiritual and aesthetic end. The point of Wittgenstein’s thought is rather to overcome philosophy altogether in the service of a greater supra-philosophical understanding, and a deeply personal quest for meaning. This is true not only in the details of his philosophy, but in his very conception of philosophical inquiry and its tasks. Wittgenstein recognizes that philosophy can only be overcome from within; that it must collapse of its own weight when its respective components and their permissible combinations are correctly understood. Only then can we become free of philosophy’s devastating charms, and we can walk away with no regrets. Here we are reminded of Wittgenstein’s response to David Hilbert’s challenge in his 1925 lecture ‘On the Infinite’, declaring that: ‘No one shall drive us out of the paradise Cantor has created for us.’17 To this, Wittgenstein, in his Lectures on the Foundations of Mathematics, poignantly replies: ‘I would say, “I wouldn’t dream of trying to drive anyone from this paradise”. I would do something quite different: I would try to show you that it is not a paradise – so that you’ll leave of your own accord. I would say, “You’re welcome to this; just look about you.”’18 The marvelous detail in Wittgenstein’s writings, early and later, can often look much like conventional philosophy being pursued and developed, and certainly expressed, in an unconventional way. If I understand Wittgenstein correctly, however, then his anti-philosophy is at once something less and a great amount more, with a fundamentally different aim and orientation than can reasonably be attributed to any other thinker in the history of philosophy, analytic, continental, pragmatic, rationalist, empiricist or the like. Wittgenstein’s anti-philosophy turns philosophy against itself. Philosophy defeats itself as a consequence of its own internal deficiencies, under a correct understanding of the possibilities of meaningfulness in language limited by logic and picture theory (Tractatus) or philosophical grammatical (Philosophical Investigations) constraints. This is the sense in which Wittgenstein’s anti-philosophy, in both its

Wittgenstein: Trans-Analytic-Continental

171

early and later phases, transcends analytic and continental, and, ultimately, all other philosophy. That Wittgenstein did not succeed in these aims in either major period of his work, that he was fated to be profoundly misunderstood by conventional philosophy that seems unable to interpret his ideas except in terms of the received categories he was purposefully trying to overcome, and that his best efforts still resulted in another special kind of philosophy, are all important further chapters in the story of Wittgenstein’s radically antiphilosophical enterprise.

Notes 1

2

3

4

5

6

Dale Jacquette, Wittgenstein’s Thought in Transition (West Lafayette, IN: Purdue University Press, 1998), 160–192. Bertrand Russell, My Philosophical Development (New York: Simon & Schuster, 1959), 216–217. Wittgenstein’s reading of Kant’s Critique of Pure Reason is discussed by Newton Garver, This Complicated Form of Life: Essays on Wittgenstein (LaSalle, IL: Open Court, 1994), 91–101. Also see Rudolf Haller, ‘Was Wittgenstein a Neo-Kantian?’, in Questions on Wittgenstein (Lincoln: University of Nebraska Press, 1988), 44–56, and Dale Jacquette, ‘Haller on Wittgenstein and Kant’, in Austrian Philosophy Past and Present: Essays in Honor of Rudolf Haller, ed. Keith Lehrer and Johan Christian Marek Boston Studies in the Philosophy of Science (Dordrecht: Kluwer Press, 1997), 29–44. Recognition of Schopenhauer’s influence on Wittgenstein is found in: Erik Stenius, Wittgenstein’s ‘Tractatus’: A Critical Exposition of Its Main Lines of Thought (Ithaca: Cornell University Press, 1960), 214–226; Alan S. Janik and Stephen Toulmin, Wittgenstein’s Vienna (New York: Simon & Schuster, 1973); D. A. Weiner, Genius and Talent: Schopenhauer’s Influence on Wittgenstein’s Early Philosophy (Rutherford: Farleigh Dickinson University Press, 1992); Allan Janik, ‘Schopenhauer and the Early Wittgenstein’, Philosophical Studies 15 (1966): 96–95; M. S. Engel, ‘Schopenhauer’s Impact on Wittgenstein’, The Journal of the History of Philosophy, 7 (1969): 285–302; Hans-Johann Glock, ‘Schopenhauer and Wittgenstein: Representation as Language and Will’, in The Cambridge Companion to Schopenhauer, ed. Christopher Janaway (Cambridge: Cambridge University Press, 1999), 422–458; Dale Jacquette, The Philosophy of Schopenhauer (Chesham: Acumen Publishing Limited, 2005), 249–261. Wittgenstein’s acquaintance with some of Kierkegaard’s writings is evident in Culture and Value, trans. Peter Winch (Chicago: University of Chicago Press, 1984). The importance of Dostoyevsky in Wittgenstein’s thought is examined by John Gibson, ‘Reading for Life’, in The Literary Wittgenstein, ed. Gibson and Wolfgang Huemer (London and New York: Routledge, 2004), 109–124. Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. C. K. Ogden (London: Routledge and Kegan Paul, 1922), Section 6.54; 7. The recent growth industry of debate surrounding the so-called ‘New Wittgenstein’ has taken its point of departure from problems in interpreting 6.54. The New Wittgensteinians are sometimes described as including Cora Diamond, Thomas Ricketts, Warren Goldfarb, Juliet Floyd and James Conant; while on the other side of the debate

172

7

8

9

10

11

12

13

14

15

16

17

18

Postanalytic and Metacontinental

and often in heated opposition to the New Wittgensteinians we find the equally respected early Wittgenstein scholars, among others, P. M. S. Hacker, Jaakko Hintikka, Ian Proops, Hans Glock and Danièle Moyal-Sharrock. The New Wittgensteinians have sought to articulate an alternative reading of 6.54 that would raise doubts about the adequacy of what they regard as an orthodoxy of standard interpretations. These are the readings originally associated with Bertrand Russell, Frank Ramsey, G. E. M. Anscombe, P. T. Geach, Norman Malcolm, G. H. von Wright, Anthony Kenney, David Pears, Hacker and Gordon Baker. This in turn is to say sometime in the aftermath of the publication of Cora Diamond’s 1984 essay, ‘Throwing Away the Ladder: How to Read the Tractatus’ in Cora Diamond, The Realistic Spirit: Wittgenstein, Philosophy, and the Mind (Cambridge: MIT Press, 1996), 179–204. Ludwig Wittgenstein, Letters from Ludwig Wittgenstein with a Memoir by Paul Englemann, ed. Brian McGuinness (Oxford: Basil Blackwell, 1967), 143. See U. T. Place, ‘Is Consciousness a Brain Process?’, British Journal of Psychology 47 (1956): 44, and David M. Armstrong, A Materialist Theory of the Mind (London: Routledge, 1968), 10. Theodor W. Adorno, Against Epistemology, trans. Willis Domingo (Cambridge: MIT Press, 1982), 42. Stanley Rosen, The Limits of Analysis (New Haven: Yale University Press, 1980), 182. Marjorie Perloff, Wittgenstein’s Ladder: Poetic Language and the Strangeness of the Ordinary (Chicago: University of Chicago Press, 1996), 66–67. Newton Garver and Seung-Chong, Derrida & Wittgenstein (Philadelphia: Temple University Press, 1995). Nicholas F. Gier, Wittgenstein and Phenomenology: A Comparative Study of the Later Wittgenstein, Husserl, Heidegger, and Merleau-Ponty (Albany: SUNY Press, 1981). See also Chris Lawn, Wittgenstein and Gadamer: Towards a Post-analytic Philosophy of Language (London: Continuum, 2005). Stanley Cavell, The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy (New York: Oxford University Press, 1999), xvii. Richard Rorty, Philosophy and the Mirror of Nature (Princeton: Princeton University Press, 1979), 367–368. David Hilbert, ‘On the Infinite’. Trans. Erna Putnam and Gerald J. Massey, in Paul Benacerraf and Hilary Putnam, eds, Philosophy of Mathematics: Selected Readings (Cambridge: Cambridge University Press, 1983), 191. Ludwig Wittgenstein, Wittgenstein’s Lectures on the Foundations of Mathematics, ed. Cora Diamond (Chicago: University of Chicago Press, 1976), 103.

Chapter 11

Disenfranchising Film? On the AnalyticCognitivist Turn in Film Theory Robert Sinnerbrink

The past two to three decades have seen an extraordinary surge of interest in the relationship between philosophy and film. While film theory has retreated from psychoanalytical and semiotic forms of theorizing, philosophers have discovered the manifold joys of film. We can now speak of the ‘philosophy of film’ as an independent area of inquiry with its own methodologies, competing schools, theoretical debates and active research programmes.1 Although Stanley Cavell published works on the topic during the 1970s, contemporary philosophy of film emerges in the Anglophone world as a specialty area only during the late 1980s to mid-1990s.2 Much like the post-war emergence of analytic philosophy, the philosophy of film owes its self-definition to the rejection of what David Bordwell and Noël Carroll have called ‘Grand Theory’: the oncedominant paradigm of 1970s film theory drawing on psychoanalytic, semiotic, structuralist, Marxist, critical theory and post-structuralist perspectives.3 In what follows, I take contemporary philosophy of film – what we could call the analytic-cognitivist turn in film theory – as a case study in aesthetics that exemplifies the fraught relationship that still exists between ‘analytic’ and ‘Continental’ approaches.4 This is a problematic distinction to be sure, but one that points to two distinct ways of understanding the relationship between philosophy and its object, two ways of understanding what the task of philosophical reflection should be. While crudely mapping onto what we might call ‘analytic’ versus ‘Continental’ film theory, the crucial distinction, I will argue, is better understood as that between a theoretical explanatory approach to analysing the nature of film (philosophy of film), and an aesthetic, reflective, interpretative approach that puts philosophy in dialogue with film as an alternative way of thinking (film-philosophy). While the analytic-cognitivist turn in film theory offers some powerful explanatory theories, it also repeats what I call the ‘philosophical disenfranchisement of film’; the reduction of film to an object of philosophical analysis or the attempted integration of it into philosophical or scientific theories as superior forms of knowledge and understanding. The philosophy of film, I suggest, thus still requires the aesthetic, hermeneutic, and

174

Postanalytic and Metacontinental

critical supplementation offered by film-philosophy (F-P), an approach that treats film as capable of engaging in philosophical reflection via cinematic means, and which opens up the possibility that philosophy might be transformed through its encounter with film.

1. A (movie) house divided Perhaps more dramatically than comparable areas of the humanities (like literary theory) the discipline of film studies has been undergoing something of a paradigm shift in recent decades. Indeed, the 1996 publication of David Bordwell and Noël Carroll’s important collection of essays, Post-Theory: Reconstructing Film Studies, can be taken as marking the new paradigm’s arrival.5 Its exponents include philosophically trained film theorists (notably David Bordwell and Noël Carroll), analytic aestheticians (Richard Allen, Berys Gaut, Carroll, Paisley Livingstone and Murray Smith), and a number of cognitivist film theorists (Gregory Currie, Torben Grodal and Carl Plantinga). Whatever their theoretical orientations, the new wave of ‘post-theory’ philosophers of film defined themselves against a superseded paradigm – the institutionalized film theory of the 1970s and 80s inspired by psychoanalysis, structuralism, semiotics, cultural theory and various strands of German critical theory and French post-structuralism.6 On this score, the title of Noël Carroll’s 1988 book says it all: Mystifying Movies: Fads and Fallacies in Contemporary Film Theory.7 The new philosophy of film that challenged the prevailing model of film theory styled itself as analytic rather than ‘Continental’ in inspiration; cognitivist rather than psychoanalytic in approach; scientistic rather than hermeneutic in style; concerned with framing and testing empirical hypotheses rather than engaging in speculation or interpretation; it aimed at the rational understanding of film rather than at plumbing unconscious mechanisms of desire; and was concerned to explain our enjoyment of movies and their philosophical significance in plain language and theoretical arguments rather than via metaphysical jargon or ideologico-political analysis. With its twofold strategy of debunking the ‘fads and fallacies’ of ‘Continental-style’ film theory, and proposing analytical and empirically testable models of film theorization instead, the analytic-cognitivist turn in film theory has succeeded in establishing itself as arguably the dominant paradigm in film theory today.8 Noël Carroll’s essay ‘Prospects for Film Theory: A Personal Assessment’ provides one of the definitive statements of the critical rejection of ‘Grand Theory’, identifying the source of its dubious theoretical consequences in an uncritical commitment to various eclectic strains of continental philosophy.9 Carroll identifies five obstructions to fruitful theorizing on film that all stem from the flawed paradigm of continental theory: 1. a monolithic conception of film theory, according to which a single theoretical paradigm has to account for all relevant aspects of the complex phenomenon

The Analytic-Cognitivist Turn in Film Theory

2.

3.

4.

5.

175

of film; this is linked with an implausible ‘medium-essentialism’, which assumes that all such phenomena must be explained in light of the essential features of the film medium; the conflation of film theory with film interpretation, in which film scholars confuse theorizing about film with providing interpretations of particular films using various forms of theoretical jargon; typically, so the charge goes, film theorists assume the truth of a theoretical framework (Lacanian psychoanalysis, for example), and then proceed to ‘confirm’ the theory in question by finding its concepts or ideas instantiated in various carefully chosen film examples; political correctness, in which the progressive ethico-political claims of film theory are bolstered by underlining its solidarity with various emancipatory social-political movements of the 1960s and 70s, with the corollary that criticism of ‘Theory’ is regarded as tantamount to social conservatism or political reaction; charges of formalism, according to which any alternative version of theorizing about film that lacks a suitable ‘political’ or ideological focus, so Carroll contends, is dismissed as merely ‘formalist’ or as lacking substantive content; biases against truth, which refers to film theory’s allegiance to the alleged postmodernist dismissal of truth as an ideological construct, a claim that rests on what Carroll identifies as an untenable ‘argument from absolute truth’ (any truth claim about film presupposes an absolutist conception of truth; there is no such absolute conception of truth; ergo truth claims made about film are ideologically suspect).10 Taken together, Carroll concludes, these five obstructions have diminished the prospects of any plausible philosophical theorization of film; hence the need for film theory to make the transition to a more theoretically congenial paradigm – the analytic-cognitivist philosophy of film.11

A more aggressive attack on the pernicious influence of continental philosophy on film theory can be found in Richard Allen and Murray Smith’s ‘Film Theory and Philosophy’, the Introduction to their 1997 collection of essays on analyticcognitivist approaches to film. For Allen and Smith, as for Bordwell and Carroll, the parlous state of film theory can be blamed on the sophistical and dogmatic character of the continental philosophy that inspired it.12 To this end, they identify three basic strategies in continental thought that have had a deleterious effect on film theory: 1. ‘a disabling habit of deference and quotation with respect to authoritative figures’, such as the oft-criticized tendency towards uncritical dependence on master theorists, like Freud or Lacan, with little critical analysis or argumentation to support their respective theories;13

176

Postanalytic and Metacontinental

2. ‘the misleading use of examples (and a related confusion between analogy and example),’ according to which continental-style film theorists make generalizations on the nature of cinema on the basis of a single cinematic exemplar; for example, claiming that film in general displays a ‘sadistic’ gaze on the basis of a psychoanalytic reading of certain Hitchcock films, a move which the authors dub the ‘fallacy of exemplification’;14 3. ‘a strategic use of obscurity’, which is to say deliberate but unacknowledged obfuscation;15 for example, the continental theorist’s alleged reliance on punning and conceptual free association; a pernicious proclivity towards pseudo-argumentation that runs ‘counter to the clarificatory impulse of the analytic tradition’. Film theory’s underlying ‘psycho-semiotic theory of subject positioning’, according to Allen and Smith, has a dual provenance, deriving from Hegel, Sartre and psychoanalytic thought on the one hand, and from Husserl, Derrida and Kristeva on the other.16 On this basis, they develop a fundamental critique of ‘Continentally inspired film theory’ for its uncritical dependence upon the following three assumptions. 1. According to continental philosophers, the purpose of philosophy is to critique epistemology; continental philosophy is thus committed to a sceptical (and self-refuting) denial of truth, with a correlated scepticism towards knowledge claims more generally, failings that are dutifully transmitted to film theory; 2. Continental philosophy’s critique of epistemology is identified with the ethical critique of modernity; this sceptical critique of truth and knowledge is undertaken in the name of an ethical critique of the ideological dimensions of supposed truth claims and their masking of hidden power relations or vested interests; 3. the ethical critique of modernity in continental film theory is promoted through a theorization of cinema as a model of, or process governing, knowledge and/or the critique of knowledge. Due to the psychoanalytic emphasis on the unconscious processes structuring our desire, coupled with film theory’s claims concerning the ideological manipulation of consciousness through the cinematic apparatus, film is taken to be a privileged site for the aforementioned critique of truth, subjectivity and ideology, as well as providing an antidote, in the form of modernist cinema, for the ideological effects of mainstream Hollywood film.17 There is no doubt a grain of truth in some of these claims concerning the failings of certain strands of ‘Grand Theory’. Carroll is right, for example, to criticize the older film theory’s tendency towards tiresomely circular repetitions of psychoanalytic and/or ideological readings of carefully chosen films, and to

The Analytic-Cognitivist Turn in Film Theory

177

question its lazy eschewal of more recent work in empirical psychology or in other areas of the humanities and sciences that might advance our understanding of film (not to mention the shameful neglect of film aesthetics). Nonetheless, the dismissal of any more considered reflection on the aims, methodology and content of prevailing forms of film theory reflects a certain resistance to the idea that there can be alternative ways of responsibly going on in philosophy. The more strident claims against continental film theory enumerated above, for example, are given cogency through the use of caricatures of the theoretical positions being attacked, emphasizing their dogmatism, obscurantism and intellectual incoherence (such as the claims made concerning Derrida’s ‘epistemic atheism’). We are told, for example, that continental philosophers such as Theodor Adorno, Jacques Derrida and Michel Foucault have all attempted to demonstrate ‘the impossibility of knowledge’; despite the self-undermining nature of this claim, they have, moreover, ‘embraced this contradiction as the defining feature of philosophy and the only legitimate path that philosophy can take in response to modernity’.18 I would challenge the authors of this claim to produce convincing textual evidence to support the assertion that Adorno, Derrida and Foucault argue for ‘the impossibility of knowledge’, and hence freely embrace ‘self-contradiction’ as the only legitimate philosophical response to modernity. As radical heirs to the post-Kantian, post-Hegelian tradition of European philosophy, these thinkers argue, rather, for expanding our conception of ways of knowing and understanding – emphasizing how art, literature, even politics should be included in an expanded conception of reason – and to this end explore diverse, alternative ways of engaging in philosophical critique as an ethically motivated response to the ‘nihilism’ of modern rationality. Unfortunately, Allen and Smith bluntly assert the irrationality of these thinkers with little argumentative evidence. What is striking in both Carroll’s, and Allen and Smith’s, attack on continental film theory is the way they rehearse what Simon Glendinning has identified as the caricatured dismissal of continental philosophy promulgated by the defenders of the analytic movement in the 1950s (such as Gilbert Ryle, R. M. Hare and Geoffrey Warnock).19 Indeed, the strident dismissal of continental film theory is a good example of the analytic construction of continental philosophy as mere sophistry, rhetoric and poetry, as opposed to the rationality, moral probity and intellectual integrity of analytic philosophy. As an example of this kind of caricatural dismissal, we might consider R. M. Hare’s remarkable lecture on British philosophy, ‘A School for Philosophers’, delivered to various German audiences in 1957.20 Like Gilbert Ryle, Hare contrasts two ways of doing philosophy, as different as chalk and cheese, namely British philosophy (as practised in Oxford and Cambridge) and German philosophy (Heideggerian phenomenology). Deploying a series of rhetorical tropes that have become depressingly familiar (witness Carroll, Allen and

178

Postanalytic and Metacontinental

Smith), Hare valorizes the British intellectual virtues of ‘clarity, relevance, and brevity’, where arguments can be defended or refuted concerning philosophical works that present an ‘unambiguously stated thesis’.21 These edifying Oxbridge virtues stand in stark contrast to the Germanic propensity to indulge in creating ‘monstrous philosophical edifices’ that produce ‘verbiage’ disguised as ‘serious metaphysical inquiry’, and that are expressly designed to turn ‘philosophy into mystique’.22 Such Teutonic tomes, Hare intones, aim to seduce and intoxicate followers, creating ‘a private coterie’ dedicated to promulgating the master’s word. Hare thus summarizes German philosophy – by which he seems to mean Heidegger – as characterized by ‘ambiguities and evasions and rhetoric’; precisely those characteristics that ‘British philosophers’ regard as ‘the mark of a philosopher who has not learned his craft’.23 We need only substitute ‘Grand Theory’ or ‘Continental film theory’ for ‘German philosophy’ in these passages in order to produce a passable model of the Carroll/Allen and Smith attack on continental film theory. Sadly, the fact that caricature and cant are not what the ‘British philosopher’ is supposed to be about, that these represent a serious lapse from the ‘rigorous standards’ of argument defining his craft, seems to have curiously escaped Hare’s attention, not to mention that of his epigones. My point here is not simply to question the stereotypical dismissal of continental philosophy for its obscurity, dogmatism and incoherence, but rather to suggest that there are deeper issues at stake in the construction of film theory as hopelessly muddled and disreputable. In a gesture that Simon Glendinning has identified as constitutive to the analytic/continental divide, analytic philosophy of film has largely defined itself in opposition to continental film theory. In so doing it rehearses the way the post-war analytic movement defined itself in opposition to then contemporary ‘Continental’ style of philosophizing. Indeed, continental philosophy, as Glendinning argues, is defined by analytic philosophy as that which is not-part of philosophy as such (namely sophistry, rhetoric, poetry); hence as that which must be expelled, precisely in order to shore up the identity of analytic philosophy, which is then identified with the inheritance of philosophy as such.24 This dispute concerning the legacy of the philosophy of film today points to deeper questions concerning the nature of philosophy, and what the relationship between philosophy and its object should be. Is philosophy primarily an explanatory enterprise, producing testable theories in the manner of a rigorous science or contributing to a naturalistic explanation of, in this case, artistic and cultural phenomena? Or is it rather an interpretative, hermeneutic, reflective enterprise, modelled on humanistic forms of inquiry, and hence concerned to reflect on what we experience, to enhance our understanding of meaningful phenomena, or to question our assumptions regarding theories of human nature? What is at stake is the possibility of philosophical pluralism: that whatever else philosophy claims to be, it can also be an anti-naturalistic form of inquiry that argues for the specificity of cultural phenomena – such as art and film – that demand reflection on questions of meaning and value that cannot

The Analytic-Cognitivist Turn in Film Theory

179

be exhaustively explained by naturalistic or scientific explanations. Not that these are necessarily mutually exclusive aims; in the best cases philosophy can successfully combine both perspectives, or bring them into productive tension, reflecting upon the scope and limitations of each, and perhaps offering a sense of how they might better interact. It is rather that the analytic-cognitivist turn in film theory, at least in its more strident, partisan guise, seeks to eliminate this alternative approach from the discussion altogether, or to discredit any meaningful philosophical contribution it might make to our understanding of film, not to mention aesthetic experience and cultural meaning more generally.

2. From philosophy of film to film-philosophy Does this dispiriting fable of the vanquishing of decadent, continental-style film theory by analytic philosophy of film tell the whole story? Fortunately there is another way of theorizing the relationship between philosophy and film than the rather rigid opposition between continental and analytic approaches that I have been discussing thus far. Here we can identify three major currents: 1. the pioneering work of Stanley Cavell – the first Anglophone philosopher of note to take film seriously enough to devote entire books of philosophy to its varieties – and what we might call the Cavellian school that is now emerging in his wake, the most notable figure here being Stephen Mulhall;25 2. the philosophical work on cinema undertaken by French philosopher Gilles Deleuze, and the Deleuzian school now gathering momentum, elaborating Deleuze’s monumental Cinema books – two volumes on the movement-image and the time-image respectively – with reference to contemporary forms of cinema and film theory;26 3. the Wittgensteinian philosophers of film who take as their inspiration Wittgenstein’s pragmatism, his therapeutic dispelling of the rigidified metaphors and fixed habits of mind that mislead our thinking, in order to return us to the everyday, now transfigured by philosophical reflection; film offers an experience of precisely the kind of reflection on and retrieval of the ordinary that prompts philosophical reflection of this kind.27 What brings together these disparate approaches to film is their rejection of the idea that we need a philosophically (or cognitively) grounded theory of film; one that takes film theory to be a species of applied epistemology or applied philosophy of mind (or of cognitive science or of evolutionary biology). Constructing such theories is one of the primary tasks of what I shall call the ‘philosophy of film’, a title that resonates with the more familiar forms of the ‘philosophy of X’ (e.g. of science, of law, of biology, of economics and so on). In the ‘philosophy of X’ approach, philosophy reflects upon, analyses, and theorizes about its object, precisely because its object cannot or will not engage

180

Postanalytic and Metacontinental

in this kind of specifically conceptual self-reflection on its own (unless prompted by philosophers to do so). Call this ‘philosophy of X’ model of theorization the Philosophy of Film approach (PF). PF can be understood as a traditional philosophical ‘theory of X’ that seeks to provide, depending on one’s orientation, a conceptual definition of, empirical investigation into, or philosophical criticism aiming at theories claiming to account for X (where ‘X’ means film, motion pictures, moving images and so on). To this end, PF engages in high-level theorization on film as its object, exploring questions such as whether film is an art and how it is to be distinguished from cognate artforms (photography, theatre, painting), or what the ontology of the moving image might be. It might investigate the way our perception of images works (drawing, say, on empirical psychology or cognitive science), or analyse how film narrative is understood (for example, David Bordwell’s and Noël Carroll’s puzzle-solving or ‘question and answer’ models of narrative). Or it may reflect on why film seems more immediately intelligible than, say, poetry or painting, how it appears to emulate natural perception rather than relying primarily on the application of learned conventions, and so on. This is a powerful and flourishing research paradigm that is making significant contributions to our understanding of film. But it is also one that is clearly attempting to eliminate its rivals, those approaches that challenge the idea that what understanding film requires above all is a theory of film, one that (analytic-cognitivist) philosophy is best equipped to provide. The rival position is what I shall call ‘Film-Philosophy’ (F-P), a style of philosophical reflection on and with film that questions the common tendency to philosophically privilege conceptual theorization over film aesthetics.28 F-P argues that film should be regarded as engaging in philosophically relevant reflection via the medium of film, or even as engaging in a distinctively cinematic kind of thinking. It is a way of aesthetically disclosing, and thereby critically transforming, our experience of the modern world; one that raises philosophical questions in its own right, and even forces philosophy to reflect upon its own nature and limits. In contrast to the PF model (defined as a philosophy of film), F-P (defined rather by the relation between film and philosophy) does not presuppose superior knowledge of its object, but engages rather in a mutual reflection or dialogue with its object with the aim of responding to film’s nature, significance, and possibilities in such a way as to open up a space of mutual reflection and understanding. In what follows, I want to defend this alternative philosophical approach to film, and to argue that it is a legitimate, even necessary approach, particularly in its endorsement of the possibility that philosophy might be transformed by its encounter with film. The advantage of distinguishing PF from F-P, I suggest, is that it avoids the temptation of mapping the two styles of film philosophizing directly on to the analytic/continental divide, as though ‘Grand Theory’ and its descendants simply correlated with continental philosophy, while ‘post-Theory’

The Analytic-Cognitivist Turn in Film Theory

181

philosophy of film corresponded to analytic and/or cognitivist approaches. While there is some undeniable overlap between these two camps – the new philosophy of film drawing on analytic and cognitivist approaches clearly defines itself against continental-style film theory – there are also significant differences and crossovers (some ‘Continental’ philosophers engage in theorization typical of the PF approach, while there are theorists aligned with analytic philosophy or even cognitivism who also draw on work from the ‘Continental’ tradition). The important difference between PF and F-P, I want to suggest, is that the former subordinates film as the object of theoretical analysis or explanation, while the latter places film and philosophy on the same level as engaging in a mutually transformative reflection. Put differently, we could say that PF is analytic, conceptual and explanatory in nature, whereas F-P is reflective, hermeneutic and aesthetic in orientation. Once again, there will be interesting cases of overlap – for example, Deleuze’s film theory purports to think the concepts of the cinema but does so in an aesthetic-constructivist manner – but the distinction I am proposing enables us to acknowledge explanatory models of film without condemning reflective, hermeneutic or aesthetic-constructivist approaches to philosophical perdition. To elaborate these claims, I will draw on Arthur Danto’s famous thesis concerning the ‘philosophical disenfranchisement of art’, arguing that Danto’s thesis is also applicable to contemporary PF.29 The point is not to deny the legitimacy of this theoretical enterprise, but rather to point out its presuppositions, its limitations or omissions, and to suggest the need for a more pluralist stance when it comes to philosophizing on, about, or with film. Such a stance is ably represented by F-P; it provides a salutary way of overcoming PF’s tendencies to philosophically disenfranchise film, and thereby opens up the possibility of a philosophical re-enfranchisement of film.30

3. Disenfranchising film? The remainder of my discussion will reflect primarily upon the question of how to think the relationship between film and philosophy, a question invoking the vexed question of the relationship between philosophy and art. Here we might ask a question: why the surprisingly happy nuptials between film and philosophy, when the history of the relationship between philosophy and art, historically speaking, has been so unhappy, even antagonistic? Such a question assumes, of course, that we know what philosophy is; that we can distinguish it from non-philosophy, from that which pretends to be philosophy but is merely sophistry, rhetoric or poetry. This is, in effect, the founding gesture of philosophy with the Greeks; the distinguishing of philosophy from its rival claimants to knowledge and truth. And it continues to be a defining gesture of much mainstream philosophical discourse, whether in the ‘analytic’ versus ‘Continental’

182

Postanalytic and Metacontinental

debate (with the latter playing the role of sophistry, rhetoric or poetry), or in the way mainstream philosophy maintains a certain aloofness from other humanities disciplines, while at the same time defending its privileged relationship with the sciences (e.g. the recent alignment of film theory with cognitive science and dismissal of psychoanalysis and literary theory). This exclusionary self-definition of philosophy and its others continues in recent attempts to bring philosophical and scientific legitimacy to the interdisciplinary matrix of film theory and the philosophy of film. We might even say that there is a kind of activism evident among philosophers of film today, marching confidently into the hitherto foreign territory of film speculation and engaging in a kind of ‘humanitarian intervention’ to civilize the barbarians. Like ‘academic SAS squads’, as Daniel Frampton quips, our brave philosophers ‘come in to sort out the mess left by film studies’.31 The combination of dispassionate reason and cultural-political activism in Frampton’s witty simile is highly pertinent here. For what we are dealing with is precisely what I am calling the philosophical disenfranchisement of film. Now, as is well-known, Plato (or rather Socrates) tells us in The Republic that there is ‘an ancient quarrel between philosophy and poetry’.32 In Book X, Plato famously criticizes art (techne) and poetry (in a broad sense encompassing literary and dramatic narrative) for its epistemic deceptiveness, its irrational appeal to the senses and the emotions, hence its pedagogical and moral dubiousness. Socrates famously does dialectical battle with the sophists, pitting his ironic dialectic and rational search for truth against rhetorical persuasion and the seduction of appearances. Even the famous image of Plato’s cave, as numerous theorists have observed, bears an uncanny resemblance to the cinema. In any event, for Plato, the founding gesture defining the philosophical grounding of the polis was essentially prejudicial, based upon a series of defining exclusions: a prejudice against sensuous perception, against the emotions, against rhetoric, art and poetry, that is, against non-conceptual forms of experience and understanding – all of which culminate in Plato’s famous exclusion of the poets. Such was the warrant for the philosophical and moral-political censorship of art, which in turn laid the foundation for the subsequent philosophical subordination of art and poetry to conceptual knowledge as the privileged means of attaining rational mastery over nature and ensuring the rational governance of society. The Platonic attack on art, which neutralizes it by arguing against art’s epistemic inferiority and moral dangerousness, sets the scene for art’s subsequent philosophical disenfranchisement. Indeed, this philosophical neutralization and subordination of art, as Danto argues, is effectively the defining gesture of philosophy as such: for if art can be transferred ontologically to the sphere of secondary and derivative entities – shadows, illusions, delusions, dreams, mere appearances, and sheer reflections – well, this is a brilliant way to put art out of harm’s way

The Analytic-Cognitivist Turn in Film Theory

183

if we can get people to accept a picture of the world in which the place of art is outside it. And since Plato’s theory of art is his philosophy, and since philosophy down through the ages has consisted in placing codicils to the platonic testament, philosophy itself may just be the disenfranchisement of art – so the problem of separating art from philosophy may be matched by the problem of asking what philosophy would be without art.33 The contradictory characterization of art as both ineffectual (art is mere illusory representation) and as dangerous (art’s power of illusion can corrupt our rational judgment) is a symptom of the traditional philosophical suspicion and prejudice towards art and its aesthetico-political power. In fact the view that art is ineffectual (that it lacks power to effect change in the world) is precisely a response to the worry that art is dangerous (that it has the power to effect precisely such a change). As Danto explains, the Platonic attack against art consists in two related stages, both of which are evident in Plato, and elaborated further by philosophers from Kant to Carroll: (1) art is degraded to a deficient and defective way of knowing reality (‘reality is logically immunized against art’); and (2) art is then subsumed within philosophy but as an inferior way of knowing (the rationalization of art through which ‘reason bit by bit colonizes the domain of feelings’).34 Philosophy to this today has persistently maintained this battle against art, what Nietzsche called the triumph of aesthetic Socratism that precipitated the death of tragedy, eventually culminating in the happy nihilism of modern culture. How has this disenfranchisement proceeded? For Danto, philosophy disenfranchises art via two distinct but related strategies: (1) the ephemeralisation of art (paradigmatic here is the Kantian aestheticist view art is only truly art insofar as it is the object of a disinterested pleasure); and (2) the philosophical takeover of art (paradigmatic here is the Hegelian view that modern art becomes self-reflexive and is eventually superseded by the philosophy of art, which comprehends what art actually is). On the one hand, we have an aesthetic formalism coupled with a privileging of the ‘aesthetic attitude’ that together neutralize art’s cognitive as well as its moral-political power; and on the other, we have the subsuming of art within the edifice of conceptual knowledge, effectively relegating art to a ‘primitive’ form of knowledge or inadequate form of understanding that is destined to be completed by philosophical (or scientific) analysis. From this point of view, Danto argues, we can begin to see art as ‘the reason philosophy was invented’.35 Philosophy, from its inception, seeks to protect us from the deceptions and seductions of art, as well as from its ‘pseudophilosophical’ guises in rhetoric and sophistry. Art can exist as the object of philosophical analysis provided it is ‘domesticated’ by rational understanding. If this is so, then we might come to see philosophical systems, Danto remarks, as ‘penitentiary architectures it is difficult not to see as labyrinths for keeping

184

Postanalytic and Metacontinental

monsters in and so protecting us against some deep metaphysical danger’.36 The question, of course, is what kind of “metaphysical danger” art is presumed to pose. Traditionally, the defining ‘classical’ philosophical anxiety about art is that it concerns mere appearances, that it engages in deception or plays with semblances, that it appeals to the senses, feeling, the emotions or non-conceptual forms of understanding, that it articulates the world ambiguously, equivocally, without certainty or exactitude – in short, that it resists ready translation into philosophy, itself defined as the other of art, rhetoric and poetry. This gesture of philosophical disenfranchisement established the relationship between philosophy and art as inherently antagonistic, exclusionary and irreconcilable. We might call this the birth of aesthetics out of the suppression of art; a long history of conceptual mastery over art, of keeping it – philosophically and politically – in its proper place. Even today, the Minotaur of art remains safely sequestered within a conceptual labyrinth carefully constructed for it by so many philosophical Daedaluses. The name of this labyrinth, since Kant’s Critique of Judgment, is aesthetics; and the interesting hybrid beast it now attempts to domesticate is called the cinema. My contention here is that contemporary philosophy of film repeats the philosophical disenfranchisement of art by theoretically reducing the polyvalent medium of film to a univocal conceptual discourse. This gesture maintains the hierarchy of philosophy over art, curtailing the possibility of film being capable of independent philosophical reflection, or even questioning philosophy from the perspective of cinematic art. (As an example, we might cite philosophy’s repeated questioning whether film is really ‘philosophical’, which assumes that we know what philosophy is or should be and are authorized to speak in its name, rather than asking whether genuinely engaging with film transforms what philosophy might be, indeed whether the encounter with art might force philosophy to question its meaning, nature and possibilities). The problem with such approaches – as Simon Critchley remarks of philosophical readings of Samuel Beckett’s work – is that they say both too much and too little.37 The philosophical overtaking of film says too much by translating the cinematic work into a conceptual meta-language that foregrounds the ‘theoretical’ concerns of the film, its ‘argument’ or ‘position’ on various philosophical problems, or its role as helpful example illustrating such problems, thus ignoring the specifically cinematic aspects of the film. And it says too little by simultaneously limping behind the film, clumsily translating themes into ‘philosophese’ that are already articulated in complex non-conceptual ways through cinematic means (the typical emphasis on narrative, action, and character in philosophical readings of film at the expense of visual, aural, affective, temporal and symbolic dimensions). As more generally with the relationship between philosophy and art, the philosophical disenfranchisement of cinema can be seen as a response to the alleged ‘metaphysical danger’ provoked by film. Here we can point to cinema’s

The Analytic-Cognitivist Turn in Film Theory

185

perceptual, affective, non-conceptual forms of presentation, or even its rhetorical form of ‘argumentation’ through (visual and temporal) re-description; the revealing and thematizing of alternative descriptions or points of view upon a recognizable yet unfamiliar or novel world. For the singular power of cinema as a visual, aural and temporal artform, I would suggest, lies in its creation of unique film-worlds that we temporarily inhabit during the immersive experience of film viewing (further compounded by its powerful imprinting of individual and collective memory). It is perhaps this immersive experience of the cinema – the relinquishing of one’s engaged perception, thought and action in favour of a sensuous receptivity and reflective openness towards a virtual cinematic world – that philosophy finds so fascinating yet unsettling, philosophically rich yet anxiety-inducing, about the cinema. It’s Plato’s cave all over again. Underlying this philosophical anxiety, expressed through the disenfranchisement of cinema, is the Platonic schema that hierarchizes conception over perception, and that suspects (cinematic) art because of its aestheticopolitical power, its inherently democratic potential as the egalitarian, technological art of modernity. Given the overtly ‘political’ context of much film studies – the conjunction of Freudian-Lacanian psychoanalytical theory, feminist analysis, Althusserian-Marxist theories of ideology – we might also venture the hypothesis that the philosophical disenfranchisement of cinema is an attempt to quarantine film’s ethico-political dimensions. Whereas film theory was once at the centre of debates on desire, gender, subjectivity, not to mention ideology, today one can safely venture through various philosophical discussions of film and only rarely encounter such anachronistic ‘political’ dinosaurs – as though the phenomena that earlier theorists attempted to understand through film have been happily dissolved with the banishment of continental-style theory. Unfortunately, in sequestering the ‘extra-philosophical’ dimensions of film in this way, philosophy of film remains stubbornly unchanged, fixed in its reactive self-definition. Far from being transformed through its encounter with film, philosophy only reinforces its own exclusionary self-definition. It is in this sense that contemporary philosophy of film runs the risk of participating in the disenfranchisement of cinema: quarantining the ethico-political, affective, imaginative exploration of cinematic experience, and traducing cinema’s aesthetico-political power by subsuming it into various kinds of technical and theoretical meta-languages. To return to my earlier contrast between PF and F-P, we can see that the danger that PF courts is to disenfranchise film philosophically: subordinating film as an inferior, inadequate, or inarticulate form of understanding, one in need of philosophical clarification, conceptualization and explanation. This disenfranchisement of film is then overlaid by the dismissal of continentalinspired film theory, which is construed as ideologically-freighted sophistry,

186

Postanalytic and Metacontinental

a pseudo-theorization of film that needs to be purged through analysis and/ or science. The alternative approach, F-P, tries to reverse this philosophical disenfranchisement of film. It does not presuppose the conceptual superiority of philosophy over cinema, avoids the reduction of film to whatever translates into a philosophical meta-language, and opens philosophy up to transformation through its encounter with film. It suggests the possibility of engaging in a kind of philosophy that no longer needs to subordinate or disenfranchise art. Retrieving the possibility of philosophical engagement with art – with film – that re-enfranchises, rather than disenfranchises it, that is transformative rather than reductive, that is oriented towards the disclosure of new aspects of experience rather than the explanation of what we already know, that takes seriously the possibility that there are things that art, indeed film, can show that cannot be said in philosophy, or that philosophy can learn to say only by allowing itself to be affected by what film can do – such would be the promise of a philosophical re-enfranchisement of film.

Notes 1

2

3

4

5

6

7

See Thomas Wartenberg, ‘Philosophy of Film’, Stanford Encyclopedia of Philosophy: http://plato.stanford.edu/entries/film/; revised November 2008. 411. See Stanley Cavell, The World Viewed: Reflections on the Ontology of Film (Cambridge, MA.: Harvard University Press, 1979), and the work of Noël Carroll. See David Bordwell and Noël Carroll, eds, Post-Theory: Reconstructing Film Studies (Madison: University of Wisconsin Press, 1996). I am concerned here with the debate that has emerged between self-identified philosophers of film and film theorists whose work is explicitly informed by philosophy. While using the generic term ‘film theory’ I confine my discussion to this crossover between film theory and philosophy rather than referring to ‘film theory’ in general. Noël Carroll’s earlier essay, ‘The Power of Movies’. Daedalus, 114 No. 4 (1985): 79–104, outlined the research program that he was to develop in the following decade. David Bordwell’s Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985) pioneered the cognitivist approach. See also Gregory Currie’s Image and Mind: Film, Philosophy, and Cognitive Science (New York: Cambridge University Press, 1995). Carroll defines ‘Theory’ as ‘a classy continental number, centrally composed of elements of Louis Althusser, Jacques Lacan, and Roland Barthes, often with optional features derived, often incongruously, from Michel Foucault, Julia Kristeva, Pierre Bourdieu, Gilles Deleuze, and (maybe sometimes) Jacques Derrida . . .’. Noël Carroll, ‘Prospects for Film Theory: A Personal Assessment’, in Post-Theory, David Bordwell and Noël Carroll, eds (Madison: University of Wisconsin Press, 1996), 37. Noël Carroll, Mystifying Movies: Fads and Fallacies in Contemporary Film Theory (New York: Columbia University Press, 1988). See also Carroll’s Philosophical Problems of Classical Film Theory. Princeton: Princeton University Press, 1988.

The Analytic-Cognitivist Turn in Film Theory 8

9 10 11 12

13 14 15 16

17 18 19

20 21 22 23 24 25

26

27

187

My use of the hybrid rubric ‘analytic-cognitivist’ follows the Introductions to the volumes edited by Bordwell and Carroll (1996), and Allen and Smith (1997), which defined the new philosophy of film as a) anti-‘Continental’, b) as drawing on the techniques of analytic philosophy, and c) as inspired (for some) by developments in cognitive psychology as a rival to psychoanalytic theory. It does not mean that all such theorists are ‘analytic’ or ‘cognitivist’ in orientation. Carroll, ‘Prospects for Film Theory’, 37–68. Ibid., 38–56. Ibid., 56–68. Richard Allen and Murray Smith, ‘Introduction: Film Theory and Philosophy’, in Film Theory and Philosophy, ed. Richard Allen and Murray Smith (Oxford: Oxford University Press, 1997), 1–35. Ibid., 6. Ibid. Ibid. Ibid., 9. Allen and Smith propose an implausible genealogy of ‘Continental’ film theory: they mention Sartre, who is completely absent from ‘Continental’ film theory, as is Hegel; Husserl and Derrida are cited as progenitors of one key tradition, even though Husserlian phenomenology has been largely ignored, and Derrida’s work is virtually absent; Kristeva is influenced by psychoanalytic feminist thought and has little connection with either Husserl or Derrida. This ‘straw-man’ version of ‘Continentally inspired film theory’, however, allows Allen and Smith to attack the ‘epistemic atheism’ they attribute to Derrida, and by implication, to all continental film theorists. Allen and Smith, ‘Introduction’, 11–22. Ibid., 10. Simon Glendinning, The Idea of Continental Philosophy: A Philosophical Chronicle (Edinburgh: Edinburgh University Press, 2006), 69–84. Ibid., 74–78. Ibid., 74–75. Ibid., 75. Ibid., 75. Ibid., 12–13, 83–84. See Stanley Cavell, Pursuits of Happiness: the Hollywood Comedy of Remarriage (Cambridge, MA: Harvard University Press, 1981); Cavell, Themes out of School: Effects and Causes (Chicago: University of Chicago Press, 1988); Cavell, Contesting Tears: the Melodrama of the Unknown Woman (Chicago University of Chicago Press, 1996) and Stephen Mulhall, On Film, Second expanded edition, 2008 (London: Routledge. 2001). See Gilles Deleuze, Cinema 1: The Movement-Image, trans. Hugh Tomlinson and Barbara Habberjam (Minneapolis: University of Minnesota Press, 1985); Deleuze, Cinema 2: The Time-Image, trans. Hugh Tomlinson and Robert Galatea (London: Athlone Press, 1989); Gregory Flaxman ed. The Brain is the Screen: Deleuze and the Philosophy of Cinema (Minneapolis: University of Minnesota Press, 2000) and David N. Rodowick, Gilles Deleuze’s Time Machine (Durham NC: Duke University Press, 1997). See Rupert Read and Jerry Goodenough, eds, Film as Philosophy: Essays on Cinema after Wittgenstein and Cavell (Basingstoke/New York: Palgrave, 2005).

188 28

29

30

31 32 33 34 35 36 37

Postanalytic and Metacontinental

I would include here, for example, the work of Leo Bersani and Ulysse Dutoit, Stanley Cavell, Gilles Deleuze, Daniel Frampton, Stephen Mulhall, D. N. Rodowick, Jacques Rancière and Vivian Sobchack. See Arthur Danto, The Philosophical Disenfranchisement of Art (New York: Columbia University Press, 1986). I discuss this idea in ‘Re-enfranchising Film: Towards a Romantic Film-Philosophy’, in Film and Philosophy, Havi Carel and Greg Tuck, eds (Basingstoke/New York: Palgrave MacMillan, 2011). Daniel Frampton, Filmosophy (London: Wallflower Press, 2006), 9. Plato, Republic, trans. Benjamin Jowett (Mineola, NY: Dover, 2000), 607b5–6. Danto, The Philosophical Disenfranchisement of Art, 7. Ibid., 6. Ibid., 12. Ibid. See Simon Critchley, Very Little . . . Almost Nothing: Death, Philosophy, Literature (London: Routledge, 2004), 165–169.

Chapter 12

The Possibility of German Idealism after Analytic Philosophy: McDowell, Brandom and Beyond Paul Redding

The late Richard Rorty was no stranger to provocation, and many an analytic philosopher would surely count as extremely provocative comments he had made on Robert Brandom’s highly regarded book from 1994, Making It Explicit.1 Brandom’s book was, Rorty asserted ‘an attempt to usher analytic philosophy from its Kantian to its Hegelian stage’.2 The reception of Kant within analytic philosophy has surely been, at best, patchy, but if it is difficult to imagine exactly what Rorty could have had in mind by analytic philosophy’s ‘Kantian phase’, the idea of an immanent Hegelian one would strike many as ludicrous. Given that the beginnings of analytic philosophy are conventionally described in terms of the radical break initiated by Russell and Moore with the Hegel-inspired idealism of their teachers at Cambridge in the closing years of the nineteenth century, the distinctly anti-Hegelian character of analytic philosophy has been held to be central. Moreover, the increasing naturalistic tenor of recent analytic philosophy would seem hardly propitious for a revival of nineteenth-century idealism. And yet Rorty’s description should not be dismissed as mere provocation. While Brandom’s references to Hegel in Making It Explicit are only fleeting, what he does say there clearly signals the view that Hegel should not only be seen as one of the key figures in the development of modern philosophy, but also as particularly relevant to the contemporary state of core disciplines like logic and semantics. Moreover, neither was Brandom alone in the 1990s in expressing such a view, as 1994 had also witnessed another highly regarded analytic philosopher professing a type of ‘Hegelian turn’. Thus John McDowell in Mind and World, also somewhat fleetingly, appealed to Hegel’s ‘absolute idealism’ as offering solutions to problems at the heart of contemporary analytic philosophy.3 Moreover, the claims of McDowell and Brandom were not unprecedented. From the side of Hegel scholarship, in 1975 Charles Taylor had portrayed Hegel as having many intelligible and interesting things to say to contemporary philosophy despite what Taylor took to be his metaphysics,4 but

192

Postanalytic and Metacontinental

from the late 1980s and early 90s, a quite new take on Hegel had been coming into view which promised the good bits of Taylor’s Hegel without the bizarre metaphysical commitments. In Robert Pippin’s path-breaking Hegel’s Idealism of 1989,5 Hegel was seen as taking the idealism of Immanuel Kant in a direction something like the direction in which analytic philosophy had been taken by the mid-twentieth century by the likes of Wittgenstein and Sellars. While Sellars was only mentioned in passing by Pippin, Terry Pinkard, in Hegel’s Phenomenology (1994) utilized Sellars’s ideas more extensively. Given that it is Sellars’s work that has inspired the Hegelian turns of both McDowell and Brandom,6 perhaps it is just possible to think of analytic philosophy as capable of an ‘Hegelian stage’, and of Hegelian philosophy as capable of an ‘analytic form’. In the following section I briefly map the paths by which McDowell and Brandom have found their way to Hegel.

1. The paths from Pittsburgh to Berlin The egg from which this strange mutation of ‘analytic Hegelianism’ has started to hatch is, undoubtedly, Wilfrid Sellars’s well-known epistemological critique of the ‘myth of the given’, as set out in his ‘Empiricism and the Philosophy of Mind’. The target to Sellars’s critique was the type of ‘givenness’ with which many are familiar from Russell’s seminal work from 1912, The Problems of Philosophy.7 Like traditional empiricists, Russell had appealed to immediate knowledge of or ‘acquaintance’ with sensory givens (‘sense data’) in order to provide certain foundations for empirical knowledge of the world, but Sellars pointed out the difficulties of doing this in the context of a philosophy being transformed by the adoption of the type of Fregean logic that Russell himself had advocated and developed. Very broadly, Russell’s appeal to sense-data was an appeal to the ultimate status of just the sorts of independent ‘things’ [Dinge] that Wittgenstein had attacked in the opening sentences of the Tractatus Logico-Philosophicus when he declared that the world was ‘the totality of facts [Tatsachen], not of things [Dinge]’.8 What motivated Wittgenstein’s claim here was the ‘context principle’ that he had taken from Frege: ‘[O]nly in the nexus of a proposition has a name meaning.’9 Sellars, who had a detailed knowledge of the history of philosophy, was well aware that this critique of ‘the given’ threatened to lead from a type of Kantianism that he endorsed, towards a type of Hegelianism that he did not.10 However, neither McDowell nor Brandom share this worry, intending, rather, to undermine the assumptions upon which Sellars’s worry had relied, assumptions they see Sellars himself as having implicitly dismantled. But from Sellars’s starting point, each, I suggest, have taken a somewhat different route in their journey from Pittsburgh to Berlin. While taking Sellars’s classic critique of myth of the given as his starting point in Mind and World, McDowell quickly signals the obvious danger awaiting the

McDowell, Brandom and Beyond

193

thinker rebounding from the lure of the given. Abandoning the myth of some passively received items capable of rationally constraining the active application of concepts in perceptual judgments can lead to the embrace of an equally implausible conception of concept application entirely freed from external constraint. Here ‘exercises of concepts threaten to degenerate into moves in a self-contained game.’11 Indeed, McDowell finds this danger threatening in the approach of another critic of the given, Donald Davidson, who, like Sellars, had been critical of the role of ‘intermediaries’ between mind and world, and had attempted to hold onto the idea of the world’s constraining ‘friction’ on thought, by stressing the causal constraints exercised by the world on judgment.12 Any such account, thinks McDowell, cannot capture the normative role that experience plays in providing thought with its objective purport: what is needed is a way of maintaining the idea of experience as exercising rational, and not simply causal, constraint on belief. Rather than show how experience can justify belief, Davidson’s causal account at best shows how it can offer ‘exculpations’.13 We must, thinks McDowell, retain the minimally empiricist idea that we are, in our perceptual beliefs, answerable to experienced reality.14 Of course, any such minimal empiricism must come unencumbered by the ‘mythical’ interpretation of the given as some non-conceptual ‘ultimate ground’ or ‘bare presence’ to which we can gesture in justifying our claims. Davidson had expressed his rejection of the myth of the given in terms of the idea that the only thing capable of justifying a belief is another belief,15 but McDowell thinks that this move effaces the obvious distinction between judgments inferentially arrived at and the non-inferential judgements of perception. Insisting on experiential justification, McDowell’s answer is to regard the ‘deliverances of experience’ or ‘bits of experiential intake’16 as necessarily already conceptual. If the contents of experience are conceptual, then we can see how they can maintain the appropriate justificational relation to beliefs – and it is this idea that points McDowell in the direction of Hegel. According to McDowell, Kant himself had been on the verge of a philosophy freed from the intolerable oscillation afflicting recent epistemology between the myth of the given and the idea of thought as a ‘frictionless spinning in a void’ that results when experience is deprived of the role of rationally constraining thought.17 But Kant was still ensnared in a version of the myth courtesy of his theory of sensory ‘intuitions’, regarded as non-conceptual mental representations. At one level his idea that thought is constrained by the representations of the receptive faculty – intuitions – seems innocent enough. ‘From the standpoint of experience’, Kant did not conceive of intuitions as making a separable contribution to the joint activity of the receptive and conceptual faculties, and so for him, ‘experience does not take in ultimate grounds that we could appeal to by pointing outside the sphere of thinkable content’18 But Kant joined to the account given from this standpoint one described from another, ‘transcendental’, standpoint and ‘in the transcendental perspective there does seem to be an isolable contribution from receptivity. In the transcendental perspective, receptivity

194

Postanalytic and Metacontinental

figures as a susceptibility to the impact of a supersensible reality, a reality that is supposed to be independent of our conceptual activity in a stronger sense than any that fits the ordinary empirical world.’19 However, Hegel had ‘urged that we must discard the supersensible in order to achieve a consistent idealism’, and this move, thinks McDowell, ‘frees Kant’s insight so that it can protect a common-sense respect for the independence of the ordinary world.’20 It is this idea of the thoroughgoing conceptual nature of experience that leads McDowell to embrace the ‘Hegelian’ thought of the world itself as ‘made up of the sort of thing that one can think’.21 Thus at the end of Lecture II, McDowell makes his approximation to Hegel explicit when he notes that ‘it is central to Absolute Idealism to reject the idea that the conceptual realm has an outer boundary, and we have arrived at a point from which we could start to domesticate the rhetoric of that philosophy.’22 In Making It Explicit, Robert Brandom invokes quite different aspects of Hegelianism to those invoked by McDowell. According to Brandom Hegel initiated a radical ‘inferentialist’ alternative to the dominant ‘representationalist’ paradigm of modern philosophical semantics, an alternative that reappeared in analytic guise in the work of Wittgenstein, Sellars and others in the midtwentieth century. Representationalists typically think of sentences or thoughts as composed of items that represent or stand for worldly things, properties, relations, etc. – they have a basically naming conception of the way language or thought and world come together. But Brandom develops an ‘inferentialist’ approach that starts from the ‘context principle’ of Frege and Wittgenstein. Rather than being the ultimate units of reference, a name has a meaning ‘only in the nexus of a proposition’.23 But in his later writings, Wittgenstein had widened the relevant context within which names find their meaning from single sentences or propositions to bigger stretches of discourse, and a similar direction had been taken by Sellars and Quine around the same time. Like McDowell, Brandom thinks that Kant was on the verge of this insight of contemporary philosophy. Kant recognized the ‘primacy of the propositional’ with his idea that the ‘fundamental unit of awareness or cognition, the minimum graspable, is the judgement’.24 Some readers of Kant have taken his employment of intuitions as a type of semantic device – singular representations which, by mediating between concepts and world, ultimately tie concepts to those worldly items that satisfy them.25 Seen in this way, Kantian intuitions resemble Russell’s ‘logical proper names’ – a resemblance which fits with Russell’s identification of Kant’s intuitions with his ‘sense-data’. But Brandom finds this aspect of Kant, along with conventional model-theoretic semantics that is used in conjunction with Frege’s own logic, to be ultimately at variance with the spirit of the context principle, especially in its expanded form. Rather, the progressive aspect of Kantian thought lies in the holistic idea that locates any judgment within the community of logically cohering judgements signalled by the notion of the ‘transcendental unity of apperception’,26 and Brandom finds the resources for a holistic critique of model-theoretical semantics in Frege himself.

McDowell, Brandom and Beyond

195

In the Begriffsschrift (Concept-script) of 1879 Frege had suggested that the only features relevant to a sentence’s conceptual content were those that affected the possible inferences that could be drawn from that sentence.27 While Frege himself had later retreated from this inferentialist view to a more representationalist one, his former student Rudolf Carnap was to maintain the inferentialist approach by defining the semantic content of a sentence in terms of the class of its consequences that held other than for reasons of logical form.28 Sellars had adopted this approach and it forms the basis of Brandom’s own inferentialist semantics in which the semantic content of a sentence is entirely derived from the totality of “material” inferences within which an asserted sentence stands.29 As, following the context principle, the semantics of all the sub-sentential parts of a sentence are derived from the semantic content of the sentence itself, the meaning of those parts will in turn be explained inferentially. That is, no atomistic theory of names is necessary and the representationalist outlook is vanquished. For Brandom, this was just Hegel’s way beyond the purported semantic contribution of intuitions in Kant. Thus, while the inferentialist stance was only implicit in Kant ‘it remained for Hegel . . . to complete the inversion of the traditional order of semantic explanation by beginning with a concept of experience as inferential activity and discussing the making of judgments and the development of concepts entirely in terms of the roles they play in that inferential activity.’30 In this way, the post-Fregean inferentialist movement towards a type of conceptual holism found in Wittgenstein, Sellars, Quine, Davidson, Rorty and others effectively reprised the move found within post-Kantian idealism away from Kant’s focus on judgements towards Hegel’s on inferences, a move summed up by Hegel’s dictum that the syllogism is the ‘truth’ of the judgment, rather than something that consists of judgments.31 In Mind and World and Making It Explicit, all references to Hegel are very general and might be read as merely adding a frisson of the exotic to these works. But is there really the proximity of these issues to Hegel’s philosophy as they suggest? A closer look at Hegel suggests that there is.

2. Hegel’s history of analytic philosophy from Russell to Quine and beyond Smitten fans of Hegel often attribute to him the most implausible of predictive powers, but it can be difficult not to read the first four chapters of his Phenomenology of Spirit as alluding to issues that were to become central to the history of analytic philosophy.32 In chapters 1 to 3 of this work, Hegel aims to demonstrate the inadequacy of the idea that knowledge can be founded on the pure givenness to consciousness of ‘objects’ of various kinds. In the first of these chapters, ‘sense certainty’, the particular object given to consciousness is meant to be a simple non-conceptualized singular item, something akin to Kant’s idea of an empirical intuition considered in isolation from any concept, and able to be

196

Postanalytic and Metacontinental

picked out by the use of a bare demonstrative, ‘this’! As Sellars had implicitly acknowledged, Hegel’s critique is close to his own critique of epistemological ‘givens’, for example, the ‘sense data’ that the early Russell had claimed could be known with certainty in ‘acquaintance’ and picked out by demonstratives considered as ‘logically proper names’.33 In Hegel’s account, the very idea of a bare singular presence knowable non-conceptually, is shown to be riven by contradiction. The epistemological outlook of sense-certainty had conceived of the bare ‘this’ as given in an immediate way without the participation of any general concept, but effectively Hegel suggests that such a ‘this’ in order to be cognitively relevant must, at the same time, be taken as an instance of a more general category – we might say, taken as an instance of ‘thisness’. In the object that comes to replace it in Chapter 2, the object of the shape of consciousness that Hegel calls ‘perception’ (Wahrnehmen), the fact that it instantiates some general kind is made explicit, and so this object is effectively conceived in the way that Aristotle conceived of a substance – that is, as an instance of a general kind, a ‘this such’ In Chapter 2, the concept of such a self-subsistent object of ‘perception’ with its particular categorical constitution undergoes a similar collapse and is replaced in Chapter 3 by a conception of something much more like a theoretically posited object found in modern scientific explanations of the world – the notion of a ‘force’ for example. This outlook Hegel calls, ‘the understanding’, and what ultimately exists for the understanding are no longer simply everyday things perceived as instances of kinds. The understanding’s ‘forces’ are not ‘perceived’ directly at all; they are posited in attempts to explain certain observable effects, and the contrast between ‘the understanding’ and ‘perception’ aligns with the type of difference that Sellars talked of in terms of different ‘scientific’ and everyday ‘manifest’ images of the world.34 I suggest that Hegel’s series of postulated ‘objects’ in these opening chapters might be aligned with the array of purported referring terms discussed in the early history of analytic philosophy. I have mentioned the parallel between ‘sense-certainty’ and Russellian sense-data qua ‘logically nameable’ givens of immediate acquaintance, but the ‘this such’ objects of Hegel’s ‘perception’ too seem to have an analogue, as objects picked out by definite descriptions,35 a purported species of referring term that Russell eliminated by analysis in his classic paper of 1905, ‘On Denoting’.36 Following Frege, Russell had been concerned with the possibility of empty terms such as the proper name ‘Pegasus’ or the definite description ‘the current King of France’. Earlier he had criticized what he took to be the misleading syntactic structure of universal affirmative sentences which seemed to model sentences like ‘All Greeks are mortal’ on singular sentences like ‘Socrates is mortal.’ Employing a form of paraphrase that utilized Frege’s terminology of quantifiers and variables, ‘All Greeks are mortal’ thus came to be replaced by a sentence with a conditional form something like ‘For all things, if that thing is Greek, then it is mortal.’ In ‘On Denoting’ he extended this technique to sentences with definite descriptions in subject place, such that

McDowell, Brandom and Beyond

197

‘The present King of France is bald’ became paraphrased in a way that eliminated the apparently referring noun phrase, ‘the present King of France’. In short, with the technique of logical paraphrase Russell eliminated subject terms from sentences that he thought liable to be misunderstood as types of names – the phrase ‘all unicorns’, for example, as liable to be conceived as a name of a collective subject, or the phrase ‘the present king of France’ as the name of a particular person. But he still held to the idea that naming was an essential dimension of the way language connected to the world – one simply had to find the right type of name such as ‘logically proper names’. Hence his recourse to the objects of ‘sense certainty’. However, by mid-century Quine challenged the model itself: for him, a singular term ‘need not name to be significant’,37 and this opened up a deep disagreement at the heart of the analytic project. For Quine, talk of singular reference was, as he put it, ‘only a picturesque way of alluding to the distinctive grammatical roles that singular and general terms play in sentences. It is by grammatical role that general and singular terms are properly to be distinguished’.38 Russell’s attack had been directed at understanding the sentence ‘All Greeks are mortal’ as structurally akin to ‘Socrates is mortal’, and so treating ‘all Greeks’ as a collective proper name. But Quine was to undercut the very contrast by treating ‘Socrates is mortal’ in just the same way! For Quine, ‘Socrates is mortal’ was to be effectively treated in terms of a bound quantifier that ‘ranged’ over a domain of discourse and as stating that if something is found that socratizes, then that thing is mortal. As Quine was well aware,39 treating proper names as predicates was just how medieval scholastic logicians had gotten around Aristotle’s prohibition on using singular judgements within syllogisms, as they had treated singular terms as ‘universals’. I have suggested that a definite description, in employing a subject term that included some general concept to designate an individual thing, is the sort of phrase that captures what Hegel thought of as a typically Aristotelian object of ‘perception’, the successor of the purported quasi-object of ‘sense certainty’.40 Russell’s technique in ‘On Denoting’, then, might be thought of as enacting the analysis of the ‘perceived’ object of the Phenomenology into the objects of ‘sense certainty’. But while Russell ‘regressed’ within the Hegelian series to a position more like that of sense certainty, Quine might be seen as having progressed within the series, to replace the objects of ‘perception’ by the posits of ‘the understanding’. But a consequence of this has been the problem diagnosed by McDowell as the picture of concept application free from empirical constraint, in which concept application threatens ‘to degenerate into moves in a self-contained game’.41 McDowell’s response to the Quinean ‘degeneration’ is effectively to dig in at the level of Hegel’s conception of the perceptual object and regard perceptual experience as thoroughly conceptual. Brandom’s response is different. Considered against the background of chapters 1 to 3 of Hegel’s Phenomenology of Spirit, Brandom’s attempt to go beyond the dilemma diagnosed by McDowell has him reformulating Hegel’s solution to the collapse of the given

198

Postanalytic and Metacontinental

objects of consciousness in chapter 4. Hegel’s solution to the problems that flow from the understanding is found in his radically anti-Cartesian account of the primacy of the relation of reciprocal recognition holding between cognitive subjects. Crudely, Hegel attempts to show that if all I can know are my own posits, then what I posit cannot simply be up to me. If we think of positing an object as something like issuing a decree or rule about what in experience is to count as an instance of that object, then for this notion to have any content, a subject would have to be able to hold herself to that decree or rule in her subsequent behaviour. That is, the relation a conscious subject bears to its posited object must be understood as holding internal to that subject’s normative relation to herself, a practical capacity to hold herself to her own rules – her ‘self-consciousness’. Next, such a holding oneself to a rule can only be understood within an overarching relation between subjects. I can only hold myself to a particular rule if I belong to a community of agents who can also recognize the rule and can hold me to it, thereby recognizing me as a rule-following subject.42 Thus, what separates out human beings from the rest of nature as the ‘spiritual’ (geistig) dimension of their existence is this capacity for ‘reciprocal recognition’ that Hegel famously introduces in his account of the simple model of social life with the master and slave.43 With this, Hegel’s phenomenology leaves the Cartesian domain of consciousness and self-consciousness and becomes, as the title conveys, a phenomenology of ‘spirit’ (Geist). That is, it becomes a phenomenology tracing the dynamics of concrete systems or networks of such recognitively mediated social relations that Hegel sees as operative in different societies in history. Brandom’s solution to the problems of reference within analytic philosophy is thus effectively Hegel’s solution to the problems of consciousness transposed into the key of a pragmatics of language use. While there is nothing simply ‘given’ from the world to constrain the semantic content of our claims, there are nevertheless social norms concerning the way we put words together in sentences and the way we put our sentences together in patterns of material inference. For Brandom to make an assertion is to make an undertaking to another: it is to commit oneself to giving reasons to one’s interlocutor for believing it, should she bring it into question. This brings out the primacy of the relations of material inference: to give a reason is to offer a further assertion from which the original can be materially inferred. Moreover, inferential relations figure further in that to make an assertion is not only to commit oneself to the truth of the sentence itself but implicitly to the further sentences that can be materially inferred from it. Conversations are exchanges in which we ‘keep score’ of each other’s changing inferential commitments and entitlements, and at the heart of this relation is the Hegelian stance of ‘reciprocal recognition’. In our asserting we are not answerable to the world as such, but to each other and so must recognize each other as beholden to the same rules that are immanent to our social practices and able to be made explicit, reflected upon and critically transformed within those social practices. In this way, Brandom has embedded central technical issues that have occupied many analytic philosophers for the last hundred years within a meta-conception of

McDowell, Brandom and Beyond

199

normative social practices mediated by institutionally encoded ‘normative statuses’ within which individuals stand – practices and institutions that he sees as at the heart of what Hegel thinks of as rationally self-transforming ‘spirit’ or ‘Geist’, and whose ‘shapes’ are charted in subsequent parts of the Phenomenology of Spirit. From such a perspective, philosophical practice cannot remain at the level of such technical problem-solving, say those of logical semantics or epistemology – a limitation which continental philosophers often see as a stifling characteristic of analytic philosophy. Philosophy must reflect upon the historically given normative relations within which modern discourses like that of philosophy itself function. Ethical and political considerations, thus, could no longer be confined to mere subject areas of philosophy to be approached with a conceptual armature constituted prior to and free from such considerations. With the work of McDowell and Brandom, then, we get a sense of how the traditional antagonism between Hegelian and analytic philosophy might be overcome.

3. But is it Hegel? Of course neither McDowell nor Brandom would want to be held to the plausibility of all of Hegel’s claims. As Hegel had put it, philosophy is ‘its time’ raised to the level of thought, and Hegel’s time was clearly not ours. So, for any reading of Hegel there will arise the question of what can, and what cannot, be separated from what we might take to be the contingencies of the time and culture to which he belonged, and we might think of our relation to Hegel here in ways broadly parallel to those that Hegel employed with respect to philosophers of his past. But beyond this generally historicist point, it is already possible to see a tension between what lessons Brandom and McDowell themselves wish to extract from Hegel, a tension between what we might call their respective ‘rationalist’ and ‘romantic’ interpretations. Here Brandom is helpfully explicit with respect to what he takes to be Hegel’s relation to the ‘Romantic expressivist tradition that he inherited’.44 Hegel’s version of this was, he points out, a rationalist approach which ‘gives pride of place to reasoning in understanding what it is to say or do something’.45 I suspect that from the side of the more conventional Hegelians, however, Brandom’s analytic translation might condemn Hegel to a framework of which he was consistently critical – the framework that typified Kant’s philosophy, ‘the understanding’, while Hegel’s were offered from the standpoint of ‘reason’.46 Elsewhere I have suggested that McDowell’s concern with the more perceptual aspects of experience link him to a less rationalistic form of thought, one more characteristic of those early romantic attempts to counter what they saw as rationalism’s threat to value.47 Does this mean, then, that an analytically reconstituted Hegel is in danger of being torn into irreconcilable opposed halves? This would be a fate akin to that of the historical Hegel, whose philosophy fell prey to warring ‘left’ and ‘right’ factions over the issue of religion. I suggest that Hegel himself may provide a way around this.

200

Postanalytic and Metacontinental

The part of Hegel’s Phenomenology of Spirit that coheres with McDowell’s employment of Sellars’s critique of the ‘myth of the given’ is, as I have suggested, the critique of ‘sense-certainty’ that leads Hegel to the consideration of ‘perception’. In contrast, Brandom is more concerned with the progression beyond ‘perception’ to ‘the understanding’ as well as the answer Hegel provides to the problems of it with his ‘recognitive’ account of the basis of self-consciousness. But in all such transitions, we are meant to see how each superseded “shape” of consciousness or self-consciousness is, while negated, still somehow incorporated into the superseding shape – the process that Hegel captures with the term ‘Aufhebung’. I suggest that Brandom’s social pragmatic interpretation may give us the means to make sense of this: might it not be that we employ different types or styles of reasoning in different situations or contexts? While reflecting on the nature of our reasoning in natural-scientific contexts, we may situate ourselves within the more rationalistic forms of thought like that of ‘the understanding’, but in other contexts, say reflecting on the nature of more value-laden forms of relation to the world such as literature or history, we might want to preserve forms of reasoning that stick closer to the shape of ‘perception’. Thus, just as in some areas of contemporary analytic philosophy, appealing to some form of ‘contextualism’ here – a ‘cognitive contextualism’ – may be what is needed.48 After a century of non-communication, the emerging dialogue between analytic philosophy and German idealism promises a rich future. Regardless of the directions that such an engagement may take, it will be difficult to ignore the path-breaking work of Brandom and McDowell.

Notes 1

2

3 4 5 6

7 8

9 10

11 12

Robert B. Brandom, Making It Explicit (Cambridge, MA: Harvard University Press, 1994). Richard Rorty, ‘Introduction’, to Wilfrid Sellars, Empiricism and the Philosophy of Mind (Cambridge, MA: Harvard University Press, 1997), 8–9. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). Charles Taylor, Hegel (Cambridge: Cambridge University Press, 1975). Robert B. Pippin, Hegel’s Idealism (Cambridge: Cambridge University Press, 1989). Terry Pinkard, Hegel’s Phenomenology: The Sociality of Reason (Cambridge: Cambridge University Press, 1994). Bertrand Russell, The Problems of Philosophy (Oxford: Oxford University Press, 1959). Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. C. K. Ogden (London, Routledge and Kegan Paul, 1922), §§ 1 & 1.1. Ibid., § 3.3. Thus Sellars referred to his lectures as his ‘incipient Meditations Hegeliènnes’ [sic]. Sellars, Empiricism and the Philosophy of Mind, § 20. Sellars actually advocated a hybrid of Kantian and scientific realist thought. McDowell, Mind and World, 5. Indeed, such a focus on causal constraint in the effort of find a way around the problems of the Russellian approach to sense-data has been a prominent feature

McDowell, Brandom and Beyond

13 14

15

16

17 18 19 20 21 22 23 24

25

26

27 28

29

30

31

32

33

34

201

of much work in analytic epistemology in the second half of the twentieth century. McDowell, Mind and World, 8. Thus McDowell talks of making room ‘for a different notion of givenness, one that is innocent of the confusion between justification and exculpation’ (as above, 10). Donald Davidson, ‘A Coherence Theory of Truth and Knowledge’, in Dieter Henrich (ed.), Kant oder Hegel? (Stuttgart: Klett–Cotta, 1983), 427–428. McDowell uses the phrases ‘experiential intake’ or ‘bits of experiential intake’ as a paraphrase of the Kantian term ‘intuition’ in a number of places in Mind and World (e.g. 4, 6, and 9). Ibid., 11. Ibid., 41. Ibid. Ibid., 44. Ibid., 27–28. Ibid., 44. Wittgenstein, Tractatus Logico-Philosophicus, § 3.3. Robert Brandom, Articulating Reasons (Cambridge, MA: Harvard University Press, 2000), 159–160. Such an approach was suggested in the early 1970s by Manley Thompson. See his ‘Singular Terms and Intuitions in Kant’s Epistemology’, Review of Metaphysics 26 (1972–3): 314–343. More recently, similar views have been advanced in Béatrice Longuenesse, Kant and The Capacity to Judge: Sensibility and Discursivity in the Transcendental Analytic of the Critique of Pure Reason, trans. Charles T. Wolfe (Princeton: Princeton University Press, 1998) and Robert Hanna, Kant and the Foundations of Analytic Philosophy (Oxford: Clarendon Press, 2001). ‘The subtlety and sophistication of Kant’s concept of representation is due in large part to the way in which it is integrated into his account of the inferential relations among judgments.’ Brandom, Making It Explicit, 92. Brandom, Making It Explicit, 94. Brandom, Making It Explicit, 96. That is, the class of its non-valid consequences. With this Carnap separated out conceptual content from logical form. A ‘material inference’ is simply an inference that does not hold because of logical form, for example, the inference from ‘Wollongong is to the south of Sydney’ to ‘Sydney is to the north of Wollongong.’ Brandom, Making It Explicit, 96. In particular, Brandom finds in Hegel’s methodological use of the combination of ‘mediation’ and ‘determinate negation’, ideas about the implicit structuring of the linguistic practice of the asking for and giving of reasons which are at the heart of his own rationalist pragmatism. G. W. F. Hegel, Science of Logic, trans. A. V. Miller (London: Allen and Unwin, 1969), 669. G. W. F. Hegel, Phenomenology of Spirit, trans. A. V. Miller (Oxford: Oxford University Press, 1977). See, for example, Russell, The Problems of Philosophy, 12. Russell had believed that Kant’s account of intuition was essentially in agreement with his position on sense-data – cf. 85. Wilfrid Sellars, ‘Philosophy and the Scientific Image of Man’, in Science, Perception and Reality (London: Routledge and Kegan Paul, 1963), 5 and passim.

202 35

36

37 38

39

40

41 42

43

44 45 46

47 48

Postanalytic and Metacontinental

A definite description, I suggest, in employing a subject term that included some general concept to designate an individual thing, is just the sort of phrase that captures what Hegel thinks of as a typical object of ‘perception’, an object characterized in Aristotelian way as an instance of a kind. Aristotle in fact discussed this type of judgment as a particular judgment, restricting syllogisms to universal and particular judgment types, and disallowing singular judgments, such as judgments with proper names in subject position. See my Analytic Philosophy and the Return of Hegelian Thought (Cambridge: Cambridge University Press, 2007), 89–91. Bertrand Russell, ‘On Denoting’, in Logic and Knowledge: Essays 1901–1950 (London: Allen and Unwin, 1956). W. V. O. Quine, From a Logical Point of View (New York: Harper and Row, 1961), 9. W. V. O. Quine, Word and Object (Cambridge, MA: MIT Press, 1960), 96, quoted in Brandom, Making It Explicit, 361. As Quine had earlier put it, a singular term ‘need not name to be significant’. From a Logical Point of View, 9. Thus ‘logicians in past centuries . . . commonly treated a name such as “Socrates” rather on a par logically with “mortal” and “man,” and as differing from these latter just in being true of fewer objects, viz. one.’ Quine, Word and Object, 181. Reference by definite description is what in Russell corresponds to the Aristotelian approach of picking out an object as instance of a kind in a particular judgment form. See above, footnote 35. McDowell, Mind and World, 5. Without this, the gap between the facts of my behaviour and the implicit norms being followed in the behaviour would collapse. Hegel, Phenomenology of Spirit, §§ 185–196. For an interpretation see my ‘The Independence and Dependence of Self-Consciousness: The Dialectic of Lord and Bondsman in Hegel’s Phenomenology of Spirit’, in F. Beiser ed., The New Cambridge Companion to Hegel and Nineteenth Century Philosophy (Cambridge: Cambridge University Press, 2008), 94–110. Brandom, Articulating Reasons, 34. Ibid. Here the thorny question of Hegel’s relation to the romantics is crucial. Rather than rejecting Enlightenment reason, the Jena romantics of the 1790s were rather trying to accommodate it within a broader conception of reasoning, one that could find a place for the truths of literature and religion. Redding, Analytic Philosophy and the Return of Hegelian Thought, chs. 5 and 6. Kant was an idealist, rather than a realist, about logical form, and it is just this idealism that will be useful for Hegel’s cognitive contextualism. Were one a realist about form, one could ask which of the cognitive forms, for example, that of perception or that of the understanding, gets the world right by capturing its basic form. But this question is meaningless for the idealist. On the relevance of Kant’s formal idealism for this tradition, see my Continental Idealism: Leibniz to Nietzsche (London: Routledge, 2009).

Chapter 13

On the Problem of Selection in Events for Deleuze and Davidson James Williams

1. A kind of event There are thin paint blobs down the back of your coat, so you set your internal mind-video to ‘play’. You rewind and search for an event. A detective working through close-circuit images, you search for clues, freezing frames, fast-forwarding, rejecting some scenarios and privileging others. The builders on that scaffolding were using the wrong colour. The school whose gates you pass was not on a break; no juvenile pranksters can be blamed. Then the critical moment comes into focus. As you passed under North Bridge, a painter was stencilling a slogan on grey beams. The event now has a frame. It can be bracketed in time by the commencement of an incautious flick and by the last instants of an irreversible spray. Spatially, it is restricted to a box, wide as your shoulders, with a top edge at the paint dispenser, a bottom one at the last fine speckle on cloth, a left one on the worker’s restraining hand as he leans from the cradle, and a right one at the tip of your nose.1 The splattering takes place in that enclosure, though it seems to articulate few of its many points. The metal bridge might be blameless, along with the seagull drawn from over-fished seas. Yet the wind shear round the bridge could bend the arc of paint. The painter might shift responsibility to the bird, to its unsettling shriek, and to the last time his scalp was skewered. As soon as the event is drawn into that causal network its well-defined boundaries begin to fade, diluted into long chains of cause and effect, or into a series of minute spatio-temporal points within the frame, a pointilliste representation but shorn of unity and meaningfulness.2 If the event is reduced to the points, it loses the essential movement and change captured in the gerund: a splattering, a dynamic and cohesive extension through time.3 If the event is hung on causes and effects, its frame is broken and any new edge must be a contingent limit on chains stretching into the distance. This is the fork any philosophy of events finds itself on; it is all the more damaging because the causal net and the space-time points are in alliance, the first explaining how the latter acquire

204

Postanalytic and Metacontinental

movement and change, the latter providing a real fundament for theoretical and hypothetical entities.4 Yet we describe events. We value them and ascribe roles to them in decisions and explanations. The extended event of the ruining of the coat articulates the snapshots of virgin state before the accident and dried paint afterwards. It focuses the lines of causes and effects as they draw away to the wasting of a sea bird’s habitat and to a painter’s nerves shredded by gull attacks. But is that focused event only a subjective entity? Remove thought, decisions, actions and sentience from the picture and the event, in its well-determined location and significance seems to fade away again. Why prioritize the splattering over one drop, or none, or over a seagull, changing course and spotting the botulism-soaked mouldy bread that will slowly kill it? This multiplication or disappearance of perspectives is damaging to more than events. There is a problem for causal and fact-based accounts too around which principles allow for determinacy and selection in them. As we move from comprehensive and abstract theory to local and involved practice, we fall back on events to set the scale for any investigation, for its priorities and its exclusions, for where a starting point is set, for goals and for directions of enquiry.5 When thrown into an open-ended map of all facts and corresponding scientific theories, a forward path is set by processes external to the enquiry: the motives of researchers, the priorities of societies, accidents and an intuitive feel for the merits of competing scientific theories. These processes can be described in terms of their characteristic events, for instance, happenstance, eureka moments, decisions, economic penury and opportunity, shifts in Zeitgeist, novel dispositions, fashion and changes in ideology. Whether scientific accounts have an internal selection principle allowing for practical distinctions of importance between alternative scales and locations, between the selection of which objects are important and which merely subservient, is not the main interest here. There are candidates for solutions in terms of contingency and necessity in a given causal process, taken from a practical angle or through scientific accounts of ideology, practical methods and social movements. However, even on this manifold and comprehensive scientific picture, and even if we accept that a set of statements of matters of fact are true and a group of theories valid, we still lack a principle for deciding which frames, for instance which deaths or endings in an overall picture, are to be selected as the significant practical ones. These principles could be specific to particular selections, that is, allowing frames to be set appropriate to a specific target in each instance, or they could be more general, providing meta-level guides suited to types of situation, or overall rules guiding different kinds of selection. It will be claimed later in this chapter that the latter two possibilities rely in some way on the first, on a singular fit of selection and event, rather than on overarching rules. This in turn raises the objection that non-specific rules are in fact adequate, to which the response will be that the selection of these rules itself requires a singular moment of selection determined by some kind of

Selection in Events for Deleuze and Davidson

205

interest thereby in effect making the claim that the rule-based, general, approach falls foul of some kind of regress. This response is of course only preliminary and would require an elaboration well-beyond the scope of the chapter. Setting aside this objection, the claim is that we require principles because without them the selection of entities according to a hierarchy of significance becomes arbitrary. Events remain useful because they can be described to allow for explicit accounts and assessments of frames and focus points. In this context, Donald Davidson claims a special relation of events to descriptions, acts and intentions.6 This allows for pragmatic accounts of determinacy, interest and focus absent from neutral or objective models, yet Davidson does not take account of a series of problems and characteristics of the selection of events as explained by Gilles Deleuze. To begin a reflection on the connections and contrasts between their two positions, this enquiry will attempt answers to the following question on the relation of language to events in terms of the pragmatic setting of focus points and priorities: ‘How are events selected and what are the consequences of this selection?’7 The point is not to give a full interpretation of the philosophy of events given by both philosophers. It is to give a Deleuzian account of the selection of events, inspired by his work but nonetheless distant in style, and to then consider it in relation to a position derived from Davidson.

2. Searching for something We can understand the selection of events generally and see it as a property of the philosophical definition of events. Thus, according to which definition is selected, different entities or processes qualify as events. Yet selection can be understood in a more bottom up manner where selection or principles of selection imply a set of events for later definition. In this case, how individual events are selected comes prior to their definition. It is this second approach and its many implied problems that are of interest here. It could be objected that this choice is mistaken, since philosophical analyses of events will have working models and intuitions of the definition of an event before selecting them. Even if we take our appeal to events in ordinary language or in different everyday practices as a test of the philosophical definition of events, the everyday still presupposes a general concept or idea of the event, or more minimally a word for the event, prior to looking at different forms of usage. Nonetheless, here, the issue is not how we select events as conceptualized in one way or another, but rather how certain selections draw out things from a bigger or wider picture that might later be described as events, but whose first determination does not assume a generic ‘event’. This explains why we began with the example of the splattering including the retrospective search for the spraying of paint on coat. The ongoing combination of search and splattering is not an event, but a complex and fluid relation of selections relating a ruined

206

Postanalytic and Metacontinental

garment to its causal web and to the emotions and judgements surrounding them. Even if the search arrives at the event in its well-defined box, that box is determined by a prior set of selections. We then have an event in the context of a selection, rather than as free-standing. The observation of the selection of individual events changes the problem of events from the technical difficulties of their definition to a tracking of the implications of different types and forms of selection. An assumption of this approach is that there is always some kind of pre-selection of individual events in any definition of the event, for instance, when a philosopher has a set of phenomena in mind as the general definition is being developed, or when a philosopher opts for a narrow set of examples of uses of the word ‘event’ in ordinary language, in everyday practices, or in specialized discourse, whether philosophical or scientific. The first claim to be made is that this pre-selection is not neutral with respect to the emerging theory of events. The second is that the functioning of this selection is not only of critical interest in terms of philosophies of the event, but is also a way of understanding how the selection of events conditions thought about them and reflection on their importance. The third claim is that it is not possible to escape the problems of selection, that is, there is no theory of events independent of prior conditions set by their selection and there are no events that can be determined independent of a prior selection. This leads to a further distinction between selection in terms of individuals, in terms of types and in terms of form. Each of these matters, but is important for different reasons and to different degrees. The selection of an individual, or individuation, is the picking out of an individual event.8 This does not necessarily mean the framing of an individual with well-defined boundaries; on the contrary, it could also imply the tracing of a set of infinite paths through a wider and more chaotic state – such as an endless branching of tumbling dominos. The selection of a type is a move to generality allowing an individuation to be classified among others as this or that category of event, such as a splattering, and up to distinctions such as political and social events, or normal and extraordinary ones. Form in selection is neither about a single selection, nor about collecting them into groups. Form is how selection sets out conditions for any event. These conditions are not causal, in the sense where a given selection could eliminate or make possible given events by taking away or adding in necessary causes. Rather, formal conditions have to be thought as revealing the necessary form all events must take. Form is about the ontological conditions for being an event, whereas type is epistemological, it sets the laws allowing for knowledge about the sets specific events belong to. Individuation is both ontological and epistemological. It is the condition for the being of an individual event and sets down a model for knowledge about it. We can see therefore that individuation, typology and formal conditions are not completely distinct; they depend upon

Selection in Events for Deleuze and Davidson

207

one another. The problem of the selection of events concerns each one in different yet connected ways. Form is transcendental but in a special meaning of this ambiguous term. The form implied by a selection does not set the general content of events, its parts or elements, or their relations to other types of content, but rather the formal relations that must obtain to give rise to an event: how an event comes about rather than what an event is. In this sense transcendental form is about the conditions for genesis.9 Though we could muster a good degree of acceptance for the importance of transcendental methods and the role of genesis in Deleuze’s philosophy around the late-60s and in Difference and Repetition and The Logic of Sense, there are deep problems of interpretation at the level of the complexity of his system, for instance in the triple form of asymmetric geneses in The Logic of Sense, and severe difficulties in describing his methodology due to its manifold nature, for instance in its claims to combine transcendental, empirical and pragmatic philosophy. Yet one of the claims of this chapter is that Deleuze’s insights have a wide import and relevance to contemporary philosophical debates, notably but not exclusively on the concept of the event. This raises the challenge characteristic of debates between traditions of how to translate between approaches that do not share a common philosophical language and method. It also raises the opposed commonplace objections against any such attempt that, either, the translation is a betrayal of the internal rigour and conceptual depth of one or other position, or, that one or other position has the last word on what such a correct translation should be, thereby inherently devaluing the other. My approach in this chapter is to translate and reinterpret Deleuze’s work on what I take to be its core directions in terms of the definition of the event. It is also to give a, no doubt simplified, example of Deleuze’s method in practice. This is to allow the originality and broad applicability of his philosophy to come through, while also providing a starting point that is consistent with more complex pictures of his philosophy of events, including the original texts that I refer to in footnotes. In what follows there is then what I take to be the spirit of Deleuze’s method and the bare bones of his conceptual frame. They are of course my versions of both of these and as such should be viewed as at best incomplete and, at worst, erroneous interpretations of his work. Even granted these objections, I nonetheless hope to have sketched an interesting position inspired by Deleuze’s philosophy. In that vein, we can think of Deleuze’s take on transcendental philosophy as a derivation of formal conditions from singular experiences with at least some speculative claim to universal applicability. This derivation relies heavily on ideas of genetic conditions rather than ones concerned with necessary content or essences. That a book should have a thousand pages, because every known book has, or because it is analytically necessary, are then not formal conditions. That it should have an author because, deriving from current experience, we cannot conceive of books not coming about without an author is a formal

208

Postanalytic and Metacontinental

condition. This deduction of formal genetic conditions from the observation of instances of events is determined empirically: the deduction begins with singular selections. It is therefore open to the accusation that what is deduced cannot reveal necessary conditions. This is correct and it would therefore be more accurate to call the necessity by the paradoxical name of speculative necessity, a claim to necessity open to revision through novel encounters. The argument about form therefore takes the following steps. Selection reveals something necessary about events, however this necessity is itself only arrived at through the observation of actual selections and events, so the necessity is speculative and open to change, but it warrants its name because the speculation is about formal conditions that are taken to hold for all selections and events. The strength of this approach is that it does not have to posit an abstract selection or event. Its weakness is that any claim to necessity is at best a wager and at worst illegitimate even at the time of the bet. The example given at the outset describes the selection of an individual splattering that fits under a number of possible types (splatterings, events involving intention and responsibility, actual rather than ideal events). As argued earlier, it is prone to difficulties with respect to its boundaries, in particular in relation to selection and this in turn makes the attribution to types vague and temporary. It is here that an appeal to form can help by drawing out three formal conditions. First, the selection is arbitrarily motivated; there is a minimal condition of some kind of interest, or movement, or inclination in the event. This does not have to be a conscious motivation. A dreamer’s hand dropping from the side of a bed into a glass of water can pick out many events; so can a change in current through a circuit board directing an automaton; and so can the eye of a bird as it responds to a glint on water. What there has to be, though, is that alteration: a change in intensity, here glossed as a motivation.10 Intensity explains a selection, a shift in a pattern against a homogeneous background, such as law-driven successions or stasis. As the description of the relation between two states, intensity is arbitrary. It distinguishes two sets: one (loosely called) prior and one subsequent to the appearance of intensity. Though it is easier to think of these relations in time, all that is required is the distinction between the states, rather than any temporal ordering or even spatial organization. There is arbitrary change in intensity as a condition for the selection of an event because a distinction is drawn free of any necessary relation between two sets. The arbitrariness and change are important for picking out an event – for instance, the event of a snake lashing out at an ankle due to an unpredictable shift in intensities governing coil and recoil. However, they are also points for raising objections against the notion of events, since, if we remove the arbitrariness and the motivations accompanying it and replace them by a necessarily determined unfolding, nothing seems to be lost other than the distinction and the mistaken ascription of arbitrariness where there is really cause and effect. The formal conditions for events neither justify practical reference to them, nor their

Selection in Events for Deleuze and Davidson

209

preponderance in wider epistemological issues; they only support their ontological legitimacy. Even this formal quality is fragile, since if a critical approach can question the presence of arbitrary motivation, it can eliminate the event as selected, for instance, where the snake strike is accounted for completely in causal terms. According to a second formal condition, we have a feedback process and emerging structure in the selection of the event: feelers are sent out and return in multiple and complex cycles constituting an ever-changing structure.11 Though intensity provides the initial explanation for selection, it requires and does not remain intact in the feedback process; like a bored CCTV operator, we need the apparatus to follow a hunch about a possible shoplifter but the intensity of the hunch might fade as we end up following an attractive stranger instead. There is no necessary limit in the number of loops, nor a necessary hierarchy in them, nor an intrinsic limit to their extent or content. All we can say is that expeditions and returns take place and these draw up manifold temporary structures. These cross and transform one another. They also carry, interfere with and transform intensities. These loops do not presuppose homogeneous space-time. Like film, they can jump, cut, split and assemble. Unlike film though, there is no intrinsic limit formed either by the medium, or by its history, or by its social context, or by its economic constraints.12 The feedback is not for anything other than the feedback itself and its connection to the intensity which becomes decentred and multiple within the structured loops. Intensity is then a contingent and mobile starting point destined to dispersal within feedback loops that double back to it. It is therefore a mistake to speak of a single intensity; there are multiple points of intensity travelling along multiple feedback loops. This is what an event is: intensities shooting through webs of feedback in relations of reciprocal determination. The structures and intensities determine one another, neither one taking the upper hand or grounding role. Long after the splattering, when you quiz the painter, fear in his eyes might be accompanied by unforeseen novel intensities. It becomes pointless to search for the original decision in all this, as if it had some kind of priority over others, since what matters is the ongoing review and transformation of intensities and feedback. The right approach is instead to describe and participate in the shifting intensities and the passages they make across loops folding back on each other. The Deleuzian moral question, on the horizon of his work on the event, is then ‘Are there principles to guide this description and novel participation or “counteractualisation”?’13 We can now begin to trace an answer to objections concerning the space-time location of the event. Space and time are constituted by the event. They do not contain it because the event is determined by the passage of intensities along paths and through structures that are not limited a priori by any given space within any given time-frame.14 Spaces are only drawn after the passage and provide only a partial representation of a place where the event is no longer

210

Postanalytic and Metacontinental

fully present, for instance, at the murder site where a search rested for a while but then passed on to look for a motive, and then on to a pardon or an execution. This is because intensity and feedback loops are by their nature mobile and dynamic; their essence is to pass or belong elsewhere, when they rest they fade, when they move they bifurcate and transform themselves. Similarly, time as a single well-ordered succession fails to take account of the multiplicity of loops and their different yet connected successions; it also fails in its measured ordering because the fading and raising of intensity, its durations, are prior to and cannot be accounted for according to the homogeneity and equivalences of measured divisions. Times and spaces appear with the event, which means that the point-like representations of events and causal chains always miss something essential about events.15 Events create complex spaces and times; they only correspond to ordered space as they leave it in their wake; they stretch, concentrate and distend times, refusing even a single time of succession through the way they draw different time-directions together. What ensures the connection between motivation and structure, between mobile intensities and the loops doubling back on them, while shifting and surpassing them? In answering this question, the human focus of my opening example is at its most misleading, in particular in terms of the discernment of different kinds of conditions and their relation to the (speculatively) necessary form of events. It could seem that some kind of consciousness is required to store, assess and act upon kinds of feedback gleaned from a mode of enquiry with a view to satisfying an initial motivation. This consciousness would use language to record and assess information returning along feedback loops. It would also need linguistic reference to retain the initial motivation on departure and along the way. Some kind of memory and linguistic marker holds an initial injury when we pursue the original accident that led to them. A similar marker carries evidence back from distant enquiries allowing the initial intensity – perhaps a desire for revenge – to be satiated, changed or increased. These markers and their relation to intensities and to motivations need to be related and computed, if not by human minds, at least by some kind of machine that can record and calculate. Formally, though, consciousness is supernumerary to structure, intensity and event. All that is required are coding, recording, locating and innovating processes. We need coding or denoting to identify parts or structures and particular intensities.16 These then need to be recorded in a separate structure that registers a given state and also registers its differences with others. There has to be some way of locating processes, or anchoring them. This self-identification for itself and for others is a form of manifestation or registering of a location. Finally, since coding, recording and manifesting are essentially forward looking, or themselves interested, that is, part of structure and intensity, there has to be a condition for this move forward or innovation, for the new intensities appearing in the process to be related to those already present as something new; this is significance or sense. Deleuze calls this set of processes ‘language’.

Selection in Events for Deleuze and Davidson

211

It adds denoting, signification, manifestation and sense to intensity and structure. Thanks to language, the third and final formal condition for events, there can be an identification of parts, a differentiation of structures, a settling of their location and, in sense, a condition for innovation (where sense17 is not meaning but rather an open potential for novel structures and novel significance, alongside novel intensities).

3. Davidson’s restriction on Deleuze’s conditions If we follow Deleuze’s lead, the necessary and sufficient formal conditions for the selection of events are intensity (a motivation), structure (a frame for feedback) and language (a recording, differentiating and innovating process). Any restriction on these conditions for the genesis of an event is an unnecessary restriction on events; that is, an illegitimate narrowing down of events. It is possible to see Davidson’s work on events as in tune with these conditions. Like Deleuze, he associates events with language – with descriptions, parts of sentences and logical formalization. More significantly this association of event with language sets up a series of strong relations between event and language such that, for instance, events have to be defined in a way consistent with the linguistic fact that events can be redescribed in open and multiple ways. Like Deleuze, Davidson also associates events with kinds of motivations, with actions and intentions, but not always and not necessarily. Again, this association strongly conditions the definition of events, though, unlike Deleuze, not all and not in the same way. Nonetheless, there are categories of events important for Davidson (acts and intentions) that replicate the role played by intensity and motivation in Deleuze’s model. Finally, something like structural conditions is given by Davidson in his work on the individuation of events and their association with a novel semantics. As we shall see, these structural conditions are more varied than Deleuze’s and are much more strongly conditioned by an engagement with scientific or fact-based critiques of the concept of an event, but they are structural conditions connected to the limits to be ascribed to events (for instance in terms of different types of explanation) or to the relation of events to space-time (the location and divisibility of events). Yet each of these connections leads to wide divergences and these can be analysed in terms of the selection of events: formally and as individual events. In turn, many of these distinctions have repercussions on what can constitute an event, but this is less significant than the source of the divergence in the different problems treated by each philosopher. For Deleuze, the problem of how to select events, formally and individually, is the deepest driver behind his work on events. He is concerned with the presuppositions underlying any selection and with the question of how to construct the most open form for events, one that imposes the least limitations, yet one that provides the most precise guide as to how to do so.

212

Postanalytic and Metacontinental

Davidson’s motivations are outside the event proper, they can be found instead by looking at the titles of his influential essays on events: ‘agency’, ‘logical form of action sentences’, ‘individuation’ (in a sense quite different from Deleuze’s), ‘events as particulars’, ‘eternal’, ‘ephemeral’ and ‘mental’ events. Two dominant traits emerge from the concerns captured by these titles. The first is that events, to be defined in a certain way, are necessary for a translation from natural language into a logical form: ‘There remains, however, a more direct consideration [. . .] in favour of an ontology of events, which is that without events it does not seem possible to give a natural and acceptable account of the logical form of certain sentences of the most common sorts [. . .].’18 The second is that events have to be consistent with some other philosophical positions, for instance regarding actions, intentions and mental events. In addition to this consistency, events should be defined in such a way as to help with the resolution of difficulties in relation to these wider philosophical positions. This latter requirement overlaps with the first trait, since one such difficulty, involves the semantics of natural language: ‘[. . .] there is a relation between a person and an event, when it is his action, that is independent of how the terms of the relation are described [. . .] we have discovered no analysis of this relation that does not appeal to the concept of intention.’19 Another difficulty, perhaps in tension with the first, is that events should cohere with our scientific knowledge of the world, yet also mark the boundaries of such knowledge: ‘Mental events as a class cannot be explained by physical science; particular mental events can when we know particular identities.’20 The following general points can therefore be made with respect to the different approaches to the selection of events in Deleuze and Davidson. For Deleuze, language and event determine one another because the selection of events depends on selections in each of the main components of language as process. Changes in language – in sense, denotation, signification and manifestation – are a condition for the genesis of events. Yet these changes must accompany events: the emergence of intensity in structures determines language as process. For Davidson, reference to events in natural language and translation into a logical form guide the definition of events.21 Language and logic are first and pre-select what events can be. For Deleuze, there will be many references to events in natural language that should be viewed as incomplete when viewed in terms of the form events must take. Logical formalization will be judged according to how well it can cohere with this speculative determination of events. For Davidson, the demands of natural language and logical form take precedence but at the cost of drawing in a large array of preconceptions about the event in terms of usage in natural language and with respect to how events can and cannot be formalized logically. In terms of the relation to science, Davidson’s position will be constrained by general theoretical tenets, for instance in terms of causality, and more local restrictions, for instance regarding measurement and measurability in terms

Selection in Events for Deleuze and Davidson

213

of mental events. Yet there is a connection to Deleuze’s position since for Davidson mental events as a class are beyond the explanatory power of science and this is consistent with some aspects of Deleuze’s approach, where intensity and sense are beyond the explanatory scope of science. The difference between the two lies, however, in the universal scope of Deleuze’s position, taking in science, when opposed to Davidson’s setting of proper boundaries for different modes of theoretical approaches and kinds of object. This does not mean that neither philosopher is informed by science, they both are. It is rather that the independence claimed for their definition of events from particular scientific theories and observations is different. Davidson’s strategy draws limits for science in relation to events, whereas Deleuze is designing a metaphysical position that precedes any such delimitation and denies it for certain modes of thought in relation to events, which remain universally applicable. This difference in the relation of theories of the event to science can be seen at work in the contrast between the two in terms of intensity or motivation and in terms of the individuation of events. For Deleuze, the spark of intensity can be expressed in many different actual occurrences. All that is required is the appearance of a variation in intensity, a shift or turning point, that we can associate with an observable change in a particular system or type of system (biological or mental, for instance) but only incompletely and indirectly. The intensity is deduced from the observation, rather than known in it. Intensity takes the same form for all events and this form is extra-scientific. For Davidson, the individuation of events is explicitly scientific and constrained by a causal frame, implying that causality is a prior condition for events: ‘[. . .] events are identical if and only if they have exactly the same causes and effects. Events have a unique position in the framework of causal relations [. . .].’22 This is in direct contradiction with Deleuze’s setting of events as prior to a position in a causal frame; the frame always comes after and is insufficient for determining the event.23 Moreover, causality can only ever be a partial aspect of the structure of the event. It is insufficient for individuation without the three other necessary conditions of intensity, structure and language, none of which are causal. Once again, the imposition of a causal frame of events is a restriction on a wider possible selection of events when viewed according to Deleuze’s model. There is a general pattern of such restrictions in Davidson’s work when compared with Deleuze’s. However, if we then turn to a critique of Deleuze based on Davidson’s work on the event, we see that a high price is paid for keeping events as free of restrictions as possible, except for those conditions necessary for their selection. The relation of event to natural language is then broken. Events become hard to formalize logically. Perhaps most seriously, the relation of philosophy to science is even more problematic than in Davidson’s position, since instead of limiting the claims of science to certain (large) areas, Deleuze’s view of events takes certain things central to the selection of all events (intensity and sense) as explicitly outside the realm of science.

214

Postanalytic and Metacontinental

Notes 1

2

3

4

5

6

7

8

9

10

11

12

For a critique for this spatial account of events see P. M. S. Hacker, ‘Events and Objects in Space and Time’, in Events, ed. Achille Varzi and Roberto Casati (Aldershot: Dartmouth, 1996), 434: ‘Events do not consist of matter [. . .] they do not have any spatial dimensions.’ For an exemplary account of this kind of role given to space-time for objects as opposed to events, see W. V. Quine, ‘Events and Reification,’ in Events, ed. Achille Varzi and Roberto Casati, 107–116. Aldershot: Dartmouth, 1996. 113: ‘Space-time is the matrix on which we can draw for all our reifications of concrete objects . . .’ For analyses of the importance of the gerund in expressing change in events as opposed to imperfect nominals see Jonathan Bennett, Events and Their Names, (Oxford: Clarendon, 1988), 4–6. For a broad statement combining the space-time location of events with causal relations, see Kent Bach ‘Actions are Not Events’, in Events, 344: ‘I assume that events exist in space and time and that they enter into causal relations as causes and effects.’ Davidson speaks of the ‘nomological slack’ between the mental and physical to explain the heterogeneity of the two realms for instance with regard to intentions. It is this slack and shift that allows for perspective and focus in the latter realm and not in the former. See Donald Davidson, ‘Mental Events’, Essays on Actions and Events (Oxford: Clarendon, 2001), 213. For an account of the relation between descriptions, events and agency, see Donald Davidson, ‘Agency’, in Davidson, Essays on Actions and Events, 46–51. One merit of drawing this comparison lies in the lack of work done on contrasts between philosophies of the event in the continental and analytic traditions. For instance, in their collection on the event, Varzi and Casati cite no contemporary continental philosophers in what they claim to be a comprehensive list, despite work published at the time by Deleuze, Derrida, Lyotard and Badiou, to name but some (see Achille Varzi and Roberto Casati, 50 Years of Events: An Annotated Bibliography 1947–1997 (Bowling Green, OH: Bowling Green State University, 1997), 2–4). Deleuze’s account of individuation, in its more strongly biological version, can be found in Gilles Deleuze, Difference and Repetition, trans. P. Patton (New York: Columbia University Press, 1994), 251. For an excellent account of the role of genesis in Deleuze’s transcendental thought, see Miguel de Beistegui, Truth and Genesis (Bloomington: Indiana University Press, 2004), 248–250. See also my Gilles Deleuze’s Logic of Sense: A Critical Introduction and Guide (Edinburgh University Press, 2008), 194–201. For a full account of intensity see Difference and Repetition, 232–233, and my Gilles Deleuze’s Difference and Repetition: A Critical Introduction and Guide (Edinburgh University Press, 2003), 178–184. Gilles Deleuze, The Logic of Sense, trans. M. Lester and C. Stivale (New York: Columbia University Press, 1990), 63–67 has the most full account of the role of structure in relation to events. This reference to film in relation to space and time is very important for understanding Deleuze’s full conception of the event as resistant to a situation in

Selection in Events for Deleuze and Davidson

13

14

15

16

17

18 19 20 21

22 23

215

homogeneous space-time. See his two books on film, Cinéma 1 (Paris: Minuit, 1983) and Cinéma 2 (Paris: Minuit, 1985). See Deleuze, Logic of Sense, 143–151 for an extended account of the event in relation to morals. The idea of spaces and distributions free of prior grids is important to Deleuze, for instance in his concept of smooth space from Gilles Deleuze and Félix Guattari, A Thousand Plateaus: Capitalism and Schizophrenia, trans. B. Massumi, (Minneapolis: University of Minnesota Press, 1987), 474–500. Note how this view runs counter to Hacker’s account: ‘Events, unlike objects, are directly related to time. They occur before, after, or simultaneously with other events’ (P. M. S. Hacker, ‘Events and Objects in Space and Time’, 445). See also Hacker’s claims in ‘Events, Ontology and Grammar’: ‘Objects are immediately related to space, events are immediately related to time.’ (‘Events, Ontology and Grammar’, in Events, 82). For an account of the importance of coding and a view of structure different from Logic of Sense where desire takes the place of intensity, see Gilles Deleuze and Félix Guattari, Anti-Oedipus: Capitalism and Schizophrenia, trans. R. Hurley, M. Seem, and H. Lane (New York: Viking Press, 1977), 36–42. Sense must not be confused with meaning here, it is rather a condition for the production of novel forms of significance, see The Logic of Sense, 22. Donald Davidson, Essays on Actions and Events, 166. Ibid., 61. Ibid., 225. For a position that draws distinctions between different linguistic registers, ordinary language and the language of science, see F. I. Dretske, ‘Can Events Move?’, in Events, 427. For a rejection of Davidson’s enterprise on the logical transcription of events, see P. M. S. Hacker, ‘Events, Ontology and Grammar’, 88. For a different approach to that transcription, based around concerns with events as repeatable – something more distant from a Deleuzian approach than Davidson – see Roderick Chisholm, ‘Events and Propositions’, in Events, 91. See also Bennett’s doubts regarding the usefulness of events that constitutes a critical position against Davidson and Deleuze on the grounds that events are ‘essentially imprecise and uninformative’ (Events and Their Names). As a counter to this, see Thalberg’s arguments for pro-event metaphysics based on the detection of eventlike categories in no-event-metaphysics: Irving Thalberg, ‘A World Without Events’ in Essays on Davidson Actions and Events, ed. Bruce Vermazen and Merrill B. Hintikka (Oxford: Clarendon, 1985), 150. Thalberg’s approach offers a promising strategy for a similar counter based on Deleuze’s work. Davidson, Essays on Actions and Events, 179. This means that Deleuze’s position sets him not only against Davidson’s position on individuation, but others based around different yet still prior identity conditions, such as Kim’s, based on properties (Jaegwon Kim, ‘Events as Property Exemplifications’, in Events, 119). However, both Deleuze and Davidson fall foul of Anscombe’s scepticism of any principle of individuation in ‘Under a Description’ because it must lack ‘criteria for what constitutes a single action or event’ since these are uncountable (Elizabeth Anscombe, ‘Under a Description’ in Events, 313).

Chapter 14

Virtual Platonisms: Lautman and Gödel Pierre Cassou-Noguès

The aim of this chapter is to compare Lautman’s and Gödel’s Platonism. Alongside Cavaillès and Bachelard, Lautman is part of an epistemological movement initiated by Brunschvicg that looks for the objectivity of mathematics in its history. For Gödel, as for other logicians of the same period, such as Bernays, the problem of Platonism is first related to the foundations of mathematics. Nevertheless, Lautman and Gödel eventually reach similar positions, which can be described, in Deleuze’s terminology, as a virtual Platonism. I will contrast two kinds of Platonism: naive and virtual Platonism. This terminology, naive versus virtual, is mine. One could find in the literature examples of both naive and virtual Platonism simply considered as ‘Platonism’. Nevertheless, the term ‘Platonism’, in contemporary philosophy, usually refers to what I would call a naive Platonism. In a naive Platonism, mathematical theories are taken to describe an ideal reality. These ‘ideas’ are the objects of the theory and their domain would constitute a model for the theory, were it perfectly worked out. Examples can be found in some of Gödel’s texts: (1) ‘Mathematics describes a non-sensual reality, which exists independently both of the acts and [of] the dispositions of the human mind and is only perceived, and probably perceived very incompletely, by the human mind.’1 This ideal reality is a well-determined domain that our mathematical theories mirror more or less adequately. In contrast, a virtual Platonism situates ideal reality not in the objects of the theory, in the sense of a model the theory describes, but rather at the root of mathematical activity. Thus, the virtual Platonist holds that there is an ideal reality at the basis of arithmetic, or set theory, and one cannot therefore account for mathematics from a purely formalist perspective. But this ideal reality need not consist in integers, or sets, as these are described by the theory. Ideas may not be referred to in the formal theory. This distinction between naive and virtual Platonism will be more amply worked out in the course of this chapter. Several examples will be given which will hopefully make it clearer. I must also add that this distinction mirrors Deleuze’s distinction between possible and virtual. However, I do not claim that my sense of virtual is strictly Deleuzian. My aim is to show how Lautman and Gödel, though coming from

Virtual Platonisms: Lautman and Gödel

217

different philosophical traditions, reach analogue positions. The interest of this virtual Platonism is that it lends objectivity to mathematics. In the first section, I present Lautman’s Platonism in the context of historical epistemology. I show that Lautman shares a common perspective with Brunschvicg, Cavaillès and Bachelard. Lautman’s singularity, in this philosophical tradition, is to relate objectivity in history to an ideal reality, thus reconciling two seemingly antagonist positions, historicism and a kind of Platonism. In the second section, I discuss Gödel’s Platonism. As a logician, Gödel produced his best-known results, the theorem of incompleteness and the proof for the relative consistency of the continuum hypothesis and the axiom of choice, in the 30s (1931 for the first result, 1938 for the second). After his emigration to the USA in 1940, Gödel turns more and more to philosophy. He is usually considered as holding a strong and unqualified Platonism. In previous papers, I have tried to show that Gödel pursued a long reflection on the problem of Platonism and that his position changed several times; here, I make free use of Gödel’s texts in order to discuss what I take to be Gödel’s main argument for his Platonism.2 The third section brings together results from the first two, thereby showing how Gödel rejoins Lautman. I then return to the Brunschvicgian tradition and offer another reading of it from Gödel’s point of view.

1. An historical epistemology Starting with Brunschvicg, there is a French epistemological tradition relying on a consideration of the history of sciences – where ‘science’ refers both to mathematics and to physics. Brunschvicg is not the first to use history of science; rather, he seems to crystallize a methodology. All the authors of the following generation, Cavaillès, Lautman and Bachelard, refer to Brunschvicg, while in turn they are reference points for Althusser, Foucault and Deleuze. Brunschvicg is at the root of a tree of explicit references right up to contemporary thinkers. The tradition Brunschvicg initiates is centred on a methodological postulate that Bachelard, Cavaillès and Lautman all take up.

1.1 Brunschvicg (2) ‘The philosopher works on history as the scientist operates in his laboratory, because they cannot do otherwise: humanity is only given to reflection through history.’3 (3) ‘Science considered independently of its becoming’ is ‘an abstraction’.4 (4) ‘To relate the conquests of the present to the difficulties of the past, only this operation assures us to have caught the true meaning of both, only this operation confers an objective meaning to the course of history.’5

218

Postanalytic and Metacontinental

Brunschvicg remarks that history is material and means for philosophical thought.6 His historical method depends on a postulate not justified by an argument other than that of quotation (2): the philosopher cannot but turn his attention to history. Contemporary theories only make sense when related to their past. It is necessary to know the history of science, in order to understand the present state of science. For example, when learning mathematics we repeat, however vaguely, the history of mathematics. Therefore, in considering contemporary theories independently of their history we are only dealing with an ‘abstraction’ without support. Yet, it is only on the ground of a certain conception of science – referring to its history – that Brunschvicg argues for the use of history of science. Thus, if one interprets a statement like (4) as an argument supporting the historical method, the argument seems to be circular. Quotation (4) introduces the idea of ‘objectivity’ which, with certain variants such as ‘objective meaning’, is frequent in the author’s under consideration. Its meaning is rather vague and the word seems to denote a common notion which does not need explanation. The objectivity of science does not mean reference to independent objects, to the empirical world of physics or to an ideal reality in mathematics. Neither does it mean a foundation for science, in the sense in which one thinks of ‘foundation’ in relation to Hilbert’s program. The problem is not to establish the consistency of the theories in question nor even their acceptability. Inasmuch as naive Cantorian set theory, with its paradoxes, is a step in the formation of contemporary axiomatized set theory, Cavaillès, who considers that the history of mathematics has a necessity and that none of its former steps can be bypassed, would have to accept that naive inconsistent theory has the same objectivity as axiomatized theory. An inconsistent theory may have objectivity in as much as it is a step in the formation of a consistent theory. However, for obvious reasons, it is not thus rendered satisfactory, nor can one speak of it as giving a ‘foundation’ to an inconsistent theory. The objectivity Cavaillès and Lautman are looking for does not consist in a foundation. It refers instead to the ‘solidity’ of mathematics, to the way an unproved conjecture resists the mathematician, and to the absence of arbitrariness. Part of the ambiguity of the idea of objectivity also comes from the fact that Brunschvicg associates two different factors in the objectivity of science. Indeed, his aim is to (5) ‘ground the reality of knowledge on the reciprocal adaptation of experience and reason, which turns experience into an intelligence in acta and which gives reason control over the things themselves.’7 History of science is a dialogue or a struggle between experience and reason, with experience encompassing the whole field of phenomena confronting reason. Each step in the history of science is the result of a confrontation that always has the same structure. Reason first faces an experience which appears problematical: phenomena lack unity and reason cannot embrace them in their diversity,

Virtual Platonisms: Lautman and Gödel

219

or they have no meaning in the current categories of reason. Human reason then invents a new concept uniting the phenomena. However, the concept still needs to be applied in experience through gestures Brunschvicg calls ‘practice’. These enable one to verify in concreto that the different phenomena indeed fall under the unity of the concept. However, in this practice, the concept is embodied in experience and the experience is transformed. Phenomena are no longer considered for themselves but only as representations of the concept. In fact, a new plane of experience has been created, whose terms are not the initial phenomena but the concepts which the phenomena now represent. In this new plane of experience, new difficulties appear, requiring new concepts, and so on. The initial experience an infant faces is a multiplicity of sensations. The first concept produced by human reason is that of an object as the unity of various sensations. Handling the things given to the child enable it to verify that its sensations may be organized around such unities. The child only then considers experience as made of things: no longer simple sensations but sensations organized in the unity of an object. This is a second plane of experience which will be the ground for arithmetic and geometry. The child realizes different collections of things are equivalent from the point of view of a new concept: number, put in place through the practice of one-to-one exchange. The child buys an apple for a coin. The collection of apples and the collection of coins are now related and may be considered as representing the same number. This constitutes a new plane of experience whose elements are no longer things but numbers, classes of collections of things which may be exchanged one-to-one. In the same way, in geometry, drawing enables the child to grasp the concept of exact form as a permanent feature belonging to the thing itself and underlying different appearances. For the child who draws, there are no longer things but exact forms, the space of geometry. From Brunschvicg’s perspective, the child’s learning process reproduces the history of humanity. These first steps initiate the learning of mathematics as well as the constitution of science. Each step follows the same scheme: a problematic experience, the invention of a concept, the application of the concept in experience through practice, and the constitution of a new plane of experience. In this dialogue shaping the history of science, experience and reason induce a twofold factor of objectivity. The objectivity of science can be referred to reason, in as much as it invents new concepts which make experience intelligible, as well as to experience, in as much as it embodies the concepts of reason and thus distinguishes them from pure chimera. Thus: (6) ‘Mathematics unites rationality and objectivity as reciprocal and interdependent functions which cannot be separated from one another, because, contrary to the dreams of dogmatic Realism and dogmatic Idealism, rationality cannot transcend itself in the absolute of a pure reason, in pure evidence, and objectivity cannot transcend itself in the absolute of a pure object which could be immediately apprehended.’8

220

Postanalytic and Metacontinental

Here, experience produces the objectivity of science. However, in Lautman’s reading: (7) ‘There is no philosopher who nowadays has better than Brunschvicg developed the idea that the objectivity of mathematics is the result of intelligence in its effort to triumph over the resistances in the material on which it works.’9 Lautman relates the objectivity of mathematics to reason rather than to experience. I do not think that this shift is the sign of a disagreement between Lautman and Brunschvicg. It is the result of an ambiguity in Brunschvicg’s system. The objectivity of science can be attributed both to experience and to reason. Thus Brunschvicg’s inheritance is, first, in a methodological postulate placing the objectivity of science in the history of science and, second, in the distinction between two factors of objectivity, intellectual and empirical. None of Brunschvicg’s successors, Bachelard, Cavaillès and Lautman, will take up Brunschvicg’s conception exactly, but they will all accept his methodological postulate and face up to the problem of this twofold objectivity of science.

1.2 Bachelard and Cavaillès Bachelard is close to Brunschvicg on the structure of the steps of the history of science. He takes up the idea of dialogue between experience and reason, but he adds a third term, imagination. The experience reason faces is never pure but always already informed by imagination: a world view that is implicit yet structured, which impends the development of science. Imagination, which is innate, is organized around the four elements, earth, water, air and fire. Each has a constellation of vague meanings and leads us to conceptions from which science must liberate us. The history of science, in Bachelard’s view, is both liberation from imagination and rationalization of experience. Bachelard, who opposes science and imagination, a world view whose constitution cannot be traced introspectively, opens the way for Althusser’s notion of ideology or, to some extent, Foucault’s notion of episteme. Nevertheless, concerning the mechanism of the history of science and the problem of the objectivity of science, Bachelard remains in the wake of Brunschvicg. Cavaillès diverges from Brunschvicg in two ways. First, Cavaillès attempts to circumvent Brunschvicg’s reference to reason, to consciousness. As quotation (4) shows, Brunschvicg relates the present of science to its past. Cavaillès starts by giving more weight to this relation. He sees in the emergence of new notions in mathematics, a necessity emanating from open problems. These problems require notions that will give them a solution, they explain the emergence of the notions or, to use Cavaillès’ own term, they seem to ‘engender’ or to ‘give birth’ to new notions and operations. Thus, concerning set theory, (8) ‘Certain problems of analysis have given birth to the essential notions and have engendered certain operations [. . .] defined by Cantor.’10

Virtual Platonisms: Lautman and Gödel

221

Whereas, for Brunschvicg, a new concept is the result of an invention of human reason, a reason that faces an experience but which remains free, for Cavaillès, it is entirely explained by the open problems of the mathematical field. Open problems call for new notions which in turn engender new problems, so that the philosopher need no longer relate the mathematical development to a mathematical mind, reason or consciousness. As in Spinoza, the ‘ideas’ that participate in the mathematical becoming call for one another, so that the mathematical mind can be considered an element in this development and no longer its source. It is this autonomous development that Cavaillès calls in ‘Transfini et continu’ and in his philosophical testament, Sur la logique et la theorie de la science, the ‘dialectic of concepts’. The intellectual factor of objectivity, corresponding to reason in Brunschvicg’s conception, disappears in Cavaillès’ epistemology. Or rather, there is something like an intellectual objectivity of mathematical theories, which will be opposed to an experimental objectivity. But this intellectual objectivity only comes from the integration of contemporary theories in a necessary becoming. In the introduction to his first thesis, in 1938, Cavaillès writes: (9) ‘There is an objectivity, founded mathematically, of mathematical becoming.’11 The history of a theory, like set theory, is not accidental but is related to and grounded in the mathematical reality of the theory. However, in his latter writings, it is the mathematical reality of the theory which should be grounded in and related to its history. The formula is reversed: there is an objectivity, founded historically, of mathematical theories. Parallel to his analyses on the history of mathematics, Cavaillès takes up the Brunschvicgian idea according to which mathematical development relies on an experience, a practice or what Cavaillès calls a ‘gesture’. However – and this is the second point of opposition with Brunschvicg – instead of referring to planes of phenomena, such as the multiplicity of the child’s sensations, Cavaillès refers to experiments on signs in what he calls a ‘combinatory space’. Mathematical development depends on a plane of experience but this plane does not consist of sensations, or phenomena informed by reason, it is made of the signs, the arithmetical formulas or the geometrical figures, which the mathematician uses. These ‘combinatory spaces’ represent a second factor of objectivity. In fact, when delineating his ‘modified formalism’ in his second thesis of 1938, Cavaillès acknowledges that: (10)

‘[the theory of signs] justifies the objectivity of the system’12

Cavaillès inherits Brunschvicg’s twofold notion of objectivity. Clearly, quotations (9) and (10) involve two different kinds of objectivity. There is in (9) an intellectual objectivity: mathematical theories have objectivity in as much they are part of a necessary development of problems and notions. In (10) there is

222

Postanalytic and Metacontinental

an empirical objectivity: mathematical theories have objectivity in as much they are supported by experiments on signs. Cavaillès will never solve the problem of the relationship between these two planes, the intellectual plane of the ‘dialectic of concepts’ and the empirical plane of the gestures on signs or what Cavaillès calls the ‘intuitive superposition’. In fact, in one of his last texts, Cavaillès considers this relationship as ‘the fundamental problem of mathematical philosophy’.13

1.3 Lautman Like Bachelard and Cavaillès, Lautman takes up Brunschvicg’s methodological postulate, or Brunschvicg’s historical turn, and inherits the problem of the twofold objectivity of science. In this context Lautman’s Platonism appears in its singularity. Lautman’s Platonism is a Platonism of mathematical history, of the mathematical movement. Compared with Brunschvicg, Lautman interprets the intellectual factor of objectivity, related to reason for Brunschvicg and Bachelard and to the necessity of history for Cavaillès, as the participation of mathematical theories in an ideal reality. Lautman’s ‘Ideas’ express themselves in exactly the plane that Brunschvicg, Bachelard and Cavaillès turn their attention to, that is the history, the movement, the activity of mathematics, rather than its axioms. This distinguishes Lautman’s Platonism from any naive Platonism. However, in this context, Lautman has to face another problem which is to interpret the second factor of objectivity, empirical objectivity. This problem, to elucidate the true ‘matter’ in which the ‘Ideas’ incarnate themselves so as to produce mathematics properly speaking, returns throughout Lautman’s career. I only touch on this aspect. I first present Lautman’s Platonism generally and then come back to the question of the relationship to Brunschvicg in order to distinguish Lautman’s Platonism from a naive Platonism. (11) ‘The objectivity of mathematical beings only reveals its true meanings in a theory of the participation of mathematics to a higher reality.’14 (12) ‘The reality inherent to mathematical theories comes from their participation to an ideal reality which dominates mathematics but can only be known through mathematics.’ Lautman introduces above mathematical theories other more general entities he calls dialectical Ideas. These Ideas go in opposites, global versus local, continuous versus discrete, essence versus existence. Mathematical theories represent various ways of expressing these oppositions. They intervene in every domain of thought but mathematics gives them their most precise expression. For instance, the ontological argument for the existence of God goes from the perfection of God to his existence because perfection is opposed to non-existence. Lautman tries to find procedures that would illustrate a similar relationship, or

Virtual Platonisms: Lautman and Gödel

223

a passage from ‘essence’ to ‘existence’, in mathematics. There is, for example, the semantic completeness of predicate calculus, which enables one to conclude from the consistency of a number of axioms to the existence of a domain of objects satisfying the axioms. Lautman gives other illustrations of such passages, in particular in the theory of algebraic functions. In the same way, mathematics of the nineteenth century offer different versions of the opposition of continuous versus discrete: the definition of real numbers as series of natural numbers can be seen as an attempt to master the continuous by means of the discrete, whereas the use of analysis in number theory shows the continuous as a background for the discrete. As regards the opposition of local and global, Lautman studies the interplay in contemporary mathematics (of the 1930s) between differential geometry where one can find a passage from the local to the global (if the variety is first defined by its local properties, such as the gradient, or the formula giving the distance between two points infinitely near to each other) and topology which rather goes from the global to the local (since the topology is defined by the whole set of open neighbourhoods in the space considered). The relationships between essence and existence, continuous and discrete, local and global represent universal problems: dialectical Ideas in Lautman’s terminology. Mathematical theories appear in order to express these oppositions. They constitute various ways to relate their terms. Objective reality, in Lautman’s view, belongs to the dialectical Ideas rather than to our mathematical theories. Dialectical Ideas transcend mathematical theories. They have an objective existence, independent of our acts, of the faculties of the mind and of our mathematics. In fact, the human mind creates mathematical theories in order to express or, in a sense, to incarnate the Ideas that it can only dimly see outside mathematics. Thus, mathematical theories are mere creations of the human mind and they have objectivity only insofar as they express Ideas that transcend them. The point I want to emphasize is that Lautman does not look for these Ideas in the axioms, in the principles of mathematical theories, but rather in their ‘movement’, or in the ‘activity’ of the mathematician. Lautman thereby retains Brunschvicg’s historical perspective. However, there is an ambiguity. In the discussion that follows the joint lecture of Cavaillès and Lautman at the Société française de Philosophie, E. Cartan and M. Fréchet distinguish Cavaillès and Lautman’s perspectives, because the latter considers the ‘present state of a certain number of mathematical theories’ as ‘not a historical point of view’.15 Contrary to Brunschvicg, who in Les étapes de la philosophie mathématique, starts with Greek mathematics and follows the history of mathematics up to the theory of groups, or Cavaillès, who, in his first thesis, discusses a whole century of analysis and set theory, Lautman does not study mathematical theories over long and definite periods. Nevertheless, the way in which Lautman considers mathematical theories depend on Brunschvicg’s historical method. Lautman always stresses that dialectical ideas only express themselves and reveal themselves in the ‘movement of a mathematical theory’,16 that is not in the axioms but in the

224

Postanalytic and Metacontinental

way the theory develops. It is a history that is punctual, so to speak, or instantaneous, as one speaks of the instantaneous speed of a mobile. But the ‘method of descriptive analysis’ that Lautman uses on contemporary theories remains, in its inspiration, historical. He acknowledges this methodological debt when he defines his position by associating Brunschvicg with Hilbert. From Hilbert, Lautman borrows the idea that mathematical theories may be analysed in a meta-theory that reveals their ‘logical properties’. These ‘logical properties’ are quite different from those that Hilbert had in mind in his formalist program, for, contrary to properties such as consistency, which a theory verifies, Lautman refers to properties susceptible of an ‘infinity of degrees’.17 These ‘logical properties’ are in fact dialectical Ideas. What Lautman takes from Hilbert, and which is indeed an ingredient of the formalist program, is the notion that mathematics, when seen from a non-mathematical point of view, may reveal certain features which the mathematician cannot see. From Brunschvicg, Lautman learns to turn his attention to the ‘movement’ of mathematics. His aim is to: (13) ‘develop a conception of the mathematical reality which would associate the fixity of logical notions with the movement which makes the theories alive.’18 The reference to the ‘movement’ of mathematical theories comes from Brunschvicg and his historical method. The difference with Brunschvicg and Cavaillès, from this point of view, is simply that Lautman studies the actual, or recent development of mathematical theories. Thus the dialectical ideas incarnate themselves in a movement, or an activity, even though they dominate this activity: (14) ‘The reality of mathematics is not the act of the intelligence that creates or understands, but it is in this act that it appears to us.’19 (15) ‘Beyond the temporal conditions of the mathematical activity but in this activity itself, the lineaments appear of an ideal reality which dominates the mathematical material.’20 When Lautman criticizes Brunschvicg or Cavaillès, he only reproaches them for missing this ideal reality but, in so doing, he acknowledges that the locus of ideas is the field itself, mathematical history, mathematical experience, as studied by Brunschvicg and Cavaillès. (16) Against Cavaillès: ‘One should see in experience something else and something more than experience.’21 (17) Against Brunschvicg: ‘Thus one can see the effort of intelligence for creating or understanding as a spiritual experience, but this experience has another content than the mathematics which develops at the same time.’22

Virtual Platonisms: Lautman and Gödel

225

In these quotations, Lautman clearly turns his attention to the same field, to the same ‘experience’ as Cavaillès and Brunschvicg. The shift which defines Lautman’s Platonism in this historical epistemology seems simply to consist in interpreting the intellectual factor of objectivity, which, for Brunschvicg and Bachelard, is reason, but for Cavaillès is the necessity of mathematical history as the expression of an ideal reality dominating mathematics. For example, (18) ‘The objectivity of mathematical beings only reveals its true meaning in a theory of the participation of mathematics to a higher reality.’23 This shift leaves Lautman with the problem of interpreting the second factor of objectivity which for Brunschvicg is experience. I will not go into this problem, but it is the main difficulty Lautman encounters all along his philosophical career and which he will never solve in an entirely satisfactory manner. As I have argued elsewhere,24 one can find at least three different attempts at a solution: in 1937, Lautman considers, in a manner similar to Cavaillès, that in mathematics experience is made of signs;25 his 1938 thesis relates the ‘procession of ideas’ to mathematical examples as the matter foreign to Ideas but in which they incarnate themselves; and in his last paper published posthumously in 1946, Lautman refers to the actual matter of the empirical world. I will make two concluding remarks on Lautman’s Platonism. The first point, which I have discussed at length, is that the locus of Ideas in Lautman’s Platonism is mathematical ‘movement’, the mathematical ‘activity’, mathematical ‘experience’, to use Cavaillès term which Lautman sometimes takes up, rather than the axioms of theories. Dialectical Ideas express and reveal themselves in mathematical activity. Lautman’s Platonism is a way to understand mathematical activity or mathematical history. And, because his point of view is historical and gives the same objectivity to the different theories following from one another, Lautman no more than Cavaillès can give a foundation to mathematical theories, a ‘foundation’ as we usually understand the word, referring for example to a programme like Hilbert’s. In fact, Lautman’s Platonism does not aim to give a foundation to mathematical theories but aims at describing their development. The second point follows from the first. Ideal reality, in Lautman’s Platonism, is not in the objects of mathematical theories but in the notions that the development of these theories involves. The objects of set theory are sets. The universe which the theory describes is a universe of sets. However, in Lautman’s Platonism, it is not the universe of sets which is supposed to have an independent reality, but the dialectical opposition of continuous and discreet to which set theory seems to give a new form. The ideal opposition of continuous versus discreet, does not even appear in the primitive notions of set theory. Ideas are, in Lautman’s view, a non-sensible reality from which theories are created, ideal problems to which mathematical theories give different solutions, but not necessarily primitive notions of the theory nor literally objects of the theory.

Postanalytic and Metacontinental

226

Lautman’s Platonism is not a naive Platonism according to which mathematical theories are taken to describe and duplicate an ideal reality. The ideal reality is rather that from which mathematical theories are created or that which governs their development. Lautman, with Cavaillès, rejects naive Platonism: (19)

Cavaillès: ‘I believe that the conception of systems of mathematical objects existing in themselves is in no way necessary to ground mathematical reasoning. Lautman: I acknowledge with Cavaillès the impossibility of such a conception of an immutable universe of ideal mathematical beings.’26

Lautman is explicit: (20) ‘We do not understand by Ideas models of which mathematical beings would only be copies.’27 Again, the problem for Lautman is not to give a foundation to mathematical theories by duplicating their objects in an ideal reality. It is rather to reveal the necessary requisites of their development as an ideal reality.

2. Gödel’s Platonism There is no doubt that Gödel comes from a different tradition than Lautman. The recourse to Platonism is first motivated by a foundational purpose: to give a foundation to mathematical theories. I have tried elsewhere to analyse Gödel’s arguments for his Platonism28 and I will not repeat this analysis here. Put simply, from the 1930s to1950s, we find the following argument across different texts. The second theorem that Gödel proves in 1931 establishes that if arithmetic is consistent, in order to prove its consistency, one needs arguments that cannot be translated into arithmetic and that are, therefore, at least as powerful as those of arithmetic. The same goes for any formal system capable of expressing arithmetic. That implies what Gödel calls in the 1950s the ‘non eliminability of mathematical content’.29 Hilbert’s idea, in the 1920s, was to give a foundation to mathematical theories by proving the consistency of their axioms, but in order to prove the consistency of a system of axioms, one needs to use arguments that are as powerful as those that are formalized in the axioms. Furthermore, in this proof of consistency, one must give intuitive evidence for these arguments. Thus one must acknowledge that certain non-trivial, non-finitist mathematical propositions have an intuitive evidence. In Gödel’s view, this evidence comes from the fact that these propositions express an ideal reality, a sort of mathematical realm that we intuit. Or, more precisely, Gödel distinguishes two ways in which mathematical axioms may be founded. Axioms may first be founded by

Virtual Platonisms: Lautman and Gödel

227

their evidence and that means, for mathematical axioms properly speaking, by the intuition of the ideal reality to which they refer. But axioms may also be founded by a process similar to induction. In cases when an axiom implies several propositions which are proven on a different basis, or arithmetical propositions which we can verify on numerical examples, then we are justified in accepting the axiom, just as, in physical science, we may accept a law which has observable consequences. This second criterion for the admissibility of an axiom is then always provisional, for one may eventually discover that the axiom in question either has some unlikely consequences or may be supervened by a more general proposition. In any case, (21) ‘For these axioms, there exists no other rational (and not merely practical) foundation except either that they (or propositions implying them) can be directly perceived to be true (owing to the meaning of the terms or by an intuition of the objects falling under them), or that they are assumed (like physical hypotheses) on the grounds of inductive arguments, e.g. their success in the applications.’30 Gödel’s Platonism is related to the problem of the foundations of mathematics and the kind of Platonism to which Gödel is referring is naive. It is to the objects of the theory, to the universe that the theory seems to describe, that an ideal reality is attributed. The ideal reality is mirrored, duplicated in the theory. Quotation (1) also seemed to allude to such naive Platonism. Nevertheless, in the draft of a lecture dating from 1951 and again in the conversations with H. Wang, from the 1970s, we find another argument leading to another kind of Platonism. I have argued elsewhere that it is only this more qualified Platonism that Gödel supports after 1964. I will try to show here that there is, in Gödel’s texts, beside the reference to a naive Platonism, another line of thought which would lead what I call virtual Platonism. Here is the argument. Mathematics is not ‘our own creation’: (22) ‘For the creator necessarily knows all the properties of his creatures, because they can’t have any others except those he has given to them. [. . . Therefore,] mathematical objects and facts (or at least something in them) exist objectively and independently of our mental acts and decisions, that is to say, [it seems to imply] some form or other of Platonism or “realism” as to the mathematical objects. [. . .] One might object that the constructor needs not necessarily know every property of what he constructs. For example, we build machines and still cannot predict their behavior in every detail. But this objection is very poor. For we don’t create the machines out of nothing, but build them out of some given material. If the situation were similar in mathematics, then this material or basis for our constructions would be something objective and would force some realistic viewpoint upon us even if certain other ingredients of mathematics were our own creation. The same would be

228

Postanalytic and Metacontinental

true if in our creations we were to use some instrument in us but different from our ego (such as “reason” interpreted as something like a thinking machine)’.31 The penultimate sentence was first written as: (22’) ‘Similarly, of course, if mathematics were not created by our acts (i.e. conscious acts of our ego) but by some entity in us called reason, mathematical facts would be something objective namely properties of this mind, which we could in no way determine by our choices.’32 The turning point in the argument is that, when one truly creates an object, one completely knows this object: one knows what one has put in the object so that the object cannot have properties that one does not know about. Thus, since mathematical objects, numbers for example, have properties which we do not know – there are yet undecided mathematical conjectures – therefore mathematical objects cannot have been truly created by us. However, Gödel leaves open two possibilities: that mathematical objects have been created on the basis of something else, which give them their objectivity, or that they have been created by the human mind but unconsciously, in such a way that one cannot re-enact this creation and find out what has been put into the object. The argument calls for several remarks. Above all, the criterion determining the objectivity of mathematical objects is the existence of undecided mathematical propositions: the fact that we cannot prove all the theorems of arithmetic or that there are properties of which we do not know whether numbers possess them. It is this indecision that enables Gödel to conclude that mathematical objects cannot have been created by the human mind. There are several ways to interpret this indecision. Gödel makes the hypothesis that there are absolutely undecidable mathematical propositions, propositions that the human mind will never be able to decide, and the argument is supposed to show that that would entail the objectivity of mathematics. The incompleteness theorem Gödel published in 1931 establishes, in any consistent system containing arithmetic, the existence of proposition that is undecidable on the basis of the axioms. But it does not entail the existence of absolutely undecidable propositions, because a proposition undecidable in a given system may be decided in a stronger system, which will itself contain undecidable propositions or propositions decidable only in a yet stronger system. Gödel denies the existence of absolutely undecidable mathematical propositions and believes that there is always a way to complete a formal system and, eventually, to decide any clearly formulated mathematical problem on the basis of sound axioms. Thus, the above argument is made under a hypothesis that Gödel does not accept. Later on in the text, however, Gödel uses the same argument with reference to our actual ignorance. After the reflection that has been carried out on the

Virtual Platonisms: Lautman and Gödel

229

foundations of mathematics and the ‘perfect exactness’ achieved concerning the reasonings of elementary mathematics, there still remain undecided arithmetical propositions. This shows that numbers cannot have been freely created by the human mind.33 Thus, the incompleteness of our knowledge establishes the objectivity of mathematics: the fact that mathematical objects are not adequately mastered and escape adequate knowledge. This implies, according to the above argument, that mathematical objects cannot have been truly created or, more precisely, that one cannot re-enact their constitution. Gödel’s Platonism, in this argument, lies entirely in the idea that the constitution of mathematical objects bears an opacity. It cannot be completely analysed. The analysis of the constitution of mathematical objects must eventually meet an obstacle, either a kind of datum that is received from the outside or an element which the human mind has created unconsciously and whose creation cannot be completely explained. This second possibility would make mathematics analogous to a dream, a story that is unconsciously produced and of which we do not know the end, nor the meaning. Mathematical axioms would be like the background of the dream: a series of postulates that determine a universe, which we have invented but which continues to surprise us or that we cannot completely account for. Gödel explores this possibility in some remarks of his philosophical notebooks: (23) ‘In mathematics, [the] question is to find out what we have perhaps unconsciously created.’34 (23’) ‘The method for the foundation of knowledge is then psychoanalysis.’35 Lautman’s Platonism exemplifies the first possibility, the creation of mathematical objects out of something else. Lautman acknowledges that mathematical theories are creations of the human mind, but he insists that, nevertheless, mathematical theories incarnate, and are governed by, Ideas which have an independent reality. Platonistic Ideas are an ideal reality of which we have a dim intuition. They are ideal problems we aim at giving new expressions and new solutions to. We create mathematical theories as an answer to these problems, as an expression for these ideas and, in that sense, on the basis of this ideal reality. The Ideas are an ingredient which we cannot account for or which has an independent reality in the constitution of mathematics. It is a ‘something out of which’ mathematical objects may be created. Gödel notes a passage from Lautman’s conference of 1937 similar to quotation (12) illustrating this common point between Gödel and Lautman. In 1964, Gödel himself advocates a view illustrating the idea of a creation of mathematical objects out of something else, incompatible with naive Platonism, and according to which mathematical theories literally reflect an independent reality. At first sight, Gödel seems to make strong Platonic claims. The question

230

Postanalytic and Metacontinental

of mathematical objects is considered as analogous to the question of sense objects, the objects of perception: (24) ‘The question of the objective existence of the objects of mathematical intuition [. . .] is an exact replica of the question of the objective existence of the outer word.’ ‘I don’t see any reason why we should have less confidence in this kind of perception, i.e. in mathematical intuition, than in sense-perception.’36 Mathematical objects must then have the same reality as sense objects. But, precisely, Gödel denies that sense objects as they appear to us have an independent reality. Sense objects are constituted on the basis of sensations. We then contribute, through certain mental processes, to shaping our objects from the data that we receive from the outside. The same goes for mathematical objects: (25) ‘It should be noted that mathematical intuition need not be conceived of as a faculty giving an immediate knowledge of the objects concerned. Rather it seems that, as in the case of physical experience, we form our ideas also of those objects on the basis of something else, which is immediately given. Only this something else here is not, or not primarily, the sensations.’37 We have a mathematical intuition which, for Gödel, is ‘a mere psychological fact’.38 But that does not imply that the objects of our theories have an independent reality. The objects are already informed by the subject, just like the objects of perception. Mathematical objects are merely constituted by the subject on the basis of some datum, ‘something that is immediately given’. Gödel does not deny that there is an ideal reality at the basis of mathematics. But he cannot hold that mathematical theories literally describe that ideal reality. Mathematics may rely on but do not describe an independent reality. What it seems to describe, a universe of sets, does not exist independently of the mind. In fact, in Gödel’s ‘Platonism’, mathematical objects, such as sets, may be seen as ‘mixture of subjective and objective elements’: (26) ‘The second alternative [that mathematical objects are created out of something else] is perfectly sufficient to prove the Platonic point of view. For they then exist in the same sense as physical objects since our ideas of these no doubt are a mixture of objective and subjective elements.’39 His position is similar to Lautman’s: there is in mathematical theories the expression of an ideal reality but this may not look like the universes that our theories describe, because these universes are simply human creations made on the basis of, or out of, the ideal reality. The shift away from naive Platonism illustrated by quotation (1) does not necessarily entail the abandonment of the

Virtual Platonisms: Lautman and Gödel

231

foundational perspective associated with naive Platonism. Gödel still considers that ideal reality, though it is not adequately mirrored in our theories, may serve as justification for our axioms and for the rationalistic belief that every adequately posed mathematical question can be answered. Thus, Gödel conjectures that the continuum hypothesis is undecidable in the usual set theory – he himself has proved in 1938 that it cannot be refuted on the basis of ZF axioms – but he believes that the existence of an ideal reality out of which set theory has been constituted, even though this ideal reality may not be the universe of sets that our theory seems to describe, is sufficient to justify the belief that the continuum hypothesis can be unequivocally decided in a satisfying extension of set theory: (27) ‘The axioms force themselves upon us as being true.’40 (27’) ‘The existence of an intuition which is sufficiently clear to produce the axioms of set theory and an open series of extensions of them suffices to give meaning to the question of the truth or falsity of propositions like Cantor’s continuum hypothesis.’41 Gödel’s move from naive Platonism in the 1960s and 1970s is related to speculative aspects of his philosophy and, more precisely, to an investigation of the possibility of non-human rational beings and their mathematics. On the one hand, there is evidence that Gödel believes in the existence of such rational beings.42 On the other hand, Gödel acknowledges that some aspects of our mathematics are specifically human, such as those related to time. Some of these other rational beings would not know time,43 or not know the same kind of timeness, so that they would not have the same mathematics. Thus, if our mathematics refers to an ideal reality or if our mathematics and those of these higher beings have a common basis, then this ideal reality cannot be adequately mirrored in our mathematics but can only be considered as a source, an objective root out of which we develop our specifically human mathematics. These considerations are distant from Lautman’s. Nevertheless, it is an important concern in Lautman’s texts that the reference to Ideas may provide the same objectivity and relate different theories in the history of mathematics. The thesis that the ideal reality, as a source for mathematics rather than its object, may express itself differently in various theories, or in various mathematics, again seems to be common to Lautman and Gödel. My aim was to show that Gödel moves away from naive Platonism or at least that there are elements in Gödel’s texts leading in a different direction. The 1951 argument discussed above does not entail naive Platonism. Its core is that the incompleteness of mathematical knowledge implies that the constitution of mathematical theories cannot be completely traced. Mathematical theories may have been created by the human mind but this creation cannot be completely re-enacted. Either it is radically unconscious or – as Gödel prefers – it involves

232

Postanalytic and Metacontinental

an ideal reality, which expresses itself in our mathematics but need not be adequately mirrored in them and may express itself differently in various mathematics. This ideal reality is not the universe that our theories describe: it is rather a source which mathematical knowledge requires or something in mathematical activity which cannot be accounted for.

3. Conclusion: The notion and the source of objectivity in mathematics In naive Platonism, reality is in the universes of objects that our theories describe: a universe of sets for set theory, the domain of natural numbers for arithmetic. In virtual Platonism, these universes may be considered as artefacts, the reality is rather at the roots of the theory, in Ideas that are expressed by the theory, incarnated in the theory but not literally mirrored in it. We have here an example of an opposition, which concerns both the historical epistemology initiated by Brunschvicg and the foundational tradition. Though coming from different traditions, Lautman and Gödel both delineate virtual Platonisms. For Lautman, virtual Platonism is related to the historical method of his epistemology. The universes of our theories appear as historical products, corresponding to a stage of the theories which may be transitory. Like Cavaillès, Lautman refuses to consider the universes of our theories as an immutable reality. Like Brunschvicg, Bachelard or Cavaillès, the field of epistemology is, for Lautman, the history of sciences. It is therefore in the development of mathematics, rather than in its products, that Lautman looks for an ideal reality which can only express itself in this becoming and is never the true object of a particular theory. Gödel remains attached to the foundational perspective but he moves away from naive Platonism. Eventually, he accepts that mathematical objects as we know them, the universe of sets as set theory describes it, may be human creations, so long as there remains an opacity in this creation. It is this opacity that lends objectivity to mathematics. In the 1951 argument discussed earlier, Gödel considers as a form of Platonism or Realism any doctrine that acknowledges that mathematical theories are not created consciously and out of nothing. The objectivity of mathematics is in the (impossible to re-enact) creation of mathematics: it blocks somewhere, there is something that cannot be accounted for, and this something is what gives mathematics objectivity in Gödel’s sense. In fact, Gödel does not, or very rarely, use the term ‘objectivity’, but his argument leads to such an analysis: the objectivity of mathematics is what renders impossible a reflexive account of the development of mathematics. In Brunschvicg’s Idealism, objectivity in this sense is entirely in the experience. The development of mathematics is a dialogue between reason and experience. But every step of the development of reason can be reflexively elucidated. Experience escapes this reflexive analysis, in its initial nature, as a flux of sensation in the

Virtual Platonisms: Lautman and Gödel

233

first stage of the human development, and in its responses to the concepts of reason, in later stages. Experience cannot be understood and remains an opacity in the products of science, something for which no reason and no genesis can be given. Similarly, in Bachelard’s account, the objectivity of science comes from the imagination from which science slowly liberates us, but which remains as an opacity which science has never got rid of. As we have seen, Cavaillès distinguishes two factors of objectivity in mathematical becoming. The first, corresponding to Brunschvicg’s experience, is in the signs and the figures that the mathematician uses, the second, corresponding to Brunschvicg’s reason, is in the necessity of the history of mathematical becoming. Both these factors represent an opacity for the mind and, therefore, a source of objectivity in Gödel’s sense. The fact that the mathematical object is always incarnated in a system of signs or of figures is interpreted by Cavaillès as an inadequacy in its givenness: mathematical objects are never adequately given and therefore may be subjected to a becoming. Speaking of the Kantian notion of construction in intuition, Cavaillès claims that the insertion of the object in a multiplicity gives it a sort of independence with respect to the actual content of thought. There is, in the singular mathematical object, an affinity with what is not itself: the object goes beyond its definition and justifies progress outside itself.44 The history of mathematics is not transparent to reflection and Cavaillès opposes this to Husserl’s ‘The Origin of Geometry’. Husserl thought that the meaning of a concept contained sediments of the whole history of the concept, so that one only needed to analyse the concept as it is given in order to find schematically its history and its constitution. Cavaillès’ criticism is that the history of mathematics erases its steps. There are upheavals in the history of mathematics where a concept that is at the very end of the development is taken as a primitive notion and used as a building block in a new architecture. The previous history of the concept in question is erased and no longer appears in the concept, taken as a primitive notion: ‘One of the main problems for a theory of science is that progress is not growth in volume by simple juxtaposition, the old subsisting with the new, but perpetual revision of the contents by deepening and barring.’45 The past of the concept is not contained in the meaning of the concept as it intervenes in contemporary mathematics. Their insertion in history means that the constitution of mathematical theories cannot be analysed reflexively: mathematical theories as they are given to us do not reveal their past or the steps necessary for their elaboration. Their history gives objectivity, in Gödel’s sense, to mathematical theories. Here, the objectivity of mathematics is purely immanent to mathematical development. It comes from the way mathematical history folds into itself to mask itself behind its later stages. In Brunschvicg’s account, the objectivity of mathematics, in Gödel’s sense, comes from experience. In Lautman’s account, it comes from an ideal reality that expresses itself in

234

Postanalytic and Metacontinental

mathematics. For Cavaillès, it comes from the structure of mathematical becoming. Gödel’s logical results, in particular his incompleteness theorem, have been variously interpreted in French thought, but very little attention has been paid to Gödel’s philosophy. Gödel’s analysis of objectivity in mathematics, though stemming from a different tradition, meets, and at least allows an analysis of the historical epistemology initiated by Brunschvicg.

Notes 1

2

3

4

5 6 7

8 9

10

11

12

13

14 15 16

17

18 19 20 21 22

Kurt Gödel, ‘Some Basic Theorems on the Foundations’, in his Collected Works (Oxford: Oxford University Press, 5 volumes, 1986–2003), III, 323. Among these texts there are reflections bringing Gödel close to Lautman and virtual Platonism. In fact, Gödel read some of Lautman’s texts. In his notes, he copied a quotation from Lautman illustrating what I take to be their virtual Platonism. The notes on Lautman are two pages long, so I do not claim that he influenced Gödel. Léon Brunschvicg, Ecrits philosophiques (Paris: PUF., trois tomes, 1951–1958), II, 119. Léon Brunschvicg, L’expérience humaine et la causalité physique (Paris: Alcan, 1922), 552. Brunschvicg, Ecrits philosophiques, II, 116. Ibid., II, 118. Léon Brunschvicg, Les étapes de la philosophie mathématique (Paris: Blanchard, 1993), 498. Brunschvicg, L’expérsience humaine et la causalité physique, 584. Albert Lautman, ‘Essai sur les notions de structure et d’existence’ (1938), in his Les mathématiques, les idées et le réel physique (Paris: Vrin, 2006), 128. Jean Cavaillès and Albert Lautman, ‘La pensée mathématique’ in Œuvres complètes de philosophie des sciences, Jean Cavaillès ed. (Paris: Herman, 1994), 601. Jean Cavaillès, Remarques sur la formation de la théorie abstraite des ensembles (1938), in Œuvres complètes de philosophie des sciences, 226. Jean Cavaillès, Méthode axiomatique et formalisme (1938), in Œuvres complètes de philosophie des sciences, 179. Jean Cavaillès, ‘Transfini et continu’ (1947), in Œuvres complètes de philosophie des sciences, 471. Cavaillès and Lautman, ‘La pensée mathématique’, 605. Ibid., 609, 617. Albert Lautman, ‘De la réalité inhérente aux théories mathématiques’ (1937), in Les mathématiques, les idées et le réel physique, 66. Albert Lautman, ‘Essai sur les notions de structure et d’existence’ (1938), in Les mathématiques, les idées et le réel physique, 131. Ibid., 131. Lautman, ‘De la réalité inhérente aux théories mathématiques’, 66. Lautman, ‘Essai sur les notions de structure et d’existence’, 229–230. Cavaillès and Lautman, ‘La pensée mathématique’, 630. Lautman, ‘Essai sur les notions de structure et d’existence’, 229.

Virtual Platonisms: Lautman and Gödel 23 24

25 26 27

28

29

30 31 32 33 34 35

36

37 38 39 40 41 42

43

44 45

235

Cavaillès and Lautman, ‘La pensée mathématique’, 605. See Pierre Cassou-Noguès, ‘Le platonisme de Lautman dans la tradition d’épistémologie en France’ Philosophiques, Numéro spécial sous la direction J.-P. Marquis, forthcoming. Lautman, ‘De la réalité inhérente aux théories mathématiques’, 68. Cavaillès and Lautman, ‘La pensée mathématique’, 603, 605. Albert Lautman, ‘Nouvelles recherches sur la structure dialectique des mathématiques’ (1939), in Les mathématiques, les idées et le réel physique, 238. Pierre Cassou-Noguès, ‘Gödel and the Question of the “Objective Existence” of Mathematical Objects’, History and Philosophy of Logic 26 (2005): 211–228; Pierre Cassou-Noguès, ‘On Gödel’s Platonism’, unpublished manuscript. Kurt Gödel, ‘Is Mathematics Syntax of Language?’ (1953–1959), in his Collected Works (Oxford: Oxford University Press, 5 volumes, 1986–2003), III, 345. Gödel, ‘Is Mathematics Syntax of Language?’, 347. Gödel, ‘Some Basic Theorems on the Foundations’, 312. Gödel’s Papers, box 8b, folder 93, item 040294. Gödel ‘Some Basic Theorems on the Foundations’, 14. Gödel’s Papers, box 8c, folder 117, item 040403. Gödel’s Papers, Philosophical notebook, transcription from Gabelsberger by C. Dawson, V, 344. Kurt Gödel, ‘What is Cantor’s Continuum Problem?’, Supplement to the second edition (1964), in his Collected Works, II, 268. Ibid. Ibid. Gödel’s Papers, box 20, from Gödel’s hand on a text by Wang, my emphasis. Gödel, ‚What is Cantor’s Continuum Problem?’, 268. Ibid., 268. For example, ‘there are other worlds and other rational beings of a different and higher kind’, conversation with Wang, quoted in Hao Wang, A Logical Journey (Cambridge, MA: MIT Press, 1996), 316. These weird aspects of Gödel’s thought are further described in Palle Yourgrau, Gödel Meets Einstein: Time Travel in the Gödel Universe (Chicago: Open Court, 1999) and World Without Time (New York: Basic Books, 2005), and Pierre Cassou-Noguès, Les démons de Gödel, (Paris: Seuil, 2007). ‘For a being which had no sensibility at all (i.e. no contact through sensations with reality) but only “pure understanding”, no time at all would exist.’ Gödel, ‘Some observations . . .’, appendix A (1946–1949), in his Collected Works, III, 427. Cavaillès, ‘Transfini et continu’, 469. Jean Cavaillès, Sur la logique et la théorie de la science (1947), in Œuvres complètes de philosophie des sciences, 560.

Bibliography

Adler, Jonathan. ‘An Overlooked Argument for Epistemic Conservatism’. Analysis 56 (1996): 80–84. Adorno, Theodor W. Against Epistemology. Translated by Willis Domingo. Cambridge MA: The MIT Press, 1982. Alcoff, Linda. Real Knowing: New Versions of the Coherence Theory. Ithaca: Cornell University Press, 1996. Allen, Richard and Murray Smith. ‘Introduction: Film Theory and Philosophy’. In Film Theory and Philosophy, edited by Richard Allen and Murray Smith, 1–35. Oxford: Oxford University Press, 1997. Allison, Henry E. Kant’s Theory of Freedom. Cambridge: Cambridge University Press, 1990. —Kant’s Transcendental Idealism. Newhaven: Yale University Press, 1986. Ameriks, Karl. Kant and the Historical Turn. New York: Oxford University Press, 2006. Anscombe, Elizabeth. ‘Under a Description’. In Events, edited by Achille Varzi and Roberto Casati, 303–318. Aldershot: Dartmouth, 1996. Antiseri, D. The Weak Thought and Its Strength. Ashgate: Aldershot, 1996. Apel, Karl. ‘The Problem of Philosophical Foundations in Light of a Transcendental Pragmatics of Language’. In After Philosophy: End or Transformation, edited by Kenneth Baynes, James Bohman and Thomas A. McCarthy. Cambridge MA: MIT Press, 1987. Armstrong, David M. A Materialist Theory of the Mind. London: Routledge & Kegan Paul, 1968. Austin, J. L. Philosophical Papers. Oxford: Clarendon Press, 1961. Avramides, Anita. Other Minds. London: Routledge, 2001. Ayer, A. J. Language, Truth and Logic. London: Gollancz, 1946. Babich, Babette. ‘On the Analytic-Continental Divide in Philosophy: Nietzsche’s Lying Truth, Heidegger’s Speaking Language, and Philosophy’. In A House Divided: Comparing Analytic and Continental Philosophy, edited by C. G. Prado, 63–103. Amherst, NY: Prometheus/Humanity Books, 2003. Bach, Kent. ‘Actions Are Not Events’. In Events, edited by Achille Varzi and Roberto Casati, 413–429. Aldershot: Dartmouth, 1996. Baird, Bryan. ‘The Transcendental Nature of Mind and World’. Southern Journal of Philosophy 44, No. 3 (2006): 381–398. Baker, Deane-Peter, and Patrick Maxwell, eds. Explorations in Contemporary Continental Philosophy of Religion. Amsterdam: Rodopi, 2003. Baldwin, Thomas. ‘Death and Meaning’. In Arguing with Derrida, edited by S. Glendinning, 89–102. London: Blackwell, 2001. Balfour, A. J. ‘Transcendentalism’. Mind 3 (1878): 480–505.

Bibliography

237

Bealer, George and Strawson, Peter. ‘The Incoherence of Empiricism’. Proceedings of the Aristotelian Society, Supplementary Volume 66 (1992): 99–143. Bennett, Jonathan. Events and their Names. Oxford: Clarendon, 1988. Bennington, Geoffrey. ‘For the Sake of Argument (Up to a Point)’. In Arguing with Derrida, edited by Simon Glendinning, 34–51. Oxford: Blackwell, 2001. —Other Analyses: Reading Philosophy. CreateSpace, 2008. Bermudez, Jose. ‘The Phenomenology of Bodily Awareness’. In Phenomenology and Philosophy of Mind, edited by D. Smith, and A. Thomasson, 295–316. Oxford: Oxford University Press, 2006. —‘Transcendental Arguments and Psychology: The Example of O’Shaughnessy on Intentional Action’. Metaphilosophy 26, No. 4 (1995): 379–401. Bernstein, Richard. ‘McDowell’s Domesticated Hegelianism’. In Reading McDowell: On Mind and World, edited by Nicholas Smith. London: Routledge, 2002. Biletzki, Anat. ‘Bridging the Analytic-Continental Divide’. International Journal of Philosophical Studies 9, No. 3 (2001): 293–294. BonJour, Laurence. In Defense of Pure Reason. Cambridge: Cambridge University Press, 1998. —The Structure of Empirical Knowledge. Cambridge, MA: Harvard University Press, 1985. Bordwell, David and Noël Carroll, eds. Post-Theory: Reconstructing Film Studies. Madison: University of Wisconsin Press, 1996. Borradori, Giovanna. ‘Two Versions of Continental Holism: Derrida and Structuralism’. Philosophy & Social Criticism 26, No. 4 (2000). Borradori, Giovanna and Rosanna Crocitto. The American Philosopher. Chicago: University of Chicago Press, 1994. Bostrom, Nick. ‘Are You Living in a Computer Simulation?’ Philosophical Quarterly 53 (2003), 243–255. Brandom, Robert B. Articulating Reasons. Cambridge, MA: Harvard University Press, 2000. —‘The Pragmatist Enlightenment (and its Problematic Semantics)’. European Journal of Philosophy 12, No. 1 (2004): 1–16. —Making It Explicit. Cambridge. MA: Harvard University Press, 1994. Brandom, Robert B. ed. Rorty and His Critics. Oxford: Wiley-Blackwell, 2001. —‘Some Pragmatist Themes in Hegel’s Idealism: Negotiation and Administration in Hegel’s Account of the Structure and Content of Conceptual Norms’. European Journal of Philosophy 7, No. 2 (2000): 164–189. —Tales of the Mighty Dead. Cambridge. MA: Harvard University Press, 2002. Brunschvicg, Léon. Ecrits philosophiques. Paris: PUF., trois tomes, 1951–1958. —L’expérience humaine et la causalité physique. Paris: Alcan, 1922. —Les étapes de la philosophie mathématique. Paris: Blanchard, 1993. Caird, Edward. ‘Mr Balfour on Transcendentalism’. Mind 4 (1879): 111–114. Canguilhem, Georges, Etudes d’histoire et de philosophie des sciences concernant les vivants et la vie [1968]. Paris: Vrin, 1994. Carman, Taylor. ‘On the Inescapability of Phenomenology’. In Phenomenology and Philosophy of Mind, edited by D. Smith and A. Thomasson, 67–89. Oxford: Oxford University Press, 2006. Carnap, Rudolf. ‘Intellectual Autobiography’. In The Philosophy of Rudolf Carnap, edited by Paul Schilpp, 3–87. La Salle: Open Court, 1963.

238

Bibliography

Carroll, Noël. Mystifying Movies: Fads and Fallacies in Contemporary Film Theory. New York: Columbia University Press, 1988. —‘The Power of Movies’. Daedalus, 114.4 (Fall 1985): 79–104. —Philosophical Problems of Classical Film Theory. Princeton: Princeton University Press, 1988. —‘Prospects for Film Theory: A Personal Assessment’. In Post-Theory, edited by David Bordwell and Noël Carroll, 37–68. Madison: University of Wisconsin Press, 1996. Cassou-Noguès, Pierre. ‘Gödel and the question of the “objective existence” of mathematical objects’. History and Philosophy of Logic (2005): t.26, 211–228. —‘Le platonisme de Lautman dans la tradition d’épistémologie en France’, Philosophiques. Numéro spécial sous la direction J.-P. Marquis, forthcoming. —Les démons de Gödel. Paris: Seuil, 2007. —‘On Gödel’s Platonism’, unpublished manuscript. Cavaillès, J. Œuvres complètes de philosophie des sciences. Paris: Hermann, 1994. Cavaillès, J. and A. Lautman. ‘La pensée mathématique’. Œuvres complètes de philosophie des sciences, edited by J. Cavaillès, 593–630. Paris: Hermann, 1994. Cavell, Stanley. Contesting Tears: The Melodrama of the Unknown Woman. Chicago: University of Chicago Press, 1996. —The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy. New York: Oxford University Press, 1999. —Pursuits of Happiness: The Hollywood Comedy of Remarriage. Cambridge, MA: Harvard University Press, 1981. —Themes out of School: Effects and Causes. Chicago: University of Chicago Press, 1988. —The World Viewed: Reflections on the Ontology of Film. Cambridge, MA: Harvard University Press, 1979. Chisholm, Roderick. ‘Events and Propositions’. In Events, edited by Achille Varzi and Roberto Casati, 89–98. Aldershot: Dartmouth, 1996. —The Problem of the Criterion. Milwaukee: Marquette University Press, 1973. —‘A Version of Foundationalism’. Midwest Studies in Philosophy 5 (1980): 551–552. Christensen, David. ‘Clever Bookies and Coherent Beliefs’. Philosophical Review 100 (1991): 229–247. —‘Diachronic Coherence vs. Epistemic Impartiality’. Philosophical Review 109 (2000): 349–371. Churchland, Paul. ‘Eliminative Materialism and the Propositional Attitudes’. Journal of Philosophy 78 (1981): 67–90. Clark, A. Being There: Putting Brain, Body, And World Together Again. Cambridge MA: MIT Press, 1997. Cohen, J. L. The Dialogue of Reason: An Analysis of Analytic Philosophy. Oxford: Clarendon, 1986. Critchley, Simon. Very Little . . . Almost Nothing: Death, Philosophy, Literature. London: Routledge, 2004. Currie, Gregory. Image and Mind: Film, Philosophy, and Cognitive Science. New York: Cambridge University Press, 1995. —‘The Long Goodbye: The Imaginary Language of Film’. The British Journal of Aesthetics 33.3 (July 1993): 207–219. —‘Visual Fictions’. Philosophical Quarterly, 41.163 (April 1991): 129–143.

Bibliography

239

Daniels, Norman. ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’. Journal of Philosophy 76 (1979): 256–282. Danto, Arthur C. 1986. The Philosophical Disenfranchisement of Art. New York: Columbia University Press. Davidson, Donald. ‘A Coherence Theory of Truth and Knowledge’. In Kant oder Hegel?, edited by Dieter Henrich, 427–428. Stuttgart: Klett–Cotta, 1983. —Essays on Actions and Events. Oxford: Clarendon, 2001. —Davidson, Donald. ‘On the Very Idea of a Conceptual Scheme’. Proceedings and Addresses of the American Philosophical Association 47 (1974): 14–16. —Subjective, Intersubjective and Objective. Oxford: Oxford University Press, 2001. —‘The Third Man’. Critical Inquiry 19, No. 4 (1993): 607–608. de Beistegui, Miguel. Truth and Genesis. Bloomington, IN: Indiana University Press, 2004. De Preester, H. ‘From Ego to Alter Ego: Husserl, Merleau-Ponty and a Layered Approach to Intersubjectivity’. Phenomenology and the Cognitive Sciences 7, No. 1 (2008): 133–142. Deleuze, Gilles. Cinéma 1. Paris: Minuit, 1983. —Cinema 1: The Movement-Image. Translated Hugh Tomlinson and Barbera Habberjam. Minneapolis: University of Minnesota Press, 1985. —Cinéma 2. Paris: Minuit, 1985. —Cinema 2: The Time-Image. Translated Hugh Tomlinson and Robert Galatea. London: Athlone Press, 1989. —Difference and Repetition. Translated by Paul Patton. New York: Columbia University Press, 1994. —Différence et répétition. Paris: PUF, 1969. —Expressionism in Philosophy: Spinoza. Translated by Martin Joughin. London: Zone, 1990. —Kant’s Critical Philosophy. Minneapolis: University of Minnesota Press, 1985. —The Logic of Sense. Translated by M. Lester and C. Stivale. New York: Columbia University Press, 1990. Deleuze, Gilles and Félix Guattari. A Thousand Plateaus: Capitalism and Schizophrenia. Translated by B. Massumi. Minneapolis: University of Minnesota Press, 1987. —Anti-Oedipus: Capitalism and Schizophrenia. Translated by R. Hurley, M. Seem & H. Lane. New York: Viking Press, 1977. Dennett, Daniel. ‘Cognitive Wheels: The Frame Problem of AI’. In Minds, Machines and Evolution: Philosophical Studies, edited by C. Hookway, 129–151. Cambridge: Cambridge University Press, 1984. —Consciousness Explained. Boston: Little, Brown and Company, 1991. Derrida, Jacques. The Ear of the Other. Translated by Peggy Kamuf. Lincoln: University of Nebraska Press, 1988. —On Touching – Jean Luc Nancy. Translated by C. Irizarry. Palo Alto: Stanford University Press, 2005. —Writing and Difference. Translated by Alan Bass. Chicago: University of Chicago Press, 1981. Descartes, René. Oeuvres de Descartes (11 volumes). Edited by Charles Adam & Paul Tannery. Paris: Vrin, 1996.

240

Bibliography

—The Philosophical Writings of Descartes (3 volumes). Translated by John Cottingham, Robert Stoothoff, Dugald Murdoch and Anthony Kenny. Cambridge: Cambridge University Press, 1991. Devitt, Michael. Realism and Truth. Princeton: Princeton University Press, 1984. Diamond, Cora. ‘The Importance of Being Human’. In Human Beings, edited by D. Cockburn, 35–62. Cambridge: Cambridge University, Press 1991. —The Realistic Spirit: Wittgenstein, Philosophy, and the Mind. Cambridge: The MIT Press, 1996. Dombrowski, D. Rethinking the Ontological Argument: A Neoclassical Theistic Response. Cambridge: Cambridge University Press, 2006. Domenech-Oneto, P. and Roque, T. ‘L’objectivité des problèmes et la question du sujet. Considérations sur l’Idée dans la philosophie de Deleuze’. In Le concept, le sujet et la science, edited by P. Cassou-Noguès, P. Gillot, Paris: Vrin, 2009 (forthcoming). Dretske, F. I. ‘Can events move?’ In Events, edited by Achille Varzi and Roberto Casati, 413–429. Aldershot: Dartmouth, 1996. —Perception, Knowledge and Belief. Cambridge: Cambridge University Press, 2000. Dreyfus, Hubert. Being-in-the-World: A Commentary on Heidegger’s Being and Time, Division I. Cambridge, MA: MIT Press, 1990. —What Computers Still Can’t Do: A Critique of Artificial Reason. Cambridge, MA: MIT Press, 1992. —‘Why Heideggerian AI Failed and How Fixing It Would Require Making It More Heideggerian’. In The Mechanical Mind in History, edited by. P. Husbands, O. Holland and M. Wheeler, 331–371. Cambridge, MA: MIT Press, 2008. (Alternative versions of this paper appear under the same title in Philosophical Psychology 20, No. 2 (2007): 247–268; and in Artificial Intelligence 171 (2007): 1137–1160.) Dreyfus, Hubert. and S. Kelly. ‘Heterophenomenology: Heavy-handed Sleightof-hand’, Phenomenology and Cognitive Sciences 6 (2007): 45–55. Duffy, Simon. ‘Deleuze and the Mathematical Philosophy of Albert Lautman’. In Deleuze’s Heritage, edited by G. Jones and J. Roffe. Edinburgh: Edinburgh University Press, 2009. Dummett, Michael. Elements of Intuitionism. Oxford: Oxford University Press, 1977. —The Logical Basis of Metaphysics. Cambridge, MA: Harvard University Press, 1991. —Truth and Other Enigmas. London: Duckworth, 1978. —Truth and the Past. New York: Columbia University Press, 2004. Engel, M. S. ‘Schopenhauer’s Impact on Wittgenstein’. The Journal of the History of Philosophy 7 (1969): 285–302. Engel, Pascal. ‘Analytic Philosophy and Cognitive Norms’. The Monist 82, No. 2 (April 1999): 218–234. —‘Continental Insularity: Contemporary French Analytical Philosophy’. In Contemporary French Philosophy, edited by A. Griffiths, 1–20. Cambridge: Cambridge University Press, 1989. —‘Is There Any Such Thing as Post-Analytic Philosophy?’ In Conférence Clermont Ferrand Société de Philosophie d’Auvergne. 1995. See http://www. unige.ch/lettres/philo/enseignants/pe/Engel%201995%20post%20analytic %20philosophy.pdf. Everitt, N. The Non-Existence of God. London: Blackwell, 2004.

Bibliography

241

Finke, Ståle. ‘Concepts and Intuitions: Adorno after the Linguistic Turn’. Inquiry: An Interdisciplinary Journal of Philosophy 44, No. 2 (2001): 171–200. Flaxman, Gregory. The Brain Is the Screen: Deleuze and the Philosophy of Cinema. Minneapolis: University of Minnesota Press, 2000. Fodor, Jerry. The Language of Thought. Cambridge, MA: Harvard University Press, 1975. —The Modularity of Mind. Cambridge, MA: MIT Press, 1983. Foley, Richard. ‘Epistemic conservatism’. Philosophical Studies 43 (1983): 165–182. Foucault, Michel. Archaeology of Knowledge. Translated by A. Sheridan. London: Routledge, 2002. —The Order of Things: Archaeology of the Human Sciences. New York: Vintage Books, 1970. Frampton, Daniel. Filmosophy. London: Wallflower Press, 2006. Franceschini, N., J. M. Pichon and C. Blanes. ‘From Insect Vision to Robot Vision’. Philosophical Transactions of the Royal Society, series B. 337 (1992): 283–294. Freeman, Walter. How Brains Make Up Their Minds. New York: Columbia University Press, 2000. Freud, Sigmund. Die Traumdeutung. Leipzig/Vienna, Franz Deuticke, 1900. Friedman, Michael. ‘Exorcising the Philosophical Tradition’. In Reading McDowell: On Mind and World, edited by Nicholas Smith, London: Routledge, 2002. Fumerton, Richard. ‘Epistemic Conservatism: Theft or Honest Toil?’ Oxford Studies in Epistemology 2 (2008): 63–86. Gallagher, Shaun. ‘Are Minimal Representations Still Representations?’ International Journal of Philosophical Studies, 16, No. 3 (2008): 351–369. Special issue on ‘Situated Cognition: Perspectives from Phenomenology and Science’, edited by M. Ratcliffe and S. Gallagher. —How the Body Shapes the Mind. Oxford: Oxford University Press, 2006. Gallagher, Shaun, and Zahavi, Dan. ‘Reply: A Phenomenology with Legs and Brains’. Abstracta special issue 2 (2008): 86–107. Garver, Newton. This Complicated Form of Life: Essays on Wittgenstein. LaSalle, IL: Open Court, 1994. Garver, Newton and Seung-Chong Lee. Derrida and Wittgenstein. Philadelphia: Temple University Press, 1995. Genesereth, M. R. and Nilsson, N. J. The Logical Foundations of Artificial Intelligence. Los Altos: Morgan Kauffman, 1987. Gibson, John. ‘Reading for Life’. In The Literary Wittgenstein, edited by John Gibson and Wolfgang Huemer, 109–124. London and New York: Routledge, 2004. Gier, Nicholas F. Wittgenstein and Phenomenology: A Comparative Study of the Later Wittgenstein, Husserl, Heidegger, and Merleau-Ponty. Albany: State University of New York Press, 1981. Gillot, Pascale. L’esprit. Figures classiques et contemporaines. Paris: CNRS Editions, 2007. Glendinning, Simon ed. Arguing with Derrida. Oxford: Blackwell, 2001. —The Edinburgh Encyclopedia of Continental Philosophy. Edinburgh: Edinburgh University Press, 1999. —The Idea of Continental Philosophy: A Philosophical Chronicle. Edinburgh: Edinburgh University Press, 2006.

242

Bibliography

—In the Name of Phenomenology. London: Routledge, 2007. Glock, Hans-Johann. ‘Schopenhauer and Wittgenstein: Representation as Language and Will’. In The Cambridge Companion to Schopenhauer, edited by Christopher Janaway, 422–458. Cambridge: Cambridge University Press, 1999. —What Is Analytic Philosophy? Cambridge: Cambridge University Press, 2008. Gödel, Kurt. Collected Works (5 volumes). Oxford: Oxford University Press, 1986–2003. Gödel’s Papers, box 8c, folder 117, item 040403, Firestone Library of Princeton University. —Box 20, from Gödel’s hand on a text by Wang, my emphasis. —Philosophical notebook, transcription from Gabelsberger by C. Dawson, V, 344, Institute for Advanced Study, Princeton University. Goldman, Alvin I. ‘A Priori Warrant and Naturalistic Epistemology’. In Pathways to Knowledge: Private and Public, edited by A. I. Goldman, 24–50. New York: Oxford University Press, 2002. Grodal, Torben. Embodied Visions. Oxford: Oxford University Press, 2009. Habermas, Jürgen. ‘From Kant to Hegel: On Robert Brandom’s Pragmatic Philosophy of Language’. Translated by Maeve Cooke. European Journal of Philosophy 8, No. 3 (2000): 322–355. —The Philosophical Discourse of Modernity. Cambridge, MA: MIT Press, 1987. Hacker, P. M. S., ‘Events and Objects in Space and Time’. In Events, edited by Achille Varzi and Roberto Casati. 428–447. Aldershot: Dartmouth, 1996. —‘Events, Ontology and Grammar’. In Events, edited by Achille Varzi and Roberto Casati, 79–88. Aldershot: Dartmouth, 1996. Haller, Rudolf. Questions on Wittgenstein. Lincoln: University of Nebraska Press, 1988. Hanna, Robert. Kant and the Foundations of Analytic Philosophy. Oxford: Clarendon Press, 2001. Hans, Beatrice. Michel Foucault’s Critical Project: Between the Transcendental and the Historical. Palo Alto: Stanford University Press, 2002. Harman, Gilbert. Change in View. Cambridge MA: MIT Press, 1986. —‘Knowledge, Inference, and Explanation’. American Philosophical Quarterly 5 (1968): 164–173. —‘The Inference to the Best Explanation’. Philosophical Review 74 (1965): 88–95. Hegel, G. W. F. Phenomenology of Spirit. Translated by A. V. Miller. Oxford: Oxford University Press, 1977. —Science of Logic. Translated by A. V. Miller. London: Allen and Unwin, 1969. Heidegger, Martin. Basic Writings, edited by D. Farrell Krell. London: Routledge, 1993. —The Basic Problems of Phenomenology. Bloomington: Indiana University Press, 1982. —Being and Time. Translated by J. Macquarrie and E. Robinson. Oxford: Basil Blackwell, 1927. —The Essence of Human Freedom: An Introduction to Philosophy. London: Continuum, 2002. —The Metaphysical Foundations of Logic. Bloomington: Indiana University Press, 1984. Hilbert, David. ‘On the Infinite’. Translated by Erna Putnam and Gerald J. Massey. In Philosophy of Mathematics: Selected Readings, edited by Paul Benacerraf and Hilary Putnam, 183–201. Cambridge: Cambridge University Press, 1983.

Bibliography

243

Hintikka, Jaakko. ‘The Emperor’s New Intuitions’. Journal of Philosophy 96 (1999): 127–147. Hollinger, Robert, ed. Hermeneutics and Praxis. Notre Dame: University of Notre Dame Press, 1985. Hookway, Christopher. ‘Modest Transcendental Arguments and Sceptical Doubts: A Reply to Stroud’. In Transcendental Arguments: Problems and Prospects, edited by R. Stern, 173–188. Oxford: Oxford University Press, 1999. Huemer, Michael. Skepticism and the Veil of Perception. Oxford: Rowman and Littlefield, 2001. Hutto, Daniel D. Wittgenstein and the End of Philosophy: Neither Theory nor Therapy. Hampshire: Palgrave Macmillan, 2006. Jackson, Frank. From Metaphysics to Ethics: A Defence of Conceptual Analysis. Oxford: Oxford University Press, 2000. Jacquette, Dale. ‘Haller on Wittgenstein and Kant’. In Austrian Philosophy Past and Present: Essays in Honor of Rudolf Haller, edited by Keith Lehrer and Johan Christian Marek, 29–44. Boston Studies in the Philosophy of Science. Dordrecht: Kluwer Press, 1997. —Wittgenstein’s Thought in Transition. West Lafayette: Purdue University Press, 1998. James, William. Essays in Radical Empiricism. London: Longmans, Green, and Co., 1912. —The Will to Believe and Other Essays in Popular Philosophy. New York: Dover Publications, 1956. Janik, Allan. ‘Schopenhauer and the Early Wittgenstein’. Philosophical Studies 15 (1966): 76–95. Janik, Alan S. and Stephen Toulmin. Wittgenstein’s Vienna. New York: Simon & Schuster, 1973. Kant, Immanuel. Critique of Practical Reason. Indianapolis: Bobbs-Merril, 1956. —Groundwork of the Metaphysic of Morals. New York: Harper and Row, 1964. Kelly, Sean. ‘Seeing Things in Merleau-Ponty’. In Cambridge Companion to Merleau-Ponty, edited by Taylor Carman and Mark Hansen, 74–110. Cambridge: Cambridge University Press, 2005. Kemp Smith, Norman. Commentary to Kant’s ‘Critique of Pure Reason’. Amherst NY: Prometheus Books, 1991. Kim, Jaegwon. ‘Events as Property Exemplifications’. In Events, edited by Achille Varzi and Roberto Casati, 117–135. Aldershot: Dartmouth, 1996. Kneale, William and Martha Kneale. The Development of Logic. Oxford: Oxford University Press, 1984. Körner, Stephan. ‘The Impossibility of Transcendental Deductions’. The Monist 51, No. 3 (July 1967): 317–331. —‘Transcendental Tendencies in Recent Philosophy’. The Journal of Philosophy 63 (1966): 551–561. Lautman, A. Les mathématiques, les idées et le réel physique. Paris: Vrin, 2006. Lawn, Chris. Wittgenstein and Gadamer: Towards a Post-analytic Philosophy of Language. London: Continuum, 2005. Levinas, Emmanuel. Totality and Infinity. Translated by A. Lingis. Pittsburgh: Duquesne University Press, 1969. Lewis, C. I. An Analysis of Knowledge and Valuation. LaSalle, IL: Open Court, 1946.

244

Bibliography

Lewis, David. Counterfactuals. Cambridge, MA: Harvard University Press, 1973. —‘Devil’s Bargains and the Real World’. Papers in Ethics and Social Philosophy, edited by David Lewis, 201–218. Cambridge: Cambridge University Press, 2000. —On the Plurality of Worlds. Oxford: Blackwell, 1986. Livingston, Paul. ‘Wittgenstein, Kant and the Critique of Totality’. Philosophy and Social Criticism 33, No. 6 (2007): 691–715. Longuenesse, Béatrice. Kant and the Capacity to Judge: Sensibility and Discursivity in the Transcendental Analytic of the Critique of Pure Reason. Translated by Charles T. Wolfe. Princeton: Princeton University Press, 1998. Lycan, William. Judgement and Justification. Cambridge: Cambridge University Press, 1988. —‘Moore against the New Skeptics’. Philosophical Studies 103 (2001): 35–53. MacIntyre, Alisdair. After Virtue. London: Duckworth, 1984. Mackie, John. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin, 1977. Malpas, Jeff. Donald Davidson and the Mirror of Meaning: Holism, Truth and Interpretation. Cambridge: Cambridge University Press, 1992. —‘Introduction’. In From Kant to Davidson: Philosophy and the Idea of the Transcendental, edited by J. Malpas. Cambridge: Cambridge University Press, 2005. —‘The Transcendental Circle’. Australasian Journal of Philosophy 75, No. 1 (1997): 1–20. Mares, Edwin. ‘A Paraconsistent Theory of Belief Revision’. Erkenntnis 56 (2002): 229–246. —Relevant Logic: A Philosophical Interpretation. Cambridge: Cambridge University Press, 2004. Martin-Löf, Per. Intuitionistic Type Theory. Naples: Bibliopolis, 1984. Mates, Benson. The Philosophy of Leibniz: Metaphysics and Language. Oxford: Oxford University Press, 1986. McCumber, John. Philosophy and Freedom: Derrida, Rorty, Habermas, Foucault. Bloomington: Indiana University Press, 2000. McDowell, John. ‘The Content of Perceptual Experience’. The Philosophical Quarterly 44, No. 175 (1994): 190–205. —Mind and World. Cambridge, MA: Harvard University Press, 1994. —‘Transcendental Empiricism’. Paper presented at the third Athens-Pittsburgh Symposium, Rethymnon, October 2000. Greek translation in Deuka-lion, vol. 21 (2003). Merleau-Ponty, Maurice. Phenomenology of Perception. Translated by C. Smith. London: Routledge, 1962. —The Structure of Behaviour. Pittsburgh: Duquesne University Press, 1983. Morris, David. ‘Continental Philosophy of Mind’. Columbia Companion to Twentieth Century Philosophies. Edited by C. Boundas, xx–xx. New York: Columbia University Press, 2007. Mulhall, Stephen. Inheritance and Originality. Oxford: Oxford University Press, 2001. —On Film. London: Routledge. Second expanded edition, 2008. —Wittgenstein’s Private Language. Oxford: Oxford University Press, 2007. Murray, Alan. ‘Philosophy and the Anteriority Complex’. Phenomenology and the Cognitive Sciences 1 (2002): 27–47. Nagel, Thomas. The View from Nowhere. Oxford: Oxford University Press, 1986.

Bibliography

245

Nelson, Kai. After the Demise of the Tradition: Rorty, Critical Theory, and the Fate of Philosophy. Boulder: Westview Press, 1991. Niznik, Josef and John T. Sanders, eds. Debating the State of Philosophy: Habermas, Rorty and Kolakowski. Westport: Praeger Publishers, 1996. Nozick, Robert. Philosophical Explanations. Oxford: Oxford University Press, 1981. O’Connor, D. J. ‘Causal Statements’. Philosophical Quarterly 6 (1956): 17–26. Oppy, Graham. Ontological Arguments and Belief in God. New York: Cambridge University Press, 1995. Overgaard, Søren. ‘The Ethical Residue of Language in Levinas and Early Wittgenstein’. Philosophy and Social Criticism 33, No. 2 (2007): 223–249. Perloff, Marjorie. Wittgenstein’s Ladder: Poetic Language and the Strangeness of the Ordinary. Chicago and London: University of Chicago Press, 1996. Philipse, Herman. Heidegger’s Philosophy of Being. Princeton: Princeton University Press, 1998. Pihlstrom, Sami. Naturalising the Transcendental. Amherst, NY: Prometheus Books, 2003. —‘Recent Reinterpretations of the Transcendental’. Inquiry 47 (2004): 289–314. Pinkard, Terry. Hegel’s Phenomenology: The Sociality of Reason. Cambridge: Cambridge University Press, 1994. Pippin, Robert B. Hegel’s Idealism. Cambridge: Cambridge University Press, 1989. Place, U. T. ‘Is Consciousness a Brain Process?’ British Journal of Psychology 47 (1956): 44–50. Plantinga, Carl. Moving Viewers: American Film and the Spectator’s Experience. Berkeley: University of California Press, 2009. Plato, Parmenides. Translated by A. Whitaker. Newburyport, MA: Focus Publishing, 1996. —Republic. Translated by Benjamin Jowett. Mineola, NY: Dover, 2000. Pradhan, S. ‘Minimalist Semantics: Davidson and Derrida on Meaning, Use and Convention’. Diacritics 1, No. 1 (1986): 66–77. Priest, Graham. Beyond the Limits of Thought. New York: Oxford University Press, 2003. —‘Motivations for Paraconsistency: The Slippery Slope from Classical Logic to Dialetheism’. In Frontiers of Paraconsistent Logic, edited by Dederik Batens, Chris Mortensen, Graham Priest and Jean-Paul van Bendigem, 223–232. Baldock: Research Studies Press, 2000. Putnam, Hilary. ‘A Half Century of Philosophy: Viewed from Within’. Daedalus 12 (1997): 175–208. —‘After Empiricism’. In Post-Analytic Philosophy, edited by John Rajchman and Cornel West, 21–30. New York: Columbia University Press, 1985. —Mind, Language and Reality. Philosophical Papers, vol. 2. Cambridge: Cambridge University Press, 1975. —Realism with a Human Face. Cambridge, MA: Harvard University Press, 1990. —The Threefold Cord. Mind, Body, and World. New York: Columbia University Press, 1999. Quine W. V. O. ‘Events and Reification’. In Events, edited by Achille Varzi and Roberto Casati, 107–116. Aldershot: Dartmouth, 1996. —From a Logical Point of View. New York: Harper and Row, 1961.

246

Bibliography

—Word and Object. Cambridge, MA: MIT Press, 1960. Quine, W. V. and Ullian, J., The Web of Belief. Second edition. New York: Random House, 1978. Rajchman, John and Cornel West, eds. Post-Analytic Philosophy. New York: Columbia University Press, 1985. Read, Rupert and Jerry Goodenough, eds. Film as Philosophy: Essays on Cinema after Wittgenstein and Cavell. Basingstoke/New York: Palgrave, 2005. Redding, Paul. Analytic Philosophy and the Return of Hegelian Thought. Cambridge: Cambridge University Press, 2007. —Continental Idealism: Leibniz to Nietzsche. London: Routledge, 2009. —‘The Independence and Dependence of Self-Consciousness: The Dialectic of Lord and Bondsman in Hegel’s Phenomenology of Spirit’. In The New Cambridge Companion to Hegel and Nineteenth Century Philosophy, edited by F. Beiser, 94–110. Cambridge: Cambridge University Press, 2008. Reynolds, Jack. ‘Chickening Out and Continental Philosophy’. International Journal of Philosophical Studies 17, No. 2 (2009): 255–272. —‘Dreyfus and Deleuze on l’habitude, Coping, and Trauma in Skill Acquisition’. International Journal of Philosophical Studies 14, No. 4 (2006): 563–583. —‘Reply to Glendinning’. International Journal of Philosophical Studies 17, No. 2 (2009): 281–287. Rietveld, E. ‘Context-switching and Responsiveness to Real Relevance’. In Heidegger and Cognitive Science, edited by J. Kiverstein and M. Wheeler, Basingstoke: Palgrave Macmillan, forthcoming. Rizzolatti, G. and C. Sinigaglia. ‘Mirror Neurons and Motor Intentionality’. Functional Neurology, 22, No. 4 (2007): 205–210. Rodowick, David. N. Gilles Deleuze’s Time Machine. Durham, NC: Duke University Press, 1997. Rogers, A. K. ‘Professor Royce and Monism’. Philosophical Review 12 (1903): 47–61. —‘Cultural Politics and the Question of the Existence of God’. In Radical Interpretation in Religion, edited by Nancy Frankenberry, 53–77. Cambridge: Cambridge University Press, 2002. —‘Introduction’. In Empiricism and the Philosophy of Mind by Wilfrid Sellars, 8–9. Cambridge, MA: Harvard University Press, 1997. —Philosophy and the Mirror of Nature. Princeton: Princeton University Press, 1979. —‘Transcendental Arguments, Self-Reference, and Pragmatism’. In Transcendental Arguments and Science, edited by Peter Bieri, Rolf-Peter Horstmann and Lorenz Krüger. Dordrecht: R. Reidel, 1979. Rorty, Richard and Eduardo Mendieta. Take Care of Freedom and Truth Will Take Care of Itself: Interviews with Richard Rorty. Palo Alto: Stanford University Press, 2006. Rosen, Stanley. The Limits of Analysis. New Haven and London: Yale University Press, 1980. Rueda, L. ‘“Ontology of Events Vs. Ontology of Facts: About the Current Fissures Between the Continental and Analytic Traditions’. Translated by M. Weigelt. Journal of the British Society for Phenomenology 37, No. 2 (May 2006): 120–137. Rumana, Richard. Richard Rorty: An Annotated Bibliography of Secondary Literature. Amsterdam: Rodopi, 2002. Russell, Bertrand. Human Knowledge: Its Scope and Limits, 446. London: Routledge, 2009.

Bibliography

247

—Logic and Knowledge: Essays 1901–1950. London: Allen and Unwin, 1956. —My Philosophical Development. New York: Simon & Schuster, 1959. —The Philosophy of Leibniz. London: George Allen and Unwin, 1958. —The Problems of Philosophy. Oxford: Oxford University Press, 1959. Ryle, Gilbert. Collected Papers. London: Hutchinson, 1971. —The Concept of Mind. London: Hutchinson, 1949. Saatkamp, Herman J. ed. Rorty and Pragmatism: The Philosopher Responds to His Critics. Nashville: Vanderbilt University Press, 1995. Sacks, Mark. ‘Naturalism and the Transcendental Turn’. Ratio 19 (2006): 92–106. —‘The Nature of Transcendental Arguments’. International Journal of Philosophical Studies 13, No. 4 (2005): 439–460. Searle, John. Intentionality. An Essay in the Philosophy of Mind. Cambridge: Cambridge University Press, 1983. —Philosophy in a New Century: Selected Essays. Cambridge: Cambridge University Press, 2008. —‘Minds, Brains, and Programs’, Behavioral and Brain Sciences, No. 3 (1980), 417–424. —Mind. A Brief Introduction. Oxford: Oxford University Press, 2004. —The Mystery of Consciousness. New York: New York Review of Books, 1997. —The Rediscovery of the Mind. Cambridge, MA: MIT Press, 1992. Sellars, Wilfrid. Empiricism and the Philosophy of Mind. Cambridge, MA: Harvard University Press, 1997. —Science, Perception and Reality. London: Routledge and Kegan Paul, 1963. Shanahan, M. ‘The Frame Problem’. The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), edited by E. N. Zalta. http://plato.stanford.edu/archives/fall2008/ entries/frame-problem Sheppard, Darren. ‘The Reading Affair: On Why Philosophy Is Not “Philosophical” Through and Through’. In Arguing with Derrida, edited by Simon Glendinning, 121–134. Oxford: Blackwell, 2001. Sidgwick, Henry. ‘A Criticism of the Critical Philosophy’. Mind 8 (1883): 69–91. Sklar, Lawrence. ‘Methodological Conservatism’. Philosophical Review 84 (1975): 374–400. Smith, Nicholas H. Charles Taylor: Meaning, Morals and Modernity. Cambridge: Polity, 2002. Sobel, J., Logic and Theism. New York: Cambridge University Press, 2004. Sorensen, Roy. Thought Experiments. Oxford: Oxford University Press, 1992. Staten, Henry. Wittgenstein and Derrida. Lincoln: University of Nebraska Press, 1986. Stenius, Erik. Wittgenstein’s “Tractatus”: A Critical Exposition of Its Main Lines of Thought. Ithaca: Cornell University Press, 1960. Stern, Robert. Transcendental Arguments and Scepticism. Oxford: Oxford University Press, 2000. Stiegler, Bernard. ‘Technics, Media, Teleology’. Theory, Culture and Society 24, No. 7–8 (2007): 334–341. Strawson, Peter. The Bounds of Sense. London: Methuen, 1966. —‘Imagination and Perception’. Kant on Pure Reason, edited by R. C .S. Walker, Oxford: Oxford University Press, 1982. —Individuals. London: Methuen, 1959. —Skepticism and Naturalism: Some Varieties. New York: Columbia University Press, 1985.

248

Bibliography

Stroud, Barry. ‘Transcendental Arguments’. The Journal of Philosophy 65 (1968): 241–256. Symons, John. ‘Intuition and Philosophical Methodology’. Axiomathes 18 (2008): 67–89. Talbott, W. J. ‘Two Principles of Bayesian Epistemology’. Philosophical Studies 62 (1991): 135–150. Tarski, Alfred. Logic, Semantics, Metamathematics. Indianapolis: Hackett, 1984. Tartaglia, James. Routledge Philosophy Guidebook to Philosophy and the Mirror of Nature. London: Routledge, 2007. Taylor, Charles. ‘Action as Expression’. In Intention and Intentionality, edited by Cora Diamond and Jenny Teichmann, 73–90. Ithaca: Cornell University Press, 1979. —‘Foundationalism and the Inner-Outer Distinction’. In Reading McDowell: On Mind and World, edited by Nicholas H. Smith. London: Routledge, 2002. —‘Genesis’. Review of The Structure of Behaviour by Maurice Merleau-Ponty. New Statesman 70, 3 (September 1965): 326–327. —Hegel. Cambridge: Cambridge University Press, 1975. —Human Agency and Language. Cambridge: Cambridge University Press, 1985. —‘The Importance of Herder’. In Isaiah Berlin: A Celebration, edited by Edna Margalit and Avishai Margalit. Chicago: University of Chicago Press, 1991. —Philosophical Arguments. Cambridge, MA: Harvard University Press, 1995. —Sources of the Self. Cambridge: Cambridge University Press, 1989. Taylor, Charles and Michael Kullman. ‘The Pre-Objective World’. The Review of Metaphysics, No. 12 (1958): 103–132. Taylor Mill, Harriet. The Enfranchisement of Women. London: Virago Press, 1983. Thalberg, Irving. ‘A World without Events’. In Essays on Davidson: Actions and Events, edited by Bruce Vermazen and Merrill B Hintikka, 137–155. Oxford: Clarendon, 1985. Thompson, Manley. ‘Singular Terms and Intuitions in Kant’s Epistemology’. Review of Metaphysics 26 (1972–1973): 314–343. Thomson, J. ‘A Defence of Abortion’. Philosophy and Public Affairs 1 (1971): 47–66. Trakakis, Nick. The End of Philosophy. London: Continuum, 2008. Vahid, Hamil. ‘Varieties of Epistemic Conservatism’. Synthese 141 (2004): 97–122. Van Inwagen, P. The Problem of Evil, 37–55. Oxford: Oxford University Press, 2006. Varzi, Achille and Roberto Casati, eds. Events. Aldershot: Dartmouth, 1996. —50 Years of Events: An Annotated Bibliography 1947–1997. Bowling Green, OH: Bowling Green State University, 1997. Wang, Hao. A Logical Journey. Cambridge, MA: MIT Press, 1996. Wartenberg, Thomas E. 2004. ‘Philosophy of Film’. Stanford Encyclopedia of Philosophy, November 2008. http://plato.stanford.edu/entries/film/. Weinberg, Jonathan. ‘How to Challenge Intuitions Empirically Without Risking Skepticism’. Midwest Studies in Philosophy 31 (2007): 318–343. Weiner, D. A. Genius and Talent: Schopenhauer’s Influence on Wittgenstein’s Early Philosophy. Rutherford: Farleigh Dickinson University Press, 1992. Westphal, Kenneth R. ‘Contemporary Epistemology: Kant, Hegel, McDowell’. European Journal of Philosophy 14, No. 2 (2006): 274–301. Wheeler, Michael. ‘Cognition in Context: Phenomenology, Situated Robotics and the Frame Problem’. International Journal of Philosophical Studies 16, No. 3 (2008):

Bibliography

249

323–349. Special issue on ‘Situated Cognition: Perspectives from Phenomenology and Science’, edited by. M. Ratcliffe and S. Gallagher. —‘The Problem of Representation’. In The Handbook of Phenomenology and Cognitive Science, edited by S. Gallagher and D. Schmicking, chapter 18. Berlin: Springer, 2009, chapter. —Reconstructing the Cognitive World: The Next Step. Cambridge MA: MIT Press, 2005. Williams, Bernard. ‘Contemporary Philosophy: A Second Look’. In The Blackwell Companion to Philosophy, edited by N. Bunnin and E. P. Tsui-James, 23–34. Oxford: Blackwell, 1996. —Ethics and the Limits of Philosophy. Cambridge: CUP, 1985. —Gilles Deleuze’s Difference and Repetition: A Critical Introduction and Guide. Edinburgh: Edinburgh University Press, 2003. —Gilles Deleuze’s Logic of Sense: A Critical Introduction and Guide. Edinburgh: Edinburgh University Press, 2008. Williamson, Timothy. ‘Philosophical “Intuitions” and Scepticism about Judgement’. Dialectica 58 (2004): 109–153. Wittgenstein, Ludwig. Culture and Value. Translated by Peter Winch. Chicago: University of Chicago Press, 1984. —Letters from Ludwig Wittgenstein with a Memoir by Paul Englemann. Edited by Brian McGuinness. Oxford: Basil Blackwell, 1967. —On Certainty. London: Blackwell, 1969. —Philosophical Investigations. Oxford: Blackwell, 1953. —Recollections of Wittgenstein, edited by Rush Rhees. Oxford: Oxford University Press, 1984. —Tractatus Logico-Philosophicus. Translated by C. K. Ogden. London: Routledge and Kegan Paul, 1922. —Wittgenstein’s Lectures on the Foundations of Mathematics, Cambridge 1939. From the notes of R. G. Bosanquet, Norman Malcolm, Rush Rhees and Yorick Smythies, edited by Cora Diamond. Chicago: University of Chicago Press, 1976. Wrathall, Mark, ‘The Conditions of Truth in Heidegger and Davidson’, The Monist, No. 82 (1999), 304–323. Wright, Crispin. ‘Human Nature?’. In Reading McDowell: On Mind and World, edited by Nicholas Smith. London: Routledge, 2002. Yourgrau, Palle. Gödel Meets Einstein: Time Travel in the Gödel Universe. Chicago: Open Court, 1999.

This page intentionally left blank

Index

Adams, Ernest 61 Adler, Jonathan 94 Adorno, Theodor W. 164, 165, 177 aesthetics 4, 5, 159, 173, 174, 177, 180 Alexander, Samuel 160 Althusser, Louis 185, 186n6, 217, 220 Ameriks, Karl 30, 32 anti-realism 42, 47, 56–7 Apel, Karl-Otto 37, 50n38 argument a priori 31, 36, 40, 42, imagination 76, 77–8, 80 narrow 5, 73, 74–80 non-argumento-centric 80 ontological 30, 34, 222 and persuasion 74, 182 transcendental see transcendental argument Armstrong, David M. 162 art 4, 23, 146, 152–4, 163, 177, 179, 180, 181, 182–6 artificial intelligence 108, 125–7, 130, 134, 137, 140n2 Augustine, Saint 11, 74 Austin, J. L. 31–2, 49n19, 49n21, 81 Avramides, Anita 47 Bachelard, Gaston 216, 217, 220–2, 225, 232 Badiou, Alain 4, 214n7 Bataille, Georges 158 Bealer, George 90, 100, 103n24 Beckett, Samuel 184 Bennington, Geoffrey 71, 72–4, 82n11, 83n17, 83n18 Bergson, Henri 46 Berlin, Isaiah 145 Bernays, Paul 216 bivalence 56 body see embodiment Boolos, George 100 Bordwell, David 173–4, 175, 180 Bostrom, Nick 86 Bradley, F. H. 160

Brandom, Robert 4, 5, 7, 19–20n8, 21n27, 40, 152, 192, 195 expressivism 145–8 on Hegel 154, 191–2, 194, 195, 197–9, 201n30 on Kant 191, 194 Brunschvig, Léon 37, 217–20, 221 Canguilhem, Georges 117 Carnap, Rudolf 1, 8, 31, 93, 158, 159, 195, 201n28 Carroll, Noël 5, 173, 174–5, 177, 183, 186n6 Cassirer, Ernst 37 Cavaillès, Jean 216–18, 220–2, 224–6, 232–3 Cavell, Stanley 4 and film 173, 179 and Wittgenstein 165, 166–8 Chisholm, Roderick 86–7, 89, 93, 215n21 Churchland, Paul 89 cognitive science 5, 41, 47, 107, 108, 115, 117, 125, 130–1 context-dependent relevance sensitivity (CDRS) 126–7, 131, 134, 140, 141n3, 179, 180, 182 frame problem 5, 126, 127, 128–30, 131, 133–4, 135, 137, 140 Cohen, L. Jonathan 83n19, 95 Collingwood, R. G. common sense 28, 30, 36, 86, 88–9, 91–2, 99, 100–1, 102n9, 115, 119, 161, 167, 194 Conant, James 18, 24n58, 171n6 conceptual schemes 11–12, 50n30 conservatism analytic dialogue common assumption 5, 97–9, 101 inferential links 5, 96–101 non-deference 98–101 non-hierarchical 96, 99–101 epistemic comparative 87, 93 generator 86 perseverance 86, 89

252 conservatism (Cont’d) institutional 85, 99, 101 reasons for 86, 88, 93–4, 99 technique appeals to intuitions 88–9, 91, 94 common sense 88–9, 91, 94 local coherence-building 88–9 regimentation practices 89 contextualism 34, 200, 202n48 coping background 132, 135–6, 138–40 skillful 39, 128, 132–40 Critchley, Simon 184 Currie, Gregory 174, 186n5 Danto, Arthur 181, 182–3 Davidson, Donald 4, 5, 6, 7, 9, 10, 11–12, 14–16, 21n25, 22n42, 22n43, 28, 122n22, 193, 195 anomalous monism 112 and events 205, 211–13, 214n5, 215n21, 2115n23 transcendental arguments 34, 48n6 De Unamuno, Miguel 158, Deleuze, Gilles 4, 155n5, 217 and events 205, 207–8, 210–11, 211–13 and film 179, 181, 186n6 and transcendental arguments 37, 42, 43–5, 46 virtual Platonism 216 Dennett, Daniel 35, 47, 90, 114, 118, 122n16 Derrida, Jacques 3, 15, 17, 21n25, 28, 42, 158, 165, 176, 177, 186n6, 187n15 argument 37, 45, 73–4 and the transcendental 17, 28, 42, 45, 51n57 Descartes, Rene 5, 126, 127 neuropsychology 109, 111, 119–20 substance dualism 107, 109–11, 126 Diamond, Cora 18, 24n58, 75–7, 79, 171n6 Dilthey, Wilhelm 163 Dostoyevsky 159, 171n5 Dreyfus, Hubert 5, 39–40, 47, 127, 128–9, 131–2, 133–4, 135–7 dualism 11, 12–14 mind-body see mind scheme-content 8, 10, 11, 16, 20n12 subjective-objective 8 Dummett, Michael 5 and logic 53–7 and meaning 55–7, 64, 68n5, 70n41

Index embodiment 150–1, 153 and intentionality 38 and perception 38 and transcendental arguments 28, 38 empiricism 8, 14, 28, 30–1, 35, 37, 39, 45–6, 51n57, 193 Engel, Pascal 8, 28 epistemology see knowledge essentialism and the divide 1–2 events see metaphysics expressivism and the Enlightenment 145–7, 149, 154 meta-ethics 145, 149 modes of expression 152 pragmatism 145–6, 149, 152, 154 and Romanticism 146, 152, 155n4 family resemblance 10, 11, 38, 96, 166 Fichte, J. G. 37, 158, 164, 168 film analytic-cognitivist turn 5, 173, 174, 179 disenfranchisement of 5, 173, 182–4, 185 film-philosophy (F-P) 174, 180, 181, 185, 186 ‘Grand Theory’ 173 criticisms of 174, 176, 178, 180 philosophy of (PF) 180–1, 185 and psychoanalysis 174, 175, 182 Fodor, Jerry 90, 141n2 Foley, Richard 86 Foucault, Michel 4, 38, 50n43, 158, 177, 186n6, 217, 220 Frege, Gottlob 1, 8, 59, 61, 70, 158, 159, 160, 162, 168, 194, 195, 196 Freud, Sigmund 121, 175, 185 Fumerton, Richard 91 Gadamer, Hans-Georg 15, 158 Gallagher, Shaun 129 genealogy 42, 48, 110, 112, 187n15 genetic fallacy 42 Gier, Nicholas F. 165 given, myth of the 13, 192–3, 200 Glendinning, Simon 2, 4, 5, 177, 178 Glock, Hans-Johann 4, 10, 95, 101, 172n6 Gödel, Kurt 216, 217, 226–32, 233 Goodman, Nelson 8 Habermas, Jürgen 37, 95, 145, 146 Hanna, Robert 4 Hare, R. M. 177–8 Harman, Gilbert 86, 87, 94

Index Hegel, G. W. F. 1, 3, 5–6, 17, 37, 118, 158, 159, 176, 183, 187n15, 191–2, 199–200 and expressivism 145, 147, 150, 153–4, 202n35 Heidegger, Martin 3, 4, 15, 17, 19, 21n25, 37, 46, 158, 168, 169, 177–8 being-in-the-world 38–9, 135–6, 139 and cognitive science 5, 41, 140 and ground 40, 51n52, 58–9, 60, 61, 63 present-to-hand (un-ready-to-hand) 39, 129, 132, 133, 135, 136, 138–9, 142n24 ready-to-hand 39–41, 132, 135, 138 Henry, Michel 37 Herder, Johann Gottfried 147, 150–1 hermeneutics 37, 42 Hilbert, David 170, 218, 224–5, 226 Hölderlin, Friedrich 147 Horkheimer, Max 85 Huemer, Michael 91 humanities 4, 47, 174, 177, 182 Hume, David 32, 35, 43, 118 Husserl, Edmund 37, 41–2, 46, 141n13, 158, 176, 187n15, 233

253

instrumentalist 11–12, 149 intersubjective 12, 14 Körner, Stephen 28, 34–6, 42, 44, 45 Kristeva, Julia 176, 186n6, 187n15

James, William 48, 113–14, 169

Lacan, Jacques 175, 185, 186n6 language 11–13, 16, 18, 31–2, 54, 73–4, 108, 114, 118, 158–9, 161, 163–5, 166, 194, 210–3 expressivism 145, 147, 149, 150–2 linguistic analysis 55 and normativity 151 representationalism 12, 145 Lautman, Albert 218, 220, 222–6, 230, 233–4 Platonism 216–17, 229, 231, 232 Leibniz, Gottfried 54, 55, 57, 59–60, 69n19 Levinas, Emmanuel 18, 155n5, 159 Lewis, C. I. 61 Lewis, David 36, 60, 66, 89, 91–2, 93, 94, 98, 103n20 literature 3, 4, 72, 79–80, 163, 177, 200, 202n46 logic deductive 38, 40, 42, 54, 55 and freedom 58 interpretations of 37, 54, 63, 160–1 intuitionist 54, 55–6, 63–4 and metaphysics 5, 24n57, 60–2, 162 modal (possible worlds) 35, 54, 60–2, 65, 90 pragmatic consistency 47, 62–3, 170 and reflective equilibrium 5, 53, 62–3, 67, 68, 90, 131–2 Lycan, William 88, 91, 94, 102n9 Lyotard, Jean-François 45–6, 158, 164, 214n7

Kant, Immanuel 1, 8, 13, 17, 18, 27, 28, 29, 33, 37, 41, 49n12, 50n33, 58, 69n16, 77, 93, 118, 145, 159, 171n3, 183, 184, 191, 192, 193–4, 195, 199, 201n33, 202n48, 233 Copernican revolution 37, 42 Kantianism 8, 37, 192 Kelly, Sean 39 Kierkegaard, Søren 159, 169, 171n5 knowledge 6, 8, 11–14, 17, 39, 47, 88, 96–7, 116, 128–9, 135, 146–9, 154, 158, 166–7, 177, 181–3, 192, 206, 212, 218, 229, 231–2 coherentist 13, 31 foundationalist 12

Mackie, J. L. 89 Marion, Jean-Luc 37 Martin-Löf, Per 64 Marx, Karl 158, 173, 185 McDowell, John 4, 5, 7, 8, 9, 10, 11, 13–14, 16–19, 21n39, 22n43, 23n51, 28, 46 on Hegel 191–2, 193–4, 199–200 on Kant 8, 13, 193 theory of perception 130–2, 197 meaning justificationism picture theory of truth conditional theory of use theory of

idealism 33, 47, 107, 111, 154, 191, 192, 194, 200, 202n48, 219, 232 British 30 German 6 transcendental 32 imagination 35, 76, 77–80, 153, 169, 220, 232 inference to the best explanation 29, 39, 46, 47, 82n11, 89, 93 intentionality 37, 38–40, 46, 108, 114, 116, 118, 120, 121, 135, 137 Irigaray, Luce 37

254

Index

Merleau-Ponty, Maurice 39–41, 81, 158 and cognitive science 46–7, 128–9, 131–2, 136–8, 140 and expressivism 147, 150, 155n4 metaphysics 4–5, 32–3, 42, 45–62, 67–9 of events relationship with science 117, 119, 212–13 selection of formal conditions 206–8, 210, 211, 213 individuals 117, 206 types 204, 206–8, 211 space-time location 203, 209, 211 method 9, 11, 14, 27, 30, 31, 37, 47, 55–6, 58, 73, 81, 84n34, 85, 88–94, 100–1, 114, 127, 158, 160–1, 168, 204, 207, 217–18, 224, 229 mind computational theory of 107, 108, 111, 121 consciousness 107–11, 114–18, 120–4 externalism 114, 119 functionalism 89, 96, 108, 111, 119 identity theory of 59, 109, 111, 113, 121 internalism 108, 115, 118, 120 materialism 107–8, 111, 114–15 mind-body dualism 107, 109–10, 113 modality impossibility 34–5, 36, 177 necessity 28–9, 34–7, 40, 61, 65, 204, 208, 218, 220, 222 possibility 13, 27, 29, 39, 43, 45, 139 modernity 2, 27, 80, 85, 176, 177, 185 Moore, G. E. 1, 30, 88, 94–5, 102n9, 158, 160–1, 191 Mulhall, Stephen 4, 82n3, 179 Murdoch, Iris 77 Nagel, Thomas 112, 114, 122n21 Nancy, Jean-Luc 37 naturalism 13, 16, 27, 46, 107, 109, 111, 112, 119, 130, 146 Nietzsche, Friedrich 17, 44, 74, 158, 183 ordinary language 31, 32, 85, 90, 93, 133, 155n15, 205, 206 perception and the body 38–41, 46, 113, 119 direct realism 113, 117, 118 representationalism 118–19, 131, 136–7

phenomenology 4, 27, 37–9, 41–2, 46, 69, 81, 84n42, 84n43, 125, 128–30, 131, 136, 158, 165, 169, 177, 187n15, 198 naturalism 27, 46, 130 philosophy postanalytic 9–13, 28 specialization in 6, 85 style 8–9, 73–4, 80, 113, 178, 180, 200 as therapy 93, 162, 169 Pinkard, Terry 192 Pippin, Robert 192 Plato 13, 69n25, 168, 182, 185 and art 182–3 mathematical Platonism see metaphysics politics 4, 78, 177 positivism 163 and conservatism 85, 93 post-structuralism 42, 43, 174 pragmatism 8, 13, 20n12, 28, 145, 146, 149, 152, 154, 166, 179, 201n30 Priest, Graham 3, 66, 67 propositional content 38 Putnam, Hilary 3, 7, 8, 20n12, 28, 34, 107, 108, 113–15, 118–19 Quine, W. V. 8, 31, 35, 60–1, 65, 87, 90, 91–3, 100, 158, 194, 195, 197 Rancière, Jacques 4 reflective equilibrium 5, 39, 41, 53, 62–3, 67, 68, 89, 90, 93 reliablism 34, 103n25 Rorty, Richard 5, 7–9, 10, 11–14, 17, 19, 20n12, 22n43, 95, 154, 191, 195 and Wittgenstein 165, 168–9, 170 Rosen, Gideon 61 Rosen, Stanley 164–5 Russell, Bertrand 1, 8, 30, 34, 36, 49n16, 54, 65, 69n19, 91, 94, 158, 160, 162, 163, 168, 172n6, 191, 192, 194, 195–7, 200n12, 201n33, 202n40 Ryle, Gilbert 81, 93, 112, 114, 115, 158, 177–8 Sacks, Mark 28, 38, 40 Sartre, Jean-Paul 37, 158, 176, 187n15 scepticism 13, 88, 100, 176 Schelling, F. W. J. 37, 158 Schleiermacher 147, 163 Schlick, Morris 1 Schopenhauer, Arthur 158, 159, 162, 171n4

Index science 2, 7, 12, 37, 41, 46, 47, 73, 80, 81, 85, 88, 93, 100, 109, 110, 119, 126, 130, 140, 146, 149, 152, 153, 158, 162, 163–4, 168, 177, 178, 179, 186, 212, 213, 215, 217–18, 219, 220, 222, 227, 232, 233 Searle, John 5 biological naturalism 107, 109, 111–12, 119 intentionality 108, 113–14, 115, 116, 118, 120–1 Sellars, Wilfrid 8, 20n13, 93, 192–4, 195–6, 200 situated thought 5, 28, 38 Sklar, Lawrence 87, 92 Smart, J. J. C. 109 Sorensen, Roy 34 Stern, Robert 28, 45 Strawson, Peter 27, 28, 32–4, 35, 77, 103n24 Stroud, Barry 28, 33–5, 41, 42, 45–7 style see philosophy Sylvan, Richard 66 synthetic a priori 29, 30–1, 42, 49n16 Tarski, Alfred 64 Taylor, Charles 4, 5, 47, 191–2 and expressivism 145–7, 150–2, 153–4 theology 4 Thompson, Judith 90 time consciousness of 41 historicism 217 synthesis 43–4 and transcendental arguments 28, 38, 41–4

255

transcendental arguments and the body 37–8, 46 and idealism (anti-realism) 30, 33, 47 modality in 30, 35–6 modest 33, 38, 42, 47 and scepticism 28, 48 universals 31–2, 60, 197 Van Inwagen, Peter 36 Vattimo, Gianni 37 verificationism 33 vitalism 46 Warnock, Geoffrey 177 Wheeler, Samuel 4 Williams, Bernard 3, 72–3, 76, 82n5 Williamson, Timothy 94 Wittgenstein, Ludwig 8, 18–19, 31, 55, 71, 77, 81, 82n3, 84n40, 84n41, 113–14, 145, 147, 192, 194, 195 as an analytic philosopher 1, 9, 10, 11, 93, 157–8 as bringing analytic and continental philosophy together 24n57 and film theory 179 philosophical influences 159 as transcending the two traditions 5, 7, 12–13, 14, 17, 159, 166 Wright, Crispin 17, 23n51, 49n7 Zeno of Elea 54