The Burden of Proof upon Metaphysical Methods 3031277767, 9783031277764

Table of contents :
The Burden of Proof upon Metaphysical Methods
Acknowledgements
Contents
List of Tables
1: Dialogue and Persuasion
1.1 The Problem of Methodological Conservatism
1.2 Commitments in Attitude and Dialogue
1.3 An Analysis of Philosophical Dialogues
Types of Dialogue
Philosophers’ Self-Conceptions
Methodology
Example Analysis: Willoughby (2012)
Results
Attitudinal Goals
1.4 Looking Ahead
References
2: The Burden of Proof
2.1 Egalitarianism versus Foundationalism
2.2 Allocating the Burden of Proof
2.3 The Prudence of Dialogical Egalitarianism in Philosophy
References
3: Evidence, Inference, and Empiricism
3.1 Meeting the Burden of Proof
3.2 Evidence
3.3 Implication Barriers
3.4 Addressing Counter-Examples
3.5 Quantitative and Qualitative Barriers
3.6 Empiricism
References
4: Philosophical Methods between Content and the World
4.1 Consequences of Dialogical Empiricism
4.2 The Content/World Separation
4.3 Hypotheses in Philosophy
Psycho-Linguistic and Non-Psycho-Linguistic Subject Matter
Concepts
Cognition
Necessary Truth
4.4 Evidence in Philosophy
Intuitions
Common Sense
Ordinary Language Usage
Theoretical Virtues
Method of Cases
Reflective Equilibrium
Conceptual Analysis and the Canberra Plan
Experimental Philosophy
4.5 Against the Content/World Separation
The Easy Approach
Content Externalism
Agential Realism
Calibration
Mentalism
References
5: Metaphysical Hypotheses
5.1 Substantive Consequences
5.2 Normativity
5.3 Necessity
5.4 Epistemology
5.5 Ontology
5.6 Quietism
References
6: Escaping Dialogical Empiricism
6.1 Looking Behind
6.2 A Way Around?
6.3 A Way Out?
References
Appendix A Reputational Journal Ranking
Appendix B Population Size
Appendix C Example Analyses
Appendix D Analysis Results
Index

Citation preview

The Burden of Proof upon Metaphysical Methods Conny Rhode

The Burden of Proof upon Metaphysical Methods

Conny Rhode

The Burden of Proof upon Metaphysical Methods

Conny Rhode Stirling, UK

ISBN 978-3-031-27776-4    ISBN 978-3-031-27777-1 (eBook) https://doi.org/10.1007/978-3-031-27777-1 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

A wise man, therefore, proportions his belief to the evidence. —David Hume

To anyone looking in amazement at metaphysical debates1

 Cf. Stich (1990: 3).

1

The Burden of Proof upon Metaphysical Methods

Who carries the burden of proof in analytic philosophical debates, and how can this burden be satisfied? As it turns out, the answer to this joint question yields a fundamental challenge to the very conduct of metaphysics in analytic philosophy. Empirical research presented in this book indicates that the vastly predominant goal pursued in analytic philosophical dialogues lies not in discovering truths or generating knowledge, but merely in prevailing over one’s opponents. Given this goal, the book examines how most effectively to allocate and discharge the burden of proof. It focuses on premises that must prudently be avoided because a burden of proof on them could never be satisfied, and in particular discusses unsupportable bridge premises across inference barriers, like Hume’s barrier between ‘is’ and ‘ought’, or the barrier between the content of our talk or thought, and the world beyond such content. Employing this content/world barrier for a critical assessment of mainstream analytic philosophical methods, this book argues that we must prudently avoid invoking intuitions or other content of thought or talk in support of claims about the world beyond content, that is, metaphysically significant claims. Yet as content-located evidence is practically indispensable to metaphysical debates throughout analytic philosophy, from ethics to the philosophy of mathematics, this book reaches the ix

x 

The Burden of Proof upon Metaphysical Methods

startling conclusion that all such metaphysical debates must, prudently, be terminated. Conny Rhode’s research has focused on philosophical methodology and argumentation theory, with forays into the philosophy of science, post-Kantian philosophy, and political and moral philosophy, often employing a feminist perspective. In light of the conclusion derived in this book, Conny has left academic philosophy and now insists on appropriate evidence in accountancy instead.

Acknowledgements

The present work would almost certainly never have been undertaken without both my sheer luck of living in societies where very many aspects of my identity are privileged, and my sheer luck of being born in a society that is itself globally privileged. Without the latter, I would likely not have had the benefit of high levels of safety and stability, nor of a well-­ funded healthcare system that was able to save my life, or of an education and student funding system that made it possible for me to pursue higher education abroad. And without my various privileges in the societies I have called home, I would likely have had markedly less opportunity for and perhaps interest in devoting over 10% of my expected lifespan to something as comparatively trivial and practically irrelevant as the study of analytic philosophy (or let alone of its methodology). It must therefore be acknowledged that the present work would almost certainly not have been undertaken without the immense luck of inter- and intra-societal privilege that I have had in my life—though the manifold violences of oppression accompanying these privileges could not possibly be cause for gratitude. This work would also almost certainly never have been undertaken without the inspiration, support, and encouragement that I received from several people before even embarking upon the project presented here. It was my mother, Sabine Rhode, who taught me the joy of learning, and it is rather unlikely that I would ever have gone to university had xi

xii Acknowledgements

it not been for the way in which she had fundamentally shaped my interests. My desire to move to Scotland for university was ultimately due to my reading the now famous literary works of Joanne Rowling (1997ff). And I would never have studied philosophy without much encouragement to try it from Ruby Chan, then an ardent admirer of Plato. Yet I eventually grew deeply frustrated with analytic philosophy’s significant and systemic methodological deficiencies (arguably inherited, ironically, from Plato), and I would have left philosophy even in spite of the fact that these profound reservations were thankfully echoed by some other philosophers, particularly Robert Cummins (1998), David Papineau (2006), and James Ladyman et al. (2007) (and similarly, earlier, Jürgen Hartmann 1997). Over many discussions and many more cups of coffee, however, Sonja Erikainen and I instead developed our shared metaphilosophical concerns into an explicit methodological critique, distilling the status-quo-­ favouring allocation of the burden of proof upon philosophical methods as the apparent root of analytic philosophy’s copious methodological deficiencies. Soon after, this critique acquired a particular shape and trajectory when inference barriers became equally central to this project, inspired by the outstanding collection Hume on Is and Ought (Pigden 2010b). Between this inception and the present realisation of these two ideas, many more people have contributed to the development of the surrounding considerations. Most significant among them were my doctoral supervisors, Keith Allen and Mary Leng, whose unwavering constructive criticism both saved this work from many an error or overly concise explanation and gave rise to numerous considerations in these pages. A variety of solutions throughout this book further owe their quality, and often their very existence, to the brilliant thinking and incisive critiques contributed by Emmanouela Kritikaki. The project also benefited from very valuable commentary and advice by Michael Beaney, Alexander Bird, Jessica Brown, Ruby Chan, James Clarke, Catarina Dutilh Novaes, Sonja Erikainen, Brendan George, Christopher Jay, Finn Spicer, Alessandra Tanesini, Giulia Terzian, Sara Uckelman, Douglas Walton, and Dagmar Wilhelm. Important improvements to the project equally

 Acknowledgements 

xiii

arose from comments by referees for Argumentation,1 the Australasian Journal of Philosophy, and Palgrave Macmillan, and by audience members at the Universities of Bochum, Bristol, Dublin, Glasgow, Granada, Leeds, Sussex, and York. I am immensely grateful to each and every one of the people just mentioned for their genuinely invaluable inspiration, advice, feedback, encouragement, and support—and most particularly so to Emmanouela Kritikaki and Sonja Erikainen, and to Keith Allen and Mary Leng. Emmanouela, Keith, Mary, and Sonja each played a crucial role in the development of this project, and I cannot thank them enough for the innumerable ways in which they have shaped the present considerations.2 Yet I owe my deepest gratitude to my family—including my mother, Sabine Rhode, my father, Jörg Rhode, and my sister, Justeen Rhode, and including my very dear friends, Ruby Chan, Sonja Erikainen, Emmanouela Kritikaki, Nicola McCallum, Lubna Nowak, and Naomi Phillips.3 I am immeasurably rich to have each one of them as part of my family. My love is theirs.

 Chapters 1 and 2 of this book contain revised material previously published in Argumentation (Rhode 2017), which is reused here without permission (since none is required under its open access license; http://creativecommons.org/licenses/by/4.0/). I gratefully acknowledge the support of the University of York, who paid Springer Nature’s remarkable open access fee. 2  Authorship note: In the absence of generally accepted criteria for authorship in academic philosophy, corresponding criteria proposed by David Resnik and Zubin Master (2011) for conceptual work in bioethics have been employed instead. According to Resnik and Master, authorship credit requires substantial contribution to at least two of the following tasks (2011: 17): ‘(1) identifying a topic, problem, or issue to study; (2) reviewing and interpreting the relevant literature; (3) formulating, analyzing, and evaluating arguments that support one or more theses; (4) responding to objections and counterarguments; and (5) drafting the manuscript and approving the final version.’ Keith Allen and Mary Leng have each contributed substantially to (4) and partially to (2) and (3). Emmanouela Kritikaki has contributed substantially to (3) and partially to (4). Sonja Erikainen has contributed substantially to (1) and (3) and partially to (2) and (4). My own contribution has been substantial to (1)-(4) and exclusive to (5). Thus, both Sonja Erikainen and I qualify for authorship credit. Sonja, however, maintains that their contribution to (3) was partial rather than substantial, and they have therefore declined co-authorship. 3  In the acknowledgements to my 2017 article in Argumentation, Naomi’s surname missed one ‘l’. This is entirely my fault, and I sincerely apologise for this error! 1

xiv Acknowledgements

References Cummins, R. 1998. Reflections on Reflective Equilibrium. In Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry, eds. M.R. DePaul and W. Ramsey, 113–127. Lanham, MD: Rowman and Littlefield. Hartmann, J. 1997. Wozu politische Theorie? Eine kritische Einführung für Studierende und Lehrende der Politikwissenschaft. Opladen: Westdeutscher Verlag. Ladyman, J., and D. Ross with D. Spurrett and J. Collier. 2007. Every Thing Must Go: Metaphysics Naturalized. Oxford: Oxford University Press. Papineau, D. 2006. The Tyranny of Common Sense. The Philosophers’ Magazine 34: 19–25. Pigden, C.R. ed. 2010b. Hume on Is and Ought. Basingstoke: Palgrave Macmillan. Resnik, D.B., and Z. Master. 2011. Criteria for Authorship in Bioethics. The American Journal of Bioethics 11(10): 17–21. Rhode, C. 2017. The Burden of Proof in Philosophical Persuasion Dialogue. Argumentation 31(3): 535–554. Rowling, J.K. 1997. Harry Potter and the Philosopher’s Stone. London: Bloomsbury. Stich, S.P. 1990. The Fragmentation of Reason: Preface to a Pragmatic Theory of Cognitive Evaluation. Cambridge, MA: MIT Press.

Contents

1 D  ialogue and Persuasion  1 1.1 The Problem of Methodological Conservatism   1 1.2 Commitments in Attitude and Dialogue   4 1.3 An Analysis of Philosophical Dialogues   7 1.4 Looking Ahead  27 References 32 2 The  Burden of Proof 35 2.1 Egalitarianism versus Foundationalism  35 2.2 Allocating the Burden of Proof  38 2.3 The Prudence of Dialogical Egalitarianism in Philosophy  47 References 50 3 Evidence,  Inference, and Empiricism 53 3.1 Meeting the Burden of Proof  53 3.2 Evidence  54 3.3 Implication Barriers  58 3.4 Addressing Counter-Examples  63 3.5 Quantitative and Qualitative Barriers  67 3.6 Empiricism  71 References 74 xv

xvi Contents

4 Philosophical  Methods between Content and the World 77 4.1 Consequences of Dialogical Empiricism  77 4.2 The Content/World Separation  78 4.3 Hypotheses in Philosophy  84 4.4 Evidence in Philosophy  89 4.5 Against the Content/World Separation 104 References113 5 M  etaphysical Hypotheses117 5.1 Substantive Consequences 117 5.2 Normativity 118 5.3 Necessity 130 5.4 Epistemology 136 5.5 Ontology 140 5.6 Quietism 147 References152 6 E  scaping Dialogical Empiricism157 6.1 Looking Behind 157 6.2 A Way Around? 160 6.3 A Way Out? 164 References173 Appendix A Reputational Journal Ranking175 Appendix B Population Size177 Appendix C Example Analyses179 Appendix D Analysis Results189 I ndex203

List of Tables

Table 1.1 Table 1.2 Table A.1 Table B.1 Table D.1

Types of dialogue 8 Frequency of types of dialogue 23 Combined Reputational Journal Ranking175 Number of Original Articles per Journal177 Types of Non-Embedded Dialogues in Sampled Journal Articles189

xvii

1 Dialogue and Persuasion

1.1 The Problem of Methodological Conservatism Take any apparent staple among analytic philosophical methods, and you find debates about the nature, philosophical role, and evidential value of the same sprinkled generously across the methodological literature of the last few decades. In the critique and defence of such commonplace philosophical devices as intuitions, thought experiments, or reflective equilibria, perspectives as disparate as the scientism of James Ladyman et al. (2007) or the rationalism of Elijah Chudnoff (2013) have been pursued, among many others, leaving hardly any methodological stone unturned in the process.1 As the methodological foundation stones of contemporary analytic philosophy are thus by and large all up in the air, one might think that analytic philosophers beyond metaphilosophy would turn at least partially from their various substantive investigations and attend to these methodological issues as a matter of priority: Surely philosophers 1

 For a very useful overview of recent work in these methodological debates, see Daly (2015).

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_1

1

2 

C. Rhode

would want to make sure that their tools are useful before trying to use these tools again? What can be observed, though, is quite the contrary, as the vast majority of analytic philosophers evince no such desire. It is business-as-usual in virtually all substantive sub-disciplines of analytic philosophy, where the various challenges to and critiques of the methods employed in these same substantive investigations are, in effect, being ignored. Reflective equilibria continue to be pursued; thought experiments continue to be wrought; and things that look a lot like intuitions continue to be invoked.2 Quite regardless of the serious concerns raised about common philosophical methods, substantive philosophical debates simply continue unperturbed. The attitude that appears to be at the root of this indifference is perhaps best illustrated in a remark by George Bealer, made with regard to analytic philosophers’ employment of intuitions as evidence. Bealer claimed that ‘[i]t is standard justificatory practice to use intuitions evidentially. Unless and until a reason for departing from this standard practice is produced, we are entitled—indeed, obligated—to continue using intuitions as evidence’ (1996: 30 n. 15). This sentiment, which may adequately be characterised as methodological conservatism (cf. Papineau 2006), appears to hold sway with regard to other standard philosophical methods as well, to the effect that analytic philosophers’ employment of these methods may supposedly continue ‘[u]nless and until a reason for departing from this standard practice is produced’. This conservative default position has the effect of placing the burden of proof firmly upon anyone challenging the methodological status quo, that is to say, upon the critics of standard philosophical methods, to the benefit of the defenders of these methods. The present work originated as a response to this conservative allocation of the burden of proof, in order to show that the burden of proof must in fact lie upon the defenders of standard philosophical methods,  To underline this ignorance, it suffices to note the following: In a representative sample of 110 articles recently published in leading Anglophone analytic philosophy journals, only five articles engaged with these methodological debates at least peripherally. (The details of this sample will be given in Sect. 1.3. The foregoing five articles were the 98th, 101st, 102nd, 108th and 109th article, as listed in Appendix D.) 2

1  Dialogue and Persuasion 

3

and hence that no philosophical method should be employed again ‘unless and until’ its evidentiary pedigree has been tested and established. In its current form, however, this work also considers how the burden of proof in methodological and other analytic philosophical dialogues might be satisfied. In addressing both the allocation and the satisfaction of the burden of proof, I develop a form of empiricism that I call dialogical empiricism, which states that the burden of proof for an asserted and questioned proposition prudently (a) lies upon the asserting party and (b) must be satisfied without crossing any inference barrier. Consider, for instance, the proposition that S: The employment of standard philosophical methods is permissible, and assume that S has been asserted by one party in a philosophical dialogue and questioned by another party in that dialogue. In this scenario, dialogical empiricism states that (a) the asserting party must prudently support or else retract S, no matter how basic or obvious S may seem. And dialogical empiricism furthermore states that, (b) in supporting S, the asserting party must prudently not invoke evidence propositions that differ in kind from S, such as descriptive propositions in support of the normative proposition S, because that would entail the crossing of the corresponding inference barrier between these two kinds (here: Hume’s is/ought barrier). The present thesis thus adopts a core tenet of empiricism, viz. that inferences beyond what is warranted by the evidence should be avoided, and applies this core tenet to the conduct of dialogue in philosophy—whence the name ‘dialogical empiricism’. In this book, dialogical empiricism lies at the argumentative centre of a series of considerations that, I believe, are notable for their metaphilosophical consequences as well as their argumentation-theoretic starting points. The first of these starting points is the logical separation between commitments in dialogue and commitments in attitude: My assertion that P, for instance, does not entail my belief that P, nor vice versa. The second starting point is the goal pursued in philosophical dialogue: As I will demonstrate empirically, it is the predominant goal of interlocutors in analytic philosophical dialogues to persuade their opponents—rather than, say, to prove or disprove a contested thesis. These two starting points together give rise to dialogical empiricism, and dialogical empiricism in turn yields as a metaphilosophical conclusion that metaphysical

4 

C. Rhode

debates must prudently be excised from analytic philosophical dialogue: For as I will discuss later in this book, analytic metaphysical debates cannot be conducted without reliance on the crossing of inference barriers, which is prudently prohibited by dialogical empiricism. I introduce the foregoing two argumentation-theoretic starting points, respectively, in the next two sections, first separating commitments in attitude from commitments in dialogue, and then assessing the goal of philosophical dialogues. In the final section of this chapter, I offer an overview of the considerations resulting from these two starting points, which will take us from the allocation of the burden of proof that gives rise to dialogical empiricism through to the recommendation that analytic metaphysical dialogue must prudently be terminated.

1.2 Commitments in Attitude and Dialogue For any proposition P, I can commit to either or neither of the potential truth-values of P. I can, that is, commit to P, or to ¬P, or to neither P nor ¬P. This is exemplified when believing that P (thus committing to P), disbelieving that P (and so committing to ¬P), or wondering whether P (thereby committing to neither P nor ¬P). Each of the truth-value commitments just exemplified is here made in, or ‘located’ in, attitudinal states, such as beliefs. Call them, therefore, attitudinal commitments. Apart from their attitudinal location, however, commitments regarding the truth-value of P can also be made in dialogue. I can, for instance, assert that P, deny that P, or ask/question whether P, thus committing dialogically to P, to ¬P, or to neither P nor ¬P.3 (It may be possible in principle to commit to both P and ¬P, for example by dyserting that P, where ‘to dysert’ is defined as ‘to simultaneously deny and assert’. My present concern, however, is solely with commitments in practice. It will become clear below that, for practical purposes, it is not possible to dysert.)  It is of course possible to assert that P (or deny that P or ask whether P) without engaging in a dialogue, for example by assertively uttering P to oneself. That notwithstanding, I subsequently employ ‘asserting that P’ and cognate expressions solely to denote ‘committing dialogically to the truth of P’ and cognate expressions, and I employ the other foregoing examples likewise exclusively to denote the other dialogical and attitudinal commitments. 3

1  Dialogue and Persuasion 

5

Commitments to either or neither of the possible truth-values of P can thus be made in attitude and in dialogue.4 This separation between attitudinal and dialogical truth-value commitments is itself founded in a distinction between targets, or recipients, of such commitments. A dialogical truth-value commitment is directed towards, or made to, someone else (lest it would not be located in a dialogue), whereas an attitudinal truth-value commitment is directed towards or made to oneself (being itself inaccessible to others). Strictly, then, the target of any of one’s own truth-value commitments either is or is not identical with oneself, thus dividing truth-value commitments exclusively and exhaustively into those made in attitude and those made in dialogue. (At this point, it should perhaps be noted that the attitude/ dialogue separation is not co-extensive with any internal/external separation, not least because it is possible to engage in an internal, imaginary dialogue with a fictional interlocutor.) Given this separation, no attitudinal commitment necessitates any dialogical commitment, nor vice versa (Aijaz et al. 2013; see also Alston 1976: 176–9; Audi 1993: 118–23; Rescorla 2009a: 93–94, 2009b). I can, as a matter of fact, deny in dialogue that Knowledge is an anthropogenic kind, yet wonder attitudinally whether this proposition might not be true after all. Equally, I can believe that Implicit biases contribute significantly to philosophy’s relative demographic homogeneity without asserting the same to an interlocutor, most obviously when I am not in the presence of any potential interlocutor. Dialogical and attitudinal commitments are thus logically independent, in that a given truth-value commitment to P in one location does not entail the same truth-value commitment to P (or indeed any other commitment) in the other location. (Identifying a truth-value commitment thus strictly requires some degree of specification of its target, that is, of the party to whom the commitment is made.) Of course, most people’s dialogical truth-value commitments correspond to their attitudinal truth-value commitments in  The terminology for this separation, ‘attitudinal’ vs. ‘dialogical’, is adapted from Imran Aijaz, Jonathan McKeown-Green and Aness Webster (2013). I use the terms ‘dialogue’ and ‘dialogical’ instead of the more commonly used terms ‘discourse’ and ‘dialectical’ due to the simplicity of cognates and in order to preclude any resultant possibility of misinterpretation in light of twentieth century French philosophy (‘discourse’) or nineteenth century German philosophy (‘dialectical’). 4

6 

C. Rhode

most circumstances. That correspondence, however, is entirely contingent: No logically necessary connection between dialogical and attitudinal commitments arises from this statistical correlation. It might be claimed that this independence of commitments in different locations is curtailed by norms prescribing permissible combinations of commitments. The social norm that One ought not to lie, for instance, proscribes making a dialogical commitment without having the corresponding attitudinal commitment. Equally, the so-called knowledge norm of assertion requires that one ‘assert that P only if one knows that P’ (Brown and Cappelen 2011: 1; cf. Williamson 2000: 11), thus just as well proscribing the making of a dialogical commitment without the corresponding attitudinal commitment. I grant for present purposes that there may well exist social, ethical, epistemic, linguistic, rational or other norms regarding the appropriateness of, for instance, asserting that P without believing that P, as well as that such norms may be employed as premises in such inferences as from one’s assertion that P to one’s belief that P. Of course, norms linking dialogical commitments to attitudinal commitments may themselves be rejected. But this is not the place to decide whether such a norm connecting the two commitment locations should be accepted or rejected. (I will touch upon that question again towards the end of the present chapter, and respond fully in Chap. 6, Sect. 6.3 by effectively rejecting such norms.) At this point, I merely seek to emphasise that such norms would be connecting two commitment locations that are themselves logically independent, such that no truth-­ value commitment in one location by itself necessitates any truth-value commitment in the other location. I stress this perhaps uncontroversial logical independence in order to concentrate solely upon dialogical commitments in my subsequent considerations. These considerations will address commitments in the academic discipline of analytic philosophy; and that discipline is conducted mostly (if not entirely) by way of dialogue between philosophers, rather than via mere attitudinal musings. My considerations regarding commitments in analytic philosophy must thus inevitably address commitments in philosophical dialogue—and hence dialogical commitments. This significant disregard for questions of attitudinal commitment, particularly including epistemological questions, in favour of questions

1  Dialogue and Persuasion 

7

of dialogical commitment is one respect in which the present work looks to depart from much of the current literature in philosophical methodology. The other key point of departure is my aim throughout this work, not to establish some or another truth, but merely to advise on the most prudent, or instrumentally rational5 conduct of analytic philosophy— and specifically on the most prudent conduct of philosophical dialogue with regard to dialogical commitments. In order to determine the instrumentally most rational conduct of such dialogue, it is of course necessary to first determine the goal to the attainment of which a given conduct would be so instrumental. That is to say, it is necessary for the advisory purposes of this work to identify the goal of philosophical dialogue. This is the purpose of the next section of this chapter, where I draw on Douglas Walton’s typology of dialogues for a content analysis of a representative sample of philosophical dialogues.

1.3 An Analysis of Philosophical Dialogues Types of Dialogue Through decades spent analysing the progression of observed dialogues across numerous formats, Douglas Walton and colleagues have distinguished seven practically mutually exclusive but not necessarily exhaustive ideal types of dialogue, abstracted from real-life dialogues imperfectly instantiating one or more ideal types (Walton 2014: 33; cf. 2016: 65–6). The ideal types, catalogued with key characteristics in Table 1.1,6 are distinguished according to both the goals pursued in a given type of dialogue and the rules of engagement accompanying these goals (Walton and Krabbe 1995: 66). The relative significance of different goals, however, appears to diverge between adversarial and collaborative types of dialogue. An adversarial type of dialogue can be delineated essentially by the goal of each individual participant, while what Walton calls the goal  I employ the Machiavellian conception of prudence as mere instrumental rationality, rather than its wider, classical conception as a civic virtue (cf. Hariman 2003). 6  Where necessary, I have adapted the description of characteristics to take account of the separation between attitudinal and dialogical commitments. 5

8 

C. Rhode

Table 1.1  Types of dialogue Type of dialogue

Relation of Initial situation participants

Information Need for information Inquiry

Discovery

Persuasion

Collaborative

Need for Collaborative conclusive proof or disproof of a hypothesis Need for Collaborative explanation of a phenomenon Conflict of Adversarial dialogical commitments

Deliberation Need for action

Collaborative

Goal of any participant

Goal of the dialogue

Acquire or Exchange give information information Find and verify Prove or disprove evidence hypothesis

Find and assess potential explanation Get opponent to adopt your dialogical commitment Coordinate goals and actions Maximise your own benefit

Negotiation Conflict of Adversarial interests and and expected collaborative benefit of agreement Eristic Personal Adversarial Verbally hit conflict out at opponent

Choose best explanation for phenomenon Resolve or clarify conflict of commitments

Agree a course of action for implementation Reach a mutually acceptable settlement

Reveal deeper basis of conflict

Adapted from Walton and Krabbe (1995: 66), Walton (2013: 200–2, 2014: 34), Groarke (2017). Essential goals are italicised

of the dialogue overall is a side benefit not necessarily pursued by participants. By contrast, in a collaborative type of dialogue the overall goal of the dialogue is essential to the dialogue and is shared by each participant, while participants’ individual goals (aside from this shared goal) merely contribute to the pursuit of the shared goal.7 Given the different  Collaborative types of dialogue are not adversarial, while adversarial types of dialogue are collaborative at least to such degree as is required for the very conduct of dialogue. It would thus be more 7

1  Dialogue and Persuasion 

9

characteristics associated with each goal, it is possible to determine the type of a given dialogue from its characteristics and thereby identify the defining, underlying kind of goal of that dialogue. This is what I will do shortly. Yet before this typology can be employed for the analysis of the types instantiated by philosophical dialogues, and hence of the goals pursued in such dialogues, several further characteristics of various types must be noted (cf. Walton 2014: 34–6, 42, 236–8). Both inquiry dialogues and discovery dialogues aim at the production of evidentially supported knowledge. However, an inquiry dialogue aims to conclusively establish whether it is the case that P (or otherwise to show that this cannot currently be established), whereas a discovery dialogue aims to explain why it is the case that P. Consequently, a hypothesis to be tested is already given in an inquiry dialogue, while one or more hypotheses must first be arrived at in a discovery dialogue before these are tested (McBurney and Parsons 2001: 417–8). Once identified, however, it is characteristic of both inquiry and discovery dialogues that a hypothesis is submitted to testing in order to gather supporting or detracting evidence. Yet the standard of evidential support to be satisfied in order to successfully conclude an inquiry dialogue is much higher than the corresponding standard in a discovery dialogue, since achieving the goal of the former but not of the latter requires that subsequent retraction of its conclusion is precluded. In contrast to dialogues of these two types, deliberation dialogues and negotiation dialogues are not aimed at acquiring knowledge but instead at the consensual coordination of subsequent action. Between themselves, they differ centrally in the compatibility of participants’ preferences: Participants engage in deliberation dialogue with each other if their individual preferences are jointly realisable, and in negotiation dialogue if they are not. Furthermore, achieving a deliberated consensus that actually satisfies each participant’s preferences requires the full disclosure of any such preferences, whereas maximising one’s preference satisfaction as part of a negotiated consensus requires the strategic and likely limited disclosure of one’s preferences. Turning now to persuasion dialogue, it is the essential characteristic of accurate to classify types of dialogue as either ‘adversarial’ or ‘non-adversarial’, though I retain Walton’s original terminology here due to its obvious benefit of clarity.

10 

C. Rhode

such dialogue that either one or each party to the dialogue attempts to get her opponent to abandon his initial dialogical truth-value commitment regarding some proposition P in favour of her own commitment.8 Such a change of commitment is pursued in this ideal type by presenting a chain of arguments culminating in one’s initial commitment regarding P, based only on commitments made or assumed to be made by one’s opponent. The goal of persuasion dialogue is commonly considered (including by Walton) to be a change in the opponent’s attitudinal rather than dialogical truth-value commitment. Pursuit of an attitudinal change, however, practically requires pursuit of a dialogical change: Any pursuit of an attitudinal change in one’s interlocutor requires some speech act to the potential effect of that attitudinal change;9 and such a speech act equally has the potential to effect a dialogical change.10 The immediate and indispensable goal of persuasion dialogue, then, is simply to lead one’s opponent to adopt one’s own dialogical commitment and/or to not be influenced likewise oneself, though one’s mediate goal may of course still be a change in one’s opponent’s attitudinal commitment.11 Persuasion dialogue can thus be located between the foregoing two pairs of types of dialogue. Consistent with inquiry and discovery dialogue, the object of the conclusion of a persuasion dialogue is a proposition rather than a set of actions. (Note that a persuasion dialogue is not itself action-guiding: Even when the proposition at issue concerns what ought to be done, the dialogue differs from a deliberation or negotiation dialogue, as the latter concern what will be done.) Further, in line with deliberation and negotiation dialogue, the successful conclusion of a persuasion dialogue requires consensus rather than evidentially supported  Given one’s initial commitment to P, one’s opponent may either dissent from P, by committing to neither P nor ¬P, or he may dispute that P, by committing to ¬P and attempting to get one to commit to ¬P as well. Any persuasion dialogue features one or the other initial situation (cf. van Eemeren and Grootendorst 1983: 82). 9  I do not assume that pursuit of something is necessarily intentional, thus understanding ‘pursuit of ’ minimally as ‘action to the potential effect of ’. 10  This distinction is not to be confused with Michael Dummett’s (1978: 295–6) distinction between suasive and explanatory arguments. An explanatory argument would be made, as Dummett states, when one’s interlocutor has already committed to the truth of its conclusion; yet the aim of persuasion dialogue is to bring about such commitment in the first instance. 11  The label of persuasion for this type of dialogue is therefore slightly misleading, though I retain it for the sake of consistency with the literature. 8

1  Dialogue and Persuasion 

11

knowledge (though knowledge may of course contribute to reaching consensus). As such, it is neither necessarily nor typically part of a persuasion dialogue to submit a hypothesis to tests in order to gather supporting or detracting evidence. Moreover, consistent with inquiry dialogue, the proposition at issue in a persuasion dialogue is given from the start. In line with discovery dialogue, achieving the essential goal of a persuasion dialogue does not necessitate any exceedingly high standard of support and so does not preclude retraction. It is worth noting that the object of any dialogue of the above types may be conditional or unconditional. In the case of types of dialogue whose object is a proposition (i.e. inquiry, discovery, and persuasion dialogue), the proposition or hypothesis at issue may be conditional or unconditional. An inquiry dialogue, for instance, may be aimed at proving the truth or falsity of a conditional hypothesis, such as the hypothesis that If an armchair is burned, then greenhouse gases are released in the process. Equally, an inquiry dialogue may be aimed at proving or disproving an unconditional hypothesis, such as that The Earth is flat. The same applies to, say, a persuasion dialogue, as indeed the foregoing hypotheses may just as well be the objects of persuasion dialogues. In the case of types of dialogue whose object is a set of actions for subsequent implementation (i.e. deliberation and negotiation dialogue), the agreed course of action may similarly be conditional or unconditional. My friend and I might conditionally decide to take a trip to Loch Ossian in the Scottish Highlands on Saturday, or alternatively, to go shopping in Glasgow in the event that Friday’s weather forecast predicts torrential rain for Loch Ossian. After a couple of hours of window shopping on Saturday (as the Scottish weather turned out true to form), we might decide unconditionally to take a break in a coffee house. As these examples serve to illustrate, none of the above types of dialogue are restricted in scope with regard to whether their objects are conditional or unconditional. Consequently, the analysis of philosophical dialogues that is to follow below will be unaffected by the conditionality or otherwise of the objects at issue in these dialogues. Finally, it should be noted that the pursuit of the essential goal of a dialogue of one type often requires a temporary shift to a separate dialogue of a different type, whereby this new dialogue is embedded within

12 

C. Rhode

the original dialogue for the purpose of advancing the latter (Walton 2014: 37). In such a case, the conclusion of the embedded dialogue serves as a contribution to the overarching dialogue, thus helping participants achieve the essential goal of that overarching dialogue. Embedded within a (long) deliberation dialogue, for instance, may be several persuasion dialogues and information dialogues (Walton 2014: 212, 217). Crucial to note, however, is that the type of an embedded dialogue has no bearing on the type of the dialogue wherein it is embedded, nor vice versa. The only impact between these dialogues lies in the embedded dialogue being potentially curtailed by, having its rules of engagement altered by, or the pursuit of its goal made conditional upon, the pursuit of the goal of the non-embedded dialogue.

Philosophers’ Self-Conceptions In light of this typology of dialogues, one might already suspect some or another particular type to be prevalent in philosophical dialogue, potentially due to one’s particular professional self-conception. It is worth examining such professional self-conceptions for a moment, as they might offer a preliminary indication of the type of dialogue prevalent in philosophy. According to Walton (1992), two very broad groups of such professional self-conceptions may be distinguished in philosophy, viz. those conceiving of philosophy as a cognitive enterprise, and those conceiving of philosophy as a critical enterprise. On the former conception, philosophy is in some sense related to, and perhaps continuous with, the natural and/or social sciences. Here, philosophy increases our knowledge or understanding through the assessment, explanation, or clarification of facts, truths, or meanings. Originating with Plato, this picture of philosophy is, and has been since the Enlightenment, integral to the professional self-conception of arguably the majority of philosophers. Indeed, one might ask, what else could philosophers possibly be doing when they examine the nature of mental content, spell out the prescriptions of justice, or explicate the passage of time? Surely, one may hold, philosophy must be in the business of advancing our knowledge—very broadly like any other cognitive, that is,

1  Dialogue and Persuasion 

13

truth-­directed, enterprise. This cognitive conception would suggest that the types of dialogue predominantly instantiated in philosophy are those featuring a cognitive goal, viz. discovery dialogue and/or inquiry dialogue. On the conception of philosophy as a critical enterprise, the disciplinary goal is to interrogate beliefs, thus to unearth and challenge errors, hidden assumptions and inconsistencies in them. This may indirectly lead to greater personal insight, thereby laying the groundwork for other, cognitive pursuits—as the avoidance of error and inconsistency, as well as awareness of hidden assumptions, is of obvious benefit to any cognitive pursuit. Notwithstanding these indirect fruits of such an enterprise, much of philosophers’ work is negative on this Socratic and indeed sophistic conception of philosophy.12 As Walton (1992: 137) explains (citing Henry Johnstone 1978): ‘Unlike a scientific inquiry, philosophical argumentation [thus conceived] is typically directed towards refuting opposed views.’ On this critical conception, according to Walton, philosophy both incubates and critically interrogates competing arguments on the ‘marketplace of persuasion’ (Walton 1992: 126). Arguments are developed and challenged in attempts to persuade one another of one of a number of competing views (cf. Cohen 1986: 3). The predominant type of dialogue in philosophy would thus be persuasion dialogue.13 If these two conceptions of philosophy reflect at least broadly the actually predominant professional self-conceptions of philosophers (which I need not assume to be the case), and if such self-conceptions are furthermore indicative of the predominant types of dialogue in philosophy, then we should expect that discovery and/or inquiry dialogue, or persuasion dialogue, is prevalent in philosophy. It must be emphasised, however, that one’s conception of the type of the dialogue one is participating in need not be indicative of the actual type instantiated by that dialogue. It is entirely possible, for instance, to believe oneself to be engaged in an inquiry dialogue when the dialogue in question actually instantiates the type of persuasion dialogue (cf. Walton 1992). As such, the foregoing conceptions of philosophy can at most serve to facilitate the formulation  Note that Walton’s reference to sophism is non-pejorative.  Whether a particular philosophical programme (such as, say, the Canberra Plan) would belong to one or the other conception is an interesting but presently irrelevant question. 12 13

14 

C. Rhode

of hypotheses to be tested. The hypotheses thus suggested are: Hid: Philosophical dialogues instantiate the type of inquiry dialogue and/ or the type of discovery dialogue significantly more frequently than they instantiate any other type of dialogue. Hp: Philosophical dialogues instantiate the type of persuasion dialogue significantly more frequently than they instantiate any other type of dialogue. The corresponding null hypotheses are: H0(id): Philosophical dialogues do not instantiate the type of inquiry dialogue and/or the type of discovery dialogue significantly more frequently than they instantiate any other type of dialogue. H0(p): Philosophical dialogues do not instantiate the type of persuasion dialogue significantly more frequently than they instantiate any other type of dialogue. Of course, the assessment of the prevalence of different types of dialogue in philosophy is not a correlational study, so the formulation of hypotheses may seem somewhat contrived. Beyond their clarificatory value, however, testing these hypotheses will not least reveal whether either of the foregoing professional self-conceptions of philosophers happens to be accurate. Nonetheless, my primary concern here is not with the assessment of the professional self-conceptions described by Walton, but rather with the identification of the goal of philosophical dialogue. The competing professional self-conceptions are each associated with certain types of dialogue (as specified in the above hypotheses), and thus indirectly with certain goals; but these professional self-conceptions still need not be indicative of the actual goal of philosophical dialogue. That goal (or goals) must be identified by directly examining philosophical dialogue itself. For this purpose, I will now conduct an empirical analysis of philosophical dialogues, in order to determine the relative prevalence of the different types of dialogue in philosophy, and by extension the relative prevalence of the different dialogical goals.

1  Dialogue and Persuasion 

15

Methodology Philosophical dialogues are located variously, as Jonathan Cohen put it (Cohen 1986: 3), ‘in informal conversation, or in formal debate, or between imagined dramatis personae, or in successive publications, with one contributor taking over from another, through minutes, months, years, or even centuries’. Such dialogues are thus highly spread out— across time, across venues, and across interlocutors. Since my aim in this work is to offer prudential recommendations for future engagements in such dialogue, it is most conducive to focus on contemporary philosophy for the analysis of the relative prevalence of different types of dialogue. Even so, philosophical dialogues are taking place across a multitude of venues, such as imaginary debates, chats over a coffee, seminar discussions, conference Q&As, unsolicited journal articles, symposia, edited volumes, and monographs. Articles in peer-reviewed journals (outside symposia or special issues),14 however, both provide a ready cross section of philosophical dialogues and have been subjected to the most rigorous peer-review, and they therefore offer a sufficiently broad snapshot of contemporary philosophical dialogues. There are three specific aspects to this suitability. First, journal articles are very often developed out of (actual or hypothetical) relatively brief discussions, and they precede the further elaboration of their contents in any subsequent monographs. Journal articles are thus contributions to ongoing dialogues (between parties whose membership is ever-changing) which are themselves conducted predominantly through journal articles, comparatively briefer discussions, and monographs, the latter two feeding into or drawing upon journal articles. It is therefore prima facie likely that the types of dialogue observable in journal articles match the types observable in other venues that feed into or draw upon journal articles. Second, journal articles undergo a more rigorous and more competitive process of selection and refinement than dialogue contributions  I subsequently use ‘journal articles’ and cognate expressions to denote original, competitively selected journal articles published in peer-reviewed journals outside any symposium, special issue, or similar format. 14

16 

C. Rhode

made in other venues. For a journal article to successfully pass through the various rounds of peer-review and editorial selection, objections and improvements by parties other than the authors’ are integrated into and addressed in the article, either in advance of or during the peer-review process. The contributions and role of other parties in the dialogue are thereby internalised into the article, so that a journal article reflects a sustained dialogue between the authors’ party and other parties, which is itself a contribution to a longer, ongoing dialogue (Dutilh Novaes 2016).15 By contrast, seminar or conference contributions are typically regarded as being, to some extent, works in progress, while monographs enjoy somewhat greater liberty in the choice of which other parties to internalise and to what extent—in both cases because the pressure of peer-review to internalise other parties is arguably not applied with the same reliability and/or strength as it is in the case of journal articles. Journal articles therefore offer a greater concentration of moves in their respective dialogues, thus making them more useful for an analysis of the types of such dialogues. Third, a journal’s peer-review process of selection and refinement not only enforces this internalisation but (to some extent thereby) also legitimises the eventual articles as recognised contributions to philosophical debates—more so than a conference’s or book publisher’s review process does (cf. Hemmings 2005: 117–8; Erikainen 2020: 6). An article’s publication in, say, a top-3 journal confers greater prestige than a presentation’s inclusion in a top-3 conference, or a monograph’s publication with a top-3 press. As such, the venues regarded as admitting only the ‘best’ pieces of philosophical work are themselves journals, rather than conferences or book publishers. These journals thereby create and reproduce a legitimised standard which most philosophers seek to emulate. In order to identify the types of dialogue recognised as satisfying this standard, it is therefore more effective to look to the standard-setting venues themselves, rather than to venues where this standard is merely being accepted and approximated. So, in summary, since journal articles are situated at the centre of the development of philosophical ideas and at the peak of  A dialogue is thus conducted between parties rather than between members of parties—though I will subsequently often write in ways that suggest the latter, for the sake of expressive simplicity. 15

1  Dialogue and Persuasion 

17

enforced rigour, a representative sample of such journal articles should also be indicative of the relative prevalence of different types of dialogue in philosophy. As the third point just mentioned already indicated, however, not any odd set of journal articles will do—since the demandingness of peer-­ review reputedly varies between journals, with an article’s publication in a highly reputed journal being considered an indication of the article’s satisfaction of the highest philosophical standards. If the selected articles were largely regarded as falling short of such standards, then any recommendations for the optimal conduct of philosophy that I would subsequently draw from the analysis of these articles may be undermined. To avoid such difficulties, I select articles for analysis exclusively from those journals that are listed among the top-25 in both of the leading reputational journal rankings in philosophy, viz. the Brooks (2011) and Leiter (2013) rankings.16 There is considerable overlap between these rankings, with 22 journals being among the top-25 in both of them.17 Given the aforementioned focus on contemporary philosophy, it is reasonable to sample journal articles merely from the most recent decade preceding this research (which was conducted in 2015 and has also been published in Rhode 2017). To estimate the number of journal articles published in that period in the identified 22 journals, I have manually counted their number for the years 2007, 2010 and 2013, and extrapolated the result to a period of ten years. On the basis of this estimate, approximately 6,960 journal articles have been published in the identified 22 journals in the ten-year period to 2015.18  Given their reputational ranks, these journals are relatively most likely to publish work instantiating what is predominantly regarded as meeting the highest philosophical standards. Of course, this implies that they publish work actually meeting the highest philosophical standards only insofar as ‘the highest philosophical standards’ is defined by a disciplinary mainstream. 17  These 22 journals include: American Philosophical Quarterly, Analysis, Australasian Journal of Philosophy, British Journal for the Philosophy of Science, Canadian Journal of Philosophy, Erkenntnis, Ethics, European Journal of Philosophy, Journal of Philosophy, Journal of the History of Philosophy, Mind, Noûs, Pacific Philosophical Quarterly, Philosophers’ Imprint, Philosophical Quarterly, Philosophical Review, Philosophical Studies, Philosophy & Phenomenological Research, Philosophy & Public Affairs, Philosophy of Science, Proceedings of the Aristotelian Society, Synthese. Their ranks in both rankings, and the ranks of journals ranked among the top-25 in only one of the two rankings, are reproduced in Appendix A. 18  The number of journal articles counted in each journal in each of the three years selected is documented in Appendix B. Outlier results within any given journal are excluded, as also documented. 16

18 

C. Rhode

I will analyse a representative sample of journal articles from among these 6,960 articles in up to two stages, beginning with an exploratory sample, and then expanding to a wider sample unless saturation is attained. For the initial representative sample, I select one article from each journal from every even-numbered year in the ten-year period to 2015, thus selecting 22 (journals) × 5 (even-numbered years) = 110 articles. The frequency of each type of dialogue found in such a sample of 110 articles will be representative of the frequency of these types of dialogue in the population of 6,960 articles at a confidence level of 95% with a margin of error of ±9.3% (or at 90% confidence with a ±7.9% margin). I will then conduct a second round of analyses, expanding the sample size beyond these 110 articles, in order to decrease the margin of error—unless saturation is attained within these 110 articles, that is, unless the results from the initial representative sample are so uniform that there is no basis for expecting different results from an expanded sample (cf. Saunders et al. 2018). In selecting a specific set of 110 articles I disregard not only articles published within special issues and symposia, but also discussion notes, reviews, editorials, and errata, due to the same considerations regarding the pressures of the peer-review process already outlined. I also disregard any obviously satirical articles.19 Among the journal articles thus remaining, I select, arbitrarily, the third article published in the given journal in the given year’s volume(s). Among the set of 110 articles thus selected, none turned out to be in so-called continental philosophy or in any non-­ Western philosophy,20 and all were written in English. The sample thus exclusively represents current so-called analytic Anglophone philosophy, though it should be expected in light of shared methods and topics that it is also representative of non-Anglophone analytic philosophy. Analysing the sampled articles, I identify the exemplified type of dialogue in two steps. First, the non-embedded dialogue in the article must be distinguished from any other dialogues embedded within it. Many articles, for instance, provide initial definitions and/or an overview of the state of the debate on an issue. This provision of background information  The only satirical article thus disregarded is Willoughby (2012), which I will return to shortly.  As a consequence, I subsequently employ ‘philosophy’ and ‘analytic philosophy’, as well as cognate expressions of either, to denote ‘Western analytic philosophy’ and cognate expressions thereof. 19 20

1  Dialogue and Persuasion 

19

(itself seemingly instantiating information dialogue) serves to facilitate the dialogues that follow it, and is thus embedded within one of these subsequent dialogues. Equally, an article may incorporate an inquiry dialogue wherein a given hypothesis is being tested, with the test result being employed to advance some other dialogue in the article and the inquiry thus being embedded within that other dialogue. As such, the task in this step is, first, to assess each paragraph or similar unit of text in an article for its function or purpose within the article as a whole. Second, it is necessary to identify the passages or paragraphs that together serve the same function or purpose, thus identifying the different dialogues in the article (such as a series of paragraphs that provide background information on existing arguments, followed by several paragraphs attempting to undermine those arguments). Third, from among the dialogues thus distinguished it is necessary to identify the dialogue that is facilitated or advanced by all others without itself facilitating or advancing any other dialogue in the article. On the basis of the functions of the various paragraphs of an article I thus identify the one dialogue (i.e. series of passages or paragraphs) in which all other dialogues are embedded. (No article contains a plurality of such non-embedded dialogues; for if it did, it would be pursuing a plurality of separate and wholly independent conclusions or purposes—presumably barring it from publication at least in any of the above 22 most highly reputed journals.) The next task is to identify the type of the non-embedded dialogue just distinguished. Here I assess the given non-embedded dialogue against the characteristics of each type listed above (including in Table 1.1). These characteristics, though quite clearly characterised, are themselves qualitative rather than quantitative, so the assessment of their respective instantiation or non-instantiation is to a degree interpretive. To mitigate against undue bias, I disregard any characteristics in the analysis of a given non-­ embedded dialogue to the extent that their instantiation is not unambiguous and readily apparent. The analysis of a given non-embedded dialogue thus relies upon the characteristics listed above solely to the extent that their instantiation is clear and unequivocal—which turned out to be true in most cases. On this basis, I judge the type of a given non-embedded dialogue according to the preponderance of evidence from the individual characteristics. For most dialogues, though, the type

20 

C. Rhode

was so clear as to leave very little room for doubt, where the assessed characteristics effectively admitted of the identification of only one type.21 To uncover any significant level of error in this analysis, I repeated the analysis for 20% of the sample (viz. one article selected at random from each of the 22 journals) one month after carrying out the original analysis. This led to no corrections of the initial results.

Example Analysis: Willoughby (2012) I will now illustrate this analysis using the example of Willoughby (2012)—a satirical article published in the Journal of Philosophy. This article is of particular present interest because any successful satire must contain a recognisable depiction of the target to be satirised—so the type of dialogue of this article is potentially indicative of the type that is prevalent among the targeted philosophical dialogues. Authored under the pseudonym of New York City subway stop Myrtle Willoughby and entitled ‘Stoppism: Retrospects and Prospects’, Willoughby (2012) offers an attack on stoppism: the thesis that trains are metaphysically dependent upon stops, such that a train is a local (or express) train because it calls at local (or express) stops. That thesis is contrasted in the introductory section of the article from theses affirming dependence in the reverse direction (train-realism), no dependence at all (coincidentalism), dependence of both trains and stops upon some other thing ‘inside the subway’ (naturalism) and dependence of both upon some other thing ‘outside the subway’ (theological views; 2012: 282). With stops understood roughly as the places where one boards or alights a train, the author outlines fictional historical debates regarding stoppism in Sect. 1, discussing such problems as the obvious fact of trains sometimes being motionless between stops. This fact seemed inexplicable under stoppism prior to the invention of precise instruments for time-­ keeping which enabled us to recognise a train’s stop between stops as caused by some other train’s stop at a stop—thus reviving stoppism. In  I suspect that the very high standards of writing applied to publications in top-ranked philosophy journals provide at least part of an explanation for this relative clarity of the types of dialogue of the sampled articles. 21

1  Dialogue and Persuasion 

21

Sect. 2 the author discusses four generally known facts that are difficult for stoppists to accommodate, including, for instance, that stops change under certain conditions from being local to being express stops or vice versa, which is inexplicable if stops are metaphysically basic. The author here also considers the naturalistic thesis of station-stoppism advocated by one Davis Wilde (an anagram of ‘David Lewis’), which regards stations as metaphysically basic entities possessing varying, numerous and fine-­ grained stop-properties (that are themselves left for scientists to investigate). This thesis is rejected, as it would leave us unable to specify the identity conditions of any single station. Other naturalistic deflations of stoppism are discussed in Sect. 3, all being rejected due to their failure to provide ‘a set of natural facts [consistent with stoppism] on which to rest the local-stop/express-stop distinction’ (Willoughby 2012: 293). In the concluding Sect. 4, the author thus rejects stoppism. Turning to the analysis of this satirical article, it should initially be noted that satire depicts as laughable a target that is taken seriously by certain people, thus effectively depicting as laughable the people taking said target seriously. Depicting an interlocutor as laughable may be regarded as a contribution to an eristic dialogue. But for present purposes, this satire is more instructive if we identify its type of dialogue on the counterfactual assumption that the article is serious rather than satirical. (After all, the philosophical dialogues targeted by this satire are not themselves satirical—or so I assume.) Given the assumption that the article is serious rather than satirical, we must first distinguish the different dialogues within the article, in order to identify the non-embedded dialogue. The introductory section and Sect. 1 together provide a basic overview of the logical space and fictional historical development of stoppism, seemingly for the purpose of ensuring that readers possess certain background information. In contrast, Sects. 2 and 3 set out various considerations against stoppism (which are distilled in Sect. 4), with the clear purpose of undermining stoppism. Thus, there are just two dialogues in this article, distinguished by their respective goals. If one dialogue furthers the other one and not vice versa, then that dialogue is embedded within the other dialogue. And this is clearly the case: The dialogue in the earlier sections advances the dialogue thereafter by preparing the ground for the latter. So we need not be concerned further with the earlier

22 

C. Rhode

dialogue (which appears to be an information dialogue) and can focus upon the later sections. The object of the conclusion pursued in that non-embedded dialogue is a proposition (viz. the thesis that stoppism is false) rather than a set of actions, so this dialogue instantiates inquiry, discovery, or persuasion dialogue, rather than deliberation or negotiation dialogue. As such, the non-­ embedded dialogue has the goal of (dis)proving some hypothesis, choosing the best explanation for some phenomenon, or getting the author’s opponents to adopt the author’s dialogical commitment. We can rule out discovery dialogue, both because there is no single, overarching phenomenon at issue here that would demand explanation, and because the proposition at issue is given from the start. But the essential goals, though perhaps suggestive, are not obviously decisive between inquiry and persuasion dialogue: Is it the goal of this dialogue to disprove stoppism or to get the author’s opponents to reject stoppism? We need to assess the dialogue against the other characteristics of inquiry and persuasion dialogue to answer this question. The initial situation of the dialogue appears to be a conflict of commitments rather than a need for conclusive proof of a hypothesis (as there is no such apparent need); and the relation between participants is clearly adversarial rather than collaborative. Both of these characteristics indicate that the dialogue instantiates persuasion dialogue. Moreover, the characterisation of persuasion dialogue as offering a chain of arguments seems to match exactly the non-embedded dialogue of this article. Lastly, recall the conception of philosophy as a critical enterprise, which is associated with persuasion dialogue. On that conception, ‘[an argumentation] is typically directed towards refuting opposed views’ (Walton 1992: 137), and the present dialogue obviously aims at refuting stoppism. So we can conclude on the preponderance of evidence that the non-embedded dialogue of this satirical article instantiates persuasion dialogue. A small selection of additional example analyses is provided in Appendix C. These example analyses further illustrate the process of identifying the type of dialogue instantiated by the non-embedded dialogue of an article. The articles analysed in Appendix C are drawn from the present sample and include one article for each of the types found to be represented in the present sample.

1  Dialogue and Persuasion 

23

Results Does the type of dialogue of the foregoing satire, then, prove to be indicative of the prevalent type of dialogue among the sampled articles analysed here? It very much does, as the results in Table 1.2 show.22 Of the non-embedded dialogues of the 110 sampled articles, 95.5% are found to instantiate persuasion dialogue. Given the margin of error of ±9.3%, we can say with 95% confidence that at least 86.2% of the 6,960 journal articles in the population equally exemplify persuasion dialogue (and we can say with 90% confidence that the proportion is at least 87.6%). (Hence Hid and H0(p) can be rejected, while Hp and H0(id) are strongly supported.)23 Moreover, the non-­embedded dialogues of the 110 sampled articles so uniformly instantiate persuasion dialogue that there is no basis for expecting a different result from an expanded sample, that is to say, that saturation has been attained. The present analysis therefore indicates that the vast majority of philosophical dialogues represented here are essentially characterised by the goal of each participant with an initial dialogical truth-value commitment regarding a given proposition P to influence her opponent by way of argument to the effect that the opponent abandons his initial Table 1.2 Frequency of types of dialogue Type of dialogue

Number of non-embedded dialogues instantiating this type of dialogue

Information Inquiry Discovery Persuasion Deliberation Negotiation Eristic

0 3 0 105 2 0 0

 A breakdown of the results by article is provided in Appendix D.  It is interesting that, of the five articles not instantiating persuasion dialogue, four are located in one or another sub-discipline of the philosophy of science and mathematics, while the fifth is located in metaphilosophy. The remainder of the 26 philosophy of science and mathematics articles in the sample (or 28 if metaphilosophy is included), however, all instantiate persuasion dialogue. As I subsequently focus on articles that do instantiate persuasion dialogue, the prevalence of articles in philosophy of science and mathematics among those not instantiating persuasion dialogue is of no relevance to the present work. 22 23

24 

C. Rhode

dialogical truth-value commitment regarding P and instead adopts her initial dialogical truth-value commitment regarding P. In effect, the essential dialogical goal of the vast majority of dialogues in contemporary analytic philosophy is to defeat one’s opposing party in argument—by technical knockout, one might say. Of course, this result applies to analytic philosophical dialogues that have been found to instantiate persuasion dialogue, and it does not apply to such dialogues as have been found to instantiate a type other than persuasion dialogue. Yet where the type of the analytic philosophical dialogue one is currently engaged in has not yet been identified, it is exceedingly likely (given the foregoing empirical results) that this dialogue too instantiates persuasion dialogue. Considering the practical difficulties of analysing the type of one’s own, ongoing dialogue with another party, it is thus prudent to assume the instantiation of persuasion dialogue for any analytic philosophical dialogue one engages in. For all practical purposes, then, it is the goal of dialogue in contemporary analytic philosophy to defeat one’s opponent in argument and not be defeated oneself.24

Attitudinal Goals In light of this conclusion, one might perhaps insist that contemporary analytic philosophy is a cognitive enterprise after all, although its cognitive pursuit of truth or knowledge is attitudinal rather than dialogical. Philosophy, one might hold, pursues the essential goal of inquiry dialogue and/or discovery dialogue, but it does so by way of interrogating hypotheses in persuasion dialogue. This conception of philosophy would acknowledge the observed prevalence of persuasion dialogue, thus granting that the critical conception of philosophical dialogue is accurate, while nonetheless upholding an ultimately and attitudinally cognitive conception of philosophy. Of course, philosophers’ self-conception of the type of dialogue they are engaged in, as noted earlier, is not necessarily indicative of the actual type of their dialogue. But might a cognitive  An exception to this rule is provided by analytic philosophical dialogues that, by their very design, instantiate a type other than persuasion dialogue. I will discuss this exception in Chap. 6, Sect. 6.2. 24

1  Dialogue and Persuasion 

25

self-­conception not be indicative of an attitudinally cognitive purpose motivating their engagement in such dialogue? That is to say, perhaps a cognitive self-conception does not indicate a cognitive type of dialogue but rather a cognitive attitudinal goal underlying the prevalent persuasion dialogue. We should remember here, first of all, that one’s own conception of the attitudinal goal pursued through philosophical dialogue is not necessarily shared by other philosophers. It is entirely possible (and, I suspect, actually the case) that philosophers, though predominantly engaged in the same type of dialogue, are motivated by a colourful variety of different attitudinal goals. Some philosophers might simply want to clarify their own commitments in order to render them consistent (cf. Scanlon 2003). Other philosophers might wish to convince their colleagues of some fundamental truth that answers many a question in a particular sub-discipline (e.g. Elder 2004). Yet other philosophers might believe themselves to be contributing to society’s stock of knowledge in much the same way as scientists are (cf. Papineau 2015). And yet further philosophers might seek to dissolve linguistic confusions arising in both philosophy and the sciences (cf. Hacker 2009). Some philosophers might also quite simply want to play the argumentation game of philosophy for their own enjoyment, seeking no other achievement than their mind’s pleasure (including asteistic pleasure from analytical care and interpretive charity). And there may also be many philosophers who are motivated quite sufficiently by their occupational goal of securing tenure, thank you very much (cf. Gettier 1963; Schukraft 2017: 9–10). So, in short, one’s own attitudinal goal, whether cognitive or otherwise, should not be considered to be indicative of other philosophers’ attitudinal goals that motivate their engagement in philosophical persuasion dialogue. (An empirical examination of philosophers’ explicitly declared attitudinal goals through both comprehensive surveys and extensive textual analyses may establish the relative prevalence of different attitudinal goals. But such research would itself likely be similar in extent to the present work and is at the same time unnecessary for present purposes.) Aside from this cautioning note, there are also some significant considerations against a cognitive attitudinal goal. On the one hand, conceiving of philosophical persuasion dialogue as being motivated by a cognitive

26 

C. Rhode

goal at the attitudinal level would ascribe to philosophers the imprudence of pursuing an attitudinal goal related to one type (or set of types) of dialogue through an entirely different type of dialogue, whose associated rules arise from and are optimised for the pursuit of an entirely different kind of goal. It is at least unclear how exactly one might get closer to attaining an attitudinally cognitive goal by persuading someone else of some particular claim. Indeed, it is not even clear whether engagement in persuasion dialogue would actually advance the pursuit of any attitudinally cognitive goal at all. On the other hand, such insistence upon an attitudinally cognitive goal pursued through persuasion dialogue would also ascribe to philosophers the remarkable epistemic immodesty of holding that their own collective failure to find persistent fault with a hypothesis in and of itself constitutes evidence of the truth of that hypothesis. We need only apply this evidential standard to an unfalsifiable hypothesis, such as the theistic hypothesis of creation and evolution by intelligent design, to recognise its inadequacy: No expert may be able to refute the hypothesis of intelligent design, but the inability of experts to refute an unfalsifiable hypothesis would not lend any support to that hypothesis. If philosophers, however, collectively and consistently fail to refute intelligent design, then this is evidence of the truth of intelligent design!—or so this attitudinally cognitive conception of philosophy would have it. Besides, philosophers’ excellent track-record at finding persistent fault with the vast majority of philosophical hypotheses would then also mean that philosophy has hardly ever produced any knowledge at all, and thus that philosophy is altogether a failure as a cognitive enterprise (cf. Walton 1992: 137). As such, conceiving of philosophy as attitudinally cognitive but dialogically critical (aimed at persuasion) would amount to conceiving of analytic philosophers as methodologically imprudent, epistemically rather presumptuous, and all along singularly unsuccessful. This downright quixotic conception of philosophy seems very uncharitable indeed, and philosophers who are pursuing a cognitive attitudinal goal through persuasion dialogue may be well advised to engage in a cognitive type of dialogue instead.25 I would conjecture, moreover, that this quixotic con I will discuss the feasibility of opting for such different types of dialogue in Chap. 6, Sect. 6.2.

25

1  Dialogue and Persuasion 

27

ception is also inaccurate.26 I therefore disregard speculation about the attitudinal goal of philosophical dialogue here, although I will return to the foregoing questions about that attitudinal goal towards the end of the present work. At this point and for present purposes, it remains the vastly predominant dialogical goal of philosophical dialogue to defeat one’s opponent; and it is this dialogical goal that I will draw upon in subsequent discussions.

1.4 Looking Ahead In the foregoing sections, I hope to have established the following two premises. First, a truth-value commitment made in dialogue does not necessitate any (corresponding or otherwise) truth-value commitment in attitude, nor vice versa. It is thus possible to examine commitments in dialogue independently of any potential attitudinal implications of such commitments, since there are no such necessary implications. Second, it is the dialogical goal of participants in analytic philosophical dialogue to persuade (that is, defeat) their respective argumentative opposing party. Chiefly on the basis of these two premises, I will offer considerations regarding the instrumentally most rational conduct of philosophical dialogue in pursuit of the above goal of persuasion. I will not seek to argue that a certain conduct or goal ought to be pursued, nor that any particular conduct or goal of philosophical dialogue is good, right, true, real, or in any other sense appropriate. Rather, I will presuppose the dialogical goal of persuading one’s opponent (as just established), and I will seek to argue that the adoption of certain conduct would be practically most conducive to the attainment of this goal. (As such, the present work will not instantiate persuasion dialogue but rather deliberation dialogue—so, notably, the conclusions to be drawn here for philosophical persuasion dialogue do not apply to this work.) These purely advisory purposes of the present work will proceed from a response to the methodological conservatism described at the outset, according to which the  Support for this conjecture would have to be provided by an empirical examination of philosophers’ explicitly declared attitudinal goals as mentioned above. 26

28 

C. Rhode

employment of standard philosophical methods may continue ‘[u]nless and until a reason for departing from this standard practice is produced’ (Bealer 1996: 30 n. 15). This methodological conservatism imposes an uneven allocation of the burden of proof that privileges the methodological status quo over challenges to the same. So the first question to be addressed here (in Chap. 2) concerns the allocation of the burden of proof in philosophy. If one party to a philosophical dialogue asserts that P (e.g. that The employment of standard philosophical methods is permissible) and another party asks whether P is indeed the case, then does the burden of proof lie with the proponent of P, or is this burden incurred by the questioning respondent instead? I will discuss several arguments in favour of either answer, such as the contention that placing the burden upon the respondent would be dogmatic, and the claim that requiring the proponent to shoulder the burden of proof would lead to scepticism. On the basis of the attitude/ dialogue separation and the goal of persuasion, I will reject each of the existing arguments for either answer. To resolve the question of the allocation of the burden of proof, I will then consider which allocation would be more conducive to the persuasion of one’s opponents. As I will show, it is prudent in this regard for the proponent to take on the burden of proof herself, rather than (as methodological conservatism would have it) seeking to evade that burden by placing it upon the respondent. Since the burden of proof thus cannot prudently be evaded, the next question to be examined (in Chap. 3) is how it can be met. In order to support a questioned hypothesis H, one must provide evidence— abstractly expressed here by proposition E—and one must claim that E supports H, which I symbolise as E:H. Of course (due to the conclusion of Chap. 2), the party that questioned H may go on to question E or E:H, which would in turn need to be supported. It is thus prudent to avoid not least the employment of any E:H that cannot be supported when questioned. This includes in particular any E:H that crosses an inference barrier, such as Hume’s is/ought separation or, indeed, the foregoing attitude/dialogue separation. In distinguishing different types of

1  Dialogue and Persuasion 

29

inference barriers27 and defusing counter-examples to them, I will argue that a claim E:H must prudently be avoided in supporting H if the propositions E and H are of respectively different kinds, for example where E is a descriptive proposition while H is a normative proposition. This is because a burden of proof upon such an E:H could never be satisfied. In consequence, Chaps. 2 and 3 together yield the intermediate conclusion that, prudently, the burden of proof for an asserted and questioned proposition (a) lies upon the asserting party and (b) must be satisfied without crossing any inference barrier. Being recognisably empiricist in its prescription though dialogical (rather than epistemological) in its scope, I call this intermediate conclusion dialogical empiricism. In order to distil the implications of dialogical empiricism for philosophical methods (in Chap. 4), I will focus on one specific inference barrier, which separates the psychological or linguistic content of talk or thought from the world beyond such content. As I will explain, both the objects of philosophical hypotheses, and the types of evidence employed in support of such hypotheses, can each be divided along this content/ world separation. In this way, for instance, anthropogenic kinds (located in psycho-linguistic content) can be distinguished from natural kinds (located in the world beyond any psycho-linguistic content). Equally, an intuition employed in support of a hypothesis about kinds (where the evidence is part of one’s psycho-linguistic content) can be distinguished from an empirical observation in support of such a hypothesis (where the evidence is part of the world beyond such content). Given the content/ world separation, then, an intuition (located in psycho-linguistic content) cannot prudently be employed in support of a hypothesis about natural kinds (which are located in the world beyond any such content). I will draw similar distinctions with regard to other forms of evidence employed in philosophy, and other objects of philosophical hypotheses, detailing the psycho-linguistic content that prudently can no longer be employed in support of hypotheses about the world beyond such content. I will thus argue not only that standard philosophical methods and the evidential relations employed in them must be supported when 27

 Note that I distinguish between, among other things, inference barriers and implication barriers.

30 

C. Rhode

questioned (contrary to methodological conservatism), but also that supporting many of them is effectively impossible, since such support would require the crossing of the inference barrier between psycho-linguistic content and the world beyond such content. The results up to this point have considerable impact on substantive philosophical debates, as I will illustrate next (in Chap. 5). I will discuss how hypotheses expressing a normative claim, a necessity claim, an epistemological claim or an ontological claim could be supported when questioned, given not least that such support must not involve the crossing of the content/world barrier. As I will argue, it is left practically impossible to support such hypotheses if the claims made in these hypotheses are about the world beyond content—that is, if these hypotheses are metaphysically significant. If, for instance, the necessity of the truth of a hypothesis was taken to be part of the very fabric of the world independent of psycho-linguistic content, then no intuitional or other psycho-­ linguistic content could be employed in support of assertions of that necessity; nor would there seem to be any non-psycho-linguistic evidence (like empirical measurements) available to us to support the claimed necessity; nor would it be open to the proponent of this necessity to seek to evade the burden of proof imposed by the respondent who questioned that necessity. By discussing hypotheses expressing such necessity claims, as well as hypotheses expressing normative, epistemological or ontological claims, I will show that metaphysically significant hypotheses that are supported only by psycho-linguistic content—roughly comprising metaphysical hypotheses that are of philosophical as opposed to scientific interest—are left prudently unsupportable in analytic philosophical dialogue. Hence, I will reach the rather radical conclusion that it is prudent to terminate philosophical dialogue about any exclusively philosophically interesting metaphysical hypothesis. Each of these chapters builds on the conclusions of previous chapters, including not least the attitude/dialogue separation and the goal of persuasion in philosophical dialogue that I have established in the present chapter. The key conclusions of the chapters just outlined may thus be summarised as follows:

1  Dialogue and Persuasion 

31

Chapter 2: Any proposition asserted by one party and questioned by another party in an analytic philosophical dialogue must prudently be supported by the asserting party, regardless of how basic the questioned proposition may appear to be. Chapter 3: The asserting party must prudently avoid employing propositions of one kind in support of a questioned proposition of a different kind, such as descriptive propositions employed in support of a questioned normative proposition. Chapter 4: Many analytic philosophical methods involve the employment of propositions expressing some psychological or linguistic content in support of propositions expressing some aspect of the world beyond such content (thus being of a different kind), and these methods must prudently be avoided in philosophical dialogue. Chapter 5: Without these methods, many propositions expressing an aspect of the world, that is, metaphysically significant propositions, are practically unsupportable in analytic philosophical dialogue and must prudently be excised from such dialogue. Beyond a similar summary of the considerations just outlined, the final chapter will be devoted to assessing potential responses to the same that may be open to methodological conservatives or to analytic philosophers wishing to make metaphysically significant claims. As I will discuss, the imprudence of employing such metaphysically significant propositions may be accommodated, for instance, by employing such propositions only within a coherentist framework—although this would leave analytic philosophers unable to make any claims about reality beyond psycho-linguistic content. Accommodation may also be achieved by discontinuing philosophical persuasion dialogue about metaphysically significant propositions and discussing such propositions within other types of dialogue instead—which would, however, seem variously either impractical or incoherent. Rather than accommodating the above results, one might instead prioritise some personal goal outside of the foregoing types of dialogue, and one might thereby fundamentally undermine these persuasion-oriented results. One may, for instance, seek to avoid having one’s dialogical position challenged at all, by permanently silencing one’s opponents, or one might give priority to the avoidance of

32 

C. Rhode

unemployment, in order to continue debating metaphysically significant propositions. Such overriding non-dialogical goals may be regarded as intellectually questionable or odious; but they cannot be foreclosed by my present considerations. These considerations are premised upon the goal of persuading one’s opponent, and they do not offer any resources for informing the conduct of philosophical dialogues under a wholly different, non-dialogical goal, nor, I should add, for resolving priority conflicts between competing goals. It is only on the basis of the goal of persuasion (and of the attitude/ dialogue separation) that I here consider the prudent conduct of philosophical dialogue. As stated earlier, this prudent conduct concerns, first and foremost, the allocation of the burden of proof. And it is to the question of that allocation to which I will now turn.

References Aijaz, I., J. McKeown-Green, and A. Webster. 2013. Burdens of Proof and the Case for Unevenness. Argumentation 27: 259–282. Alston, W.P. 1976. Two Types of Foundationalism. Journal of Philosophy 73 (7): 165–185. Audi, R. 1993. The Structure of Justification. Cambridge: Cambridge University Press. Bealer, G. 1996. On the Possibility of Philosophical Knowledge. Philosophical Perspectives 10: 1–34. Brooks, T. 2011. In Your Judgment, Which of the Following is the Better Philosophy Journal? All Our Ideas. http://www.allourideas.org/philosophyjournals/results?all=true. Accessed 17 June 2015. Brown, J., and H. Cappelen. 2011. Assertion: An Introduction and Overview. In Assertion: New Philosophical Essays, eds. J. Brown and H. Cappelen, 1–17. Oxford: Oxford University Press. Chudnoff, E. 2013. Intuition. Oxford: Oxford University Press. Cohen, L.J. 1986. The Dialogue of Reason: An Analysis of Analytical Philosophy. Oxford: Clarendon Press. Daly, C. 2015. Introduction and Historical Overview. In The Palgrave Handbook of Philosophical Methods, ed. C. Daly, 1–30. Basingstoke: Palgrave Macmillan. Dummett, M. 1978. The Justification of Deduction. In Truth and Other

1  Dialogue and Persuasion 

33

Enigmas, 290–318. Cambridge, MA: Harvard University Press. Dutilh Novaes, C. 2016. Reductio Ad Absurdum from a Dialogical Perspective. Philosophical Studies 173 (10): 2605–2628. van Eemeren, F.H., and R. Grootendorst. 1983. Speech Acts in Argumentative Discussions. Dordrecht: Foris. Elder, C. 2004. Real Natures and Familiar Objects. Cambridge, MA: MIT Press. Erikainen, S. 2020. Gender Verification and the Making of the Female Body in Sport: A History of the Present. Abingdon: Routledge. Gettier, E.L. 1963. Is Justified True Belief Knowledge? Analysis 23 (6): 121–123. Groarke, L. 2017. Informal Logic. In The Stanford Encyclopedia of Philosophy, Spring 2017 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/ spr2017/entries/logic-­informal/. Hacker, P.M.S. 2009. Philosophy: A Contribution, not to Human Knowledge, but to Human Understanding. In Conceptions of Philosophy, Royal Institute of Philosophy Supplements, ed. A. O’Hear, vol. 65, 129–153. Cambridge: Cambridge University Press. Hariman, R. 2003. Theory Without Modernity. In Prudence: Classical Virtue, Postmodern Practice, ed. R. Hariman, 1–32. University Park, PA: Pennsylvania State University Press. Hemmings, C. 2005. Telling Feminist Stories. Feminist Theory 6 (2): 115–139. Johnstone, H.W., Jr. 1978. Validity and Rhetoric in Philosophical Argument. University Park, PA: Dialogue Press of Man & World. Ladyman, J., D. Ross, D. Spurrett, and J. Collier. 2007. Every Thing Must Go: Metaphysics Naturalized. Oxford: Oxford University Press. Leiter, B. 2013. Poll Results: Which are the ‘best’ (Highest Quality) Philosophy Journals (Regardless of Area)? In Condorcet Internet Voting Service. http://www. cs.cornell.edu/w8/~andru/cgi-­perl/civs/results.pl?id=E_46442506716654dd. Accessed 17 June 2015. McBurney, P., and S. Parsons. 2001. Chance Discovery Using Dialectical Argumentation. In New Frontiers in Artificial Intelligence, eds. T. Terano, Y. Ohsawa, T. Nishida, A. Namatame, S. Tsumoto, and T. Washio, 414–424. Berlin: Springer. Papineau, D. 2006. The Tyranny of Common Sense. The Philosophers’ Magazine 34: 19–25. ———. 2015. Naturalism. In The Stanford Encyclopedia of Philosophy, Fall 2015 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/fall2015/entries/ naturalism/. Rescorla, M. 2009a. Shifting the Burden of Proof? Philosophical Quarterly 59 (234): 86–109.

34 

C. Rhode

———. 2009b. Epistemic and Dialectical Regress. Australasian Journal of Philosophy 87 (1): 43–60. Rhode, C. 2017. The Burden of Proof in Philosophical Persuasion Dialogue. Argumentation 31 (3): 535–554. Saunders, B., J. Sim, T. Kingstone, S. Baker, J. Waterfield, B. Bartlam, H. Burroughs, and C. Jinks. 2018. Saturation in Qualitative Research: Exploring its Conceptualization and Operationalization. Quality & Quantity 52 (4): 1893–1907. Scanlon, T.M. 2003. Rawls on Justification. In The Cambridge Companion to Rawls, ed. S. Freeman, 139–167. Cambridge: Cambridge University Press. Schukraft, J. 2017. Is Justified True Belief Knowledge? London: Macat International. Walton, D. 1992. After Analytic Philosophy, What’s Next?: An Analytic Philosopher’s Perspective. Journal of Speculative Philosophy 6 (2): 123–142. ———. 2013. Methods of Argumentation. Cambridge: Cambridge University Press. ———. 2014. Burden of Proof, Presumption, and Argumentation. Cambridge: Cambridge University Press. ———. 2016. Interview with Douglas Walton. Interviewed by C. Rhode. The Reasoner 10 (8): 64–68. Walton, D., and E.C.W. Krabbe. 1995. Commitment in Dialogue. Albany, NY: SUNY Press. Williamson, T. 2000. Knowledge and its Limits. Oxford: Oxford University Press. Willoughby, M. 2012. Stoppism: Retrospects and Prospects. Journal of Philosophy 109 (4): 281–294.

2 The Burden of Proof

2.1 Egalitarianism versus Foundationalism According to dialogical egalitarianism, any proposition asserted in dialogue, if questioned, must be supported or else retracted.1 Hence, the burden of proof for any asserted and questioned proposition lies upon the asserting party. According to dialogical foundationalism, at least some propositions are privileged over this burden, such that the asserting party need only support the proposition if and when the questioning party has defended her challenge by providing support for the proposition’s negation.2 In effect, if you assert some putatively privileged proposition P, such as The sky is blue by day, and another party to the dialogue questions whether P (without providing support for ¬P), then you are thus required to defend your commitment to P under dialogical egalitarianism but not under dialogical foundationalism. If, however, the proposition P that you assert is not putatively privileged and is being questioned (without provision of support for ¬P), then you are required to defend your  Note that criteria of sufficiency of such support are extraneous to the present considerations.  These competing theses have first been distinguished by Michael Rescorla (2009a).

1 2

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_2

35

36 

C. Rhode

commitment to P under either thesis. For an exact definition of the privileged status of P, assume that party a to dialogue δ commits to P in δ (e.g. by asserting that P) and party b to δ commits to neither P nor ¬P in δ (e.g. by questioning whether P, without providing support for ¬P). Then P is privileged just in case it is true that: a is required to provide support for or to retract her commitment to P in δ if, and only if, b provides support for ¬P in δ.3 Thus, dialogical foundationalism affirms and dialogical egalitarianism rejects the existence of privileged propositions and of the resultant uneven allocation of the burden of proof. Though rarely made explicit in substantive analytic philosophical debates, dialogical foundationalism has surfaced particularly in more recent disputes in philosophical methodology. Its ascription of privileged status is exemplified in Roderick Chisholm’s contention that ‘[a]ny philosophical theory which is inconsistent with any of [the propositions of common sense] is prima facie suspect. The burden of proof will be upon the man [sic] who accepts any such theory’ (1976: 18). Chisholm’s advocacy of privileged status for common-sense propositions is echoed in George Bealer’s aforementioned remark that ‘[u]nless and until a reason for departing from this standard practice is produced, we are entitled— indeed, obligated—to continue using intuitions as evidence’ (1996: 30 n. 15; emphasis added). Timothy Williamson holds the closely related view that ‘it would be foolish to change a well-established methodology [like philosophers’ use of thought experiments] without serious evidence that doing so would make the discipline better rather than worse’ (2011: 217). Chisholm, then, claims privileged status for any common-sense proposition, while Bealer and Williamson seem to claim privileged status for any proposition that sanctions some ‘standard’ or ‘well-established’ evidential procedure, including in particular the procedures of evidentially employing intuitions or thought experiments.

 Propositions that, due to their privilege, are in no need of support in δ unless and until support for their negation is provided in δ should not be confused with propositions that, due to support already provided in δ, are in no need of further support in δ unless and until support for their negation is provided in δ. Only the former status depends upon dialogical foundationalism. 3

2  The Burden of Proof 

37

These remarks by Chisholm, Bealer, and Williamson illustrate both dialogical foundationalism and methodological conservatism,4 so it is worth noting the relation between these theses. Dialogical f­ oundationalism states that some assertions, if questioned, need not be supported unless and until support for their negation is provided. Methodological conservatism states that assertions of the appropriateness or suitability of (certain) standard philosophical methods need not be supported unless and until support for their negation is provided.5 As such, methodological conservatism is an instance of dialogical foundationalism. Indeed, methodological conservatism is clearly incompatible with dialogical egalitarianism, which would require that assertions of the appropriateness of standard philosophical methods (just like any other assertion) be supported when questioned. By contrast, methodological conservatism is entirely compatible with dialogical foundationalism, which permits treating assertions of such appropriateness as privileged and in no need of support in the absence of support for their negation. Thus, methodological conservatism is logically dependent upon dialogical foundationalism, thereby presupposing either the truth or, minimally, the adoption of dialogical foundationalism. I will argue in the present chapter, however, that philosophers are prudently required to adopt dialogical egalitarianism, with the notable implication that methodological conservatism must prudently be rejected: The burdens of proof recently placed upon the employment of common sense, intuitions and thought experiments (and other standard methods) in philosophical dialogue cannot be evaded and must instead be met. To this end, I will discuss existing arguments in favour of either dialogical foundationalism or dialogical egalitarianism (in Sect. 2.2), rejecting each  It is not necessary for present purposes to claim that Chisholm, Bealer, or Williamson subscribe to either of these two theses, as their remarks are cited principally for their illustrative value. 5  Methodological conservatism thus characterised should not be confused with epistemic conservatism, which states, effectively, that the fact of one’s believing that P provides one with reason for believing that P (cf. e.g. Daly 2015: 13). Epistemic conservatism thus differs from methodological conservatism in three significant respects: It concerns (a) attitudinal rather than dialogical commitments, (b) the question of how a given (attitudinal) commitment is or can be supported rather than whether a given (dialogical) commitment need be supported, and (c) any (attitudinal) commitment at all rather than only (dialogical) commitments regarding the appropriateness of standard philosophical methods. 4

38 

C. Rhode

one of them, principally on the basis of the attitude/dialogue separation and the goal of persuasion. That very same goal, I then argue (in Sect. 2.3), renders the adoption of dialogical egalitarianism in analytic philosophy prudently inescapable.

2.2 Allocating the Burden of Proof Variations of three arguments—all in their basic form directed against dialogical egalitarianism—have found employment in the debate between advocates of dialogical foundationalism and dialogical egalitarianism in philosophy. These three arguments invoke, respectively, practical need, the threat of scepticism, and putatively foundational propositions. I will discuss them in turn. The assessment of alternative courses of action is frequently based on incomplete information, yet a decision between the alternatives must nonetheless be taken. In a criminal trial, deciding whether and how to punish a defendant cannot be postponed until sufficient evidence is available for the conclusive proof of guilt or innocence, not least since the potentially life-long waiting time (in custody or freedom) may itself constitute punishment. Instead, decision-making to guide action is often based on stipulated default outcomes, and thus on presumptions, placing the burden of proof, not upon the party in a dialogue that asserts, but upon the party that remains non-committed towards or denies the proposition expressing the default outcome. In criminal trials, this is exemplified by the presumption of innocence, which, for the deliberation of a verdict and potential sentence, prescribes the default outcome of acquittal and non-punishment. Practical need thus seems to give rise to dialogical foundationalism in criminal legal proceedings, as well as in a good many other contexts. Unless philosophy is relevantly disanalogous, dialogical foundationalism would be equally warranted in philosophy. Yet philosophy is indeed relevantly disanalogous. The failure to resolve a philosophical question is of scarce import to any real-life and thus relatively urgent problem. As Brian Treanor (2006) aptly remarked, ‘I’m a professor of philosophy, not a cardiac surgeon. How urgent can it be?’ Nothing of non-academic value will be jeopardised by failure to answer a

2  The Burden of Proof 

39

philosophical question; hence practical need does not warrant the adoption of presumptions in order to preclude such failure in philosophical dialogue (cf. Sander 2003: 84; Dyke 2013: 174). Though there may certainly be benefits from answering a philosophical question, there is no practical need for an answer (despite grant applications sometimes suggesting otherwise). This disanalogy with dialogues guided by practical need may be clarified further by attending to their respective types of dialogue. Legal and similar action-directed dialogues would arguably typically instantiate deliberation dialogue, with a persuasion dialogue embedded within said deliberation dialogue (cf. Hahn and Oaksford 2007; Walton 2014: 211–44). The conduct of such an embedded persuasion dialogue would be governed and indeed curtailed by the pursuit of the fundamental goal of the non-embedded deliberation dialogue, including the need to reach a timely decision for implementation, thus giving rise to the need for presumptions. By contrast, philosophical dialogues are not typically embedded within a wider deliberation dialogue (or dialogue of any other type) whose fundamental goal might warrant the adoption of presumptions or otherwise curtail the conduct of philosophical dialogues. Perhaps the adoption of presumptions in philosophical dialogue may instead be warranted by cognitive need, that is, the need for discovering truths or generating knowledge. This warrant would suggest that a presumption is functionally equivalent to a key thesis accepted within a Kuhnian paradigm (Kuhn 1962). Indeed, according to Neil Levy (2003), the work of Frege and Russell is treated as the defining exemplar of analytic philosophy, or as the paradigm in a narrower sense of that word (cf. Godfrey-Smith 2003: 77), while each sub-discipline or research programme within analytic philosophy is characterised additionally by its own exemplars. Exemplars provide a field of research with its key methodological and ontological theses that underlie and significantly facilitate all subsequent work within that field. Such key theses effectively function as presumptions, whose unquestioned acceptance in a given field is warranted at least by the need for generating knowledge. As such, cognitive need may seem to warrant the adoption of presumptions, and thus of dialogical foundationalism, in analytic philosophical dialogue.

40 

C. Rhode

That, however, is not the case, for the following reason. If knowledge as generated in analytic philosophy was regarded as an end in itself, then it is unclear why there would be any need for the attainment of that end. That is, it is unclear why the generation of philosophical knowledge or the discovery of philosophical truth would be needed. If, on the other hand, analytic philosophical knowledge was regarded as a means to some other end (akin to medical or climatological knowledge), then this argument from cognitive need would collapse into the above argument from practical need. Moreover, we have already seen (in Chap. 1, Sect. 1.3) that analytic philosophical dialogue is not in fact aimed at the generation of knowledge, and would thus in any case fail to satisfy any cognitive need. As such, the adoption of presumptions in current analytic philosophical dialogue is not warranted by cognitive need, nor, as discussed above, is it warranted by practical need. These putative needs, however, also seem to give rise to a further consideration in favour of dialogical foundationalism. In a great many dialogues, some propositions or other are presumed true due to the subject of the dialogue itself, such that the utility of continuing a given dialogue would be undermined if it turned out that the parties to the dialogue do not share some key presumption. For instance, there is little point in continuing a negotiation dialogue about a pay rise if it turns out that the employee is actually using the occasion to resign, as they have had enough of the firm’s persistent toleration of sexism. The presumption that the employee will remain at the firm if they are just paid enough thus turns out not to be shared. Similarly, a party to a dialogue about the exact half-­ life of the isotope tellurium-128 will quickly consider discontinuing the dialogue if they discover that the opposing party does not share the presumption that tellurium-128 is radioactive.6 Leaving certain presumptions unquestioned, it seems, is necessary for the very conduct of many dialogues. If the same holds true with regard to philosophical dialogues, then dialogical foundationalism must practically be accepted in order to conduct philosophical dialogues at all.  Tellurium-128 admittedly has the longest calculated half-life of any radioactive isotope, at 2.2 × 1024 years, or approximately the current age of the universe multiplied by 160 trillion (Emsley 2011: 532). 6

2  The Burden of Proof 

41

A notable feature of philosophical dialogues, however, is their potential longevity. These persuasion dialogues are conducted, as Jonathan Cohen noted (1986: 3), ‘through minutes, months, years, or even centuries’, ‘with one contributor taking over from another’. If one’s persuasive success in such a dialogue is built upon certain shared presumptions, then that success will be dependent upon the continued sharing of these presumptions. It appears exceedingly optimistic, however, to rely on the faint chance that all such presumptions are shared and accepted, not just by the contemporary members of one’s opposing party, but also by any future members of that opposing party, down the centuries. Yet if such presumptions do not continue to be shared, then one does not actually succeed in persuading one’s opposing party—that is, one does not actually succeed in the given philosophical persuasion dialogue. In effect, building one’s apparent persuasive success upon the faint chance that certain shared presumptions remain unquestioned is highly inadvisable—as is, therefore, the very employment of such presumptions. If anything, then, this consideration would count in favour of dialogical egalitarianism. Alternatively, one might merely seek to persuade one’s opposing party of a conditional proposition, whereby that conditional proposition features as its antecedent a conjunction of presumptions. Perhaps it could be argued in this way, somehow, that the need for presumptions in such persuasion dialogue supports dialogical foundationalism? Alas, it cannot be, since acceptance of such antecedent presumptions would not be necessary for engagement in the given persuasion dialogue about such a conditional proposition. One can easily entertain certain presumptions, and on that basis engage in persuasion dialogue about a given conditional proposition, without thereby accepting these presumptions. If acceptance of the given presumptions is unnecessary, however, then this consideration regarding conditional propositions in persuasion dialogue does not lend any support to dialogical foundationalism (nor to dialogical egalitarianism). The possibility of merely seeking to persuade one’s opponent of a conditional proposition is thus irrelevant here (though I will return to the same in Chap. 6, Sect. 6.2). A more popular argument for dialogical foundationalism holds that dialogical egalitarianism implies global epistemological scepticism and

42 

C. Rhode

must thus be rejected (Rescorla 2009b: 48). According to this argument, egalitarianism requires us to provide support for every single proposition whose truth we commit to in dialogue. Yet this would also include any proposition employed in support of another proposition, and hence this requirement would apparently launch an infinite regress of supporting propositions. Now, it is well-known, the argument goes, that an infinite regress would leave us unable to justify any proposition at all. Being thus unable to justify any proposition whatsoever, we cannot justifiably believe any proposition to be true. That, however, amounts to an endorsement of global epistemological scepticism, which must obviously be rejected, along with any thesis that leads to such scepticism. Hence, one might claim, dialogical egalitarianism must be rejected. Though this argument may appear prima facie plausible, it fails in at least three respects. First, it presupposes that global scepticism is unacceptable, thus begging the question against those dialogical egalitarians who are also global sceptics, viz. the Pyrrhonians. (In pursuit of the dialogical goal of persuading one’s opponent, begging the question against that opponent will of course be persuasively useless.) Second, the argument misrepresents dialogical egalitarianism, which states that the requirement to support (or else retract) a dialogical commitment is conditional upon that commitment’s being challenged by another dialogue party, rather than unconditional as the argument not only suggests but indeed requires. The alleged regress of supporting propositions would thus only ensue if one was faced with a persistent interlocutor (Rescorla 2009b: 54–6). Yet, third, such a regress of supporting propositions cannot lead to global scepticism, due to the separation of dialogical from attitudinal commitments established above. Even if I find myself unable to establish P to a persistent interlocutor, this does not imply that I take an agnostic attitude towards P, attitudinally committing to neither P nor ¬P. It might instead be held to imply that I ought to commit in attitude to neither P nor ¬P, due to some prescription to the effect that one’s inability to establish P to someone else renders it impermissible to believe P oneself. But it is far from obvious why dialogical egalitarians (or anyone else, for that matter) should accept such a prescription. Even if such a prescription was insisted to be among the rules of the game of analytic philosophy, it is open to dialogical egalitarians (and others) to favour a

2  The Burden of Proof 

43

different characterisation of analytic philosophy, thus rendering such insistence both arbitrary and persuasively useless. Dialogical egalitarianism thus does not lead to global scepticism. A parallel, exclusively dialogical argument holds that dialogical egalitarianism leaves us unable to establish the truth of any proposition in dialogue if faced with a persistent interlocutor. In that event, one cannot escape the ensuing dialogical regress in any way other than by abandoning the attempt to establish the given proposition to one’s interlocutor. This argument, however, fails by begging the question, since dialogical egalitarians regard the inability to establish a proposition to a persistent interlocutor as unproblematic. Few interlocutors are actually persistent in their challenges, and those few merely demonstrate ‘the unfortunate fact that there is simply no reasoning with some people’ (Rescorla 2009b: 47). For dialogical foundationalists, by contrast, it is unacceptable that one’s attempt to establish to an interlocutor the truth of any proposition at all may falter at the whim and mercy of that interlocutor. Yet dialogical foundationalists cannot invoke this unacceptability in an argument against dialogical egalitarianism, on pain of begging the question. We can thus conclude that arguments against dialogical egalitarianism from the threat of scepticism and from a parallel, exclusively dialogical threat fail as well. The last argument against dialogical egalitarianism invokes propositions ostensibly so basic that they should be accepted by default when asserted, placing the burden of proof upon anyone attempting to detract from dialogical commitment to their truth. Effectively the same eight types of such putatively foundational propositions have been catalogued by Michael Rescorla (2009a: 89) and (as a taxonomy of Wittgensteinian hinges) by Danièle Moyal-Sharrock (2005: 102–3). These include (with examples in parentheses): introspective reports (I feel happy); perceptual reports (I see a ship7); autobiographical reports (I have never been to Tasmania); reports of one’s immediate environment (The armchair in front of me is burning); reports of knowledge that is basic within one’s  The truth of such putatively privileged propositions is irrelevant here, as is, therefore, the possibility that I am not seeing a ship but rather a crashed UFO cleverly disguised as a ship by the Fake Barn County Intelligence Agency. 7

44 

C. Rhode

specifically located community (The Earth is flat); reports of knowledge that is basic in all communities (The sky is blue by day); basic conceptual truths (All kangaroos are animals); and basic logical and mathematical truths (1–1 = 0).8 According to the present argument, asserted ­propositions of any of these types must obviously be privileged, as it would seem confused, misguided, or in some other way inappropriate for one’s interlocutor to demand support for the asserted proposition. Equally, then, it would seem counter-intuitive to accept a thesis that permits such preposterous questioning, and dialogical egalitarianism should therefore be rejected (cf. Rescorla 2009a). Again, however, this argument begs the question against dialogical egalitarians, who reject the putative privilege of these propositions. The charge that dialogical egalitarians fail to recognise the inappropriateness of questioning these propositions absent support for their negation is plainly circular, as they deny that very inappropriateness. It remains circular even with regard to what may be considered the most obviously foundational propositions, viz. basic conceptual, logical, and mathematical truths. Take the mathematical truth that 1–1 = 0. Dialogical egalitarians would deny the impermissibility of questioning its assertion, and would consequently accept the burden of supporting it if needed. To support it, they might cast 1–1 = 0 as the symbolisation in Arabic numerals of the proposition that One, less one, is none. And of course, that proposition might in turn be questioned. Since dialogical egalitarians do not regard it as problematic that such putatively foundational propositions as 1–1 = 0 can be questioned in the absence of support for their negation, assuming that it is problematic simply begs the question against dialogical egalitarianism. Indeed, dialogical egalitarians can turn the argument on its head, by charging its foundationalist proponents with sheer dogmatic insistence on propositions of the foregoing eight types (or else on their privileged status; cf. Rescorla 2009b: 46, 51). On the assumption that dogmatism is in some sense inappropriate, dialogical foundationalists must avoid  One might argue that such basic conceptual, logical, and mathematical truths are actually part of community-dependent basic knowledge (as exemplified by 1–1 = 0), rather than being types of putatively foundational propositions in their own right. But such a classificatory difference is of no present relevance. 8

2  The Burden of Proof 

45

dogmatically insisting on propositions of the above types. Yet that assumption—that dogmatism about (the privileged status of ) a proposition is inappropriate—would in turn be denied by dialogical foundationalists. Indeed, such dogmatism is precisely stipulated by dialogical foundationalism, as the latter states in effect that we may permissibly be dogmatic about some particular propositions when these propositions are questioned without support for their negations. Turning the preceding argument against dialogical foundationalism instead would thus be equally question-begging.9 Whether directed against dialogical egalitarianism or dialogical foundationalism, then, the argument from putatively foundational propositions offers no support for either thesis. A potential variation on this argument would substitute putatively foundational propositions by self-referential propositions that are proved by the very performance of uttering or writing them—as outlined by Edward Lemmon (1962; cf. Hintikka 1962: 7–9; Apel 2001: 43–5; Hedberg 2005: 83). Consider, for instance, sentences expressing a truth-­ value commitment to a proposition where the proposition itself describes the making of that same truth-value commitment, such as ‘It is the case that I am asserting.’. Utterance or writing of this sentence provides sufficient support for the proposition at its end (marked in italics). The same applies to the sentences ‘Is it the case that I am asking?’ and ‘It is not the case that I am denying.’. Similar propositions performatively proved in dialogue derive from indispensable aspects of dialogue itself, including ‘The present dialogue is being conducted between a plurality of interlocutors.’ (cf. ‘διά’: between) and ‘The present dialogue is being conducted in a language.’ (cf. ‘λογος’: speech, language). An indefinite number of similar sentences can be constructed from a truth-value commitment to a proposition which itself describes the making of that same truth-value commitment within a specified medium (such as language). Those sentences include, for instance, ‘It is the case that I am asserting in an academic work.’ and ‘It is not the case that I am denying in written English.’. Surely, one might hold, these self-referentially proved propositions must be privileged in dialogue, as it would seem nothing short of contradictory to  The charge of dogmatism is sometimes regarded as distinct from the charge of arbitrariness concerning the selection of propositions to be privileged. However, as should be clear, the latter charge is as question-begging as the former charge. 9

46 

C. Rhode

question a proposition that is self-proving in dialogue. In consequence, there exist privileged propositions, and dialogical egalitarianism must be rejected. The flaw in this argument concerns the claim to privilege. As defined in Sect. 2.1, privilege amounts to an initial exemption from the burden of proof. Yet a performatively self-proving proposition already provides conclusive proof of its truth, with no need for any exemption from an already satisfied burden of proof that might be imposed. The pre-emptive satisfaction of any potential burden of proof thus does not require any privileged status. Without the need for privilege, the foregoing performatively self-proving propositions are thus equally compatible with dialogical foundationalism and dialogical egalitarianism, thereby failing to provide support for either thesis. Overall, the relative predominance of arguments against dialogical egalitarianism may suggest that dialogical egalitarianism is assumed to hold by default, while dialogical foundationalism would need to be established against it. Beyond the appeal to the impermissibility of dogmatism that has already been rejected, this apparent default status of dialogical egalitarianism may also be due to an appeal to the theoretical virtue of simplicity, as arguably instantiated to a higher degree by dialogical egalitarianism. Yet such an appeal would again be question-begging, as dialogical foundationalists may simply attach more weight to the theoretical virtue of conservatism (which ostensibly favours dialogical foundationalism). Moreover, any unsupported claim in the face of challenges that dialogical egalitarianism—of all theses!—should be accepted by default would be patently self-undermining. Dialogical foundationalists might correspondingly contend that the thesis of dialogical foundationalism itself is among the propositions that, according to dialogical foundationalism, should be accepted by default; but that contention would be just as obviously question-begging. We are thus left without any conclusive argument in favour of either dialogical egalitarianism or dialogical foundationalism, and even worse, without any default position to fall back upon. When engaging in philosophical dialogue, however, one cannot but adopt one or the other thesis at least implicitly and in practice. A resolution of the debate is therefore required, not least if we wish to avoid

2  The Burden of Proof 

47

perennially returning to this debate when discussing more substantive issues.10 Yet absent a successful argument for either thesis, we can only resort to deliberating about which of the two theses it is instrumentally more rational, that is, prudent, to accept, employ and presume in practice in our philosophical dialogues. In place of principled arguments with unrestricted scope of application, we must turn to pragmatic considerations about philosophical dialogue, thereby restricting the scope of application of any subsequent conclusion about the competing theses to analytic philosophy only.

2.3 The Prudence of Dialogical Egalitarianism in Philosophy In order to determine which of the two competing theses it is instrumentally more rational to adopt in philosophical dialogues, we need to know the goal to the attainment of which one or the other thesis would be so instrumental. That is, in order to determine whether it is prudent to adopt dialogical egalitarianism or dialogical foundationalism, we must know the goal of analytic philosophical dialogue. As shown in Chap. 1, Sect. 1.3, for all practical purposes, dialogue in contemporary analytic philosophy is essentially characterised by the goal of each participant with an initial dialogical truth-value commitment regarding a given proposition P to influence her opponent by way of argument to the effect that the opponent abandons his initial dialogical truth-value commitment regarding P and instead adopts her initial dialogical truth-value commitment regarding P.  As such, it is the essential dialogical goal of analytic philosophical dialogues to persuade, that is, defeat, one’s opponent in argument, and not be defeated oneself. Given this goal, we need to ask whether the adoption in analytic philosophical dialogues of either dialogical egalitarianism or dialogical foundationalism is more conducive to the persuasion of one’s opposing party.  For examples of more substantive discussions in philosophy turning towards this dialogical question, see Chisholm (1976), Bealer (1996) and Williamson (2011) above, as well as Aijaz, McKeown-­ Green and Webster (2013: 260). 10

48 

C. Rhode

The answer to this question is easily found. Egalitarianism and foundationalism differ with regard to the putatively privileged status of certain ostensibly foundational propositions. Take, then, any such putatively privileged proposition P. Imagine you dialogically commit to P, and your interlocutor questions whether P (committing dialogically to neither P nor ¬P) without presenting any support for ¬P. Which reaction on your part is more likely to lead to your interlocutor’s also dialogically committing to P? That is, which reaction on your part is less likely to be detrimental to your attempt to persuade your opposing party to the ongoing dialogue that P? Is it, either, your insisting that they should not question whether P without sufficiently supporting ¬P; that they should simply accept that P absent such support; and that they should only expect you to offer support for P after they have provided sufficient support for ¬P? Or is it, alternatively, your offering support for P? The prudence of the latter reaction is obvious: Insisting on the privileged status of P, as in the former reaction, is useless (or positively counter-productive) as a means of persuading the members of the opposing party who have already questioned whether P. That is, insisting on the putative privilege of P is simply ineffective for the purpose of succeeding against one’s opponents in philosophical dialogue. Such privilege is, therefore, practically non-existent, whereupon dialogical foundationalism collapses, for practical purposes, into dialogical egalitarianism in philosophical dialogue. Prudently, then, the adoption of dialogical egalitarianism in analytic philosophical dialogues is inescapable: If one has committed to P in such a dialogue, and if P is subsequently questioned by another party, then the burden of proof thus imposed prudently cannot be evaded. Instead, one’s commitment to P must prudently be supported, regardless of the putative privilege of P. Dialogical egalitarianism, however, raises a seemingly lethal problem for the foregoing considerations upon which it has been supported. For, given dialogical egalitarianism, even the presumptions that underlie the considerations raised up to this point may be questioned, including in particular the deductive and other inferential rules that I relied upon. In the event that all rules of inference are indeed questioned, any argument in defence of such questioned rules would seem unavailable, since an argument must itself employ such a rule in order to infer a conclusion

2  The Burden of Proof 

49

from a given set of premises. Nor would claims to the privileged status of these inference rules seem available, because such claims would apparently contravene the dialogical egalitarianism advocated in this chapter. Hence, a challenge to all rules of inference, it seems, could neither be answered nor could it be deflected here, thus apparently undermining the foregoing considerations in favour of dialogical egalitarianism on the basis of these same considerations. Of course, such a challenge would equally undermine any argument for dialogical foundationalism in philosophical persuasion dialogues. Aside from that little problem, however, this challenge would also misconstrue the nature and conclusion of the considerations offered thus far. These considerations have been premised explicitly on the overarching dialogical goal of persuading one’s opponent, so they do not apply to dialogues that follow a different overall goal, such as persuasion dialogues embedded within a dialogue of a different type. That is to say, my present considerations are applicable only to non-embedded persuasion dialogues in philosophy. Yet no such non-embedded persuasion dialogue is to be found in the present work. The overarching dialogical goal of my considerations throughout this book is not one of persuasion but rather of deliberation: As already indicated in Chap. 1, Sect. 1.4, my considerations offer advice on how best (or most prudently) to attain a given goal. They do not consider the normative or evaluative status of that or any other goal, nor of any course of action in pursuit of the given goal. Rather, they merely seek to aid in deciding upon the most effective or instrumentally rational course of action that will be taken in pursuit of this given goal. As such, my present considerations quite clearly instantiate, and are designed to instantiate, deliberation dialogue.11 Though Sect. 2.2 above is plausibly characterised as a contribution to a persuasion dialogue, this contribution is itself embedded within the overarching deliberation dialogue of this chapter and of the present book more generally. Now, since my considerations recommend dialogical egalitarianism for non-­ embedded philosophical persuasion dialogues, they do not apply to the non-embedded deliberation dialogue of this work, and hence do not here

11

 Cf. fn. 24 in Chap. 1, Sect. 1.3.

50 

C. Rhode

require me to adopt dialogical egalitarianism.12 Consequently, my foregoing considerations are not rendered self-undermining by a challenge to all rules of inference. Let me thus turn to the consequences of the adoption of dialogical egalitarianism in analytic philosophical dialogues. The prudential adoption of dialogical egalitarianism is also the prudential rejection of dialogical foundationalism. Seeing as methodological conservatism presumes dialogical foundationalism, the rejection of dialogical foundationalism unsurprisingly implies the rejection of methodological conservatism. With this rejection comes the rejection of any privileged status of propositions affirming the methodological appropriateness of common sense, intuitions, thought experiments, or other standard philosophical methods. Pace Chisholm’s, Bealer’s and Williamson’s comments cited above, such propositions are not privileged and, having recently faced ample questioning, must prudently be supported. The next question to be addressed, then, is how these and other propositions can be supported. Since the burden of proof imposed upon a proposition questioned in philosophical dialogue cannot be evaded, the next question is how it can be met. Options for meeting the burden of proof will be examined in the following chapter, before I return to philosophical methods in Chap. 4.

References Aijaz, I., J. McKeown-Green, and A. Webster. 2013. Burdens of Proof and the Case for Unevenness. Argumentation 27: 259–282. Apel, K.-O. 2001. The Response of Discourse Ethics to the Moral Challenge of the Human Situation as Such and Especially Today. Leuven: Peeters. Bealer, G. 1996. On the Possibility of Philosophical Knowledge. Philosophical Perspectives 10: 1–34.  This is not to say that it would be prudent (or in any other way more appropriate than otherwise) to adopt dialogical foundationalism in non-embedded philosophical deliberation dialogue. Though it is of no present import, neither dialogical egalitarianism nor dialogical foundationalism is applicable to deliberation dialogue, as there does not exist any burden of proof in such dialogue (cf. Walton 2014: 232–6). 12

2  The Burden of Proof 

51

Chisholm, R.M. 1976. Person and Object: A Metaphysical Study. La Salle, IL: Open Court. Cohen, L.J. 1986. The Dialogue of Reason: An Analysis of Analytical Philosophy. Oxford: Clarendon Press. Daly, C. 2015. Introduction and Historical Overview. In The Palgrave Handbook of Philosophical Methods, ed. C. Daly, 1–30. Basingstoke: Palgrave Macmillan. Dyke, H. 2013. On Methodology in the Metaphysics of Time. In The Future of the Philosophy of Time, ed. A. Bardon, 169–187. New York: Routledge. Emsley, J. 2011. Nature’s Building Blocks. 2nd ed. Oxford: Oxford University Press. Godfrey-Smith, P. 2003. Theory and Reality: An Introduction to the Philosophy of Science. Chicago: University of Chicago Press. Hahn, U., and M.  Oaksford. 2007. The Burden of Proof and its Role in Argumentation. Argumentation 21 (1): 39–61. Hedberg, P. 2005. What are Performative Self-contradictions? Sats—Nordic Journal of Philosophy 6 (1): 66–91. Hintikka, J. 1962. Cogito, Ergo Sum: Inference or Performance? Philosophical Review 71 (1): 3–32. Kuhn, T.S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press. Lemmon, E.J. 1962. On Sentences Verifiable by Their Use. Analysis 22 (4): 86–89. Levy, N. 2003. Analytic and Continental Philosophy: Explaining the Differences. Metaphilosophy 34 (3): 284–304. Moyal-Sharrock, D. 2005. Understanding Wittgenstein’s On Certainty. Basingstoke: Palgrave Macmillan. Rescorla, M. 2009a. Shifting the Burden of Proof? Philosophical Quarterly 59 (234): 86–109. ———. 2009b. Epistemic and Dialectical Regress. Australasian Journal of Philosophy 87 (1): 43–60. Sander, T. 2003. Beweislastverteilung und Intuitionen in philosophischen Diskursen. Journal for General Philosophy of Science 34 (1): 69–97. Treanor, B. 2006. Slow University: A Manifesto. http://faculty.lmu.edu/briantreanor/slow-­university-­a-­manifesto/. Accessed 17 June 2015. Walton, D. 2014. Burden of Proof, Presumption, and Argumentation. Cambridge: Cambridge University Press. Williamson, T. 2011. Philosophical Expertise and the Burden of Proof. Metaphilosophy 42 (3): 215–229.

3 Evidence, Inference, and Empiricism

3.1 Meeting the Burden of Proof In the previous chapter, I argued that a proposition asserted in philosophical dialogue, if questioned, must prudently be supported, and that this burden of proof cannot be evaded by claiming the asserted proposition to be privileged. As the burden of proof prudently cannot be evaded, it must instead be met. That is, the asserting party must provide support for the questioned proposition. The purpose of the present chapter is to examine the required support more closely. I will principally discuss here what conditions must prudently be satisfied by such support in order to help persuade one’s opponent of a questioned proposition. I will not, however, seek to identify conditions whose satisfaction is singly or jointly sufficient for attaining this dialogical goal. The identification of sufficient conditions for persuading an opponent is both unnecessary for the purposes of this work and, I suspect, impossible, owed to the obvious fact that an opponent may or may not elect to challenge the support offered.1  Performatively self-proving exceptions to this impossibility have been indicated in Chap. 2, Sect. 2.2 above, though such exceptions still do not provide a basis for deducing propositions beyond those included among these exceptions. 1

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_3

53

54 

C. Rhode

Rather, I merely seek to identify certain prudently necessary conditions whose non-satisfaction is perfectly detrimental to the attainment of the goal of persuading a philosophical opponent. In order to support a questioned proposition, it would not suffice to invoke some piece of evidence. One must also claim, at least implicitly, that the invoked piece of evidence is evidence in favour of the questioned proposition. Consequently, any support provided for a proposition must consist of two things: the item of evidence, and the evidential relation linking that item to the proposition to be supported. In Sect. 3.2, I will focus on items of evidence themselves, explaining not least that my subsequent considerations are independent of any thesis concerning their nature. In Sects. 3.3-3.5, I then turn to the evidential relation, which I will discuss in relation to inference barriers (such as Hume’s is/ought separation). I will argue there that philosophers must prudently avoid the employment of any support for a challenged proposition if the support to be employed violates such an inference barrier. In the final section of this chapter, I will then conjoin this requirement with the dialogical egalitarianism established above, to yield a form of empiricism that I call dialogical empiricism.

3.2 Evidence A colourful variety of theses regarding the nature of evidence has been postulated by analytic philosophers over the course of the last century. Evidence has variously been argued to consist in sense data, observation statements, sensory stimulations, propositions, mental states, and several other kinds of entities (cf. Kelly 2014). In each case, accounts of the nature of evidence are embedded within and to varying degrees integral to wider epistemological frameworks. The thesis that evidence consists in sense data, for instance, was crucial to Bertrand Russell’s empiricist attack on British idealism, just as observation statements were central to logical positivism. W. V. O. Quine’s account of evidence as the stimulation of sensory receptors played no small role in his own empiricism, while the ideas that evidence consists in the totality of one’s known propositions or in the totality of one’s current mental states are closely linked to,

3  Evidence, Inference, and Empiricism 

55

respectively, epistemological primitivism and evidentialism. Fascinating as these various epistemological frameworks may be, the present considerations are unaffected by them or their accounts of the nature of evidence. My concern is not with evidence as it informs beliefs, but with evidence as it is used in dialogue. And given the separation between commitments in attitude and in dialogue, epistemological (and hence attitudinal) theses regarding evidence have no implications for theses regarding evidence in dialogue. As will become clear below, epistemological theses about evidence are therefore effectively irrelevant to theses about the employment of evidence in dialogue. What, then, can be said about evidence in dialogue? This much at least: In dialogue, evidence will only serve to support a challenged proposition to one’s interlocutor if what is invoked as evidence is communicated to one’s interlocutor. The need for communication in dialogue implies the need for assertion (i.e. for positive truth-value commitment).2 When we invoke a piece of evidence in a dialogue, we cannot but make an assertion regarding that item of evidence, as the item invoked will only serve as evidence to the extent that one dialogically commits to a proposition (or a statement) expressing the same. So it is a precondition for the employment of some given evidence in dialogue that a proposition expressing or describing it be asserted. Note that the possibility of a proposition’s describing or expressing some piece of evidence is independent of the nature of that evidence, which may itself be a sense datum, a proposition, or any other entity or event in the broadest sense. The exact nature of evidence is thus irrelevant to the employment of evidence in dialogue. The proposition expressing or describing some given evidence, call it E, need not describe a given item of evidence in its entirety. E need only express the relevant aspects of a given piece of evidence when E is asserted in support of some challenged proposition. In order to support the claim that prime numbers do not occur entirely at random, for instance, it is necessary to assert the existence of at least one pattern among prime numbers, but not to specify exhaustively every prime number matching that pattern. Similarly, the exact typeface used in the production of a  Cf. fn. 3 in Chap. 1, Sect. 1.2.

2

56 

C. Rhode

bank statement is an aspect of the item of evidence that is usually of no significance in a trial for tax evasion. Equally, in a piece of false evidence, such as a colleague’s fictional family commitment invoked as a pretext for missing a professional networking event, the alleged time and date of the putative family commitment is likely to be of significance, whereas the employee’s cited mode of transport that evening is not typically relevant to the support of the claim that the employee is unable to attend this networking event. More generally, all that is used as evidence in dialogue, and consequently all that needs to be expressed by E, is some putatively salient aspect of an item of evidence. It should be noted, however, that the claim to salience of the aspect expressed by E does not refer to some property of E but is instead a separate claim in its own right. The claim that the aspect expressed by E is salient in supporting a challenged proposition, a hypothesis H, amounts to the claim that E supports H (which I will symbolise as ‘E:H’). For clearly, salience is a relation: It makes no sense to claim that the aspect or property described by E is salient without specifying at least implicitly what this salience pertains to, that is, what the property expressed by E is claimed to be relevant to. We can thus define E and E:H in the following way: The proposition E stands in the prima facie supporting relation E:H to the hypothesis H iff an instance i of a property p is both expressed by E and is evidence in favour of H.

In order to support H in dialogue, one must therefore commit dialogically both to the claim that E and to the claim that E:H.3 Given this formalisation, we can identify two pairs of conditions that must prudently be met in order to support a questioned proposition, viz. the hypothesis H, to one’s opponent in a persuasion dialogue. First, the opponent needs to accept, that is, dialogically commit to, both E and E:H—lest the attempt to persuade said opponent that H on grounds of E and E:H will predictably fail. Second, and as a precondition to the first,  We could further distinguish between E:H and the stronger claim E::H, with the latter implying that there is no evidence EX non-identical with E such that it is true that EX:¬H is stronger than E:H. As such, E:H amounts to the claim that E prima facie supports H (or simply that E supports H), while E::H amounts to the claim that E ultima facie supports H (or simply that E proves H). E::H plays no role in my subsequent considerations, as the latter only concern the claim that E:H. 3

3  Evidence, Inference, and Empiricism 

57

both E and E:H must be acceptable to the opponent—by which I mean the following. Let us assume that C is some proposition that one’s opponent had already committed to earlier in the dialogue (though C is not identical with ¬H). If either E or E:H were inconsistent with C (whether that inconsistency be discovered immediately or subsequently), then the opponent would be highly likely to commit to at most one of the two inconsistent claims. Given the prior commitment to C, the opponent would also be very likely to maintain commitment to C and reject E or E:H (whichever is inconsistent with C). That is, given the opponent’s prior commitment to C, and given that either E or E:H is inconsistent with C, the opponent would be very likely to reject E or E:H, making it prudent to avoid any E or E:H that is inconsistent with any C. If E or E:H were inconsistent with C, then we may say that E/E:H would not be acceptable to the opponent, and the opponent would in all likelihood not accept E/E:H. Supporting a questioned proposition H in a philosophical dialogue thus prudently requires that both E and E:H be accepted by, and hence be acceptable to, one’s opponent.4 One might object to the requirement that E and E:H be acceptable to one’s opponent. After all, one’s opponent might be a sceptic or a persistent interlocutor (and might have declared as much), so next to no proposition might ultimately be acceptable to this opponent. Why should we have to bother with a sceptic at all—and hence, why should we endorse the requirement that E/E:H be acceptable to the opponent? Is it not perfectly reasonable to insist at the very least on some particularly basic propositions that it would be preposterous to question, regardless of what one’s interlocutor might say? If such insistence upon certain basic propositions was indeed perfectly reasonable, then the requirement that any E or E:H should be acceptable to one’s opponent must be rejected, as it prohibits such reasonable insistence (cf. Williamson 2007: 210-2). The problem with this objection has already been addressed in the previous chapter: These basic propositions would in effect have privileged status, such that they ought ‘reasonably’ to be accepted unless and until they are sufficiently undermined. Such privileged status itself depends upon the  The requirement that E and E:H should be acceptable to one’s interlocutor is also implied by what Williamson (2007: 210-2) calls the thesis of evidence neutrality (cf. Weatherson 2009). 4

58 

C. Rhode

acceptance of dialogical foundationalism. Since dialogical foundationalism has already been rejected in the previous chapter, this objection to the acceptability requirement must be rejected as well. In another objection, one might claim that the universal requirement that E and E:H should be acceptable to one’s opponent seems not to hold in the sciences, where there is no obligation to take an opponent seriously if he is explicitly sceptical of evolution by natural selection, or of the existence of radioactive elements—hence, the universal requirement would seem falsified (cf. Williamson 2007: 210-2). However, the requirement for E and E:H to be acceptable to one’s opponent is not universal. It is based on the considerations offered earlier in the present section, which in turn apply only to philosophical (persuasion) dialogues, not to scientific dialogues. It thus remains the case that E and E:H must prudently be both accepted by and acceptable to one’s philosophical opponent if one is to succeed in persuading said opponent that H on grounds of E and E:H. With these two pairs of conditions in place, I will now turn to a closer examination of the supporting relation E:H.

3.3 Implication Barriers As with E, one’s opponent would need to accept E:H if one is to succeed in supporting H to this opponent on grounds of E and E:H. If the opponent did not accept both E and E:H, then the attempt to support H in this way would fail. So, as with E, the obvious problem is that the opponent, having already questioned H, may also request support for E:H, the proposition that E supports H. If the opponent were to question E, then E would in effect become the hypothesis at issue in a separate but embedded dialogue (with the possibility of a regress, as discussed in Chap. 2, Sect. 2.2). The same, then, would happen if one’s opponent chose to request support for E:H—one would be under the prudential obligation to support E:H. I will not discuss exactly how this request for support might be satisfied—both because the possibility of a regress entails that no algorithm would guarantee success in supporting E:H, and because it is in any case unnecessary for my present considerations. Given dialogical

3  Evidence, Inference, and Empiricism 

59

egalitarianism, the persuasive risk of an opponent’s questioning E:H can never be eliminated. But as discussed in Chap. 2, Sect. 2.2, this does not mean that no assertion could ever be supported to an opponent’s satisfaction. Some evidential support for a questioned proposition might satisfy an opponent, and other support might not. Equally, some evidential support for a questioned supporting relation E:H might satisfy a given opponent while other support does not. The persuasive risk of further challenge by an opponent will always exist. All that we can hope to do in argumentative practice is to minimise this persuasive risk. There may of course be various steps that we could take in order to minimise the risk to persuasion. A particularly profitable step, however, would be to avoid the employment of any E or E:H that, if questioned, practically cannot be supported. After all, if one were to employ a practically unsupportable E or E:H in an effort to persuade one’s opponents who have already questioned H, then one would all but invite these opponents with open arms to go on and question the unsupportable E or E:H. My subsequent focus will therefore be upon any E:H that would be unsupportable if questioned. I will examine under what circumstances it is impossible in argumentative practice to satisfy an opponent’s request for support for E:H. That is to say, I will examine under what conditions E:H is associated with a barrier to successful persuasion in philosophical dialogue. As I will explain shortly, E:H faces such a persuasion barrier if E:H crosses an inference barrier, also known as an implication barrier. An implication barrier is the separation between propositions of two different types, and an implication barrier thesis is a thesis expressing such a separation. It states, abstractly, that one cannot derive propositions of one type from propositions of the other type. Arguably the most prominent example of an implication barrier thesis is Hume’s First Law, famously introduced in the demand that ‘a reason should be given, for what seems altogether inconceivable, how this new relation [of ought, or ought not,] can be a deduction from others [viz., is, or is not], which are entirely different from it’ (1978 [1740]: 469-70). Other implication barrier theses state, for instance, that one cannot derive propositions of necessary truth from propositions of contingent truth (Kant’s Law); that one cannot derive propositions of temporally unrestricted scope from propositions of temporally restricted scope (Hume’s Second Law); and that one

60 

C. Rhode

cannot derive propositions of unrestricted scope from propositions of restricted scope ((Bertrand) Russell’s Law; Restall and Russell 2010: 243).5 Such barrier theses still require some further specification, in two different respects. A barrier thesis may be either logical or semantic, and it may be either epistemological or dialogical. To illustrate this, consider Hume’s First Law, the is/ought separation. This barrier thesis may be logical, stating that one cannot derive normative propositions from descriptive propositions with the aid of logical deduction alone; or it may be semantic, stating that one cannot derive normative propositions from descriptive propositions with the aid of logical deduction and analytic bridge principles, that is, certain propositions expressing conceptual truths, because there are no analytic bridge principles connecting descriptive concepts to normative concepts. Another way of putting the same distinction is to say that deriving normative propositions from descriptive propositions cannot be accomplished by means of formally valid deduction (logical barrier), or that it cannot be accomplished by means of formally or materially valid deduction (semantic barrier; Pigden 2010a: 7, 13). Either way, however, we may define implication barrier theses somewhat more precisely as stating, abstractly, that validly deducing propositions of a given type from propositions of a different specified type cannot be accomplished. One may engage in deductive inference, furthermore, in either of two locations, and one may consequently encounter an implication barrier and thus fail to accomplish the attempted deduction validly in either of these two locations. One might encounter an implication barrier epistemologically or dialogically, that is, either in reasoning as part of some belief-forming or other attitudinal process, or in reasoning as part of some dialogue. Considering Hume’s First Law as an example again, I might seek to deduce normative propositions from descriptive propositions within my own belief-forming or other attitudinal processes,  The formulations chosen here illustrate that Hume’s Second Law is but a special case of (Bertrand) Russell’s Law (cf. (Gillian) Russell 2010: 156). Historically more faithful formulations would concern inferences from propositions about the past to propositions about the future (Hume’s Second Law) and inferences from particular propositions to general propositions (Russell’s Law). Note that the person-specific labels for the foregoing implication barrier theses are drawn from Gillian Russell (2010). 5

3  Evidence, Inference, and Empiricism 

61

perhaps when reflecting on my moral obligations. Aside from that, I might also try to deduce normative propositions from descriptive propositions within a dialogue, perhaps when trying to persuade a philosophical interlocutor that her dialogical truth-value commitments to certain descriptive propositions imply commitment to some specific moral thesis. Given the separation between attitudinal and dialogical commitments, Hume’s First Law’s holding among attitudinal commitments does not necessitate its holding among dialogical commitments, nor vice versa. That is, Hume’s First Law may hold in one location regardless of whether it holds in the other location. The same independence of implication barrier theses of different locations applies with regard to any other barrier thesis, as any other barrier thesis may hold dialogically regardless of whether it holds attitudinally, and vice versa. (Hence the separation between attitudinal and dialogical commitments itself constitutes an implication barrier.) I stress this separation between attitudinal and dialogical implication barriers in order to subsequently disregard attitudinal barriers and to focus entirely upon dialogical barriers. More narrowly still, I am only concerned with implication barriers as they pertain to persuasion dialogue, where one party seeks to get her opponent to dialogically commit to H, pursuing this goal by citing E and E:H. The opponent may in turn question whether E, or he may question whether E:H, and in each case, this could launch a regress of supporting propositions. As indicated above, however, there is one crucial difference between, on the one hand, a regress arising from E:H where E:H crosses an implication barrier, and, on the other hand, a regress arising from E or else from E:H where E:H does not cross an implication barrier. In the latter case (of E or a non-barrier-crossing E:H), the questioned proposition (E or E:H) might be supported in a multitude of different ways, not logically restricted to one particular kind of support. But in the former case (of a barrier-crossing E:H), the questioned proposition (E:H) can only be supported, in effect, by being reformulated and/or re-asserted. To illustrate this, consider the descriptive proposition D1, Lying tends to undermine trust between people, along with the descriptive proposition D2, It tends to be prudent to not lie if one seeks to be trusted, as well as the normative proposition N1, One ought not to lie. If my non-barrier-­crossing assertion that D1:D2 was questioned, then I might, for instance, conduct

62 

C. Rhode

a series of laboratory experiments to test D1:D2; or I might invoke existing social psychological research findings indicating the truth of D1:D2; or I might offer analyses of the concepts employed in D1 and D2 to the effect that D1 implies D2;6 or I might assert some reformulation of D1:D2.7 However, if my barrier-crossing assertion that D1:N1 was questioned, then I could not produce or invoke experimental results in support of D1:N1, as it is practically impossible to obtain such results. Equally, I could not offer any conceptual analyses in support of D1:N1, as this would require analytic bridge principles that are unavailable (or if they are available, they are themselves practically unsupportable). All that I could do to support D1:N1 is to re-assert it in a different formulation, or perhaps to assert some similar proposition DA:NA that implies D1:N1, where DA:NA would be equally difficult to support. The same holds with regard to any barrier-crossing E:H, as no such E:H can be supported empirically or conceptually. Yet, simply re-­asserting the barrier-crossing E:H (or asserting some different claim EA:HA that is so closely related to the questioned claim E:H as to imply the latter) would be exceedingly unlikely to be accepted by an opposing party that has already questioned E:H, and is thus imprudent in any dialogue aimed at persuading that philosophical opponent. With the options for supporting E:H exhausted, we must conclude that any implication-barrier-­ crossing E:H is practically unsupportable if questioned. That is to say, a burden of proof imposed in philosophical dialogue upon E:H cannot be satisfied if E:H crosses an implication barrier. Given such an unsupportable proposition among the support offered for H, any philosophical opponent reluctant to dialogically commit to H in a persuasion dialogue would predictably question the unsupportable E:H, thus making it practically impossible to support H by invoking E and a barrier-crossing E:H. This effective barrier to persuasion, I suspect, may constitute the idea at the heart of any implication barrier. Indeed, the present  Note that the employment of conceptual analysis may in fact be imprudent here, as will become clear in the next chapter. 7  The provision of support should not be confused with the attempted insulation from any need for support, as may be pursued by appeals to authority or by appeals to putative obviousness. Either of these responses would implicitly allege that it is inappropriate to question D1:D2 and would thus presume dialogical foundationalism. 6

3  Evidence, Inference, and Empiricism 

63

persuasion-centred interpretation of implication barriers has already been advocated in similar terms by Annette Baier (2010), in her interpretation of Hume’s First Law. Baier held that Hume did not postulate a law but merely highlighted an argumentative gap, where the burden of proof is incurred by whoever sets out to bridge this gap.8 Baier—and quite possibly Hume—thus reduced this implication barrier to a persuasion barrier. In the following, I adopt this interpretation of Hume’s First Law for implication barriers in general: Given the goal of persuasion, an implication barrier is relevant in philosophical persuasion dialogue only to the extent that it is a persuasion barrier. Whether an implication barrier highlights a logical divide or metaphysical separation between two propositions is, on its own, quite irrelevant to one’s persuasive success in a philosophical dialogue. What ultimately seems to make an implication barrier relevant for the purposes of persuasion is its effect of acting as a persuasion barrier. In any attempt to support a questioned hypothesis H in a philosophical persuasion dialogue, it is therefore prudently required to avoid the employment of any implication-barrier-crossing proposition E:H. Of course, this requirement presupposes that prominent counter-­ examples to implication barriers are ineffective in philosophical persuasion dialogue. So I will demonstrate next that such counter-examples are indeed ineffective.

3.4 Addressing Counter-Examples A variety of counter-examples have been raised against the above implication barriers, yet none of these counter-examples serves to undermine their operation as persuasion barriers in philosophical persuasion dialogue. It will suffice to demonstrate as much with the example of Hume’s First Law, as the deflection of counter-examples to other implication barrier theses is effected analogously. Illustrative counter-examples to the is/ ought separation have been proposed by Arthur Prior (1960) and Gideon  Baier also held that this burden could be satisfied, e.g. by citing accepted social conventions – a possibility I will address in the following section. 8

64 

C. Rhode

Rosen (see Russell 2010: 154). They are effectively representative of other counter-examples, with Prior’s counter-example being a formally valid is/ ought-deduction, and Rosen’s counter-example being a materially valid is/ought-deduction. First, consider Prior’s formal counter-example. Take any unambiguously descriptive proposition D, such as Tea-drinking is common in England, and any unambiguously normative proposition N, such as All New Zealanders ought to be shot (Prior 1960). Now consider the following two valid arguments: D ∴ D∨N, and D∨N, ¬D ∴ N. Assuming that any proposition is either normative or descriptive, the proposition that D∨N must also be either normative or descriptive. If D∨N is normative, then the first of these two arguments constitutes a counter-example to the is/ought-separation, as the normative proposition D∨N is validly deduced from the descriptive proposition D. If, however, D∨N is descriptive, then the second of the two arguments constitutes a counter-example to the is/ought-separation, as the normative proposition N is validly deduced from the descriptive propositions D∨N and ¬D. Either way, then, it is quite easy to validly deduce a normative conclusion from exclusively descriptive premises. This counter-example can be addressed by considering the deployment of the second argument (D∨N, ¬D ∴ N) in a philosophical persuasion dialogue. In such a dialogue, the party asserting that N would, if N was questioned, here provide D∨N and ¬D in support of N. It is highly likely that the opponent, who has already proved reluctant to accept N, would then question D∨N, since it seems nearly impossible to provide support for the challenged proposition D∨N. The only support that could possibly serve to satisfy one’s opponent would be either D or N, since D implies D∨N, just as N implies D∨N. Yet if D was invoked in support of D∨N, then the original argument—D∨N, ¬D ∴ N—would be rendered trivial and thus quite obviously useless in any persuasive effort. And if N was invoked in support of D∨N, then the original argument—D∨N, ¬D ∴ N—would be rendered circular, and equally useless (while also undermining its status as a counter-example). Since neither a trivial nor a circular argument will be of any use in a philosophical persuasion dialogue,

3  Evidence, Inference, and Empiricism 

65

the argument form of Prior’s second argument will not serve to cross the is/ought separation in such dialogue. Given the goal of persuading one’s philosophical opponent, Prior’s formally valid counter-example thus does not undermine the is/ought persuasion barrier in argumentative practice. Next, consider the materially valid is/ought-deduction offered by Rosen (as cited in Russell 2010). Rosen asks us to imagine the act of ‘flurging’, which is defined as ‘[doing] something that one ought not to do in front of children’ (Russell 2010: 154). He then offers the following materially valid argument: ‘Jones is in the presence of children. [Therefore,] Jones ought not to flurg’ (Russell 2010: 154). The validity of this argument depends upon the definition of ‘flurging’, which in turn depends upon the normative claim that One ought not to flurg in the presence of children. If this normative claim was false, then the definition of flurging, as presented by Rosen, would be false as well, and the argument would be materially invalid. Hence, this normative claim is actually an implicit premise in the argument. Now, any philosophical opponent reluctant to commit to the argument’s conclusion would be likely to question the inference from the premise to the conclusion, and would thus in effect demand that the argument be rendered formally valid. And as with any merely  materially valid argument, this argument can be rendered formally valid by making the implicit premise explicit. Yet if the implicit premise is made explicit, then the argument’s normative conclusion is deduced from one descriptive and one normative premise, so it no longer provides a counter-example to the is/ought separation. Rosen’s materially valid counter-example thus equally fails to undermine the is/ought persuasion barrier in argumentative practice.9 From the brief discussion of these two putative counter-examples to the is/ought-separation, we can derive several lessons with regard to counter-examples to any implication barrier thesis. Formally valid deductions crossing an implication barrier, though perhaps logically 9  John Searle’s (1964) counter-example is more complex than but parallel to Rosen’s (and is therefore only footnoted here). In essence, Searle invokes the social institution of promising in order to materially validly deduce the obligation to act according to one’s promise from the description of the utterance of this promise. Like Rosen’s, Searle’s counter-example would be undermined by making its implicit definition of ‘promising’ explicit, as this would amount to adding the normative premise that Anyone who utters (or: credibly utters) a promise to perform some action is obligated to perform that action.

66 

C. Rhode

unobjectionable, are useless in philosophical persuasion dialogue. Aside from Prior’s putative counter-example, the deduction from a proposition of one type, together with that proposition’s negation, to a proposition of a different type (e.g. D, ¬D ∴ N) is just as valid and just as useless in philosophical persuasion dialogue. The logical form is independent of the particular barrier in question, which is why the same uselessness in persuasion dialogue applies to formally identical counter-examples to any other implication barrier. It can furthermore be expected that the failure of these counter-examples would be reproduced with regard to any other formal counter-examples to implication barrier theses. Materially valid deductions crossing an implication barrier fare no better, being equally useless in any persuasive effort, since they only actually cross a given implication barrier as long as their implicit premise remains just that—an implicit, unacknowledged premise. Once the implicit premise that they depend upon is made explicit due to an interlocutor’s questioning of the argument’s validity, a formerly materially valid barrier-­crossing deduction no longer crosses any implication barrier. This dissolution of material counter-examples appears to be independent of both the particular barrier in question and the particular counter-example at issue, so the same uselessness in persuasion dialogue affects any other case of a materially valid barrier-crossing deduction. Both formal and material counter-examples to implication barrier theses can thus be deflected by focusing upon implication barriers as they occur in argumentative practice, and specifically in philosophical persuasion dialogue. Though such counter-examples may undermine implication barriers in logic, they have no similar impact on implication barriers in argumentative practice, that is, on persuasion barriers. Hence, counterexamples to implication barrier theses do not undermine the previous conclusion that the employment in philosophical persuasion dialogue of any implication-barrier-crossing proposition E:H in support of H must prudently be avoided. In the remainder of this chapter I now want to spell out this conclusion a little further, which will require the introduction of a rather underappreciated distinction among implication barriers.

3  Evidence, Inference, and Empiricism 

67

3.5 Quantitative and Qualitative Barriers This underappreciated distinction has to date only been identified by Alexander Bird (2007: 62-3) with respect to underdetermination problems in the philosophy of science. It is well-known that a scientific theory can never be conclusively proved to be true, no matter how much evidence one may accumulate. The evidence invariably underdetermines the choice between the theory at issue and at least one other theory that is equally consistent with the evidence. Bird distinguishes two types of considerations in favour of this underdetermination thesis. On the one hand, he identifies quantitative considerations to the effect that the evidence available to us, due to its quantity, will inevitably fail to rule out all but one theory consistent with that evidence. So, for instance, the evidence available to us at any specific point in time will never include evidence obtained in the future, hence we cannot prove the truth of a scientific theory ostensibly holding at all times (including the future). On the other hand, Bird identifies qualitative considerations, according to which there is a qualitative difference between our evidence (or the propositions describing that evidence) and the theory to be supported by it. ‘To take an extreme example, if evidence could only consist of mathematical propositions we would think that our evidence underdetermines conclusions concerning organic chemistry’ (Bird 2007: 63). Our evidence thus underdetermines a scientific theory either because the evidence is not of the appropriate extent—according to quantitative considerations—or because it is not of the appropriate kind—according to qualitative considerations. The same distinction can be drawn among implication barriers. (Indeed, underdetermination problems themselves can be reconceived as implication barriers, so the distinction may already have been drawn among the latter.) Hume’s two barrier theses will serve to illustrate this distinction between quantitative and qualitative barriers. Hume’s Second Law (as I formulate it here) states that we cannot validly deduce propositions of temporally unrestricted scope from propositions of temporally restricted scope. The reason for this inability is our lack of future evidence, evidence that simply has not been gathered yet. As such, the

68 

C. Rhode

extent, or quantity, of evidence available to us is insufficient for deductively supporting a temporally unrestricted hypothesis. Contrast this with Hume’s First Law, the thesis that one cannot validly deduce normative propositions from descriptive propositions. The inability to accomplish such a deduction is not due to the insufficient quantity of the available evidence but is, rather, due to the nature of the evidence itself. No matter how much descriptively expressed evidence we gather, this evidence could not possibly serve to deductively support a normative hypothesis, since the latter is of a different kind from the evidence. Aside from Hume’s First Law, such qualitative implication barrier theses further include Kant’s Law (stating that we cannot validly deduce propositions of necessary truth from propositions of contingent truth). And aside from Hume’s Second Law, quantitative implication barrier theses also include Russell’s Law (stating that we cannot validly deduce propositions of unrestricted scope from propositions of restricted scope). (Other barrier theses may be added to these lists.) If we considered all and only the propositions of a specific quality or kind to be members of a specific domain, then we might say, for illustrative benefit, that inferences merely across a quantitative barrier are intra-domain inferences, while those across a qualitative barrier are inter-domain inferences. What distinguishes these two types of barriers appears to be the logical possibility or impossibility, respectively, of crossing them when discharging a burden of proof.10 If it is logically possible to gather sufficient evidence to prove a hypothesis by way of a given implication-barrier-crossing inference, then the implication barrier crossed by that inference is quantitative. Consider, for instance, Hume’s Second Law: It is logically possible to gather evidence from every point in time (since it is logically possible to travel into the past and the future), so a temporally unrestricted hypothesis could be proved by temporally restricted evidence thus gathered. By contrast, if it is logically impossible to gather sufficient evidence to prove a hypothesis by way of a given implication-barrier-­ crossing inference, then the implication barrier crossed by that inference is qualitative. This may again be illustrated with Hume’s First Law: No  I assume that it is logically possible that P iff P is not a contradiction according to the rules of the language of P. 10

3  Evidence, Inference, and Empiricism 

69

matter how much descriptively expressed evidence has been obtained, it is still logically impossible to prove a normative hypothesis with the evidence thus obtained. Most importantly for present purposes, quantitative and qualitative implication barriers are also distinguished by their different scope of application with regard to types of inferences in persuasion dialogue. Qualitative barriers apply to both deductive and ampliative inferences in persuasion dialogue; yet quantitative barriers apply only to deductive inferences in such dialogue. Consider qualitative barriers first: In the attempt to get an opponent to dialogically accept a hypothesis of some specific kind, evidence of a wholly different kind (or propositions expressing that evidence) would predictably fail to have the desired influence on the opponent—regardless of the type of inference ostensibly connecting the evidence to the hypothesis. Whether one claims, for instance, that the descriptive propositions expressing some evidence deductively imply some normative hypothesis or that they inductively support that normative hypothesis will be irrelevant: An opponent who has thus far been unwilling to accept this normative hypothesis in a philosophical persuasion dialogue will very likely be unimpressed by the descriptive evidence, irrespective of the type of inference. Equally, an opponent in a persuasion dialogue who has already questioned a putatively necessary truth will be unlikely to accept this supposedly necessary truth on the basis of contingent truths presented in its support—regardless of whether the presented inference from the contingent truths to the putatively necessary truth be deductive or ampliative. While qualitative barriers hold for any type of inference in persuasion dialogue, quantitative barriers only carry any significance with regard to deductive inferences. Most obviously, Hume’s observation that inductive inferences are always deductively invalid does nothing to detract from the potential persuasive strength of an inductive inference across a quantitative barrier when that inference is not billed as a deduction. Quantitative barriers thus only apply to deductive inferences in persuasion dialogue. Indeed, one can easily avoid a quantitative barrier in an attempt to persuade one’s opponent by presenting the inference in question as an ampliative rather than a deductive inference. Qualitative barriers, however, cannot be avoided in this fashion in persuasion dialogue, as explained

70 

C. Rhode

above. Consequently, in persuasion dialogue, qualitative barriers may be more properly described as inference barriers, applying to any type of inference, while quantitative barriers are implication barriers in the strict sense, applying only to deductive inferences. (A notable further difference concerns the directionality of such barriers. Qualitative barriers are bi-directional, as they block inferences across these barriers in either direction. Quantitative barriers, by contrast, are uni-directional, since they block inferences from a more narrowly scoped proposition to a more widely scoped proposition (within the same domain) but not vice versa.) Combining this distinction with the previous result, we can conclude that the employment of any proposition E:H in support of H, where E:H crosses an inference barrier (i.e. a qualitative barrier), must prudently be avoided in any philosophical persuasion dialogue. In respect of this conclusion, an important difference should be noted between inference barriers in logic and inference barriers in argumentative practice (i.e. persuasion barriers). Inference barriers in logic have traditionally been considered to block arguments for a conclusion of a given kind that exclusively employ premises of another kind. Yet in persuasion dialogue, inference barriers block any employment of premises of a given kind in support of a conclusion of another kind. The difference here concerns arguments that employ premises of kind a together with premises of kind b to support a conclusion of kind b. Such mixed-premises arguments do not exclusively employ premises of one kind in support of a conclusion of another kind, and hence, they are not (or not obviously) blocked by inference barriers in logic. In persuasion dialogue, by contrast, these mixed-premises arguments are blocked: The opposing party that has already questioned a hypothesis H, when offered a mixed-premises argument in support of H, would be very likely to question the salience of those premises that differ in kind from H. That is, an opposing party would be very likely to question the implied or explicit claim that these premises lend support to H.  Any support that could then be offered for that claim would itself cross the given inference barrier and would thus be effectively unsupportable. As such, mixed-premises arguments for H imply an inference-­ barrier-­crossing premise E:H and must thus prudently be avoided in any philosophical persuasion dialogue. They may be innocuous in purely logical terms, as seen in the above counter-examples, but they falter at an

3  Evidence, Inference, and Empiricism 

71

inference barrier in argumentative practice. (As I am not concerned with inference barriers in logic, I will subsequently use ‘inference barriers’ only to denote persuasion barriers, that is, inference barriers in argumentative practice.)

3.6 Empiricism I want to close this chapter by distilling what I take myself to have established in the present work up to this point. In Chap. 1, I have introduced the logical separation between attitudinal and dialogical truth-value commitments. (This separation is clearly qualitative, so in the terminology just adopted it constitutes an inference barrier.) I have also shown empirically that defeating one’s opposing party is the vastly prevalent goal of philosophical dialogues. On this basis, I have argued in Chap. 2 that attaining this goal prudently requires the adoption of dialogical egalitarianism in philosophy. A burden of proof thus cannot be evaded and must instead be met. In attempting to meet the burden of proof upon a questioned hypothesis in philosophy, as I have argued in the present chapter, it is prudently necessary that one does not cross any qualitative or inference barrier (e.g. Hume’s First Law or Kant’s Law). In essence, then: If party a commits to proposition P in philosophical dialogue δ, and party b commits to neither P nor ¬P in δ, then a must prudently support P to b in δ without crossing any inference barrier. This thesis is recognisably empiricist, not least as it is recognisably Humean. The most prominent of inference barriers—Hume’s is/ought-­ separation—is of course reaffirmed here as a persuasion barrier (though not as a logical barrier, due to the above counter-examples)11. Yet beyond that particular implication, the above thesis is empiricist principally because it coheres in spirit with perhaps the most quintessentially empiricist principle, famously encapsulated in Hume’s remark that ‘A wise man […] proportions his belief to the evidence’ (1975 [1748]: 110). According to Hume, it is unwise to believe that which is not warranted by the evidence, and therefore unwise to infer from the evidence what is not 11

 For an outstanding examination of its status as a logical barrier, see Pigden (2010b).

72 

C. Rhode

warranted by it. Hume’s concern here is not (in contrast to many theorists following in his path) with the nature of evidence, but rather with the inferences that may or may not be drawn from available evidence. The same applies to my foregoing thesis, which is silent with regard to the nature of evidence and rather addresses the inferences (not) to be drawn from some given evidence in philosophical persuasion dialogue. It is for this reason that I consider my above thesis to be adequately described as empiricist. Yet my thesis differs significantly from other forms of empiricism, being thrice restricted by its prudential force, its metaphilosophical scope, and its dialogical scope. Consider the first of these restrictions. The above thesis is binding only prudentially, as a matter of instrumental rationality or goal-directed action. If it is the goal of the dialogue you are engaged in to defeat your opponent in argument, then the thesis applies to you and prescribes certain actions. Its force is thus conditional. However, I have also shown that the antecedent of this conditional prescription is true for nearly every philosophical dialogue, as circa 95% of non-embedded philosophical dialogues appear to instantiate persuasion dialogue; and I have argued that the truth of this antecedent must therefore be presumed in all philosophical dialogues. Hence, while its force is conditional, that condition has been shown to be satisfied, thus rendering the thesis binding on prudential grounds upon any party to a philosophical dialogue.12 Secondly, the sample in the empirical research in Chap. 1, which showed the antecedent to be satisfied almost universally, was itself restricted to philosophical dialogues. It did not include dialogues in legal contexts, in organic chemistry, in social history, or over in the pub. Consequently, my empiricist thesis does not apply to any discipline or field other than current Western analytic philosophy. As the present work will principally seek to address metaphilosophical questions, however, this does nothing to detract from its potential significance with regard to the conduct of philosophy. Perhaps most crucially, the above empiricist thesis applies only to philosophical dialogue. It is entirely silent with regard to recognisably philosophical beliefs and other attitudinal states, making no claim as to how philosophical thought should be conducted or how philosophical  Cf. Chap. 1, Sect. 1.3, especially fn. 23.

12

3  Evidence, Inference, and Empiricism 

73

knowledge or understanding should be acquired. Its scope is restricted to (current Western analytic) philosophy as it is conducted in various forms of dialogue, ranging from oral discussions and conference presentations over peer-reviewed journal articles to monographs.13 Contrary to any other form of empiricism, it is thus an exclusively dialogical thesis, rather than an exclusively epistemological thesis or a thesis that is both epistemological and dialogical. (In the same fashion, as explained above, the present empiricist thesis is not logical either, as it does not affirm or even concern inference barriers in logic. Instead, it is only concerned with dialogical persuasion barriers.) To emphasise this important yet otherwise easily overlooked restriction, I will call the above thesis dialogical empiricism.14 To repeat, then, the thesis distilling the considerations raised up to this point states that: Dialogical Empiricism: If party a commits to proposition P in philosophical dialogue δ, and party b commits to neither P nor ¬P in δ, then a must prudently support P to b in δ without crossing any inference barrier.

One may plausibly wonder at this point why I have spent three chapters establishing such a relatively basic and perhaps quite commonsensical thesis. Could I not simply have posited the thesis and then argued less lengthily that it better coheres with our intuitions regarding the conduct of philosophy than any rival account does? Well, the answer is no—for two reasons. First, a thesis regarding the conduct of philosophy that itself depends on some controversial thesis regarding the conduct of philosophy would be of little if any deliberative value in guiding said conduct. The thesis that A theory is preferable (ceteris paribus) over a rival theory if the former better coheres with our intuitions than its rival does is itself very controversial. If that thesis was rejected, then the value of any  Insofar as the conduct of philosophical thought is considered to be dependent upon the conduct of philosophical dialogue, my empiricist thesis may of course have consequences for the former. Such consequences, however, are not my concern here. Cf. the discussion of attitudinal goals in Chap. 1, Sect. 1.3. 14  If other forms of dialogical empiricism are distinguished in the future (e.g. dialogical empiricism applying to social history) then my thesis would have to be renamed, in accordance with its other restrictions, as prudential metaphilosophical dialogical empiricism – a label I choose to avoid here for obvious reasons of economy. 13

74 

C. Rhode

argumentation I might offer for the intuitive superiority of dialogical empiricism over any rival theories would also be undermined. Arguing for the intuitive superiority of dialogical empiricism would thus potentially be a complete waste of time. To avoid wasting my time, any deliberative advice I offer will have to be grounded independently of any metaphilosophical thesis or, insofar as this is not possible, any metaphilosophical thesis my advice depends upon will need to be thoroughly supported. That is the first reason why I have derived dialogical empiricism on an independent basis over the previous three chapters. The second reason for why I could not simply have posited dialogical empiricism and assessed it against its coherence with our intuitions is that, in a sense, I intend to do the reverse—in the next chapter: to assess the importance of a theory’s coherence with our intuitions against the independently established demands of dialogical empiricism.

References Baier, A.C. 2010. Hume’s Own ‘Ought’ Conclusions. In Hume on Is and Ought, ed. C.R. Pigden, 49–64. Basingstoke: Palgrave Macmillan. Bird, A. 2007. Underdetermination and Evidence. In Images of Empiricism: Essays on Science and Stances, with a Reply by Bas C.  Van Fraassen, ed. B. Monton, 62–82. Oxford: Oxford University Press. Hume, D. 1975 [1748, 1751]. Enquiries Concerning Human Understanding and Concerning the Principles of Morals. Ed. L.A. Selby-Bigge, rev. P.H. Nidditch, 3rd ed. Oxford: Clarendon Press. ———. 1978 [1739–40]. A Treatise of Human Nature. Ed. L.A. Selby-Bigge, rev. P.H. Nidditch, 2nd ed. Oxford: Clarendon Press. Kelly, T. 2014. Evidence. In The Stanford Encyclopedia of Philosophy, Fall 2014 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/fall2014/entries/ evidence/. Pigden, C.R. 2010a. Introduction. In Hume on Is and Ought, ed. C.R. Pigden, 1–38. Basingstoke: Palgrave Macmillan. ———., ed. 2010b. Hume on Is and Ought. Basingstoke: Palgrave Macmillan. Prior, A.N. 1960. The Autonomy of Ethics. Australasian Journal of Philosophy 38: 199–206.

3  Evidence, Inference, and Empiricism 

75

Restall, G., and G. Russell. 2010. Barriers to Implication. In Hume on Is and Ought, ed. C.R. Pigden, 243–259. Basingstoke: Palgrave Macmillan. Russell, G. 2010. In Defence of Hume’s Law. In Hume on Is and Ought, ed. C.R. Pigden, 151–161. Basingstoke: Palgrave Macmillan. Searle, J. 1964. How to Derive ‘Ought’ from ‘Is’. Philosophical Review 73: 43–58. Weatherson, B. 2009. Evidence Neutrality [PowerPoint presentation]. Arché Summer School, 29 June–3 July 2009, St Andrews. http://brian.weatherson. org/ENw.pdf. Accessed 17 June 2015. Williamson, T. 2007. The Philosophy of Philosophy. Oxford: Oxford University Press.

4 Philosophical Methods between Content and the World

4.1 Consequences of Dialogical Empiricism Over the course of the preceding three chapters I have presented various considerations regarding the instrumentally rational conduct of philosophical dialogue, with particular attention devoted to the goal of such dialogue (Chap. 1), the burden of proof (Chap. 2), and inference barriers (Chap. 3). This culminated in the formulation of a thesis I called dialogical empiricism, which effectively states that any proposition questioned in philosophical dialogue must be supported (a) by the dialogue party that asserted the proposition and (b) without crossing any inference barrier. In the next three chapters, I want to discuss some of the consequences of dialogical empiricism. As dialogical empiricism is a metaphilosophical thesis, I could not hope to provide a comprehensive account of its implications across the entirety of philosophy in anything less than a multi-­ volume work. The selection of consequences to be discussed in the following chapters is thus ultimately arbitrary, guided by my own interests. Nevertheless, I hope that the discussions in these chapters will serve to be indicative of many of the consequences of dialogical empiricism not discussed here. Not least to this end, I begin in the present chapter by © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_4

77

78 

C. Rhode

examining some of the consequences of dialogical empiricism for certain prominent philosophical methods. Taking into account these methodological consequences, I will devote Chap. 5 to the discussion of consequences for a selection of more substantive philosophical issues. Chapter 6 will then conclude the present work with an assessment of potential avenues of escape from some of these consequences. In the present chapter I will begin (in Sect. 4.2) by introducing one particularly significant inference barrier that, according to dialogical empiricism, must not be crossed in philosophical dialogue. In order to assess philosophical methods, I will then apply this inference barrier to hypotheses in philosophy (Sect. 4.3) and to evidence in philosophy (Sect. 4.4), before addressing several objections to this inference barrier (Sect. 4.5).

4.2 The Content/World Separation Possibly the most consequential inferential barrier for philosophical dialogue is the separation between content and the world. This barrier divides explicit or implied mental or linguistic content, on the one hand, from the world beyond such content, on the other hand. According to this content/world separation, one cannot infer from the explicit or implied content of one or more people’s mind or language to reality beyond such psychological or linguistic (for short: psycho-linguistic) content.1 That is to say, one cannot employ propositions expressing elements of psycho-­ linguistic content in support of propositions expressing aspects of the world beyond such content. It should be noted that psycho-linguistic content is not co-extensive with mental content, though the latter is part of psycho-linguistic content, of course. To the extent that mental content is propositional (which  I am implicitly assuming, in line with virtually all analytic philosophers, that one’s access to at least some psycho-linguistic content is relatively less problematic than one’s access to reality beyond such content. Note, however, that my considerations only concern the relation between psycho-linguistic content and reality beyond such content. They are therefore independent of one’s own relation to either, that is, independent of the cited assumption. (For a critique of this assumption see e.g. Rouse 1996: 209.) 1

4  Philosophical Methods between Content and the World 

79

I will subsequently assume to be the case universally, for the sake of expressive simplicity), we may note that a proposition is just a linguistic device that is cognised and thus contained among mental events; and to the extent that mental content is non-propositional, we should note that such qualitative or experiential content is invariably dependent upon and thus located within an agent’s experience. But mental content only comprises the content of thought and of other psychological events—it comprises neither the content of talk or of other unambiguously linguistic events, nor the occurrence of thought or talk or of other psychological or linguistic events. Psycho-linguistic content, however, comprises both the content and the occurrence of thought and talk, in particular because the occurrence of thought or talk is itself situated, or contained, within the totality of someone’s psychological or linguistic events. As such, mental as well as linguistic content, and the occurrence of either, all constitute psycho-­linguistic content. The content/world separation thus has the appearance of being co-­ extensive with the related but separate mind-dependent/mind-­ independent distinction (cf. e.g. Brock and Mares 2007: 4). After all, the content and occurrence of thought and talk are equally mind-dependent and content-located, while all else beyond is both mind-independent and world-located. So is the content/world separation a mere reformulation of the mind-dependent/mind-independent distinction—a cumbersome attempt to reinvent the wheel? I admit that this is very nearly the case, and the distinction between what is mind-dependent and what is mind-­ independent may in fact serve as a familiar and convenient approximation of the different sides to the content/world separation. Nonetheless, there are two notable differences between the mind-dependent/mind-­ independent distinction on the one hand, and the content/world separation on the other hand. The first difference concerns philosophical zombies—hypothetical human-like beings that, roughly, lack a mind but engage in dialogue (cf. Chalmers 1996). The linguistic events of such a zombie, including both their content and their occurrence, cannot be dependent upon the zombie’s mind, given that these linguistic events occur even though there is no mind to give rise to them, and so they must be mind-independent. The content/world separation, however, locates a zombie’s linguistic engagement in dialogue within content, rather than

80 

C. Rhode

within the world beyond, as content is not restricted to mental content but also includes linguistic content. While this may be an interesting point, it will become clear below that hypothetical zombies have no import for present purposes. The second difference, though, is more significant. The mind-­ dependent/mind-independent distinction concerns the ontological status of putative entities: The existence of some things, such as the god Neptune, may be dependent upon their featuring in some minds, whereas the existence of other things, such as the planet Neptune, might be independent of their being found in any mind. Yet the content/world separation does not concern the ontological status of entities but rather the persuasive effectiveness of inferences between propositions. So although the mind-dependent/mind-independent distinction and the content/ world separation are almost entirely co-extensive (save for zombies’ engagement in dialogue), the former is ontological whereas the latter is functional, being concerned with persuasive effectiveness in philosophical dialogue. Does this barrier between psycho-linguistic content and the world beyond it, then, hold for all types of inference or only for deductive inferences—that is to say, is it an inference barrier or merely an implication barrier? As discussed in Chap. 3, Sect. 3.5, it is logically possible to gather sufficient evidence to prove a hypothesis via an implication barrier, but logically impossible to gather sufficient evidence to prove a hypothesis via an inference barrier. And as is readily clear, the content/world separation is an inference barrier: No increase in the number of propositions expressing an element of psycho-linguistic content would confer any degree of salience upon these propositions in the inference of a proposition expressing an aspect of reality beyond such content, since the premises would be of a different kind from the conclusion. Or to put it differently, any claim to the effect that a proposition expressing an element of psycho-linguistic content supports a hypothesis expressing an aspect of reality beyond such content would, if questioned, face an insurmountable burden of proof. Hence, the separation between psycho-linguistic content and the world beyond it is indeed an inference barrier. Varying versions of this content/world separation have been employed in the past, such as by epistemological sceptics and by logical positivists

4  Philosophical Methods between Content and the World 

81

(both of which I will touch upon again below). Yet, for present purposes, it suffices to briefly consider two examples for illustration. The first example is Heather Dyke’s critique of what she calls ‘the representational fallacy’ in metaphysics (2008; see also Baz 2017). As Dyke explains, it is common in Anglophone analytic metaphysics to employ facts about the English language as reasons in support of claims about the world beyond any language. In this way, for example, the (supposed) fact that the representation of reality in English is irreducibly tensed (as talk of past events, say, cannot be reduced to tenseless talk) has been argued to provide support for the claim that reality itself is tensed rather than tenseless, and that therefore, the past (and present and/or future) exists in some metaphysically significant way. The linguistic evidence invoked in this case is some proposition, allegedly implied in talk of past events, such as that The past exists (distinctly from the present and/or future). By inferring this proposition’s metaphysical truth from our ineliminable implicit linguistic commitment to it, metaphysicians commit the fallacy of supporting claims expressing an aspect of reality with claims expressing contents implied in language.2 In so doing, they depend on the (as Dyke argues, false) assumption that language is necessarily importantly representative of reality. Being a critique of inferences from contents implied in language to claims about reality, Dyke’s critique of the representational fallacy thus clearly espouses a version of the content/world separation between psycho-linguistic content and reality beyond such content. In a similar vein, Alvin Goldman and Joel Pust (1998) critique what they call ‘extra-mentalism’ in metaphilosophy. Assuming that intuitions are psychological states and find evidential employment in philosophy, Goldman and Pust distinguish between ‘mentalism’, the thesis that the targets of philosophical research are ‘in-the-head psychological entities’, and ‘extra-mentalism’, the thesis that these targets are ‘outside-the-head nonpsychological entities’ (1998: 183). Their principal objection to extra-mentalism is the problem of how psychological states like intuitions  In the terminology of Rudolf Carnap’s (1950) distinction between questions internal to a linguistic framework and questions external to any linguistic framework, metaphysicians commit the fallacy of supporting the purported external truth of a proposition with some proposition’s internal truth. I will fully contrast dialogical empiricism from both logical positivism and Carnap’s logical empiricism in Chap. 5, Sect. 5.6.

2

82 

C. Rhode

could reliably indicate non-psychological aspects of reality like natural kinds or modal equivalences.3 In effect, they thus posit an inference barrier between mental contents and reality beyond such contents, much like Dyke’s posited inference barrier between implied linguistic contents and reality beyond. Clearly, both barriers are versions of the content/ world separation between psycho-linguistic content and the world beyond such content. Before going on to apply this separation, however, I want to remove some potential confusions about this barrier—because what I here call the content/world separation has historically found generous employment in arguments for epistemological scepticism. The psycho-linguistic content inferentially separated from reality by this barrier includes perceptual content, such as the contents provided by one’s visual, auditory, or tactile faculties. It might therefore be argued that acceptance of the content/world separation would require acceptance of global epistemological scepticism, as no knowledge about the world could supposedly be derived from our perceptions if the barrier was accepted. Obviously, this objection would only carry any force against the content/world separation if it was granted that such scepticism is either false or unacceptable. Yet even granting that assumption, scepticism would be no threat to the present inference barrier, since the content/world separation does not operate at the epistemic but rather at the dialogical level. Recall that the context under investigation is the prudent conduct of philosophical persuasion dialogue. In this context, the content/world separation (in conjunction with dialogical empiricism) says no more than that it is persuasively imprudent in philosophical dialogue to support a challenged hypothesis expressing an aspect of the world beyond psycho-linguistic content with propositions expressing elements of psycho-linguistic content. As this persuasion barrier thus concerns nothing other than commitments in dialogue, it cannot necessitate acceptance of the attitudinal thesis of global epistemological scepticism. Given the attitude/dialogue separation, it is furthermore not the case that dialogical commitment to the content/world separation implies attitudinal commitment to the  More recently, Serena Nicoli (2016) has delineated a useful set of conditions for, and advocated the adoption of, what is here called ‘mentalism’. 3

4  Philosophical Methods between Content and the World 

83

latter barrier—so scepticism is not entailed in this indirect way either. Nor could the reasons establishing the content/world separation be applied analogously at the attitudinal level, as these reasons consist in the entire derivation of dialogical empiricism, including not least my empirical research of selected dialogues—all of which would have to be replicated attitudinally, which I strongly suspect to be effectively impossible. Finally, the content/world separation should equally not be confused with the attitude/dialogue separation. Each of these barriers could be described mistakenly (using Goldman and Pust’s terminology) as a barrier between what is ‘in-the-head’ (content, attitude) and what is ‘outside-­ the-­head’ (world, dialogue). This conflation would be straightforwardly mistaken, as the two inference barriers have different scope. The attitude/ dialogue separation divides commitments in attitude, that is, psychological truth-value commitments, from commitments in dialogue, that is, linguistic truth-value commitments. Such commitments are contents of thought or talk, thus qualifying as psycho-linguistic content. The content/world separation in turn distinguishes psycho-linguistic content from reality beyond such content. To illustrate this, consider the proposition that The Earth is flat. One’s belief (i.e. attitudinal commitment) that The Earth is flat does not make it a reality that The Earth is flat. Equally, one’s assertion or even collective acceptance (i.e. individual or collective dialogical commitment) that The Earth is flat still does not flatten the planet (regardless of whether such commitment may be in some important sense ineliminable from one’s individual or collective language). In short, the two inference barriers should not be conflated because the attitude/dialogue separation subdivides one side (viz. content) of the content/world separation. The purpose of the following two sections of this chapter is to apply the content/world separation for an assessment of a selection of philosophical methods. As my concern here is solely with methods that may be employed in support of a challenged hypothesis in dialogue, a method can suitably be characterised for present purposes (using the formalism introduced in Chap. 3) as the employment of some E together with some E:H in support of some H. In these terms, the content/world separation lies between any E where E expresses an element of psycho-linguistic content and any H where H expresses an aspect of the world beyond such

84 

C. Rhode

content. The question to be examined, then, is whether a given method (that is, a given employment of E and E:H in support of H) depends upon the crossing of the content/world barrier. To answer this question, it is necessary to examine each kind of E and each kind of H independently, thus to determine whether the evidence expressed by the given E is located in psycho-linguistic content or in reality beyond such content, and equally whether the given H expresses an element of psycho-­linguistic content or an aspect of reality beyond it. Once the location of a given method’s E and H has been determined relative to the content/world separation, we can see whether the method’s E:H crosses this inference barrier. In the next section, I will begin this examination by discussing several types of hypotheses at issue in philosophical dialogues. In the section thereafter, I will discuss various types of evidence employed in philosophical dialogue, while relating that evidence, respectively, back to the foregoing types of hypotheses in order to assess the philosophical methods characterised by the given pairing of E and H.

4.3 Hypotheses in Philosophy Hypotheses, for present purposes, include any proposition that is ever or has ever been asserted by one party to a philosophical dialogue and questioned by another party to that dialogue. As such, it is obviously impossible to examine every one of these indefinitely many hypotheses with respect to whether the given hypothesis expresses an element of psycho-­ linguistic content or an aspect of the world beyond such content. To render the present endeavour at all feasible as well as informative with regard to current philosophy, we will have to divide hypotheses along the lines of some manageably scoped groups. For this purpose, a variety of possible groupings would be available. One could, for instance, examine unconditional hypotheses and conditional hypotheses separately, thus conveniently exhausting the logical space of hypotheses. Yet it is entirely unclear why this way of grouping hypotheses would make it any easier to distinguish hypotheses according to whether they express psycho-linguistic content or reality beyond such content. Another way to group hypotheses would be to distinguish

4  Philosophical Methods between Content and the World 

85

between appropriately philosophical and not appropriately philosophical hypotheses. Along such lines, for instance, Hilary Kornblith argues (e.g. in his 2014) that the proper target of philosophical research for any philosophically interesting notion n is not the concept of n but rather n itself; hence, discussions of hypotheses expressing the (in my terms, psycho-­ linguistic) concept of n, or conceptual analyses of n, would ‘amount to little more than social psychology done badly’ (2014: 207). Kornblith’s account of appropriate philosophical hypotheses is of course widely contested, and it is highly likely that any other account to the effect that certain hypotheses of significant philosophical interest are inappropriate targets of philosophical research would be similarly contested. This approach to grouping hypotheses may thus invite challenges to any results obtained on its basis and therefore seems equally ineffective for present purposes. The recent literature in philosophical methodology provides some seemingly more fruitful ways of dividing up hypotheses. As outlined in the previous section, Goldman and Pust (1998) distinguish between (hypotheses about) psychological and non-psychological entities. In what is here effectively a co-extensive distinction, Daniel Cohnitz and Sören Häggqvist (2009: viii-xiii) find the subject matter of philosophy (and hypotheses about the same) to be variously either linguistic or extra-­ linguistic. Goldman (2007) further notes five seemingly prominent targets of hypotheses, including Platonic forms, Fregean concepts, natural kinds, personal concepts, and social concepts. To this could be added hypotheses about the processes of human engagement with psycho-­ linguistic content and with the world beyond, as studied not least in experimental philosophy (Alexander 2012: 3), as well as hypotheses expressing modal equivalences, metaphysical necessities, and mathematical, logical and conceptual truths (Goldman and Pust 1998: 185-6; Fischer and Collins 2015: 10; Pust 2016). These groups of hypotheses, though neither mutually exclusive nor clearly jointly exhaustive, still provide sufficient variety for a feasible and informative examination of hypotheses. So I will now turn towards assessing, for each of the above groups, whether the hypotheses in question express elements of psycho-­ linguistic content or aspects of the world beyond such content.

86 

C. Rhode

Psycho-Linguistic and Non-Psycho-Linguistic Subject Matter Consider first the groups of pairs delineated by Goldman and Pust as well as by Cohnitz and Häggqvist. As already touched upon above, one side of these pairs includes hypotheses about psychological (Goldman and Pust 1998) and linguistic (Cohnitz and Häggqvist 2009) subject matter, while the other side concerns hypotheses about non-psychological (Goldman and Pust 1998) and extra-linguistic (Cohnitz and Häggqvist 2009) matters. Take, for instance, the hypothesis that Saying ‘x is good’ means nothing more or less than saying ‘I approve of x’. The subject matter of this hypothesis is the meaning or content of a certain utterance (viz. the utterance ‘x is good’), so this hypothesis expresses psychological and linguistic subject matter, as well as an element of psycho-linguistic content. In contrast, take the hypothesis that Mereological simples are the only things that exist in the world. This hypothesis makes a claim regarding things existing in the world, so the hypothesis clearly expresses non-psychological and extra-linguistic subject matter, as well as an aspect of reality beyond any psycho-linguistic content. The two sides of the present pairs of types of hypotheses thus appear to be co-extensive with the two sides of the content/world separation, making the former’s identification in terms of the latter rather easy.

Concepts Next, consider Goldman’s (2007) five conceptions of the nature of concepts, which effectively amount to five groups of hypotheses mutually differentiated by their targets. Platonic forms, first, are eternal, without spatial location, yet causally tightly connected to the world. Being eternal, they exist independently of humans and their cognitions. Consequently, hypotheses describing a Platonic form straightforwardly concern reality beyond any psycho-linguistic content, thus expressing an aspect of the world beyond content rather than expressing any such content. Concepts in the Fregean sense, unlike Platonic forms, are acausal,

4  Philosophical Methods between Content and the World 

87

but in keeping with the latter, these abstract objects lack spatial location and, crucially, exist independently of human cognition. As such, hypotheses describing a concept in the Fregean sense are hypotheses about reality beyond any psycho-linguistic content, that is, they are hypotheses expressing an aspect of the world beyond content rather than expressing any such psycho-linguistic content. Natural kinds, in turn, are spatio-temporally located, specifically in the set of their instantiations. In this way, for instance, the natural kind star is located in the set of all its instantiations, including not least the star typically occupying the barycentre of our own solar system. If such kinds are indeed natural, then they exist and would persist independently of humans. In this case, hypotheses describing a natural kind would express an aspect of the world beyond content rather than expressing any psycho-­ linguistic content. If, however, such kinds are in some significant sense anthropogenic (perhaps because, as I suspect, most kinds are ontologically dependent upon their agential delineation), then such kinds would not persist independently of humans, even though their putative instantiations or material manifestations (such as the rocks comprising a planet, or the buildings of a university) may well persist independently of humans. Hypotheses describing such social or anthropogenic kinds would thus be hypotheses expressing some explicit or implied content of human classifications, thereby concerning psycho-linguistic content rather than the world beyond such content.4 Concepts can also be interpreted in what Goldman calls a psychological sense, where they may be held either personally or socially. A personal concept constitutes part of the content of some single person’s thought or talk, while a social or shared concept constitutes part of the content of a plurality of people’s talk and thought. In either of these psychological senses, concepts are psychological or linguistic contents. Hypotheses  To illustrate why many kinds, such as university, might be better conceived of as anthropogenic rather than natural regardless of their material manifestations, consider the kind planet. Planets seem entirely natural, so we may be tempted to regard planet as a natural kind. Yet in 2006, the number of planets in our solar system was reduced from nine to eight, due to an authoritative (re-) definition of ‘planet’ (Inman 2006). Such a change in the number of planets is only possible if planet is actually an anthropogenic kind. If, by contrast, planet were a natural kind, then reducing the number of planets in our solar system would require a Death Star from Star Wars (Lucas 1977).

4

88 

C. Rhode

expressing concepts in these psychological senses, then, would be hypotheses expressing elements of psycho-linguistic content rather than any aspect of the world beyond.

Cognition A further target of philosophical research is the variety of cognitive processes involved in one’s engagement with psycho-linguistic content and with reality beyond such content, as well as the potential impact of that same research for the instrumental suitability of philosophical methods (Alexander 2012: 3). This research, prominent in experimental philosophy as well as cognitive science, investigates hypotheses such as that Concept acquisition requires a significant number of concepts and/or cognitive systems that are both innate and mutually distinct (Laurence and Margolis 2015), or that Expert intuitions in philosophy are sensitive to experts’ linguistic background (Vaesen et al. 2013). Being concerned with such objects as innate concepts or experts’ intuitions, hypotheses in this group clearly express elements of psycho-linguistic content.5

Necessary Truth The last group of hypotheses to be taken up here concerns putatively necessary truths. Included in this group are hypotheses expressing modal equivalences (e.g. that Justification, truth, and belief are individually necessary and jointly sufficient for knowledge), hypotheses expressing metaphysical necessities (e.g. that Water necessarily is H2O), and hypotheses expressing conceptual, logical and mathematical truths (e.g. that P implies ¬¬P). Such hypotheses expressing necessary truths can be interpreted in either way with regard to the content/world separation (akin to hypotheses about natural or anthropogenic kinds). On the one hand, the necessity of their truth may be taken to be a metaphysically insignificant feature of our language, to the effect that the necessary truth of a 5  In light of the examples of Hollebrandse et al. (2014) and Frigg et al. (2014) in Appendix 3, some of these hypotheses are potentially investigated in types of dialogue other than persuasion dialogue, thus being in any case untouched by the present considerations.

4  Philosophical Methods between Content and the World 

89

hypothesis H should be read as According to certain aspects of our language or thought, necessarily H. In this case, hypotheses expressing the necessity of a given truth would be hypotheses expressing some explicit or implied content of language, rather than any aspect of the world beyond. On the other hand, the necessity of such a truth may be taken to be metaphysically significant, most obviously so if necessary truths are taken to require the (non-fictional) existence of possible worlds. In that case, hypotheses expressing the necessity of a given truth would be hypotheses expressing an aspect of the world beyond psycho-linguistic content rather than expressing any element of such content. To summarise, hypotheses of the various types outlined here can be divided into types of hypotheses expressing an element of psycho-­ linguistic content, and types of hypotheses expressing an aspect of reality beyond such content. Hypotheses expressing elements of psycho-­ linguistic content include those concerning psychological or linguistic subject matter, concepts in the (personal or social) psychological sense, anthropogenic kinds (as opposed to natural kinds), cognitive processes, and necessary truths in the metaphysically insignificant sense. Hypotheses expressing aspects of the world include those concerning non-­ psychological or extra-linguistic entities, Platonic forms or concepts in the Fregean sense, natural kinds, and necessary truths in the metaphysically significant sense. With the application of the content/world separation to hypotheses in philosophy thus illustrated, the next task is to similarly apply this inference barrier to different types of evidence, as well as to assess pairings of evidence and hypotheses against the content/world separation.

4.4 Evidence in Philosophy Evidence propositions, generically labelled ‘E’ as above, include for present purposes any proposition expressing some evidence and putatively standing in the prima facie supporting relation E:H to some hypothesis H (as defined at the start of the previous section). Consequently, evidence includes anything that is ever or has ever been invoked in support of

90 

C. Rhode

some hypothesis questioned in a philosophical dialogue, so it is obviously impossible to examine every one of these indefinitely many items of evidence. Instead, and in keeping with the approach taken in the previous section, I will only examine a non-exhaustive and non-exclusive selection of types of evidence that have been discussed in the recent literature in philosophical methodology. Bearing in mind that different types of evidence have not always been distinguished from methods incorporating such evidence, the selected types of evidence (and evidential methods) to be examined here include intuitions, common sense, ordinary language usage, theoretical virtues, the method of cases, reflective equilibrium, conceptual analysis, and methods employed in experimental philosophy. I will examine each of these types (or methods) with respect to whether their evidence propositions express elements of psycho-linguistic content or aspects of the world beyond such content. As this question is largely unrelated to most current methodological questions about the selected types of evidence, I will not delve into those latter questions in much if any detail when discussing these selected types of evidence, beginning with intuitions.

Intuitions The questions surrounding intuitions mainly concern (a) their evidential centrality to philosophy, (b) the epistemological appropriateness of that putative centrality, and (c) their exact nature. The present work effectively sidesteps each of these three current debates, however; so let me just briefly explain why responding to these issues is unnecessary for present purposes. First, the centrality of intuitional evidence to analytic philosophy has been challenged perhaps most prominently by Herman Cappelen (2012), who provides several close analyses of philosophical works and finds that intuitions as evidence are not central to such works at all. I suspect (with David Chalmers 2014) that an unduly narrow conception of intuition does at least some of the work for Cappelen. But even if Cappelen’s rejection of the centrality of intuitions was granted, the present considerations remain unaffected by it, since the present considerations do not solely concern intuitions. Rather, they concern a variety of

4  Philosophical Methods between Content and the World 

91

methods, at least a significant number of which does actually find widespread employment in analytic philosophy (as I will illustrate below). Second (and assuming the centrality rejected by Cappelen), various challenges to the evidential employment of intuitions in philosophy have been raised and countered by a similarly numerous range of defences of intuitional evidence. Robert Cummins (1998), for instance, argued that the reliability of intuitions, as much as of any other source of evidence, must be demonstrated by way of calibration. Yet, Cummins argues, the calibration of intuitions is either impossible or useless, respectively due to the absence or presence of an independently trusted source of evidence against which to calibrate intuitions. (Absent such evidence, calibration is impossible; yet given such evidence, why calibrate?) An objection to this argument (already indicated at the end of Cummins’s article) is that the demand for calibration would generalise across any source of evidence whatsoever, to the effect that the reliability of not only intuitions but also perception, memory, etc. would need to be but cannot be demonstrated by way of calibration. The demand for calibration, so the objection, would thus engender global epistemological scepticism. In a different challenge to intuitional evidence, Jonathan Weinberg, Shaun Nichols, and Stephen Stich (2001) have found experimentally that at least among non-philosophers some intuitions are sensitive to epistemologically irrelevant factors, such as the intuiter’s cultural background. In response, it has been claimed that the expertise of trained philosophers safeguards them from epistemologically irrelevant biases (Nado 2014b). (I will return to both calibration and experimental philosophy in more detail below.) What the vast majority of challenges to and defences of intuitional evidence have in common, though, is their exclusive concern with the epistemic status of intuitions.6 This is exactly a concern that I do not share, as my sole regard is with the dialogical role of intuitions, and specifically with the question of whether that role is affected by dialogical empiricism and the content/world separation. Moreover, conclusions drawn with regard to the epistemic status of intuitions could only be claimed to inform the dialogical employment of intuitions if one were to  Exceptions to this pattern include Joshua Earlenbaugh and Bernard Molyneux (2009) as well as Cappelen (2012).

6

92 

C. Rhode

claim that this epistemic status has implications for such dialogical employment—but that claim would require breaching the attitude/dialogue separation. Third, several different accounts have been advocated regarding the nature of intuitions. The accounts offered variously hold that intuitions are beliefs, that they are dispositions to believe, that they are judgements, that they are sui generis attitudinal states, and that they are the outputs of a cognitive faculty of intuition (cf. Lycan 2011; Fischer and Collins 2015). These various accounts might yet all turn out to be accurate to some extent, as ‘intuition’ seems to denote a set of separate and very diverse psychological states that are more or less compatible with different accounts of the nature of intuitions (Nado 2014a). It is agreed by all accounts, however, that intuitions are somehow psychological, in a sense that can be clarified by attending to the distinction between the intuited and the intuiting. The intuited is the content of a given intuition—that which has been intuited. The intuiting is the psychological event of intuiting that same content, that is, the occurrence of finding said content to be intuitive. As noted earlier, both the content and the occurrence of psychological events like intuitions are elements of psycho-linguistic content. Hence, intuitions constitute elements of psycho-linguistic content. To see exactly how the content/world separation pertains to intuitional evidence, then, it is worth recalling the dialogical situation at issue in this work: Party a to philosophical dialogue δ has asserted hypothesis H, and party b to philosophical dialogue δ has questioned hypothesis H. According to dialogical empiricism, a must prudently support H to b in δ by asserting some E and E:H, without thereby crossing any inference barrier, an example of which is the content/world separation. Since the evidence expressed by E in the case of intuitions constitutes psycho-­ linguistic content, the import of the content/world separation turns on the question of whether H also expresses an element of psycho-linguistic content or rather an aspect of the world beyond such content. If a given hypothesis H expresses an element of psycho-linguistic content, then the employment of the present E and E:H does not cross the content/world barrier and is thus innocuous (at least with regard to that particular barrier). If H expresses an aspect of the world beyond any psycho-linguistic content, however, then invoking an intuition in support of H amounts to

4  Philosophical Methods between Content and the World 

93

crossing the content/world separation, in contravention of dialogical empiricism. It is thus prudently prohibited in philosophical dialogue to employ an intuition in support of a questioned hypothesis H expressing an aspect of reality beyond any psycho-linguistic content, such as a hypothesis about a Platonic form or a natural kind.7 This is the first concrete methodological restriction to result from dialogical empiricism. More will follow in the examination of other types of evidence, the next one of which addresses common sense.

Common Sense Reference to common sense is frequently made in philosophy in a way that makes it quite interchangeable with reference to intuitions. Indeed, common sense as a type of evidence shares with intuitions the supposed characteristics of being very widely believed among non-philosophers within one’s particular community, not having been accepted as a result of reflection or study, and being affirmed effectively instantaneously when raised (Jenkins 2014: 92-4). There are, however, some supposed characteristics of (some) intuitions not necessarily shared by common-­ sense beliefs, such as that of seeming true by necessity and being grounded in linguistic competence (Jenkins 2014: 93). An important feature for present purposes that is shared with intuitions is the psycho-linguistic location of common sense: Both a given element of the content of common sense, and the commonsensicality of said element, are contained within one or more subjects’ psychological states and/or linguistic practices. Thus, common sense, just like intuitions, constitutes elements of psycho-linguistic content. It follows correspondingly that the import of the content/world separation for the evidential employment of common sense in philosophical dialogue turns on the question of whether H expresses an element of psycho-linguistic  Notably, Earlenbaugh and Molyneux (2009) also hold that this barrier-crossing evidential employment of intuitions is persuasively bound to fail. They, like Cappelen (2012), consider this problem to be resolved by the rejection of centrality. As indicated, however, my present concern is not restricted to intuitions but encompasses a variety of methods, whose combined centrality will become evident principally in the remainder of this section, and certainly in the remainder of Chaps. 4 and 5.

7

94 

C. Rhode

content or an aspect of the world beyond. In the former case, the content/world separation does not have any implications for the foregoing evidential employment of common sense. But in the latter case, the content/world separation prudentially bars the employment of common sense: If the questioned hypothesis H expresses an aspect of the world beyond any psycho-linguistic content, then common sense must prudently not be invoked as evidence in support of H.

Ordinary Language Usage Next, consider evidence provided by ordinary language usage. Ordinary language philosophers held, roughly, that the inappropriate use of a term (in the ears of ordinary speakers of a language) robs its containing sentence of meaning. Thus, when a sentence ‘sounds odd or jarring to the ear of an English gentleman’ (which was at the time considered the epitome of an ‘ordinary’ speaker of the English language; Daly 2015: 10), then this was taken to indicate that the sentence is really meaningless. The evidence, in this case, consists in verdicts about language—and neither the verdict nor the piece of language it is about can be found in the world beyond psycho-linguistic content. Hence, the evidence is constituted by psycho-linguistic content, and an evidence proposition expressing it would thus express such content. Given the content/world barrier, that evidence proposition can be employed in support of hypotheses expressing an element of psycho-linguistic content but not in support of hypotheses expressing an aspect of reality beyond such content.

Theoretical Virtues The next type of evidence to be examined here is the instantiation of theoretical virtues. A theoretical virtue is an ostensibly desirable attribute of a hypothesis or theory, though the basis of the desirability of such an attribute depends on the virtue at issue. While some theoretical virtues are considered to be truth-indicative, other theoretical virtues might be associated with the fruitfulness of a hypothesis instead. (The latter

4  Philosophical Methods between Content and the World 

95

include, among others, greater scope of a theory’s predictions and greater potential explanatory power over as yet unobtained evidence; Douglas 2013.) The instantiation of virtues associated with fruitfulness is not employed in order to support a questioned hypothesis (and indeed it would seem frivolous to attempt such employment), so these virtues are of no present concern. My concern here is only with theoretical virtues that have been or can be invoked in support of a hypothesis questioned by another dialogue party, where a theoretical virtue is understood as supporting the acceptance of a given theory (potentially over a rival theory) insofar as that theory instantiates this virtue (to a greater degree than its rival does). What, then, are these theoretical virtues that might be invoked in support of a questioned theory instantiating them—and is their capacity to provide such support barred by the content/world separation? Three potential theoretical virtues have already been discussed, viz. consistency with intuitions, consistency with common sense, and consistency with ordinary language usage. I will very briefly assess a selection of further putative virtues against the content/world barrier (ignoring other issues with these putative virtues; cf. Douglas 2013). Consistency with prior beliefs (cf. Daly 2015: 13-4): Where this virtue is employed, the evidence consists in the non-contradictory relation between the set of beliefs already present among someone’s attitudinal commitments and a potential new belief. Any proposition expressing this evidence (regardless of whether that be the respective beliefs or their mutual relations) would be expressing an element of psycho-linguistic content. The restriction found to apply to intuitions and common sense thus applies analogously. Empirical adequacy (Douglas 2013): The evidence in this case consists in empirically observed states of affairs, and would be expressed by a proposition expressing an aspect of reality beyond psycho-linguistic content. Such a proposition can therefore be employed in support of a questioned hypothesis expressing an aspect of reality beyond content without crossing the content/world separation. (Note that the evidence here does not consist in the relation between empirically observed states of affairs and a theory, even though the present theoretical virtue strictly identifies this relation. Hence, the evidence here does not consist in the theoretical virtue but is merely highlighted by it.)

96 

C. Rhode

Unification of diverse evidence (Douglas 2013): Here the evidence consists, not in the evidence that is unified, but in the great explanatory power of the instantiating theory across a wide scope of evidence. That is, the low number of theories (viz. one) needed to explain the diverse evidence itself constitutes evidence. How many theories are required in order to account for some given evidence (whether that evidence be an element of psycho-linguistic content or an aspect of reality beyond such content), however, is a product of the ways in which we make sense of evidence, and ways of sense-making are themselves implied in our language or shared thoughts. Hence, a proposition expressing the low number of required theories at issue here would be expressing an element of psycho-linguistic content. Consequently, such an evidence proposition can be employed in support of a hypothesis equally expressing an element of psycho-linguistic content, but not in support of a hypothesis expressing an aspect of reality beyond such content, due to the content/world barrier. Ontological parsimony (Baker 2013; cf. Keele 2010: 91-94): This essentially comparative virtue is instantiated by a theory insofar as this theory posits fewer kinds of existing entities than a rival theory does. The evidence provided by this virtue neither consists in the kinds of entities posited by the theory, nor in the number of these kinds, but in the fact that the number of kinds which the theory is positing is comparatively low (vis-à-vis a rival theory). The emphasis, however, must be on positing: To posit the existence of a kind of entity is to commit to that existence. A commitment, as explained above, can be attitudinal or dialogical but is invariably located in psycho-linguistic content. As such, a proposition expressing the evidence provided by the present theoretical virtue would be expressing an element of psycho-linguistic content. This evidence proposition could be employed in support of a hypothesis that also expresses an element of psycho-linguistic content, but not in support of a hypothesis expressing an aspect of reality beyond such content, as the latter evidential relation is barred by the content/world separation.

4  Philosophical Methods between Content and the World 

97

Method of Cases Following these basic types of evidence, I now turn to evidential methods that incorporate such evidence—beginning with the method of cases, where a ‘case’ is simply the subject of a thought experiment.8 Here, a hypothetical situation is presented, with respect to which (a) a given theory at issue implies some proposition P, and (b) we have the (strong) intuition that ¬P. This intuition is taken to constitute (strong) evidence against the theory at issue. To briefly illustrate this, simply consider one of the original Gettier cases (1963: 122). The theory at issue states that a subject s knows that P iff P is true, s believes that P, and s is justified in believing that P. So let Smith and Jones be two applicants for the same job on the day of their interviews. Smith has seen that Jones has ten coins in their pocket, and Smith has been informed by the company director that Jones will get the job. So Smith infers that J: The applicant who gets the job has ten coins in their pocket. Smith believes that J, Smith is justified in believing that J, and, as it turns out, J is true—not because Jones gets the job, but because Smith themselves gets the job and also happens to have ten coins in their pocket. Intuitively, Smith did not know that J, yet the theory at issue implies that they knew that J. The intuition that Smith does not know that J is thus employed as evidence against the theory whose implication is inconsistent with this intuition. As this example illustrates, the method of cases is distinguished from other methods by its employment of thought experiments. It does not appear to be distinguished, however, by the employment of a type of evidence peculiar to this method. Rather, it seems to employ intuitional evidence, which I have already examined above. And yet it might perhaps be argued successfully (albeit, I suspect, with some difficulty) that the nature of intuitions must be understood very narrowly, to the effect that  The term ‘thought experiment’, in contrast to the term ‘method of cases’, is also familiar to certain scientists, as they too employ thought experiments. It has been observed that ‘thought experiment’ lends an air of scientific credibility to the present philosophical method that ‘method of cases’ does not offer, which may be why the former label is preferred in the literature in philosophical methodology (Pust 2016). My present considerations do not apply to the sciences in any case (as I have not established that the sciences are predominantly conducted through persuasion dialogue), hence such terminological or even methodological similarity or dissimilarity with the sciences is without present import. 8

98 

C. Rhode

the method of cases does not in fact employ intuitions as evidence, and instead employs some other psychological event as evidence in support of a given proposition. However, as long as this other psychological event remains just that—a content or occurrence contained within the totality of someone’s present psychological activity (or linguistic practice), and thus an element of psycho-linguistic content—the same restriction will apply to it as has been found to apply to intuitions.

Reflective Equilibrium The method of cases can itself be embedded within a further method, viz. that of reflective equilibrium (though this label may denote either the method in general or its outcome in particular). The method of reflective equilibrium proceeds by subjecting a theory to intuitive counter-­examples (that is, by applying the method of cases) repeatedly while taking into account the theory’s instantiation of theoretical virtues. With every new counter-example, it must be decided whether to refine the theory in order to accommodate the counter-example or else to retain the theory and disregard the counter-example—though the latter is prima facie taken to constitute a cost rather than a benefit, so unaccommodated counter-­ examples of sufficient number and/or strength will render a theory unacceptable. The purpose of this method is to find a sufficiently refined and virtuous theory to which no further counter-example can be provided, in which case the theory would have reached a position of equilibrium following this extensive philosophical reflection: a reflective equilibrium. Another way to describe the method of reflective equilibrium is, simply, as the iterative application of the method of cases with consideration of theoretical virtues, until the lack of further intuitive counter-examples makes it impossible to apply the method of cases. As such, the method of reflective equilibrium works in a manner analogous to the elimination of competing theories in science, where a theory is subjected to empirical testing and refined (or abandoned) when faced with falsifying evidence (that cannot be explained away). A theory’s being in reflective equilibrium (and also to some extent a scientific theory’s having survived empirical testing) is typically taken to constitute significant indication of the

4  Philosophical Methods between Content and the World 

99

truth of the theory. However, I am not concerned with the question of whether a theory’s being in reflective equilibrium is truth-indicative, but rather with the reflective process of reaching equilibrium in the first instance. As reflective equilibrium is fundamentally an iterated version of the method of cases (with consideration of theoretical virtues), the restriction found to apply to the former will also apply to reflective equilibrium. The evidence employed in reflective equilibrium, just as in the method of cases, consists in either an intuition or some other psychological or linguistic event that is then employed as the verdict in a putative counter-­example to the given theory at issue. In light of the content/world barrier, such evidence can only be employed if the theory or hypothesis thus countered expresses an element of psycho-linguistic content.9 The method of reflective equilibrium thus cannot be employed in support of (or against) a hypothesis expressing an aspect of reality beyond psycho-­linguistic content.

Conceptual Analysis and the Canberra Plan Reflective equilibrium can be pursued in principle for any type of hypothesis; but the perhaps most quintessentially analytic philosophical hypotheses to be refined through the method of reflective equilibrium are conceptual analyses. Traditionally, a successful analysis of a concept is a biconditional proposition, with the concept (in its most basic employment in a proposition) on the one side, and the individually necessary and jointly sufficient conditions of the concept on the other side. The pre-Gettier conceptual analysis of the concept knowledge, for instance, was roughly that: S knows that P iff P is true, s believes that P, and s is justified in believing that P. The evidence employed in pursuit of conceptual analyses (or in the so-called method of conceptual analysis) remains the same as in the method of cases and in reflective equilibrium, viz. the psycho-linguistic contents that are intuitions (or relevantly similar 9  On some accounts (cf. Scanlon 2003) this is indeed the only purpose of the method of reflective equilibrium, such that a theory’s being in reflective equilibrium is of merely elucidatory value, clarifying one’s (individual or collective) commitments. I will return to this possibility in Chap. 6, Sect. 6.3.

100 

C. Rhode

psychological or linguistic events). Whether such evidence can be employed in support of the hypotheses at issue, that is, of the conceptual analyses, is dependent on whether the concepts involved are taken to be psycho-­linguistic contents (e.g. as anthropogenic kinds, or as concepts in the psychological senses) or aspects of reality beyond such content (e.g. as natural kinds, or as concepts in the Fregean sense). An analysis of a concept in which the concept is understood as an element of psycho-­linguistic content is a hypothesis expressing an element of psycho-linguistic content. Such an analysis can of course be supported by evidence propositions expressing elements of psycho-linguistic content, for example, intuitions. However, an analysis of a concept in which the concept is understood as an aspect of reality beyond any psycho-linguistic content is a hypothesis expressing an aspect of reality beyond content. An analysis of this sort cannot be supported by evidence propositions expressing intuitions or any other elements of psycho-linguistic content, as this evidential relation would cross the content/world separation. Perhaps the restriction on content/world inferences via conceptual analyses can be circumvented by turning away from the traditional method of conceptual analysis just outlined, and opting instead for the Canberra Plan of conceptual analysis. The aim of the Canberra Plan is not so much to analyse a given concept as to identify the referent, if any, of a given concept (cf. Braddon-Mitchell and Nola 2009; Papineau 2015).10 To this end, a Canberra-style treatment of a given target concept proceeds in two steps. In the first step, a wide range of intuitions or common sense outputs about or involving the concept are gathered, thus collecting a variety of predicates which the concept intuitively or commonsensically satisfies. The collected predicates are then conjoined in a Ramsey sentence, whereby a Ramsey sentence states that there exists some unique kind that satisfies this conjunction of predicates. (In effect, these predicates function as the individually necessary and jointly sufficient conditions for the identification of a unique kind satisfying them.) In the second step, one turns to the sciences in order to find out whether the Ramsey sentence is indeed made true by some unique kind, and if it is,  Calling this programme a version of conceptual analysis may thus be a misnomer, and ‘conceptual referent identification’ may be more descriptively adequate. 10

4  Philosophical Methods between Content and the World 

101

what that kind is, thus identifying the unique referent (if any) of the target concept under examination. Yet the Canberra Plan does not serve as a means of crossing the content/world barrier, since the Plan itself does not involve any crossing of this barrier in persuasion dialogue. A successful Canberra-style analysis of a concept is simply a Ramsey sentence that is found to be true. Accordingly, in a persuasion dialogue, the hypothesis at issue would be expressed by a Ramsey sentence, and the support provided for this hypothesis would be whatever is invoked in finding that Ramsey sentence to be true. What is so invoked are not the predicates collected from intuitions and common sense, of course. These predicates are employed in formulating the Ramsey sentence, so it would be persuasively useless to invoke them in support of that same sentence. What is invoked in finding the Ramsey sentence to be true is, rather, the support provided through scientific investigation. Does such support, then, cross the content/world barrier? If the kind that is characterised in the Ramsey sentence is a natural kind, then the scientific support for the Ramsey sentence would almost certainly consist in empirical observations of nature, which are expressed in propositions expressing aspects of reality beyond any psycho-linguistic content. In this case, propositions expressing aspects of reality beyond psycho-linguistic content would be employed in support of a hypothesis (expressed by the Ramsey sentence) expressing aspects of reality beyond psycho-linguistic content—so the content/world separation would remain untouched. And if the kind characterised in the Ramsey sentence is an anthropogenic kind, then the scientific support for the Ramsey sentence would be similarly likely to consist in results obtained in psychology, sociology, or linguistics through the study of thought and talk, and such results would be expressed by propositions expressing elements of psycho-linguistic content. Hence, propositions expressing elements of psycho-linguistic content would be employed in support of a hypothesis (expressed by the Ramsey sentence) expressing elements of psycho-­ linguistic content—and the content/world separation would again remain untouched. Either way, then, the Canberra Plan does not offer an avenue for conceptual analysis to circumvent the content/world barrier, as it never actually crosses that barrier.

102 

C. Rhode

Experimental Philosophy Of increasingly similar philosophical prominence to conceptual analysis is the programme called experimental philosophy. Among the various projects pursued under this heading, three are sometimes considered central. These include conceptual analysis, methodological critique, and cognitive science (Alexander 2012: 2-3; cf. Fischer and Collins 2015: 21-3). As I will outline presently, these projects largely investigate elements of psycho-­linguistic content and are thus at no risk of crossing the content/ world barrier with the employment of psycho-linguistic evidence. Within the project of conceptual analysis, experimental philosophers share the same goals as philosophers in traditional conceptual analysis. The key difference is that they do not draw on their own or their colleagues’ intuitions as evidence, but rather conduct surveys in order to gather intuitional verdicts about hypothetical scenarios, particularly from people without philosophical training. Experimental philosophers then use their findings in order to inform the selection and refinement of philosophical theories within the pursuit of reflective equilibrium. As such, the foregoing restrictions on conceptual analysis, reflective equilibrium and, ultimately, intuitions equally apply to this x-phi project. The intuitions gathered through surveys can be employed (in an evidence proposition expressing them) in support of a hypothesis expressing an element of psycho-linguistic content; but they cannot be employed in support of a hypothesis expressing an aspect of reality beyond any psycho-linguistic content. The latter employment of these survey results would cross the content/world separation. The intuitions of non-philosophers thus surveyed have been found to differ markedly in many cases from what trained philosophers considered intuitively obvious, while furthermore being sensitive to philosophically supposedly irrelevant situational and demographic variables (some of which have been found to be effective among philosophers too). The ensuing project of methodological critique has undermined the assumption that people’s intuitions are by and large homogenous, raising questions about the evidential value of philosophers’ intuitions. Drawing on people’s intuitions as evidence, this project supports its conclusions with elements of psycho-linguistic content; hence, these conclusions must

4  Philosophical Methods between Content and the World 

103

prudently not express aspects of the world beyond such content. (Indeed, they do not express such aspects of the world, as the conclusions thus drawn only concern the same psycho-linguistic content, viz. people’s— and philosophers’—intuitions.) In a further project, experimental philosophers have augmented their own research with studies from, and they have conducted studies in, psychology and other cognitive sciences. This is partly in order to understand the aetiology of intuitions and perhaps identify the conditions under which intuitions do constitute reliable evidence, and partly to more generally contribute to the study of our cognitive processes. Whether the object of study be intuitions or other psychological events or processes, this object of study is equally part of psycho-linguistic content, and hypotheses about the same are supported here through direct or indirect evidence of talk or thought. Hence, such studies are at no risk of crossing the content/world barrier.11 To summarise, in the present section we have examined eight different groups of evidence (or methods) that may be employed in support of a questioned hypothesis. As discussed, intuitions, common sense and ordinary language usage, as well as the evidence employed in three other theoretical virtues, all constitute psycho-linguistic content, while evidence employed in the theoretical virtue of empirical adequacy constitutes aspects of reality beyond such content. The method of cases employs intuitions (or similar psycho-linguistic content) as its evidence, and reflective equilibrium as a method is merely an iterated version of the method of cases with consideration of theoretical virtues, thus employing the same evidence. Conceptual analysis is but a particular instance of the method of reflective equilibrium and employs the same evidence, while experimental philosophy employs various psycho-linguistic content as evidence. In the previous section, we have already examined different types of hypotheses, finding that hypotheses variously express either an element of psycho-linguistic content or an aspect of reality beyond such content.  Some such studies do not seem to engage in persuasion dialogue but rather in inquiry or discovery dialogue, and would thus in any case remain unaffected by these considerations. Cf. fn. 5 in Sect. 4.3.

11

104 

C. Rhode

Beyond the outline of the content/world separation, the foregoing examination of different types of hypotheses and different types of evidence has hopefully illustrated in some detail the application of this inference barrier. We have now seen various examples of evidence propositions expressing elements of psycho-linguistic content, and of hypotheses expressing aspects of reality beyond such content, whose inferential pairing is foreclosed by the content/world separation. That is to say, in conclusion: Psycho-linguistic content expressed by an evidence proposition (including intuitions, common sense, ordinary language usage and certain theoretical virtues—whether by themselves or within methods such as the method of cases, reflective equilibrium, conceptual analysis, or within experimental philosophy) cannot prudently be employed in support of a questioned hypothesis if this hypothesis expresses an aspect of reality beyond psycho-­ linguistic content (including aspects of non-psychological or extra-linguistic entities, Platonic forms, concepts in the Fregean sense, natural kinds, and necessary truths in the metaphysically significant sense). I could not pretend that the examination in this chapter took account of all possible pairings of types of evidence and hypotheses (lest the foregoing statement of its conclusion would be even longer!). Nor do I claim that the inferences not restricted by the content/world separation are entirely unproblematic. Neither of these claims would follow from the above discussions. Furthermore, the content/world separation is but one of various inference barriers that could be established from dialogical empiricism. Other inference barriers include the is/ought separation (i.e. Hume’s First Law) and the contingent/necessary separation (i.e. Kant’s Law; see Chap. 3), which I will return to in the next chapter. Yet before moving on to that examination, I want to discuss a small selection of objections that one might raise against the content/world separation and/ or its significance for philosophical methodology.

4.5 Against the Content/World Separation Given the metaphilosophical scope of my considerations in this chapter, a small mountain of objections could likely be raised against the content/ world separation, only a handful of which I can discuss here. I hope,

4  Philosophical Methods between Content and the World 

105

however, that my following responses to the selected objections will provide sufficient indication of responses that could be given to objections not discussed here. The objections to be addressed concern the easy approach to ontology, content externalism, agential realism, calibration, and mentalism.

The Easy Approach From the content/world separation it follows that no conclusion about reality beyond psycho-linguistic content may prudently be derived from premises expressing such content. It might be objected, however, that it is perfectly unproblematic and easy to derive such an ontologically significant conclusion from an innocuous proposition expressing an element of psycho-linguistic content. Consider, for instance, the following argument: P1: The first sentence of the present paragraph contains 23 words. C1: The number of words in the first sentence of the present paragraph is 23. C2: There are numbers. This effectively neo-Fregean line of argument is an example of what Amie Thomasson (e.g. in her 2014) calls the easy approach to ontology. Of course, if C2 expresses an element of psycho-linguistic content, then C2 may indeed be supported by P1 according to the content/world separation. Yet the point pursued by Thomasson is that even if C2 expresses an aspect of reality beyond psycho-linguistic content and is thus ontologically significant (which she would hold it to be), C2 is nonetheless also necessitated by P1, with the effect that P1 may well be employed in support of C2. As such, Thomasson claims that P1 expresses an element of psycho-linguistic content and implies C2, whereby C2 expresses as aspect of reality beyond such content. Underlying this inference is Thomasson’s view of the relation between existence and reference. Thomasson effectively holds that K1: Things of a certain kind k exist iff the term ‘k’ refers, as well as that K2: The term ‘k’ refers iff ‘k’ is used in accordance with the rules for the proper use of ‘k’ (Thomasson 2008). In effect, Thomasson thus holds that K: Things of a certain kind k

106 

C. Rhode

exist iff the term ‘k’ is used in accordance with the rules for the proper use of ‘k’. Hence, if the term ‘number’ is used in accordance with the rules for the proper use of ‘number’, then numbers exist. The usage of the term ‘number’ in C1 very much seems to be in accordance with these rules— so according to K, C1 implies that there are numbers, that is, it implies C2. (I here grant that P1 implies C1.) Given K, furthermore, C1 expresses an element of psycho-linguistic content: Any term (including ‘number’) obviously constitutes such an element of content (and the subject of C1 is in any case a particular element of language, viz. the words in a sentence). At the same time, Thomasson holds that C2 expresses an aspect of the world beyond any such content. So in essence, K acts as a bridge premise in the inference from the content-located C1 to the world-­ located C2. And therein lies the problem for Thomasson’s easy approach. The easy approach depends upon K; but K crosses the content/world barrier, rendering K effectively unsupportable in philosophical persuasion dialogue. Employing K in the easy approach thus assumes that crossing the content/world separation is not persuasively imprudent after all, and hence, that this persuasion barrier does not hold. Consequently, Thomasson’s easy approach is not an objection to this barrier but merely a denial of the barrier. That is to say, employing Thomasson’s easy approach as an objection to the content/world separation simply begs the question against this inference barrier, and does absolutely nothing to address the demonstrated persuasive imprudence of inferring across this separation in philosophical dialogue. (This issue not least implies that the easy approach fails not just as an objection but altogether within philosophical dialogue, which is after all its only context of application.)

Content Externalism One might wonder whether the content/world barrier could instead be undermined by externalism about psycho-linguistic content, in which case the individuation of such content would depend at least partly upon circumstances in the world, outside of the human bearers of psycho-­ linguistic content (Lau and Deutsch 2014). This would seem to provide

4  Philosophical Methods between Content and the World 

107

a link across the content/world separation, as content would depend at least in part upon the world. The most prominent argument for such externalism is of course Hilary Putnam’s (1975) Twin Earth thought experiment, which also serves to illustrate the thesis for present purposes. Imagine that water is, and has all its familiar properties as, H2O; and further imagine a time when the chemical composition of water was not yet known. There exists, in this thought experiment, a Twin Earth that is identical to our Earth except that the chemical composition of what is there referred to as ‘water’ is actually XYZ.  Now, since water is really H2O, it intuitively follows that the people on Twin Earth are not actually referring to water when they talk about what they call ‘water’. As such, the meaning of ‘water’ depends upon the chemical composition of what is referred to as ‘water’, thus depending upon circumstances in the world, external to the bearers of psycho-linguistic content. Yet if that is so, then content is connected to, indeed determined (at least partly) by, the world, thus seemingly undermining the separation between the two. The problem with invoking content externalism in an attempt to undermine the content/world separation is that content externalism and the content/world separation are perfectly consistent, as they respectively concern different sorts of relations (very much as already discussed in Sect. 4.2 with regard to the mind-dependent/mind-independent distinction). Content externalism is a semantic thesis about the supervenience of content on the world, whereas the content/world separation is a functional thesis about persuasively effective inference from content to the world. One might nonetheless try to furnish content externalism into a bridge premise across the content/world barrier for use in a philosophical persuasion dialogue. What would that look like? Using Twin Earth again, consider the meaning of ‘water’ prior to the discovery of the chemical composition of water, which was roughly that which has all the familiar properties of what we refer to as ‘water’. Could we, at that time, have employed the meaning of ‘water’ in support of the hypothesis that water is H2O? That would be plainly unpersuasive to anyone who has already questioned said hypothesis, as they would simply go on and question the implicit premise that the meaning of ‘water’ is even relevant to the nature of water. What if we added the premise that the meaning of ‘water’ was

108 

C. Rhode

at least partly determined by the nature of water, and we explained as much by way of the Twin Earth thought experiment? That would again be unpersuasive, as the chemical composition of water may perhaps determine one person’s meaning of ‘water’, but someone’s meaning of ‘water’ does not thereby determine the chemical composition of water. If it did, then someone’s use of the word ‘Earth’ to mean that which is flat upon which we all live would thereby determine that the Earth is flat, and the seismic upheaval of the instant flattening of the planet would probably end all life on its surface.12 Content/world supervenience is thus simply ineffective in bridging the content/world inference barrier.

Agential Realism A further potential line of argument towards a rejection of the content/ world separation is provided by the agential realism developed by Karen Barad (2007). Barad draws significantly on the work of Niels Bohr, who explained that it is a mistake to think of quantum-level particles as ‘objects’ that are separate from and independent of ‘agencies of observation’ (roughly: acts of and tools for observation; Barad 2007: 31). Rather, the metaphysically primitive kind is the phenomenon, an intra-active relation whose ontologically non-independent relata are agencies of observation and objects.13 Barad constructs parallel accounts of various other applications, ranging across the natural and social sciences, thus generally concluding that any endeavour of observing an aspect of reality, that is, any epistemic performance, is not an inter-action between, but rather an intra-action among the agencies of observation and the aspect of reality under observation. Reality thus just is the totality of all phenomena, that

 We might further note that ‘that which has all the familiar properties of what we refer to as “water”’ is not actually H2O by itself, but a liquid of constantly forming, dissipating and reforming polymers such as (H2O)2 and (H2O)3 (Ross, Ladyman and Spurrett 2007: 21). 13  This is best illustrated by the double-slit experiment: A single electron travelling through either of two slits actually is but a probability wave travelling through both slits, only forced into an unambiguous location by the interference of observation on a screen behind the slits. The phenomenon, the particle, is thus emergent from the intra-action of the probability wave and the observation. 12

4  Philosophical Methods between Content and the World 

109

is, performances.14 The separation between the act of observing and the object observed, or more generally between psycho-linguistic content and the world beyond, is merely conceptual, emerging through the intra-­ acting of its ontologically non-independent relata. Just as the independence of energy and mass is conceptual but not metaphysical (since E=mc2), so the independence of psycho-linguistic content and the world beyond is a product of our conceptions that does not reflect metaphysical reality. Hence, the separation between psycho-linguistic content and the world beyond is really a misconception. Let me grant for present purposes that the content/world separation does indeed fail to capture what are fundamentally phenomena rather than independent objects and agencies of observation. However, the content/world separation is not a metaphysical thesis, as my concern is not with the exact nature of psycho-linguistic content or reality beyond it. Indeed, the content/world separation is entirely silent with regard to the fundamental metaphysical phenomenality or separateness of content and the world beyond. I treat content and the world as separate and independent because my considerations operate posterior to what Barad describes as the intra-active conceptual emergence of this separation. The content/ world barrier merely concerns issues of the prudent conduct of dialogue in light of that emergent conceptual separation, so invoking the metaphysical non-separateness of content and the world beyond against the content/world barrier misses its target.

Calibration Perhaps but not necessarily on the basis of some metaphysical connection like the one implied by agential realism, or some supervenient connection such as suggested by content externalism, it could be argued that certain elements of psycho-linguistic content reliably indicate certain aspects of reality beyond such content, thus overcoming the ­content/  Barad is at pains to distinguish her theory from the most extreme social constructivism that has acquired infamy in the science wars. Her theory is decidedly realist in its affirmation of the object, while emphasising the object’s dependence upon agency (in terms of both observation and conceptual delineation)—hence, agential realism. 14

110 

C. Rhode

world separation in persuasion dialogue. Of course the claim cannot be that, for any proposition P, the fact that P is part of one’s psycho-­linguistic content reliably indicates the metaphysical truth of P—for plainly, neither my belief that The Earth is flat nor my belief that The Earth is spherical would be indicative of the planet’s geometry. Rather, it would be necessary to determine for any given case or type of case whether the psycho-linguistic content at issue is a reliable indicator of an aspect of reality beyond any such content. Specifically, when invoking an element of psycho-linguistic content in support of a hypothesis expressing an aspect of reality beyond content, one would (given dialogical empiricism and the content/world barrier) very likely have to provide support for the claim that the invoked psycho-linguistic content is indeed indicative of the aspect of reality beyond psycho-linguistic content that is at issue. The seemingly most promising way to establish this reliable indicatorship to one’s opponent is to calibrate the source of the given psycho-­ linguistic content against some independently accepted source of evidence about at least relevantly similar aspects of reality beyond content. In the case of an intuition, for instance, the intuition would need to be calibrated against an independently accepted source of evidence. In order to calibrate a source of evidence about a given target, one would have to compare its output about that kind of target with the output of an independently accepted source of evidence about the same kind of target, and if necessary, tweak the source of evidence at issue until its outputs are systematically congruent with the outputs of the independently accepted source of evidence. To overcome the content/world separation in this way, one would thus have to calibrate the source of the psycho-linguistic content at issue against an independently accepted source of evidence about the same kind of aspect of reality; and for that source to be accepted independently of (effectively: any) psycho-linguistic content, its outputs would itself need to be an aspect of reality. As already indicated above, this approach fails in either of two ways, as pointed out by Robert Cummins (1998). If, on the one hand, an independently accepted source of evidence about the target at issue is not available, then calibration could simply not take place. Indeed, this seems

4  Philosophical Methods between Content and the World 

111

to be the case for most targets at issue in philosophical dialogues (such as knowledge, justice, necessity, beauty, meaning, etc.). Yet if, on the other hand, an independently accepted source of evidence about the given target is available, then the question arises why one should bother calibrating another source of evidence about that same target. If, say, one already accepts the general theory of relativity (due to its own empirical support) as a source of evidence about the nature of time, then what would be the point of trying to calibrate one’s pre-reflective intuitions about the nature of time? Moreover, it is far from clear how exactly a source like intuitions could just be ‘tweaked’ in order to alter its outputs, and thus it is far from clear whether such calibration efforts would ever be successful—which it seems entirely safe to doubt (cf. Weinberg et al. 2012). Calibrating the source of a given kind of psycho-linguistic content is thus either impossible, or seemingly pointless and rather hopeless, rendering such calibration practically ineffective as a way of overcoming the content/world barrier. Against demands for calibration, or generally for demonstrating the reliability of a source of evidence, it has of course been argued that such a demand would generalise (absent arbitrary and thus putatively impermissible restrictions) to any source of evidence whatsoever, thus leading to global epistemological scepticism. Cummins’s requirement for the calibration of intuition has already been widely rejected (e.g. by Goldman 2007) because of its alleged sceptical consequences. It is worth remembering, however, that the demand does not generalise all by itself in a persuasion dialogue. If one’s intuitions about the nature of time were calibrated successfully against the implications of general relativity, for instance, then it is not clear that a philosophical opponent would proceed to demand that general relativity be calibrated as well (though this further demand is of course possible). Furthermore, even if the demand did generalise across all sources of evidence, scepticism still would not follow from this generalisation. As explained repeatedly at earlier points in the present work, my concern is with the prudent conduct of dialogue, and no attitudinal thesis, not even a form of scepticism, follows from any thesis about such conduct.

112 

C. Rhode

Mentalism The last objection to the considerations in this chapter that I want to briefly discuss here grants the content/world separation but claims that this barrier is innocuous. As outlined earlier, Goldman and Pust (1998) argued against what they called extra-mentalism and for mentalism, with the latter encompassing ‘[those] views of philosophical analysis […] that take the targets of such analysis to be in-the-head psychological entities’ (1998: 183). To slightly broaden the scope of methods and targets, we can substitute mentalism by what might be called psycho-linguism, which encompasses those views of philosophical research that take the targets of such research to be psycho-linguistic contents. And we can equally substitute extra-mentalism by extra-psycho-linguism, where the latter correspondingly encompasses those views of philosophical research that take the targets of such research to be aspects of reality beyond any psycho-­ linguistic content. If psycho-linguism is correct with regard to all targets of philosophical research, then this would mean that aspects of reality beyond psycho-­ linguistic content are never among the targets of such research, and hence that the employment of propositionally expressed psycho-linguistic content as evidence about these targets never crosses the content/world separation. It is perhaps best, however, to interpret Goldman and Pust as holding that the targets of philosophical research should be restricted to psychological entities (or to psycho-linguistic contents), as there are examples aplenty of philosophers advocating hypotheses that very clearly express aspects of the world beyond any psycho-linguistic content. Just consider, for one, Crawford Elder (2004), who hypothesises about ‘the subatomic microparticles that future physics will discover to be the truly fundamental building blocks of the physical world’ (2004: 51; cf. Ross, Ladyman and Spurrett 2007: 23). It seems impossible to interpret Elder in any way other than as speculating about aspects of reality beyond psycho-­ linguistic content—and Elder is far from alone in debating hypotheses that express such aspects of the world. Further examples demonstrating the at least partial truth of extra-psycho-linguism will be raised in the next chapter, with regard to a selection of more substantive

4  Philosophical Methods between Content and the World 

113

philosophical debates. There I will examine how the selected substantive dialogues are affected by dialogical empiricism, the content/world separation, and the resultant restrictions on philosophical methods that have been derived in the present chapter.

References Alexander, J. 2012. Experimental Philosophy: An Introduction. Cambridge: Polity Press. Baker, A. 2013. Simplicity. In The Stanford Encyclopedia of Philosophy, Fall 2013 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/fall2013/entries/ simplicity/. Barad, K. 2007. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning. Durham, NC: Duke University Press. Baz, A. 2017. The Crisis of Method in Contemporary Analytic Philosophy. Oxford: Oxford University Press. Braddon-Mitchell, D., and R. Nola. 2009. Introducing the Canberra Plan. In Conceptual Analysis and Philosophical Naturalism, eds. D. Braddon-Mitchell and R. Nola, 1–20. Cambridge, MA: MIT Press. Brock, S., and E. Mares. 2007. Realism and Anti-Realism. Durham: Acumen. Cappelen, H. 2012. Philosophy without Intuitions. Oxford: Oxford University Press. Carnap, R. 1950. Empiricism, Semantics, and Ontology. Revue Internationale de Philosophie 4 (11): 20–40. Chalmers, D.J. 1996. The Conscious Mind: In Search of a Fundamental Theory. New York: Oxford University Press. ———. 2014. Intuitions in Philosophy: A Minimal Defense. Philosophical Studies 171: 535–544. Cohnitz, D., and S.  Häggqvist. 2009. The Role of Intuitions in Philosophy. Studia Philosophica Estonica 2 (2): i–xiv. Cummins, R. 1998. Reflections on Reflective Equilibrium. In Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry, eds. M.R.  DePaul and W.  Ramsey, 113–127. Lanham, MD: Rowman and Littlefield. Daly, C. 2015. Introduction and Historical Overview. In The Palgrave Handbook of Philosophical Methods, ed. C. Daly, 1–30. Basingstoke: Palgrave Macmillan.

114 

C. Rhode

Douglas, H. 2013. The Value of Cognitive Values. Philosophy of Science 80 (5): 796–806. Dyke, H. 2008. Metaphysics and the Representational Fallacy. New  York: Routledge. Earlenbaugh, J., and B. Molyneux. 2009. If Intuitions Must Be Evidential then Philosophy Is in Big Trouble. Studia Philosophica Estonica 2 (2): 35–53. Elder, C. 2004. Real Natures and Familiar Objects. Cambridge, MA: MIT Press. Fischer, E., and J.  Collins. 2015. Rationalism and Naturalism in the Age of Experimental Philosophy. In Experimental Philosophy, Rationalism, and Naturalism: Rethinking Philosophical Method, eds. E. Fischer and J. Collins, 3–33. New York: Routledge. Frigg, R., S. Bradley, H. Du, and L.A. Smith. 2014. Laplace’s Demon and the Adventures of His Apprentices. Philosophy of Science 81 (1): 31–59. Gettier, E.L. 1963. Is Justified True Belief Knowledge? Analysis 23 (6): 121–123. Goldman, A. 2007. Philosophical Intuitions: Their Target, Their Source, and Their Epistemic Status. Grazer Philosophische Studien 74: 1–26. Goldman, A., and J. Pust. 1998. Philosophical Theory and Intuitional Evidence. In Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry, eds. M.R. DePaul and W. Ramsey, 179–197. Lanham, MD: Rowman and Littlefield. Hollebrandse, B., A. van Hout, and P.  Hendriks. 2014. Children’s First and Second-order False-belief Reasoning in a Verbal and a Low-verbal Task. Synthese 191: 321–333. Inman, M. 2006. Pluto Not a Planet, Astronomers Rule. In National Geographic News. http://news.nationalgeographic.com/news/2006/08/060824-­pluto-­ planet.html. Accessed 17 June 2015. Jenkins, C.S.I. 2014. Intuition, ‘Intuition’, Concepts and the A Priori. In Intuitions, eds. A.R. Booth and D.P. Rowbottom, 91–117. Oxford: Oxford University Press. Keele, R. 2010. Ockham Explained: From Razor to Rebellion. Chicago: Open Court. Kornblith, H. 2014. Is There Room for Armchair Theorizing in Epistemology? In Philosophical Methodology: The Armchair or the Laboratory? ed. M.C. Haug, 195–216. New York: Routledge. Lau, J., and M.  Deutsch. 2014. Externalism About Mental Content. In The Stanford Encyclopedia of Philosophy, Summer 2014 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/sum2014/entries/content-­externalism/. Laurence, S., and E. Margolis. 2015. Concept Nativism and Neural Plasticity. In The Conceptual Mind: New Directions in the Study of Concepts, eds. S. Laurence and E. Margolis, 117–147. Cambridge, MA: MIT Press.

4  Philosophical Methods between Content and the World 

115

Lucas, G. dir. 1977. Star Wars [Film]. Universal City, CA: Lucasfilm. Lycan, W.G. 2011. Epistemology and the Role of Intuitions. In The Routledge Companion to Epistemology, eds. S.  Bernecker and D.  Pritchard, 813–822. New York: Routledge. Nado, J. 2014a. Why Intuition? Philosophy and Phenomenological Research 89 (1): 15–41. ———. 2014b. Philosophical Expertise. Philosophy Compass 9 (9): 631–641. Nicoli, S.M. 2016. The Role of Intuitions in Philosophical Methodology. London: Palgrave Macmillan. Papineau, D. 2015. Naturalism. In The Stanford Encyclopedia of Philosophy, Fall 2015 Edition, ed. E.N.  Zalta. http://plato.stanford.edu/archives/fall2015/ entries/naturalism/. Pust, J. 2016. Intuition. In The Stanford Encyclopedia of Philosophy, Spring 2016 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/spr2016/entries/ intuition/. Putnam, H. 1975. The Meaning of ‘meaning’. In Philosophical Papers, Vol. II: Mind, Language, and Reality, 215–271. Cambridge: Cambridge University Press. Ross, D., J. Ladyman, and D. Spurrett. 2007. In Defence of Scientism. In Every Thing Must Go: Metaphysics Naturalized, eds. J.  Ladyman, D.  Ross, D. Spurrett, and J. Collier, 1–65. Oxford: Oxford University Press. Rouse, J. 1996. Engaging Science: How to Understand Its Practices Philosophically. Ithaca, NY: Cornell University Press. Scanlon, T.M. 2003. Rawls on Justification. In The Cambridge Companion to Rawls, ed. S. Freeman, 139–167. Cambridge: Cambridge University Press. Thomasson, A.L. 2008. Existence Questions. Philosophical Studies 141: 63–78. ———. 2014. The Easy Approach to Ontology: A Defense. In Philosophical Methodology: The Armchair or the Laboratory? ed. M.C.  Haug, 107–125. New York: Routledge. Vaesen, K., M. Peterson, and B.V. Bezooijen. 2013. The Reliability of Armchair Intuitions. Metaphilosophy 44 (5): 559–578. Weinberg, J.M., S.  Crowley, C.  Gonnerman, I.  Vanderwalker, and S.  Swain. 2012. Intuition and Calibration. Essays in Philosophy 13 (1): 256–283. Weinberg, J.M., S.  Nichols, and S.  Stich. 2001. Normativity and Epistemic Intuitions. Philosophical Topics 29 (1–2): 429–460.

5 Metaphysical Hypotheses

5.1 Substantive Consequences Dialogical empiricism, as distilled at the end of Chap. 3 above, requires that a proposition asserted and questioned in philosophical dialogue be supported by the asserting party without crossing any inference barrier. One such inference barrier separates psycho-linguistic content from the world beyond such content, and on its basis, I have discussed certain methodological consequences of dialogical empiricism in the previous chapter. The present chapter will turn from methodological to substantive consequences of dialogical empiricism, which I want to discuss in light of not only the content/world separation (and its methodological implications) but also other inference barriers, including the is/ought separation and the contingent/necessary separation. As before, however, a complete account of the substantive consequences of dialogical empiricism cannot possibly be provided here. What I therefore hope to do in this chapter is to discuss the implications of dialogical empiricism for a selection of comparatively broad issues that cut across many different philosophical sub-disciplines. I will occasionally address particular implications arising from these broader issues, but © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_5

117

118 

C. Rhode

any further particular consequences beyond these occasions will only be indicated through the discussion of these broader questions. The selected broad issues to be examined here include normativity, necessity, epistemology, and ontology. Each of these broader issues will be discussed, respectively, in one of the following four sections. In Sect. 5.6 I will then draw a rather sweeping conclusion, based on both the discussion of the foregoing four broad issues and the conclusions of the preceding four chapters. This conclusion states that it is prudently necessary to excise from analytic philosophical dialogue any debate about hypotheses that express aspects of the world beyond content and find their only support in evidence propositions expressing elements of psycho-linguistic content. In effect, I will conclude in the present chapter that the vast majority of analytic philosophical debates about the world beyond content, particularly in the sub-discipline of analytic metaphysics, must prudently be terminated.

5.2 Normativity Dialogical empiricism bars inferences from propositions of one kind to propositions of a different kind. More specifically, within philosophical dialogue, dialogical empiricism gives rise to several barriers to the employment of evidence of one kind in support of a hypothesis of a different kind. One such barrier is the content/world separation discussed above, which bars the employment of psycho-linguistic content in support of hypotheses about the world beyond such content. Another barrier is the is/ought separation famously introduced by David Hume (1978 [1740]: 469-70). In its form as a persuasion barrier, the is/ought separation states that one must prudently not employ descriptive propositions (or propositions expressing descriptive evidence) in support of a normative hypothesis (or a hypothesis expressing a putative norm) that has been questioned in a philosophical dialogue. This separation clearly constitutes an inference barrier as opposed to an implication barrier: No increase in the quantity of descriptive evidence would render that evidence any more

5  Metaphysical Hypotheses 

119

salient with regard to the questioned norm, so descriptive evidence differs from normative hypotheses in kind rather than extent.1 Given the imprudence of adducing descriptive evidence in support of a questioned norm, then, how could one prudently support a norm that one’s interlocutor in a philosophical dialogue has questioned? The exact answer will depend on whether norms are elements of psycho-linguistic content or aspects of reality beyond such content, since the range of evidence prudently available in support of a questioned norm is sensitive to the location of that norm in relation to the content/world barrier, as indicated in the previous chapter. Yet rather than pronouncing on the nature of norms here, I will merely examine how a questioned norm could prudently be supported in either case—first on the assumption that norms are aspects of the world beyond any psycho-linguistic content, and then on the assumption that norms are elements of psycho-­ linguistic content. If norms are aspects of the world beyond psycho-linguistic content (and thus, if propositions expressing them express such aspects of the world), then both the content/world barrier and the is/ought barrier must be taken into account when selecting evidence to support a questioned norm. Consider, for instance, the employment of intuitional evidence in support of a questioned norm, as advocated most explicitly by ethical intuitionists. According to ethical intuitionism, ‘intuition can provide rather strong warrant for the normative judgments that are its ethically relevant upshot’ (van Roojen 2014: 149), and such warrant should be accounted for ‘in terms of a process or state which can plausibly be construed as hooking us up to a mind-independent ethical reality’ (Cowan 2015: 1107). Ethical intuitionists thus claim that intuitions provide evidence in support of norms that are aspects of the world beyond psycho-linguistic content. Yet the occurrence of a certain element of psycho-­linguistic content—in this case, the intuiting from which the normative intuited arises—is both content-located and factual. Hence, its employment in support of a world-located norm is blocked by the content/world barrier as well as by the is/ought barrier. Aside from intuitions, common sense as a form of evidence in support of such a norm  Cf. Chap. 3, Sect. 3.5.

1

120 

C. Rhode

will fare no better—for common sense (as explained in Chap. 4, Sect. 4.4) is equally located in psycho-linguistic content and equally factual. The same applies to traditional conceptual analyses, which are ultimately based upon such factual, psycho-linguistic content and are thus equally barred here by both the is/ought separation and the content/world separation. Elements of psycho-linguistic contents like the above would therefore be of no use in supporting a questioned norm if that norm is an aspect or reality beyond psycho-linguistic content. Maybe the regularity of behaviour in accordance with the questioned norm could instead by invoked as evidence for the norm. After all, if a behaviour is both relatively prevalent and identified by its adherence to a certain pattern, it is also called ‘normal’ behaviour. In this way, maybe the question ‘Why should I accept the norm N?’ could effectively be answered by saying ‘Because N is how it is done!’? Observed behavioural regularity, however, would obviously constitute descriptive evidence and would thus be useless in supporting a normative hypothesis, due to the is/ought separation. Still, it might be countered that expressions of what is ‘normal’ are simultaneously descriptive and normative, and hence that such expressions serve to bridge the is/ought separation.2 Against this response, consider how that would play out in a philosophical persuasion dialogue: The questioned hypothesis N is supported by the contention that N expresses both what happens regularly and what ought to happen. It can be expected with considerable confidence that the party that questioned N would offer something like the following two-fold response to this argument: ‘Firstly, can you please provide evidence in support of the apparent premise that the fact of this regularity supports N; and secondly, can you please explain how exactly the normativity of behaviour in accordance with N constitutes effective evidence in support of N – seeing as this amounts to invoking N in support of N?’ Plainly, neither the descriptive nor the normative face of what is deemed ‘normal’ would be persuasively effective in support of a questioned norm, which is why invoking this thick concept is persuasively useless. This uselessness is notably

 This approach has already been rejected in Chap. 3, Sect. 3.4 but may need rejecting again for the sake of clarity. 2

5  Metaphysical Hypotheses 

121

independent of the concept at issue, as it only depends upon the thickness of the concept.3 Hence other thick concepts would be equally useless. To support a questioned norm located in the world beyond psycho-­ linguistic content, perhaps we must look to that same world beyond content for evidence. For instance, scientific evidence about nature, as invoked in Canberra-style conceptual analyses, might be employed in support of a world-located norm, seeing as both share the same location relative to the content/world separation. Equally, non-natural or super-­ natural circumstances may be invoked for the same purpose. We might try to support a questioned norm, say, by invoking some (relevant aspect of a) particular Platonic form or other abstract entity, such as (to adapt a commonly used example) the entity called the command that you save this drowning child that is floating outside the spacetime of this universe. Or we might offer as support, for a questioned norm, a deity’s reported utterance commanding that you save this drowning child. Yet facts about scientifically examined nature, about abstract entities, or about divine pronouncements are still just that—facts, expressed in descriptive propositions. And as explained by the is/ought separation, employing a descriptive proposition in support of a questioned normative hypothesis will be persuasively useless. A further option for supporting a questioned norm that is an aspect of the world beyond psycho-linguistic content would be to take a contractualist or a contractarian approach. Either of these kinds of social contract theories builds on a hypothetical scenario in which rational agents jointly commit themselves to a set of norms. Contractualism speculates as to what norms would be accepted if such joint commitment was made with due regard to everyone’s equal moral standing as a rational agent, to the effect that these norms must be rationally justifiable to every person. By contrast, contractarianism considers what norms self-interested rational agents would jointly commit to in order to maximise the satisfaction and avoid the frustration of their self-interest. Now, contractualism already depends upon certain normatively loaded premises concerning the equal moral standing of agents, which are just as precarious and  Again, this reiterates substantially what has been discussed in Chap. 3, Sect. 3.4.

3

122 

C. Rhode

difficult to support as the contractualist support thereby offered for the norm that was originally questioned. What is of greater significance here, however, is an aspect shared by both contractualist and contractarian approaches, viz. the hypothetical status of its contracting scenario and contract. Being hypothetical, the contract and contracting scenario are imaginary, existing only within people’s (or really, philosophers’) thought or talk. That is, if a hypothetical social contract is invoked in support of a questioned norm, then the evidence thus employed constitutes an element of psycho-­linguistic content: The contract cannot be found in the world beyond such content; it is entirely an element of that content. With the questioned norm being (ex hypothesi) an aspect of reality beyond any psycho-­linguistic content, invoking such evidence breaches the content/world barrier, thus being persuasively useless in supporting the norm in question. Besides, the is/ ought separation is crossed just as well here: The hypothetical actions of fully prudent or instrumentally rational agents, or of fully (non-instrumentally) rational agents, are invoked in support of a norm about how not necessarily fully rational agents ought to act in actual scenarios. Invoking how some people would act in support of how other people ought to act evidently crosses the is/ought barrier. For the persuasive purposes of philosophical dialogue, then, it seems that hypothetical contracts are indeed not worth the paper they are not written on. Increasingly running out of options, one might resort to claiming that we can support a questioned norm by merely employing arguments for this norm. Ostensibly in line with most of the rest of analytic philosophy (at least according to Deutsch 2015), we do not need to draw upon intuitive, naturalistic, contract-based or any other evidence in order to support a norm that has been questioned, as we can instead draw upon arguments in place of evidence. I take it to be plain that the persuasive effectiveness of an argument depends not least upon one’s interlocutor’s acceptance of the premises of that argument, and that said acceptance will in turn depend upon the evidence in favour of those premises (cf. Chap. 3, Sect. 3.2; Ball 2017). One might hold, however, that such evidence is not required, as the support for those premises would yet again be provided by arguments rather than evidence, and so on ad infinitum (Deutsch 2015). But, if one’s interlocutor questions a norm one asserted, and if one attempts to provide

5  Metaphysical Hypotheses 

123

an infinite chain of arguments in order to persuade that interlocutor of the norm in question, then one will by definition never finish, and hence one will never succeed in persuading that interlocutor. This kind of normative infinitism is therefore of no use in philosophical persuasion dialogue. (In practice, one may of course also find it difficult to muster even a long chain of arguments in support of a questioned norm, as one might soon reach the normative bedrock of one’s fundamental convictions, at which point such an infinitist approach may fail anyway.) With seemingly all avenues of supporting a questioned norm exhausted, it looks as though norms cannot be supported in a persuasively effective way if they constitute aspects of the world beyond psycho-linguistic content. Yet if one were to insist that norms are aspects of the world beyond content, then the result that they cannot be effectively supported would seemingly entail that they cannot be justified, which may in turn lead to normative scepticism. Normative scepticism, however, is usually held to be a deeply implausible thesis, one which must of course be rejected— along with any argument entailing such scepticism. So, perhaps the appeal to the sheer implausibility of normative scepticism will eventually enable us to defend a given norm from questioning, on the basis that such questioning would open the floodgates to this most implausible thesis? This appeal unfortunately gives rise to a dilemma. On the one hand, the appeal might rely on the privileged status of the proposition that normative scepticism must be rejected, and thus on dialogical foundationalism, which has itself already been rejected in Chap. 2. Or on the other hand, if this appeal does not rely on such privilege, then it instead relies on the implausibility of normative scepticism as the crucial piece of evidence. Yet such implausibility cannot be found in the world outside our psycho-linguistic content, and hence the evidence at issue is an element of psycho-linguistic content, employed in support of a norm located among aspects of reality beyond such content. Employment of that evidence thus crosses the content/world separation and is therefore persuasively as useless as an insistence on privilege. It also crosses the is/ought separation, of course, as the fact of finding something implausible is still a fact, expressed by a descriptive proposition. And, of course, the inability to support a norm in dialogue does not even give rise to the attitudinal thesis of normative scepticism, due to the attitude/dialogue separation.

124 

C. Rhode

In conclusion, if norms are aspects of the world beyond psycho-­ linguistic content, then there appears to be no avenue of supporting them that does not breach one or another inference barrier, in contravention of dialogical empiricism. It is therefore prudent to avoid any assertion of such world-located norms in philosophical dialogue. Without assertion, however, one can practically only fail to persuade one’s opponents, which is to be avoided. So, unable to assert such norms, it is prudent to excise norms from philosophical dialogue to the extent that norms are aspects of the world beyond psycho-linguistic content. This prudential recommendation arose from two inference barriers, viz. the content/world separation and the is/ought separation. It is worth considering, then, whether a similar recommendation arises from these inference barriers on the assumption that norms are elements of psycho-­ linguistic content, rather than aspects of the world beyond such content as had been assumed thus far. In this scenario, the content-located types of evidence discussed above may well be employed in support of a questioned norm—but only if such employment does not breach the is/ought barrier, as it remains imprudent to support a hypothesis expressing a (content-located) normative requirement by invoking a (content-located) descriptive proposition. Inconveniently, all the evidence discussed above with regard to world-located norms was, in each case, also blocked by the is/ought barrier (or by the persuasive uselessness of infinite chains of arguments). In consequence, the employment of norms in philosophical dialogue must prudently be avoided even if such norms are merely elements of psycho-linguistic content. Quite independently of that, it would anyway seem rather unappealing to conceive of norms, and especially of moral norms, as existing merely within—and thus carrying force merely within—our psychological states or linguistic practices. So conceived, a normative claim would effectively be prefaced by something like the qualification that According to what happen to be our prevalent linguistic practices and/or my current psychological states, [you ought (not) to do x]. In this way, moral requirements would at best amount to rules of grammar, concerned only with the meanings of morally significant concepts in our language. They would not concern what we intuitively think is actually right or wrong regardless of any coincidences of language. A moral obligation thus qualified

5  Metaphysical Hypotheses 

125

would invite the response that ‘This is how you interpret moral concepts—why does that bind me?’ But such relativism appears wholly incompatible with the absoluteness of our ordinary moral convictions. As Brad Hooker wrote (2000: 12): ‘[S]ome moral claims seem overwhelmingly compelling. […] Suppose bored soldiers entertain themselves by torturing their prisoners. This is obviously wicked.’ At the very least, it is an open question as to whether it would be tenable in the face of such strong moral convictions to engage with norms in philosophical dialogue on the express premise that they hold merely within the boundaries of our psycho-linguistic content. And in any case, even within these boundaries, the employment of normative claims must prudently be avoided, due to the is/ought separation. So it seems prudently impossible to debate normative hypotheses in philosophical dialogue. There is, however, one kind of prescription that is in principle supportable when questioned even though it holds regardless of and beyond any psycho-linguistic content. This kind of prescription encompasses Machiavellian prudential recommendations, as employed paradigmatically in Niccolò Machiavelli’s The Prince (1988 [1532]; cf. Femia 2012). Such prescriptions are hypothetical imperatives that apply only if the antecedent condition is actually satisfied, that is, if I actually find myself in the situation described by the antecedent condition. This condition must also specify a goal or ranking of goals, and the prudential recommendation thus only applies if I have the specified goal or ranking of goals. Such Machiavellian prudential recommendations therefore take the following general form: If you are in situation a, and you have goal c, then you should prudently perform the set of actions b to attain c. As such, prudence in the Machiavellian sense4 avoids both the is/ought separation and the content/world separation. On the one hand, such a prudential recommendation is not itself binding, lacking as it does any normative force of its own. A recommendation of this kind does not render it impermissible to act in a manner contrary to what it recommends, leaving its addressee free to pursue their goal in an imprudent fashion or not pursue it at all. Yet although the  This is distinguished from the Ancient sense of prudence as a civic virtue (cf. Hariman 2003; see also fn. 5 in Chap. 1, Sect. 1). 4

126 

C. Rhode

addressee is free to act otherwise, if the addressee’s goal and current situation are contained among the antecedent conditions of this recommendation (c and a above), and if the recommendation is true (in that b does actually tend to be the most effective way of attaining c out of situation a), then the addressee will likely be motivated by their goal to act in the manner recommended. To illustrate this goal sensitivity, consider the recommendation that If you live in the U.S. and want to avoid going to prison, then you should not rob a bank. This recommendation did not motivate Richard James Verone to refrain from robbing a bank, as Mr Verone lived in the U.S. but did not seek to avoid going to prison. Indeed, he robbed a bank in North Carolina precisely in order to go to prison—and there obtain healthcare (Pilkington 2011). In contrast to a Machiavellian prudential recommendation, however, the force of a norm is goal-insensitive. After all, you should save the oft-invoked drowning child when you become aware of its plight and are able to save it, no matter what goal you might or might not have on the day. A norm thus renders it impermissible to act contrary to its prescription: The addressee of a norm is not at moral liberty to disregard its demand. In consequence, a norm is (obviously) normative in the sense of the is/ought separation; but a Machiavellian prudential recommendation is descriptive, specifying what action, as a matter of fact, is most likely to bring about the desired outcome. Understood in this way, a prescription is prudential (and hence descriptive) iff it is goal-sensitive, and a prescription is normative iff it is goalinsensitive. Making a prudential recommendation thus does not involve the crossing of the is/ought separation. On the other hand, making such a recommendation does not involve crossing the content/world separation either. To see why that is the case, it is worth distinguishing such Machiavellian prudential recommendations from prescriptions derived through contractarian theorising. Demonstrated paradigmatically in Thomas Hobbes’s Leviathan (2012 [1651]), contractarianism considers a hypothetical scenario and concludes on its basis that it would be rational for agents in the pursuit of their self-interest—in effect, it would be prudent—to jointly adopt certain specified prescriptions that are discoverable through mere consideration of the hypothetical scenario and the nature of the agents and their self-interests. Since this would seemingly be rational for any agent, it is

5  Metaphysical Hypotheses 

127

rationally required for non-hypothetical, actual people to abide by these specified prescriptions. Now, although this contractarian approach also focuses on prudent action, it is distinguished in one crucial respect from the Machiavellian prudential recommendations outlined above. Hobbesian prescriptions are based on a hypothetical scenario, but they are ostensibly binding for rational agents whether or not they find themselves in the described scenario. The Hobbesian approach thus employs aspects of this hypothetical scenario to yield prescriptions that are ostensibly binding for anyone, whereby this binding status holds regardless of any psycho-linguistic content. As already pointed out above, hypothetical scenarios are imaginary, and their evidential employment in support of a thus world-located prescription involves the crossing of the content/ world barrier. Machiavellian prudential recommendations, however, employ actual past events as their evidence and counsel a certain course of action only in light of the specified situation and goal, that is to say, only for agents who do actually find themselves in the specified situation and with the specified goal. Here, neither the evidence nor the antecedent situations exist merely within people’s psycho-linguistic content.5 Rather, both are located in the world beyond such content—so Machiavellian prudential recommendations do not rely on the crossing of the content/world separation. More fundamentally, however, Machiavellian prudential recommendations are not even located in persuasion dialogue. Seeing as their addressee is free to choose an imprudent action, and seeing that such prudential considerations seek to inform the addressee’s decision-making regarding a set of actions in an actual situation, rather than persuade them to accept a certain norm or hypothesis, invoking Machiavellian prudential considerations quite clearly takes place within deliberation dialogue.6 (This contrasts with contractarianism, where it is attempted to  Note that I am not concerned with people’s accurate or inaccurate representations of either, but rather only with the evidence and scenarios themselves. False beliefs about either may of course lead to mistaken prudential recommendations that counsel an actually imprudent course of action. But such false beliefs are irrelevant to prudential recommendations based on the actual evidence and scenario. 6  As with any type of dialogue, such a deliberation dialogue may be embedded within a different dialogue and/or have a different dialogue embedded within it. 5

128 

C. Rhode

persuade the addressee of some particular prescription.) As a result, the foregoing two inference barriers do not even have any significance for Machiavellian prudential considerations, as their significance has only been established for persuasion dialogue.7 And yet, even this prudential approach to deriving prescriptions in philosophical dialogue has an important drawback. Machiavellian prudential considerations of the above form (viz., If you are in situation a, and you have goal c, then you should prudently perform the set of actions b to attain c) will often concern matters of political, social, economic or other fact. The paradigm case of Machiavelli’s Prince (1988 [1532]) is indeed a case in point, being concerned exclusively with political facts. To establish such matters of fact, empirical investigation of some form will often (though not necessarily always) be required. Yet many analytic philosophers (not unlike many economists) still seem to harbour a certain distaste for such ‘worldly concerns’ as to find out whether a certain course of action does actually statistically tend to result in a certain outcome. So philosophers who want to derive Machiavellian prudential recommendations would probably sooner or later have to get their hands dirty and conduct empirical research. They might even need to train in one or two disciplines outside analytic philosophy, or indeed switch to a different discipline altogether, in order to ensure that their research is conducted according to the respective discipline’s standards of rigour. Let me now take stock of the conclusions established thus far with regard to norms and prescriptions. We have discussed three different conceptions of prescriptions: as normative aspects of the world beyond psycho-­linguistic content; as normative elements of psycho-linguistic content; and as Machiavellian prudential recommendations. First, if norms are aspects of reality beyond any psycho-linguistic content, then a questioned norm effectively cannot be supported in philosophical persuasion dialogue. Here, every avenue for supporting the questioned norm must prudently be avoided, in each case due to its violation of the is/ ought barrier and often also of the content/world barrier. Consequently, it is prudent in philosophical persuasion dialogue to not even seek to  As stated in Chap. 2, Sect. 2.3, this Machiavellian prudential approach is adopted throughout the present work. Cf. fn. 12 in that section. 7

5  Metaphysical Hypotheses 

129

persuade an interlocutor of any such norm, as one effectively cannot succeed in such an effort, and hence it is prudent to not even debate such norms. Second, if norms are elements of psycho-linguistic content, then a questioned norm effectively cannot be supported in philosophical persuasion dialogue, owed to the separation between descriptive and normative propositions. Every avenue for supporting such a norm would cross this is/ought barrier (or fail on independent grounds), and must thus prudently be avoided. (Norms so conceived would in any case be rather unappealing, as they only hold within the boundaries of what happen to be our psychological states or linguistic practices and carry no force independent of such content.) Since supporting a norm so conceived must prudently be avoided, it is again prudent in philosophical persuasion dialogue to not even seek to persuade an interlocutor of any such norm, and hence to not even debate such norms. Third, if we employ Machiavellian prudential recommendations, then we could support a questioned prescription on grounds that it is prudent in light of an interlocutor’s current situation and goal to act in accordance with the questioned prescription. The evidence inductively invoked here would be an observed statistical relation between certain actions and certain outcomes. This evidential relation would not cross either of the foregoing two inference barriers—not least since these barriers have no bearing in the deliberation dialogue that characterises such prudential recommendations. Obtaining the necessary evidential support, however, may often require empirical research and detailed familiarity with one or more academic disciplines outside of analytic philosophy. Which one of these three avenues is least unappealing I leave to the reader. My intention here was only to advise on the most effective conduct of philosophical dialogue with regard to normative or prescriptive claims, and with particular focus on conduct that is prudently to be avoided. This focus will continue in the next section of this chapter, though I will there turn from issues of normativity to questions concerning necessity.

130 

C. Rhode

5.3 Necessity We have seen that dialogical empiricism gives rise to both the content/ world separation and the is/ought separation. Beyond these two barriers, there was a third persuasion barrier identified in the preceding chapters, viz. the contingent/necessary separation. This inference barrier states that one must prudently not employ contingent propositions (or propositions expressing contingent truths) in support of a putatively necessary hypothesis (or a hypothesis expressing a necessary truth) that has been questioned in a philosophical dialogue. This is an inference barrier (rather than an implication barrier), as no increase in the quantity of contingent evidence would render that evidence any more salient with regard to the questioned necessary hypothesis.8 Now, if one cannot prudently support a questioned necessary truth by invoking contingent truths, then the question arises as to what other evidence could prudently be invoked to support a necessary truth to a questioning interlocutor. For the sake of clarity in answering that question, it will be useful to again consider necessary truths separately on the assumption that they are aspects of the world beyond any psycho-linguistic content, and thereafter on the assumption that they are elements of psycho-linguistic content. First, then, assume that necessary truths are aspects of reality beyond any psycho-linguistic content. In this event, crossing the content/world separation, just like the other foregoing inference barriers, must prudently be avoided when supporting a questioned necessary truth. So, what evidence could potentially be employed in support of a necessary truth if necessary truths are aspects of the world beyond psycho-linguistic content? Intuitions seem to be a frequent choice of evidence here (cf. e.g. Vaidya 2015). Assuming that necessary truths are aspects of reality beyond any psycho-linguistic content, however, employing an intuition in support of a questioned necessary truth will be persuasively useless, as intuitions are elements of psycho-linguistic content. A breach of the content/world separation would equally occur if the evidence invoked was some  Cf. Chap. 3, Sect. 3.5. Note that the present barrier lies between necessary truths and contingent truths, not possible truths. Inferences from necessary truths to possible truths thus do not undermine the bi-directionality of the contingent/necessary separation. (I assume here that a truth is contingent iff that truth obtains, but could logically also have not obtained, in the actual world.) 8

5  Metaphysical Hypotheses 

131

common sense verdict, or a traditional analysis of a concept, since both common sense and the evidence employed in traditional conceptual analyses are located among psycho-linguistic content. The same holds true with regard to claims to plausibility: The plausibility of a necessary truth is not found in the world beyond content but rather within psycho-linguistic content, and consequently the employment of the plausibility of a necessary truth as evidence for that questioned necessity would breach the content/world barrier.9 Similarly, beliefs constitute elements of psycho-­linguistic content, and their evidential employment in support of a questioned necessary truth would thus cross the content/world barrier. Notably, the employment of any such psycho-linguistic content would simultaneously also cross the contingent/necessary barrier, since the occurrence of any particular psycho-linguistic content is clearly contingent, while the hypothesis to be supported is putatively necessary. Violations of the content/world separation may of course be avoided if the evidence invoked in support of a necessary truth that is an aspect of reality beyond content was itself an aspect of reality beyond psycho-­ linguistic content. Such world-located evidence may be found, for example, in empirically observed exceptionless regularities. Consider, for instance, Newton’s law of gravity, according to which the force of gravity acting between two masses is equal to the gravitational constant multiplied by the quotient of the product of their masses (as the numerator) and the square of their distance (as the denominator). There have been innumerable observations of the regularity described by this law, and this observed regularity might well be employed in an attempt to support the claim that Newton’s law of gravity is a necessary truth.10 Although such  Equally, considerations of the plausibility of a mechanism by which necessary truths may be known are of no present utility. For its prudential employment in philosophical persuasion dialogue, such a mechanism would have to provide support for world-located necessary truths without crossing the content/world barrier in the process. Yet as I will discuss shortly, the knowledge that it supposedly yields very much seems to be an element of psycho-linguistic content. So neither the output of such a mechanism (viz. knowledge), nor the plausibility of that mechanism, can prudently be employed in support of a questioned world-located necessary truth. Such a mechanism might yield world-located evidence in addition to its content-located output; but then this content-­ located output still remains prudently unusable. 10  The regularity expressed by this law is of course not actually exceptionless, just as the gravity described by this law is not actually a force, though I will ignore both of these issues here due to the general familiarity and thus illustrative value of Newton’s law. 9

132 

C. Rhode

employment (also via the Canberra Plan) would not cross the content/ world separation, it would violate the contingent/necessary separation: Any observed event consistent with Newton’s law could hypothetically have occurred differently (including in a manner not consistent with the law) and is therefore contingent. Given the contingent/necessary barrier, such contingent evidence must prudently not be employed in support of a purported necessary truth. Hence, an empirically observed regularity, even an exceptionless regularity, will not serve to support a questioned necessary truth. It would only serve to support a description of the contingent relation that it constitutes. This point notably has direct implications for the debate regarding the status of laws of nature, as already indicated by the present discussion on Newton’s law of gravity. On instrumentalist views, laws of nature are not truth-apt but are rather characterised by their utility as relatively simple tools for describing and predicting observations. Such views therefore do not posit any necessary truths and are thus unaffected by the present point regarding regularity as evidence for necessity. On regularity views, in turn, laws of nature are truth-apt and describe in a summarised, generalised, simplified or abstracted form a certain set of natural facts or observations, though such laws do not amount to claims that these facts could not have been otherwise. Such regularity views therefore do not imply that laws of nature are necessarily true, and thus remain equally unaffected by the present point.11 Necessitarian views, however, do posit the necessity of laws of nature, to the effect that such laws ‘are facts over and above the facts they govern’ (Lange 2014: 241). The evidence for there being such necessary truths is variously taken to include the facts or observations governed by the laws of nature, the plausibility or explanatory simplicity of this conception of laws of nature, and the putative counter-intuitiveness of non-necessitarian views or of the latter’s implications. Yet the natural facts governed by the laws of nature are themselves contingent, as is the regularity of their occurrence, and contingently true propositions expressing these facts or their regularity cannot prudently be  As will become clear below, instrumentalist and regularity views are nonetheless affected by the considerations of this chapter to the extent that they claim that there do not exist any necessarily true laws of nature as part of the world beyond psycho-linguistic content. But neither of the positive characterisations of laws of nature that are offered by these views strictly implies this negative claim. 11

5  Metaphysical Hypotheses 

133

employed to support the claim to the necessity of a truth. At the same time, both plausibility (as well as explanatory simplicity) and intuitiveness or counter-intuitiveness are elements of psycho-linguistic content12 and can therefore not be employed in support of laws of nature that are facts in their own right ‘over and above the [natural] facts they govern’.13 Being elements of psycho-linguistic content, their occurrence is also contingent, and their evidential employment here is thus also barred by the contingent/necessary separation. Together, the content/world barrier and the contingent/necessary barrier thus make it effectively impossible in philosophical dialogue to support the claim that a given law of nature is necessarily true. In consequence, it is prudently necessary (pun intended) to refrain from positing necessitarian views of laws of nature. But perhaps the key evidence that can be invoked in defence of a questioned world-located necessary truth may be drawn quite literally from another world—or many other worlds. After all, a claim to the necessity of a particular truth is predominantly considered to be equivalent to a claim to the non-existence of a possible world wherein that truth does not obtain. Hence, if it was found that there is no possible world where a given fact does not obtain, then this would seemingly provide entirely conclusive evidence for the necessity of that fact. However, the employment of other possible worlds as evidence faces one particular dilemma, regarding the ontological status of other possible worlds. If such possible worlds do not exist as aspects of reality but merely as convenient fictions, manners of speaking, or in any case as elements of psycho-linguistic content, then invoking them in support of a questioned necessary truth that does exist in the world beyond mere talk or thought would violate the content/world separation. Yet if such possible worlds are claimed to exist as aspects of reality beyond content, then the interlocutor who challenged the necessary truth in question would very likely proceed to request evidence for the existence and exact nature of such possible worlds as aspects of reality beyond content. Of course, the intuitiveness, plausibility, convenience, or any other theoretical attribute of their existence could not prudently be  On intuitiveness and explanatory simplicity or force cf. Chap. 4, Sect. 4.4, and on plausibility cf. Sect. 5.2. 13  Cf. Chap. 4, Sect. 4.3. 12

134 

C. Rhode

invoked in support of said existence, as that would involve crossing the content/world barrier. The evidence for the world-located existence of other possible worlds would instead have to consist in aspects of reality beyond psycho-linguistic content. Yet there is only one account of additional worlds comprising all possible states of affairs that arguably enjoys such evidential support—viz. the many-worlds interpretation of quantum mechanics. In contrast to the currently dominant Copenhagen interpretation of the results obtained in quantum-physical experiments, the mathematically more literal many-worlds or Everettian interpretation (after Hugh Everett III) resolves apparent incoherencies in quantum physics by way of positing a vast number of real worlds existing in parallel to each other. According to the results of quantum physical experiments, a particle that can be in a plurality of mutually exclusive states actually is in a plurality of mutually exclusive states—which seems flatly contradictory.14 The Copenhagen interpretation resolves this seeming contradiction by regarding that particle as a mere set of potentialities, specifically a probability wave, only collapsing into a single, actual state (and particle) upon observation. In contrast, the Everettian interpretation takes the particle’s plurality of mutually exclusive states at face value, and instead resolves the apparent contradiction by claiming that every such particle is also actually observed in each of its mutually exclusive states—albeit by separate, mutually inaccessible and parallelly existent observers, in separate, mutually inaccessible and parallelly existent realities.15 The resultant boundless multiplication of parallel realities, however, is unfortunately of no use to philosophers seeking to support a questioned necessary truth by invoking possible worlds. For one, the Everettian parallel realities comprise all and only physically possible histories of the universe, which of course do not neatly map onto logically possible histories as putatively found across different possible worlds. But more importantly, parallel realities as posited by the many-worlds interpretation are mutually inaccessible. Yet in order to employ parallel worlds as evidence  This is best illustrated by the double-slit experiment; cf. fn. 13 in Chap. 4, Sect. 4.5.  For a slightly more extensive sketch of the Everettian interpretation, see Papineau (2006: 22-5). For a comprehensive overview and defence, see Wallace (2012). 14 15

5  Metaphysical Hypotheses 

135

in support of the necessity of some particular truth, one would require access to all other parallel realities, so as to investigate whether there is indeed no such reality wherein the truth in question does not obtain. With such access being unavailable, it is practically impossible to employ possible worlds that are aspects of reality beyond content in support of a questioned putatively necessary truth that is also an aspect of reality beyond psycho-linguistic content. It appears, then, that there is no way to support a necessary truth to a questioning interlocutor in philosophical dialogue if necessary truths are aspects of reality beyond psycho-linguistic content—in which case it is prudent in philosophical dialogue to refrain altogether from claiming that some proposition is true by necessity (lest one’s efforts of persuasion are bound to fail). These difficulties do not arise in exactly the same way if necessary truths are elements of psycho-linguistic content, such that their necessity obtains only within and as a result of the coincidences of people’s talk or thought, lacking any import about reality beyond psycho-­ linguistic content. Specifically, on the assumption that the necessity of a truth is an element of psycho-linguistic content, the content/world separation does not pose an obstacle to supporting a questioned necessary truth by invoking psycho-linguistic content as evidence, since such support would never cross this inference barrier. One must nevertheless prudently avoid crossing the contingent/necessary separation when supporting a questioned, content-located necessary truth, as it remains imprudent to support a hypothesis expressing a (content-located) necessary truth by invoking a (content-located) contingent truth. It is still possible to offer as evidence other necessary truths or reformulations of the questioned necessity (e.g. in terms of content-located possible worlds); but such evidence would just as well invite further questioning from the interlocutor who has already proved reluctant to accept the purported necessary truth at issue. In a manner that parallels the practical unsupportability of content-located norms, content-located necessary truths are rendered practically unsupportable by the contingent/necessary barrier. In consequence, the employment of necessary truths in philosophical dialogue must prudently be avoided even if such necessities are merely elements of psycho-linguistic content.

136 

C. Rhode

It is thus prudently impossible in philosophical dialogue to support a questioned necessary truth, whether necessary truths be elements of psycho-­linguistic content or aspects of reality beyond such content. (This prudential restriction applies to any truth—such as conceptual, logical, or mathematical truths—to the extent that such a truth is a necessary truth.) Having thus discussed the implications of dialogical empiricism for claims to the necessary truth of a proposition, I will turn in the next section to certain consequences of dialogical empiricism for epistemology.

5.4 Epistemology In order to examine the implications of dialogical empiricism for debates in epistemology, it is in this case useful to consider whether the objects of such debates are elements of psycho-linguistic content or aspects of reality beyond such content. The classical object of epistemological debates is of course knowledge, which may or may not imply the other central notion of epistemology, justification. Let us assume for the moment that it does, and specifically that what is often considered the pre-Gettier analysis of knowledge as justified true belief is accurate. The question is, then, whether knowledge so analysed is an element of psycho-linguistic content or an aspect of the world beyond such content. Being a belief, it seems clear that the former is the case, since beliefs are elements of psycho-linguistic content. It might perhaps be argued that a belief ’s truth and/or justification may be determined by, and dependent upon, some aspect of the world beyond psycho-linguistic content, and that therefore such a belief itself must be regarded as an aspect of the world beyond content. However, a relation of dependence of a belief ’s truth or justification upon some aspect of reality beyond content does not entail that said belief is itself part of the world beyond content. It is entirely possible for a justified true belief to be an element of psycho-­ linguistic content while the things that makes it justified is not, seeing as the epistemic relation of justification is itself independent of the respective metaphysical nature of that which justifies and that which is thereby justified. Equally, it is entirely possible for a justified true belief to be an

5  Metaphysical Hypotheses 

137

element of psycho-linguistic content while the thing that makes it true is not, as the semantic relation of truth is equally independent of the respective metaphysical nature of the truth-maker and the belief. Hence, neither the source of truth of a given belief, nor the source of justification of said belief, can turn this belief into an aspect of reality beyond psycho-­ linguistic content. Post-Gettier analyses of knowledge have posited further conditions, sometimes in addition to, and at other times in partial substitution of, the foregoing three conditions, with particular focus on the justification condition in the pre-Gettier analysis of knowledge, resulting in analyses of knowledge as true belief that also satisfies some set of conditions which together are broadly functionally equivalent to justification. In light of this functional equivalence, I very strongly suspect that any such functionally quasi-justificational set of conditions would equally fail to turn a belief into an aspect of the world beyond content. On any pre-Gettier or post-Gettier analysis of knowledge, then, knowledge seems clearly an element of psycho-linguistic content. And even if knowledge is a primitive, unanalysable concept (cf. Williamson 2000), it still seems to be something held by an agent among the agent’s other mental and thus psycho-­ linguistic content. Perhaps the clearest challenge to the characterisation of knowledge as an element of psycho-linguistic content has been offered by Hilary Kornblith (see e.g. his 2014). As he explains in reference to results from cognitive science, our concept of something may bear little to no relation to the thing so conceptualised. Kornblith therefore argues that philosophers should not be concerned with our concept of knowledge but rather with knowledge itself, which he considers to be a kind. Though he has expressed openness to conceiving of knowledge as an anthropogenic kind, he has argued that it is a natural kind (2007: 36-7), in which case knowledge would be an aspect of the world beyond any psycho-linguistic content. Kornblith essentially holds that, akin to instances of other natural kinds, instances of knowledge also exhibit a high degree of consistency of their properties: Knowledge is very similar in appearance across all its instances (2002). For this analogy with natural kinds, he effectively relies on two premises: First, that knowledge is indeed very similar across all its instances; and second, that such similarity across all instances is the

138 

C. Rhode

defining characteristic of natural kinds. This argument from analogy therefore relies on two premises that amount to nothing other than the very conclusion to be established by the analogy: It is circular. Charitably interpreted, then, Kornblith principally relies on the plausibility of the analogy, thus relying upon psycho-linguistic content as evidence in order to support a hypothesis about the world beyond content (and specifically, about a putative natural kind). Not only in light of this circularity and apparent content/world inference, but also in light of the works of Michel Foucault (e.g. 1972 [1969]; cf. Gutting 2001: 258-88), I am very much inclined to disagree with Kornblith and instead consider knowledge an anthropogenic kind. To see the implications of dialogical empiricism for debates about knowledge, let us first assume for the moment that Kornblith is right and knowledge is a natural kind after all. If that is true, then knowledge exists as an aspect of the world beyond any content of talk or thought, and consequently no such content could prudently be invoked as evidence in support of hypotheses about knowledge. In particular, intuitions, conceptual analyses, thought experiments, reflective equilibria and various other kinds of philosophical evidence and methods, as well as people’s implicit or explicit reports about such psycho-linguistic content as they may call ‘knowledge’, would be effectively useless in philosophical debates about knowledge, due to the content/world separation. There also appears to be no world-located evidence available to support hypotheses about the natural kind knowledge, as this natural kind (like most if not all natural kinds) does not readily lend itself to empirical observation or other forms of evidence gathering from aspects of reality beyond content, seeing as the instances of this putative natural kind seem invariably contained within people’s talk or thought.16 It is therefore prudent in philosophical dialogue to refrain from hypothesising about knowledge if knowledge was indeed a natural kind or in any case an aspect of reality beyond psycho-linguistic content. This restriction would not arise from dialogical empiricism, however, if Kornblith is mistaken and knowledge is an anthropogenic kind (or in any case an element of psycho-linguistic content). In that case,  Cf. Chap. 4, Sect. 4.3.

16

5  Metaphysical Hypotheses 

139

hypotheses about knowledge would be expressing an element of psycholinguistic content, and the employment of intuitional and other contentlocated evidence in support of such hypotheses would not involve the crossing of the content/world barrier (nor indeed would it unavoidably involve the crossing of either the is/ought barrier or the contingent/necessary barrier). Beyond these brief considerations regarding the nature of knowledge and the resultant implications of dialogical empiricism for debates about knowledge, there is one consequence arising from dialogical egalitarianism (and thus from dialogical empiricism) that should also be noted for present purposes. This consequence concerns global epistemological scepticism. In discussing scepticism here, we must distinguish between Academic scepticism and Pyrrhonian scepticism.17 For any proposition P, Academic scepticism holds that one does not know that P, and anti-­ scepticism holds for some but not all P that one does know that P. By contrast, Pyrrhonian scepticism holds neither that one does nor that one does not know that P. As such, when discussing their respective positions in philosophical dialogue, anti-sceptics and Academic sceptics make a commitment regarding knowledge that P, while Pyrrhonian sceptics make no such commitment regarding knowledge that P. Given dialogical egalitarianism, any commitment made in philosophical dialogue can be challenged and must then prudently be supported. But one cannot incur such a burden of proof if one does not make any dialogical commitment in the first instance. Hence, anti-sceptics and Academic sceptics can each incur a burden of proof in philosophical dialogue about knowledge that P, but Pyrrhonian sceptics cannot. By implication, anti-sceptics and Academic sceptics can fail in their efforts to persuade their opponent of the (anti-)sceptical position to which they have committed themselves, but Pyrrhonian sceptics cannot fail likewise, as they do not seek to persuade any opponent. Pyrrhonian scepticism is therefore easier to ‘hold’ in philosophical dialogue (though one never commits to it) than either  Academic scepticism is also known as Cartesian scepticism, despite the fact that the method of doubt at the centre of the latter had already been deployed by Abū Ḥ āmid Muḥammad ibn Muḥammad al-Ṭ ūsī al-Ghazālī over half a millennium before René Descartes (cf. Najm 1966, Moad 2009). So perhaps ‘al-Ghazālian scepticism’ would be a more appropriate name than ‘Cartesian scepticism’. 17

140 

C. Rhode

anti-­scepticism or Academic scepticism.18 This particular consequence of dialogical empiricism of course does not mean that the latter implies any of the preceding three attitudinal positions, neither anti-scepticism nor any form of global scepticism, since dialogical empiricism does not imply any attitudinal thesis at all. It does indicate, however, that a thesis that makes no commitment in a given debate may, perhaps rather obviously, be easier to hold than any thesis making a positive or negative commitment within that debate. Beyond epistemology, the advantageous dialogical position of non-commitment also applies in debates regarding abstract entities—which will be examined in the next section.

5.5 Ontology The final illustration of the consequences of dialogical empiricism that I want to provide here concerns debates between realism and anti-realism. Just like questions of normativity, necessity, and epistemology, such ontological questions arise right across the sub-disciplinary landscape of analytic philosophy. Debates between realists and anti-realists can be found in metaethics, aesthetics, the philosophy of colour, the philosophy of mathematics, and in general philosophy of science, to name but a few. For simplicity, I will here only consider such ontological debates insofar as they pertain to entities rather than to facts (cf. Brock and Mares 2007: 2). And for reasons that will become clear, it is useful to discuss ontological debates about abstract entities in separation from ontological debates about concrete entities. Though there are a variety of purported abstract entities whose ontological status is at issue in philosophical debates, it will suffice to examine key arguments in one such ontological debate as an example, since the implications of dialogical empiricism for such an example debate apply equally to debates regarding other abstract entities. Correspondingly, it will suffice to examine one debate regarding the ontological status of concrete entities. For this purpose, I will examine,  In persuasion dialogue, Pyrrhonian scepticism could of course not be a position in a dispute but only (implicitly) in a dissent, as only the latter includes a non-commitment position; cf. fn. 8 in Chap. 1, Sect. 1.3. 18

5  Metaphysical Hypotheses 

141

first, mathematical realism and anti-realism, and second, scientific realism and anti-realism. Let mathematical realism be the commitment to the existence of abstract mathematical entities (such as numbers) as aspects of reality beyond psycho-­linguistic content.19 And let mathematical nominalism be the non-commitment to the existence of abstract mathematical entities as aspects of reality beyond psycho-linguistic content. Mathematical nominalism can then be divided into anti-realism and agnosticism, whereby mathematical anti-realism is the commitment to the non-existence of abstract mathematical entities, while mathematical agnosticism is the non-commitment to both the existence and the non-existence of abstract mathematical entities (perhaps because such entities are considered to be mere elements of psycho-linguistic content). It suffices for present purposes to briefly consider the most prominent arguments in the debate between realism and anti-realism about abstract mathematical entities. In Chap. 4,  Sect. 4.5 we already discussed one such line of argument for mathematical realism, viz. Amie Thomasson’s (2014) neo-Fregean easy approach to ontology. According to Thomasson and other neo-Fregeans, the following argument is valid and proceeds from an ontologically insignificant premise to an ontologically significant conclusion: P1: The first sentence of the present paragraph contains 23 words. C1: The number of words in the first sentence of the present paragraph is 23. C2: There are numbers. As explained above, this argument for realism founders on the content/world separation. C1 expresses an element of psycho-linguistic content, while C2 expresses an aspect of reality beyond such content—hence C1 or similar hypotheses cannot prudently be employed in support of C2 if C2 is questioned in philosophical dialogue. A different line of argument for mathematical realism claims that realism is preferable because it allows for the employment of the same semantics in both mathematics and science, such that existential quantifications over abstract entities in mathematics are true in just the same manner as 19

 For simplicity, I will ignore realism about concrete mathematical entities (cf. Maddy 1990).

142 

C. Rhode

(and can be taken literally just as much as) existential quantifications over concrete entities in science (Bueno 2014). Now, it is of course very questionable whether such external consistency (and consequent simplicity) of mathematical semantics is indeed preferable (cf. Longino 1996). But quite apart from that, the semantic evidence being invoked here in support of mathematical realism straightforwardly and rather obviously consists of facts about semantics. Hence this evidence is an element of psycho-linguistic content. Employing such evidence in support of mathematical realism would therefore cross the content/world separation and must prudently be avoided. Perhaps the most popular argument for realism in the philosophy of mathematics, however, is the indispensability argument. We must commit to the existence (beyond mere psycho-linguistic content) of all such entities as are indispensable to our best theories of the world, the argument holds, and given the indispensable role of quantification over mathematical entities in science, we must commit to mathematical entities. Yet even if quantification over mathematical entities is indispensable to the formulation of our best theories of the world, this indispensability would be a product of the way we conceptualise the world and formulate our theories about the same. It would, that is, be a product of the ways in which our language operates.20 These ways themselves constitute elements of psycho-linguistic content, and employing such psycho-­linguistic content in support of mathematical realism would thus require crossing the content/world barrier, which must prudently be avoided. In summary, the foregoing three arguments do not offer a single argument for mathematical realism, agnosticism, or anti-realism that is not blocked by the content/world separation, and thus in effect by the prudential requirement to avoid unsatisfiable burdens of proof. Yet the burden of proof is of significance here not only through an inference barrier but also on its own. If a commitment to the existence of an abstract mathematical entity is questioned, then evidence must prudently be offered in support of the existential claim at hand. Abstract entities, however, are distinguished from concrete entities by the former’s lack of both spatio-temporal location and participation in causal  Cf. the discussion of ontological parsimony in Chap. 4, Sect. 4.4.

20

5  Metaphysical Hypotheses 

143

relations. So evidence in support of an existential claim about a worldlocated abstract entity would inevitably be unavailable to such concrete biological entities as humans engaged in philosophical dialogue (cf. Liggins 2010). To put the same point differently: How am I, as a concrete human being, supposed to be able to provide evidence to an interlocutor about abstract entities that I can neither causally interact with nor locate in the universe? Thus, if an existential claim about such an abstract entity is questioned, then it seems impossible to provide evidential support and thus impossible to meet the burden of proof. Note, however, that this difficulty affects not only mathematical realists but also anti-realists, as they too make an existential claim about an abstract entity, viz. a claim to the world-located non-existence of such an entity. Mathematical anti-­ realists thus also face the seemingly insurmountable problem of accessing evidence for the claimed non-existence in the world beyond content of abstract mathematical entities. The only position in the present debate that remains unaffected by the challenge to provide evidence about causally and spatio-temporally inaccessible entities is mathematical agnosticism, since this is the only position to not make any commitment regarding the existence of abstract mathematical entities as aspects of reality beyond content in the first instance. It should be noted here that such mathematical agnosticism may of course be combined with commitment to the existence of fictional mathematical entities as elements of psycho-linguistic content. That combination, advocated by Otávio Bueno (2009), refrains from commitment about the existence or otherwise of world-located abstract mathematical entities whilst committing to the existence of content-located mathematical entities. Again, the content/world separation (also with regard to philosophical methods) would pose little difficulty to supporting hypotheses about mathematical entities if these entities are mere elements of psycho-linguistic content without any import about reality beyond such content. In summary, then, dialogical empiricism appears to render both mathematical realism and mathematical anti-realism prudently unsupportable, though it has no similar implication with regard to mathematical agnosticism. It is thus prudent to refrain from advocating (and by extension, debating) both mathematical realism and mathematical anti-realism

144 

C. Rhode

insofar as these theses concern abstract entities that are aspects of reality beyond psycho-linguistic content. But this prudential restriction does not extend to mathematical agnosticism, nor to commitments regarding the existence of merely content-located mathematical entities. Having examined the implications of dialogical empiricism with regard to an example of an ontological debate about abstract entities, let us consider what implications dialogical empiricism might have for an example of an ontological debate about concrete entities. The example I want to discuss here concerns concrete unobservable entities, which provide a key focus for the debate between scientific realists and scientific anti-realists. Not only but especially in physical research, it is common practice to employ observable entities as evidence for conclusions about unobservable entities, such as black holes. In particular, observable entities are employed as evidence for the existence of such unobservable entities. In light of the foregoing results concerning commitments to the existence or non-existence of abstract entities, we could well expect that commitments to the existence or non-existence of such concrete unobservable entities must prudently be avoided. Before drawing that conclusion just yet, let us examine in a little more detail the implications arising from dialogical empiricism for existential claims about concrete unobservables. Dialogical empiricism implies that one must prudently refrain from invoking evidence of one kind in support of claims of a different kind. As explained in Chap. 3, Sect. 3.5, the evidence proposition is of a different kind from the hypothesis in question iff it is not logically possible to gather sufficient evidence of the kind exemplified by the evidence proposition in order to deductively prove the questioned hypothesis from that evidence. By contrast, if it is logically possible to gather sufficient evidence of the kind exemplified by the evidence proposition in order to deductively prove the hypothesis from the evidence thus gathered, then the evidence proposition and the hypothesis do not differ in kind. The extent of evidence that is currently available may of course still be insufficient in order to prove the questioned hypothesis from that evidence; but it may nonetheless be logically possible to gather evidence of sufficient extent for that purpose. According to the terminology introduced in Chap. 3, Sect. 3.5, in cases where gathering sufficient evidence of the kind exemplified by

5  Metaphysical Hypotheses 

145

the evidence propositions for the proof of the hypothesis in question is logically impossible, there is an inference barrier between the evidence proposition and the hypothesis; and where gathering sufficient evidence of this kind for the proof of the hypothesis in question is logically possible, there is (at most) an implication barrier between the evidence proposition and the hypothesis. And implication barriers can easily be circumvented by presenting the inference in support of a hypothesis as being ampliative rather than deductive, thus not claiming the hypothesis to be implied by the evidence propositions provided. To determine the consequences arising from dialogical empiricism for the debate between scientific realism and anti-realism, it is therefore necessary to assess whether (salient aspects of ) concrete observable entities (i.e. the evidence invoked) and propositions describing the same differ in kind from concrete unobservable entities (whose existence or otherwise is to be supported) and propositions describing them. To answer that question, we must consider whether it is logically possible to gather sufficient evidence in the form of aspects of concrete observables in order to prove the existence of some particular concrete unobservable entity, such as the hypothesised supermassive black hole at the centre of the Messier 87 galaxy, called Pōwehi. Is it logically possible to prove from aspects of concrete observables that there exists a seemingly unobservable black hole at the centre of Messier 87? Of course it is—because it is logically possible to travel any distance through spacetime nearly effortlessly, cross the event horizon of a black hole, pass through its spacetime singularity (all the while measuring Pōwehi’s properties and effects), re-emerge through the event horizon on the other side, and create some clever technology which enables the people so travelling to be neither disintegrated by the extreme curvature of spacetime of a black hole nor crushed by the relentless floods of matter against Pōwehi’s inner horizon. It is thus logically possible to fly out to Pōwehi and directly observe this seemingly unobservable entity— or, in short, to make this unobservable black hole observable. The more general lesson from this scientifically ridiculous thought experiment is the well-known relativity of the observable/unobservable separation: Whether a particular concrete entity is unobservable is wholly dependent upon the capacities of the would-be observer (cf. Maxwell 1962). (Indeed, these same capacities have seen recent improvements among humans,

146 

C. Rhode

leading to Pōwehi’s direct observation from Earth; Drake 2019.) In effect, then, it is logically possible to gather sufficient evidence from aspects of concrete observable entities in order to prove the existence of a particular, concrete, (previously) unobservable entity. Hence observables and unobservables do not differ in kind, nor do, therefore, propositions describing them. This continues to hold when we consider concrete observables and concrete unobservables in light of the inference barriers applied above. Both observables and unobservables like Pōwehi are aspects of the world beyond psycho-linguistic content rather than elements of such content;21 both are typically the subjects of descriptive propositions rather than of normative propositions; and both are typically the subjects of contingently true propositions rather than of necessarily true propositions. As it thus remains the case that concrete observables and concrete unobservables do not differ in kind, any inferential barrier between them would be an implication barrier and not an inference barrier. The prudential recommendations of dialogical empiricism, however, are only concerned with the avoidance of crossing inference barriers. Thus, dialogical empiricism remains entirely silent with regard to inferences from concrete observables to concrete unobservables (or between propositions respectively describing the same). Of course, this does not imply that such inferences are unproblematic, as that would equally not follow from dialogical empiricism. It simply means that dialogical empiricism does not have any apparent implications for the debate between scientific realism and scientific anti-realism. A final remark in this respect is due in order to answer potential concerns regarding evidence about concrete observable entities. The literally most obvious (and indeed obviously most literal) evidence about observables would of course consist in observation, or more generally in perception. But perception is an element of psycho-linguistic content, so invoking perception as evidence about world-located concrete observable entities would clearly cross the content/world barrier. To that I agree: If I  I disregard the idealist possibility that concrete entities are mere elements of psycho-linguistic content. 21

5  Metaphysical Hypotheses 

147

tried to persuade my interlocutor of the existence of the sadly empty cups of coffee on the table between us, I could invoke my or her perception of these cups. If I opted for my own perception, then it would be impossible to provide this evidence to my interlocutor. If I instead opted for her own perception (on the guess that she is perceiving these cups too), then, having already questioned the existence of the cups despite her perception of them, she would likely refocus her questioning of my claim to that existence by requesting support for the reliability of her perception as evidence about the world beyond perception (or other content). And that I would find effectively impossible to provide. Now, it might be argued yet again that this failure to persuade one’s interlocutor leads to global epistemological scepticism—for, if one cannot justify to an opponent something as obvious as a pair of cups on a table right in front of that opponent, then one also cannot justifiably believe in the existence of these cups, or of almost anything else. I would, however, remind anyone putting forth such an argument of the attitude/dialogue separation. As discussed particularly in Chap. 2, Sect. 2.2 above, failure to persuade someone else of P does not entail that one no longer attitudinally commits to P oneself, nor that one ought to refrain from continued commitment to P oneself.

5.6 Quietism Let us review the consequences of dialogical empiricism that have been outlined in the present chapter. First, norms are effectively unsupportable in philosophical dialogue, whether they be aspects of reality beyond psycho-­linguistic content or elements of such content. Neither descriptive evidence will be of any use, due to the is/ought separation, nor content-­located evidence, due to the content/world separation. Yet if norms are substituted by Machiavellian prudential recommendations, then they can nonetheless be supported without crossing any inference barrier, by invoking (typically social scientific) facts as evidence within a deliberation dialogue. Second, necessary truths are effectively unsupportable in philosophical dialogue, whether they be aspects of reality beyond psycho-linguistic

148 

C. Rhode

content or elements of such content. Neither contingent nor content-­ located evidence will be of any use, due to the contingent/necessary separation and the content/world separation. Third, if knowledge were an aspect of reality beyond psycho-linguistic content, then hypotheses about knowledge would be effectively unsupportable in philosophical dialogue, since content-located evidence would prudently have to be avoided while world-located evidence would seem unavailable. But as knowledge very much appears to be an element of psycho-linguistic content, content-located evidence can nonetheless be employed in support of hypotheses about knowledge (so long as the breaching of further inference barriers is avoided). Fourth, if abstract entities are aspects of reality beyond psycho-­ linguistic content, then claims to their existence or non-existence are effectively unsupportable in philosophical dialogue, though this is not the case if such entities are fictional elements of psycho-linguistic content. (No parallel implications arise with regard to concrete unobservable entities.) In light of these consequences, it is worth reiterating that dialogical empiricism is not itself a normative, modal, epistemic or ontological thesis. Dialogical empiricism does not imply that some particular action is forbidden, that some particular truth is necessary, that some particular fact is unknown, or that some particular entity exists, nor any other substantive conclusion. Instead, dialogical empiricism is a purely methodological thesis. Yet even so, the four consequences just summarised all exemplify the general metaphilosophical lesson arising from dialogical empiricism. Let us first note that hypotheses (or propositions) expressing an aspect of reality beyond psycho-linguistic content may adequately be described as metaphysically significant hypotheses (or propositions), or simply as metaphysical hypotheses (or propositions). I will therefore take metaphysical hypotheses to encompass all and only those hypotheses that express an aspect of reality beyond psycho-linguistic content. Given this characterisation, the general lesson from the consequences of dialogical empiricism is the following: Parties to an analytic philosophical dialogue must prudently refrain from committing to the truth or falsity of any metaphysical

5  Metaphysical Hypotheses 

149

hypothesis if this hypothesis lacks evidential support expressed in metaphysical propositions.22 It is worth noting that metaphysical hypotheses which do not lack evidential support from metaphysical propositions are more typically investigated in the natural sciences, rather than in analytic philosophy.23 (This not least includes hypotheses about some concrete, putatively unobservable entities like black holes.) By contrast, metaphysical hypotheses which do lack evidential support from metaphysical propositions rarely receive treatment in the natural sciences and are often of significant philosophical interest instead. (Among their number are hypotheses about worldlocated norms, necessities, knowledge, and abstract entities.) As Gary Gutting aptly noted, ‘every modern philosophical enterprise has had to guarantee a place for itself by showing that there is something for it to know that escapes the grasp of empirical science’ (2001: 50). Of course, not all metaphysical hypotheses that lack evidential support from metaphysical propositions are of philosophical interest—as the metaphysically significant contents of, say, religious dogmata are seldom if ever the subject of analytic philosophical debate (though their nature and status may well be). So we may characterise metaphysical hypotheses which (a) lack evidential support from metaphysical propositions, and which (b) are actually debated by parties to an analytic philosophical dialogue, as metaphysical hypotheses of philosophical interest, or simply as philosophically interesting metaphysical hypotheses. This terminological choice excludes both metaphysical hypotheses that do not lack evidential support from metaphysical propositions, as typically investigated in the natural sciences, and those that do lack such support but are also not the subject of any analytic philosophical debate.24 On this basis, we can restate the

 A hypothesis which actually lacks such support, but does not lack it necessarily, must nonetheless be avoided. After all, you are very unlikely to successfully persuade another party that H by invoking the mere possibility that there may be appropriate evidence in support of H. 23  Such hypotheses are also sometimes investigated in the philosophy of science, although such research is typically conducted through types of dialogue other than persuasion dialogue, as I will discuss in Chap. 6, Sect. 6.2. Cf. fn. 27 in Chap. 1, Sect. 1.3. 24  My concern here is with the actual rather than the possible availability of evidence (and equally with the actual interest of dialogue parties) because, in essence, the aim of this work remains to contribute to deliberation, rather than to persuasion or other goals; cf. fn. 22 in this section. 22

150 

C. Rhode

general lesson in the following way: Parties to an analytic philosophical dialogue must prudently refrain from committing to the truth or falsity of any philosophically interesting metaphysical hypothesis. On this general lesson it should further be noted that the prudential requirement to avoid committing to the truth or falsity of a hypothesis renders it prudent to avoid engaging in persuasion dialogue about said hypothesis altogether. For if one does not make a dialogical truth-value commitment regarding the given hypothesis, then one can practically only fail to persuade one’s opponents in the game of persuasion, which is of course to be avoided. To give the final statement of the general conclusion, then: Parties to an analytic philosophical dialogue must prudently refrain from engaging in dialogue about any philosophically interesting metaphysical hypothesis.

Or, to give this conclusion its terminologically deconstructed statement: Parties to an analytic philosophical dialogue must prudently refrain from engaging in dialogue about any hypothesis expressing an aspect of reality beyond psycho-linguistic content if that hypothesis lacks evidential support from propositions expressing an aspect of reality beyond psycho-linguistic content.

The obvious implication of this general conclusion is that any debate within analytic philosophy that concerns the truth or falsity of what I have called philosophically interesting metaphysical hypotheses must prudently be terminated. This means not least that analytic metaphysics must prudently be excised from philosophical dialogue.25 But it also means that philosophical discussions outside of analytic metaphysics must equally be discontinued if they debate world-located norms, necessities, knowledge, or abstract entities, among other things. However, this general conclusion does not imply the termination of philosophical  Analytic metaphysics would exclude naturalistic metaphysics as advocated by Ladyman et al. (2007, cf. Maclaurin and Dyke 2012), but of course only to the extent that naturalistic metaphysics is free from debates about philosophically interesting metaphysical hypotheses as defined above. 25

5  Metaphysical Hypotheses 

151

debates that exclusively have elements of psycho-linguistic content as their object (so long as the breaching of further inference barriers is avoided, as might be achieved, for instance, in debates about the content-­ located kind knowledge). I have little doubt that the excision of philosophically interesting metaphysical debates from current analytic philosophy may adequately be regarded as the most notable recommendation arising from dialogical empiricism. But in light of this metaphysically quietist recommendation, dialogical empiricism may itself be regarded as more closely related to other anti-metaphysical programmes, such as naturalism (cf. Papineau 2015) or logical positivism (cf. Ayer 1936), than it actually is. On the one hand, naturalists maintain that philosophical research, in metaphysics and elsewhere, ought to be properly informed by, consistent with, and/or continuous with research in the natural sciences. Metaphysical debates about hypotheses which lack evidential support from metaphysical propositions fail to satisfy this naturalistic demand, with the consequence that such debates ought to be discontinued (cf. Maclaurin and Dyke 2012). Although I have considerable sympathy for naturalism, I sought to refrain from relying on any distinctly naturalistic premises in the derivation and subsequent application of dialogical empiricism, so as not to risk undermining the present work if any such naturalistic premises were to be rejected.26 Besides, naturalists strongly tend to espouse a cognitive (rather than critical) philosophical self-conception, which is not supported by the evidence presented in Chap. 1, Sect. 1.3. On the other hand, dialogical empiricism also fundamentally differs from logical positivism—which difference becomes clear when we note that the metaphysical quietism espoused by the logical positivists was a form of semantic quietism (as epitomised by Ludwig Wittgenstein; cf. Virvidakis 2008: 162; Macarthur 2017: 251-2). It stipulated that an expression is meaningful iff it is either analytically true or synthetically verifiable—whereupon metaphysical hypotheses lacking any possibility of evidential support from metaphysical propositions were deemed

26

 Cf. Chap. 3, Sect. 3.6.

152 

C. Rhode

meaningless.27 In contrast to this semantic form, the metaphysical quietism arising from dialogical empiricism is not semantic but aporetic, following in the Pyrrhonian tradition.28 Under dialogical empiricism, participation in analytic metaphysical dialogues is identified, not as meaningless, but as imprudent in the reader’s pursuit of the dialogical goal of persuasion in analytic philosophical dialogue.29 It may also be noted that dialogical empiricism is not vulnerable to the major criticisms raised against logical positivism, neither being reflexively self-defeating (like verificationism)30 nor depending upon the analytic/synthetic distinction.31 Notwithstanding these differences, I am not denying32 that logical positivism, naturalism, and dialogical empiricism do share a metaphysically quietist conclusion: Like the others, dialogical empiricism (with its associated inference barriers) equally seeks to stand as the proverbial thin red line, between metaphysicians with their psycho-linguistic shadow images in their Platonic cave, and the real world beyond.

References Ayer, A.J. 1936. Language, Truth and Logic. London: Gollancz. Ball, D. 2017. The Myth of the Intuitive: Experimental Philosophy and Philosophical Method, by Max Deutsch. Australasian Journal of Philosophy 95 (2): 414–415. Brock, S., and E. Mares. 2007. Realism and Anti-Realism. Durham: Acumen.

 Carnap’s (1950) logical empiricism equally qualifies as espousing a form of semantic quietism, through his distinction between questions that are internal or external to a linguistic framework. Cf. fn. 2 in Chap. 4, Sect. 4.2. 28  In practice, this Pyrrhonian or aporetic quietism does not generalise explosively across all areas of philosophical dialogue, as explained in Chaps. 3 and 2, Sects. 3.3 and 2.2. 29  Cf. fn. 7 in Sect. 5.2. 30  Cf. Chap. 2, Sect. 2.3. 31  Note that dialogical empiricism also neither presupposes nor entails the a priori/a posteriori distinction. Note also that this latter distinction is not co-extensive with the content/world separation, as exemplified not least by experimental philosophers’ a posteriori research of people’s psycho-­ linguistic content. 32  So, evidently, I am denying – cf. Chap. 2, Sect. 2.2. 27

5  Metaphysical Hypotheses 

153

Bueno, O. 2009. Mathematical Fictionalism. In New Waves in Philosophy of Mathematics, eds. O. Bueno and Ø. Linnebo, 59–79. Basingstoke: Palgrave Macmillan. ———. 2014. Nominalism in the Philosophy of Mathematics. In The Stanford Encyclopedia of Philosophy, Spring 2014 Edition, ed. E.N. Zalta. http://plato. stanford.edu/archives/spr2014/entries/nominalism-­mathematics/. Carnap, R. 1950. Empiricism, Semantics, and Ontology. Revue Internationale de Philosophie 4 (11): 20–40. Cowan, R. 2015. Clarifying Ethical Intuitionism. European Journal of Philosophy 23 (4): 1097–1116. Deutsch, M. 2015. The Myth of the Intuitive: Experimental Philosophy and Philosophical Method. Cambridge, MA: MIT Press. Drake, N. 2019. First-ever Picture of a Black Hole Unveiled. In National Geographic. https://www.nationalgeographic.com/science/article/first-­ picture-­b lack-­h ole-­revealed-­m 87-­e vent-­h orizon-­t elescope-­a strophysics. Accessed 17 July 2019. Femia, J.V. 2012. Pareto, Machiavelli, and the Critique of Ideal Political Theory. In Vilfredo Pareto: Beyond Disciplinary Boundaries, eds. J.V.  Femia and A.J. Marshall, 73–83. Farnham: Ashgate. Foucault, M. 1972 [1969]. The Archaeology of Knowledge. Trans. A.M. Sheridan Smith. London: Tavistock. Gutting, G. 2001. French Philosophy in the Twentieth Century. Cambridge: Cambridge University Press. Hariman, R. 2003. Theory Without Modernity. In Prudence: Classical Virtue, Postmodern Practice, ed. R. Hariman, 1–32. University Park, PA: Pennsylvania State University Press. Hobbes, T. 2012 [1651]. Leviathan. In The Clarendon Edition of the Works of Thomas Hobbes, Vols. 3–5, ed. N. Malcolm. Oxford: Oxford University Press. Hooker, B. 2000. Ideal Code, Real World: A Rule-Consequentialist Theory of Morality. Oxford: Oxford University Press. Hume, D. 1978 [1739–40]. A Treatise of Human Nature. Ed. L.A. Selby-Bigge, rev. P.H. Nidditch, 2nd ed. Oxford: Clarendon Press. Kornblith, H. 2002. Knowledge and Its Place in Nature. Oxford: Clarendon Press. ———. 2007. Naturalism and Intuitions. Grazer Philosophische Studien 74: 27–49. ———. 2014. Is There Room for Armchair Theorizing in Epistemology? In Philosophical Methodology: The Armchair or the Laboratory? ed. M.C. Haug, 195–216. New York: Routledge.

154 

C. Rhode

Ladyman, J., D. Ross, D. Spurrett, and J. Collier. 2007. Every Thing Must Go: Metaphysics Naturalized. Oxford: Oxford University Press. Lange, M. 2014. Laws of Nature. In The Routledge Companion to Philosophy of Science, eds. M. Curd and S. Psillos, 2nd ed., 235–244. Abingdon: Routledge. Liggins, D. 2010. Epistemological Objections to Platonism. Philosophy Compass 5 (1): 67–77. Longino, H.E. 1996. Cognitive and Non-Cognitive Values in Science: Rethinking the Dichotomy. In Feminism, Science, and the Philosophy of Science, eds. L.H. Nelson and J. Nelson, 39–58. Dordrecht: Kluwer. Macarthur, D. 2017. On Metaphysical Quietism and Everyday Life. In The Cambridge Companion to Philosophical Methodology, eds. G.  D’Oro and S. Overgaard, 249–273. Cambridge: Cambridge University Press. Machiavelli, N. 1988 [1532]. The Prince. Eds. and Trans. Q.  Skinner and R. Price. Cambridge: Cambridge University Press. Maclaurin, J., and H.  Dyke. 2012. What is Analytic Metaphysics For? Australasian Journal of Philosophy 90 (2): 291–306. Maddy, P. 1990. Realism in Mathematics. Oxford: Clarendon Press. Maxwell, G. 1962. The Ontological Status of Theoretical Entities. In Scientific Explanation, Space, and Time, eds. H.  Feigl and G.  Maxwell, 3–27. Minneapolis: University of Minnesota Press. Moad, O.E. 2009. Comparing Phases of Skepticism in Al-ghazālī and Descartes: Some First Meditations on Deliverance from Error. Philosophy East and West 59 (1): 88–101. Najm, S.M. 1966. The Place and Function of Doubt in the Philosophies of Descartes and Al-Ghazālī. Philosophy East and West 16 (3–4): 133–141. Papineau, D. 2006. The Tyranny of Common Sense. The Philosophers’ Magazine 34: 19–25. ———. 2015. Naturalism. In The Stanford Encyclopedia of Philosophy, Fall 2015 Edition, ed. E.N. Zalta. http://plato.stanford.edu/archives/fall2015/entries/ naturalism/. Pilkington, E. 2011. US Man Stages $1 Bank Robbery to Get State Healthcare. In The Guardian. https://www.theguardian.com/world/2011/jun/21/verone-­ one-­dollar-­robbery-­healthcare. Accessed 17 June 2015. Thomasson, A.L. 2014. The Easy Approach to Ontology: A Defense. In Philosophical Methodology: The Armchair or the Laboratory? ed. M.C. Haug, 107–125. New York: Routledge.

5  Metaphysical Hypotheses 

155

Vaidya, A. 2015. The Epistemology of Modality. In The Stanford Encyclopedia of Philosophy, Summer 2017 Edition, ed. E.N. Zalta. http://plato.stanford.edu/ archives/sum2017/entries/modality-­epistemology/. van Roojen, M. 2014. Moral Intuitionism, Experiments, and Skeptical Arguments. In Intuitions, eds. A.R. Booth and D.P. Rowbottom, 148–164. Oxford: Oxford University Press. Virvidakis, S. 2008. Varieties of Quietism. Philosophical Inquiry 30 (1–2): 157–175. Wallace, D. 2012. The Emergent Multiverse: Quantum Theory According to the Everettian Interpretation. Oxford: Oxford University Press. Williamson, T. 2000. Knowledge and Its Limits. Oxford: Oxford University Press.

6 Escaping Dialogical Empiricism

6.1 Looking Behind1 Metaphysically inclined philosophers are understandably unlikely to welcome the recommendation that their work be excised from analytic philosophy, and may seek to resist such excision. The purpose of this final chapter is to outline various avenues that are and remain open to metaphysicians in order to mitigate or escape the consequences of dialogical empiricism. First, however, it will be useful to briefly remind ourselves of the series of considerations that gave rise to this metaphysical quietism.

 The titles of  Chap. 1, Sect. 1.4 and  Sect. 6.1—‘Looking Ahead’ and  ‘Looking Behind’—are of course inspired by a memorable exchange between Thorin and Gandalf in J. R. R. Tolkien’s The Hobbit, taking place after the company defeated the three mountain trolls. As Tolkien writes at the end of chapter 2 (1937: 42): 1

‘“Where did you go, if I may ask?” said Thorin to Gandalf as they rode along. “To look ahead,” said he. “And what brought you back in the nick of time?” “Looking behind,” said he.’

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1_6

157

158 

C. Rhode

In Chap. 1 I identified an inference barrier separating commitments to the truth-value of a proposition that are made in attitude from those that are made in dialogue, and I explained that no commitment in one of these two locations implies the corresponding commitment or any other commitment in the other location. With this separation in place, my focus was entirely upon commitments in dialogue, with no need to consider the attitudinal (and especially potentially sceptical) implications of the subsequent considerations, because these considerations would not have any such implications. In that chapter I also reported original empirical research of recent journal articles in analytic philosophy, conducted on the basis of the typology of dialogues developed by Douglas Walton and colleagues. The results of this empirical research showed that it is the vastly prevalent dialogical goal of parties to analytic philosophical dialogues to persuade their opponent, or in effect to defeat their opponent in argument, and that this goal must prudently be presumed to underlie any given analytic philosophical dialogue. Chapter 2 was devoted to the question of whether the burden of proof in philosophical dialogue is incurred equally by any party that asserts a proposition and finds the same questioned by their opponent, or whether certain propositions, perhaps those that are very basic, are privileged to the effect that any party questioning them itself incurs the burden of proof for the proposition’s negation. I have discussed existing arguments in favour of either an even or a privileged allocation of the burden of proof, such as the contention that an even allocation would lead to scepticism, or the charge that a privileged allocation would be dogmatic, and I have dismissed all of the existing arguments—principally on grounds of the attitude/dialogue separation and the goal of persuasion. On the basis of said goal, I then argued that it is prudently necessary to allocate the burden of proof evenly without privilege in philosophical dialogue, as a privileged, uneven allocation would be persuasively useless. In Chap. 3 I considered how the burden of proof imposed upon a proposition can be met, and in particular what forms of evidential support must prudently be avoided due to their making it practically impossible to meet that burden of proof. As I have argued there, in supporting a questioned proposition of one kind (such as a normative proposition), one must prudently avoid employing evidence that is expressed by

6  Escaping Dialogical Empiricism 

159

propositions of a different kind (such as descriptive propositions). The reason for this prudential requirement is that breaching the inference barrier between two kinds is persuasively ineffective. Combining this requirement with the even allocation of the burden of proof, I then formulated the thesis of dialogical empiricism, which states that the burden of proof for an asserted and questioned proposition in philosophical dialogue (a) lies upon the asserting party and (b) must be satisfied without crossing any inference barrier. Chapter 4 served to spell out some of the consequences of dialogical empiricism for philosophical methods, for which purpose I focused on the inference barrier between the content of our talk or thought on the one hand, and the world beyond such psycho-linguistic content on the other hand. Examining both evidence and hypotheses in philosophy, I argued that evidence consisting of psycho-linguistic content must prudently not be invoked in support of hypotheses expressing an aspect of the world beyond such content. In particular, such evidence as intuitions, common sense, reflective equilibria, or traditional conceptual analyses prudently cannot be employed in support of a questioned hypothesis if that hypothesis concerns such things as Platonic forms, natural kinds, or any other aspect of reality beyond psycho-linguistic content. In Chap. 5 I turned from methodological to substantive consequences of dialogical empiricism, employing not merely the content/world barrier but also other inference barriers. In light of these barriers (and resultant restrictions upon philosophical methods already specified), I examined normativity, necessity, epistemology and ontology. I found that hypotheses about what is right or wrong, what is necessarily true, or what abstract entities do or do not exist, among others, are left effectively unsupportable if they are taken to hold true as aspects of the world beyond the confines of people’s talk or thought. This unsupportability extends to any hypothesis about aspects of reality beyond content that is not supported by propositions expressing aspects of reality beyond content; or, in approximate effect, it extends to any metaphysically significant hypothesis of philosophical interest. As it is persuasively imprudent to even debate unsupportable hypotheses, I thus concluded that any debate in analytic philosophy that concerns philosophically interesting metaphysical hypotheses must prudently be terminated.

160 

C. Rhode

This is a stark result, particularly but not only for philosophers working in analytic metaphysics. And yet, not all is lost. There are some more or less appealing avenues of escape from dialogical empiricism and its quietist consequences. It is the purpose of the remainder of this final chapter to outline and briefly assess these escape routes—including ways to accommodate the prudential excision of metaphysics (in the next section), as well as ways to prudently disregard this excision altogether (in the final section).

6.2 A Way Around? Metaphysicians might try to accommodate in some way or another the imprudence of conducting their work, and here I want to sketch a few ways in which this could be done. The first such way is also arguably the most obvious one, which is to refrain from hypothesising about aspects of reality beyond content. That is to say, one way for metaphysicians to accommodate the foregoing conclusion is to restrict their philosophical interests to psycho-linguistic content, such as the analysis of content-­ located concepts. (Indeed, this is precisely the restriction favoured by Alvin Goldman and opposed by Hilary Kornblith; cf. Kornblith 2007). No standard philosophical support for hypotheses expressing an element of psycho-linguistic content would be at risk of crossing the content/ world separation. Restriction could be minimised even further if one were to discuss philosophically interesting hypotheses only under a coherence theory of truth. For instance, one might discuss such hypotheses only with regard to their inclusion and position within a Quinean web of belief, thus embedding such discussion within an already well-used philosophical framework. According to Quine’s model of a web of belief, as well as according to any other theory relying on a coherence theory of truth, the truth condition of a given proposition is, roughly, the absence of any mismatch between this proposition and other propositions regarded as true. Coherence theories thus entrap truth within content, since propositions regarded as true are simply propositions whose truth has been committed to. As such, making a claim about the world beyond content

6  Escaping Dialogical Empiricism 

161

under a coherence theory of truth amounts to something like making a claim about the world beyond content as depicted within our psycho-­ linguistic content. So while coherence-based hypotheses might seem like the proverbial get-out-of-jail-free card, even they share the not insignificant drawback that is inherent to this first way of accommodating the foregoing results. This drawback is that, by refraining from hypothesising about aspects of reality beyond content, claims about reality as it really is outside of psycho-­linguistic content are nonetheless left prudently prohibited. After all, this approach merely offers an accommodation, rather than a refutation, of the imprudence of debating philosophically interesting metaphysical hypotheses. I suspect, however, that many moral philosophers, for instance, would find this accommodation quite unsatisfactory, as they would be unwilling to only make normative claims on the express restriction that anyone not sharing the relevant psycho-linguistic content would not be bound by those norms.2 The same unwillingness will likely be found across various other philosophical sub-disciplines. It is worth noting an unsuccessful variation on the preceding approach that, rather than opting for a coherence theory of truth, instead restricts claims about reality beyond content to in some sense conditional claims. Much philosophical dialogue is spent on debates about conditional claims. To take but one example, Justin Clarke-Doane (2012) argues that if one accepts the evolutionary challenge to moral realism, then one ought also to accept the evolutionary challenge to mathematical realism.3 Restricting their claims about the world to conditional claims might thus allow philosophers to retain large parts of metaphysics by spinning a web of complex interconnected conditionals that nonetheless reflects reality—even though this web would always hang in suspension (like a huge spider web between trees) and never touch real ground, as it does not include any unconditional claim about reality.4  Such moral philosophers also have to contend with the is/ought barrier, of course, as discussed in Chap. 5, Sect. 5.2. 3  Cf. Appendix C. 4  Of course this approach cannot be to make two separate conditional claims with mutually negating antecedents and identical consequents—e.g. If A then B; furthermore, if ¬A then B. This would practically amount to the unconditional claim that B and would here be unpersuasive. 2

162 

C. Rhode

Conditionality in any case requires an antecedent claim and a consequent claim, and in any case, a conditional claim about reality beyond content must have as its ultimate consequent an unconditional claim about reality beyond content.5 The antecedent of such a conditional claim could nonetheless be either about content or about reality beyond, thus yielding two possible kinds of conditional claims about reality beyond content. First, in the case of a conditional claim about reality beyond content with an antecedent claim about content, this conditional claim would be an inferential bridge premise breaching the content/ world barrier. Such a claim would state that if some particular element of psycho-linguistic content obtains then some particular aspect of reality beyond content obtains as well, thus crossing the inference barrier between content and the world. And second, in the case of a conditional claim about reality beyond content with an antecedent claim about reality beyond content, this claim would be altogether about reality, which must prudently be avoided for philosophically interesting claims. So a web of interconnected conditionals does not offer a way of debating reality beyond psycho-linguistic content whilst accommodating the quietism derived above. A perhaps more successful and appealing way of accommodating the foregoing results arises when we remember that these results have only been obtained for philosophical persuasion dialogue. Dialogical empiricism does not entail any prudential restrictions on analytic philosophers that would bar them from discussing the world beyond content within dialogues not instantiating persuasion dialogue. Instead, dialogical empiricism is entirely silent about the conduct of philosophy in types of dialogue other than persuasion dialogue. So, which other types of dialogue might lend themselves to philosophical engagement with aspects of reality beyond psycho-linguistic content?

 The consequent may itself be a conditional claim, which may in turn have as its consequent a conditional claim, and so forth. But still, the original and widest-scoped conditional claim would not be about reality beyond content unless the ultimate and most narrowly-scoped conditional claim within it had as its consequent an unconditional claim about reality beyond content. The ultimate and final consequent must thus be an unconditional claim about reality beyond content in order for the original conditional claim to be about reality beyond content. 5

6  Escaping Dialogical Empiricism 

163

Information dialogue, negotiation dialogue and eristic dialogue all seem manifestly unsuitable. The mere gathering of information would hardly warrant the sustenance of dozens of professorships in metaphysics, particularly in light of the ease of finding information through Google and Wikipedia.6 Bartering about joint acceptance of the existence of colours in exchange for joint acceptance of the non-existence of possibilia, as in negotiation dialogue, would equally deprive metaphysics of any semblance of a raison d’être that it might seek to preserve. And engaging in any non-negligible amount of eristic dialogue in one’s work as a philosopher would also provide a promising means of initiating a career change. Inquiry dialogue and discovery dialogue, in turn, would demand that a hypothesis be submitted to systematic testing against salient evidence. Although I have not investigated inquiry or discovery dialogue in the present work, it seems highly likely that such salient evidence for a metaphysically significant hypothesis would have to consist in aspects of reality beyond content—in which case the given hypothesis would allow for scientific treatment. Some hypotheses investigated in the philosophy of science do satisfy this condition, as illustrated by instances of inquiry dialogue within the empirical sample in Chap. 1, Sect. 1.3.7 Yet anywhere else in analytic philosophy, metaphysically significant hypotheses lack support from propositions about aspects of reality beyond content, and so inquiry or discovery dialogue could simply not be carried out successfully about those hypotheses (cf. Kornblith 2007: 30–5). Such dialogue would thus not enable metaphysicians to engage with hypotheses of their current interest. This leaves deliberation dialogue, a type that appears somewhat incoherent when applied to metaphysical hypotheses of philosophical interest. For instance, it makes no sense to ask whether it is prudently advisable that the number 23 exists as an aspect of reality beyond anyone’s psycho-­ linguistic content. (What, or who, would the existence of 23 be advisable for?) As explained earlier, deliberation dialogue has as its object a set of  Like countless young academics today, I owe a very special thanks to the generous people at Wikipedia, whose work has helped my understanding during my undergraduate days far more often than I was allowed to acknowledge. 7  Cf. fn. 23 in Chap. 1, Sect. 1.3. 6

164 

C. Rhode

actions rather than a proposition, rendering deliberation dialogue about a proposition incoherent. One might suspect that deliberation dialogue may allow moral and political philosophers, among others, to nonetheless debate and derive prescriptions of action that are in some sense deemed appropriate—and indeed, this is the case. Yet this option comes with the significant limitation that the applicability of any prescriptive conclusion to be derived would depend upon the reader’s actual situation and actual goal, as explained in Chap. 5, Sect. 5.2. That would be unlikely to satisfy many moral and political philosophers who seek to derive prescriptions that are binding regardless of people’s goals and situations (cf. Femia 2012). With the alternative types of dialogue exhausted, this second way of accommodating the imprudence of conducting analytic metaphysics and related debates thus seems rather unappealing as well.

6.3 A Way Out? If accommodating the prudential excision of metaphysics proves elusive, then one might seek to altogether reject this prudential recommendation. As just reiterated, prudential recommendations are goal-dependent—so the obvious angle of attack against my foregoing quietist conclusion would be to take issue with the goal that I have presumed throughout the present work. To reiterate, the goal of any party to an analytic philosophical dialogue can safely be assumed to be to defeat her opponent in argument (and not be defeated herself ). This goal has been discovered empirically through the analysis of a representative sample of (contributions to) analytic philosophical dialogues, and it would seem difficult to argue such empirical results away. Yet this goal is a dialogical goal: It only expresses the goal pursued in dialogue, but it does not express the attitudinal goal that might lead someone to engage in a given dialogue in the first instance. Perhaps the purpose of engaging in analytic philosophical dialogue about the world beyond psycho-linguistic content is to create new knowledge about the world as it really is. Yet if that is indeed one’s motivation for engaging in philosophical dialogue, then (as already explained in Chap. 1, Sect. 1.3) one would be conducting a consensus-directed dialogue for truth-directed

6  Escaping Dialogical Empiricism 

165

purposes, which would turn out to be both imprudent and unsuccessful. It would be imprudent because a truth-directed dialogue would be far more conducive to such truth-directed purposes, while using a consensus-­ directed dialogue would presume without obvious warrant that philosophical consensus about such issues as the existence of ultimate smallest particles was indicative of that same existence (regardless of quantum-­ physical evidence to the contrary). And it would be unsuccessful because enduring philosophical consensus has forever proved elusive, receding before us like the green light across the bay. So if metaphysically inclined philosophers do indeed have truth-directed purposes, then there are independent prudential reasons for abandoning what would thus be the misdirected failure called current analytic philosophical dialogue about philosophically interesting metaphysical hypotheses. As discussed in Chap. 1, Sect. 1.3, it seems at least more charitable to assume that the attitudinal goal motivating metaphysicians’ participation in analytic philosophical dialogue is, in line with the dialogical goal, directed towards achieving consensus. But perhaps the motivation for engaging in such dialogue at all is the desire to clarify one’s commitments (whether personal or shared, whether attitudinal or dialogical), to understand their implications, and to resolve contradictions among them. Indeed, Thomas Scanlon (2003) considers such elucidation to be the very purpose of the method of reflective equilibrium that many philosophers employ on a daily basis.8 Maybe the pursuit of consistency, or at least transparency, among one’s commitments is the principal reason for philosophers to not just employ this method but to engage in philosophical and especially metaphysical dialogue at all. If that was true, then engaging in philosophical persuasion dialogue about philosophically interesting metaphysical hypotheses would seem neither imprudent nor obviously unsuccessful. One might very well get clearer about one’s commitments and their implications by ostensibly trying to persuade others and thereby testing one’s commitments in what Walton (1992: 126) calls the ‘marketplace of persuasion’. If all this is the case, then we should expect most analytic philosophical dialogues to instantiate persuasion dialogue—as has indeed been established in Chap.  Cf. fn. 9 in Chap. 4, Sect. 4.4.

8

166 

C. Rhode

1, Sect. 1.3. And yet, this prevalent type of dialogue would not imply or even indicate that persuading one’s philosophical opponent was the attitudinal goal of such dialogue, thus undermining the foundation of the quietism advocated here. This may indeed be the case for many participants to various philosophical debates, particularly including those with a more therapeutic perspective on philosophy. It looks, though, as if this is not the case in the debates principally affected by the quietist consequences of dialogical empiricism. In analytic metaphysics, for instance, it would seem as if parties that defend claims about reality beyond content are hoping to influence the commitments of members of other parties (perhaps, ostensibly, by creating ‘knowledge’) and not merely to clarify their own commitments. Extensive use of the word ‘surely’ must surely be accepted as a good general indicator to that effect, as this word seems to be employed chiefly when one wishes to have a claim accepted without subjecting this claim to debate in the ‘marketplace of persuasion’. The critical comments in the defence of scientism by James Ladyman and colleagues (Ross et al. 2007) reinforce the impression that the goal pursued in metaphysical debates is one of persuasion rather than elucidation, illustrated particularly well in reference to Elder hypothesising about ‘the subatomic microparticles that future physics will discover to be the truly fundamental building blocks of the physical world’ (2004: 51).9 Similarly, in applied ethics and normative moral theory, the goal of dialogue parties strongly appears to be to convince their interlocutors that, for instance, abortion ‘really is wrong’. This goal seems to go well beyond mere clarification of a given party’s own commitments. More targeted empirical research of particularly the attitudinal goals of analytic philosophers in these sub-­ disciplines may be beneficial here, examining their explicitly declared attitudinal goals through both comprehensive surveys and extensive textual analyses. But I am quite confident in the conjecture that philosophers participating in debates about what I have termed philosophically interesting metaphysical hypotheses strongly tend to desire and pursue through such dialogue an attitudinal and dialogical goal of persuasion. If this conjecture is accurate, then it seems that no invocation of merely  Cf. Chap. 4, Sect. 4.5.

9

6  Escaping Dialogical Empiricism 

167

elucidatory goals will serve to undermine the prudential force of the quietism that arises from dialogical empiricism. Instead of invoking different attitudinal goals, metaphysically inclined philosophers might simply give priority to personal goals that are essentially unconnected to commitments in either dialogue or attitude. As explained, dialogical empiricism is premised upon the dialogical goal of persuasion; so, its prudential recommendations neither apply under a different goal, nor does it offer any resources for resolving priority conflicts between this goal and any other goal. Hence, if a personal goal is given priority over the dialogical goal of persuasion, then any prudential recommendations arising from this personal goal would also take priority over the prudential recommendations arising from the goal of persuasion, and thus they would take priority over dialogical empiricism. One such personal goal may be the goal of avoiding challenges to one’s commitments. In this regard, metaphysically inclined philosophers might try to ensure that certain propositions are never actually questioned, so a burden of proof could never be incurred in the first instance. Such silencing of opposition might be accomplished by more-or-less explicitly stipulating the proposition to be insulated, and by making it difficult for challenges to this proposition to receive a serious hearing within the discipline. Dissenting philosophers might be tacitly excluded from relevant sub-disciplinary circles (cf. Cummins 1998: 115–6), and dissenting would-be philosophers might be discouraged from continuing to study philosophy (cf. Buckwalter and Stich 2014).10 Justification for such exclusion may be sought under the pretext of a Kuhnian paradigm,11 whereby those who do not accept the key methodological and ontological theses arising from certain paradigmatic works of philosophy may supposedly legitimately be excluded from the discipline. If the exclusion of dissenting voices is sufficiently effective (and if dispute about the exact formulation of the key methodological and ontological theses within a paradigm is sufficiently restricted), then one may indeed avoid challenges to certain commitments.  Such disciplinary exclusion may or may not occur in precisely the manner suggested by these publications—though it would be surprising, from a sociology of knowledge perspective, if such advertent or inadvertent exclusion was entirely absent. 11  Cf. Chap. 2, Sect. 2.2; Levy (2003). 10

168 

C. Rhode

Among the propositions thus protected might importantly be inferential bridge premises, such as the knowledge norm of assertion,12 that enable philosophers to cross certain inference barriers. If certain bridge premises are indeed insulated from challenge, then the role that the possibility of their questioning has played in the above derivation of my quietist conclusion would be undermined, as would be that same quietism. So, for instance, the possibility of questioning the knowledge norm of assertion had the consequence of allowing me to rely on the separation between attitude and dialogue as an inference barrier throughout the present work. But if the knowledge norm was successfully protected from challenge, then this norm could be used for inferences between attitudinal and dialogical commitments, which would nullify the eponymous inference barrier and thus fundamentally undermine dialogical empiricism. Prioritising the goal of avoiding challenges to certain commitments (such as certain norms or other bridge premises) over the dialogical goal of persuading one’s opponents may thus undermine dialogical empiricism and its quietist consequences. It will only do so, however, if one manages to effectively silence any potential challenge to certain propositions. Since philosophical dialogues may continue over centuries, ‘with one contributor taking over from another’ (Cohen 1986: 3), the silencing of any potential challenge to selected propositions would have to be enforced practically forever. Yet to rely on the successful silencing of a questioning party over many centuries would seem a little optimistic, to put it mildly. In practice, one cannot realistically expect such perpetual silencing to succeed, and hence one cannot prudently rely on the protection from challenge of the knowledge norm or other inferential bridge premises. As such, the goal of avoiding challenges to one’s commitments seems practically unattainable. (Hence, notably, inferential bridge premises like the knowledge norm of assertion effectively cannot be employed to undermine dialogical empiricism.) Another personal goal that may be given priority over the dialogical goal of persuasion is that of avoiding unemployment. According to the quietist implication of dialogical empiricism, it is imprudent to engage in analytic philosophical dialogue about philosophically interesting  Cf. Chap. 1, Sect. 1.2.

12

6  Escaping Dialogical Empiricism 

169

metaphysical hypotheses. However, analytic metaphysicians, and others, are earning a living by doing just that, and for them it may be occupationally imprudent to not engage in analytic philosophical dialogue about philosophically interesting metaphysical hypotheses. (Such disengagement may be all the more imprudent if, to paraphrase Kornblith (2014: 207), analytic philosophers’ dialogue about non-metaphysical hypotheses (or philosophically uninteresting metaphysical hypotheses) were indeed to amount to little more than social (or natural) science done badly.) Giving priority to the goal of avoiding philosophical unemployment over the dialogical goal of persuasion would thus equally allow metaphysically inclined philosophers to disregard the quietist prudential recommendations arising from dialogical empiricism. Prioritising such personal goals (like avoiding challenges or unemployment) over the substantive professional goal of persuading other philosophers may be considered intellectually questionable or odious. But as explained, such an assignment of priorities cannot be foreclosed by my above considerations surrounding dialogical empiricism. Of course, philosophers rarely face such a stark choice between, on the one hand, pursuing their dialogical goal of persuasion and, on the other hand, pursuing their personal goal of avoiding challenges or unemployment. Very often, a philosopher finds herself in a position where she can work towards persuading the opposing party and thereby also work towards remaining in academic philosophical employment. Perhaps these two goals (and other combinations of a dialogical and a personal goal) are jointly attainable after all, such that questions of their relative priority would never even arise? This would be very convenient indeed, as one could then pursue the goal of persuading one’s opponents without jeopardising one’s job. It is not actually the case, however, that there are no such priority conflicts. This can be illustrated in two ways. If persuading one’s opposing party (which, recall, may be in existence for several centuries) practically requires that one does not engage in dialogue about philosophically interesting metaphysical hypotheses, but remaining in philosophical employment practically requires that one does engage in such dialogue, then one will at some point have to compromise on one of these goals in order to preserve the possibility of attaining the other. On the one hand, one

170 

C. Rhode

might choose to remain in philosophical employment and restrict one’s dialogical goal to merely persuading those among one’s contemporary opponents who accept without question, say, some particular norm. In this way, one would accept that one’s current metaphysical work will almost certainly be undermined through challenges raised by other contemporary or future members of one’s opposing party.13 That is to say, one might choose to remain in philosophical employment at the price of failing to actually and sustainably persuade one’s opposing party. On the other hand, one might choose to preserve the possibility of persuading one’s opposing party and accept that one could no longer find philosophical employment in analytic metaphysics. In essence, when push comes to shove, one goal has to give in such nevertheless existent priority conflicts. Another way to illustrate that priority conflicts nevertheless exist is by the reductio ad absurdum of supposing their non-existence. Let us assume that, in practice, one can attain both the dialogical goal of persuading one’s opponents with regard to a given philosophically interesting metaphysical hypothesis, and attain the personal goal of retaining one’s philosophical employment. From this it would follow that one could, in practice, meet the requirements for attaining both of these goals. As discussed above, however, these requirements would mean that one has to simultaneously engage and not engage in philosophical dialogue about the given philosophically interesting metaphysical hypothesis. Yet it is unlikely, to say the least, that debating and not debating some given philosophically interesting metaphysical hypothesis would be persuasive in any way. But if such contradictory (non-)engagement in philosophical dialogue would not be persuasive, then the joint attainment of the dialogical goal of persuasion and the personal goal of continued employment entails the non-attainment of the dialogical goal of persuasion, by requiring a kind of conduct in philosophical dialogue that undermines the attainment of the goal of persuasion. Thus, joint attainment of these dialogical and personal goals is practically impossible in the above context. Hence the idea of jointly attaining these goals does not offer an escape from the quietism counselled above, as that idea itself is a  Cf. Chap. 2, Sect. 2.2.

13

6  Escaping Dialogical Empiricism 

171

non-starter. But as explained, prioritising such a personal goal over the dialogical goal of persuasion does offer an escape from metaphysical quietism—even though this assignment of priorities may be intellectually highly questionable. A further personal goal that might be prioritised over the dialogical goal of persuasion is the goal of affecting the outcome of the constant cultural negotiation of the ‘common sense’ by which people conduct their lives. This goal may be particularly salient with regard to the moral and political norms commonly accepted by people in a given society. The ultimate goal here would not be to persuade an opponent, but rather to render certain conduct societally untenable—and thereby to force people to not engage in that conduct. Given this goal, a philosophical argument for the conclusion that abortion ‘really is wrong’ would actually be advanced in order to influence the cultural negotiation regarding the permissibility of abortion to the effect of preventing people from having abortions. Similarly, the goal pursued by arguing that hate speech ought to be made a crime may be to influence the cultural negotiation of the permissibility of hate speech to the effect that acts of hate speech are actually punished by the state. Or, to take an example from the philosophy of science, arguments for the conclusion that creationism is not a science might be advanced in order to render the teaching of creationism in schools societally untenable, and thereby to halt such teaching. What appears to be a given philosophical persuasion dialogue would thus be embedded within a society-wide negotiation dialogue, with the ultimate purpose of enacting one’s personal attitudes by way of a negotiated consensus.14 The goal of persuading one’s philosophical opponents would here be of at most secondary relevance, and the quietism arising from this goal would thus be equally irrelevant. Of course this escape would only be of any use to philosophers who wish to debate metaphysically significant hypotheses that are the subject of cultural negotiation—and not, for instance, to mathematical realists. To the extent that the former philosophers do indeed seek to influence cultural negotiation through their philosophical work, however, they  The goal of such negotiation dialogue thus is not a philosophical consensus but a wider, cultural consensus (that is then enacted). 14

172 

C. Rhode

would appear to have made a rather imprudent choice of occupation and audience. The impact on the cultural negotiation of values, norms and other concerns that is available to those working in politics, journalism, NGOs, social scientific research, or the judiciary rather clearly exceeds any impact that can realistically be expected to be available to analytic philosophers. While the former groups address their work to (other) societal decision-makers or the wider public, analytic philosophers address their work most of all to themselves, and they consequently have a comparatively negligible audience among societal decision-makers or the wider public (cf. Femia 2012). It would thus have been independently prudent for philosophers who seek to influence the cultural negotiation of norms, values, etc. to pursue a higher-impact occupation, rather than preaching principally to their philosophical colleagues. So it would be quite uncharitable towards philosophers to take them to have made such a remarkably futile choice of occupation in pursuit of the implementation of their personal attitudes. Charitably interpreted, enforcing their personal attitudes by way of a negotiated social consensus would thus not be the goal pursued in such philosophical work. In consequence, an impotent role in the cultural negotiation of values, norms and other concerns does not provide an escape from the foregoing quietism. There is, however, one last and very easy way to escape from dialogical empiricism and its resultant quietism—viz. by ignoring this book. The present author is very much unknown within the discipline, which is unlikely to change in the foreseeable future, as this author has left academia. So it is relatively unlikely, ceteris paribus, that this book would have any particularly significant impact. At the time of writing, the quietism counselled here is furthermore far from being practiced, as debates about all sorts of metaphysically significant hypotheses of philosophical interest are thriving across large parts of analytic philosophy. Conserving this status quo might thus be accomplished most conveniently by paying no heed whatsoever to the pragmatic imprudence of engaging in dialogues about philosophically interesting metaphysical hypotheses—and keeping all fingers crossed that one’s dialogical opponents will do so too, lest one were to easily lose in the persuasion game of analytic philosophical dialogue. If such be your choice, then good luck.

6  Escaping Dialogical Empiricism 

173

References Buckwalter, W., and S.  Stich. 2014. Gender and Philosophical Intuition. In Experimental Philosophy, Vol. 2, eds. J.  Knobe and S.  Nichols, 307–346. Oxford: Oxford University Press. Clarke-Doane, J. 2012. Morality and Mathematics: The Evolutionary Challenge. Ethics 122 (2): 313–340. Cohen, L.J. 1986. The Dialogue of Reason: An Analysis of Analytical Philosophy. Oxford: Clarendon Press. Cummins, R. 1998. Reflections on Reflective Equilibrium. In Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry, eds. M.R.  DePaul and W.  Ramsey, 113–127. Lanham, MD: Rowman and Littlefield. Elder, C. 2004. Real Natures and Familiar Objects. Cambridge, MA: MIT Press. Femia, J.V. 2012. Pareto, Machiavelli, and the Critique of Ideal Political Theory. In Vilfredo Pareto: Beyond Disciplinary Boundaries, eds. J.V.  Femia and A.J. Marshall, 73–83. Farnham: Ashgate. Kornblith, H. 2007. Naturalism and Intuitions. Grazer Philosophische Studien 74: 27–49. ———. 2014. Is There Room for Armchair Theorizing in Epistemology? In Philosophical Methodology: The Armchair or the Laboratory? ed. M.C. Haug, 195–216. New York: Routledge. Levy, N. 2003. Analytic and Continental Philosophy: Explaining the Differences. Metaphilosophy 34 (3): 284–304. Ross, D., J. Ladyman, and D. Spurrett. 2007. In Defence of Scientism. In Every Thing Must Go: Metaphysics Naturalized, eds. J.  Ladyman, D.  Ross, D. Spurrett, and J. Collier, 1–65. Oxford: Oxford University Press. Scanlon, T.M. 2003. Rawls on Justification. In The Cambridge Companion to Rawls, ed. S. Freeman, 139–167. Cambridge: Cambridge University Press. Tolkien, J.R.R. 1937. The Hobbit, or There and Back Again (5th ed.). Reprint 2011. London: HarperCollins. Walton, D. 1992. After Analytic Philosophy, What’s Next?: An Analytic Philosopher’s Perspective. Journal of Speculative Philosophy 6 (2): 123–142.

Appendix A Reputational Journal Ranking

Table A.1  Combined Reputational Journal Ranking, based on the reputational rankings by Brooks (2011) and Leiter (2013), includes all journals ranked among the top-25 by at least one ranking, and here ranked by the journals’ respective lowest rank in either ranking. Ranks below rank 25 in parenthesis. Journal Journal of Philosophy Noûs Mind Philosophy & Phenomenological Research Ethics Australasian Journal of Philosophy Analysis Philosophers’ Imprint Philosophy of Science Synthese Proceedings of the Aristotelian Society

Brooks rank

Highest Leiter rank rank

Lowest rank

1 2 3 4

2 3 4 5

1 2 3 4

2 3 4 5

5 7

6 8

5 7

6 8

6 11 8 12 10

10 9 13 15 16

6 9 8 12 10

10 11 13 15 16 (continued)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1

175

176 

Appendix A Reputational Journal Ranking

(continued) Journal British Journal for the Philosophy of Science Philosophical Studies Erkenntnis Philosophical Review American Philosophical Quarterly Philosophy & Public Affairs Canadian Journal of Philosophy Pacific Philosophical Quarterly Philosophical Quarterly European Journal of Philosophy Journal of the History of Philosophy Monist Mind and Language Journal of Philosophical Logic British Journal for the History of Philosophy Midwest Studies of Philosophy Southern Journal of Philosophy

Brooks rank

Highest Leiter rank rank

Lowest rank

16

14

14

16

17 15 18 9

7 17 1 18

7 15 1 9

17 17 18 18

21 22 14 24 20 25

12 19 23 11 24 20

12 19 14 11 20 20

21 22 23 24 24 25

13 (26) (29) (30)

(26) 22 21 25

13 22 21 25

(26) (26) (29) (30)

19 23

(32) (not ranked)

19 23

(32) (n/a)

References Brooks, T. 2011. In your judgment, which of the following is the better philosophy journal? In All Our Ideas. http://www.allourideas.org/philosophyjournals/results?all=true. Accessed 17 June 2015. Leiter, B. 2013. Poll Results: Which are the ‘best’ (highest quality) philosophy journals (regardless of area)? In Condorcet Internet Voting http://www.cs.cornell.edu/w8/~andru/cgi-­perl/civs/results. Service. pl?id=E_46442506716654dd. Accessed 17 June 2015.



Appendix B Population Size

Table B.1  Number of Original Articles per Journal, as published in the volumes of 2007, 2010 and 2013  in the journals listed in Appendix A.  Discussion notes, reviews, editorials, errata, and articles in special issues and symposia excluded. Journal

2007

2010

2013

Average

Journal of Philosophy Noûs Mind Philosophy & Phenomenological Research Ethics Australasian Journal of Philosophy Analysis Philosophers’ Imprint Philosophy of Science Synthese Proceedings of the Aristotelian Society British Journal for the Philosophy of Science Philosophical Studies Erkenntnis Philosophical Review American Philosophical Quarterly

25 31 21 36 13 34 63 9 23 59 14 29 52 21 17 29

26 30 19 50 14 37 71 12 26 55 15 31 86 35 14 28

25 36 22 40 9 40 55 24 27 146 16 35 78a 47a 12 30

25.33 32.33 20.67 42.00 12.00 37.00 63.00 15.00 25.33 86.67 15.00 31.67 72.00 34.33 14.33 29.00 (continued)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Rhode, The Burden of Proof upon Metaphysical Methods, https://doi.org/10.1007/978-3-031-27777-1

177

178 

Appendix B Population Size

(continued) Journal

2007

2010

2013

Average

Philosophy & Public Affairs Canadian Journal of Philosophy Pacific Philosophical Quarterly Philosophical Quarterly European Journal of Philosophy Journal of the History of Philosophy Subtotal, top-25 in both rankings Monist Mind and Language Journal of Philosophical Logic British Journal for the History of Philosophy Midwest Studies of Philosophy Southern Journal of Philosophy Subtotal, top-25 in only one ranking Total

15 23 28 28 13 26 609 – 25 27 27 17 27 123 732

12 24 27 36 22 18 688 – 18 21 34 – 18 91 779

12 25 28 35 27 22 791 – 14 27 47 – 28 116 907

13.00 24.00 27.67 33.00 20.67 22.00 696.00 0.00 19.00 25.00 36.00 5.67 24.33 110.00 806.00

 Outlier article counts for a given year were substituted by the count for the following year iff the count for the following year was not also an outlier. Outlier article counts were defined as either: >2.5 times a journal’s smallest count and at a greater distance from the mid-size count than the distance between the mid-size and the smallest count; or: