On certainty and other philosophical essays on cognition 9783110319774, 9783110319293

On Certainty continues Rescher’s longstanding practice of publishing occasional studies that form part of a wider progra

247 45 2MB

English Pages 94 [107] Year 2011

Report DMCA / Copyright


Polecaj historie

On certainty and other philosophical essays on cognition
 9783110319774, 9783110319293

Table of contents :
Chapter Five APORY
Name Index

Citation preview

Nicholas Rescher On Certainty And other Philosophical Essays on Cognition

Nicholas Rescher

On Certainty And other Philosophical Essays on Cognition

Bibliographic information published by Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic data is available in the Internet at http://dnb.ddb.de

North and South America by Transaction Books Rutgers University Piscataway, NJ 08854-8042 [email protected] United Kingdom, Eire, Iceland, Turkey, Malta, Portugal by Gazelle Books Services Limited White Cross Mills Hightown LANCASTER, LA1 4XS [email protected]

Livraison pour la France et la Belgique: Librairie Philosophique J.Vrin 6, place de la Sorbonne; F-75005 PARIS Tel. +33 (0)1 43 54 03 47; Fax +33 (0)1 43 54 48 18 www.vrin.fr

2011 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm www.ontosverlag.com ISBN 978-3-86838-105-4 2011 No part of this book may be reproduced, stored in retrieval systems or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use of the purchaser of the work Printed on acid-free paper FSC-certified (Forest Stewardship Council) This hardcover binding meets the International Library standard Printed in Germany by CPI buch bücher dd ag

Contents ON CERTAINTY And other Philosophical Essays on Cognition PREFACE

Chapter 1: On Certainty


Chapter 2: Practical Inference in Theoretical Contexts


Chapter 3: On the Complications of Epistemic Quality


Chapter 4: Persuasion


Chapter 5: Apory


Chapter 6: Cognitive Obligation


Chapter 7: Semantical Approximation


Chapter 8: Cognitive Reflexivity and the Limits of Knowledge


Chapter 9: Issues of Ultimate Explanation




Name Index:




he present book continues my longstanding practice of publishing occasional studies that form part of a wider program of investigation of the scope and limits of rational inquiry in the pursuit of understanding. And pragmatism forms a subtextual Leitmotiv of these essays, seeing that the linking idea at work throughout is that knowledge is a tool for the management of our theoretical and practical affairs, and that what we ask of it is serviceability for the uses we have in view. I am grateful to Estelle Burris for helping me to put this material into a form suitable for publication.

Nicholas Rescher Pittsburgh PA November 2010



eople—even philosophers—are often confused about certainty. All too often they overlook the vast gulf of difference that obtains between saying “I am certain” and “It is certain.” “It is certain” is a matter of an objective certainty. By contrast, “I think it is certain” is on only a matter of subjective certitude. When someone says “I am certain that p” the power response is “You sound like an interesting person, tell me more about yourself.” For “I am certain that p” is, in the end (and indeed the beginning as well) merely a claim about me, and as such it says very little—if anything—about p. Granted, when I am certain that p (that the cat is in the mat) it is likely, indeed inevitable, that I think that it is certain. But of course thinking something to be so does not make it so. There is even a significant gulf between “I feel certain” and “I am certain.” Saying that “I feel certain that Smith is honest” indicates a degree of hesitancy that is absent from “I am certain that Smith is honest.” But of course both alike remain at the level of subjectivity. Under what circumstances is the inference from “I am certain” to “It is certain” actually justified? Given that “I think that p” (e.g. that it is certain that the cat is on the mat) there undeniably follows the conclusion “Some people think that p.” But this does not go very far to showing that p is true (let alone certain). To move deductively from “X is certain that p” to “It is certain that p” requires a supplementary premiss on the order of “Whatever X thinks to be so is indeed so.” This clearly is not a very plausible prospect. For the fact of it is that there just is no plausible premiss to effect that transition in a valid, deductively conclusive way. With respect to certainty, one must also distinguish between an ontological and an epistemic mode, that is between • It is certain that p

On Certainty and Other Philosophical Essays on Cognition

and • X is certain that p. Ontology is a matter of how things stand; epistemology a matter of the resource and evidential indications that are available to people with regard to how things stand. Many is the fact for which people have no evidence whatsoever. Our cognition hinges on the available evidential indications that entail people to think something to be so. And it is perfectly possible for all of the available evidence to point in an incorrect direction. So people are, regrettably, often entitled—rationally and evidentially to regard as certain something that actually fails to be so. For it can transpire that not only are all available evidential indications such as to support the conclusion that p is certainly the case. In particular this happens when not only is there a substantial body of evidence in p’s favor, but further there is, throughout a vast range of available fact, no case-specific indication whatever that p might not hold. And when someone (X) is in this position that it can happen that • X is entitled to think (believe, accept) that p is certain. The reality of it is that the only resource that can carry us from “X is certain that p” to “It is certain that p” is not deductive logic but generosity. For what we can do here is to extend X credit (in the epistemic rather than monetary sense). It is by seeing X as a cognitive paragon—one would not possibly accept something that is not so—that we take on the risk of this inferential transit. Now in our own case we are inclined to be generous. With us ourselves the difference between “I am certain” and “It is certain” seems vanishingly small. But there is some hybris if not worse (i.e. megalomania) at work here, since in actuality there is always a gap between the subjectivity of one’s being certain and the objectivity of something indeed being so. But then are we ever rationally entitled to lay a claim to objective certainty? Sceptics since classical antiquity have always denied it. However, that is going too far. One is rationally entitled to claim as



true those theses for which one has a sizable body of substantiating evidence and proceeds in the absence of any counter-indications. For by and large all will go well here. A claim to certainty can prove to be reasonable even in its absence. For here as elsewhere there is something of a gap because even a rationally warranted claim can fail to hold good. After all, there is little if any probative gap between the claim “p is true” and the claim “It is certain that p.” Whatever is actually true is certainly so, and conversely. But of course this only holds in the objective order of things. Subjectively a substantial difference remains. It makes perfectly good sense to say “I think that p is true but am not certain of it.” However, in the mode of objectivity matters stand differently. It makes no sense to say “I know that p is true but am not certain of it.” I am certain that every claim made is the preceding discussion is true. But is it really so? No further claim that I myself can possibly make will establish this. It is something that readers will have to judge for themselves.




onsider the following line of reasoning:

• X acts in a way that seems φ . • There is no good reason for thinking that in the present case things are not as they seem. ∴ [It is appropriate to think that in this present case] X acts in a way that actually is φ. What this mode of reasoning obviously does is to effect an inferential transit for what merely seems to what actually is in situations where there is no apparent reason for thinking otherwise. To all appearances, the reasoning proceeds on the basis of the principle: To accept the merely seeming as actually authentic whenever due diligence indicates that there are no case-specific counterindications to doing so.

One might designate this as the Principle of Rational Supposition (PRS). The object of the present discussion is to inquire into the grounds and the consequences of this principle. Note that the rational basis for this principle is not provided by a tacitly presupposed generalization about the actual facts: Absent counterindications, whatever seems to be so always is.

On Certainty and Other Philosophical Essays on Cognition

This, of course, is very far from true. To be sure, it must be acknowledged that the principle could scarcely be maintained were it not the case that: Very often—and indeed generally—what seems to be is.

But this circumstance does not—and cannot—of itself validate the principle. It may be a necessary condition for its tenability but it is hardly sufficient. So what is it that speaks for the principle? Note, to begin with, that the principle is in a way, self-instantiating and self-substantiating. It can be argued for as follows: • Reality so functions as to make it appear that the principle holds. • There is no good reason to think that in this context things are not as they seem. ∴ It is appropriate to think that that principle actually obtains. This self-sustainingness is something of a positive feature, but it is hardly decisive. The salient consideration is that the principle at issue is not actually a principle of logic—or indeed of theoretical reason at large, but is, rather, a matter of practical reason. And as such, its rationale is not a matter of demonstratable fact, but one of practical policy. And this policy itself is justified by a process that effectively serves our interests in practical matters—one that proceeds via two fundamentally practical considerations: 1. That it affords good promise for the realization of our objectives and 2. That the eventual experience shows it to be effective in actual practice



That is to say there is both the initial justification of procedure via praise, and the ex-post-facto retrojustification of working out to good effect in actual application. And this here as elsewhere, underwrites the rational appropriateness of a practical policy. What is thus at issue is a principle of rational economy—a matter of cost-effectiveness in relation to the realization of our ends. Applicative efficacy is the contributing factor here, and the validating rational is of a pragmatic rather than a theoretical cast. On this basis the Principle of Rational Supposition comes to the pragmatic injunction: If all the available indications—all the indications I could reasonably expect to have—speak for something, then, barring the other obstacles, it is rationally appropriate to accept it. It rests on a commitment to the idea that the best we can possibly do in the circumstance should be seen as sufficient for practical purposes—that here as elsewhere obligation does not outrun the limits of the possible. (Ultra posse nemo obligatur as the Roman legal dictum had it.) So much for validation, let us now turn to the potential uses of the Principle of Rational Supposition. It is readily seen that this practical principle has an impressive range of promising implementations. In specific it can serve to close the epistemic gap that lies • between appearance and reality, between subjective impressions and objective facts. • between limited observed generality and unlimited objective universality. • between apparent universality and lawful necessity. • between the subjectivity of the “moral sense” and objectivity of moral judgment. • between preference and preferability (“Mill’s paradox”), and between fact and value more generally.


On Certainty and Other Philosophical Essays on Cognition

All too often philosophers argue that such conceptual gaps are impassible—that there is no theoretically validatable way to get from here to there. But even were this so, this would not mean that there is no rational transition at all. For practical rather than purely theoretical considerations can be brought into play, and practical reason can enter where theoretical reason fears to tread. But are we actually entitled to rely on conclusions whose justificatory bases lies in the realm of the practical rather in that of strictly theoretical/evidential considerations. The answer is that we not only are justified, but are actually so situated in the world’s scheme of things that we cannot but be. For we are creatures not only of thought but also of action. And while our actions are—and should be—guided by our thoughts, those thoughts are—or should be—so constituted as to serve to guide us adequately in matters of their implementation in action. We humans are creatures that have not only beliefs but needs—a need for theory-based knowledge included. And the two have to be coordinated into one unified manifold in which the demands that each component makes upon the others are duly accommodated.




ur present concern is with the merit of knowledge that is known rather than with that of the knower. It is the cognitive merit of the claims rather than that of the claimants that will be at issue here. Clearly a prime and essential merit for claims to knowledge is their truth. One cannot know that which is false: to say “X knows that p but it isn’t so” is to commit a contradiction in terms. (What one could and should say is “X thinks he knows that p, but it isn’t so.”) But truth as such is not the end of the matter. Our cognitive commitments can have different degrees of merit and can be superior or inferior to one another in a considerable variety of ways. Here the following items stand at the forefront: • certainty • probability (likelihood) • evidentiation (substantiation) • informativeness/novelty • precision (exactness of detail) • systemic fit (coherence) • applicability/utility

On Certainty and Other Philosophical Essays on Cognition

In accessing the comparative merit of various available answers to a question all on these cognitive virtues can come to the forefront. And they are related to one another in a complex variety of ways. Consider a controversial yes-or-no question that is being investigated in a setting of ever-expanding information. At stage 1, the answer seems to be Yes. But with the advance to stage 2, new counterinformation comes to light to indicate that the previous appraisal was premature and that the answer is in fact No. But stage 3 of the investigation once more reverses matters. So as we continue to realize an ever-enlarging base of information, our view of the matter moves through ongoingly variant stages, so that, we arrive at the following history of issue resolution: YNYYNN... In this sort of situation, things just need not settle down to an evercontinuing series of Ys and Ns. Ongoingly enhanced information can in theory bring ongoing destabilization in its wake. This is, to be sure, a merely theoretical prospect. But its very existence as such suffices to indicate that we cannot identify evidentiation with actual likelihood. A contention’s substantiation doubtless indicates how likely this contention seems. But how likely it actually is cannot safely be inferred on this basis. Here again objectivity and subjectivity can go separate ways. We would like to think that there is a tight linkage between stronger evidentiation and higher probability. And ideally it might to be so. But things do not always happen in an ideal way. Added evidence can—in theory and sometimes in practice—destabilize a correct conclusion by providing mistaking counter-indications. Those complex cases where the indications at hand sometimes point in one direction and at other times in the other are bound to undermine our confidence in a secure linkage between evidentiation and probability. Then too there is the dauntingly significant consideration that informativeness and precision are at odds with probability. It is a basic principle of epistemology that increased confidence in the correctness of our estimates can always be secured at the price of decreased accuracy. For in general an inverse relationship obtains be-



___________________________________________________ Display 1 COGNITIVE COMPLEMENTARITY locus of feasible combinations precision


___________________________________________________ tween the definiteness or precision of our information and its substantiation: detail and security stand in a competing relationship. We estimate the height of the tree at around 25 feet. We are quite sure that the tree is 25±5 feet high. We are virtually certain that its height is 25±10 feet. But we can be completely and absolutely sure that its height is between 1 inch and 100 yards. Of this we are “completely sure” in the sense that we are “absolutely certain,” “certain beyond the shadow of a doubt,” “as certain as we can be of anything in the world,” “so sure that we would be willing to stake your life on it,” and the like. With any sort of estimate whatsoever there is always a characteristic trade-off relationship between the evidential security of the estimate, on the one hand (as determinable on the basis of its probability or degree of acceptability), and on the other hand its contentual detail (definiteness, exactness, precision, etc.). And so a complementarity relationship of the sort depicted in Display 1. obtains. This was adumbrated in the ideas of the French physicist Pierre Maurice Duhem (1981-1916) and may accordingly be called “Duhem’s Law.”1 In his classic work on the aim and structure of physical theory,2 Duhem wrote as follows: A law of physics possesses a certainty much less immediate and much more difficult to estimate than a law of common sense, but it surpasses the latter by the minute and detailed precision of its predictions. . . The laws of physics can acquire this minuteness of detail only by sacrificing something of the fixed and absolute certainty of common-sense laws.


On Certainty and Other Philosophical Essays on Cognition

There is a sort of teeter-totter of balance between precision and certainty: one cannot be increased except to the detriment of the other.3

In effect, these two factors—security and detail—stand in opposition. Then again consider the linkage between truth-likelihood and coherence—that is fit with already well-secured information. By and large when a proposition fits smoothly among the environing context of what we already accept. And this is often counted in its evidential favor. But this is very problematic. After all, coherence comes down to redundancy and is inherently in conflict with infiniteness. Accordingly, much the same sort of conflict that obtains with regard to the detail/likelihood complementarity recurs in the coherent/innovation case. A crucial consequence of these considerations is the challenging nature of life in a difficult world. There is no guarantee that the epistemically superior thesis will actually be true and that cognitive progress may fail to move us in the direction of the actual truth. 2. DESIDERATUM COMPLEMENTARITY AS A GENERAL

PHENOMENON Cognitive complimentarity instantiates a very general situation which occurs across a wide spectrum of situations, and indeed has substantial ramifications in many areas. For lies in the nature of things that their desirable features are in general competitively interactive. A conflict or competition among desiderata is an unavoidable fact of life, seeing that since positivities cannot all be enhanced at once since more of the one can only be realized at the expense of less of the other. Desideratum complementarity thus arises when two (or more) parameters of merit are linked (be it through a nature-imposed or a conceptually mandated interrelationship) in a see-saw or teeter-totter interconnection where more of the one automatically ensures less of the other, as per the situation of Display 2. Such desideratum complementarity is pretty well inevitable with any complex, multidimensional good whose overall merit hinges on the cooperation of several distinct value-components. In all such cases we have a teeter-totter, see-saw relationship of the general sort here characterized as desideratum



________________________________________________________ Display 2 DESIDERATUM COMPLEMENTARITY VALUES

Positivity 1 ↑

→ Positivity 2

________________________________________________________ complementarity. Beyond a certain point, augmentations of the one are simply incompossible with augmentations of the other (to use Leibniz’s terminology). There is always a trade-off curve that characterizes the decrease in one parameter of value that is the unavoidably exacted price for an increase in the other. Let us consider yet another instance of this phenomena. Give the imperfections of our knowledge we must reckon with the prospect of error. And it must be noted that there are two significantly different sorts of errors, namely errors of commission and errors of omission. For it is only too clear that errors of commission are not the only sort of misfortune there are. Ignorance, lack of information, cognitive disconnection from the world’s course of things—in short, errors of omission—are also negativities of substantial proportionism, and this too is something we must work into our reckoning. Both are negativities and obviously need to be avoided insofar as possible in any sensible inquiry process. With error-avoidance in matters of cognition the tradeoff between errors of type 1 and errors of type 2—between improper negatives and false positives—is critical in this connection. For instance, an inquiry process of any realistically operable sort is going to deem some falsehoods acceptable and some truths not. And the more we fiddle with the arrangement to decrease the one sort of error, the more we manage to increase the other. The situation here is a general one that appertains not just to error but to cognitive misfortunes. (See Display 3.)


On Certainty and Other Philosophical Essays on Cognition __________________________________________________________________________ Display 3 RISK ACCEPTANCE AND MISFORTUNES Misfortune of kind 1 Misfortune of kind 2 Number of (significant) misfortunes






Increasing risk acceptance (in % of situations)


The by now familiar teeter-totter relationship obtains here once more. For unfortunately the reality of it is that any given epistemic program—any sort of process or policy of belief formation—will answer to the situation of complimentarity between the two modes of error. The question accordingly arises: How much gain in one is needed to compensate for how much loss in the other? Are we prepared to run a significantly greater risk of mistakes to secure the potential benefit of a significantly enlarged understanding? In the end, the matter is one of priorities—of safety as against information, of ontological economy as against cognitive advantage, of an epistemological risk aversion as against the impetus to understanding. The issue, in the end, is one of values and priorities, weighing the negativity of ignorance and incomprehension against the risk of mistakes and misinformation. 3. THE UNREALIZABILITY OF PERFECTION Throughout such complementarity cases we have the situation that to all intents and purposes realizing more of one desideratum entail a correlative decrease in the other. We cannot have it both ways, so that ideal of achieving the absolute perfection at issue with a concurrent maximization of every parameter of merit at one and the same time



lies beyond our grasp as a matter of principle. In the interest of viability some sort of compromise must be negotiated seeing that the concurrent maximization of desiderata is now automatically unrealizable. The interactive complexity of value is crucial for desideratum complementarity. For it is the fundamental fact of axiology that every object has a plurality of evaluative features some of which will in some respects stand in conflict. And consequently in any setting of multicriterial complementarity, “absolute” perfection is simply a piein-the-sky impossibility. What we need to do here is look at the exact picked situation in where we find ourselves, and, given its nature, to seek to identify the least acceptable value of our merit-parameters d1 and d2. What we then have on our hands is not a problem of maximization as such, but rather a problem of optimization, i.e. of maximization subject to constraints. And of course one must expect the same sort of situation to prevail further along the line. When (a value) V fissions apart into components V1 and V2, then it must be expected that these themselves will in their turn fission in the same way, respectively splitting into V11/V12 and V21/V22 in a way that once again calls for contextual optimalism rather than absolute maximization. Thus consider an automobile. Here the parameters of merit clearly includes such factors as speed, reliability, repair infrequency, safety, operating economy, aesthetic appearance, road-handle ability. But in actual practice such features are interrelated and it is unavoidable that they trade off against one another: more of A means less of B. It would be ridiculous to have a supersafe car with a maximum speed of two miles per hour. It would be ridiculous to have a car that is inexpensive to operate but spends three-fourths of the time in a repair shop. Invariably, perfection—an all-at-once maximization of every value dimension—is inherently unrealizable because of the inherent interaction of evaluative parameters. In designing a car you cannot maximize both safety and economy of operation (which demands lightness of weight). And analogously, the world is not, and cannot possibly be, absolutely perfect—perfect in every respect—because this sort of absolute perfection is in principle impossible of realization. For absolute perfection consists in the concurrent realization of every relevant parameter of merit. But whenever


On Certainty and Other Philosophical Essays on Cognition

different sorts of merit stand in a teeter-totter complementarity relationship it is clearly inevitable that they cannot both achieve a maximal degree at one and the same time. Desideratum complementarity is a clear indication that the idea of absolute perfect is altogether inapplicable and inappropriate in many evaluative situations. The holistic and systemic optimality of a complexly articulated whole will require some of its constituent value components to fall short of what would be ideally desired for them if abstractly considered in detached isolation. Sometimes, however, it just does not matter all that much. Once we establish that a source is unreliable in various cases, it does not much matter whether this manifests itself in 25 or in 50 percent of the cases—we would still stop relying on this source in any case. And of course, it makes no sense to ask for the impossible in these matters. It is an inherently inevitable features of the nature of things— a logico-conceptually inevitable “fact of life”—that with the complexity of a world at issue, value realization will always be a matter of balance, of trade-offs, of compromise because value factors always compete in matters of realization. Concurrent maximization in every relevant positivity is simply unavailable in this or indeed any other realistically conceivable world. All that one can ever reasonably ask for is an auspicious combination of values—an overall optimal profile. And what is acceptable on this basis is going to pivot in the pragmatic factor of the inherent nature of the issues that are at stake. And here the issue of the operational context becomes important. The quality that we demand for knowledge will generally depend on the context of its use. We do not need to know how much it will rain—do not need precision in the amount of rainfall—when the issue is that of determining whether or not to take an umbrella. Even in matters of cognition, the practicalities of the situation will have to play a crucial role. NOTES 1


It is alike common and convenient in matters of learning and science to treat ideas and principles eponymously. An eponym, however, is a person for whom something is named, and not necessarily after whom this is done, seeing that eponyms can certainly be honorific as well as genetic. Here at any rate eponyms are sometimes used to make the point that the work of the person at issue has suggested rather than originated the idea or principle at issue.



La théorie physique: son objet, et sa structure (Paris: Chevalier and Rivière, 1906); tr. by Philip P. Wiener, The Aim and Structure of Physical Theory (Princeton, Princeton University Press, 1954.) This principle did not elude Neils Bohr himself, the father of complementarity theory in physics: “In later years Bohr emphasized the importance of complementarity for matters far removed from physics. There is a story that Bohr was once asked in German what is the quality that is complementary to truth (Wahrheit). After some thought he answered clarity (Klarheit).” Stephen Weinberg, Dreams of a Final Theory (New York: Pantheon Books, 1992), p. 74 footnote 10.


Duhem, op. cit., pp. 178-79. Italics supplied.




ersuasion as a topic has long been the province of rhetoricians and of sociologists. But philosophers have not been all that much concerned to engage the matter. And yet the fact is that persuasion is replete with philosophical aspects and involvements, cognitive and ethical issues preeminent among them. Man, homo sapiens, is a rational animal. We act, at least sometimes, neither by instinct nor by Pavlovian conditioning, but for reasons. And then beliefs and desires conspire to shape our actions. We are thirsty, desire drink, and think our refrigerator is the most convenient place to get it—and so we get up and go into the kitchen. Intelligent beings that we are, our actions implement our thoughts. We are also social animals. We interact with others, often by communication, by the transmission of information. And personal persuasion is the process of using communication to induce someone to think or act differently from the way in which they otherwise would or well might. The aim of the enterprise is to induce people to think or act in a certain way. Persuasion can be either personal or impersonal. One can be persuaded of something by someone, or one can be persuaded of something via impersonal considerations. (There can also be mixed cases, as per “He drew my attention to certain facts which, in providing an alibi, persuaded me of his innocence.) Persuasion can be either theoretical or practical. There is cognitive persuasion and performativey persuasion: persuasion to accept something and persuasion to do something—to believe or to act. Either way, persuasion is a matter of drawing someone’s attention to otherwise overlooked considerations that should influence their deliberations. It thus requires two parties, a persuader who seeks to influence

On Certainty and Other Philosophical Essays on Cognition

someone’s proceedings, and a persuadee whom that persuader is trying to influence. Apart from threats—which are also often miscalled “persuasion”— there are basically two avenues of procedure open for rational persuasion. One can provide the persuadee with new information not previously accessible to him, or one can provide a new perspective, a variant formwork of interpretation for the information that the persuadee already has which gets him to view matters “in a different light.” Much of the time teaching is persuasion. For the teacher is, in effect, engaging in an endeavor to persuade the student to accept certain facts, to consider (and possibly even to adopt) certain points of view, and to do certain things in a certain sort of way. The extent to which people are amenable to persuasion can be measured across a spectrum ranging from gullible to bull-headed. The issue of a just right intermediation is something that will be critically dependent on the exact nature of the situation at issue. Persuasion must be somehow motivated. It always makes an appeal to some aspect of the persuadee’s make-up: his cupidity, his avarice, his public spirit, his affections. Somehow the persuader must be able to represent a failure to be persuaded as somehow injurious to the presuadee’s interest. Even in purely cognitive matters some features of the presuadee’s make-up will be in play—if only his dedication to the truth. For personal persuasion is generally advisory. The persuader in effect tells the persuadee: “There is good reason why you should think/act in a certain particular way—one in which I think you might well not choose were it not for the things I am now telling you.” Is it possible for the persuader and the persuadee to be one and the same person? Is there such a thing as self-persuasion? Ordinary discourse suggests an affirmative answer. “He persuaded himself that . . .” is a perfectly meaningful contention that qualifies as appropriate in certain circumstances. The individual who simply dismisses or wittingly discounts certain counterindications can plausibly be said to have persuaded himself of something. Persuasion is thus somewhere between pointless and unavailing when it is perfectly clear to all concerned exactly what the persuadee is going to think or do. Persuasion, like any project, can thus have three outcomes: successes, failure, and pointlessness.



People are more easily persuaded of certain things than others. And some are gullible and readily persuaded, others are stubborn and can only be persuaded with great difficulty. People vary in the extent to their attachments to beliefs, to ideas, to persons, and to policies. Persuasion also functions at levels different from that of the individual. Nations try to persuade one another to adopt favorable policies. Industries and companies try to persuade legislators to establish favorable laws. Advertisers try to persuade the public to purchase various products and services. But in the final analysis, it is the individuals who constitute those groups that really matter. A group can be persuaded—and can only be persuaded—by persuading (many of) the individuals without it. By and large, the Greek philosophers viewed mathematics as the primordial mode of knowledge and with the quest for authentic truth oriented in this mathematical direction, the processes of mere persuasion—prominently on display in law courts and in the political arena—was deemed an inferior enterprise. Accordingly, the Sophists did not get a good press. The middle ages saw revelation and its interpretation by the church as the foundation of authentic knowledge. Only with greater confidence in the common man as it emerged in the democratic aftermath of the Age of Reason did there emerge the idea that people are basically rational, and that only the real truth can carry general conviction. The idea of persuasion as an instrumentality of the quest for truth thus gained acceptance only in a significantly secular and democratic era. The high-water mark of the metaphysical use of persuasiveness was set early on in Plato’s dialogue Timaeus. Plato here viewed the physical world as a vast intelligent organism. And he saw its comportment as subject to a cosmic force of Reason. The characteristic works of this rationality impelling force is to persuade, and it endeavors, with ever-increasing success, to persuade the physical universe to comport itself with ongoingly increasing rationality. And so over time the universe has become orderly, regular, and lawful—its history seeing development of a cosmos from a chaos.


On Certainty and Other Philosophical Essays on Cognition

2. ETHICAL ASPECTS Persuasion has a prominent ethical dimension. The persuader and the persuadee alike are entitled to ask: “What’s in it for me?” and the answer can run the entire gamut ranging from the warm glow created by pure benevolence when “virtue is its own reward” to blatant selfinterest in its most crass and unworthy guise. The question of whose interest is at stake is thus always relevant with persuasion. Here the prime possibilities are as follows. (1) The interest of the persuadee himself (2) The interest of the persuader (3) The interest of certain others or of people-in-general. Case (1) can be categorized as beneficial persuasion—that is, advice. Case (2) would be either a plea or a demand, all depending. Case (3) is complex and its nature hinges on the circumstances. It is not easy to delineate with precision the difference between these cases. All may involve performing the same action—giving alms, for example. All alike involve negative consequences for failure to conform. All leave the response to the agent’s decision (in contrast, say, to a seizure of asserts). However, the salient feature of compulsion is that here the persuader is part of the causal claim that activates the negative consequences of noncompliance. Persuasion is by its very nature a matter of intervention in the affairs of another. And this circumstance should put the label PROCEED WITH CARE on the whole business. The process can qualify as ethically appropriate only when the persuasion is rational (rather than manipulative), benignly well-intended, and geared autonomously to the best interests of the persuadee or to other deserving beneficiaries. And even here there are potential problems. Both the persuader and the persuadee face substantial ethical decisions. The persuader faces the choice of whether or not to take this project on; the persuadee face the choice of whether or not to yield to the persuasion. It is clearly the persuader’s duty to ascertain that suffi-



cient appropriate interests are at stake to justify interfering with another person. And it is clearly the persuadee’s duty to ensure that there are valid ground for yielding to the persuasion—or for resisting it. For different sorts of claims and different sorts of acts are—and should be—subject to different standards of persuasiveness. Jane Austen’s classic novel Persuasion affords a telling illustration here. The friends and relatives of her protagonist unite in a thoroughly well-meaning and ultimately successful effort to dissuade her from a seemingly injudicious marriage with an apparently unqualified suitor. Happily, after that rejected swain’s circumstances change and he becomes a real “catch”, he swallows his disappointment and remains true to the heroine’s affections. And so, notwithstanding various embarrassments on the heroine’s: part, they are happily united at the end. But, well-intentions persuasion has created a great deal of trouble all round and could well have issued in misfortune. There is an object lesson here for us all. Persuasion is a purposive activity. It is always and only undertaken with a view to producing a certain result—a change of mind or of will on the persuadee’s part. And it is a procedure that is only undertaken with rational appropriateness when the persuader is humanly conical that it would be better with a view to certain interests for the persuadee to proceed conformably. With rational persuasion someone’s interests are always at stake. The persuader is in effect telling the persuadee: “If you don’t proceed (believe, do) as I am proposing, certain negative consequences will ensue [for you, for us, for everyone]”. Persuasion always looks to consequences. And the negative consequences of noncompliance cover the entire range from merely being misinformed to bringing about a major catastrophe. Already in the earliest days of philosophy, persuasion suffered a bad name. For then the art of persuasion was principally cultivated and taught by the so-called Sophists, and Plato had little good to say about then. Seeing them as unscrupulous practitioners of belief manipulation, he regarded them as malign deceivers who cared only for persuasive success and nothing for the truth. Then too there is the matter of malign persuasion of a falsehood Shakespeare’s Othello is the classic illustration here. Interestingly, there is also the prospect of an ill-intended persuasion of a truth. The


On Certainty and Other Philosophical Essays on Cognition

“good friends” who persuade the heroine of Clare Booth Luce’s play The Women to accept the (true) fact of her husband’s unfaithfulness illustrate this situation. In the final analysis, what renders a venture in cognitive persuasion wicked is the (malign) intention of the persuader and not just the truth or falsity of the claim at issue. In any community of rational beings persuasion is a useful instrument which, like instruments in general, can be put both to good and to bad uses.




n apory is a knotted tangle in the fabric of belief that cries out for unraveling. It results from the not uncommon situation when the things that we are minded to say are inconsistent with each other, with the result that they cannot all be maintained together with rational cogency. In such circumstances the interests of consistency and coherence demand that our cognitive commitments be revised and refined. The term apory (pl. apories; Greek aporia) comes from a Greek verb meaning to be baffled, at a loss, in perplexity. The situation this envisions is not just a matter of ignorance or unknowing, but of being pulled in different and conflicting directions. In particular, apory crises when the claims we are inclined to believe and accept stand in logical conflict with one another, some combing to point towards one conclusion, and others to point to one that is discordant and conflicting. What we then have is, in short, a conflict of belief inclinations. An apory is thus a puzzle of sorts, but one of a very particular sort that calls for resolving an inconsistency, readjusting inconsistent cognitive commitments so as to restore them to harmonious consistency. In invisiioning such a situation, Plato spoke of “confusions and inconsistencies” (tarachês kai aporias) at Theaetetus 167e. And Aristotle has it that aporetic perplexity results from deeming contradictory propositions as equally accessible.1 He thus speaks of aporêma as dialectical inference from contradictory propositions.2 As envisioned by Aristotle, the process of aporetic analysis begins with an information-seeking question, and then sets out a series of possible alternative responses to it. When these responses—and accordingly the complex of reasons that support them—are inconsistent, then determined though we are to get an answer, we cannot have it every which way. We must make choices—something has to give way if anything is to be salvaged from the morass of alternative possibility.

On Certainty and Other Philosophical Essays on Cognition

And only a comparative evaluation of the positive and negative features of those rival contentions can be or and in finding the best way out. Overall, the mission of aporetics is thus to provide a practicable means for coming to terms with inconsistency. Particularly prominent among the situations in which inconsistency arises are: • Conflicting information arising from discordant sources in matters of empirical inquiry. Or conflicts which arise when new information disagrees with the old. • Conflicts of putative fact with speculative supposition (i) in thought experimentation and hypothetical reasoning, (ii) in counterfactual conditionalization, and (iii) in ad absurdum and per impossible reasoning. • Paradoxes in matters of theoretical deliberations where some of our belief-inclinations disagree with others. • In speculative conjecture regarding history. • Conflicts arising in philosophy through the clash of doctrines and contentions. When confronted with an aporetic situation, we of course can, in theory, simply throw up our hands and abandon the entire cluster of thesis involved. But this total suspension of judgment is too great a price to pay. For taking this course of wholesale abandonment we would plunge into vacuity by foregoing answers to too many questions. We would curtail our information not only beyond necessity but beyond comfort as well, seeing that we have some degree of commitment to all members of the cluster and do not want to abandon more of them than we have to. Our best option—or only sensible option—is to try to localize the difficulty in order “to save what we can.” In this way aporetics is, in effect, a venture in cognitive damage control in the face of inconsistencies.



2. RECOURSE TO APORY BY PHILOSOPHERS Apories provide a prime source of philosophical problems, and from the days of Plato and Aristotle onwards, philosophers have confronted complex issues where discordant plausibilities exert an appeal that pulls thought in different directions. Two rationally practicable reactions to inconsistency present themselves when we are confronted with an aporetic conflict among our belief-commitments, namely to revise or to abandon. The standard bearer of these two approaches are Aristotle and Kant. Confronted with a philosophical problem, Aristotle set out to examine the competing alternative resolutions with a view to seeking out that which is optimal. He thus saw aporetic analysis as an instument of investigation. Kant took a very different line. Like the ancient skeptics before him, he viewed philosophical conflict not as a challenge to inquiry but as a symptom of incapacity. For him, the ever-changing conflict of philosophical doctrines is not a challenge to deeper investigation but a “scandal of philosophy”—a cancer calling for excision, a sign that the very issues themselves are inappropriate, untenable, and illusionary. Those reciprocally contradicting antinomies—those quintessential Kantian apories—are clear signs that the conceptions involved over untenable mirages of defective thought that cry out for positivistic abandonment. While for Aristotle, apories are a challenge to theory, for Kant they are a sign of the futility of theorizing, Both Aristotle and Kant viewed apories as an important resource in philosophy. But for the one they were an instrument of construction, for the other one of demolition. Other thinkers took a very different line. This is nowhere illustrated more clearly than in the work of St. Thomas Aquinas. His masterwork Summa Theoligia has the standard format of aporetics analysis. The discussion sets out from a theoretically controlled inventory of questions and makes a tour d’ horizon survey of the available answers to them. It then makes a comparative weighing of the repetitive assets and liabilities (the “costs and benefits” as it were) that accrue to each of these alternative positions with a view to determining what has the greatest balance of positivity on its side. On this basis, Aquinas’ ap-


On Certainty and Other Philosophical Essays on Cognition

proach typifies the approach to apory-analysis envisioned—and occasionally emphasized—by his master, Aristotle. Some recent philosophers, especially those in the Aristotelian tradition have made use of apories and aporetic analyses. Thus Franz Brentano, in his 1874 classic, Psychology from an Empirical Standpoint, characterizes the aporetic method as a salient procedure of philosophical investigation: This method exhibits all the various conceivable assumptions, indicates for each of them the characteristic difficulties, and in particular gives a dialectical and critical survey of all the opposing views, whether formulated by eminent men or held by the people.3

The Germanophone philosopher Nicolai Hartmann (1882-1950) was another thinker in whose work the conception of aporia figured importantly. Following in Kant’s footsteps, he saw philosophical issues as divided into two sectors: ontology which deals with experience-accessible reality and its ramification, and metaphysics which deals with transcendental (i.e., experience-detailed) issues. These latter related to issues which—like the problem of free will—outside the range which science can provide guidance. And these experience transcending issues are ultimately insoluble. Hartmann accordingly endorsed the idea of an aporetics approach that sought to distinguish thought-limits of inconsistency. But he regarded this enterprise as one in which success is achievable only in part—with those ultimate questions of traditional concern declined to remain beyond the reach of experience. It is not that those problems are meaningless (as per the logical positionists) nor based on untenable presuppositions (as per Kant). It is just that they are fundamentally intractable because any and all means of resolution lie outside our reach. It is not just that aporetics thought-perplexity is involved but insolubility as well. Over the history of philosophy apories have thus been cast in both negative (destructive) and a positive (constructive) roles. They have been used negatively by those who sought to find contradictions in the thought of their opponents. And they have been used positively by those who urged their own positions as a way of averting aporetics inconsistency.



3. HOW APORIES PERVADE PHILOSOPHY ITSELF The big issues of philosophy regarding truth, justice, meaning, beauty and the like were encapsulated in Immanuel Kant’s summary of the key questions regarding one’s place in the scheme of things as a rational free agent: What can I know? What shall I do? What may I hope? What should I to aspire to? However, thanks to the inherent complexity of the issues the elaboration and substantiation of answers to such questions inevitably result in contentions that became enmeshed in aporetic conflicts. We have many and far-reaching questions about our place in the world’s scheme of things and endeavor to give answers to them. Generally the answers that people incline to give to some questions are incompatible with those they incline to give to others. (We sympathize with the skeptics, but condemn the person who doubts in the face of obvious evidence that those drowning children need rescue.) We try to resolve problems in the most straightforward way. But the solutions that fit well in one place often fail to square with those that fit smoothly in another. Cognitive dissonance rears its ugly head and inconsistency arises. And the impetus to remove such puzzlement and perplexity is a prime mover of philosophical innovation. And so while philosophizing may “begin in wonder,” as Aristotle said, but it soon runs into puzzlement and perplexity.4 The doctrinal positions of philosophy standardly root in apories— in groups of individually plausible but collectively incompatible contentions. Just here, for example, lay the basic methodological insight of Plato’s Socrates. His almost invariable procedure was a process of “Socratic Questioning” to elicit a pre systemic apory that sets the stage for philosophical reflection. Thus in the Republic, Thrasymachus was drawn into acknowledging the aporetic triad:


On Certainty and Other Philosophical Essays on Cognition

(1) What men call justice is simply what is decreed by the authorities as being in their own interest. (2) It is right and proper (obligatory, in fact) that men should do what is just. (3) Men have no obligation to do what is in the interest of the authorities particularly since those authorities may well themselves be mistaken about what their interests really are. Over the history of philosophy apories have thus been cast in both negative (destructive) and a positive (constructive) roles. They have been used negatively by those who sought to find contradictions in the thought of their opponents. And they have been used positively by those who urged their own positions as a way of averting aporetics inconsistency. In this sort of way, the problem-context of philosophical issues standardly arises from a clash among individually tempting but collectively incompatible overcommitments. Philosophical issues that standardly center about a family of plausible theses that is assertorically overdeterminative in claiming so much as to lead into inconsistency. It lies in the logical nature of things that there will always be multiple exits from aporetic inconsistency. An aporetic cluster is accordingly an invitation to conflict: any single resolution will only be one of a coordinated group of mutually discordant doctrines (positions, teachings, doxa). In addressing philosophical apories in the light of experience, the standard policy proceeds by breaking the chain of inconsistency at its weakest link. But this can be seen in different ways by different thinkers. For in the setting of philosophical concerns this weakness is always a matter of assessed plausibility. And this issue of philosophical plausibility will here be a matter of consonance with one’s fundamental commitments which is—and is bound to be—a matter of experience. The sort of data that a philosopher’s course of experience has brought his way is going to be pivotal in this regard. The cluster accordingly sets the stage for divergent “schools of thought” and provides the bone of contention for an ongoing controversy among them.



One could, in theory, simply suspend judgment in such aporetic situations and abandon the entire aporetics cluster, rather than trying to localize the difficulty in order “to save what we can.” But this is too high a price to pay. By wholesale abandonment we lose too much through forgoing answers to too many questions. We would curtail our information not only beyond necessity but beyond comfort as well, seeing that we have some degree of commitment to all members of the cluster and would not want to abandon more of them than we have to. And this inevitably carries a doctrinal commitment in its wake—a circumstance even more emphatic in total than in partial abandonment. Thus consider the following philosophical apory about freedom of action: (1) People are free agents: they can and sometimes do act from free choice. (2) If an action issues from free choice, then it is causally unconstrained. (3) All occurrences—human actions included—are caused, that it, causally construed by antecedent occurrences. Each of these exerts the appeal of plausibility. But rejecting a thesis is tantamount to endorsing its negation and the negation of a philosophically relevant contention is itself philosophical in its bearing it follows that the pursuit of mere logical consistency involves one in endorsing a philosophical position. Consider the just-stated apory from this angle. The fact that an inconsistent triad is at issue means that one must adopt at least one of the following positions (1)-rejection: This represents a determinism that denies free choice. (2)-rejection: The acknowledgment that free actions can be caused leads straightaway to a theory of agent causation [Kant’s “causality of freedom”] as contrasted to nature-causation [Kant’s


On Certainty and Other Philosophical Essays on Cognition

“causality of nature”]. The resulting position is that of a compatibilism of freedom and causality. (3)-rejection. This involucrates a doctrine of occurrence surdity to the effect that some events are uncaused. The result is causal exceptionalism that sees some events as positioned outside the causal order. In such cases of collective inconsistency, something obviously has to go. Whatever favorable disposition there may be toward these plausible theses, they cannot be maintained in the aggregate. Too many alternatives strive for our approbation and acceptance. And this state of affairs is standard in philosophy and represents a most common and pervasive impetus to philosophical reflection. 4. APORY AND PARADOX Logic can tell us when our modes of reasoning are valid—how it is that when applied to truths they must lead to truths. But it does not— cannot—tell us that when applied to plausibilities, valid arguments cannot yield implausible (or even self-contradictory) conclusions. (After all, a conjunction is often less plausible then its conjuncts.) Logic thus provides no insurance against paradox. And the reason is simple. Paradox resolution requires a choice among alternatives for premiss abandonment. But logic is value-free. It will dictate that we must make choices in the interests of consistency resolution, but not how. It can criticize our conclusions but not our premisses. And so, as long as we are conjuring with plausibilities, the threat of paradox dogs our steps, irrespectively of how carefully and cogently we may proceed in point of logic. The long and short of it is that paradox management requires an extra- or supra-logical resource. For the way to restore consistency to an aporetic situation is to implement some sort of prioritization principle that specifies how, in a case of conflict, we should proceed in making some of the relevant claims give way to others. What is needed is a rule of precedence or right of way. Considerations of priority are needed for breaking the chain of inconsistency at its weakest link.



The guiding ideas of this approach are accordingly two: • that paradoxes of the most diverse sorts can be viewed in a uniform way as resulting from an aporetic overcommitment to theses which, albeit individually plausible, are nevertheless collectively incompatible. And on this basis— • that paradoxes of the most diverse sorts can be resolved through a uniform process of weakest-link abandonment in view of the fact that some of the conflicting theses take precedence or priority over others. And with this second point, considerations of epistemic evaluation⎯of priority determination⎯become an inevitable part of paradox management. This means that plausibility and presumption become the crucial considerations. Plausibility considerations accordingly play a very special role in the cognitive scheme of things. We take recourse to plausible propositions as premisses when working out answers to our questions. But their use in the setting of particular question-resolving contexts is not predicated on an outright and unqualified commitment to these propositions as true. For we know full well that we cannot accept all those plausibilities as truths since this can and generally would lead us into contradiction. Plausibilities are accordingly something of a practical epistemic device. We use them where this can render effective service. But we are careful to refrain from committing ourselves to them unqualifiedly and come-what-may. And we would, in particular, refrain from using them where this leads to contradiction. In sum, our commitment to them is not absolute but situational: whether or not we endorse them will depend on the context. To reemphasize: the “acceptance” that is at issue here represents no more than a merely tentative or provisional endorsement. The shift from accepting-as-true to accepting-as-plausible offers us a degree of flexibility with the claims that we make. Thus consider


On Certainty and Other Philosophical Essays on Cognition

the Exception Paradox which centers on the contention that “All generalizations have exceptions.” This leads to the following paradox: (1) All generalizations have exceptions

by hypothesis

(2) (1) is true

by (1)

(3) (1) is a generalization

from (1) by inspection

(4) (1) has exceptions

from (1), (3)

(5) Any generalization that admits

as a principle of logic of exceptions is false

(6) (1) is false

from (3), (4), (5)

(7) (6) contradicts (2)

Note, however, that this paradox is immediately dissolved when the thesis at issue in (1) is asserted not as true but merely as plausible. For then the inferential step from (1) to (2), which is essential to deriving the contradiction, is automatically invalidated. Unfortunately, life being what it is, we cannot always get away with accepting the plausible outright because actual truth is something more selective and demanding than mere plausibility, seeing that plausibilities—unlike truths—can conflict both with truths and with one another. Distributively true statements are of course collectively true: we have [T(p) & T(q)] → T(p & q). But this is emphatically not the case with plausible let alone with merely probable statements. For plausible (and probable) statements can come into conflict with one another, and thereby impel us into paradox. In paradox resolution as in aporetic analysis in general, rational procedure is a matter of breaking a chain of inconsistency at its weakest link. And this, of course, requires making evaluative appraisals of presumptive acceptability—a process that is ultimately a matter of judgment rooted in the practicalities of everyday-life experiences.



REFERENCES Hartmann, Nicolai, “Zur Methode der Philosophiegeschichte,” in his Kleinere Schriften (Berlin: De Gruyter, 1958). Hülsmann, Heinz, Die Methode in der Philosophie Nicolai Hartmanns (Duሷsseldorf: L. Schwann, 1959). Rescher, Nicholas, The Strife of Systems (Pittsburgh: University of Pittsburgh Press, 1985). Rescher, Nicholas, Aporetics (Pittsburgh: University of Pittsburgh Press, 2009). Siitonen, Arto, Problems of Aporetics (Helsinki: Suomalainen Tiedeakatemia, 1989: Annales Acdemiae Scientiarum Fennicae, Dissertationes Humanarum Litterarum, No. 50). NOTES 1

Topics, 145b2-6.


Physics 185a18.


Franz Brentano (1973), book 1, ch. IV, § 4,p. 73 (Trans. A. C. Rancurella et al). Another example is Gottfried Martin (1957), § 71: “Die Methode der aporetischen Dialektik,” pp. 326-32.


Kant wrote: “Now wonder is a shock of the moral sense, arising from the incompatibility of a representation. . . with the principles already lying at its basis, which provokes a doubt as to whether we have rightly seen or rightly judged” (Critique of Judgment, see. 62; trans. J. H. Bernard [London, 1892], p. 211). Our present construction of the term generalizes this overly narrow construction to include a conflict of “beliefs” as well as one of “representations.”


Chapter Six COGNITIVE OBLIGATION (An Essay in Epistemic Ethics) 1. BASICS


ver the past generation the idea of epistemic duties has been establishing an expanding beachhead on the shores of philosophical deliberation. But the project goes back at least as far as William Kingdon Clifford’s famous 1877 dictum that “It is wrong, always, everywhere, and for anyone to believe something on insufficient evidence.”1 The obligation to developing one’s knowledge has two prominent ethical aspects: • To enhance one’s understanding of things for the sake of self development. • To be better able to act effectively in furthering the interests of others. Duty and obligation thus enter into matters of cognition via two ethical avenues: self-enhancement and morality. The former of these reflects our self-oriented duty as intelligent beings committed by their very nature as such to avail themselves of the opportunities for the good at their disposal. The latter is a concordant of moral obligation. The route to cognitive obligation proceeds from the consideration that to the extent to which we do not know how things work in the world we are unable to render effective support and to others in their hour of need. After all, if I propose to give you medicine to relieve your toothache I had best make sure that I am not mistaking some poisonous substance for that medication. In such matters of application the obligation to get it right is simply part and parcel of our moral obligations.

On Certainty and Other Philosophical Essays on Cognition

In defaulting on our cognitive obligations we accordingly run the risk of injury to the best interests of ourselves or our fellows. Cognitive obligations are thus ethical in nature, and ethical obligations are not always moral obligations as well. For moral obligations are those involved in having the interest of others, while ethical obligations can be self-regarding. First and foremost among these is the obligation to make some use of the prospects for a realization of the good that fate puts at one’s disposal and, in particular, to take some steps to the cultivation of one’s talents. The obligation to take some steps to extend one’s knowledge simply for the sake of self-development is among these obligations. Some obligations are imposed upon us by fate, and some are humanly imposed via self or others—or even society at large. No one asked you whether you wanted to be a fellow human being, or sibling. But signing on as a sea-captain will be a matter of choice for you. And all of these conditions alike carry certain obligations in their wake— irrespective of whether these obligations are incurred on an involuntary or on a voluntary basis. Some of our obligations are existential and mandatory—we have them simply through our existence as members of homo sapiens within the world’s larger scheme of things. Other obligations are functional and optional—we have them ex efficio by virtue of the particular niche we have same occupy, be it as doctors or lawyers, as tinkers or tailors, or husbands or brothers. Our cognitive obligations call for these things 1. Knowledge amplification through inquiry: extending the range and scope of our knowledge 2. Quality control: insuring the credentials of our claims to knowledge by providing for its cogency, its reliability, and its accuracy. 3. Knowledge diffusion: making our knowledge available and accessible to others. However, none of these cognitive proceedings are unlimited and unqualified. In actual practice, all of them are profoundly constricted and circumscribed by circumstances.



No one is obligated to be a scientist or a scholar. No one is obligated to become a polymath. The imperative to knowing is dependent on one’s interests, talents, and capacity. Still there is such a thing as inexcusable ignorance. No one can appropriately fail to realize that sticking pins into people will hurt them. And no one can appropriately think that hitting people with a stick would cure them of a headache. There is a range of knowledge that is mandatory for everyone above and beyond the range of knowledge that is so for individuals who play a particular role (as present day parents, physicians, and accounts). Insofar as we are obligated to the development of knowledge, we become derivatively obligated to ensure that the things we accept as items of knowledge are actually qualified to be accounted as such. And this makes it a matter of duty to assert adequate quality control over what we accept as authentic knowledge. Moreover, knowledge does not fall upon us like manna from the heavens. Its production as a human effort and activity. And the requisite process of inquiry requires an expenditure of time, effort, and resources. The resultant economic dimension has significant ethical aspects. For the time effort and resources at our disposal are limited. In consequence, ethics and common sense alike require that they be expended effectively and efficiently. These conditions highlight a further important factor in epistemic ethics, namely importance. For items of information are not created equal. The amount of information that could be developed regarding a plot of grass is virtually endless. Each blade of grass, each spec of earth, each stick or stone could be described in endless detail, its every facet and feature characterized in detail and then the relationship of each of them to the other could be expatiated upon. Volumes without number could be devoted to the project. But to what avail? What use could possibly be made of this useless information. In theory, all information informs, but in realistic practice we are obligated to a diligent heed of significance and importance. Here as elsewhere a proper balance of costs and benefits is of the essence. However, in matters of epistemic ethics it is important to distinguish between epistemic merit (or virtue) of knowers and the epistemic merit of what is known. The application of this distinction is set out in Display 1.2


On Certainty and Other Philosophical Essays on Cognition

________________________________________________________ Display 1 COGNITIVE VIRTUES Virtues of Knowledge •

As individual items of knowledge —substantiation/evidentiation —probability —accuracy/detail —importance

As bodies of knowledge —breadth/extent —depth —overall importance

Virtues of Knowers —Reliability/Veracity —Conscientiousness/Care —Informedness/Breadth —Profundity/Depth —Openness (to new ideas and initiatives) —Accessibility (willingness to share information) —Cooperativeness —Good judgment

______________________________________________________ It is clear on this basis that insofar as people have an ethical obligation to the cultivation of knowledge they also: • Have an obligation to cultivate their virtues of knowers and • Have an obligation to see to it that their knowledge claims exhibit the virtues of knowledge It is an interesting question whether cultivating the virtues of knowers will be rewarded with products that more amply exhibit the virtues of knowledge. There is certainly no reason of logically airtight necessity why their cognitive virtue must issue in cognitive success for individuals. Able inquirers do not necessarily arrive at the truth. But there is, of course, strong reason of an inductive bearing why it



should be so. After all we would not regard those knower-virtues as such—this is, as authentic virtues—if ample experience did not indicate the cognitive appropriateness of the products that issue from their cultivation. 2. LIMITS Still, we are almost never obligated to develop the quality of our knowledge beyond a certain point. To know if the restaurant is crowded, one need not aim for a precise headcount. If you want to know Harry’s height we need not measure it to the nearest millimeter. If you want to know someone’s age one need not specify it to the nearest minute. The detail, precision, and accuracy of our knowledge need seldom be carried to extreme lengths. To determine the spelling of an English word I need not consult every extant dictionary, one or two will be quite sufficient. Those of our cognitive obligations which relate to matters of quality control are never unlimited but reach only so far. Responsibility has its limits to here. We need only ensure that our cognitive claims and our practical recommendations are well grounded, precise, etc up to a certain point—one whose position hinges on the specifics of the situations at hand. As regards such cognitive desiderata we always reach a point of sufficiency where enough is enough. And this means that cognitive ethics demands not only knowledge but good judgment as well. Pushing matters too far here is often not virtue but vice. Consider the following example: Suppose that we are sailing on the open sea on a vacation cruise ship. It is dusk, and the visibility is getting poor. As we stroll on deck along the rail of the ship, there is suddenly a shout, “Man overboard.” Someone grabs a life preserver from the nearby bulkhead and rushes with it towards the railing. Suddenly, he comes to a stop and hesitates a moment. To our astonishment he turns, retraces his steps, and replaces the life preserver, calmly proceeding step by step as the region of the incident slips away, first out of reach, then out of sight. Puzzled and chagrined, we turn to the individual and ask why he broke off the rescue attempt. The response runs as follows: “Of course, throwing that life preserver was my first instinct, as my behavior clearly showed. But then some ideas from my undergraduate


On Certainty and Other Philosophical Essays on Cognition

epistemology courses came to mind and convinced me that it made no sense to continue.” Intrigued, we ask for more details and receive the following response: Consider what we actually knew. All we could see was that something that looked like a human head was bobbing out there in the water. But the visibility was poor. It could have been an old mop or a lady's wig stand. Those noises we took for distant shouts would well have been no more than a pulsing of the engines and the howling of the wind. There was simply no decisive evidence that it was actually a person out there. And then I remembered W. K. Clifford’s aforementioned dictum: “It is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.” So why act on a belief that there was actually a human being in danger out there, when the evidence for any such belief was clearly insufficient? And why carry out a rescue attempt when you do not accept that someone actually needs rescuing? Something has clearly gone badly wrong here. We may not choose to fault our misguided shipmate as an epistemologist, yet we cannot but wonder about his moral competency. Even if I unhesitatingly accept and endorse the abstract principle that one must try to be helpful to others in situations of need, I am clearly in moral difficulty if I operate on too stringent a standard of evidence in relevant contexts—if, for example, I allow sceptical concerns about other minds to paralyze me from ever recognizing another creature as a human person. For then I will be far reachingly precluded from doing things that, morally considered, I ought to do. William James rightly noted this connection between epistemology and morality, in insisting that the sceptic rudely treads morality underfoot: “If I refuse to stop a murder because I am in [some] doubt whether it is not justifiable homicide, I am virtually abetting the crime. If I refuse to bale out a boat because I am in doubt whether my effort will keep her afloat, I am really helping to sink her. . . . Scepticism in moral 3 matters is an active ally of immorality.” There is much to be said for this view of the matter. To operate in life with epistemological principles so stringent as to impede the discharge of one's standard moral obligations is to invite justified reproach. Where the interests of others are at risk, we cannot, with moral appropriateness, deploy evidential standards of acceptabili42


ty of a higher, more demanding sort than those that are normally operative in the community in the ordinary run of cases. At this point, epistemology has moral ramifications. For morality as we know it requires a commonsense, down-to-earth epistemology for its appropriate implementation. In such a case, then, the stance we take in the one domain (epistemology) has significant repercussions for the way we can proceed in the other (ethics). The issues arising in these seemingly remote areas stand in systemic interlinkage. Externalities can come into play. A problem-solution that looks like a bargain in the one domain may exact an unacceptable price in the other. NOTES 1

W. K. Clifford, Lectures and Essays, ed. by L. Stephen and F. Pollock (London: Macmillan, 1879, 2nd ed. 1886; originally published in the Contemporary Review, vol. 30 (1877), pp. 42-54.)


On virtue epistemology at large see especially Ernest Sosa, Knowledge in Perspective (New York: Cambridge University Press, 1991).


William James, “The Sentiment of Rationality,” in The Will to Believe and Other Essays in Popular Philosophy (New York: Longmans, Green, and Co., 1897), p. 109.




e often view the world myopically, with a vision that is not infrequently blurry and imprecise. The propositions we accept are then surrounded by a penumbra of others whose content is virtually indistinguishable, each forming part of a cluster of others that are informatively proximate. Whenever this transpires to a substantial extent, we will have it that propositions “say almost the same thing”, the bulk of the information conveyed by the one being conveyed by the other as well. Thus consider the statements “John arrived [here] last Tuesday.” and “John came here the day before yesterday [today being Thursday]”. Yet while they seem equivalent, there are differences. The former, unlike the latter, strongly suggests that John is still here. And if I am even slightly uncertain about today’s being Thursday, then the former conveys more secure information about the timing than the latter. Such statements come close to being equivalent without quite bringing it off. Their equivalency is only rough and approximate. In this way, p and q are roughly equivalent—symbolically p ≈ q—iff virtually the whole of p’s assertive content is also q’s, and conversely. Roughly equivalent statements will not necessarily have the same status in point of true and falsity however, since what is being said is going to be somewhat different. When one is true it is not likely that the other is false, but it is not impossible. We will thus have it that: (∃q) (q ≈ p) is not outright incompatible with (∃q) (q ≈ ~p). This idea that approximately equivalent statements can differ in truth status may invite resistance. But consider the statements: (1) Tom was surprised that John had forgotten his name. (2) Tom was surprised that John had forgotten who he (Tom) was.

On Certainty and Other Philosophical Essays on Cognition

These might well be deemed to be approximately equivalent. Yet while we usually identify people by name, nevertheless their mode of identification might be different. For suppose that John suffers from senile dementia and that Tom knows it. Then Tom might well be notat-all surprised about that name-forgetting. Yet, nevertheless, he might still be surprised that John had forgotten him, who, after all, was his longtime best friend and the companion of his youth. Translation is a good example of assertive approximation. Even a very exact translation can have problems. Take even so simple and literal translation as •

The cat is on the mat (English)

Die Katze ist auf der Matte (German)

That “obvious” German equivalent is complicated by the fact that “Matte” can mean meadow as well as mat. 2. APPROXIMATE TRUTH STATUS A proposition p is (flatly) true—symbolically Tp—whenever this proposition p obtains—period. Accordingly, we have the Tarski equivalence: Tp iff p Extending this idea, we will have it that p is approximately true, symbolically T*p, whenever some proposition q that is roughly equivalent with p is (flatly) true: T*p iff (∃q) (q ≈ p & Tq). In effect, T*p thus says that some version of p is true. Once we adopt this idea of propositional approximation, the logical landscape blossoms. For now any one-place propositional operator F(p) will yield an approximative counterpart as per:



F*(p) iff (∃q) (q ≈ p & F(q)) The conceptions of truth, falsity, and negation provide immediate illustrations of such approximation. Given our earlier observations, it can transpire that T*P obtains even when ~Tp. However, whenever p itself is true, T*p will be so as well, seeing that p ≈ p. But if p is false, T*p may be true or false—all depending. The situation with respect to truth value is often going to be indecisive, as per the following tabulation:1 P


T* ~ p

~ T* p

~ T*~ p








T*(p & q)

T*(p ∨ q)





Appearances to the contrary notwithstanding the “approximate truth” operator T* does not address the truth-value of propositions, but merely some aspect of their truth-condition. There is also the stronger conception of holding good throughout the domain of approximation: T#p iff (∀q)(p ≈ q ⊃ p) Given our earlier observations, it can happen that even a true proposition can fail in this regard: Tp certainly does not entail T#p, although the converse is always the case.


On Certainty and Other Philosophical Essays on Cognition

3. ROUGH EQUIVALENCE For any proposition p, there will now be the set comprising all those propositions that affirm roughly the same thing: [p] = {q:q ≈ p} Accordingly, p ≈ q iff q∈ [p], that is, whenever q is some version of p. It will of course transpire that ≈ is reflexive in that: (∀p) (p ≈ p) or equivalently (∀p) (p∈ [p]) And, obviously, if p affirms roughly the same as q, then the reverse must also be the case, so that we have it that ≈ is symmetric, that: (∀p) (∀q) (p ≈ q → q ≈ p) Moreover, we also have it that whenever p ≈ q, then ~ p ≈ ~q, and conversely. However, by a slippery slope argument of the familiar kind we do not have it that ≈ is transitive. It is not the case that: If p ≈ q and q ≈ r, then p ≈ r. Accordingly, ≈-”equivalent” propositions will not be generally intersubstitutable. And further we will also not have it that: If p ≈ q and p ≈ r, the p ≈ (q & r). If both q and r are some version or other of p, their conjunction need not necessarily be so on grounds of discrepancy. To get a firm grip on ≈-equivalency, it is useful to consider some of its salient features: • p ≈ q does not entail Tp ≡ Tq. Merely approximative variants can differ in truth value.



• p ≈ q does entail Tp ≈ Tq. For we have it that p ≡ Tp (via Tarski) • p ≈ q entails F(p) ≈ F(q) whenever the propositional function F is truth functional. (However F(p) ≡ F(q) does not follow.) With the preceding concept of propositional approximation in view we can, first of all, address the issue of propositional functions of a single variable, as per F(p). For given any propositions operator F we can immediately achieve its approximative counterpart F* via the specification. F* (p) iff (∃q) (q ≈ p & F(q)) We have already seen how this works out in the case of truth. But of course we can adopt the same approach as regards falsity via the specification: Np iff ~p This now yields a concept of approximate falsity as per: N*p iff (∃q) (q ≈ p & Np) As noted above, we have it that [p], the approximationneighborhood of a proposition p, can in principle include both true and false versions of it. Accordingly the selfsame proposition can in principle be both approximately true and approximately false. For neither does T*p entail Tp, nor is the converse the case. And this holds for N*p and Np as well. Truth (or falsity) as such, and approximate truth (or falsity) are separate and distinct issues. 4. APPROXIMATE INFERENCE Let it be that we have a valid deductive argument of the format: P1, P2 ├ Q


On Certainty and Other Philosophical Essays on Cognition

Are we now entitled to reason as per: T*P1, P2 ├ T*Q? When one of the premises of a deductively valid argument is approximately true, can one infer an approximate conclusion? Consider modus ponens: p p→q ∴q This is valid. But observe that we cannot reason: T*(p) p→q ∴ T*(q) For this seemingly plausible reasoning is not valid, seeing that the conclusion q may well depend upon the exact and precise content of p and thereby become entirely destabilized and not even approximately true when p is even only the smallest bit altered. To make this graphic let p be X = 100 and let q be X3 – 1,000,000 = 0 Here clearly p ├ q. But let it be that p only approximately true with X being 99. Then nevertheless q will not be approximately true, the difference between X3 and a million being quite substantial (i.e. about 30,000). However, there indeed are cases where inference approximation works out successfully. For example consider: 50


All As and Bs All Bs and Cs ∴All As and Cs We will clearly have it that the following approximative variant holds good: Approximately/roughly: All As and Bs All Bs and Cs ∴Approximately/ roughly All As and Cs However, this inference would not be valid if the second (minor) premiss were approximate. That is, we do not have: All As are Bs Approximately/roughly All Bs are Cs ∴Approximately/roughly all As are Cs For it is perfectly possible that the As constitute those thus Bs that are the exception to the rule. In systematic inference approximate conclusions require appropriate majors. 5. APPROXIMATION-CONNECTIVITY HOMOGENEITY



As noted above we do not have approximative transitivity: (p ≈ q & q ≈ r) → p ≈ r Nevertheless the fact that p and r can be connected in this way does mean that there must be some kinship between them. For example, a certain thematic homogeneity must obtain. (For example, if p is about cats, then r must be so as well.) And any such relationship could of course extend further, extending along an entire approximation-chain of the format:


On Certainty and Other Philosophical Essays on Cognition

(p ≈ q1 & q1 ≈ q2 & q2 ≈ q3 &. . . & qn ≈ r) When this situation obtains, we may say that approximation connectivity obtains between p and r. It means that a certain kinship— modality and thematic homogeneity included—will prevail throughout the entire set of propositions [[p]] that are approximation- connectable with p. (Truth, however, does not qualify in this regard.) F is an approximation-universal feature (AUF) of a proposition p iff F characterizes each and every approximation to p: (∀q) ([F(p) & q ≈ p] → F(q)) With the classical modalities of necessity and possibility ( and ◊, respectively) we obtain such approximation-counterparts as: *p iff (∃q) (q ≈ p & q) But can be have already noted these conditions are only satisfied when p and ◊p, respectively—see section 1 above. There just are no distinct notion of approximative necessity and possibility. Such absolutes do not admit of approximation: they are subset to a decisive yes-or-no. Propositional approximation can be reflected in the equivalence: F*(p) iff (∃q) (q ≈ p & F(q)) There is, however, a far stronger conception that relates not to approximation such as, but approximation-local pervasiveness. This is based on the principle: F#(p) iff (∀q) (q ≈ p → F(q)) The idea here is that what holds of p also holds good for anything substantially like it. Thus while an approximate feature holds of some version of a proposition such a pervasive feature holds for all of its approximative variations. Necessity and possibility are like that.



On this basis, the modal status of a proposition as possible, contingent, or necessary must be taken to qualify as approximation-universal features of that proposition. And aboutness is yet another example: if p is about cats we would not and could not regard q as a close approximation to it if q were not also about cats. 6. APPROXIMATION WITH PROPOSITIONAL FUNCTIONS OF TWO VARIABLES Complications arise when approximation is carried over to twoplace propositional functions. For with R(p, q) that is the prospect of approximation regarding either of the two variables at issue. So we here arrive at a disjunctive specification: R*(p, q) iff ([(∃r) r ≈ p & R (r, q)] ∨ [(∃r) r ≈ q & R (p, r)]) This general principle gives rise to a plethora of implementations. We will for example be able to apply it to • conjunction : p & q • disjunction : p ∨ q • implication : p → q Thus, for example, approximate implication is p →* q iff [(∃r) ([r ≈ p & r → q) ∨ (∃r) (r ≈ q & p → r)] With n-place operators there are always n pathways to approximation, and with each of them we arrive at a rather “natural choice” version of the operation at issue. 7. LEXICOGRAPHIC APPROXIMATION The idea of semantical approximation is readily extended from propositions or statements to words or expressions. The key here lies 53

On Certainty and Other Philosophical Essays on Cognition

in the classical salva veritate criterion of sameness of meaning: Two words or expressions have an identical of meaning when they are intersubstitutible in statements without altering their truth status. In our present context this idea can be transposed to read: Two words or expressions have an approximate identity of meaning when they are intersubstitutible in statements without altering their approximate equivalency kinships.

Thus when statements of roughly the same meaning result systematically with the substitution of two expressions, then these expressions will have approximately the same meaning. Thus let F be a sentence that places the expression e in a sentential context F. We can then define: e1 ≈ e2 iff (∀F) (F ≈ F) This equivalence effectively specifies the approximate meaningidentity of words (expressions) in terms of what obtains among propositions. We have, in effect, shifted from a contextual standard of meaning-identity to a contextual standard of meaning-approximation 8. APPROXIMATE KNOWLEDGE It is one thing to have knowledge of an approximate truth, and something rather different to have an approximate knowledge of a truth. The former comes to Kxq & q ≈ p, the latter to Kx(q & q ≈ p). These are distinct in the latter, unlike the former, require Kx(q ≈ p). In the former case x knows a truth that approximate to p; in the latter he is also cognizant of that approximation. 9. EROTETIC ISSUES When a certain proposition affirms an (exactly) correct answer to a question there will generally also be some approximately answerequivalent proposition that provides an approximately correct answer. It is plausible to stipulate:



If p @ Q, then (∃q) (q ≈ p & q@* Q) Here p @ Q stands for “p is a correct answer to the questions Q.” However, the stronger thesis that all approximately answerequivalent propositions will so function, namely If p @ Q, then (∀q) (q ≈ p → q@* Q) Is not comparably acceptable, It’s failure follows at once from the above-noted consideration that even if q is such that (p ≈ q) & q, it may nevertheless fail to be the case that every other ≈-equivalent version of p is true. 10. CODA Why do we trade in cognitive approximations. There are in fact many reasons. Sometimes we make approximate claims because we can do no better in the circumstances. Sometimes we do so because there is safety in imprecision—to offer information “in the neighborhood” rather than something that is to secure greater assurance of correctness. And sometimes we are content with an approximation because there is no need for added precision and exactness. Especially if speed of information-access is of the essence, approximate may provide a convenience in meeting the needs of the situation.2 In matters of theoretical cognition we standardly seek precision, rigor, exactness, universality. However, in matters of practical concern we are willing to settle for less, to compromise idealities in the interest of the needs of the occasion. So it is here that approximation comes upon the scene. We would avoid it if we could—approximation is not among the idealities. It is a relevant concession we make in the interest of viable praxis. Be this as is may, propositional approximation opens the doorway to a wide variety of attention-worthy issues. The present deliberations merely create a crack for the window of opportunity that beckons here.


On Certainty and Other Philosophical Essays on Cognition


We here have a situation that is clearly not truth-functional, but rather is only indecisively quasi-truth-functional. On propositional functions of this sort see the author’s “Quasi-Truth-Functional Systems of Propositional Logical,” in The Journal of Symbolic Logic, vol. 27 (1962), pp 1-10.


Instructive considerations that bear upon this line of thought are provided in Chapter 2, “Logic and Conversation,” of Paul Grice, Studies in the Way of Words (Cambridge Mass: Harvard University Press, 1989).






eflexivity—the functioning of things in relation to themselves— pervades the universe. Physical arrangements of all sorts can operate in self-affecting ways. The ancient Greeks saw self-inaugurated motion as the defining feature of animality. And with the emergence of intelligence in the cosmos, self-conscious and self-understanding, come into being. All in all, then, reflexivity is a salient feature of the complex nature of things. Then too, the idea of a personal self has been a dominant conception in modern Western philosophy. One of the earliest philosophical injunctions arises in relation to cognitive reflexivity, encapsulated in the Socratic dictum: know thyself (gnôthi seauton). And the issue of knowledge about our knowledge comes to the fore here. The cogito of Descartes put thought at the center-stage in regard to knowledge as the fons et origo of certainty. Leibniz’s conception of apperception makes self-apprehension the core of metaphysical concepts. Hume as empiricist tries to eliminate the self as inaccessible to sensory experience. But Kant shifted the issue to the transcendental self required by a conscious awareness of facts. For he saw this as the precondition of experience—a prerequisite for having any meaningful experience at all. As Fichte then saw it, agency rather than knowing as the crux: the self as agent creates the self as knower. Hegel sought to socialize the self with the core emphasis not on oneself but on us, the community. And in this regard pragmatists emphasized the projection of the sell via communication. But later thinkers propose to see self’s salient feature not as agency but as suffering (Schopenhauer), and in particular as agonizing in the face of our mortality (Unamuno, Freud). Latterly it is rather self-assertions that has stood in the foreground. However, in the

On Certainty and Other Philosophical Essays on Cognition

present discussion it will be the well-worn Cartesian factor of reflexive cognition that will stand at the focus of concern. But just how much can a people get to know about their own knowledge—its defects included? After all, a picture cannot completely depict itself; a story cannot completely contain itself, a statement cannot completely describe itself. All such modes of self-containment are, by nature, impracticable. But while pictures, stories, and statements can refer to themselves without problem or paradox, they nevertheless cannot manage to internalize themselves. For in each such case of would-be self-embodiment, there is set afoot a process of sequential realization that cannot be competed. The process is akin to trying to effect a transit from A to B by first stepping to the halfway point A1 between A and B, and then to the halfway point between A1 and B, and so on. No matter how long we proceed in this sort of way the final destination is never reached, the complete journey never accomplished. The reflexive nature of cognitive ignorance is exhibited by the circumstance that the thesis • p is a (particular) fact that I (Rescher) do not know. represents something that I myself cannot concretely instantiate. Cashing this in by way of a particular p is doubtless unproblematic for you, but it is inaccessible to me. Unquestionably there are many facts I do not know. But I myself cannot identify any one of them. Their concrete substantiation, while unproblematic for you, is unrealizable by me. The critical difference here reflects that between Ki(∃p)~Kip and (∃p)Ki~Kip. And this betokens the circumstance that the details of my ignorance are hidden from me in an impenetrable obscurity. Perhaps the most fundamental thing that we know about our knowledge is that it is incomplete. However while we know that there are facts about which we are ignorant, we do not know in specific what they are. The contention “F is a fact that I do not know” makes no sense. For in claiming F to be a fact I already claim to know it. We humans—all of us—are finite knowers: there are facts that we just do not know. But we cannot pinpoint them.



It is often said that knowledge is a matter of agreement with reality (adequatio ad rem). But this contention has its problems. For knowledge can be imprecise. I can know that there are roughly 20 people in the room. But reality has to be exact (at any rate above the quantum level). It cannot put roughly twenty people into that room but has to make up its mind” about just how many. So among the things that I know about my knowledge is not only that it is incomplete but also that it is inexact and thereby fails to do justice to reality in this way as well. But how adequate is my knowledge of what I do actually know? 2. FINITE KNOWERS ARE COGNITIVELY LIMITED With finite knowers, the contention (∃x)(∀t)Kxt—to the effect that some knower knows all truths—is by hypothesis false. Individually no finite knower knows everything.1 And not only is this thesis false with finite knowers, but so is (∀t)(∃x)Kxt—the thesis for every truth t there is some knower x such that x knows t. There indeed are facts that finite knowers do not know, even collectively. This emerges as follows. Let x1, x2, . . ., xn be the (finite) totality of finite knowers. Since for any i there will be a truth t1 unknown to knower xi, it transpires that the conjunction of all of these ti will be unknown to the entire group. There will, accordingly be truths that are unknown to any and every knower. To be sure, a knower can, quite unproblematically, know that he does not know something. The thesis Kx(∃p)(p & ~Kxp) is perfectly plausible. However this is emphatically not the case with the thesis that there is some particular, specifically identified proposition of which someone knows that it is a truth unknown to him. (∃p)Kx(p & ~Kxp)


On Certainty and Other Philosophical Essays on Cognition

For this contention would enmesh us in selfcontradiction since our knower would have to know that what is at issue is a truth.2 No knower can know of a fact that it is a truth that is unknown to him. 3. MORE DRASTIC LIMITATIONS But not only are there truths that finite knowers do not know but even more drastically there are truths that finite knowers cannot possibly know. We thus also have: (∃t)~◊(∃x)Kxt or (∃t)…(∀x)~Kxt





The rather complex demonstration of this portentous thesis is set out in the Appendix. It will, of course, be impossible to provide a concrete example of a truth that finite knowers do not—or cannot—know. For giving such an illustration would require indicating a specifically unknown fact, and here there is a vitiating conflict between the claim to unknowedness on the one hand and factuality and truth on the other. For in claiming F to be an unknown fact you take away with one hand (unknowedness) what you offer with the other (factuality). 4. OMNISCIENCE There is, however nothing in the preceding considerations establishes that knowledge of all truth is in principle impossible, with respect to all knowers: ~◊(∃x)(∀t)Kxt For our initial purview has been explicitly limited to finite knowers. Nothing in its scope negates the possibility of an omniscient knower, ξ, for which (∀t)Kξt



But of course when omniscient knowers are introduced into the realm of discussion, the situation is bound to look very different. For the difference between finite and infinite knowers is not just a matter of what they do know, but of what they can know. 5. RAMIFICATIONS OF FINITE KNOWLEDGE From classical antiquity on, theorists have insisted that there is no inferential transit from mere possibility to authentic reality. De posse ad esse non valet consequentia was the watchword. But the medieval schoolmen insisted on one significant qualification. For as they saw it here, as elsewhere, God was the exception to the usual rules. Thus Scotus—duly followed by Leibniz—saw God’s existence as following from his mere possibility. And, in specific, with him—and him alone—the very possibility of his knowing a fact entails his actually doing so. With finite knowers there will be some truths they cannot possibly know, but with an omniscient knower there will be no unknown truths. Cognitive reflexivity is a matter of what an individual knows about his own knowledge. Now every knower certainly knows that he himself knows something. One can regard the thesis: (∀x)(∃p)KxKxp as an epistemic axoim. It is assured (inter alia) by the Cartesian Cogito: “I am.” And this means that all knowers will know that someone knows something: (∀x)Kx(∀p)(∃y)Kyp And this in turn means that there is a proposition that all knowers know in common, namely (∃p)(∃y)Kyp


On Certainty and Other Philosophical Essays on Cognition

Accordingly there is an item of universally shared knowledge, viz. a truth that is known to every knower whatsoever, seeing that everyone knows that somebody knows something. Further, it seems plausible to stipulate that every finite knowers knows that he is finite, i.e., that there is something he does not know: (∀x)Kx(∃p)(p & ~Kxp) However, while a finite knower doubtless knows that there are facts that he does not know, he cannot possibly know what those facts are. For—as already noted—to know that F is a fact that I do not know, I would have to know that F is a fact—contrary to hypothesis. No finite being can be adequately informed about the details of his own ignorance. And not only would there be facts that lie outside the cognitive reach of individual finite knowers, but there will be facts that no finite knower whatsoever can possibly know. Thus let f1 be a fact that knower1 cannot possibly manage to know, and in general let fi be a fact that knower xi cannot ascertain. And now consider the conjunction of all of the fi for all those (finitely many) knowers xi This megafact will, by its very composition, necessarily lie beyond the reach of finite knowers at large. The detail of the imperfection of human knowledge—alike individually and collectively—lies beyond the reach of human cognition. 6. THE KK THESIS When I know something do I thereby know that I know it? Do we have the KK thesis? Kip ├ KiKip with i = I myself, and more generally Kxp → KxKxp? Certainly if “knowing” is construed in an explicit, consciousawareness mode, then that statement looks to be false. I may know that your name is Henry tacitly without ever bringing this fact itself upon the agenda of conscious deliberation.



I realize more things to be so than I am overtly aware of. After all, awareness is not an automatically reflexive cognitive condition: I can be aware of something without being aware of my awareness of it. Were it not so, an infinite regress would be imposed upon us. Only if the idea of “knowing” were construed in a sufficiently liberal sense (say by including tacit knowledge and thereby also crediting me with a knowledge of the “obvious consequences” of the fact that I know) then the KK thesis will indeed obtain. But this is a rather stretched sort of “knowledge”. After all, we would, quite in general, be extremely reluctant to allow any significant conclusion to be drawn from the sheer ignorance of someone’s failing to know that he doesn’t know something. That is, one should surely look with suspicion at any proposed thesis of the format: ~Kx~P ⊃ P But once we adopt Kxp ⊃ KxKxp We are saddled with ~KxKxp ⊃ ~Kxp And now there is no escape from ~Kx~[~Kxp] ⊃ [~Kx~p] which, of course, has exactly the suspect format via the substitution of ~Kx~p for P. Or again, let us contemplate the idea that “x knows p” can be construed as “x realizes any and every immediate consequence of p to be the case.” It then clearly need not be that x realizes the immediate consequences of those immediate consequences—and so on ad indefinitum, as the KK thesis would require.


On Certainty and Other Philosophical Essays on Cognition

To be sure, it x were an omniscient being—that is if x were ξ—then we would have: p ⊃ Kξp And this would automatically assure ubiquitous cognitive reflexivity Kξp ⊃ KξKξp But to stipulate this sort of thing for finite beings would be carrying matters too far. For with finite beings, not only is knowledge limited but so is cognitive self-knowledge. Even were it the case that objective, externally oriented facts were always accessible to us (something which is surely not so, if only because the scope of such facts outruns the limit of finitude), our cognitive situation would still remain such that certain reflexive, selforiented facts in the realm of cognition are bound to be inaccessible to us. 7. EPISTEMIC POSSIBILITY With finite knowers, the concept of epistemic possibility plays an important role. It inheres in the definition that something is epistemically possible when it obtains for aught that anyone knows to the contrary: ’p iff ~(∃x)Kx~p However, it is important here that only finite knowers be at issue. If an infinite knower ξ is allowed into the x-range, then ’ becomes otiose in that ’p reduces to p itself.3 On this basis, the KK-thesis is tantamount to ’’p ⊃ ’p Moreover, since we have



‰p iff ~’~p it at one follows that epistemic necessity is simply a matter of being known. Thus ‰p ≡ (∃x)Kxp So if ξ is admitted into the x-range then ‰ becomes otiose with ‰p reducing to p itself. We then realize the Spinozistic situation that possibility and necessity are exactly one and the same. Epistemic contingency now vanishes, and p & ’~p collapses into contradiction. Appendix PROOF OF (∃t)…(∀x)~Kt FOR FINITE KNOWERS In the endeavor to the necessary limitedness of finite knowers, it will facilitate our reasoning to begin by establishing THE MODAL COLLAPSE LEMMA This lemma stipulates that whenever F is such that (C) F(p & q) ⊃ (Fp & Fq)

Conjunctivity thesis

(V) Fp ⊃ p

Veracity thesis

(P) p ⊃ ◊Fp

Possibility thesis

then p ⊃ Fp PROOF


On Certainty and Other Philosophical Essays on Cognition

(1) F(p & ~Fp) ⊃ (Fp & F~Fp) From (C) by the substitution of ~Fp/q (2) ~F(p & ~Fp) From (1) because its consequent is selfcontradicting via (V) (3) …~F(p & ~Fp) From (2) by necessitation (5) ~◊F(p & ~Fp) From (3) by modal logic (6) (p & ~Fp) ⊃ ◊F(p & ~Fp) From (P) by the substitution p & ~Fp/p (7) ~(p & ~Fp) From (5), (6) (8) p ⊃ Fp From (7) With this Lemma in hand, let us now proceed to demonstrate that for finite knowers (∃t)…(∀tx)~Kxt. We shall now let the F of the preceding discussion be specified as follows: Fp = (∃x)Kxp To begin with note that with regard to knowledge we have both (KV) Kxp ⊃ p


(KC) Kp (p & q) ⊃ (Kxp & Kxq)


The K-veracity principle maintains that knowledge relates only to truths. The very concept of knowers is such that one cannot know



falsehoods. When someone commits to a falsehood, one has to say that he merely thinks that he knows this item. Moreover, K-conjunctivity obvious on the very surface of it: the individual who does not know a conjunct does not know its cognition with other material. Let us now suppose that K-possibility also obtains in that we have (KP) p ⊃ ◊(∃x)Kxp or equivalently (∀t)◊(∃x)Kxt Under this assumption all of the premises of the Modal Collapse Theorem are satisfied, and we shall obtain (∀p)(p ⊃ (∃x)(Kxp) or equivalently (∀t)(∃x)Kxt But just this is a thesis which—with respect to finite knowers—was refuted in Section 1 above. And so we have, in effect, a reductio ad absurdum of (KP). And so with respect to finite knowers: ~(∀t)◊(∃x) or equivalently (∃t) …(∀x)~Kxt With finite knowers there will be a necessarily unknown truth. But of course no-one can ever specify it.4 WE CANNOT ASSESS THE SCALE AND SCOPE OF OUR KNOWLEDGE One important aspect of the incompleteness of our knowledge inheres in the fact that present cognition cannot speak for future knowledge. We can say nothing about the substance of future discoveries— for if we could they would not, after all, have to await the future. The discoveries of our future inquiries—be it in science or elsewhere—are, virtually by definition, unavailable to the inquirers of the present. And not only cannot the substances of future knowledge be presently discussed but neither can its scope. With geographic explanation the size of the end could be determined in advance and thereby the scale and scope of unexplained terra incognita could be assessed. Nothing like this is possible with cognitive as opposed to geographic 67

On Certainty and Other Philosophical Essays on Cognition

exploration. The knowledge of the present cannot speak for the scale and scope of future knowledge: the quantity and extent of its reach cannot be foretold. C. S. Peirce viewed scientific discovery on an analogy with global exploration. As he saw it, later is always lesser, even as geographic discovery moved from large to small: the first great explorers discovered hemispheres, their successor’s continents, later explorers discovered islands, finally the targets of exploration came down to climbing mountains and discovering stream-sources. Peirce thought that scientific discovery was much like this: adding further detail to refine predetermined magnitudes. But this is just not how things work. Smallscale findings can have massive reverberations. The perihelion of Mercury can disestablish the entire grand edifice of classical Newtonian physics. Small anomalies can upset big theories and effect radical innovations. Later will often be different but not necessarily lesser. The advance of science does not impose a fixed direction of importance. And so, not only can we not foresee the substance of future scientific discovery, we know neither its extent nor its significance. Our knowledge regarding what we know is not only imperfect and incomplete but it is also imponderable with as regards its significance. Now only must we recognize the incompleteness of our knowledge, but we have no way of knowing how important the facts that we know not are in the overall scheme of things. In inquiry as elsewhere we must struggle along doing the best we can. UNIDENTIFIABILITY AND VAGRANT PREDICATES Reality outruns the range of our knowledge of it. And thereby a peculiar and interesting aspect of our ignorance emerges with the mode of reference that occurs when an item is referred to obliquely in such a way that its specific identification is precluded as a matter of principle. This phenomenon is illustrated by claims to the existence of ⎯a thing whose identity will never be known. ⎯an idea that has never occurred to anybody. 68


⎯an occurrence that no-one has ever mentioned. —an integer that is never individually specified. There are certainly bound to be such things, but we obviously cannot identify them.5 In these cases, those particular items that render “Some x has F” true are referentially inaccessible: to indicate them individually and specifically as instances of the predicate at issue is ipso facto to unravel them as so-characterized items.6 The concept of a predicate that is somehow applicable but nevertheless noninstantiable comes to view at this point. We have a predicate F whose realization is noninstantiable because while it is true in abstracto that this property is exemplified⎯that is (∃u)Fu will be true⎯nevertheless the very manner of its specification makes it impossible to identify any particular individual u0 such that Fu0 obtains. Such predicates are “vagrant” in the sense of having no known address or fixed abode: though they indeed have applications these cannot be specifically instanced—they cannot be pinned down and located in a particular spot. Accordingly we may define: F is a vagrant predicate iff (∃x)Fx is true while nevertheless Fx0 is false for each and every specifically identified u0.7

Predicates of this sort will be such that on the basis of general principles one can show that while there must indeed be items to which they apply, nevertheless it can be shown that no such items can ever be concretely and specifically identified.8 While the predicates indeed have application but we are destined to be ignorant about where they apply.9 The following predicates represent properties that are clearly noninstantiable in this way: ⎯being an ever-unstated proposition, or theory, contention, etc.). ⎯being a never-mentioned topic (or idea, object, etc.).


On Certainty and Other Philosophical Essays on Cognition

⎯being a truth (a fact) no one has ever realized (or learned, stated). ⎯being someone whom everyone has forgotten. ⎯being a never-identified culprit. ⎯being an issue no-one has thought about since the 16th century. Noninstantiability itself is certainly not something that is noninstantiable: many instances of it are readily adduced. Vagrant predicates are by nature noninstantiable, but we can nevertheless use them to individuate items that we can never identify. Consider “the oldest unknown (i.e., never-to-be identified) victim of the eruption of Krakatoa. We can clearly make various true claims about the so-individuated person—for example that he-or-she was alive at the time of Krakatoa’s eruption. Reference is no problem here. But by hypothesis cannot manage to identify him. Predicative vagrancy thus reinforces the distinction between mere individuation and actual identification. The existence of vagrant predicates shows that applicability and instantiability do not come to the same thing. By definition, vagrant predicates will be applicable. But this is always something that must be claimed on the basis of general principles, doing so by means of concretely identified instances is, by hypothesis, infeasible. Consider an example of this sort of general-principle demonstration. There are infinitely many positive integers. But the earth has a beginning and end in time. And its overall history has room for only a finite number of intelligent earthlings, each of whom can only make specific mention of a finite number of integers. (They can, of course, refer to the set of integers at large, but they can only specifically take note of some finite number of them.) There will accordingly be some ever-unmentioned, ever unconsidered integers, integers that are individually and explicitly never taken into consideration⎯indeed an infinite number of them. But clearly no-one can give a specific example of this. (The substitutional interpretation of quantifiers will not work with these vagrant predicates.)



Or again consider: ⎯being an unstated proposition Since in the history of the species there can only be a finite number of specifically stated propositions, while actual truths must be infinite in number, we know that there will be some such unstateable. But to say specifically of a particular proposition that it is unstated is impracticable. We can allude to such items generically but cannot actually identify them specifically. In sum, then, vagrant predicates reflect a salient cognitive incapacity where such materials are at issue: we can ascertain that such predicates apply but not where they do so. Ignorance is inevitable here. And once we begin to add to the description of things such cognitive inaccessibility qualifiers as “never specifically contemplated” “never identified,” “ever-overlooked,” and the like, we push them into a cognitive black hole from which there is no exit. The very concept of instantiability/noninstantiability is epistemic in its bearing because all of the relevant procedures⎯exemplifying, illustrating, identifying, naming, and the like⎯are inherently referential by way of purporting a knowledge of identity. And since all such referential processes are mind-projected—and cannot but be so—they are epistemic in nature. Accordingly the idea of knowledge is unavoidably present throughout the phenomenon of predicative vagrancy, seeing that the factor of ignorance is essential here. And so, two features are salient with regard to vagrant predicates: (1) cognition (i.e., what can and cannot be known), and (2) reflexivity (reference to oneself at the individual level (I) or the communal level (us). Vagrant predicates always involve an ignorance of specifics. And this inevitable ignorance is just another aspect of our inevitable ignorance regarding our ignorance itself. The theme of cognitive reflexivity inexorably carries that of cognitive limitedness in its wake. References Bealls, J. C. [2000]: “Fitch’s Proof, Verificationism and the Knower Paradox,” Australasian Journal of Philosophy, 78(2), 241-247. 71

On Certainty and Other Philosophical Essays on Cognition

Brogaard, B., and J. Salerno [2002]: “Fitch’s Paradox of Knowability,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy.Stanford: Stanford University. Cited in http://plato/stanford.edu/cite.html. Cozzo, C. [1994]: “What Can We Learn from the Paradox of Knowability?” Topoi, 13(2), 71-78. Edgington, Dorothy [1985]: “The paradox of knowability,” Mind 93, 557-568. Fara, Michael [2010]: “Knowability and the capacity to know,” Synthese 173, 53-74 Grim, Patrick [1984]: “There is No Set of All Truths,” Analysis 44, 206-208. Hand, M. [2008]: “Performance and Paradox,” in New Essays on the Knowability Paradox. Oxford: Oxford University Press. Jardine, Nick [1980]: “The possibility of absolutism,” pp. 23-42 in Science, Belief, and Behaviour: Essays in Honour of R. B. Braithwaite. D. H. Mellor (ed). Cambridge University Press. Johnstone, Mark [1997]: “Manifest Kinds,” Journal of Philosophy 94, 564-583. Kvanig, J. [1995]: “The Knowability Paradox and the Prospects for Anti-Realism,” Noûs, 29(4), 4810500. Kvanig, J. [2006]: The Knowability Paradox. Oxford: Clarendon Press. Martin-Löf, P. [1998]: “Truth and Knowability: On Principles C and K of Michael Dummett,” in G. Dales and G. Oliveri (eds.), Truth in Mathematics (pp. 105-114). New York: Oxford University Press. Montague, R. and D. Kaplan [1960]: “A Paradox Regained,” Norte Dame Journal of Formal Logic, 1, 79-90. Percival, P. [1990]: “Fitch and Intuitionistic Knowability,” Analysis, 50 182-187. Rescher, Nicholas [2005]: Epistemic Logic. Pittsburgh: University of Pittsburgh Press. Rescher, Nicholas [2009]: Unknowability. Lanham, MD: Lexington Books. Salerno, Joe [2008]: “Knowability Noir,” in New Essays on the Knowability Paradox. Oxford: Oxford University Press.



Salerno, Joe [2010]: “Introduction to knowability and beyond,” Synthese 173, 1-8. Williamson, T. [1987]: “On the Paradox of Knowability,” Mind, 96, 256-261. NOTES 1

Standard logical notation apart, we shall use the variables x, y, z, etc. for knowers, and t, t′, t″, etc. for truths. Kxt abbreviates “X knows t.”


Selfcontradiction follows via theses (KV) and (KC) stipulated in the Appendix.


With ξ we have Kξp ≡ p so that ~Kξ~p is simply ~~p or equivalently p.


For variations on the theme of this chapter see the references listed above.


We have the true K(∃x)Fx but certainly not the false (∃x)KFx. So the former obtains despite ~(∃x)KFx.


We can, of course, refer to such individuals and even to some extent describe them. But what we cannot do is to identify them in the sense specified above.


Accordingly we have K(∃x)Fx but not (∃x)KFx, a situation already contemplated in chapter 1.


A uniquely characterizing description on the order of “the tallest person in the room” will single out a particular individual without specifically identifying him.


The classic Paradox of the Heap (Sorites) affords an illustration. We know abstractly there is some number n as of which collected sandgrains come to constitute a heap, but there is no specifiable number n of which we can say that this begins to be so.




he stage for the present deliberations was set by two transcendental questions put on the philosophical agenda in the seventeenth century by G. W. Leibniz. I.

Why is there anything at all in the world?

II. Why are things as they are in the world? These questions can in turn be elaborated into four: • How is it that there are actual states of affairs at all? • How is it that there are laws of nature? • Why do the actual states of affairs have the character they do? • Why do the laws of nature have the character they do? Clearly, these questions pose issues of fundamental importance for any adequate philosophy of nature.1 In addressing these ultimate questions, there are basically five options. We can (1) seek to resolve them on empiricistic and scientific grounds, (2) reject the questions themselves as meaningless and inappropriate, or (3) take the answer to be lying in super-natural and specifically theological considerations, or (4) abandon the issue as lying beyond the reach of our limited capacities and capabilities. Finally, there is (5) the prospect of resolving them by an appeal to somehow nature-transcending meta-physical facts. Accordingly, there are five salient alternatives:

On Certainty and Other Philosophical Essays on Cognition

• The scientistic course of an exclusive reliance on natural science • The positivistic and rejectionist course of issue-dismissal • the theological course of supra-natural explanation • the mystificational course of relegation to the realm of emplacement beyond the reach of human comprehension. • the metaphysical course of a science-transcending but still broadly “natural,” non-theological mode of explanation Which way is one to turn here? 2. PROBLEMS OF SCIENTISM Consider the decidedly problematic prejudgment that it should be our stance towards questions that “If science cannot answer them, they’re hopeless: science is our only resource here.” Clearly such blatant scientism is decidedly problematic. Natural science is our venture in securing observational knowledge about the world. And in general it explains states of affairs developmentally. In order to account for an aspect of the world’s arrangements, it looks to some earlier, antecedent condition of things, and then invokes the laws of nature to explain on this basis how those subsequent conditions have come to be realized. The scientific format of naturalistic explanation is thus that of a developmental transit of states of affairs under the operation of nature’s laws. However, it simply will not do to argue—as some recent cosmologists have done—that things exist, and exist in the way they do— because the laws of physics require it. For this just does not really accomplish the necessary job, seeing that it immediately raises the question of why those laws are as they do. The law-based explanation of reality would be much like that of the Indian sage who explained the earth’s place in space by maintaining that it was emplaced on the back



of a large elephant who in turn stood on the shell of an enormous turtle. But it is, or should be, clear on the very surface of it that this state of affairs means that natural science will in the final analysis be unable to resolve those matters of ultimate explanation. It is, after all, the facts of science which themselves are at issue here. For those “ultimate questions” are to be seen as transcendental because they do not, by their very nature, admit of explanation on a naturalistic basis. Seeing that the very materials of scientific explanation themselves are at issue—this standard line of procedure is simply unavailable because such explanations require as inputs the very materials whose explanation is in question. But where would the circumstance that those big transcendental questions outrun the limits of naturalistic understanding leave us in the endeavor to get a cognitive grip on the issues? What other options are available to us when confronted with such scientific insolubilia? 3. REJECTING POSITIVISTIC REJECTIONISM There is, to be sure, the option of dismissing these questions as improper and illegitimate. But such rejectionism is not a particularly appealing course. For any alternative to it has the significant merit of retaining for rational inquiry and investigation some clearly intriguing questions that would otherwise be lost. The question of “the reason why” behind existence is strikingly challenging and surely important. If there is any possibility of getting an adequate answer—by hook or by crook—it seems obvious that we would very much like to have it. Its nonstandard nature makes this difficult, but that is surely not a very satisfactory ground for abandoning the problem. There is, after all, nothing patently meaningless about this “riddle of existence.” Nor does it seem to rest in any obvious way on any particularly problematic presupposition—apart from the epistemically optimistic yet methodologically inevitable idea that there are always reasons why things are as they are (the “principle of sufficient reason”). To dismiss such a question as improper or illegitimate is implausible, fruitless, and embarrassingly reminiscent of Aesop’s “Fox and


On Certainty and Other Philosophical Essays on Cognition

Grapes” story. Try as we will to put the question away, it comes back to haunt us.2 4. PROBLEMS OF THEOLOGICAL SUPERNATURALISM In the Western tradition of thought, the theological approach to those ultimate questions has always stood at the forefront. The doctrine is that things are what and as they are because a creator God chose to have it that way. However, the difficulty with this position is that where the ways of God are concerned it is so difficult to get there from here. Moreover, relying upon recourse God to resolve our philosophical problems always seems to be somewhat overreaching. Medieval theorists of the discussion of labor had a need for it: non in philosophia recurrere est ad deum. (Roughly: “Don’t call on God to pull your philosophical Chestnuts out of the fire”) If there is any prospect of cracking this nut without record to such heavy machinery, one ought to be willing to explore it. 5. THE IMPALATABILITY OF MYSTIFICATION Closely allied to dismissalism is the course of abandonment or the issue as something that reaches beyond the limited range our human powers and capacities. On this prospective we frail humans just do not have enough smarts—enough cognitive capacity and power to work out a satisfactory answer. This response is akin to dismissalism, that time out on grounds of putative defects is the question but now on grounds of inadequacy of the questioner. But this tactic is no better than an ultimate default—a recourse to be taken only after all other possibilities have failed us. 6. A SUPRA-NATURAL APPROACH: THE IDEA PROTOLAWS AND THE METAPHYSICAL TURN


Fortunately, yet another option is available. For a variant strategy of resolution can look not to the scientific realities but to the metaphysical principles that underpin our efforts at the cognitive domesti-



cation of the observational realm. Such deliberations will move in the sphere of the theoretically possible rather than that of factually actual. But how can an explanation ever move from mere possibility to actuality by comprehensible means? Pursue this line it becomes necessary to undertake a brief excursus into the theory of possibility. The realm of the possible can be represented as a region that is divided, archery-target fashion, into concentric rings. Moving inwards we encounter first (outermost) the domain of mere logical (or “purely hypothetical”) possibility, and last (centrally) the domain of a physical possibility that reflects the mode of operation of the actual things of this world. Intermediate between them lies the realm of real (or “metaphysical”) possibility.3 That initial region of mere possibility is a matter of abstract, logical consistency—of purely theoretical prospects. And physical possibility reflects how things actually present on the world’s existential stage comport themselves subject to the laws of nature. However, metaphysical possibility is something intermediate between these two: a matter of genuine or “realistic” possibility, not in the sense of psychological imaginability, but in a way that must eventually be cashed out with respect to metaphysical proto-norms that are not laws OF the constitution of nature, but laws FOR the constitution of nature. Such protophysical laws should be understood as laying down conditions of real possibility, ruling certain theoretical (logical) possibilities out as outside the realm of realizability. They “precede” nature and delineate among all the abstractly available possibilities certain ones as alone practicable or “real,” ruling out the rest as unreal, remote, merely hypothetical or the like. 7. NOOPHELIA IS THE CRUX The root idea of such an approach goes back to Leibniz, who effectively envisioned real possibility in terms rooted in axiological considerations of evaluative optimality. To be sure, it is clear that one cannot just optimize, any more than one can just maximize or minimize. For one has to optimize something, some feature or aspect of things. And if this factor is to be something that is qualified, be accepted as self-validating and self-sustaining, then the clearly most


On Certainty and Other Philosophical Essays on Cognition

promising candidate would seem to be intelligence itself—that is to say the overall status and standing of intelligent rational beings at large. Such a philosophical perspective has it that intelligence and its concomitant rationality best qualifies as the controlling standard of value at issue with “for the best” optimality. The enhancement and diffusion of intelligence in the cosmos will be the crux here. After all, a rational being is bound to see the loss of reason as a supreme tragedy. The value at issue here with “being for the best” is a matter of being so as intelligent creatures see it—that is from the vantage point of intelligence itself. Assuredly, no intelligent being would prefer an alternative that is inferior in this regard. And so, for an intelligent being—a rational creature—intelligence itself must figure high on the scale of values. Against this background, the metaphysical approach envisioned here is directed at the optimal conditions of existence for intelligent beings at large. And at the cosmological level such an optimalism will militate towards a universe which • provides for the chance and randomness through which alone intelligent beings can emerge in the world through evolutionary processes based on chance-conditioned variation and selection. • provides for the chance-conditional novelty and innovation needed to provide an environment of sufficient complexity to be of interest for intelligent beings. • provides for the order of regularity and lawfulness needed for a universe sufficiently orderly and to allow complex creatures to develop and thrive. • provides for a lawful order in the modus operandi of nature sufficiently simple to be understood by imperfectly intelligent beings as a basis for grounding their decisions and actions in a complex world.



The arrangements of such an intelligence-friendly (“nophelic”) universe must, in sum, manage things in a way that the emergence of intelligent and rational creatures is provided for. And so an optimal world, in the metaphysical sense presently at issue, is one that achieves a condition of optimalization under constraints—these constraints being a manifold of natural law favorable to the best interests of intelligence—that is, of intelligent beings at large. Unlike a theological supernaturalism which looks to divine intervention in the ordinary course of nature, such an axiological optimalism represents a merely supra-natural—and not super-natural— position: it looks to what lies behind nature not to what lies beyond it. It makes no explicit appeal to a native transcending diversity. All that it does is to explain nature’s features in terms of an axiological protolaw that possibilizes and probabilifies arrangements conducive to the success of intelligent beings in the world (which is not quite to say that such success is guaranteed). An important terminological distinction is at work here, Supernaturalism accepts an agent or agency that affects or effects the ordinary course of nature and shapes its operations. Supra-naturalism, by contrast, leaves all the ordinary forces of nature in place and comes into play only where they leave certain matters open and unresolved, providing resolutions where these forces would otherwise leave matters open and undecided. 8. THE RATIONALE OF OPTIMALISM VALIDATION AND THE REGRESS PROBLEM However, there now looms the objection: “What can possibly explain such an axiological proto-law regarding the manifold of possibility? How could the existential impact of value and intelligence-geared optimality possibly be fitted out with a justifactory rationale?” The answer to this seeming conundrum is straightforward. If optimalism is our proto-law, then it will and must be self-sustaining. And indeed the optimality principle indeed has this feature of self support. For the obviously appropriate answer to the question “But why is it that such a principle of optimality obtains” is simply “The principle obtains because that is for the best.” The principle is self-sustaining


On Certainty and Other Philosophical Essays on Cognition

and self-explanatory. And in this present cast this is not vitiating circularity but an indispensable feature of the problem-situation—a decidedly virtuous circularity. So mandates the axiological program. To be sure, the principle of optimalism represents a contingent feature of things. And as such itself requires explanation seeing that one can and should maintain the Leibnizian Principle of Sufficient Reason to the effect that for every contingent fact there is a reason why it is so rather than otherwise. But there is no decisive reason why that further explanation has to be “deeper and different”—that is, no decisive reason why the prospect of self-explanation has to be excluded at this fundamental level.4 To be sure, the question “Why optimalism?” splits into two decidedly distinct parts, namely (1) Why does optimalism obtain?, and (2) Why is it that we should accept optimalism’s obtaining? These issues are, of course, every bit as distinct as “Why did Booth assassinate Lincoln?” and “Why should we accept that Booth assassinated Lincoln?” The former question seeks an explanation for the existence of a fact, the latter asks for the evidentiation of a judgment. The answer to the first question has just been noted. Optimalism obtains because it is self-potentiating. It is the case that what is for the best obtains because this itself is for the best. Optimalism, in sum, obtains on its own self-sufficient footing. For if we press the questions, “Why should what is for the best exist?” then the answer lies in the very nature of the principle itself. It is self-substantiating, seeing it is automatically for the best that the best alternative should exist rather than an inferior rival. Value is, or can be, an explanatory terminus: it can be regress stopping and “final” by way of self-explanation in a way that causality or purposive can never manage to be. Granted, optimalistic explanation is of a decidedly nonstandard sort. But the crucial consideration here is that ultimate questions are themselves altogether extraordinary. Usually when we ask about things and their conditions we are after a developmental account— how they got to be so given a process of transformation from some earlier condition. But this standard sort of issue-resolution is clearly impossible in the present case. The fact of it is that when we ask an extraordinary question we must be prepared for an extraordinary and indeed seemingly bizarre answer. For if an altogether basic condition



of things is to be explained this cannot be done on the basis of the machinations of some sort thing. Rather, it must be done on the basis of some sort of supra-natural principle and here a metaphysical optimalism does the task. After all, the requisite principle of ultimate explanation cannot be one that relates to actualities. This would simply carry us back to the starting point. Rather, it must relate to possibilities. It cannot deal in actually extant facts but must deal merely in what is potentially possible. And since it cannot relate to facts at all, it must relate to values. And on this basis it confronts us with the problem of factoring a transit from values to fact grounded in evaluative considerations. But such a transit is possible only with an axio-metaphysical principle of the form: Among the various alternative possibilities, it is X that qualifies as real (actual, existent) became this is for the best. Only such a principle of optimality can achieve the requisite work in relation to those ultimate questions. So that we must accordingly take the line that: the existing order of things is what it is and as it is because this is for the best: it represents the best-possible constitution that is encompassed within the overall manifold of possibility. The essential creativity of a possibility-to-actuality principle can thus be achieved along this axiological route which has it that optimality is able to provide for the needed transit from possibility to actuality. And such then is what optimalism sets out to do. So much for explaining how it is that optimalism might obtain. But why should we accept it—what sorts of substantiation could it have? Based on the realization that, since rational explanation must proceed from premises, the principle of anangkê steinai—of the need for an ultimate stopping place of self-sufficiency—has been acknowledged since classical antiquity.5 And since this sort of thing is somewhere between difficult and impossible to encounter in the sphere of fact, it seems plausible to seek it in the sphere of value. The shift from natural to the axiological order of explanation is clearly inviting. The explanation of facts cannot forever lie in the factual domain itself, it must ultimately run out into the domain of virtue. After all, if there is to be anything worthy of the name of an ultimate explanation, there is just no alternative to its being self-


On Certainty and Other Philosophical Essays on Cognition

sustaining and self-validating. And it is, for this very reason, inexorably bound to be an exceptional and altogether extra-ordinary one. For this reason self-validation is not viciously circular, but, rather, is circular in a profoundly virtuous way. After all, it lies in the very nature of a genuinely ultimate explanation that it must not require anything else for its own validation. We have here an account that reaches outside the range of scientific naturalism to a conception of nature whose evidential rationale lies not in observation but in a systematization with human experience at large. In the end, then, we must come to terms with the fact that optimalism is self-explanatory, and obtains because this itself is for the best?6 To ask for a different sort of explanation would be inappropriate. We must expect that any ultimate principle should explain itself and cannot, in the very nature of things, admit of an external explanation in terms of something altogether different. And the impetus to realization inherent in authentic value lies in the very nature of value itself. A rational person would not favor the inferior alternative; and there is no reason to think that a rational reality would do so either. So in the end what recommends acceptance is simply that it works. 9. ON RELATING FACT AND VALUE But is it practicable and legitimate to shift the domain basis of our explanations and answer a basically factual question about the nature of things by a recourse into issues of value? There is good reason to think that in the final analysis we not only can but must do so thanks to the inherent limits of naturalistic reason. The crux of the matter here goes back to the very dawn of philosophy in the days of the early Greeks. It is adumbrated in their classical distinction between physis (the realm of physical reality) and nomos (the realm of normative prescription) between what is and what ought to be—the Platonic distinction between what is actual and accessible to the eyes of the body and what is ideal and accessible only to the mind’s eye. In this light we are brought back to the neo-Platonism of Kant with its crucial distinction between the “theoretical” domain of observable fact and the “practical” domain of judgmental evaluation. Now the



______________________________________________________ Display 1 INSTRUCTIVE ANALOGIES NORMATIVE CONSTRAINTS


1. The generic grammar of discourse

The verbal make-up of particular declarations

2. The defining rules of chess

The unfolding of actual chess games

3. The laws of nature

The occurant phenomenon

______________________________________________________ crucial fact in this regard is that man is an amphibious creature who lives in both of these realms, that of observation and that of evaluation—that of observed factuality and that of assessed normativity. And it is this fundamental fact of life that accounts—easily and naturally— for the limits of natural science. For natural science by its very nature is an enterprise with a mission—an inherent telos, namely the explanation of nature’s phenomena as observation reveals them to us. The sphere of the normative—not of what is, but of what ought to be—is something that lies entirely outside this sense-accessible domain. And this sets an inherent and insurmountable limit. But of course this limit no more constitutes a limitation—an inherent defect—then it constitutes a limitation on agronomy that it does not deal with issues of philately. The two domains—scientific descriptive and normatively prescriptive simply have different regions of concern—different jurisdictions to employ one of Kant’s favorite metaphors. And it is this simple but fundamental fact that sets insuperable limits to scientific/naturalistic reason. Granted, it might still seem strange to say that while science cannot settle issues of value as such, nevertheless values can play a determinative role in relation to the scientific issues. But experience teaches that there are multiple instances where normative constraints serve to canalize and delimit factual results. Consider some analogies here as per the illustrative instances of Display 1. In each case the norms ca-


On Certainty and Other Philosophical Essays on Cognition

nalize and delimit the range of facts while nevertheless those facts fail to determine the range of norms at issue. 10. CONCLUSION In closing, let us return to the starting point. Given the orientation of our present approach, the answer to those ultimate questions about the nature of things come to be seen as rooted in evaluative considerations. It is thus clear that we are here led beyond the limits of naturalist explanation. For irrespective of whether this explanatory project is conducted along metaphysical or theological lines, it is unavoidable that the explanation of nature’s nature is an issue that outruns the reach of ordinary naturalism. And so the resulting situation seems clear. While the scientific facts regarding nature do not settle any evaluative issues, there is nevertheless no reason of fundamental principle why evaluative considerations cannot in theory play a role in delimiting and canalizing constraining (at least some of) the scientific facts. The make-up of reality—its specific constitution as contrasted with alternatives—could thus emerge from a principle of value that underlies the laws and accounts for their salient aspects. And while the normative principles can put ontological constrains upon the facts, the facts themselves will afford no more than epistemically suggestive indications regarding the norms and values at work. But in the final analysis the ultimate validation of such a nonstandard approach is rooted in functionally pragmatic rather than epistemic evidential considerations. We accept them not because of an evidentiating “you must” but because if we have questions, seek for answers and find—faute de meiux—that no rationally superior line of response is available to us. NOTES 1


To be sure, these questions relate to contingent existence. The theological issue of a necessary being remains outside the purview of present discussion. The literature of this topic is vast. An instructive recent discussion is Michael Heller, Ultimate Explanations of the Universe (Berlin: Springer, 2009). The present author’s positions was articulated in The Riddle of Existence (Lanham, MD: University Press of America, 1984).



For criticisms of ways of avoiding the question “Why is there something rather than nothing?” see Chap. III of William Rowe, The Cosmological Argument (Princeton, 1975). Cf. also Donald R. Burrill (ed.), The Cosmological Argument (Garden City, 1967), esp. “The Cosmological Argument” by Paul Edwards, as well as Adolf Grünbaum, “The Pseudo-Problem of Creation in Physical Cosmology,” Philosophy of Science, vol. 56 (1989), pp. 373-394.


In principle there might be more refinement here, with some actuality-departing possibilities being relatively more proximate or remote than others, depending on how radical the departures from existential reality. This would lead to gradations of more or less “real” possibility, depending how close one comes to the “real life” modus operandi of actual things. We here ignore this prospect and abstract from such complications which do not mind and afford the outcome of the deliberations.


After all, there is no reason of logico-theoretical principle why propositions cannot be self-certifying. Nothing vicious need be involved in self-substantiation. Think of “Some statements are true” or “There are statements that state a particular rather than universal claim.”


See John Michael Shea, “St. Thomas Aquinas on the Principle ‘Anankê stenai’,” New Scholasticism, vol. 55 (1981), pp. 139-158.


Optimalism is closely related to optimism. The optimist holds that “Whatever exists is for the best,” the optimalist maintains the converse that “Whatever is for the best exists.” However, when we are dealing with exclusive and exhaustive alternatives the two theses come to the same thing. When one of the alternatives A, A1, . . . An must be the case, then if what is realized is for the best it follows automatically that the best is realized (and conversely).


REFERENCES Aristotle, Physics. Bealls, J. C., “Fitch’s Proof, Verificationism and the Knower Paradox,” Australasian Journal of Philosophy, vol. 78/2 (2000), pp. 241-247. Brentano, Franz, (1973). Brogaard, B., and J. Salerno, “Fitch’s Paradox of Knowability,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Stanford: Stanford University, 2002). Cited in http://plato/stanford.edu/ cite.html. Clifford, William Kingdon, Lectures and Essays, ed. by L. Stephen and F. Pollock (London: Macmillan, 1879, 2nd ed. 1886; originally published in the Contemporary Review, vol. 30 (1877), pp. 42-54.) Cozzo, C., “What Can We Learn from the Paradox of Knowability?” Topoi, vol. 13/2 (1994), pp. 71-78. Duhem, Pierre Maurice, La théorie physique: son objet, et sa structure (Paris: Chevalier and Rivière, 1906); tr. by Philip P. Wiener, The Aim and Structure of Physical Theory (Princeton, Princeton University Press, 1954.) Edgington, Dorothy, “The paradox of knowability,” Mind, vol. 93 (1985), pp. 557-568. Edwards, Paul, “The Cosmological Argument,” in Donald R. Burrill (ed.), The Cosmological Argument (Garden City, 1967). Fara, Michael, “Knowability and the Capacity to Know,” Synthese, vol. 173 (2010), pp. 53-74. Grice, Paul, Studies in the Way of Words (Cambridge Mass: Harvard University Press, 1989). Grim, Patrick, “There is No Set of All Truths,” Analysis, vol. 44 (1984), pp. 206-208. Grünbaum, Adolf, “The Pseudo-Problem of Creation in Physical Cosmology,” Philosophy of Science, vol. 56 (1989), pp. 373394. Hand, M., “Performance and Paradox,” in New Essays on the Knowability Paradox (Oxford: Oxford University Press, 2008).

On Certainty and Other Philosophical Essays on Cognition

Hartmann, Nicolai, “Zur Methode der Philosophiegeschichte,” in his Kleinere Schriften (Berlin: De Gruyter, 1958). Heller, Michael, Ultimate Explanations of the Universe (Berlin: Springer, 2009). Hülsmann, Heinz, Die Methode in der Philosophie Nicolai Hartmanns (Duሷsseldorf: L. Schwann, 1959). James, William, “The Sentiment of Rationality,” in The Will to Believe and Other Essays in Popular Philosophy (New York: Longmans, Green, and Company, 1897). Jardine, Nick, “The possibility of absolutism,” pp. 23-42 in Science, Belief, and Behaviour: Essays in Honour of R. B. Braithwaite. D. H. Mellor (ed.), (Cambridge University Press, 1980). Johnstone, Mark, “Manifest Kinds,” Journal of Philosophy, vol. 94, (1997), pp. 564-583. Kant, Immanuel, Critique of Judgment, translated J. H. Bernard (London, 1892). Kvanig, J., “The Knowability Paradox and the Prospects for AntiRealism,” Noûs, vol. 29/4 (1995), pp. 481-500. ———, The Knowability Paradox (Oxford: Clarendon Press, 2006). Martin, Gottfried, (1957) Martin-Löf, P., “Truth and Knowability: On Principles C and K of Michael Dummett,” in G. Dales and G. Oliveri (eds.), Truth in Mathematics (New York: Oxford University Press, 1998). Montague, R. and D. Kaplan, “A Paradox Regained,” Norte Dame Journal of Formal Logic, vol. 1 (1960), pp. 79-90. Percival, P., “Fitch and Intuitionistic Knowability,” Analysis, vol. 50 (1990), pp. 182-187. Plato, Topics Rescher, Nicholas, “Quasi-Truth-Functional Systems of Propositional Logical,” in The Journal of Symbolic Logic, vol. 27 (1962), pp. 1-10. ———, The Riddle of Existence (Lanham, MD: University Press of America, 1984). ———, The Strife of Systems (Pittsburgh: University of Pittsburgh Press, 1985). ———, Epistemic Logic (Pittsburgh: University of Pittsburgh Press. 2005). ———, Unknowability (Lanham, MD: Lexington Books, 2009).



———, Aporetics (Pittsburgh: University of Pittsburgh Press, 2009). Rowe, William, The Cosmological Argument (Princeton, 1975) Salerno, Joe, “Knowability Noir,” in New Essays on the Knowability Paradox (Oxford: Oxford University Press, 2008). ———, “Introduction to Knowability and Beyond,” Synthese, vol. 173 (2010), pp. 1-8. Shea, John Michael, “St. Thomas Aquinas on the Principle ‘Anankê stenai’,” New Scholasticism, vol. 55 (1981), pp. 139-158. Siitonen, Arto, Problems of Aporetics (Helsinki: Suomalainen Tiedeakatemia, 1989: Annales Acdemiae Scientiarum Fennicae, Dissertationes Humanarum Litterarum, No. 50). Sosa, Ernest, Knowledge in Perspective (New York: Cambridge University Press, 1991). Weinberg, Stephen, Dreams of a Final Theory (New York: Pantheon Books, 1992. Williamson, T., “On the Paradox of Knowability,” Mind, vol. 96 (1987), pp. 256-261.


Name Index Aesop, 77 Aquinas, St. Thomas, 27 Aristotle, 25, 27-29, 89 Austen, Jane, 23 Bealls, J. C., 71, 89 Bohr, Neils, 17n2 Brentano, Franz, 28, 35n3, 89 Brougaard, Berit 72, 89 Clifford, William Kingdon, 37, 42, 43n1, 89 Cozzo, Cesare, 72, 89 Descartes, René, 57 Duhem, Pierre Maurice, 11, 17n3, 89 Edgington, Dorothy, 72, 89 Edwards, Paul, 87n2, 89 Fara, Michael, 72, 89 Fichte, J. G., 57 Freud, Sigmund, 57 Grice, Paul, 56n2, 89 Grim, Patrick, 72, 89 Grünbuam, Adolf, 87n2, 89 Hand, Michael, 72, 89 Hartmann, Nicolai, 28, 35, 90 Hegel, G. W. F., 57 Heller, Michael, 86n1, 90 Hülsmann, Heinz, 35, 90 Hume, David, 57

Nicholas Rescher • On Certainty and Other Philosophical Essays on Cognition

James, William, 42, 43n3, 90 Jardine, Nicholas, 72, 90 Johnstone, Mark, 72, 90 Kant, Immanuel, 27-29, 31, 35n4, 57, 84-85, 90 Kaplan, David, 90, 73 Kvanig, Jonathan, 72, 90 Leibniz, G. W., 57, 61, 75 Luce, Clare Booth, 24 Martin, Gottfried, 35n3, 90 Martin-Löf, Per, 73, 90 Montague, Richard, 73, 90 Peirce, C. S., 68 Percival, Percival, 73, 90 Plato, 21, 23, 25, 29, 90 Rescher, Nicholas, 35, 73, 90 Rowe, William, 87n2, 91 Salerno, Joe, 72-74, 89, 91 Schopenhauer, Arthur, 57 Scotus, John Dins, 61 Shakespeare, William, 23 Shea, John Michael, 87n5, 91 Siitonen, Arto, 35, 91 Socrates, 29 Sosa, Ernest, 43n2, 91 Tarski, Alfred, 46 Unamuno, Miguel de, 57 Weinberg, Stephen, 17n2, 91 Williamson, Timothy, 74, 91



Nicholas Rescher, Patrick Grim

Beyond Sets A Venture in Collection-Theoretic Revisionism This book is the product of a collaboration stretching the years 2007-10, whose initial fruit was a paper on “Plenum Theory” published in Nous. The work grew out of the author’s conviction that standard set theory, which had evolved to meet the needs of mathematics, was not fully adequate to the less abstractly geared and rigidly determine needs of less finalized ranges of inquiry and deliberation. About the Author Nicholas Rescher is University Professor of Philosophy at the University of Pittsburgh where he also served for many years as Director of the Center for Philosophy of Science. He is a former president of the Eastern Division of the American Philosophical Association, and has also served as President of the American Catholic Philosophical Association, the American Metaphysical Society, the American G. W. Leibniz Society, and the C. S. Peirce Society. An honorary member of Corpus Christi College, Oxford, he has been elected to membership in the European Academy of Arts and Sciences (Academia Europaea), the Institut International de Philosophie, and several other learned academies. Having held visiting lectureships at Oxford, Constance, Salamanca, Munich, and Marburg, Professor Rescher has received six honorary degrees from universities on three continents. Author of some hundred books ranging over many areas of philosophy, over a dozen of them translated into other languages, he was awarded the Alexander von Humboldt Prize for Humanistic Scholarship in 1984. In November 2007 Nicholas Rescher was awarded by the American Catholic Philosophical Association with the „Aquinas Medal“. Patrick Grim is Distinguished Teaching Professor of Philosophy at Stony Brook University. He is author of The Incomplete Universe, co-author of The Philosophical Computer, editor of Mind and Consciousness: 5 Questions, and founding co-editor of over 25 volumes of the Philosopher's Annual. Grim has produced two lecture series with the Teaching Company: Questions of Value and Philosophy of Mind. He has published widely in philosophy but also in a variety of other fields: theoretical biology, computer science, artificial life, linguistics, decision and game theory.

ontos verlag

Frankfurt • Paris • Lancaster • New Brunswick 2011. 112pp. Format 14,8 x 21 cm Hardcover EUR 59,00 ISBN 978-3-86838-100-9 Due December 2010 Please order free review copy from the publisher Order form on the next page

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34


Nicholas Rescher

Free Will An Extensive Bibliography With the Cooperation of Estelle Burris

Few philosophical issues have had as long and elaborate a history as the problem of free will, which has been contested at every stage of the history of the subject. The present work practices an extensive bibliography of this elaborate literature, listing some five thousand items ranging from classical antiquity to the present.

About the author Nicholas Rescher is University Professor of Philosophy at the University of Pittsburgh where he also served for many years as Director of the Center for Philosophy of Science. He is a former president of the Eastern Division of the American Philosophical Association, and has also served as President of the American Catholic Philosophical Association, the Americna Metaphysical Society, the American G. W. Leibniz Society, and the C. S. Peirce Society. An honorary member of Corpus Christi College, Oxford, he has been elected to membership in the European Academy of Arts and Sciences (Academia Europaea), the Institut International de Philosophie, and several other learned academies. Having held visiting lectureships at Oxford, Constance, Salamanca, Munich, and Marburg, Professor Rescher has received six honorary degrees from universities on three continents. Author of some hundred books ranging over many areas of philosophy, over a dozen of them translated into other languages, he was awarded the Alexander von Humboldt Prize for Humanistic Scholarship in 1984. In November 2007 Nicholas Rescher was awarded by the American Catholic Philosophical Association with the „Aquinas Medal“

ontos verlag

Frankfurt • Paris • Lancaster • New Brunswick 2009. 309pp. Format 14,8 x 21 cm Hardcover EUR 119,00 ISBN 13: 978-3-86838-058-3 Due December 2009

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34


Nicholas Rescher

Autobiography Second Edition

This revised edition of his Autobiography brings up-to-date Rescher’s account of his life and work. The passage of years since the publication of an autobiographical work makes for its growing incompleteness. Moreover, the passage of time is bound to bring some new perspectives to view. This new edition comes to terms with these circumstances. Since the publication of the previous version Rescher’s philosophical work has made substantial progress, betokened by the publication of over a score of new books that mark an ongoing expansion of his philosophical range. Then too, the internet has brought to light interesting new information about Rescher’s family background and antecedence. Overall the book affords a detailed, vivid, and highly personalized picture of the life and work of someone who counts as one of the most prolific and many-sided contemporary thinkers.

ontos verlag

Frankfurt • Paris • Lancaster • New Brunswick 2010. 419 Seiten Format 14,8 x 21 cm Paperback EUR 49,00 ISBN 978-3-86838-084-2

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34


Nicholas Rescher

On Rules and Principles A Philosophical Study of their Nature and Function The present book is a natural outgrowth of Rescher’s longstanding preoccupation with the rational systematization of our knowledge as manifested in such earlier works as Cognitive Systematization (Oxford: Blackwell, 1979), and Complexity (New Brunswick: Transaction Publishers, 1998). Accordingly, the role of principles in human affairs is crucial and ubiquitous. Principology, the theory of principles—underdeveloped through it may be—is accordingly bound to find a significant place in the sphere of philosophical inquiry regarding matters of thought and action. About the Author Nicholas Rescher is University Professor of Philosophy at the University of Pittsburgh where he also served for many years as Director of the Center for Philosophy of Science. He is a former president of the Eastern Division of the American Philosophical Association, and has also served as President of the American Catholic Philosophical Association, the American Metaphysical Society, the American G. W. Leibniz Society, and the C. S. Peirce Society. An honorary member of Corpus Christi College, Oxford, he has been elected to membership in the European Academy of Arts and Sciences (Academia Europaea), the Institut International de Philosophie, and several other learned academies. Having held visiting lectureships at Oxford, Constance, Salamanca, Munich, and Marburg, Professor Rescher has received six honorary degrees from universities on three continents. Author of some hundred books ranging over many areas of philosophy, over a dozen of them translated into other languages, he was awarded the Alexander von Humboldt Prize for Humanistic Scholarship in 1984. In November 2007 Nicholas Rescher was awarded by the American Catholic Philosophical Association with the „Aquinas Medal“.

ontos verlag

Frankfurt • Paris • Lancaster • New Brunswick 2010. 220 Seiten Format 14,8 x 21 cm Hardcover EUR 89,00 ISBN 978-3-86838-089-7

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34



Nicholas Rescher

Collected Paper. 14 Volumes Nicholas Rescher is University Professor of Philosophy at the University of Pittsburgh where he also served for many years as Director of the Center for Philosophy of Science. He is a former president of the Eastern Division of the American Philosophical Association, and has also served as President of the American Catholic Philosophical Association, the American Metaphysical Society, the American G. W. Leibniz Society, and the C. S. Peirce Society. An honorary member of Corpus Christi College, Oxford, he has been elected to membership in the European Academy of Arts and Sciences (Academia Europaea), the Institut International de Philosophie, and several other learned academies. Having held visiting lectureships at Oxford, Constance, Salamanca, Munich, and Marburg, Professor Rescher has received seven honorary degrees from universities on three continents (2006 at the University of Helsinki). Author of some hundred books ranging over many areas of philosophy, over a dozen of them translated into other languages, he was awarded the Alexander von Humboldt Prize for Humanistic Scholarship in 1984. ontos verlag has published a series of collected papers of Nicholas Rescher in three parts with altogether fourteen volumes, each of which will contain roughly ten chapters/essays (some new and some previously published in scholarly journals). The fourteen volumes would cover the following range of topics: Volumes I - XIV STUDIES IN 20TH CENTURY PHILOSOPHY ISBN 3-937202-78-1 · 215 pp. Hardcover, EUR 75,00

STUDIES IN VALUE THEORY ISBN 3-938793-03-1 . 176 pp. Hardcover, EUR 79,00

STUDIES IN PRAGMATISM ISBN 3-937202-79-X · 178 pp. Hardcover, EUR 69,00

STUDIES IN METAPHILOSOPHY ISBN 3-938793-04-X . 221 pp. Hardcover, EUR 79,00

STUDIES IN IDEALISM ISBN 3-937202-80-3 · 191 pp. Hardcover, EUR 69,00

STUDIES IN THE HISTORY OF LOGIC ISBN 3-938793-19-8 . 178 pp. Hardcover, EUR 69,00

STUDIES IN PHILOSOPHICAL INQUIRY ISBN 3-937202-81-1 · 206 pp. Hardcover, EUR 79,00

STUDIES IN THE PHILOSOPHY OF SCIENCE ISBN 3-938793-20-1 . 273 pp. Hardcover, EUR 79,00

STUDIES IN COGNITIVE FINITUDE ISBN 3-938793-00-7 . 118 pp. Hardcover, EUR 69,00

STUDIES IN METAPHYSICAL OPTIMALISM ISBN 3-938793-21-X . 96 pp. Hardcover, EUR 49,00

STUDIES IN SOCIAL PHILOSOPHY ISBN 3-938793-01-5 . 195 pp. Hardcover, EUR 79,00

STUDIES IN LEIBNIZ'S COSMOLOGY ISBN 3-938793-22-8 . 229 pp. Hardcover, EUR 69,00

STUDIES IN PHILOSOPHICAL ANTHROPOLOGY ISBN 3-938793-02-3 . 165 pp. Hardcover, EUR 79,00

STUDIES IN EPISTEMOLOGY ISBN 3-938793-23-6 . 180 pp. Hardcover, EUR 69,00

ontos verlag Frankfurt • Paris • Lancaster • New Brunswick 2006. 14 Volumes, Approx. 2630 pages. Format 14,8 x 21 cm Hardcover EUR 798,00 ISBN 10: 3-938793-25-2 Due October 2006 Please order free review copy from the publisher Order form on the next page

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34