Fact, Fiction, and Forecast

Fourth Edition

Nelson Goodman

• Publisher: Harvard University Press
• Year: 1983
• Print ISBN: 9780674290716, 9780674290709
• eBook ISBN: 9780674255210
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Here, in a new edition, is Nelson Goodman's provocative philosophical classic—a book that, according to Science, "raised a storm of controversy" when it was first published in 1954, and one that remains on the front lines of philosophical debate.

How is it that we feel confident in generalizing from experience in some ways but not in others? How are generalizations that are warranted to be distinguished from those that are not? Goodman shows that these questions resist formal solution and his demonstration has been taken by nativists like Chomsky and Fodor as proof that neither scientific induction nor ordinary learning can proceed without an a priori, or innate, ordering of hypotheses.

In his new foreword to this edition, Hilary Putnam forcefully rejects these nativist claims. The controversy surrounding these unsolved problems is as relevant to the psychology of cognitive development as it is to the philosophy of science. No serious student of either discipline can afford to misunderstand Goodman's classic argument.

PROJECT
1953

The three chapters to follow are somewhat revised versions of the Special Lectures in Philosophy delivered at the University of London on May 21, 26 and 28, 1953. The first of the three was repeated at Harvard University on December 2 of the same year. All are now published for the first time. The Introductory Notes to this book (pp. xvii-xxv) contain some remarks on the relationship between these three chapters and the preceding one.

II
THE PASSING OF THE POSSIBLE

1. Foreword: On the Philosophic Conscience

In life our problems often result from our indulgences; in philosophy they derive rather from our abnegations. Yet if life is not worthwhile without its enjoyments, philosophy hardly exists without its restraints. A philosophic problem is a call to provide an adequate explanation in terms of an acceptable basis. If we are ready to tolerate everything as understood, there is nothing left to explain; while if we sourly refuse to take anything, even tentatively, as clear, no explanation can be given. What intrigues us as a problem, and what will satisfy us as a solution, will depend upon the line we draw between what is already clear and what needs to be clarified.

Yet I am afraid that we are nowhere near having any sound general principle for drawing this line. Surely I need not in this place and before this audience recount the tragic history of the verification theory of meaning.¹ The failure of this gallant effort to distinguish sense from nonsense, like the failure of various worthy efforts to codify the difference between right and wrong, has encouraged in some quarters the libertine doctrine that anything goes. The perverse maxim that whatever you can get away with is right has its counterpart in the claim that whatever works is clear. So crude a pragmatism deserves mention only because it seems to be spreading. I may not understand the devices I employ in making useful computations or predictions any more than the housewife understands the car she drives to bring home the groceries. The utility of a notion testifies not to its clarity but rather to the philosophic importance of clarifying it.

In the absence of any convenient and reliable criterion of what is clear, the individual thinker can only search his philosophic conscience. As is the way with consciences, it is elusive, variable, and too easily silenced in the face of hardship or temptation. At best it yields only specific judgments rather than general principles; and honest judgments made at different times or by different persons may differ in any degree. Indeed this talk of conscience is simply a figurative way of disclaiming any idea of justifying these basic judgments. Beyond making them carefully and declaring them loudly, about all we can do is to disparage any alternatives. If your conscience is more liberal than mine, I shall call some of your explanations obscure or metaphysical, while you will dismiss some of my problems as trivial or quixotic.

All this is by way of preface to declaring that some of the things that seem to me inacceptable without explanation are powers or dispositions, counterfactual assertions, entities or experiences that are possible but not actual, neutrinos, angels, devils, and classes. Concerning the last of these, I have had a good deal to say elsewhere,² and I shall not press the point in these lectures. I shall use the language of classes rather freely because means have now been provided for giving a satisfactory interpretation of most ordinary statements about classes, and because I don’t want to grind too many axes at once. Some other items on my list—angels and devils—enter so little into my daily discourse or into scientific discussions that I can wait patiently a long time for them to be explained. As for neutrinos and some other particles of physics, I think they are as yet beyond our philosophic reach. But the interrelated problems of dispositions, counterfactuals, and possibles are among the most urgent and most pervasive that confront us today in the theory of knowledge and the philosophy of science. It is this cluster of problems that I want to discuss in these lectures.

My sample listing of suspect notions is of course far from complete. Some of my other prejudices will be revealed by what I abjure in seeking a solution to the problems mentioned. For example, I shall not rely on the distinction between causal connections and accidental correlations, or on the distinction between essential and artificial kinds, or on the distinction between analytic and synthetic statements. You may decry some of these scruples and protest that there are more things in heaven and earth than are dreamt of in my philosophy. I am concerned, rather, that there should not be more things dreamt of in my philosophy than there are in heaven or earth.

Today let us examine briefly first the problem of counterfactuals, second the problem of dispositions, and finally the problem of possible entities. The reasons for this ordering will become apparent as we proceed.

2. Counterfactuals

A common habit of speech, a recent trend in philosophy, and the apparent ease of expressing in counterfactual form what we want to say about dispositions and possible entities make it natural to begin with the problem of counterfactual conditionals. Nowadays I think few of us are any longer willing to accept a counterfactual conditional, however impressively intoned, as providing in itself an explanation that requires no further analysis. The legal mind investigating the question what is meant by the value of real estate may rest content with the pronouncement that the value is the price the property would bring if it were sold by a willing seller to a willing buyer; but the philosopher (at least I) will regard this as reframing the question rather than answering it.

Nevertheless, replacement of a statement like

k was flexible at time t

by a statement like

If k had been under suitable pressure at time t, then k would have bent

has obvious promise as a step towards clarification. The disposition-term “flexible” is eliminated without the introduction of any such troublesome word as “possible”; only non-dispositional predicates appear to remain, even if they are slightly jaundiced with a modal inflection. Moreover, the counterfactual formulation seems already to effect at least a preliminary analysis, since a conditional is made up of simpler statements. Indeed, if we interpret the counterfactual conditional as saying

If the statement “k was under suitable pressure at time i” were true, then the statement “k bent at time t” would be true,

the modality is removed from the predicates and we may focus attention upon the relationship affirmed to hold between two simple indicative statements. By thus moving to the plane of relations between statements, we feel that we have exchanged an ontological problem for a linguistic one.³ Also, we half-consciously expect that the truth-functional treatment of ordinary indicative conditionals will somehow serve as a helpful model for the analysis of counterfactuals. All these factors, I think—together with the prospect of acquiring at a single stroke the means for dealing with a whole tangle of problems—have contributed to a notable quickening of philosophical interest in the counterfactual conditional during the past few years.

Plainly, the truth-value of a counterfactual does not derive simply from the truth-value of its components; for since the antecedent and consequent of every counterfactual are both false,⁴ all counterfactuals will have the same truth-value by any truth-functional criterion. Counter- factual connection must be defined in some quite different way. Some philosophers, of course, prefer to regard counterfactuals as rules or licenses for making inferences rather than as statements that are true or false. But whether we are seeking to distinguish true from false statements or distinguish valid from invalid licenses, the task is to discover the necessary and sufficient conditions under which counterfactual coupling of antecedent and consequent is warranted.

The relationship between the component statements of a true counterfactual is seldom a matter of logical implication. The statement

Match m lit at time t

does not follow by any familiar logical principle from the statement

Match m was scratched at time t;

there is an appeal to a general physical principle about matches. But two difficulties arise.

In the first place, matches do not always light when scratched. They light only if attendant circumstances are propitious. Let us, for easy reference, give the name “S” to the counterfactual statement

If m had been scratched at t, then m would have lit.

S does not merely affirm that if the circumstances had been propitious then the match would have lit; S affirms that the circumstances were propitious. A counterfactual is true if and only if the antecedent conjoined with relevant true statements about the attendant circumstances leads by way of a true general principle to the consequent. But what statements are relevant? Surely not, in the case of S, all true statements about m at t; for some of these (e.g “m was not scratched at i”, and “m did not light at i”) are in-compatible with the antecedent or with the consequent. We soon find that other exclusions are needed; and after a long series of failures to arrive at a competent formula that is not itself counterfactual and therefore question-begging, we come to recognize that this aspect of the problem is very troublesome.⁵

In the second place, not every true general principle is capable of sustaining a counterfactual conditional. It is true that every person now in this room is safe from freezing. It is also true that every person now in this room is English-speaking. Now consider a certain Eskimo who is at this moment nearly frozen to death somewhere in the Arctic. If he were now in this room he would be safe from freezing, but he would not be English-speaking. What makes the difference? We may say that the generalization about safety from freezing expresses a causal relationship or follows from a law, while the generalization about knowledge of English is only contingently or accidentally true; but to define this distinction is a delicate matter. Since we shall soon encounter the problem again, I shall go into no details at the moment; but this second aspect of the problem of counterfactuals, like the first, is formidable enough to have defied many intensive efforts to solve it.

These difficulties and the unsullied record of frustration in attempting to meet them have pretty thoroughly deflated our initial hope of finding a relatively easy approach to our problems through the study of the counter- factual conditional. We are still a very long way from having a solution to the problem of counterfactuals;⁶ and by this time we may be ready to try another tack. After a number of years of beating our heads against the same wall and of chasing eagerly up the same blind alleys, we may welcome a change in strategy if only for its psychological benefits. But I think there are at least two better reasons for turning our attention for a while to the problem of dispositions.

First, in dealing with counterfactuals we are looking less at what is said than at the way it is said. We are expressly concerning ourselves with a form of statement; and the pattern of analysis we seek is largely dictated by the structure of the conditional. This structure, although it promised at the outset to be a valuable aid, may actually have become a hindrance. The very disanalysis effected by returning to consider dispositional statements, which are indicative and simple in form, may free us to explore a better scheme of analysis.

Second, I suspect that the problem of dispositions is really simpler than the problem of counterfactuals. This may sound strange in view of the apparent full convertibility between dispositional and counterfactual statements; but it turns out that ordinary dispositional statements often correspond to abnormally weak counterfactuals. Suppose that ω is a piece of dry wood during a given brief period of time. We commonly suppose that a statement like

ω is inflammable

amounts to some such normal counterfactual as

If ω had been heated enough, it would have burned.

Once we look more closely, however, we can readily describe circumstances—for example, a lack of oxygen near ω—under which the dispositional statement is true and the counterfactual false. For a translation guilty of no discrepancies like this we should be forced back to some such fainthearted counterfactual as

If all conditions had been propitious and ω had been heated enough, it would have burned.

To speak very loosely, the dispositional statement says something exclusively about the ‘internal state’ of ω, while our original counterfactual says in addition something about the surrounding circumstances; but the important point is that the dispositional statement is the weaker. And in the margin of difference may lie some of the obstacles that have blocked our way so far.

These, then, are some of the reasons for dropping the problem of counterfactuals for a time and seeing what can be done about the problem of dispositions; but I am by no means suggesting that this reorientation solves anything by itself or opens any royal road to progress.

3. Dispositions

Besides the observable properties it exhibits and the actual processes it undergoes, a thing is full of threats and promises. The dispositions or capacities of a thing—its flexibility, its inflammability, its solubility—are no less important to us than its overt behavior, but they strike us by comparison as rather ethereal. And so we are moved to inquire whether we can bring them down to earth; whether, that is, we can explain disposition-terms without any reference to occult powers.

Perhaps we ought to notice at the very beginning that more predicates than we sometimes suppose are dispositional. A tell-tale suffix like “ible” or “able” is not always present. To say that a thing is hard, quite as much as to say that it is flexible, is to make a...

Table of contents

• COVER
• TITLE
• COPYRIGHT
• CONTENTS
• FOREWORD TO THE FOURTH EDITION
• INTRODUCTORY NOTES
• PREDICAMENT 1946
• PROJECT 1953
• INDEX