Because Professor Stuart Hampshire of Princeton is a philosophical analyst who can write very well, who is learned without being pompous, and civilized without being an intellectual snob, he represents something extremely valuable in philosophy today. Hampshire has written on fiction, on politics, and on contrary-to-fact conditionals; he has expounded Spinoza brilliantly for Penguin Books; he has edited a Mentor paperback on the Age of Reason. And unlike so many of his contemporaries among British and American philosophers, Hampshire is neither a manqué mathematician nor a pedant, as his book Thought and Action clearly shows. He is, to his credit, a humanist who argues for his views, and he can do so effectively in a variety of situations—from All Souls to Princeton, and from the pages of the Oxford Mind, to those of London’s New Statesman. Hampshire is, I dare say, American philosophy’s most versatile English acquisition since Whitehead.

Having said this, however, I must add that this book is not in itself a sample of Hampshire’s versatility, nor, in spite of its title, an expression of his interest in politics. Furthermore, it is not so well-written as one might have expected, for it is quite obscure at many places. True, it is characterized by a good deal of the orderliness that one associates with analytic philosophy. There is much talk about what is meant by statements and what they entail; there are distinctions between what is normative and what is not; there are many confident assertions about what is and what is not a necessary truth. But over the whole tough, formidable framework there seems to hang a kind of elegant mist that often prevents us from understanding the author’s intentions. At times therefore one may feel frustrated while reading the book, but if one sympathetically recognizes how difficult Hampshire’s topic is, one will think twice and read it more than once. I approached the book in the second way and will report some of my findings and reactions, with the distinct fear that I may have misunderstood Hampshire and that I may be oversimplifying his views.

As a product of Oxford in the last generation Hampshire is of course a linguistic philosopher. By this I mean that he believes that the careful study of language can help us clarify, solve, or dissolve philosophical problems, both old and new. And like most Oxford philosophers of his generation Hampshire is deeply affected by the idea of Wittgenstein that language has many uses, only one of which is the scientific use of describing or stating facts. The idea is a banal one from one point of view, since every schoolboy knows—or used to know—that there are interrogative and imperative sentences as well as declarative sentences, and that we do not describe or state facts when we ask questions or issue commands. But banal as the idea is, it becomes very powerful when applied to certain sentences which might be supposed by a simple-minded philosopher or by a grammarian of narrow vision to state facts when they really do not. J. L. Austin was a leader in the effort to show that many philosophically controversial utterances are not descriptive, and his treatment of the words “I promise” is in a certain respect a paradigm of this kind of philosophizing. According to Austin, when a man uses these words in promising to do something, he is not rendering a report about the world, nor is he bringing us scientific, psychological news of the internal workings of his mind. In a similar spirit Hampshire believes that some of the language that philosophers treat when they discuss free will is not totally descriptive or fact-stating in character. Hampshire in this book does not make explicit use of Austin’s analysis of promising, but his approach is dominated by the related idea that in certain utterances about desires, beliefs, regrets, and intentions—the sorts of things that figure so heavily in traditional discussions of free will—there is a divergence from scientific discourse which, if taken seriously, will lead to a clarification of the concept of human freedom.

In particular Hampshire thinks that utterances in the first person like “I want to escape,” “I believe that Dallas is bigger than Washington,” and “I regret having done that” express a claim to knowledge which is different from that claimed by a scientific psychologist who asserts something about a human being in the third person. The psychologist who says “Oswald wanted to escape,” “Oswald believed that Dallas is bigger than Washington,” and “Oswald regretted having shot Kennedy,” may be asked how he knows all of these facts about Oswald and be expected to offer evidence for his assertions. But when Oswald says similar things in the first person and in the present tense, he cannot be sensibly asked how he knows, for example, that he wants to escape. Nevertheless, says Hampshire, Oswald claims to have knowledge when he utters these words, and to show what kind of knowledge he claims to have, Hampshire proceeds as follows. He argues that if Oswald should say “I believe that Dallas is bigger than Washington,” he would say by implication that this proposition is to be believed; and, if he should say “I regret having done that,” he would say by implication that his action is to be regretted. In other words, the kind of knowledge that Oswald claims to have when he uses such first-person sentences is at least partly normative. That is to say, it is knowledge of what ought to be or should be the case.


To this extent these sentences record decisions on his part, which decisions must be distinguished from any facts about him or other parts of nature. It is here that Hampshire locates the freedom of the individual. His view may therefore be contrasted with the view that we may be free even though our decisions are explained by external causes, provided that nothing prevents these decisions from issuing in actions. After all the factual returns are in, the individual makes up his mind to adopt an attitude, or to believe or to do something. When asked why he believes a given proposition he tells us why it is to be believed, he does not tell what events in his biography caused him to endorse it. When asked why he regrets a given action, he tells us why it is to be regretted and not what caused him to regret the action.

After arguing to this effect for about one hundred and eight of his one hundred and twelve pages, Hampshire unfortunately devotes only his four remaining pages to the relationship between his main thesis and a possible thesis of those who call themselves determinists. His brevity makes it difficult for me to be altogether sure of what he is saying, but it goes something like this. Hampshire construes determinism as the doctrine that we may explain all human behavior by connecting described “outputs” of the organism with described “inputs” by means of suitably framed scientific laws. Naturally, our descriptions in such a scheme will be very complicated and no one knows exactly what they will be like. But Hampshire does not press the determinist at this sore point, since he is willing to grant that the determinist might some day be able to present his laws and his descriptions in scientifically acceptable terms. What Hampshire opposes is the idea that such a deterministic scheme of explanation could ever replace—his word—the scheme he has analyzed so carefully and interestingly in the previous pages of his book. At bottom his reason seems to be that no such deterministic model of scientific explanation could eliminate the need for justifying the normative statements implied by our first-person expressions of belief, regret, and so on. No matter how successful a scientist might be in explaining deterministically why I believed that Dallas is bigger than Washington, he would not eliminate the need for a defense of the contention that this statement is true or worthy of belief.

Connected with this is the fact that the descriptive statements of a determinist who explains our beliefs and decisions will not reproduce—again Hampshire’s word—the normative element in the first-person sentences analyzed by Hampshire. Nevertheless, it seems to me that even though there is such a normative element in the speaker’s state of mind when he uses such sentences, it does not follow that a psychologist could never causally explain the speaker’s state of mind. Not only might the psychologist be able to describe and causally explain a state of mind which contains a normative element, but he might also be able to describe and explain a state of mind that might be called wholly normative. Suppose that instead of uttering the first-person sentence “I regret having shot Kennedy,” Oswald had said explicitly that his action was regrettable. A determinist could descriptively report this normative belief and then go on to explain it causally in the course of an account which rested upon facts of Oswald’s life and generalizations about him and about human behavior in general.

Hampshire might reply as follows, although I am not sure that he would. He may hold that when a psychologist uses apparently descriptive language in talking about Oswald’s partly or totally normative states of mind, even such apparently descriptive language will turn out upon analysis to be itself normative and hence not includable in a deterministic explanation. But such a reply seems to me inadequate. Hampshire’s key declaration here is that “a thought can be counted as a belief only insofar as it approximates to well-founded belief,” but this is terribly misleading. If Hampshire means that any belief—whether well-founded or not—must have more in common (from a certain point of view) with a well-founded belief than it has with, say, a sensation of a mood, I agree. But it is a far cry from this truism to the false contention that when we think we are merely reporting that a man has a belief, we are making a statement which is covertly normative because there is an indirect employment of the concept of a well-founded belief. I am perfectly aware that Hampshire insists, and properly so, that there is a difference between the attitudes of belief and regret on the one hand, and moods, sensations, and strong passion on the other. Even though we may believe something passionately, or sadly regret something, Hampshire calls attention to an important difference between a belief and a mere mood when he says that there are normative or active elements associated with the first but not with the second. A mood may just come upon us but a belief is something we adopt, yet I think Hampshire verges on, or falls into the error of, inferring from this that we cannot causally explain beliefs. In other words, I believe Hampshire is wrong if he thinks that because of their active nature, beliefs and regrets cannot be explained causally, whereas sensations, moods, and strong passions can be. Therefore, although I do not think deterministic explanation can replace rational justification, I believe they are compatible. I should add, of course, that I agree with Hampshire that a believer cannot at the same moment defend and causally explain his belief. But a belief may have a causal antecedent even though the believer is at a given moment too preoccupied to seek it out. The fact that others may seek it out is enough for the determinist.


It should be evident by now that the issues which Hampshire discusses here in an abstract, linguistic manner are not to be dismissed as matters of “mere grammar.” Only by attending carefully, as he has, to the use of language can we begin to understand some of the most baffling problems concerning human thought and action. It should also be evident that anyone who hopes to say something profound or significant about the study of man must deal with some of the questions Hampshire raises. Although he does not deal explicitly with the more general significance of his analyses, it should be clear that they seriously affect the foundations of psychological and historical studies. If the psychologist who studies attitudes like belief and regret is supposed to be unable to explain them as he can explain moods, strong passions, and sensations, then obviously the psychology of so-called higher active processes, to say nothing of history, will be sharply cut off from animal psychology as well as from biology.

Clearly, then, the value of an enormously influential tradition is at stake in what might seem to some like remote logical lucubrations. To appreciate the relevance of some of the questions Hampshire discusses one need only think of Darwin, who, although he allowed in The Descent of Man that “the difference between the mind of the lowest man and that of the highest animal is immense,” also said that “the difference in mind between man and the higher animals, great as it is, certainly is one of degree and not of kind,” and “that animals may constantly be seen to pause, deliberate, and resolve.” But if human regretting, believing, and the conscious adoption of attitudes were as sharply removed as Hampshire seems to think they are from those psychological processes or states which we share with the beasts, then the entire effort to explain the higher, active mental processes naturalistically would be called into question. As I understand Hampshire, some of these active processes cannot be causally explained at all and here I disagree with him. I see no a priori reason for supposing that we cannot causally explain active processes or the normative beliefs associated with them. For these normative beliefs, Wittgenstein might have said, are also parts of our natural history. We can causally explain why men hold certain moral convictions even though we must grant that the process of justifying them is not identical with, and hence not replaceable by, the process of explaining them causally. The fact that Hampshire seems to think otherwise is regrettable, but the fact that he has thought as deeply and as provocatively as he has about such questions is what explains the high regard in which this reviewer holds him. He is one of those rare figures in today’s intellectual world: a sensitive, cultivated philosopher who thinks it is of the first importance to address himself in a rational and rigorous way to problems of the deepest human concern.

This Issue

November 25, 1965