Published in Contemporary British Philosophy (2nd series), ed. J. H. Muirhead, 1925. Reprinted in G. E. Moore, Philosophical Papers (1959).

A Defence of Common Sense

G. E. Moore

In what follows I have merely tried to state, one by one, some of the most important points in which my philosophical position differs from positions which have been taken up by some other philosophers. It may be that the points which I have had room to mention are not really the most important, and possibly some of them may be points as to which no philosopher has ever really differed from me. But, to the best of my belief, each is a point as to which many have really differed; although (in most cases, at all events) each is also a point as to which many have agreed with me.

I

The first point is a point which embraces a great many other points. And it is one which I cannot state as clearly as I wish to state it, except at some length. The method I am going to use for stating it is this. I am going to begin by enunciating, under the heading (1), a whole long list of propositions, which may seem, at first sight, such obvious truisms as not to be worth stating: they are, in fact, a set of propositions, every one of which (in my own opinion) I know, with certainty, to be true. I shall, next, under the heading (2), state a single proposition which makes an assertion about a whole set of classes of propositions -- each class being defined, as the class consisting of all propositions which resemble one of the propositions in (1) in a certain respect. (2), therefore, is a proposition which could not be stated, until the list of propositions in (1), or some similar list, had already been given. (2) is itself a proposition which may seem such an obvious truism as not to be worth stating: and it is also a proposition which (in my own opinion) I know, with certainty, to be true. But, nevertheless, it is, to the best of my belief, a proposition with regard to which many philosophers have, for different reasons, differed from me; even if they have not directly denied (2) itself, they have held views incompatible with it. My first point, then, may be said to be that (2), together with all its implications, some of which I shall expressly mention, is true.

(1) I begin, then, with my list of truisms, every one of which (in my own opinion) I know, with certainty, to be true. The propositions to be included in this list are the following:

There exists at present a living human body, which is my body. This body was born at a certain time in the past, and has existed continuously ever since, though not without undergoing changes; it was, for instance, much smaller when it was born, and for some time afterwards, than it is now. Ever since it was born, it has been either in contact with or not far from the surface of the earth; and, at every moment since it was born, there have also existed many other things, having shape and size in three dimensions (in the same familiar sense in which it has), from which it has been at various distances (in the familiar sense in which it is now at a distance both from that mantelpiece and from that bookcase, and at a greater distance from the bookcase than it is from the mantelpiece); also there have (very often, at all events) existed some other things of this kind with which it was in contact (in the familiar sense in which it is now in contact with the pen I am holding in my right hand and with some of the clothes I am wearing). Among the things which have, in this sense, formed part of its environment (i.e. have been either in contact with it, or at some distance from it, however great) there have, at every moment since its birth, been large numbers of other living human bodies, each of which has, like it, (a) at some time been born, (b) continued to exist from some time after birth, (c) been, at every moment of its life after birth, either in contact with or not far from the surface of the earth; and many of these bodies have already died and ceased to exist. But the earth had existed also for many years before my body was born; and for many of these years, also, large numbers of human bodies had, at every moment, been alive upon it; and many of these bodies had died and ceased to exist before it was born. Finally (to come to a different class of propositions), I am a human being, and I have, at different times since my body was born, had many different experiences, of each of many different kinds: e.g. I have often perceived both my own body and other things which formed part of its environment, including other human bodies; I have not only perceived things of this kind, but have also observed facts about them, such as, for instance, the fact which I am now observing, that that mantelpiece is at present nearer to my body than that bookcase; I have been aware of other facts, which I was not at the time observing, such as, for instance, the fact, of which I am now aware, that my body existed yesterday and was then also for some time nearer to that mantelpiece than to that bookcase; I have had expectations with regard to the future, and many beliefs of other kinds, both true and false; I have thought of imaginary things and persons and incidents, in the reality of which I did not believe; I have had dreams; and I have had feelings of many different kinds. And, just as my body has been the body of a human being, namely myself, who has, during his lifetime, had many experiences of each of these (and other) different kinds; so, in the case of very many of the other human bodies which have lived upon the earth, each has been the body of a different human being, who has, during the lifetime of that body, had many different experiences of each of these (and other) different kinds.

(2) I now come to the single truism which, as will be seen, could not be stated except by reference to the whole list of truisms, just given in (1). This truism also (in my own opinion) I know, with certainty to be true; and it is as follows:

In the case of very many (I do not say all) of the human beings belonging to the class (which includes myself) defined in the following way, i.e. as human beings who have had human bodies, that were born and lived for some time upon the earth, and who have, during the lifetime of those bodies, had many different experiences of each of the kinds mentioned in (1), it is true that each has frequently, during the life of his body, known, with regard to himself or his body, and with regard to some time earlier than any of the times at which I wrote down the propositions in (1), a proposition corresponding to each of the propositions in (1), in the sense that it asserts with regard to himself or his body and the earlier time in question (namely, in each case, the time at which he knew it), just what the corresponding proposition in (1) asserts with regard to me or my body and the time at which I wrote that proposition down.

In other words what (2) asserts is only (what seems an obvious enough truism) that each of us (meaning by "us," very many human beings of the class defined) has frequently known, with regard to himself or his body and the time at which he knew it, everything which, in writing down my list of propositions in (1), I was claiming to know about myself or my body and the time at which I wrote that proposition down, i.e. just as / knew (when I wrote it down) "There exists at present a living human body which is my body," so each of us has frequently known with regard to himself and some other time the different but corresponding proposition, which he could then have properly expressed by, "There exists at present a human body which is my body"; just as I know "Many human bodies other than mine have before now lived on the earth," so each of us has frequently known the different but corresponding proposition "Many human bodies other than mine have before now lived on the earth"; just as I know "Many human beings other than myself have before now perceived, and dreamed, and felt," so each of us has frequently known the different but corresponding proposition "Many human beings other than myself have before now perceived, and dreamed, and felt"; and so on, in the case of each of the propositions enumerated in (1).

I hope there is no difficulty in understanding, so far, what this proposition (2) asserts. I have tried to make clear by examples what I mean by "propositions corresponding to each of the propositions in (1)." And what (2) asserts is merely that each of us has frequently known to be true a proposition corresponding (in that sense) to each of the propositions in (1) -- a different corresponding proposition, of course, at each of the times at which he knew such a proposition to be true.

But there remain two points, which, in view of the way in which some philosophers have used the English language, ought, I think, to be expressly mentioned, if I am to make quite clear exactly how much I am asserting in asserting (2).

The first point is this. Some philosophers seem to have thought it legitimate to use the word "true" in such a sense that a proposition which is partially false may nevertheless also be true; and some of these, therefore, would perhaps say that propositions like those enumerated in (1) are, in their view, true, when all the time they believe that every such proposition is partially false. I wish, therefore, to make it quite plain that I am not using "true" in any such sense. I am using it in such a sense (and I think this is the ordinary usage) that if a proposition is partially false, it follows that it is not true, though, of course, it may be partially true. I am maintaining, in short, that all the propositions in (1), and also many propositions corresponding to each of these, are wholly true; I am asserting this in asserting (2). And hence any philosopher, who does in fact believe, with regard to any or all of these classes of propositions, that every proposition of the class in question is partially false, is, in fact, disagreeing with me and holding a view incompatible with (2), even though he may think himself justified in saying that he believes some propositions belonging to all of these classes to be "true."

And the second point is this. Some philosophers seem to have thought it legitimate to use such expressions as, e.g. "The earth has existed for many years past," as if they expressed something which they really believed, when in fact they believe that every proposition, which such an expression would ordinarily be understood to express, is, at least partially, false; and all they really believe is that there is some other set of propositions, related in a certain way to those which such expressions do actually express, which, unlike these, really are true. That is to say, they use the expression "The earth has existed for many years past" to express, not what it would ordinarily be understood to express, but the proposition that some proposition, related to this in a certain way, is true; when all the time they believe that the proposition, which this expression would ordinarily be understood to express, is, at least partially, false. I wish, therefore, to make it quite plain that I was not using the expressions I used in (1) in any such subtle sense. I meant by each of them precisely what every reader, in reading them, will have understood me to mean. And any philosopher, therefore, who holds that any of these expressions, if understood in this popular manner, expresses a proposition which embodies some popular error, is disagreeing with me and holding a view incompatible with (2), even though he may hold that there is some other, true, proposition which the expression in question might be legitimately used to express.

In what I have just said, I have assumed that there is some meaning which is the ordinary or popular meaning of such expressions as "The earth has existed for many years past." And this, I am afraid, is an assumption which some philosophers are capable of disputing. They seem to think that the question "Do you believe that the earth has existed for many years past?" is not a plain question, such as should be met either by a plain "Yes" or "No," or by a plain "I can't make up my mind," but is the sort of question which can be properly met by: "It all depends on what you mean by 'the earth' and 'exists' and 'years': if you mean so and so, and so and so, and so and so, then I do; but if you mean so and so, and so and so, and so and so, or so and so, and so and so, and so and so, or so and so, and so and so, and so and so, then I don't, or at least I think it is extremely doubtful." It seems to me that such a view is as profoundly mistaken as any view can be. Such an expression as "The earth has existed for many years past" is the very type of an unambiguous expression, the meaning of which we all understand. Anyone who takes a contrary view must, I suppose, be confusing the question whether we understand its meaning (which we all certainly do) with the entirely different question whether we know what it means, in the sense that we are able to give a correct analysis of its meaning. The question what is the correct analysis of the proposition meant on any occasion (for, of course, as I insisted in definin (2), a different proposition is meant at every different time at which the expression is used) by "The earth has existed for many years past" is, it seems to me, a profoundly difficult question, and one to which, as I shall presently urge, no one knows the answer. But to hold that we do not know what, in certain respects, is the analysis of what we understand by such an expression, is an entirely different thing from holding that we do not understand the expression. It is obvious that we cannot even raise the question how what we do understand by it is to be analysed, unless we do understand it. So soon, therefore, as we know that a person who uses such an expression is using it in its ordinary sense, we understand his meaning. So that in explaining that I was using the expressions used in (1) in their ordinary sense (those of them which have an ordinary sense, which is not the case with quite all of them), I have done all that is required to make my meaning clear.

But now, assuming that the expressions which I have used to express (2) are understood, I think, as I have said, that many philosophers have really held views incompatible with (2). And the philosophers who have done so may, I think, be divided into two main groups. A. What (2) asserts is, with regard to a whole set of classes of propositions, that we have, each of us, frequently known to be true propositions belonging to each of these classes. And one way of holding a view incompatible with this proposition is, of course, to hold, with regard to one or more of the classes in question, that no propositions of that class are true -- that all of them are, at least partially, false; since if, in the case of any one of these classes, no propositions of that class are true, it is obvious that nobody can have known any propositions of that class to be true, and therefore that we cannot have known to be true propositions belonging to each of these classes. And my first group of philosophers consists of philosophers who have held views incompatible with (2) for this reason. They have held, with regard to one or more of the classes in question, simply that no propositions of that class are true. Some of them have held this with regard to all the classes in question; some only with regard to some of them. But, of course, whichever of these two views they have held, they have been holding a view inconsistent with (2). B. Some philosophers, on the other hand, have not ventured to assert, with regard to any of the classes in (2), that no propositions of that class are true, but what they have asserted is that, in the case of some of these classes, no human being has ever known, with certainty, that any propositions of the class in question are true. That is to say, they differ profoundly from philosophers of group A, in that they hold that propositions of all these classes may be true; but nevertheless they hold a view incompatible with (2) since they hold, with regard to some of these classes, that none of us has ever known a proposition of the class in question to be true.

A. I said that some philosophers, belonging to this group, have held that no propositions belonging to any of the classes in (2) are wholly true, while others have only held this with regard to some of the classes in (2). And I think the chief division of this kind has been the following. Some of the propositions in (1) (and, therefore, of course, all propositions belonging to the corresponding classes in (2)) are propositions which cannot be true, unless some material things have existed and have stood in spatial relations to one another: that is to say, they are propositions which, in a certain sense, imply the reality of material things, and the reality of Space. E.g. the proposition that my body has existed for many years past, and has, at every moment during that time been either in contact with or not far from the earth, is a proposition which implies both the reality of material things (provided you use "material things" in such a sense that to deny the reality of material things implies that no proposition which asserts that human bodies have existed, or that the earth has existed, is wholly true) and also the reality of Space (provided, again, that you use "Space" in such a sense that to deny the reality Of Space implies that no proposition which asserts that anything has ever been in contact with or at a distance from another, in the familiar senses pointed out in (1), is wholly true). But others among the propositions in (1) (and, therefore, propositions belonging to the corresponding classes in (2)), do not (at least obviously) imply either the reality of material things or the reality of Space: e.g. the propositions that I have often had dreams, and have had many different feelings at different times. It is true that propositions of this second class do imply one thing which is also implied by all propositions of the first, namely that (in a certain sense) Time is real, and imply also one thing not implied by propositions of the first class, namely that (in a certain sense) at least one Self is real. But I think there are some philosophers, who, while denying that (in the senses in question) either material things or Space are real, have been willing to admit that Selves and Time are real, in the sense required. Other philosophers, on the other hand, have used the expression "Time is not real," to express some view that they held; and some, at least, of these have, I think, meant by this expression something which is incompatible with the truth of any of the propositions in (1) -- they have meant, namely, that every proposition of the sort that is expressed by the use of "now" or "at present," e.g. "I am now both seeing and hearing" or "There exists at present a living human body," or by the use of a past tense, e.g. "I have had many experiences in the past," or "The earth has existed for many years," are, at least partially, false.

All the four expressions I have just introduced, namely, "Material things are not real," "Space is not real," "Time is not real," "The Self is not real," are, I think, unlike the expressions I used in (1), really ambiguous. And it may be that, in the case of each of them, some philosopher has used the expression in question to express some view he held which was not incompatible with (2). With such philosophers, if there are any, I am not, of course, at present concerned. But it seems to me that the most natural and proper usage of each of these expressions is a usage in which it does express a view incompatible with (2); and, in the case of each of them, some philosophers have, I think, really used the expression in question to express such a view. All such philosophers have, therefore, been holding a view incompatible with (2).

All such views, whether incompatible with all of the propositions in (1), or only with some of them, seems to me to be quite certainly false; and I think the following points are specially deserving of notice with regard to them:

(a) If any of the classes of propositions in (2) is such that no proposition of that class is true, then no philosopher has ever existed, and therefore none can ever have held with regard to any such class, that no proposition belonging to it is true. In other words, the proposition that some propositions belonging to each of these classes are true is a proposition which has the peculiarity, that, if any philosopher has ever denied it, it follows from the fact that he has denied it, that he must have been wrong in denying it. For when I speak of "philosophers" I mean, of course (as we all do), exclusively philosophers who have been human beings, with human bodies that have lived upon the earth, and who have at different times had many different experiences. If, therefore, there have been any philosophers, there have been human beings of this class; and if there have been human beings of this class, all the rest of what is asserted in (1) is certainly true too. Any view, therefore, incompatible with the proposition that many propositions corresponding to each of the propositions in (1) are true, can only be true, on the hypothesis that no philosopher has ever held any such view. It follows, therefore, that, in considering whether this proposition is true, I cannot consistently regard the fact that many philosophers, whom I respect, have, to the best of my belief, held views incompatible with it, as having any weight at all against it. Since, if I know that they have held such views, I am, ipso facto, knowing that they were mistaken; and, if I have no reason to believe that the proposition in question is true, I have still less reason to believe that they have held views incompatible with it; since I am more certain that they have existed and held some views, i.e. that the proposition in question is true, than that they have held any views incompatible with it.

(b) It is, of course, the case that all philosophers who have held such views have repeatedly, even in their philosophical works, expressed other views inconsistent with them: i.e. no philosopher has ever been able to hold such views consistently. One way in which they have betrayed this inconsistency, is by alluding to the existence of other philosophers. Another way is by alluding to the existence of the human race, and in particular by using "we" in the sense in which I have already constantly used it, in which any philosopher who asserts that "we" do so and so, e.g. that "we sometimes believe propositions that are not true," is asserting not only that he himself has done the thing in question, but that very many other human beings, who have had bodies and lived upon the earth, have done the same. The fact is, of course, that all philosophers have belonged to the class of human beings which exists only if (2) be true: that is to say, to the class of human beings who have frequently known propositions corresponding to each of the propositions in (1). In holding views incompatible With the proposition that propositions of all these classes kre true, they have, therefore, been holding views inconsistent with propositions which they themselves knew to be true; and it was, therefore, only to be expected that they shoiMd sometimes betray their knowledge of such propositions. The strange thing is that philosophers should have been able to hold sincerely, as part of their philosophical creed, propositions inconsistent with what they themselves knew to be true; and yet, so far as I can make out, this has really frequency happened. My position, therefore, on this first point, differs from that of philosophers belonging to this group A, not in that I hold anything which they don't hold, but only in that I don't hold, as part of my philosophical creed, things which they do! hold as part of theirs -- that is to say, propositions inconsistent with some which they and I both hold in common. But this difference seems to me to be an important one.

(c) Some qf these philosophers have brought forward, in favour of their position, arguments designed to show, in the case of some or all of the propositions in (1), that no propositions of that type can possibly be wholly true, because every such proposition entails both of two incompatible propositions. And I admit, of course, that if any of the propositions in (1) did entail both of two incompatible propositions it could not be true. But it seems to me I have an absolutely conclusive argument to show that none of them does entail both of two incompatible propositions. Namely this: All of the propositiofis in (1) are true; no true proposition entails both of two incompatible propositions; therefore, none of the propositions in (1) entails both of two incompatible propositions.

(d) Although, as I have urged, no philosopher who has held with regard to any of these types of proposition that no propositions of that type are true, has failed to hold also other views inconsistent with his view in this respect, yet I do not think that the view, with regard to any or all of these types, that no proposition belonging to them is true, is in itself a self-contradictory view, i.e. entails both of two incompatible propositions. On the contrary, it seems to me quite clear that it might have been the case that Time was not real, material things not real, Space not real, selves not real. And in favour of my view that none of these things, which might have been the case, is in fact the case, I have, I think, no better argument than simply this -- namely, that all the propositions in (1) are, in fact, true.

B. This view, which is usually considered a much more modest view than A, has, I think, the defect that, unlike A, it really is self-contradictory, i.e. entails both of two mutually incompatible propositions.

Most philosophers who have held this view, have held, I think, that though each of us knows propositions corresponding to some of the propositions in (1), namely to those which merely assert that I myself have had in the past experiences of certain kinds at many different times, yet none of us knows for certain any propositions either of the type (a) which assert the existence of material things or of the type (b) which assert the existence of other selves, beside myself, and that they also have had experiences. They admit that we do in fact believe propositions of both these types, and that they may be true: some would even say that we know them to be highly probable; but they deny that we ever know them, for certain, to be true. Some of them have spoken of such beliefs as "beliefs of Common Sense," expressing thereby their conviction that beliefs of this kind are very commonly entertained by mankind: but they are convinced that these things are, in all cases, only believed, not known for certain; and some have expressed this by saying that they are matters of Faith, not of Knowledge.

Now the remarkable thing which those who take this view have not, I think, in general duly appreciated, is that, in each case, the philosopher who takes it is making an assertion about "us" -- that is to say, not merely about himself, but about many other human beings as well. When he says "No human being has ever known of the existence of other human beings," he is saying: "There have been many other huma beings beside myself, and none of them (including myself) has ever known of the existence of other human beings." If he says: "These beliefs are beliefs of Common Sense, but they are not matters of knowledge," he is saying: "There have been many other human beings, beside myself, who have shared these beliefs, but neither I nor any of the rest has ever known them to be true." In other words, he asserts with confidence that these beliefs are beliefs of Common Sense, and seems often to fail to notice that, if they are, they must be true; since the proposition that they are beliefs of Common Sense is one which logically entails propositions both of type (a) and of type (b); it logically entails the proposition that many human beings, beside the philosopher himself, have had human bodies, which lived upon the earth, and have had various experiences, including beliefs of this kind. This is why this position, as contrasted with positions of group A, seems to me to be self-contradictory. Its difference from A consists in the fact that it is making a proposition about human knowledge in general, and therefore is actually asserting the existence of many human beings, whereas philosophers of group A in stating their position are not doing this: they are only contradicting other things which they hold. It is true that a philosopher who says "There have existed many human beings beside myself, and none of us has ever known of the existence of any human beings beside himself," is only contradicting himself if what he holds is "There have certainly existed many human beings beside myself" or, in other words, "I know that there have existed other human beings beside myself." But this, it seems to me, is what such philosophers have in fact been generally doing. They seem to me constantly to betray the fact that they regard the proposition that those beliefs are beliefs of Common Sense, or the proposition that they themselves are not the only members of the human race, as not merely true, but certainly true; and certainly true it cannot be, unless one member, at least, of the human race, namely themselves, has known the very things which that member is declaring that no human being has ever known.

Nevertheless, my position that I know, with certainty, to be true all of the propositions in (1), is certainly not a position, the denial of which entails both of two incompatible propositions. If I do know all these propositions to be true, then, I think, it is quite certain that other human beings also have known corresponding propositions: that is to say (2) also is true, and I know it to be true. But do I really know all the propositions in (1) to be true? Isn't it possible that I merely believe them? Or know them to be highly probable? In answer to this question, I think I have nothing better to say than that it seems to me that I do know them, with certainty. It is, indeed, obvious that, in the case of most of them, I do not know them directly: that is to say, I only know them because, in the past, I have known to be true other propositions which were evidence for them. If, for instance, I do know that the earth had existed for many years before I was born, I certainly only know this because I have known other things in the past which were evidence for it. And I certainly do not know exactly what the evidence was. Yet all this seems to me to be no good reason for doubting that I do know it. We are all, I think, in this strange position that we do know many things, with regard to which we know further that we must have had evidence for them, and yet we do not know how we know them, i.e. we do not know what the evidence was. If there is any "we," and if we know that there is, this must be so: for that there is a "we" is one of the things in question. And that I do know that there is a "we," that is to say, that many other human beings, with human bodies, have lived upon the earth, it seems to me that I do know, for certain.

If this first point in my philosophical position, namely my belief in (2), is to be given any name, which has actually been used by philosophers in classifying the positions of other philosophers, it would have, I think, to be expressed by saying that I am one of those philosophers who have held that the "Common Sense view of the world" is, in certain fundamental features, wholly true. But it must be remembered that, according to me, all philosophers, without exception, have agreed with me in holding this: and that the real difference, which is commonly expressed in this way, is only a difference between those philosophers, who have also held views inconsistent with these features in "the Common Sense view of the world," and those who have not.

The features in question (namely, propositions of any of the classes defined in defining (2)) are all of them features, which have this peculiar property -- namely, that if we know that they are features in the "Common Sense view of the world," it follows that they are true: it is self-contradictory to maintain that we know them to be features in the Common Sense view, and that yet they are not true; since to say that we know this, is to say that they are true. And many of them also have the further peculiar property that, if they are features in the Common Sense view of the world (whether "we" know this or not), it follows that they are true, since to say that there is a "Common Sense view of the world," is to say that they are true. The phrases "Common Sense view of the world" or "Common Sense beliefs" (as used by philosophers) are, of course, extraordinarily vague; and, for all I know, there may be many propositions which may be properly called features in "the Common Sense view of the world" or "Common Sense beliefs," which are not true, and which deserve to be mentioned with the contempt with which some philosophers speak of "Common Sense beliefs." But to speak with contempt of those "Common Sense beliefs" which I have mentioned is quite certainly the height of absurdity. And there are, of course, enormous numbers of other features in "the Common Sense view of the world" which, if these are true, are quite certainly true too: e.g. that there have lived upon the surface of the earth not only human beings, but also many different species of plants and animals, etc. etc.

II

What seems to me the next in importance of the points in which my philosophical position differs from positions held by some other philosophers, is one which I will express in the following way. I hold, namely, that there is no good reason to suppose either (A) that every physical fact is logically dependent upon some mental fact or (B) that every physical fact is causally dependent upon some mental fact. In saying this, I am not, of course, saying that there are any physical facts which are wholly independent (i.e. both logically and causally) of mental facts: I do, in fact, believe that there are; but that is not what I am asserting. I am only asserting that there is no good reason to suppose the contrary; by which I mean, of course, that none of the human beings, who have had human bodies that lived upon the earth, have, during the lifetime of their bodies, had any good reason to suppose the contrary. Many philosophers have, I think, not only believed either that every physical fact is logically dependent upon some mental fact ("physical fact" and "mental fact" being understood in the sense in which I am using these terms) or that every physical fact is causally dependent upon some mental fact, or both, but also that they themselves had good reason for these beliefs. In this respect, therefore, I differ from them.

In the case of the term "physical fact," I can only explain how I am using it by giving examples. I mean by "physical facts," facts like the following: "That mantelpiece is at present nearer to this body than that bookcase is," "The earth has existed for many years past," "The moon has at every moment for many years past been nearer to the earth than to the sun," "That mantelpiece is of a light colour." But, when I say "facts like these," I mean, of course, facts like them in a certain respect; and what this respect is I cannot define. The term "physical fact" is, however, in common use; and I think that I am using it in its ordinary sense. Moreover, there is no need for a definition to make my point clear; since among the examples I have given there are some with regard to which i I hold that there is no reason to suppose them (i.e. these particular physical facts) either logically or causally dependent | upon any mental fact.

"Mental fact," on the other hand, is a much more unusu expression, and I am using it in a specially limited sense, which, though I think it is a natural one, does need to be explained. There may be many other senses in which the term can be properly used, but I am only concerned with this one; and hence it is essential that I should explain what it is.

There may, possibly, I hold, be "mental facts" of three different kinds. It is only with regard to the first kind that am sure that there are facts of that kind; but if there we any facts of either of the other two kinds, they would "mental facts" in my limited sense, and therefore I must plain what is meant by the hypothesis that there are facts those two kinds.

(a) My first kind is this. I am conscious now; and also 11 seeing something now. These two facts are both of the mental facts of my first kind; and my first kind consists clusively of facts which resemble one or other of the two in a certain respect.

(α) The fact that I am conscious now is obviously, in certain sense, a fact, with regard to a particular individua and a particular time, to the effect that that individual is con scious at that time. And every fact which resembles this in that respect is to be included in my first kind of men fact. Thus the fact that I was also conscious at many diffe ent times yesterday is not itself a fact of this kind: but it tails that there are (or, as we should commonly say, becaxi the times in question are past times, "were") many other fa of this kind, namely each of the facts, which, at each of times in question, I could have properly expressed by "I conscious now." Any fact which is, in this sense, a fact regard to an individual and a time (whether the individu be myself or another, and whether the time be past or pr ent), to the effect that that individual is conscious at that time, is to be included in my first kind of mental fact: and I call such facts, facts of class (α).

(β) The second example I gave, namely the fact that I am seeing something now, is obviously related to the fact that I am conscious now in a peculiar manner. It not only entails the fact that I am conscious now (for from the fact that I am seeing something it follows that I am conscious: I could not have been seeing anything, unless I had been conscious, though I might quite well have been conscious without seeing anything) but it also is a fact, with regard to a specific way (or mode) of being conscious, to the effect that I am conscious in that way: in the same sense in which the proposition (with regard to any particular thing) "This is red" both entails the proposition (with regard to the same thing) "This is coloured," and is also a proposition, with regard to a specific way of being coloured, to the effect that that thing is coloured in that way. And any fact which is related in this peculiar manner to any fact of class (α), is also to be included in my first kind of mental fact, and is to be called a fact of class (β). Thus the fact that I am hearing now is, like the fact that I am seeing now, a fact of class (β); and so is any fact, with regard to myself and a past time, which could at that time have been properly expressed by "I am dreaming now," "I am imagining now," "I am at present aware of the fact that . . . ," etc. etc. In short, any fact, which is a fact with regard to a particular individual (myself or another), a particular time (past or present), and any particular kind of experience, to the effect that that individual is having at that time an experience of that particular kind, is a fact of class (β): and only such facts are facts of class (β).

My first kind of mental facts consists exclusively of facts of classes (α) and (β), and consists of all facts of either of these kinds.

(b) That there are many facts of classes (α) and (β) seems to me perfectly certain. But many philosophers seem to me to have held a certain view with regard to the analysis of facts of class (a), which is such that, if it were true, there would be facts of another kind, which I should wish also to call "mental facts." I don't feel at all sure that this analysis is true; but it seems to me that it may be true; and since we can understand what is meant by the supposition that it is true, we can also understand what is meant by the supposition that there are "mental facts" of this second kind.

Many philosophers have, I think, held the following view as to the analysis of what each of us knows, when he knows (at any time) "I am conscious now." They have held, namely, that there is a certain intrinsic property (with which we are all of us familiar and which might be called that of "being an experience") which is such that, at any time at which any man knows "I am conscious now," he is knowing, with regard to that property and himself and the time in question, "There is occurring now an event which has this property (i.e. 'is an experience') and which is an experience of mine," and such that this fact is what he expresses by "I am conscious now." And if this view is true, there must be many facts of each of three kinds, each of which I should wish to call "mental facts"; viz. (1) facts with regard to some event, which has this supposed intrinsic property, and to some time, to the effect that that event is occurring at that time, (2) facts with regard to this supposed intrinsic property and some time, to the effect that some event which has that property is occurring at that time, and (3) facts with regard to some property, which is a specific way of having the supposed intrinsic property (in the sense above explained in which "being red" is a specific way of "being coloured") and some time, to the effect that some event which has that specific property is occurring at that time. Of course, there not only are not, but cannot be, facts of any of these kinds, unless there is an intrinsic property related to what each of us (on any occasion) expresses by "I am conscious now," in the manner defined above; and I feel very doubtful whether there is any such property; in other words, although I know for certain both that I have had many experiences, and that I have had experiences of many different kinds, I feel very doubtful whether to say the first is the same thing as to say that there have been many events, each of which was an experience and and experience of mine, and whether to say the second is the same thing as to say that there have been many events, each of which was an experience of mine, and each of which also had a different property, which was a specific way of being an experience. The proposition that I have had experiences does not necessarily entail the proposition that there have been any events which were experiences; and I cannot satisfy myself that I am acquainted with any events of the supposed kind. But yet it seems to me possible that the proposed analysis of "I am conscious now" is correct: that I am really acquainted with events of the supposed kind, though I cannot see that I am. And if I am, then I should wish to call the three kinds of facts defined above "mental facts." Of course, if there are "experiences" in the sense defined, it would be possible (as many have held) that there can be no experiences which are not some individual's experiences; and in that case any fact of any of these three kinds would be logically dependent on, though not necessarily identical with, some fact of class (α) or class (β). But it seems to me also a possibility that, if there are "experiences," there might be experiences which did not belong to any individual; and, in that case, there would be "mental facts" which were neither identical with nor logically dependent on any fact of class (α) or class (β).

(c) Finally some philosophers have, so far as I can make out, held that there are or may be facts which are facts with regard to some individual, to the effect that he is conscious, or is conscious in some specific way, but which differ from facts of classes (α) and (β), in the important respect that they are not facts with regard to any time: they have conceived the possibility that there may be one or more individuals, who are timelessly conscious, and tunelessly conscious in specific modes. And others, again, have, I think, conceived the hypothesis that the intrinsic property defined in (b) may be one which does not belong only to events, but may also belong to one or more wholes, which do not occur at any time: in other words, that there may be one or more timeless experiences, which might or might not be the experiences of some individual. It seems to me very doubtful whether any of these hypotheses are even possibly true; but I cannot see for certain that they are not possible: and, if they are possible, then I should wish to give the name "mental fact" to any fact (if there were any) of any of the five following kinds, viz.

  1. to any fact which is the fact, with regard to any individual, that he is timelessly conscious,
  2. to any fact which is the fact, with regard to any individual, that he is timelessly conscious in any specific way,
  3. to any fact which is the fact with regard to a timeless experience that it exists,
  4. to any fact which is the fact with regard to the supposed intrinsic property "being an experience," that something timelessly exists which has that property, and
  5. to any fact which is the fact, with regard to any property, which is a specific mode of this supposed intrinsic property, that something timelessly exists which has that property.

I have, then, defined three different kinds of facts, each of which is such that, if there were any facts of that kind (as there certainly are, in the case of the first kind), the facts in question would be "mental facts" in my sense; and to complete the definition of the limited sense in which I am using "mental facts," I have only to add that I wish also to apply the name to one fourth class of facts: namely to any fact, which is the fact, with regard to any of these three kinds of facts, or any kinds included in them, that there are facts of the kind in question; i.e. not only will each individual fact of class (α) be, in my sense, a "mental fact," but also the general fact "that there are facts of class (α)," will itself be a "mental fact"; and similarly in all other cases: e.g. not only will the fact that I am now perceiving (which is a fact of class (β)) be a "mental fact," but also the general fact that there are facts, with regard to individuals and times, to the effect that the individual in question is perceiving at the time in question, will be a "mental fact."

A. Understanding "physical fact" and "mental fact" in the senses just explained, I hold, then, that there is no good reason to suppose that every physical fact is logically dependent upon some mental fact. And I use the phrase, with regard to two facts, F1 and F2, is logically dependent on F2," wherever and only where F1 entails F2, either in the sense in which the proposition "I am seeing now" entails the proposition "I am conscious now," or the proposition (with regard to any particular thing) "This is red" entails the proposition (with regard to the same thing) "This is coloured," or else in the more strictly logical sense in which (for instance) the conjunctive proposition "All men are mortal, and Mr. Baldwin is a man" entails the proposition "Mr. Baldwin is mortal." To say, then, of two facts, F1 and F2, that F1 is not logically dependent upon F2, is only to say that F1 might have been a fact, even if there had been no such fact as F2; or that the conjunctive proposition "F1 is a fact, but there is no such fact as F2" is a proposition which is not self-contradictory, i.e. does not entail both of two mutually incompatible propositions.

I hold, then, that, in the case of some physical facts, there is no good reason to suppose that there is some mental fact, such that the physical fact in question could not have been a fact unless the mental fact in question had also been one. And my position is perfectly definite, since I hold that this is the case with all the four physical facts, which I have given as examples of physical facts. For example, there is no good reason to suppose that there is any mental fact whatever, such that the fact that that mantelpiece is at present nearer to my body than that bookcase could not have been a fact, unless the mental fact in question had also been a fact; and, similarly, in all the other three cases.

In holding this I am certainly differing from some philosophers. I am, for instance, differing from Berkeley, who held that that mantelpiece, that bookcase, and my body are, all of them, either "ideas" or "constituted by ideas," and that no "idea" can possibly exist without being perceived. He held, that is, that this physical fact is logically dependent upon a mental fact of my fourth class: namely a fact which is the fact that there is at least one fact, which is a fact with regard to an individual and the present time, to the effect that that individual is now perceiving something. He does not say that this physical fact is logically dependent upon any fact which is a fact of any of my first three classes, e.g. on any fact which is the fact, with regard to a particular individual and the present time, that that individual is now perceiving something: what he does say is that the physical fact couldn't have been a fact, unless it had been a fact that there was some mental fact of this sort. And it seems to me that many philosophers, who would perhaps disagree either with Berkeley's assumption that my body is an "idea" or "constituted by ideas," or with his assumption that "ideas" cannot exist without being perceived, or with both, nevertheless would agree with him in thinking that this physical fact is logically dependent upon some "mental fact": e.g. they might say that it could not have been a fact, unless there had been, at some time or other, or, were tunelessly, some "experience." Many, indeed, so far as I can make out, have held that every fact is logically dependent on every other fact. And, of course, they have held in the case of their opinions, as Berkeley did in the case of his, that they had good reasons for them.

B. I also hold that there is no good reason to suppose that every physical fact is causally dependent upon some mental fact. By saying that F1 is causally dependent on F2, I mean only that F1 wouldn't have been a fact unless F2 had been; not (which is what "logically dependent" asserts) that F1 couldn't conceivably have been a fact, unless F2 had been. And I can illustrate my meaning by reference to the example which I have just given. The fact that that mantelpiece is at present nearer to my body than that bookcase, is (as I have just explained) so far as I can see, not logically dependent upon any mental fact; it might have been a fact, even if there had been no mental facts. But it certainly is causally dependent on many mental facts: my body would not have been here unless I had been conscious in various ways in the past; and the mantelpiece and the bookcase certainly would not have existed, unless other men had been conscious too.

But with regard to two of the facts, which I gave as instances of physical facts, namely the fact that the earth has existed for many years past, and the fact that the moon has for many years past been nearer to the earth than to the sun, I hold that there is no good reason to suppose that these are causally dependent upon any mental fact. So far as I can see, there is no reason to suppose that there is any mental fact of which it could be truly said: unless this fact had been a fact, the earth would not have existed for many years past. And in holding this, again, I think I differ from some philosophers. I differ, for instance, from those who have held that all material things were created by God, and that they had good reasons for supposing this.

III

I have just explained that I differ from those philosophers who have held that there is good reason to suppose that all material things were created by God. And it is, I think, an important point in my position, which should be mentioned, that I differ also from all philosophers who have held that there is good reason to suppose that there is a God at all, whether or not they have held it likely that he created all material things.

And similarly, whereas some philosophers have held that there is good reason to suppose that we, human beings, shall continue to exist and to be conscious after the death of our bodies, I hold that there is no good reason to suppose this.

IV

I now come to a point of a very different order.

As I have explained under I., I am not at all sceptical as to the truth of such propositions as "The earth has existed for many years past." "Many human bodies have each lived for many years upon it," i.e. propositions which assert the existence of material things: on the contrary, I hold that we all know, with certainty, many such propositions to be true. But I am very sceptical as to what, in certain respects, the correct analysis of such propositions is. And this is a matter as to which I think I differ from many philosophers. Many seem to hold that there is no doubt at all as to their analysis, nor, therefore, as to the analysis of the proposition "Material things have existed," in certain respects in which I hold that the analysis of the propositions in question is extremely doubtful; and some of them, as we have seen, while holding that there is no doubt as to their analysis, seem to have doubted whether any such propositions are true. I, on the other hand, while holding that there is no doubt whatever that many such propositions are wholly true, hold also that no philosopher, hitherto, has succeeded in suggesting an analysis of them, as regards certain important points, which comes anywhere near to being certainly true.

It seems to me quite evident that the question how propositions of the type I have just given are to be analysed, depends on the question how propositions of another and simpler type are to be analysed. I know, at present, that I am perceiving a human hand, a pen, a sheet of paper, etc.; and it seems to me that I cannot know how the proposition "Material things exist" is to be analysed, until I know how, in certain respects, these simpler propositions are to be analysed. But even these are not simple enough. It seems to me quite evident that my knowledge that I am now perceiving a human hand is a deduction from a pair of propositions simpler still -- propositions which I can only express in the form "I am perceiving this" and "This is a human hand." It is the analysis of propositions of the latter kind which seems to me to present such great difficulties, while nevertheless the whole question as to the nature of material things obviously depends upon their analysis. It seems to me a surprising thing that so few philosophers, while saying a great deal as to what material things are and as to what it is to perceive them, have attempted to give a clear account as to what precisely they suppose themselves to know (or to judge, in case they have held that we don't know any such propositions to be true, or even that no such propositions are true) when they know or judge such things as "This is a hand," "That is the sun," "This is a dog," etc. etc. etc.

Two things only seem to me to be quite certain about the analysis of such propositions (and even with regard to these I am afraid some philosophers would differ from me) namely that whenever I know, or judge, such a proposition to be true, (1) there is always some sense-datum about which the proposition in question is a proposition -- some sense-datum which is a subject (and, in a certain sense, the principal or ultimate subject) of the proposition in question, and (2) that, nevertheless, what I am knowing or judging to be true about this sense-datum is not (in general) that it is itself a hand, or a dog, or the sun, etc. etc., as the case may be.

Some philosophers have I think doubted whether there are any such things as other philosophers have meant by "sense-data" or "sensa." And I think it is quite possible that some philosophers (including myself, in the past) have used these terms in senses such that it is really doubtful whether there are any such things. But there is no doubt at all that there are sense-data, in the sense in which I am now using that term. I am at present seeing a great number of them, and feeling others. And in order to point out to the reader what sort of things I mean by sense-data, I need only ask him to look look at his own right hand. If he does this he will be able to pick out something (and, unless he is seeing double, only one thing) with regard to which he will see that it is, at first sight, a natural view to take that that thing is identical, not, indeed, with his whole right hand, but with that part of its surface which he is actually seeing, but will also (on a little reflection) be able to see that it is doubtful whether it can be identical with the part of the surface of his hand in question. Things of the sort (in a certain respect) of which this thing is, which he sees in looking at his hand, and with regard to which he can understand how some philosophers should have supposed it to be the part of the surface of his hand which he is seeing, while others have supposed that it can't be, are what I mean by "sense-data." I therefore define the term in such a way that it is an open question whether the sense-datum which I now see in looking at my hand and which is a sense-datum of my hand is or is not identical with that part of its surface which I am now actually seeing.

That what I know, with regard to this sense-datum, when I know "This is a human hand," is not that it is itself a human hand, seems to me certain because I know that my hand has many parts (e.g. its other side, and the bones inside it), which are quite certainly not parts of this sense-datum.

I think it certain, therefore, that the analysis of the proposition "This is a human hand" is, roughly at least, of the form "There is a thing, and only one thing, of which it is true both that it is a human hand and that this surface is a part of its surface." In other words, to put my view in terms of the phrase "theory of representative perception," I hold it to be quite certain that I do not directly perceive my hand; and that when I am said (as I may be correctly said) to "perceive" it, that I "perceive" it means that I perceive (in a different and more fundamental sense) something which is (in a suitable sense) representative of it, namely, a certain part of its surface.

This is all that I hold to be certain about the analysis of the proposition "This is a human hand." We have seen that it includes in its analysis a proposition of the form "This is part of the surface of a human hand" (where "This," of course, has a different meaning from that which it has in the original proposition which has now been analysed). But this proposition also is undoubtedly a proposition about the sense-datum, which I am seeing, which is a sense-datum of my hand. And hence the further question arises: What, when I know "This is part of the surface of a human hand," am I knowing about the sense-datum in question? Am I, in this case, really knowing about the sense-datum in question that it itself is part of the surface of a human hand? Or, just as we found in the case of "This is a human hand," that what I was blowing about the sense-datum was certainly not that it itself was a human hand, so, is it perhaps the case, with this new proposition, that even here I am not knowing, with regard to the sense-datum, that it is itself part of the surface of a hand? And, if so, what is it that I am knowing about the sense-datum itself?

This is the question to which, as it seems to me, no philosopher has hitherto suggested an answer which comes anywhere near to being certainly true.

There seem to me to be three, and only three, alternative types of answer possible; and to any answer yet suggested, of any of these types, there seem to me to be very grave objections.

(1) Of the first type, there is but one answer: namely, that in this case what I am knowing really is that the sense-datum itself is part of the surface of a human hand. In other words that, though I don't perceive my hand directly, I do directly perceive part of its surface; that the sense-datum itself is this part of its surface and not merely something which (in a sense yet to be determined) "represents" this part of its surface; and that hence the sense in which I "perceive" this part of the surface of my hand, is not in its turn a sense which needs to be defined by reference to yet a third more ultimate sense of "perceive," which is the only one in which perception is direct, namely that in which I perceive the sense-datum.

If this view is true (as I think it may just possibly be), it seems to me certain that we must abandon a view which has been held to be certainly true by most philosophers, namely the view that our sense-data always really have the qualities which they sensibly appear to us to have. For I know that if another man were looking through a microscope at the same surface which I am seeing with the naked eye, the sense-datum which he saw would sensibly appear to him to have qualities very different from and incompatible with those which my sense-datum sensibly appears to me to have: and yet, if my sense-datum is identical with the surface we are both of us seeing, his must be identical with it also. My sense-datum can, therefore, be identical with this surface only on condition that it is identical with his sense-datum; and, since his sense-datum sensibly appears to him to have qualities incompatible with those which mine sensibly appears to me to have, his sense-datum can be identical with mine only on condition that the sense-datum in question either has not got the qualities which it sensibly appears to me to have, or has not got those which it sensibly appears to him to have.

I do not, however, think that this is a fatal objection to this first type of view. A far more serious objection seems to me to be that, when we see a thing double (have what is called "a double image" of it>, we certainly have two sense-data each of which is of the surface seen, and which cannot therefore both be identical with it; and that yet it seems as if, if any sense-datum is ever identical with the surface of which it is a sense-datum, each of these so-called "images" must be so. It looks, therefore, as if every sense-datum is, after all, only "representative" of the surface, of which it is a sense-datum.

(2) But, if so, what relation has it to the surface in question?

This second type of view is one which holds that when I know "This is part of the surface of a human hand," what I am knowing with regard to the sense-datum which is of that surface, is, not that it is itself part of the surface of a human hand, but something of the following kind. There is, it says, some relation, R, such that what I am knowing with regard to the sense-datum is either "There is one thing and only one thing, of which it is true both that it is a part of the surface of a human hand, and that it has R to this sense-datum," or else "There are a set of things, of which it is true both that that set, taken collectively, are part of the surface of a human hand, and also that each member of the set has R to this sense-datum, and that nothing which is not a member of the set has R to it."

Obviously, in the case of this second type, many different views are possible, differing according to the view they take as to what the relation R is. But there is only one of them, which seems to me to have any plausibility; namely that which holds that R is an ultimate and unanalysable relation, which might be expressed by saying that "xRy" means the same as "y is an appearance or manifestation of x." I.e. the analysis which this answer would give of "This is part of the surface of a human hand" would be "There is one and only one thing of which it is true both that it is part of the surface of a human hand, and that this sense-datum is an appearance or manifestation of it."

To this view also there seem to me to be very grave objections, chiefly drawn from] a consideration of the questions how we can possibly know with regard to any of our sense-data that there is one thing and one thing only which has to them such a supposed ultimate relation; and how, if we do, we can possibly know anything further about such things, e.g. of what size or shape they are.

(3) The third type of answer, which seems to me to be the only possible alternative if (1) and (2) are rejected, is the type of answer which J. S. Mill seems to have been implying to be the true one when he said that material things are "permanent possibilities of sensation." He seems to have thought that when I know such a fact as "This is part of the surface of a human hand," what I am knowing with regard to the sense-datum which is the principal subject of that fact, is not that it is itself part of the surface of a human hand, nor yet, with regard to any relation, that the thing which has to it that relation is part of the surface of a human hand, but a whole set of hypothetical facts each of which is a fact of the form "If these conditions had been fulfilled, I should have been perceiving a sense-datum intrinsically related to this sense-datum in this way," "If these (other) conditions had been fulfilled, I should have been perceiving a sense-datum intrinsically related to this sense-datum in this (other) way," etc. etc.

With regard to this thjird type of view as to the analysis of propositions of the kind; we are considering, it seems to me, again, just possible that it is a true one; but to hold (as Mill himself and others seemj to have held) that it is certainly, or nearly certainly, true, seems to me as great a mistake, as to hold with regard either to (1) or to (2), that they are certainly, or nearly certainly, true. There seem to me to be very grave objections to it; in particular the three, (a) that though, in general, when I know such a fact as "This is a hand," I certainly do know some hypothetical facts of the form "If these conditions had been fulfilled, I should have been perceiving a sense-datum of this kind, which would have been a sense-datum of the same surface of which this is a sense-datum," it seems doubtful whether any conditions with regard to which I know this are not themselves conditions of the form "If this and that material thing had been in those positions and conditions . . . ," (b) that it seems again very doubtful whether there is any intrinsic relation, such that my knowledge that (under these conditions) I should have been perceiving a sense-datum of this kind, which would have been a sense-datum of the same surface of which this is a sense-datum, is equivalent to a knowledge, with regard to that relation, that I should, under those conditions, have been perceiving a sense-datum related by it to this sense-datum, and (c) that, if it were true, the sense in which a material surface is "round" or "square," would necessarily be utterly different from that in which our sense-data sensibly appear to us to be "round" or "square.

V

Just as I hold that the proposition "There are and have been material things" is quite certainly true, but that the question how this proposition is to be analysed is one to which no answer that has been hitherto given is anywhere near certainly true; so I hold that the proposition "There are and have been many Selves" is quite certainly true, but that here again all the analyses of this proposition that have been sugggested by philosophers are highly doubtful.

That I am now perceiving many different sense-data, and that I have at many times in the past perceived many different sense-data, I know for certain -- that is to say, I know that there are mental facts of class (β), connected in a way which it is proper to express by saying that they are all of them facts about me; but how this kind of connection is to be analysed, I do not know for certain, nor do I think that any other philosopher knows with any approach to certainty. Just as in the case of the proposition "This is part of the surface of a human hand," there are several extremely different views as to its analysis, each of which seems to me possible, but none nearly certain, so also in the case of the proposition "This, that and that sense-datum are all at present being perceived by me," and still more so in the case of the proposition "I am now perceiving this sense-datum, and I have in the past perceived sense-data of these other kinds." Of the truth of these propositions there seems to me to be no doubt, but as to what is the correct analysis of them there seems to me to be the gravest doubt -- the true analysis may, for instance, possibly be quite as paradoxical as is the third view given under IV as to the analysis of "This is part of the surface of a human hand"; but whether it is as paradoxical as this seems to me to be quite as doubtful as in that case. Many philosophers, on the other hand, seem to me to have assumed that there is little or no doubt as to the correct analysis of such propositions; and many of these, just reversing my position, have also held that the propositions themselves are not true.