Hume on The Epistemological Status of Metaphysical Statements

My paper examines critically the widespread view that Hume confines meaningful propositions to those which are analytic ( a priori ), and those which are synthetic empirical, thereby rejecting synthetic a priori propositions as meaningful. What I show is that Hume does recognize certain metaphysical synthetic a priori propositions as meaningful, thereby dispelling the traditional view that Hume rejects all synthetic a priori as meaningful. 1


Hume on The Epistemological Status of Metaphysical Statements
During the past century, commentators, especially the logical positivists, have sought an understanding of the types of propositions Hume is willing to allow as genuine, in light of his emphasis on observation and experience, and the need for the employment of the Experimental Method in Philosophy. Two interpretations emerged in this discussion. Each begins with the view that, in addition to analytic (a priori) propositions, Hume recognizes as meaningful propositions which are synthetic and empirical. 2 Where they differ is that one side holds the more extreme position, that Hume holds that there are no synthetic a priori propositions, 1 whereas the other group insists that synthetic a priori propositions are misleading and meaningless, inasmuch as no means of verifying them is possible; 2 and consequently, they are pseudoproposition. What these views have in common is that both maintain that Hume's critique of his opponents -especially the rationalists, when they are doing metaphysics -is carried out by rejecting any proposition which purports to be both synthetic and a priori. I will attempt to establish that both views are mistaken in that, on the basis of what he says, Hume must admit a class of metaphysical propositions which are synthetic and a priori, and meaningful. If the views of Macnabb and Ayer are mistaken, then a reinterpretation is in order of the last paragraph of Hume's First Enquiry, cited by A.J. Ayer in defence of his position on Hume, and which I set out in the third footnote in this article.
I will now show that both views mentioned above are fundamentally mistaken, inasmuch as on the basis of what he says, Hume must admit a class of metaphysical propositions which is synthetic and a priori, and meaningful. 3 For the purposes of this paper, I will make the well -known assumption that all propositions can be divided into those which are analytic and those which are synthetic; and that every proposition is either a priori, or empirical or a posteriori. The meanings of these four classifications will also be offered along traditional lines. Accordingly, I hold that a proposition is empirical if, in order to verify it, some recourse to observation is required. If, on the other hand, no observation or series of observations can be employed to verify a proposition, solely because it would be impossible, even in principle, to verify it in this way, then the proposition in question is a priori. Therefore, to make out whether an a priori proposition is true or false, something other than recourse to observation is required. number(s)). The two passages most often cited to support this view are contained in the first two paragraphs of Section IV, Part I of the first Enquiry, and in the very last paragraph of the same book. 1 For example, D.G.C. Macnabb in David Hume-His Theory of Knowledge and Morality (Blackwell, 1951) writes: "Hume's contention is that no a priori propositions are synthetic, all a priori are analytic, all synthetic propositions are empirical" (p.46). 2 See, for example, A. J. Ayer's Language, Truth, and Logic (Dover Publications, Inc., New York, 1946), pp. 31, 35, 53-54. On page 54, Ayer writes: "Of Hume, we may say not merely that he was not in practice a metaphysician, but that he explicitly rejected metaphysics. We find the strongest evidence of this in the passage with which he concludes his Enquiry Concerning Human Understanding. "If", he says, "we take in our hand any volume; of divinity or school of metaphysics, for instance; let us ask, DOES IT CONTAIN ANY ABSTRACT REASONING CONCERNING QUANTITY OR NUMBER? No. DOES IT CONTAIN ANY EXPERIMENTAL REASONING CONCERNING MATTER OF FACT AND EXISTENCE? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion". (E.211; S.B.E.165) Ayer comments that this is but a rhetorical version of his own thesis that a sentence which does not express either a formally true proposition or an empirical hypothesis is devoid of literal significance? Ayer, therefore, regards Hume as holding the logical positivist position in regard to the meaning of sentences. 3 This is not to say that Hume regards all metaphysical synthetic a priori propositions to be meaningful. More on this toward the end of this article.
Further, a proposition is analytic if it can be made out to be true by apprehending the meanings of the terms involved, or if it has been deduced from other propositions which have been so verified. Since all analytic judgements are verified without recourse to experience, all analytic judgements must be a priori. If, even after one has apprehended the meaning of a proposition, further steps would be required to verify it, then the proposition is synthetic.
Having now delineated three major propositional categories -the possibility of analytic empirical propositions having been shown to be impossible -I now turn to Hume's writings to see where these three classifications can be employed.
In the Enquiry Concerning Human Understanding, Hume restricts propositions concerning the relations of ideas to Geometry, Algebra, and Arithmetic, and "in short, [to] every affirmation which is either intuitively or demonstratively certain." (E.108; S.B.E.25) Such propositions can be discovered by thought alone, "without dependence on what is anywhere existent in the universe", and as a result, "though there never were a circle or triangle in nature, the truths demonstrated by Euclid would forever retain their certainty and evidence". (E.108; S.B.E.25) An example of a proposition of this type which is intuitively certain is 'a triangle is not a square', and one which is demonstratively certain is 'the square on the hypotenuse is equal to the sum of the squares on the other two sides'. (E.209; S.B.E.163) To determine whether a proposition concerns the relations of ideas, we need only determine whether the contrary is contradictory. Now, it is this contention of Hume's which reveals that such propositions are analytic, since the criterion he has set out has application only in virtue of the meanings of the terms involved. Those propositions which are verified intuitively are immediately apprehended and gain our assent. However, others are not immediately seen by the mind, and, therefore, they require additional propositions (premises) before their truth is seen. And since the propositions to be demonstrated are analytically true, the premises employed in demonstrating their truth must be of the same nature.
Propositions concerning matters of fact are not verified in the same manner. The contrary of every matter of fact is still possible, and "we should in vain, therefore, attempt to demonstrate its falsehood". (E.108; S.B.E.26) How, then, do we acquire knowledge of matters of fact? According to Hume, there are three possible means, depending on the type of problem with which we are concerned.
In certain cases, we can have recourse to the "present testimony of the senses" (E.108; S.B.E.26), and here, clearly, Hume is thinking of verifying propositions through direct observation, as, for example, that the sun is now shining, or that I am angry. That is, included in this category are matters of fact which are external to the observer or states of her/ his own consciousness, and the propositions concerning these can be confirmed or confuted by an immediate observation or observations. Now, it often happens that the propositions concerning matters of fact can no longer be verified by direct observation. Nevertheless, Hume holds that such propositions may be verified, provided that we can recall an observation of what they are about.
Hume's main concern is with determining what evidence, if any, we have for apprehending the truth or falsity of propositions for which we do not have the direct evidence of the senses. To this question, he answers, first, that I can know a matter of fact which I myself have not observed if I know that it is connected by the relation of cause and effect with some matter of fact which I have observed. For example, if I observe footprints in the sand, then I can be said to know that some other human being was present here at some time in the past. In addition, Hume asserts that, if I am to know that two facts are connected by this relation, I can do so only through the aid of my own past experience. That is, the only legitimate evidence I can acquire for assenting to a causal proposition is to have found objects like the ones I now believe to be connected causally to have been constantly conjoined in the past. Therefore, for Hume, the only tribunal available to us for settling questions of fact is observation and experience. And as such, he holds that, in principle, the types of propositions we have been discussing can be known to be true or false. Further, from what was said earlier, it is clear that all the propositions concerning matters of fact discussed thus far must be classified as being both synthetic and empirical, since they cannot be verified by attending to the meanings of the terms involved, but only be having recourse to observation and experience.
It is to this point that philosophers such as D.G.C. Macnabb and A. J. Ayer maintain that Hume is willing to go in determining the range of meaningful propositions. To see that this cannot be Hume's position, we must now push our inquiry concerning the causal relation even further.
As we have seen, reasonings concerning matters of fact are founded on the relation of cause and effect, and the foundation of such reasonings is experience. But Hume now asks for the foundation of all conclusions from experience, and he warns that this question "may be of more difficult solution and explanation", and that "even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding" (E.113; S.B.E.32).
We must first get clear on what it is that Hume is asking; and it is this. We know nothing of the powers or forces which may reside in objects. All that is ever present to us are the sensible qualities which objects possess. But even in the light of our ignorance, whenever we observe objects with sensible qualities resembling those of objects observed in the past, we assume that they also have similar secret powers "and expect that effects, similar to those which we have experienced, will follow from them" (E.114; S.B.E.33). In effect, the problem resolves itself into determining how we move from 'I have found that such an object has always been attended by such an effect' to 'I foresee, that the other objects, which are, in appearance, similar, will be attended with similar effects' (E.114; S.B.E.34); although Hume also poses the problem by asking by what argument or faculty we arrive at the conclusion that the future will resemble the past (E.117; S.B.E.37), and that similar powers will be conjoined with similar sensible qualities (E.117; S.B.E.36). 1 Concentrating initially on the first formulation of the problem, Hume points out that, since the validity of the inference between these two propositions can be doubted, the connection between these two propositions cannot be intuitive. Nor can there be any demonstration of the conclusion we draw, since "it implies no contradiction that the course of nature may change" (E.115; S.B.E.35). Since we are able to assert without contradiction that the course of nature may change, the proposition that the future will resemble the past cannot be a tautology. Hence, this proposition must be synthetic in nature.
Perhaps, then, the inference between the two propositions in the first formulation is 'experimental'. But Hume shows that it is also not in this way that the inference is made. "To say it is experimental is begging the question. For all inferences from experience suppose, as their foundation, that the future will resemble the past, and that similar secret powers will be conjoined with similar sensible qualities -It is impossible, therefore, that any arguments from experience can prove this resemblance of the past to the future; since all these arguments are founded on the suppositions of the resemblance" (E.117; S.B.E.37-38).
Accordingly, we find in Hume's philosophy that the proposition that the future will resemble the past does not fit into the analytic -synthetic empirical mould which commentators such as Macnabb and Ayer hold is to be found there. In fact, the last passage quoted above makes our belief in the past as a standard for the future to be one which is requisite if we are to acquire empirical knowledge. That is, this belief that the future will resemble the past is ultimately seen to be a principle of the possibility of empirical knowledge -an expression anticipatory of Kant's philosophy.
Other passages offered by Hume lend further evidence to such an interpretation. For example, after concluding that we come to believe that the future will resemble the past through the influence of Custom, Hume adds: Custom, then, is the great guide to human life. It is that principle alone which renders our experience useful to us. Without the influence of custom, we should be entirely ignorant of every matter of fact beyond what is immediately present to the memory and senses. We should never know how to adjust means to end, or to employ our natural powers in the production of any effect. There would be an end of all action, as well as of the chief part of speculation. (E.122-123; S.B.E.45) 1 I stated earlier that in this article the assumption would be made that every proposition is either analytic or synthetic, and either a priori or empirical. Now, since experience has been ruled out as a means of verifying the proposition that the future will resemble the past, the proposition must be non-empirical, that is, it must be a priori. And since its being analytic or tautological has also been ruled out, it follows that it must also be synthetic. In short, we have come upon a synthetic a priori proposition in Hume's philosophy.
It may be objected here that the proposition that the future will resemble the past is not a priori, since the basis for our belief in this proposition is custom or habit which is acquired from repeated experience. But this objection can have no force here. Custom, for Hume, explains how certain propositions, and the belief which attends them, come to be generated. But Custom cannot explain the specific character of such propositions as synthetic and a priori, given that Custom has no role to play in the process of verification. In fact, the enigmatic aspect of the proposition that the future will resemble the past is that no means of verifying it is available to us. However, Hume's contention is that, precisely because of its character and the important role it plays in our lives, its veracity is something to which we must always consent: no speculative argument, nor any experience, can for long dissuade us of its truth. Unlike all synthetic empirical propositions, it is both, in practice and in principle, intrinsically unverifiable. When dealing with this passage, commentators often lose sight of the context in which it is provided. For this paragraph sums up Hume's answer to the question raised two pages earlier, namely, "what are the proper subjects of science and enquiry?", that is, when is it appropriate to engage in reasoning? But it does not follow that, because our reasonings may be fruitfully employed solely with respect to quantity or number (relations of ideas) and matters of fact and existence, that these two realms provide the only genuine propositions. On the contrary, all that does follow is that it is only in these two realms that reasoning can serve a justificatory function. It is still possible that what we cannot justify or verify, namely, certain metaphysical propositions which are both synthetic and a priori, may be genuine and meaningful. And so it is with the proposition that the future will resemble the past: although we all do, and must, believe in it, we find that no process of reasoning ever can support our belief in it. Hence, what Hume wants committed to the flames are volumes which are concerned with attempting to reason about, or verify, propositions which are not the proper objects of science and inquiry. How could such volumes not but contain 'sophistry and illusion'? In short, the final paragraph is not concerned to outline the scope of genuine propositions, but rather is concerned with the types of propositions about which our reasonings can be concerned. Therefore, Hume has not ruled out the possibility that there are genuine propositions which are both synthetic and a priori in nature. And if the analysis presented here is correct, then Hume must accept certain synthetic a priori propositions as genuine propositions, in the sense that, without believing in them, empirical knowledge would, for us, be impossible. 1 Each of these beliefs is such that "Nature has doubtless esteem'd it an affair of too great importance to be trusted to our uncertain reasonings and speculations". (T.125; S.B.T.187) A metaphysical proposition which is endorsed by Nature Hume holds to be genuine and meaningful, even though the proposition is synthetic and a priori. Synthetic a priori propositions which are not so endorsed, Hume rejects as meaningless. But even these meaningless propositions Hume divides into two groups. The first of these consists of those which are necessary to satisfy, at least temporarily, the conflicting claims of reason and imagination. Such is the case with those asserting the existence of souls and substance. The act of the mind here is called 'feigning', and the object of belief is entitled a 'fiction'. The 'feigning' and resultant 'fiction' take place when we ascribe an identity to our successive perceptions, and suppose ourselves possessed of an invariable and uninterrupted existence through the whole course of our lives.
Hume illustrates this with how we come to believe in the identity of plants and animals. He explains that we have a distinct idea of an object that remains invariable and uninterrupted through a period of time (identity or sameness), and we have a distinct idea of several different objects existing in succession and connected together by a close relation (diversity). He points out that in our common way of thinking, they are generally confounded with each other: "[T]he relation facilitates the transition of the mind from one object to another, and renders its passage as smooth as if it contemplated one continuous object…In order to justify to ourselves this absurdity, we often feign some new and unintelligible principle, that connects the objects together, and prevents their interruption or variation. Thus we feign the continued existence of the perceptions of our senses, to remove the interruption; and run into the notion of soul, and self, and substance, to disguise the variation. But we may farther observe, that where we do not give rise to such a fiction…we are apt to imagine something unknown and mysterious, connecting the parts, and this I take to be the case with regard to the identity we ascribe to plants and animals….Thus the controversy concerning identity is not merely a dispute of words. For when we attribute identity, in an improper sense, to variable or interrupted objects, our mistake is not confin'd to the expression, but is commonly attended with a fiction, either of something invariable and uninterrupted, or of something mysterious and inexplicable, or at least with a propensity to such fictions. The second group of such metaphysical propositions which are rejected as meaningless are propositions expressing beliefs in the existence of matters with respect to which one can never have an idea, nor even generate a fiction, and, therefore, they are propositions which can never have meaning for us. Such is the case, for example, with regard to the claim that causal powers exist in objects.
'Tis natural for men…to imagine they perceive a connexion betwixt such objects as they have constantly found united together…But philosophers, who abstract from the effects of custom and compare the ideas of objects; immediately perceive the falsehood of these vulgar sentiments, and discover that there is no known connexion among objects… For it being usual, after the frequent use of terms, which are really significant and intelligible, to omit the idea, which we would express by them, and to preserve only the custom, by which we recall the idea at pleasure; so it naturally happens, that after the frequent use of terms, which are wholly insignificant and unintelligible, we fancy them to be on the same footing with the precedent, and to have a secret meaning, which we might discover by reflection. The resemblance of their appearance deceives the mind…and makes us imagine a thorough resemblance and conformity. (T.147-148; S.B. 223-224) Much of the Humean enterprise in the First Book of the Treatise can, therefore, be seen as an attempt to sort out metaphysical synthetic judgements a priori, in order to determine which are intelligible and which are not, and which are merely useful and which are not.
Hume's reliance on the Experimental Method does not commit him to the view that there are no synthetic propositions a priori, or that all such utterances are mere pseudo -propositions. The use of the Experimental Method is viewed by Hume as the technique for determining which propositions are meaningful and which are not, and how belief in such propositions arises.