Hume, Wittgenstein, and Kripke
In An Enquiry Concerning Human Understanding Hume writes as follows: “It must certainly be allowed, that nature has kept us at a great distance from all her secrets, and has afforded us only the knowledge of a few superficial qualities of objects; while she conceals from us those powers and principles, on which the influence of these objects entirely depends. Our senses inform us of the colour, weight, and consistence of bread; but neither sense nor reason can ever inform us of those qualities, which fit it for the nourishment and support of a human body. Sight or feeling conveys an idea of the actual motion of bodies; but as to that wonderful force or power, which carry on a moving body for ever in a continued change of place, and which bodies never lose but by communicating it to others, of this we cannot form the most distant conception. But notwithstanding the ignorance of natural powers and principles, we always presume, when we see like sensible qualities, that they have like secret powers, and expect, that effects, similar to those which we have experienced, will follow them…. It is allowed on all hands, that there is no known connexion between the sensible qualities and the secret powers; and consequently, that the mind is not led to form such a conclusion concerning their constant and regular conjunction, by anything which it knows of their nature.” (Section VI, [16])
Later he writes: “When we look about us towards external objects, and consider the operation of causes, we are never able in a single instance, to discover any power or necessary connexion; any quality, which binds the effect to the cause, and renders the one an infallible consequence of the other. We only find, that the one does actually, in fact, follow the other.” (Section VII, [6]) Again: “The scenes of the universe are continually shifting, and one object follows another in an uninterrupted succession; but the power or force, which actuates the whole machine, is entirely concealed from us, and never discovers itself in any of the sensible qualities of the body.” (Section VII, [8]) And again: “And experience only teaches us, how one event constantly follows another; without instructing us in the secret connexion, which binds them together, and renders them inseparable.” (Section VII, [13]) More: “Can there be a more certain proof, that the power, by which this whole operation [voluntary motion] is performed, so far from being directly and fully known by an inward sentiment or consciousness, is, to the last degree, mysterious and unintelligible.” (Section VII, [14]) Also: “We are ignorant, it is true, of the manner in which bodies operate on each other. Their force or energy is entirely incomprehensible.” (Section VII, [25]) As a result of our ignorance “so imperfect are the ideas which we form concerning it [causation], that it is impossible to give any just definition of cause, except what is drawn from something that is extraneous and foreign to it.” (Section VII, [29]) Referring to his own attempts to define cause, he confides: “But though both these definitions be drawn from circumstances foreign to the cause, we cannot remedy this inconvenience, or attain any more perfect definition, which may point out that circumstance in the cause, which gives it a connexion with its effect. We have no idea of this connexion; nor even any distinct notion what it is we desire to know, when we endeavour at a conception of it.”(Section VII, [29])
Hume’s position seems clear enough, being a conjunction of three main theses: (a) causation consists in the objective presence of causal power (“necessary connexion”) in the objects; (b) we have no direct knowledge of the nature of such power (no “impression” of it); and (c) all we know of causation is constant conjunction, which is “extraneous and foreign” to causation itself. It is constant conjunction, combined with our natural instinct to project it into the future, that furnishes us with the only notion of causation that we possess—a notion that fails to capture the actual nature of the causal relation. Causation essentially involves necessary connexion between individual causes and effects, but our knowledge of it is confined to a mere symptom of such connexion, viz. constant conjunction. We don’t grasp what causation objectively is, though we have causal beliefs as a result of natural propensity—as do children and animals. As Hume sometimes puts it, we don’t reason from cause to effect, but merely find ourselves inferring effects from causes as a result of our innate disposition. His position might be put this way: God has knowledge of the objective nature of causation (those actuating powers dwelling in objects) and can therefore reason a priori from cause to effect, but we have no such penetrating knowledge and hence no basis from which to reason in this way. We rely in daily life on causal inference but we have no understanding of the nature of that on which we rely—we just blindly follow the regularities we have observed.
Now I don’t wish to go into a defense of this interpretation of Hume here (others have done so), though I think it is clearly correct; my aim is to use it in application to Kripke’s discussion of Wittgenstein’s views on meaning. [1] Kripke compares what he takes to be Wittgenstein’s view of meaning with what he takes to be Hume’s view of causation, hoping thereby to illuminate Wittgenstein’s tantalizing discussion. I disagree with both of his interpretations as interpretations, but that is not my current concern; I want to develop a new and better comparison between Hume and Wittgenstein. Kripke contends that Wittgenstein’s position is that there is no individual fact of meaning but that meaning ascriptions have social assertion conditions. Analogously, Hume is said to hold that there is no individual fact of causation—nothing in the cause-effect pair considered intrinsically that could constitute causation—but that causal statements have assertion conditions that pertain to constant conjunctions, so that causation is a “social” matter. Thus there cannot be purely individual rule following and there cannot be one-off causal relations: meaning and causation are not located in individuals but in collections—societies of speakers or groups of events. We are tempted to believe that meaning is an individual matter and that causation resides in the particular events that are causally connected, but in reality there are no such localized facts—both things involve wider collections of speakers or events (via assertion conditions). They are extrinsic and plural.
Obviously the interpretation of Hume that I gave earlier contradicts these ideas: there is an individual fact of the matter about causation—actual local necessary connexion—but we have no cognitive access to it, no knowledge of it, and no definition of it. It exists all right but it doesn’t “discover” itself to us—so we can’t form an “adequate idea” of it. All we can do is follow our natural instinct to make causal inferences based on constant conjunction, even though this is “extraneous and foreign” to causation in itself. Our notion of necessary connexion, such as it is, arises through this natural instinct (what Hume calls “custom”) and not through any kind of rational examination of individual causes and effects—for we have no “impression” of necessary connexion as it exists in objects. The question I want to ask is what an analogous theory of meaning would look like: What should we say about meaning if we model it on what the real Hume says about causation? Is there anything to be said in favor of such a theory? What if we become Humeans (real Humeans) about meaning?
A real Humean about meaning accepts the following three propositions: (a) there is an objective individual fact of meaning, involving “normative connexion” (cf. necessary connexion); (b) we have no direct knowledge of this fact, no “impression” of it, and no adequate definition of it; and (c) our notion of meaning, such as it is, derives from our observation of symptoms of meaning that are “extraneous and foreign”. Meaning is thus something that undoubtedly exists in the objects (human speakers), but it doesn’t present itself to us; our idea of meaning is solely derived from circumstances merely associated with meaning. Meaning as it is in itself is “mysterious”, “incomprehensible”, and “secret”. Nature “has kept us at a great distance” from meaning, which counts among its many secrets; yet we do mean things and we can even talk about meaning. We just don’t have the kind of knowledgeof meaning that we have of other things: our idea of it is not derived from acquaintance with it—we believe in its existence but we can’t comprehend its inner nature. It is objectively real but hidden.
Hume thinks that constant conjunction results from necessary connexion—indeed, necessary connexion is the cause of constant conjunction. Constant conjunction, however, is all we are acquainted with and all that we know of causation—yet it is “extraneous and foreign”. What is the analogue in the case of meaning? Here we cannot do better than to follow Kripke’s discussion: external linguistic behavior, inner states of consciousness, and dispositions. None of these can be said to constitute meaning, though they are certainly closely “associated” with it. This is all we directly know of meaning, which lies “behind” these symptoms, and it is what enables us to talk meaningfully about meaning—but it is not what meaning is. We don’t know what meaning is, as we don’t know what causation is. We may have sensations associated with meaning, as we have sensations associated with causation—inner and outer sensations in both cases—but they are not the essence of meaning. Observing someone’s linguistic behavior over time is no more observing their meanings than observing constant conjunctions over time is observing causation. Meaning and causation transcend such observable matters. Behavior is to meaning as constant conjunction is to causal power—related but not constitutive. You can examine behavior from dawn till dusk and you will not see meaning there; and the same is true of mental images, sudden feelings of understanding, or sensations of hesitation (as the meaning of “or”). Meaning is simply not given in this way, as causation is not: it cannot be seen with the inner eye any more than causation can be seen with the outer eye.
We can compare this view of meaning and causation with another issue discussed by Hume: the self. A analogous Humean about the self holds (a) that the self undoubtedly exists as one and the same entity over time, despite psychological fluctuations; (b) that we have no direct knowledge of the self, no “impression” of it, and no definition of it; and (c) that we talk of the self by adverting to what are merely extraneous symptoms (the body, mental states). Our natural instincts lead us to believe in a continuing self, even though we have no “impression” of such continuity and no real conception of that of which we speak; our idea of the self, such as it is, doesn’t come from up-close confrontation with the self, or from a priori reasoning, but just from a natural tendency to assume personal continuity beneath psychological fluctuation. We are ignorant of the inner being of the self, but we compensate for that ignorance by blindly following our instinctive inclinations. Similarly, the real nature of meaning is hidden from us, because of what Hume describes as the “surprising ignorance and weakness of the understanding” (55). Our ignorance should on no account be confused with non-existence: there is a hard fact of meaning, as of causation and the self, but it happens not to be presented to the (“weak”) human understanding. Hume’s question was how we could have an idea of causation if we have no acquaintance with it, and this is what led to his “skeptical solution” in terms of custom and instinct. In the case of meaning we have the same question, and the analogous answer is that we are acquainted with the symptoms of meaning, though not with the thing itself, which allows us to possess a working concept of meaning (inadequate as that concept is).
It is an interesting question whether there is more than an analogy between causation and meaning. Might we not think of meaning as a kind of inner causal power? What we mean governs how we use words—the ability to mean is an ability to speak. Linguistic competence is a power to produce certain kinds of effect—meaningful utterances appropriate to the occasion (“performance”). So it is a special case of a causal power; and we don’t grasp the nature of causal power, according to Hume’s view of causation. Meaning is a kind of norm-governed potentiality—a source of unlimited linguistic use. How meaning is able to achieve this remains difficult to fathom, but it does. So maybe we should think of use (performance) as effect and meaning (competence) as cause, and then note that causal relations are ultimately impenetrable. And in addition to that general opacity, there is the problem of understanding the nature of semantic norms—how use conforms to meaning. In virtue of what is one use correct and another incorrect?
Why might we jib at this Humean picture of meaning (a form of agnostic realism)? I think it’s because we have a tendency to suppose that we know more about meaning than we really do, as Hume remarked regarding causation—we overestimate our degree of insight. But why is that? In the case of causation we have been long drilled in the practice of following constant conjunctions: that works fine for most purposes, so we don’t notice how shaky are its foundations (until someone like Hume brings up the problem of induction). We have adequate practical knowledge, and (as Hume also remarks) nothing can disrupt our instinctive tendencies. We therefore suppose that we understand more about causation than we do. In the case of meaning the source of complacency is simply that we have first-person knowledge of meaning: I know what I mean by my words. But that doesn’t imply that I know the nature of meaning in its deep essence. It is just the reflection of a practice that works and which is underpinned by instinct. I cannot gaze at my meanings and drink in their essence—as I can gaze at my pains and grasp their nature. As Wittgenstein insisted, meaning and understanding are nothing like experience and sensation—they are not “contents of consciousness”. Meanings are elusive, slippery, hard to pin down. Hence they baffle us—and hence there is an inclination towards skepticism about their very existence. Just as many people see in Hume’s reflections about the impossibility of sensing necessary connexion a reason to doubt that necessary connexion exists, so many people find the elusiveness of meaning a reason to doubt the existence of meaning. In both cases they are illicitly drawing a metaphysical conclusion from an epistemological premise—something that Hume himself emphatically does not do.
What goes for meaning presumably goes for concepts: these too are intrinsically unknowable (directly), though they are real enough, and we form ideas of them by recourse to their symptoms. What are their symptoms? Chiefly, they are thoughts: we ascribe concepts to ourselves by ascribing thoughts to ourselves; and we never encounter a concept except in the context of a thought. Thoughts are to concepts as constant conjunction is to causation—and as use is to meaning. Concepts are what make thought possible, as causation is the foundation of regularity: but both are elusive to direct detection. When was the last time you introspected a concept? [2] Concepts bear an obscure relation to thought, as meaning bears an obscure relation to use: in neither case should we try to reduce the one to the other. We can’t inspect our concepts per se (have “impressions” of them), though we know they are there, just as we can’t inspect causal necessity, though we don’t doubt its existence.
A Humean view of certain facets of reality is evidently consistent, but is it attractive? It is certainly radical, as Hume recognized with respect to his view of causation; it is not what we naturally assume. We tend to think that we can observe necessity in the objects, or else we suppose that if we can’t that must prove non-existence, leaving only constant conjunction. The idea that causal necessity is real but inaccessible, and that our causal inferences arise from animal instinct not reason, is repugnant to us. Similarly, we tend to assume that meaning is open to introspection, or can be heard in words as sounds are heard. Once this naive position is called into question there a tendency to lapse into skepticism about meaning. The idea that meaning is real but inaccessible, and that our talk of meaning is derived only from extraneous symptoms of it, is also repugnant to us. The same is true of the self, where again it is hard to accept inaccessible realities combined with a non-rationalistic account of our discourse about selves. Still, there is a lot to be said for this kind of view, especially in the light of the troubles occasioned by other views. The idea that our practice might be sustained by something other than direct knowledge of the thing in question, something merely instinctual, is particularly striking. Constant conjunction is quite far removed from the real nature of causation, and mere custom is not what we expect from causal inference; but that is what our ideas of causation ultimately involve, if we follow Hume. This is what he calls a “skeptical solution”—“skeptical” because it finds no rational basis for ordinary belief (hence the problem of induction) and a “solution” because it tells us where our concept comes from and why the practice works. An analogous view of meaning has it that our practice of speaking of meaning is one step removed from meaning itself, which remains elusive, hidden, and mysterious; instead we go by observable symptoms extraneous to meaning. This is a skeptical solution because it suggests that our talk of meaning does not result from rational insight into the nature of meaning but from non-rational natural facts.
Hume believes that causality (“in the objects”) is fully present in individual instances, so that the effect could be inferred from the cause alone had we the requisite knowledge of the cause. Constant conjunction does not create causal connexion, and indeed can never establish connexion as opposed to mere conjunction. There is, for Hume, nothing to prevent a single instance of causation that is never repeated. And yet we have no knowledge of causation except by experiencing constant conjunctions and feeling the tug of instinct to venture beyond them. In the case of meaning an analogous view would insist that meaning is present whole and entire in every act of meaning, with regular linguistic use a mere contingent accompaniment. And yet we cannot form an idea of meaning except by recourse to what is contingent and dispensable (behavior or states of consciousness), and then letting nature do its work. Meaning in itself is real, singular, and inherent, but our knowledge of it perforce relies upon symptoms that don’t do justice to its nature. If causal powers could look at constant conjunctions, they would not recognize themselves therein; likewise, if meanings could look at linguistic behavior and states of linguistic consciousness, they would not recognize themselves therein. Causation is really nothing like constant conjunction, and meaning is really nothing like its symptoms (inner or outer); yet we are condemned to conceive of them both in terms “extraneous and foreign”. [3]
[1] Saul Kripke, Wittgenstein on Rules and Private Language (Basil Blackwell: Oxford, 1982). I discuss this book in Wittgenstein on Meaning (Basil Blackwell: Oxford, 1984).
[2] I mean a concept in isolation, not connected to other concepts. It is true that we can try to analyze a concept, but when we do we position it within a thought—we can’t just gaze at it independently of the thoughts in which it might occur. Concepts come to us as constituents of propositions (this is a version of Frege’s context principle). We can’t be conscious of concepts as pre-propositional units.
[3] The same thing cannot be said of consciousness or typical conscious states: here we do have an “adequate idea” of that of which we speak. In Hume’s language, we understand the nature of “impressions and ideas”—these are not hidden like causation or the self.
