Forms of Analysis

Since Plato inaugurated conceptual analysis a certain pattern has recurred. His first stab at an analysis of knowledge broke it down into two parts: truth and belief. To know something you had to believe it and it had to be true. Neither element alone was sufficient (though both were necessary) but the conjunction of them added up to knowledge. We have a kind of conceptual equation: xplus yequals z. But then he noticed that this simple combination wasn’t enough for knowledge; it needed an extra ingredient. For it is possible to have true beliefs that aren’t knowledge, as when you accidentally hit on the truth. So he added a further element: justification. Now knowledge is a triadic concept: xplus yplus jequals z. The sufficiency of this was in turn questioned, but let us stop here for the moment. We could say that Plato discovered that truth and belief had to be coordinatedin some way in order to add up to knowledge: you have to believe the truth justifiably(rationally, non-accidentally, for adequate reasons, reliably, etc.). Truth and belief had to be suitably connectednot just exist side by side—you must have the belief becauseof the truth in order to have knowledge. Instead of belief andtruth, we need belief becausetruth. Knowledge breaks into two parts, but the parts don’t just sit there separately; they meld in some way. Knowledge is the kind of belief that resultsfrom truth. Thus a structure emerged: the concept breaks into two basic parts joined in a certain way, where this way features as an extra ingredient added to the basic ones. Knowledge is not a simple thing, but it is not a serial thing either; it is a composite thing—parts coordinated.

This structure is not confined to knowledge. What is perception? It consists of two parts: experience and object. In order to see an object you have to have an experience (a “sense-datum”) andthe experience must be veridical, i.e. there is a suitable object answering to it. You seem to see a table and there is a real table in front of you: neither is sufficient for seeing but if you combine them the upshot is seeing. There are two sides to seeing, as there are two sides to knowing–an internal side and an external side, a subjective side and an objective side. Seeing is a two-factor state, as we can see from conceptually analyzing it. But on further examination we see that seeing must be more than that, because these two conditions are not sufficient for seeing: there needs to be some connectionbetween experience and object; they can’t just be accidentally joined, as when you hallucinate a table but there happens to be a table just where you seem to see one. Thus it becomes natural to require that the two elements be causally connected: the object has to cause the experience. Again, this triadic analysis itself runs into problems of sufficiency, but let’s not be detained by that: what we must note is that perception breaks into two parts and the parts must be properly coordinated. Perception is experience becauseof object. There is an internal side and an external side, along with a relation of dependence. The form is: xbecause of yequals z. This is beginning to sound like a kind of lawof conceptual analysis—a recurrent pattern. And further inquiry confirms that diagnosis: for the same thing is true of memory. To remember a past event is to have both a memory impression and for the past to be a certain way: neither alone is sufficient for remembering but together we get memory. Mind and world supply the necessary ingredients–internal and external, subjective and objective. But again, the two elements cannot merely be conjoined, since you don’t remember something simply because you have a memory impression of it and it actually occurred—that could be so and yet you have completely forgotten the past event (the memory impression has some other source). You have to have the impression becauseof the past event (if you have it because someone randomly stimulates your brain, you don’t really remember). Once again, the concept has the form: xbecause of yequals z. Memory impression because of past event equals memory. Again, problems of sufficiency can be produced, but we won’t go into that. What we can say is that we now have three important concepts whose analysis follows the same pattern—quite an impressive record for the enterprise of conceptual analysis. Our putative law, in brief, then is this: Epistemic concepts break into two coordinated parts. Their analysis has the form: xbecause of yequals z, where xis subjective (internal) and yis objective (external).

Emboldened by this result we might wonder whether other concepts follow the same pattern. In the history of the subject this claim has not been ventured, but I propose to extend the pattern into other areas of the mind. First, and somewhat familiar, there is the concept of action: an action consists of an internal component and an external component, both necessary and (on the face of it) sufficient. To perform an action it is necessary (a) to will it and (b) for a bodily movement to occur, as when I open a car door. I don’t open the door if I merely will it and my body doesn’t move, and similarly if my body causes the door to open but not because of any decision or intention of mine (a sudden spasm, say). Action is willing plus moving—subjective and objective, inner and outer. The concept bifurcates into two. But again, these conditions need to be augmented to deal with a familiar problem, namely that both elements could occur and yet I don’t act. What if I decide to open the door and my body is caused to open it by some accidental event? Then we can’t say that I opened the door: I performed no action, though I tried to and my body did what I was trying to do (because of some random outside stimulus). Again, the cure for this is to require that the agent’s body moved becauseof the internal willing: the willing has to cause the moving. Now the causation is going from inner to outer instead of outer to inner, but the structure is the same: xbecause of yequals z. Moving because of willing equals acting. Again, there are going to be problems of sufficiency (deviant causal chains and so on), but we won’t worry about that here. The important point is that yet another concept falls under our generalization: the concept of acting emerges as a composite concept consisting of two elements, internal and external, joined by a coordinating factor. The mere conjunction of the two elements is never enough; we always need to add the extra ingredient. Is this perhaps the general form of psychological concepts? That would be an interesting discovery in conceptual science, would it not?

One might suppose that it could not be the general form of psychological concepts: for consider belief itself. Is that concept triadic in the way described? Where are the two elements here, and what might coordinate them? We now venture into virgin territory, but not without some prior preparation. Here is an analysis of belief with respectable credentials: For a subject Xto believe that pis for Xto stand in a certain relation Rto a sentence sand for sto mean that p. Intuitively, the subject assents to a sentence in the language of thought that means the content of his belief. For me to believe that the sky is blue is for me to internally assent to the sentence “the sky is blue” (or some synonym) and for that sentence to mean that the sky is blue. Thus belief is assent plus meaning: it is assenting to sentences with propositional content. These are two distinct conceptual elements that together add up to the concept of belief (we are supposing). One is psychological; the other is semantic. If you assented to a meaningless sentence, that would not be a belief, while the mere fact of a sentence meaning something confers no beliefs on anyone. Belief requires both things. But now comes the big question: do we need in addition a coordination condition? Is the mere conjunction enough? That would spoil our generalization (though not entirely), so we anxiously inquire whether our law can be preserved in this case. I think it can be preserved, happily, because the conjunction is notenough, and in a familiar way: you could assent to a sentence that means that pwithout thereby believing that pbecause you might not know what that sentence means. Suppose you are in a foreign country and hear the natives talking: you might accept what they are saying as true, and their sentences certainly have meaning, but you don’t know what they mean, and hence don’t believe what they say. You have to accept what they say becauseof what the sentences mean, not merely because the speakers look like a reliable bunch. You have to understand the sentences, not merely assent to them independently of understanding them. So the conjunction of assent and meaning is not enough.

But what if the sentence occurs in your very own language of thought? Here we must wax more reserche: suppose you have a psychological disability that prevents you from understanding the sentences coded into your genes, yet you have a credulous tendency to assent to these sentences anyway (maybe you think they wouldn’t occur in your mind if they were false, given the ways of natural selection). The sentences have meaning (inherited from your ancestors) but youdon’t grasp this meaning—yet you blithely and blindly assent. Ifthat were possible, this would be a case in which assent to sentences in your own language of thought would not suffice for having the corresponding belief; and conceptually there is clearly daylight here. What is needed to plug the gap is that your understanding of these sentences should play a role in your assent to them: that is, your assent must be becauseof their meaning (among other things). The two factors can’t just operate independently; they must be connected in the right way. Maybe we will find ingenious counterexamples even when this extra condition is added, but again that is not to the point—we have uncovered the same basic pattern in the case of belief too (given the suggested analysis of belief). Belief is assent because of meaning, to put it simply. (This means, of course, that the two-factor concept of knowledge embeds the two-factor concept of belief; or three-factor if we include the coordinating condition.) Belief might have struck us initially as logically simple, but upon analysis we see that it exhibits the same kind of structure that Plato long ago uncovered in the concept of knowledge (it only took us two thousand years). There are two parts to the concept, psychological and semantic, and a condition on their combination; put together we have the composite whole that is the concept of belief (and belief itself). Perhaps we reach conceptual bedrock with the concept of assent, or perhaps not, but there seem to be many ordinary psychological concepts that break down in the way described.[1]Just to have a grand label for our would-be law, let us call it “The Law of Coordinated Duality”, or more colloquially “The Mixed Doubles Law”. It is a law about how psychological concepts are constituted (or some of them), which is to say how the mind is constituted.

What about purely mental actions? Bodily actions divide neatly into two, inner and outer, but what about actions that go on entirely within the mind? Again, we need to get imaginative if we are to discern a comparable structure. Consider mental calculation—calculating in the head. Since this is an action, it is willed—you intend to perform a certain calculation and proceed to do it. But there is also the event of calculation: symbols going through your mind. Someone observing these processes could use them to arrive at the same result you arrive at. So there is a willing and an execution of this willing. You perform the mental act of calculation if both things go on; thus mental action has the same fundamental structure as bodily action. But could there be a case in which the two elements are not properly connected, so that it is false that the person didthe calculation? Imagine an alien scientist who uses your brain as a calculator: he punches in questions and recruits your brain circuits to perform calculations, thus sparing himself the trouble of doing them himself. From the inside you experience symbols passing through your consciousness, but no feeling of willing the process to occur. You feel, as we say, alienatedfrom the calculation, because the alien is willing it not you (compare his causing mental images in your mind against your will). A calculation was occurring in your consciousness, but it wasn’t an action of yours. This is the analogue of the externally imposed bodily movement of opening the car door. Now suppose we add to this scenario your willing to do the calculation, but this willing is not the cause of calculation itself—the cause is still the alien. Intuitively, you still didn’t do the calculation: you willed it and it was done, but you didn’t do it. It just so happened that the alien caused the calculation immediately after you willed it yourself. The two together don’t add to your doing mental arithmetic—the calculation wasn’t your action. What is missing, obviously, is that the calculating didn’t occur becauseof your willing it, but because of the alien. So we need to amend the simple two-factor account by adding that the mental event of calculation was caused by the mental event of willing it. Calculation because of willing equals performing the mental act of calculating. Suppose that the calculation would not have occurred if the alien had lost interest in it, despite the fact that you willed it (maybe your brain’s executive functions are down); then you wouldn’t have done any actual calculating. Adding the alien-caused calculation doesn’t change this; you still didn’t perform the calculation. So again we have the two-factor analysis supplemented with a coordination condition. If you perform a calculation partly in your head and partly on paper, this result is more intuitively obvious, because now we can clearly separate the two side of the action: logically, inner calculating is just like outer calculating. It’s mixed doubles in the head.

Finally, we reach the hardest case: having an experience. Does thisbreak down into two separable components coordinated together? It may not; it may just be primitive (something has to be). At first sight two logically separable elements may be discerned: the experience and the having of it. To have an experience eis for eto exist and for you to have e. Experiencing is an experience andthe having of it. But in this case there seems no logical gap between the experience and the having of it: one entails the other. There is no separating the components, as there is in all the other cases. However, consider this strange scenario: your brain is hooked up to someone else’s brain in such a way that when he has an experience you automatically do, irrespective of what else is happening in your mental life (you know this is the set-up). For example, you have an experience as of a green truck because this other guy sees a green truck (you are at home lying in bed and think, “Oh boy, here we go again!”). The experience occurred in your consciousness but was it your experience? One wants to say that it was his experience intruding on your consciousness; you endured it but you didn’t haveit—it didn’t belong to you. That may sound wrong, because you certainly were the subject of an experience as of a green truck, but the question is whether it was your experience. The case is rather like possession: you are the subject of experiences that belong to the possessing demon, but it doesn’t follow that these experiences are (experienced as) yours—they are the demon’s experiences occurring in you. If it is logically possible for someone else’s experience to occur in you, then we have a possible case in which the experience occurs in you but isn’t had by you in the relevant sense. That would be the logical analogue of truth without belief or object without percept or past event without memory or bodily movement without willing or mental calculation without mental calculating. Conceivably the mind of a baby is like this: experiences occur in its consciousness, but we can’t say that it hasthe experiences, perhaps because a self has not yet fully formed. So there could be experience without the possessing relation holding between it and the subject.[2]Conceptually, it looks as if there is a logical chink here separating an experience occurring and its being possessed by a subject. No doubt this is all very obscure and difficult to pin down, but there is some sense of the kind of structural duality I have discerned. In any case, the matter is worth considering further if we are to determine how far our law of analysis extends. It is possible that the same basic conceptual architecture exists in this case but that it differs in significant ways from case to case. That would certainly be an interesting finding of conceptual science—a kind of structural universal found across a wide range of psychological concepts. Knowledge would then not be a unique case but simply one instance of something much more general. Two factors in combination would be a general feature of mental life.

Colin McGinn
[1]We might say that members of this family of concepts have the same body-plan, to borrow a term from biology—the same architecture, the same geometry.

[2]Could the possessing relation exist without the experience? That would be the logical analogue of belief without fact or percept without object, and so on. It seems hard to make sense of, since it would be the mind shorn of all experience. But maybe it does correspond to some sort of psychological reality in that the mind presumably has a pre-existing capacity to host experiences of different kinds—something like a blank slate. Whether it could exist in a state of pure possessing without anything possessed is hard to contemplate, but conceptually it seems like a distinction exists here. There is the experience and there is the fact that I have it.

Physics and the Physical

It sounds reasonable—indeed tautological–to say that physics is about the physical, as psychology is about the psychological. But that is not clearly true. Consider Newton’s physics: it includes not only physical things in the ordinary sense but also space and time—as well as gravitational force. That last item raised eyebrows at the time owing to its “occult” nature (it wasn’t “mechanical”), but the first two items also raise questions. Are space and time physical? Intuitively they are not, but the question is clouded by lack of clarity about the meaning of the term “physical”. They certainly contrast with chunks of matter in a number of respects, according to Newton: they lack mass, they are not solid and impenetrable, they don’t move around, they can’t be sensed, they are not made of atoms, they have no shape, and they are infinite and eternal.  Space and time contrast rather sharply with matter—they are, if anything, immaterial. Yet they are essential to the way physics understands the world; in particular, they are how motionis defined, i.e. translation of place over time. Newton was not himself a materialist, believing both in the soul and in God, so he had no materialist scruples about accepting this capacious ontology: he had no wish to keep physics physical. He was not a physicalist about physics. Someone claiming to reduce the mind to physics, say, would not be a physicalist under this conception of physics, i.e. someone who believes only in physical things. In physics we have material bodies as well asspace and time (and force), the latter not being physical in the sense applicable to material bodies (mass, solidity, motion, etc.) We could say that, for Newton, the world of physical things exists within a larger world of non-physical things. And these things are not just trivially non-physical (as radiation may be said to be), but fundamentally different in nature from what is physical. The physical thus exists against a background of completely non-physical things. If anything, space and time belong on the side of God, not on the side of matter (consider their infinity and eternality)—at least as Newton sees things.

It is understandable that a physicist with empiricist and physicalist tendencies will balk at Newton’s ontology, because he includes realities that are non-physical and imperceptible. A positivist will be suspicious of such things (see Mach), especially one who wants physics not to stray from the physical. Einstein was just such a physicist: he had decided positivist sympathies and he wanted to find a “physical meaning” (his phrase) for such terms as “time” and “space”. Accordingly, in the special theory of relativity he replaced talk of time with talk of clocks—physical objects in space. No more superordinate time dimension marching on in splendid isolation from the physical world; instead there are just physical clocks and their readings. He physicalizes time; or he eliminates it in favor of clocks. As a result we get the familiar but still startling “discoveries” of special relativity, such as the relativity of simultaneity and “time dilation”. These claims are all really about the behavior of clocks in various conditions of motion. Clocks are finite, perishable, mutable, physical things, unlike the Newtonian time dimension. Motion is understood as change of position correlated with different clock readings, with each object assigned its own clock in Einstein’s thought experiment. There are thus as many “times” as there are assigned clocks, and hence “time” has not the absoluteness we might expect—or better, there are just clocks in this model with time itself eliminated from consideration.

What about space? What “physical meaning” can we give to space? The first thing is to do is make motion relative: objects only move relative to other objects not in relation to absolute space. Thus we replace the impalpable Newtonian spatial dimension with perceptible physical objects—nothing non-physical in the picture. Movement through space becomes change of position relative to a chosen physical object, as time becomes the changing behavior of clocks. Time and space, as Newton understands them, drop out, to be replaced by relations among physical objects. Hence we have physicalism about physics. Moreover, in general relativity space comes to have some of the characteristics of matter: instead of being fixed and unchangeable, it acquires the ability to bend, as steel rods may bend. Note that space doesn’t bend inspace, as if there is some extra spatial dimension behind the curvature of space; rather, space simply isthe collection of all such curves. Just as what we call “time” is regarded as a collection of clocks, so space is regarded as a collection of (gravity-induced) bends—which brings space much closer to matter than it was under the Newtonian conception. Only something physical in nature can literally bend, so we have brought space into the physical arena. The old Newtonian dualism of space and matter has been replaced by a quasi-monism of bendable being. Indeed, one might wonder how space can be curved, with corresponding causal powers, unlessit is a form of matter (of an etiolated kind, to be sure). Thus Einstein physicalizes space in the general theory as he physicalized time in the special theory. Now physics has become comprehensively physical under the new dispensation. All this might sound counterintuitive and confused, but it is the result of ruthlessly physicalizing the Newtonian system. To put it differently, this is what you get if you insist on finding “physical meaning” for the terms of standard pre-Einsteinian physics–you distort their meaning beyond recognition. What implications there might be, or not be, for the nature of matter, motion, space, and time, as they exist in nature, is very much left an open question; one might suppose, none. We have simply stopped talking about one thing (motion through space over time) and started talking about something quite different (clock readings of events in varying relative positions). Likewise, instead of referring to gravitational effects inspace, we describe space as itself curved: light rays bend in space near massive bodies because space itself curves like a physical thing; it isn’t that space remains unchanged while things move differently through it. These are all physicalizing tendencies designed to free physics from the ontological heterogeneity of classical Newtonian physics. They result from adopting physicalism about physics. No doubt this tendency reflects empiricist assumptions, given that the physical is deemed perceptible; so Einstein’s style of physics results from Newton’s physics put through the sieve of empiricism. Otherwise put, it changes the subject.[1]

Can we conclude from this that physics would look very different if it was never subjected to the physicalizing tendencies in question, themselves an offshoot of empiricist epistemology? It appears that we can. Let us imagine a world like the world Newton describes except that we stipulate that there are no physical objects in this world, i.e. no objects with mass, solidity, or perceptibility. There are, however, things that move through space over time—a bit like neutrinos, perhaps. Compared to our weighty solidities these hypothetical things are not material at all—they are wispy penetrable things. They exist in a universe of absolute space and time, which themselves are not physical. There are no sentient beings in this universe, and there are no clocks or observers of any kind. Nevertheless, there are laws of motion, mathematically expressed—let’s say Newton’s laws. Then there is in principle a physics of this world, with forces and equations governing these forces. In this world physics is entirely, not merely partially, non-physical (whatever quite the notion of the physical comes to—I have stipulated possessing mass and solidity for present purposes). That is, there is a science of motion for this world, tailored to the entities it contains; whether there is anything physicalis beside the point. So physics is not essentiallyabout the physical as such; it is about motion in space over time.[2]The label “physics” is therefore misleading in that it suggests that the science in question deals essentially with what is physical (compare “psychology”). We might want to rename it “motion science” or some such. Indeed, our actual physics contains entities often deemed non-physical (in some sense) such as fields of force and certain massless particles. Newton’s physics dealt with the motions of material bodies, but his general framework is not necessarily tied to that ontology. By contrast, Einstein’s conception of physics ties it firmly to the physical, even to the point of physicalizing time and space—or, more accurately, replacing them with surrogates deemed more “meaningful”. I think this was a mistake, but I haven’t attempted to argue that here; my aim has been rather to set out the underlying methodological and metaphysical issues more perspicuously than is usual. The relationship between physics and the physical is actually quite contentious; certainly, we must not assume that what physics deals with is ipso factophysical in any well-defined sense.[3]

Colin McGinn

[1]Here I am influenced by some unpublished work of Randolph Lundberg, though I don’t attribute my conclusions to him.

[2]If we borrowed from Descartes the notion of an immaterial substance, we could specify a world that contains such entities in a state of motion governed by laws. The science of these motions would still be physics.

[3]We may note that, according to the OED, the word “physic” was used in the Middle Ages and later to refer to medicine, and that it comes from the Greek phusismeaning “nature”. Thus it did not originally connote the idea of corporeal matter. It is certainly not analyticthat physics is the study of corporeal matter, which is why it can include what is not material. It was Einstein who edged physics towards the physical, in our modern sense, with his insistence that we find “physical meaning” for terms like “space” and “time”. Under “physical” in the OEDwe find “relating to things perceived through the senses as opposed to the mind; tangible or concrete”. This is the sense that Einstein surely intended, though he says little to clarify his meaning. Very likely he was just taking over Mach’s positivist critique of Newton.

Space, Time, and Matter: A Note

It is sometimes said, jokingly, that it is fortunate time exists or else everything would happen simultaneously. We could also say, jokingly, that it is fortunate space exists or else everything would have to be in the same place. If space existed but time didn’t, different events could occur at distinct locations but simultaneously, thus allowing for a multiplicity of events; while if time existed but space didn’t, different objects could exist at distinct times but all in the same place, thus allowing for a multiplicity of objects. But if neither space nor time existed, then everything would have to occupy a single place at a single time. All of the physical universe would have to exist at one place, and all of history would have to occur at one time. Of course, this is absurd: there could not be such multiplicities at a single place and time (a space-time point). It is the existence of space and time that makes the multiplicity of objects and events possible; they make room for the Many. When God created space and time he did so in order to allow for the possibility of many objects and many events—since you can’t have the latter without the former.

Part of the humor in the remark is that if there were no space and no time then there would not be oneplace and onetime: there would be noplaces and notimes. If that were so, there would be noobjects and noevents either—not the totality of them squeezed into one spatiotemporal location. Space and time are the preconditions for objects and events to exist at all. But the remark raises the possibility of there being a single place and a single time—which itself sounds fishy. Could there be just one place and one time in isolation? Could there be a world in which the area occupied by my head now exists but no other area of space exists, and in which last Tuesday exists but no other stretch of time? There is just one cubic foot of space and twenty-four hours of time, with nothing, so to speak, on either side.

Surely that is absurd: these entities must be embedded in a wider totality of places and times—the things we call Space and Time. Concerning this particular unit of space and time, itcould not exist without other units of space and time. Indeed, it is plausible to claim that it could not exist without allof space and time. For thispart of space and time to exist therestof space and time must exist. If space and time are infinite as a matter of metaphysical necessity, then the existence of any portion of them implies the existence of an infinite dimension: the being of one unit of space or time requires the being of infinitely many such units. This is a kind of extreme holism, analogous to the (alleged) holism of belief: just as one belief cannot exist in isolation, but requires the existence of many other beliefs, so one place or one time cannot exist in isolation, but require the existence of (infinitely) many other places or times. Places and times are essentially parts of a spatial and temporal totality, not isolable detachable atoms. They cannot wander off on their own.

What about matter? Here we find a sharp contrast: for units of matter canexist in lonely isolation. A particular collection of particles, say a small meteorite, could exist without any other matter existing—it doesn’t need to be surrounded by other small meteorites. Concerning this particular collection, itcould exist without any of the rest of the matter of the universe existing. There is a possible world in which all of the actual world’s matter has been removed leaving only this solitary piece. Bits of matter don’t imply other bits of matter. Bits of matter are not essentially parts of a totality of matter, extending outwards indefinitely. They are not subject to the “holism of the material”, analogous to the “holism of the spatial and temporal”. There is thus a deep ontological difference between matter, on the one hand, and space and time, on the other. Therefore matter cannot be some kind of configuration or condensation of space and time; and space and time cannot be some kind of rarefaction of matter. We cannot reduce one category of being to the other category: matter is not space and time congealed, and space and time are not matter etiolated.[1]

Matter is conceived as an occupantof space and time—something that “takes up” space and time. Matter is not essentially joined to other matter: one occupant does not entail the existence of another occupant. The occupants are separate existences, not just in being numerically distinct, but also in not being necessarily connected. But units of space and time are not separable occupants, modally unconnected: they are essentially elements of a whole and cannot be removed from that whole. It is not an essential property of any piece of matter that neighboring pieces of matter should exist (the existence of Earth does not logically require the existence of Mars), but it is an essential property of places and times that neighboring places and times should exist (the place of my head now implies the existence of contiguous places, and last Tuesday requires Monday and Wednesday). Places don’t occupyspace and moments don’t occupytime: consequently they cannot be sundered from other such occupants. Space and time don’t necessarily contain specific bits of matter, but they do necessarily contain specific bits of space and time—they are made up of places and times. Places and times are not atoms of space and time in the way physical particles are atoms of matter: the latter can be modally separated from each other, but the former cannot. Time is never short and space is never narrow, but matter can be rare and small.

Colin McG

[1]This poses a problem for Descartes’ view of space as a material plenum: space cannot be a form of matter or else it would not be subject to the holism of the spatial

A Problem in Hume

Early in the TreatiseHume sets out to establish what he calls a “general proposition”, namely: “That all our simple ideas in their first appearance are deriv’d from simple impressions, which are correspondent to them, and which they exactly represent” (Book I, Section I, p.52).[1]What kind of proposition is this? It is evidently a causal proposition, to the effect that ideas are caused by impressions, and not vice versa: the word “deriv’d” indicates causality. So Hume’s general proposition concerns a type of mental causation linking impressions and ideas; accordingly, it states a psychological causal law. It is not like a mathematical generalization that expresses mere “relations of ideas”, so it is not known a priori. As if to confirm this interpretation of his meaning, Hume goes on to say:  “The constant conjunction of our resembling perceptions [impressions and ideas], is a convincing proof, that the one are the causes of the other; and this priority of the impressions is an equal proof, that our impressions are the causes of our ideas, not our ideas of our impressions” (p. 53). Thus we observe the constant conjunction of impressions and ideas, as well as the temporal priority of impressions over ideas, and we infer that the two are causally connected, with impressions doing the causing. In Hume’s terminology, we believe his general proposition on the basis of “experience”—our experience of constant conjunction.

But this means that Hume’s own critique of causal belief applies to his guiding principle. In brief: our causal beliefs are not based on insight into the real powers of cause and effect but on mere constant conjunctions that could easily have been otherwise, and which interact with our instincts to produce non-rational beliefs of an inductive nature. It is like our knowledge of the actions of colliding billiard balls: the real powers are hidden and our experience of objects is consistent with anything following anything; we are merely brought by custom and instinct to expect a particular type of effect when we experience a constant conjunction (and not otherwise). Thus induction is not an affair of reason but of our animal nature (animals too form expectations based on nothing more than constant conjunction). Skepticism regarding our inductive inferences is therefore indicated: induction has no rational foundation. For example, prior to our experience of constant conjunction ideas might be the cause of impressions, or ideas might have no cause, or the impression of red might cause the idea of blue, or impressions might cause heart palpitations. We observe no “necessary connexion” between cause and effect and associate the two only by experience of regularity—which might break down at any moment. Impressions have caused ideas so far but we have no reason to suppose that they will continue to do so—any more than we have reason to expect billiard balls to impart motion as they have hitherto. Hume’s general proposition is an inductive generalization and hence falls under his strictures regarding our causal knowledge (so called); in particular, it is believed on instinct not reason.

Why is this a problem for Hume? Because his own philosophy is based on a principle that he himself is committed to regarding as irrational—mere custom, animal instinct, blind acceptance. He accepts a principle—a crucial principle–that he has no reasonto accept. It might be that the idea of necessary connexion, say, is an exception to the generalization Hume has arrived at on the basis of his experience of constant conjunction between impressions and ideas—the equivalent of a black swan. Nothing in our experience can logically rule out such an exception, so we cannot exclude the idea based on anything we have observed. The missing shade of blue might also simply be an instance in which the generalization breaks down. There is no necessityin the general proposition Hume seeks to establish, by his own lights–at any rate, no necessity we can know about. Hume’s philosophy is therefore self-refuting. His fundamental empiricist principle—all ideas are derived from impressions—is unjustifiable given his skepticism about induction. Maybe we can’t helpaccepting his principle, but that is just a matter of our animal tendencies not a reflection of any foundation in reason. It is just that when we encounter an idea our mind suggests the existence of a corresponding impression because that is what we have experienced so far—we expectto find an impression. But that is not a rational expectation, merely the operation of brute instinct. Hume’s entire philosophy thus rests on a principle that he himself regards as embodying an invalid inference.

It is remarkable that Hume uses the word “proof” as he does in the passage quoted above: he says there that the constant conjunction of impressions and ideas gives us “convincing proof” that there is a causal relation that can be relied on in new cases. Where else would Hume say that constant conjunction gives us “convincing proof” of a causal generalization? His entire position is that constant conjunction gives us no such “proof” but only inclines us by instinct to have certain psychological expectations. And it is noteworthy that in the Enquiry, the more mature work, he drops all such talk of constant conjunction, causality, and proof in relation to his basic empiricist principle, speaking merely of ideas as “derived” from impressions. But we are still entitled to ask what manner of relation this derivation is, and it is hard to see how it could be anything but causality given Hume’s general outlook. Did he come to see the basic incoherence of his philosophy and seek to paper over the problem? He certainly never directly confronts the question of whether his principle is an inductive causal generalization, and hence is subject to Humean scruples about such generalizations.

It is clear from the way he writes that Hume does not regard his principle as a fallible inference from constant conjunctions with no force beyond what experience has so far provided. He seems to suppose that it is something like a conceptual or necessary truth: there couldnot be a simple idea that arose spontaneously without the help of an antecedent sensory impression—as (to use his own example) a blind man necessarily cannot have ideas of color. The trouble is that nothing in his official philosophy allows him to assert such a thing: there are only “relations of ideas” and “matters of fact”, with causal knowledge based on nothing but “experience”. His principle has to be a causal generalization, according to his own standards, and yet to admit that is to undermine its power to do the work Hume requires of it. Why shouldn’t the ideas of space, time, number, body, self, and necessity all be exceptions to a generalization based on a past constant conjunction of impressions and ideas? Sometimes ideas are copies of impressions but sometimes they may not be—there is no a priori necessity about the link. That is precisely what a rationalist like Descartes or Leibniz will insist: there are many simple ideas that don’t stem from impressions; it is simply a bad induction to suppose otherwise.

According to Hume’s general theory of causation, we import the idea of necessary connexion from somewhere “extraneous and foreign”[2]to the causal relation itself, i.e. from the mind’s instinctual tendency to project constant conjunctions. This point should apply as much to his general proposition about ideas and impressions as to any other causal statement: but then his philosophy rests upon the same fallacy–he has attributed to his principle a necessity that arises from within his own mind. He should regard the principle as recording nothing more than a constant conjunction that he has so far observed, so that his philosophy might collapse at any time. Maybe tomorrow ideas will notbe caused by impressions but arise in the mind ab initio. Nowhere does Hume ever confront such a possibility, but it is what his general position commits him to.

Colin Mc

[1]David Hume, A Treatise of Human Nature(Penguin Books, 1969; originally published 1739).

[2]The phrase is from Section VII, [26], p. 56 of An Enquiry Concerning Human Understanding(Oxford University Press, 2007).