Hume has somewhat become a victim of his own success. By this, I only mean that when one reads the Treatise or Enquiry, there’s an incredible sense of obviousness about what he is pointing out. So far have we become accustomed to empiricism, the importance of observation, etc, that one can be forgiven for thinking perhaps we have no need for Hume any longer. His work is done.
However, an interesting moment happens on the final page of the Treatise of Human Nature. Hume suddenly realizes something very strange has emerged from his arguments, specifically his arguments concerning personal identity, the nature of the ‘self’, which he admits to having no idea how to grapple with – he leaves it as an exercise to the reader or future thinkers (himself in the future included).
“In short there are two principles, which I cannot render consistent; nor is it in my power to renounce either of them, viz. that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences. … For my part, I must plead the privilege of a sceptic, and confess, that this difficulty is too hard for my understanding. … Others, perhaps, or myself, upon more mature reflections, may discover some hypothesis, that will reconcile those contradictions.” (Treatise, p.678.)1placeholder
Though the concern emerges from the personal identity arguments that conclude book I, they are prefigured by the very opening moves of the Treatise – given everything that had been developed prior, there is nothing shocking or new introduced in the relevant section. Given this, perhaps we can explain Hume’s reticence to revise or delay the manuscript by some lethargy on his part – having made it so far only to have to go back and make alterations and redevelop ideas. So, instead he just adds a note in the appendix admitting to the problem, and then washes his hands of it. However, and this is what I want to explore here, if we work to bring the problem to the surface, and grasp it qua problem with Hume’s basic foundational arguments, we realize how far beyond our common sense intuitions concerning empiricism these foundational arguments are. That is to say, we can read from Hume’s admission, that the problem appears insoluble to him, a certain way of understanding what appear to us as relatively mild premises, but which, through being related to this problem, take on radical import.
So, let’s do that.
⋅ ⋅ ⋅
To begin at the beginning, Hume divides the contents of our minds, all that we perceive, into two kinds of things; impressions and ideas. These he distinguishes firstly by their degree of intensity. Impressions are the intense, bright, violent, lively ones, and ideas are the faint, dim, light, and airy ones. So, of everything that we are aware, some of it is quite bright and intense, some of it less so. This is so far purely definitional, and Hume thinks the distinction should be readily apparent to anyone. After all, we can all readily tell the difference between a pumpkin and the idea of a pumpkin when we entertain one then the other. The pumpkin we ‘see’ is, for starters, a brighter and more vivid orange. It’s the kind of thing you could sit down and sketch, pick up and eat, play like a drum. The idea of a pumpkin is kind of flickering, and thin. If you attempt to draw it you are prone to being dissatisfied with the result, but unable to clearly cross-reference the drawing you are now seeing with the idea floating elusively around. This definition is phenomenological in nature, it asks us to direct our attention to our experience and ‘weigh’ the relative experiential ‘intensities’ of the components of the fullness of that experience.
From a definition to an observation: impressions and ideas are not just distinguished by their degree of intensity for the purposes of identification – it seems that this really is their chief distinguishing feature: to wit, almost every impression seems to have a corresponding idea, and vice versa, differentiated by their relative intensities. Impressions and ideas form two related series. It’s not the case that there are two isolated realms – the realm of concepts and the realm of sensory experience, but rather that this material, distributed along the intensive scale, runs through and through every experience. Descartes sits at his desk, melting wax, watching the relation between impressions and ideas in that experience, eventually siding with the idea of the wax insofar as the impressions are punctured through with chaotic transformations, and his questions was “is this the same wax?” Only the idea seemed sufficient to answer to that identity. The impressions are never the same from one moment to the next.
So, almost everywhere impressions and ideas connect. The question then, is which is original and which is derived? The answer is that the origin of every idea is to be found in some past impression, and never the other way around. Given the ‘intensity’ condition, it seems the deck has already been stacked against ideas. Hume has an argument though: for every idea he has examined, there is always some trace of what was once before an impression. That is, these examined ideas seem to have their origins (or their constituents’ origins) explained through the past presence on an impression. But the reverse, impressions originating in ideas, doesn’t hold. Because we can’t get an ‘every’ from the examination of any quantity, he just puts forward a challenge: find an idea that has no corresponding trace of an impression. We only need one to refute the ‘every’. Then, travelling in the other direction, Hume argues that where a certain class of impressions are missing, the corresponding ideas will be taken to be inconceivable too: e.g., in the case of blindness, ideas of color.
However, that being said, there is another recommending feature of impressions in the question of origins and primacy. Even Descartes noted that even though inferences we draw from our present perceptions can be doubted (such as my inference that my perception of a chair-like body justifies me in believing there really is a chair over there), the pure self-presence of the perceptions cannot. When I look at the rotating snakes illusion, I would be wrong to conclude the paper is moving, but I cannot be wrong about the fact that it is currently appearing to move. This immanent, indubitable and grounded feature of our present perceptions, the lively impressions, needs to be bundled into Hume’s talk of the intensity and vibrance of impressions versus ideas. As far as starting points go, it seems a reasonable move to prioritize this self-present, immanent surety of impressions as ground, and work from there. The alternative is to treat the a priori feature of ideas as the hard kernel upon which to build, but these projects always ended up needing to invoke God to end regresses, or return us to the simplicity of the immanent. Hume may or may not have had this in mind, but it is worth mentioning, as part of the strange difficulties he gets himself into lie in the fact that he begins firmly from the immanent, and then must derive the a priori, rather than the other way around.
So, to briefly conclude the above, Hume’s project begins with classifying all that we are aware of into impressions and ideas, distinguished by their degree of intensity and liveliness, and with ideas being derived from impressions. This, then, gives us a criteria that Hume will endlessly apply moving forward: for every idea, we may ask from which impression is it derived? If one is not to be found then the idea is more or less a groundless fancy2placeholder. This not to say that we have had to have seen everything about which we think, but rather it should be possible, with the ideas that we think of, to trace the origins of their constituents back to some impression that could ground them. We can entertain the idea of a golden mountain, despite not being able to have the requisite impressions of such an object, because the idea itself is composite, and its parts we can ground in previous impressions.
⋅ ⋅ ⋅
That criteria being established, Hume then applies it to the relations proper to ideas: those that allow us to go beyond an insect-like recognition of the present given, and come to know the world, to go beyond the particular and grasp things like physical laws.
Ideas are related, without this thought would be impossible – if every idea has an impression as its origin, and this admits of complex constitutions, then it’s not the case that we are resigned to only thinking about objects that we have encountered previously. An object, like an apple, is already a complex bundle of impressions, and thought can isolate these impressions, like when we identify or imagine the flavor of an apple, or a particular hue of an apple as a color of a car. But given thought’s ability to i) form complex ideas on the basis of heterogenous impressions, ii) break down these complex ideas into their constituent simples, and then iii) recombine these simple ideas with others, it is amazing that we can think anything coherent at all. I.e., why isn’t our thought a kind of entropic noise of multitudinous deconstructed atoms? It isn’t, so there needs to be some contours, attractors, within the surgical operation of ideas. Hume lists these as resemblance, contiguity, and cause and effect.
So, to take our apple example again, as I see the light diffraction pattern across the surface of the apple, every different shade of green that I can isolate I can individuate as a simple idea, and the origin of these ideas will be the simple impressions that cluster together to form the apple in front of me. However, each shade of green is related to the others by a resemblance (the impressions are all colors, and similar colors at that), contiguity3placeholder (each shade of green is only imperceptibly different to its left and right neighbor) and cause and effect (the light diffraction pattern on a polished sphere-approximate). It is not the case that it is difficult for me to bring to my imagination this collection of green hues, as I do when I imagine a car which is the green of an apple, in fact the inverse is true: even though I can isolate each hue, it is difficult for me to combine them chaotically and pixel-like with other, randomly selected, hues, or even other sensations. It is easier to imagine a centaur, who is half of a horse attached to half of a man, than it is to break down our ideas of both man and horse into tiny, one centimeter, pieces, and imagine some other random way, patchwork quilt-like, of combining them into a static glitched horse-man.
Hume is quick to point out that the way ideas ‘hang together’ through these associative forces is not some fixed law of ideas, but merely a tendency, an inner torque that determines ease and difficulty, with our mind more often than not following the easier path. He is also quick to point out that it is the relation of cause and effect that provides the strongest connection in all of our ideas about matters of fact. The relations of resemblance and contiguity we could say give us a power of recognition, but understanding almost always involves the relating of disparate things into causal series.
So, the question then becomes where do we get our confidence in this one specific relation between ideas, keeping in mind the criteria discovered above? We can think the idea of resemblance and contiguity in general from a sort of ‘atmosphere’ among our impressions: though I have no direct impression of contiguity in addition to the impressions I am aware of, those impressions just simply are contiguous to one another. However, we cannot apply this same loose reasoning to our general understanding of the cause-and-effect relation. There are never impressions of causal powers, merely successive collections of impressions (succession being a kind of temporal contiguity). One ball moves on up to the other, there’s a tapping sound, and the second ball begins to move. We cannot derive a causal conclusion from the contiguity of one ball to the other, nor the resemblance between the first ball’s motion and the second ball’s. These might be a part of our causal understanding, our ideas concerning causes and effects, but are not sufficient to it – there is, in addition, a sense of necessity we attribute to causal powers, this is what makes our ideas concerning causality predictive and not just descriptive. It is not enough to point to contiguity and resemblance: our understanding moves from the impressions of one ball moving on its path firmly to the idea of the second ball being jolted to action upon the collision. But there is no impression to correlate to this idea beyond those concerning the mere movement of the balls. Furthermore, we could argue, even if there was – a ‘flash’ of necessary causal connection, or a unique sense providing additional impressions, then this does not help us at all, as this new impression is in no better position than the impressions of the movements and the sound of the collision, it would just be one other event thought within the ideational causal nexus.
It’s clear what dimension of experience is lacking for our ideas, in our witnessing causally linked events, from Hume’s examination (and rejection) of one possibility: that we do have direct impressions of causes when we ourselves will our bodies to move, or our thoughts to change from one thing to another. Hume offers a number of counter-arguments to this attempt, all similar in form. Had we some direct impression of causal power here, then we would as easily know the causality involved between willing and moving as we know the color of an apple we are looking right at – that is to say, instantly and directly. But we don’t, so we don’t. The fact that the mind-body problem is a problem at all speaks to how little we know about the causal processes that go on when I will a slight movement of my finger. We are left no better off considering the causal chains we play a role in than considering ones we merely observe: in both cases we are an observer who must learn what is possible merely by noting the connections between events (I will my finger to move, I see and feel my finger move).
The conclusion of this is that we cannot ground at all our mind’s predilection for associating ideas down causal lines. When it does so, it does not take its cues from some underlying order of things, and reason confidently from the perspective of them, in an originary and a priori position. The only thing we can ground in the immediacy of impressions is our bubbling memories of constant conjunctions between configurations of impressions of certain types. From this becoming accustomed, we develop an expectation, which is felt – this ‘feeling’ of anticipation then providing the impression from which our idea of causation is derived, nothing more. The thought of order, best summarized by Leibniz’s principle of sufficient reason4placeholder, is not a priori at all, but rather the sedimented layers of past impressions exhibiting a local state of remembered order, with no guarantee how far this order stretches, spatially or temporally. A gut feel.
This is all well known, the problem of induction, the fact that causal necessity requires an appeal to an ungroundable statement of the a priori regularity of things, and the attendant modesty that it recommends when we offer causal accounts, and the constant ambiguity of these accounts. But it’s worthwhile to pause here and extract Hume’s general conclusion from the failure to ground our ideas concerning causality in impressions: it is not just causal connections, we perceive no connections among impressions.
“Summing up, then: throughout the whole of nature there seems not to be a single instance of connection that is conceivable by us. All events seem to be entirely loose and separate. One event follows another, but we never can observe any tie between them.” (Enquiry, p.36)
They seem associated, but never connected. We think these connections, through the habits of mind that create creodes and contours on our thinking, allowing ideas to ‘cluster’ and ‘hang’ together and suggest one another. Our experience itself deepens these contours, and increases the speeds of the flows, but since every idea originates in an impression, and every impression is isolable in its own simple idea, then sans ideas, there are no connections at all between impressions. We can re-use my argument from above: if, in addition to two impressions connected, say, by their contiguity/proximity, I also had a third impression – the impression of nothing but their connection – this would not ‘connect’ the two impressions, but rather raise a further question about how this third impression is to connect to the other two. They can only be related through ideas – whose relations to one another are merely the well-worn ruts of past experience, accumulated repetitions, paths of least resistance carved into a geography.
There’s much more to be said on this, but this provides enough context to examine Hume’s treatment of the idea of personal identity, which will allow us to bring the problem into its own.
⋅ ⋅ ⋅
The idea we have of personal identity is of something to which our impressions and ideas are related, and exists unchanged through time. My perceptions and my memories are mine, and I remain unchanged by their variation and decay. I can forget everything, and have all of my senses permanently scrambled, and there still would be a sense in which ‘I’ am still there, just undergoing changes.
However, Hume applies his criteria: from what impression could this idea of the identity of the mind over time be derived? On the surface of it, it looks lacking almost by definition. The ‘self’ is not one continued impression alongside all of the others, but the that which all impressions, ideas, and memories are related, their common context. This common context is supposed to be identical over times (it is the substance that ‘has’ the impressions) and also to be simple (beyond the impressions that it has, its pure state does not permit of parts, it is the unity upon which complex objects are reflected/constructed). But in order for this idea of an identical and simple ‘mind’ to be coherent, it would need some impression to ground it. But, being the thing that ‘has’ the impressions, the only way we could bring it into view would be to subtract all of the impressions that it has, to get a clear ‘look’ at it in its pure state. However, having done that, we’ve annihilated it. There’s nothing there to observe.
Hume reaches his initial conclusion decisively:
“[Each of us is] nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement.” (Treatise, p.300)
That is, it is not the case that there is some mind, which then has its perceptions, some context filled with ideas and impressions but could just as happily be empty, but, rather, the ‘I’ is a label for this ‘bundle’ of present and immanent impressions and ideas, hence the neologistic name for this view: “the bundle theory of self”.
Okay, fine. This fits with Hume’s general project of not admitting more than is necessary based on the given impressions, and always referring back to them as ground. But Hume is drawn into a number of strange consequences here:
“There is properly no simplicity in [the mind] at one time, nor identity in different times; whatever natural propension we may have to imagine that simplicity and identity” (Treatise, p.301)
Bennett appends this elucidation directly after this sentence in his edition:
“That is to say: It is not strictly true that when a blue colour is seen and a whistling sound heard at the same time, one single unified mind has both these perceptions; nor is it strictly true that the mind that has a certain perception at one time is the very same mind that has a perception at another time.” (Treatise, (b), p.131)
Since there is no singular context above and beyond the impressions, then a mind is constituted for every impression. Or, more precisely, no mind is constituted, there are just the impressions. The connection between a blue color and a whistling sound (bird?) is not that both arise in a single mind, and then, given this common context they are already related one to the other in a basic proto-relation, to be strengthened and developed by ongoing repetitions of the experience, but is rather just the fact that these impressions have come in tandem, and become associated together at the level of ideas in the ways discussed above. But, where are these impressions, and where are these associations forming? Hume is agnostic here: “we haven’t the faintest conception of the place where these scenes are represented or of the materials of which it is composed.” (Treatise, (b), p.131).
Hume then sets himself the task, mirroring that done with the cause-and-effect relation, of explaining just how we can become so confident that there is an unchanging self. This moves by examining our idea of identity simply, and from where it arises, and, again, arguing that it is based on nothing more than habits of mind and expectations developed through the flow of impressions. Ultimately the identity we apply to any object is a convenient, habitual fiction. In truth,
“every distinct perception, which enters the composition of the mind, is a distinct existence, and is different, and distinguishable, and separable from every other perception” (Treatise, p.307)
And thus, we have no special reason to think that the identity we attribute to ourselves is any less fictitious. Yet,
“we suppose the whole sequence of perceptions to be united by identity—we say that the members of the sequence are all perceptions of a single person—which naturally raises a question about this relation of identity. Is it something that really binds together our various perceptions themselves, or does it only associate the ideas of them in the imagination? In other words, when we speak about the identity of a person, do we observe some real bond among his perceptions, or do we merely feel a bond among the ideas we form of those perceptions?” (Treatise (b), p.135)
Hume has already answered this question. There is no real bond between perceptions to be observed, because each impression exists distinctly. So, our sense of identity needs to be of the order of the relation of ideas – those creodes that form through constant conjunction5placeholder, a mere ‘path of least resistance’ that has been formed by repetition. But here the difficulty emerges, and Hume’s previous agnosticism on the question of the ‘where’ comes back to haunt him.
In the appendix Hume points out that these two positions, that every impression exists distinctly and independently, and that the understanding never perceives a real connection among impressions, leads him into a contradiction. He is not at his clearest in explaining just what exactly he takes the problem to be. He says:
“But all my hopes vanish, when I come to explain the principles, that unite our successive impressions in our thought or consciousness. I cannot discover any theory, which gives me satisfaction on this head.” (Treatise, p.678)
I feel the most straightforward reading here is that though our sense of identity can be explained by our reflecting on the relations (creodes and valleys) formed between past ideas, this presupposes an apparatus that can feed impressions in an orderly way into memory for the formation of ideas in the first place. That is, it might be the case that I am nothing but the relations between my ideas, formed over time, and this is what gives me a feeling of identity, but, nevertheless, here are a bundle of impressions all congruent with the idea of a perspective of a body looking at a world. What holds this ‘array’ of impressions together such that, through their orderly progression, I can begin to associate my ideas of them into a stable sense of the world? The orderly progression of impressions is assumed, but this very ordering implies a structure, which in turn implies connections, connections that impressions don’t have as distinctly and independently existing intensive ‘atoms’. So, the problem is this: what’s stopping the mind from fracturing apart and disintegrating into its constituent atoms? What’s holding our minds together? Hume laments that if it was the case that there was some unchanging substance in which impressions inhered, or, if it was the case that we could perceive some real connection between impressions, then the problem he has discovered would vanish. Either of these would provide the uniting principle for impressions, the way that habit formed the uniting principle of ideas. Obviously if there was ‘screen of consciousness’ to display the ‘stream of consciousness’ then the stream would have its principle of unification (for each, one screen). Or, if the stream itself contained its own connections and limits, we could explain and differentiate them. But, as it is, Hume does not want to budge on accepting either of these.
Now, here is what I think is interesting. Rather than raise the problem in the appendix, Hume could have martialed this objection as an argument for suggesting that we need just one more thing in the ontology beyond impressions and ideas: the apparatus through which the existence of them is grounded and activated. But he will have none of that. This, then, should make us do a double take. In the beginning, Hume points our attention to two things of which we are aware: impressions and ideas, and says everyone understands this distinction, and then carries on. When we attend to the distinction, we see it and continue on with him. However, it is very easy at that point to be smuggling in the ‘apparatus’ that Hume refuses, in the final page, to accept. Despite the fact that Hume is careful never to commit the mistake of positing minds above and beyond impressions when it matters, his often-casual prose, and turns of phrase, lulls us into preserving an assumption: an impression is just a current perception of an object out there in the world. There are minds, and then there are things, and the mind perceives the things and become conscious of impressions and then transforms them into ideas and thinks about them. However, this final moment shows us that Hume never had this picture in mind. Impressions are not representations: they are full on objectivities.
That is, for Hume, there is just a field of ‘perceptual objectivities’, impressions, fully grounded and independently existing atoms which differ from each other in regard to their intensity. The less intense being the after-glow of the more intense, and the dynamic interplay between these sensory atoms carves out a provisional, semi-stable, local order, known as the relations of ideas. The after-glows blur and mix into one another in a contoured geography, the set of total relations which guide thought from idea to idea.
At the final moment, he realizes he is missing the principle of individuation, how do objects and persons emerge out of the dynamic inter-play of these intensive, sensory objectivities? A tough problem. But his complete stubbornness to introduce an ad hoc principle of individuation (individual minds/differentiated mental apparatuses) at the end speaks to his radical empiricism. This radical empiricism does not ask us to ground our knowledge on experience as a merely prudential method for avoiding rationalist excesses, and gaining a more trustworthy and reliable knowledge, it does not marshal economic reasons to resign ourselves to ‘mere’ representations, but makes a much stronger claim: the impressions are the objects. That’s the ground.
In this reading, Hume is developing his position on a plane that lies beneath, or orthogonal to, the idea of the subject as ready-given principle of individuation. He’s not beginning from a point where people, minds, and representations are being assumed. Instead, he develops a whole theory of intensive objectivities, not a theory of representations. With that in mind, one can then start again, and find, in addition to the sensible, humorous, Hume, a radical Hume, intent on breaking out of the confines of the subject as representing Cogito. His reticence to back track attests to the fact that he is holding out for a theory of individuation that could explain how our minds emerge out of intensive sensory impressions, rather than merely assuming this individuation, under the heading of the ‘necessarily existing mind’ that Descartes clung to. This reversal, from impressions emerging from a mind, to a mind emerging from impressions, is easy to miss, but for that final page.
As a last example, perhaps we should have pricked up our ears when Hume introduces his maxim that “an object can exist, yet be no where” (Treatise, p.284). The fact that he introduces this through the discussion of locating the flavor of a fig might make us think that the import of this argument only extends to a negative claim, that not all mental representations contain spatial references to the world, only our visual and tactile impressions contain that additional reference. Taste is thus deficient in its content, duller than sight in what it communicates. But, here, we can already see the radical empiricism if we’re looking for it: space itself is only an after-glow (ideational) effect that emerges from certain classes of these intensive objectivities (visual and tactile), but there are many species of others burning just as brightly, with their own strange aspatial geographies…
A note on referencing: Treatise refers to Hume’s “A Treatise of Human Nature”, Enquiry to Hume’s “Enquiry Concerning Human Understanding”. At some points, I use Jonathan Bennett’s clarity-edit of the Treatise, which I’ll signify by appending (b) in the reference.
Or, more accurately, we examine the term, and ask if there is an idea that corresponds to it – all ideas have some impression as their origin. So, if there is not a locatable impression, then no idea, and thus the term under examination is senseless for lack of an idea that could correspond to it.
Or “nextness”.
Which can be phrased, roughly, as: for everything that is, there is a sufficient reason for why it is, in the precise way that it is, and not otherwise.
Ultimately it is the resemblance relation associated with memory, and the cause-and-effect relation based on our actions, plans and inferences about forgotten past states of ourselves.