In my previous post on James Mill, I noted that he emphasised the need for the study of the human mind as a precondition for planning education. Mill tackles this very topic in his two-volumed Analysis. First volume, which I will be studying now, deals roughly with the theoretical side of the human mind, while the consideration of the practical side is left for the second volume.
Mill’s viewpoint is clearly empiricist and he begins with an assumption of Humean undertones that everything in our experience is reducible to sensations (what Hume would call impressions), ideas, formed on basis of sensations, and groups and trains of ideas, generated from sensations and ideas through association. Starting from sensations, he enumerates the usual smell, hearing, taste and sight, divides sensations of touch into sensations of (moderate) temperatures, sensations of tactile contact, sensations of disorganisation, such as pain or itch, and sensations involving muscular exertion, and adds the sensations belonging to alimentary canal and possibly also to other intestines.
Mill’s portrayal of these various types vary from ordinary to perplexing, latter being exemplified by his insistence that the effects of opium are sensed in the alimentary canal and just strengthened by associated ideas. Indeed, one might argue that Mill should have added a separate class for clearly pleasurable sensations, such as is occasioned by stimulation of various substances and sex. In any case, Mill makes an important remark that when speaking of sensations, at least English has a deplorable lack of distinct words for a) sensation of one kind (say, a smell), b) whatever causes this sensation (also known as smell) and c) the organ for this type of sensation (yet again, smell).
Ideas, Mill continues, are then traces of sensations. While we have some say on what we sense e.g. by changing our position in relation to our environment, he affirms, we cannot control what ideas we will have or think about, because they are generated by sensations and other ideas through association. This association is governed, again in a Humean manner, by habit - if we are used to sensing various sensations or ideas together, either simultaneously or successively, we are bound to think one part of this group when another part is sensed or thought.
Although the account is rather simple, there are some questions that Mill leaves unanswered. For instance, Mill notes that in case of succession we are used to following the train of thoughts to the direction of the process involved: thus, a farmer is used to thinking of first planting the seed and then the seed sprouting. Problem is that we often think such processes in reverse and particularly do so, when looking for explanations, for instance, if we want to explain why plants grow, we think first of their growing and only then the idea of someone planting the seeds. Another problem is that Mill does not point out that there are associations, which appear to be a necessary product of the way we sense - for instance, a combination of colour with extension seems not just coincidental nor even one based on the nature of things sensed, but on a fact that we cannot sense coloured things, which are not extended.
Such groups and trains of ideas can, then, form a single idea out of many individual ideas, Mill says. Thus, since we are accustomed to sensing, say, a distinct rosy smell with a certain kind of sight, we tend to think of them as a single unit (a rose). Similarly, temporal trains of ideas should form complex ideas of processes.
The purpose of language, Mill suggests, is merely to communicate to others or to our future selves these various sensations and ideas, either individually or as complex trains of thought. Later on, he adds another purpose for the language, that is, making it possible to think of such otherwise difficult ideas like a polygon with hundred sides. This seems a very limited understanding of the purposes of language. Particularly, this appears not to note that sentences of a language are meant to indicate e.g. that something happens or happened, not just that we have sensed or thought something. Mill himself appears to be at least unconsciously aware of this, since he notes that all verbs carry with them the notion of existence. Yet, Mill uses “existence” here in a very strong sense, which makes him instantly think of this immediate connection as a detriment to language - if we speak of e.g. what chance is, Mill suggests, we immediately assume that chance is something existent, namely an individual object. It seems obvious that we need not go so far, in other words, when we say that chance is so and so, we truly affirm something about chance without assuming that chance is an individual existent.
The trains of thought we express in language indicate, Mill suggests, causal relations, relations of resemblance or different names for same things. The third option here is the most perplexing, since Mill includes in it such mathematical statements as one plus one equals two. One might accept this classification in such simple cases, but whenever the calculations become complex enough, it is doubtful that we have any concrete idea to name in both sides of the equation. For instance, can we really say we have a distinct idea of what 1000^1000 is meant to say? What this expression indicates is more like a certain task of calculation (multiply 1000 with itself 1000 times), and the sentence connecting this task to its result indicates a certain relation between numbers.
The remaining part of the first volume is dedicated to an analysis of certain central concepts of human cognition in terms of concepts introduced thus far. For instance, Mill suggests that consciousness means no feeling distinct from sensations and ideas, but is simply a general name for both of them: if I am conscious of something, I either sense or have an idea of it. Although one might agree with Mill that consciousness is nothing distinct from sensations and ideas, one might still ask in the manner of Leibniz whether consciousness - or better, conscious sensations and thoughts - could form a subspecies of all sensations and ideas. Indeed, we might assume that different sensations and ideas have various degrees of consciousness and by moving our attention to one aspect of a sensation or an idea we might become more aware of its components.
Of the concept of conception Mill notes that it simply refers to complex ideas. This is undoubtedly true description of how some thinkers have used the word or similar expressions in other language (say, Begriff in German). Yet, other thinkers seem to use concept and conception only in relation to linguistically expressed notions (thus, a mental image of a bicycle would not yet be a concept, whereas a linguistic expression “bicycle”, perhaps also with a definition of the expression, would certainly be). Related to this is the question whether conception is somehow active in comparison with mere sensation and “having an idea”. Mill suggests that the supposed activeness of conceiving is just a misdirection of the active form “I conceive” and that conceiving is as passive an event as dreaming or dying, which can also be expressed in an active form. While Mill’s point is well made, it could be that when philosophers have spoken of the activity of conception, they might have actually meant the relative independence of conceptions from external objects. After all, we can conceive things that do not exist, especially if conception is understood in a linguistic manner.
In case of imagination, Mill goes even farther in his redefinition of words and suggests that imagination is just a new term for having a train of ideas. This seems to go against all ways the word has been used. In an early modern discussion, imagination was closer to what we would nowadays call memory, meaning thinking of anything that isn’t currently present, even if we have perceived it earlier. Nowadays, we are mostly using imagination in the sense that Mill disparages as Dugald Stewart’s redefinition of the word: that is, imagination is a train of new thoughts. “It is implied in every wish of the child to fly”, Mill says, trying to be sarcastic, but we indeed speak of imaginative kids as having such inventive thoughts, often divorced from reality.
Especially confusing is Mill’s discussion of belief. He at first divides belief into three distinct classes: belief in events and real existences, belief in testimony and belief in the truth of propositions. Later he makes a reclassification, saying that belief in testimony is just a special case of a belief in events (those told by other people). Nowadays, many philosophers might want to classify all beliefs as concerning propositions, because the content of belief can be expressed as a proposition. On a closer look, Mill himself might agree, because he thinks that one part of beliefs in propositions, those concerning individual objects, are essentially beliefs in events and real existences (he would just have to make the assumption that all such beliefs can be expressed in propositions). This still leaves the problem that Mill’s account of propositions is too simple, allowing in principle only monadic predication - a problem we already saw in the case of his idea of mathematics.
An even bigger problem lies in Mill’s attempt to describe belief merely in terms of sensations, memories, associations and abstractions. For instance, he says that in the case of sensations and memories, belief means simply having them - if we sense something red or have a memory of it, we believe it. These cases already seem problematic. If Mill means just that having a sensation or having a memory is such an experience we can never doubt, we might accept that this is true, but this is, for the most part, not a very characteristic way to speak of belief (one might even say that only philosophers say things like “I believe I have a sensation of redness”, whereas usually people use the concept of belief only in cases where there is at least a possibility of things not being so). If, on the other hand, Mill would want to use the notion of belief here in a more substantial manner - say, as a belief that there is something that causes this sensation of redness or that there was something that I now remember as red - then we immediately see that belief is definitely not identical with sensation or memory. True, sensation and memory might well be evidence and occasion for such a belief, but there might still be doubt e.g. whether what we remember really was red and not blue.
Mill’s mistake is even clearer in his supposed examples of belief. If we spin around very fast and stop, we get the feeling of the earth spinning beneath us. Mill suggests that we also believe that earth is really spinning, which seems far-fetched. Even more, Mill makes a quick throwaway line that fearing something implies also believing in it - hence, he says, anyone who is afraid of ghosts must believe in them. A quick consideration makes Mill’s statement seem unfounded, even if he thought it so self-evident to need no justification. Picture a person hearing a strange noise in her house during the night. She might be afraid that there is a burglar in her house, but she might still be uncertain whether there is any burglar and thus not really believe in the existence of this burglar.
Indeed, Mill appears to make no distinction between modalities of actuality and possibility, belief being intimately related to the former - we might consider e.g. a sensation or memory, without being sure of its object being real. In the more complex cases, he says that habitual associations just are beliefs - if we tend to think of the sun rising after it has set, we believe that it will rise again after setting. Again, this regular or habitual association might be a ground for believing something, but there are cases where we might still have doubts, whether e.g. the sun will rise everyday for the whole infinity of time.
This is not to say Mill’s book is completely without merits. His account of abstraction is particularly interesting. Mill’s basic point is the difference between what calls notation and connotation. A word like “red” notes a certain property of individual objects, but it also connotes something, namely, those very objects that can be called red. In more modern terms, we might say Mill is here talking about the difference of intension and extension. Abstraction, then, means that we ignore the connotation of the word and think only of its notation: we think of redness without any reference to red things. Mill’s point is that abstractions are then nothing truly existent, but always dependent on the ignored connotation - there is no Platonic redness floating around without red objects.
