In his 1919 paper ‘On Propositions: What they are and how they mean’ Bertrand Russell attempted to solve a problem which had bothered him since 1903 but to which he could not find a solution. The problem which concerned Russell was called the problem of the unity of the proposition. Russell made various different attempts to solve this problem culminating in his 1913 ‘Theory of Judgement’ which was severely criticised by Wittgenstein and led to Russell abandoning his theory of judgement. Since his ‘Principles of Mathematics’ Russell had difficulties in accounting for the unity of the proposition. His technique of logical analysis involved him analysing propositions down to their basic atomic components. However, the problem was that once he did this he found it difficult to reconstruct the propositions in a way that had sense. If we take the proposition ‘The dog is bigger than the cat’, in the proposition one has two different nouns and a two place predicate. The problem is that if we treat the two place predicate ‘is bigger than’ and nouns ‘The dog’ and ‘The Cat’ as separate atoms, then we are left with the difficulty of how these atoms can be related to each other in a sensible manner. If we take the two nouns as separate atoms and the relational predicate as another atom; then the question becomes as to how the atoms are related together. If we say the relational predicate is related to the atoms by another relation we will go on an infinite regress of relations relating relations. But if we say that the relation intrinsically relates the two atoms we are pointing towards a mystery as opposed to solving it. Russell eventually decided that the atoms are related by a process of judgment by a subject. But he had intractable difficulties in explicating the nature of the subject or the nature of the atoms the subject was supposedly organising.
Around 1919, while in prison, Russell came to believe that solving the difficulty (which relates to symbolism) would of necessity involve psychological data. So Russell turned his attention to psychology. His primary area of interest was the behaviourism of J.B. Watson, however while Russell was impressed with Behavioural science, he found one major difficulty with it. Watson’s behaviourism involved denied that mental imagery existed and explained away mental imagery interms of muscular movements. Russell strongly disagreed with this approach:
“When Professor Watson says: “I should throw out imagery altogether and attempt to show that practically all natural thought goes on in terms of sensori-motor processes in the larynx (but not in terms of imageless thought”, he is it seems to me, mistaking a personal peculiarity for a universal human characteristic.” (Bertrand Russell ‘On Propositions’ p. 293)
Above Russell is criticising Watson for engaging in the typical mind fallacy when theorising about mental imagery. On Russell’s hypothesis; Watson has an inability to call up mental imagery and confusedly mistakes his inability as a universal trait shared by all humans. This inference of Watson’s is obviously an invalid one; and Russell cites the work of Francis Galton who demonstrated empirically that most humans were able to create mental imagery (though the degree to which they could do so varied).
Russell also noted that Galton’s study of mental imagery demonstrated that people’s ability to form mental images decreased as they got older. Russell speculated that perhaps Watson had lost the ability to form mental imagery through hard abstract work in behavioural science. This conjecture of Russell’s has been verified in a close scholarly study of Watson’s evolving views on mental imagery; ‘The First Modern Battle for Consciousness: J B Watson’s Rejection of Mental Imagery’ by Berman and Lyons. Through a close textual analysis of Watson’s writing throughout his career they demonstrated that early in his career Watson admitted to experiencing mental imagery while later in his career he claimed an inability to form mental imagery. Berman and Lyons argue convincingly that Watson’s change of mind resulted from him losing the ability to form mental imagery later in life. So Russell’s conjecture about Watson’s capacity to form mental imagery does seem to have been on the mark. Likewise Russell was surely correct to charge Watson with engaging with the typical mind fallacy.
Russell argued convincingly that the existence of mental imagery was a clear refutation of the type of behaviourism that Watson was endorsing. However, Russell made further uses of mental imagery in trying to solve the unity of the proposition which resulted in Russell having to incoherently attack his own ability to form judgements. When discussing the meaning of words Russell made the following point:
“So far we have found four ways of understanding words:
(1) On suitable occasions you use the word properly.
(2) When you hear it, you act appropriately.
(3) You associate the word with another word (say in a different language) which has the appropriate effect on behaviour.
(4) When the word is first being learned, you associate it with an object. The word ‘motor!’ can make you leap aside, just as motor can, but it cannot break your bones. ( Russell ‘On Propositions’ p. 301)

Russell’s account of the nature of words is thus far pretty comprehensive; and as he notes holds no difficulties for the behaviourist. But Russell argues that such usage of words is limited to what he calls demonstrative language; picking out and discussing objects in the immediate environment. Another important use of language is narrative use of language which involves telling someone about some remembered event. On Russell’s account a narrative use of language typically involves the formation of mental images in the speaker discussing the remembered event, and the formation of mental images in the mind of the person who listens to and understands what the speaker is saying. Russell notes:
“It is clear that, in so far as the child is genuinely remembering, he has a picture of the past occurrence, and the words are chosen so as to describe the picture; and in so far as the hearer is genuinely apprehending what is said, the hearer is acquiring a picture more or less like that of the child… it is nevertheless the possibility of a memory image in the child and an imagination image in the hearer that makes the essence of the ‘meaning’ of the words. In so far as this is absent, the words are mere counters, capable of meaning, but not at the moment possessing it. We may say that while, words used demonstratively describe and are intended to cause sensations, the same words used in narrative describe and are intended to cause images.” (Russell ‘On Propositions’ p. 302)

Based on the above line of reasoning Russell concludes that there are two other main ways that words can mean as well as the four mentioned above. (5) Words can be used to describe or recall a memory image, (6) Words can be used to describe or recall an imagination image. Russell notes that 5 and 6 are the essence of the meaning of words.
Russell uses such mental imagery in attempting to solve the problem of the unity of the proposition:
“I have a complex image, which we may analyse, for our purposes, into (a) the image of the window, (b) the image of the fire, (c) the relation that (a) is to the left of (b). The objective consists of the window and the fire with the very same relation between them. In such a case, the objective of a proposition consists of the meanings of its constituent images related (or not related, as the case may be) by the same relation as that which holds between the constituent images in the proposition. When the objective is that the same relation holds, the proposition is true; when the objective is that the same relation does not hold, the proposition is false.” (Russell ‘On Propositions’ pp. 316-318)

Russell’s attempted solution to the problem of the unity of the proposition involves the postulation of mental imagery. Likewise he argues that mental imagery is the essence of the meaning of words in most cases. This position raises a number of difficulties for Russell. As we saw above Russell argued that Watson didn’t have the ability to form mental imagery. But if we assume with Russell that Watson didn’t have the capacity to form mental imagery then this raises questions with Russell’s theory of meaning. Watson is perfectly capable of reasoning verbally about things in the past and things in the future. Russell respected Watson enough to ask him to read and comment on his ‘Analysis of Mind’. In practice Russell acted as though Watson was a thinker who was capable of meaning things by his words. Yet on Russell’s theory; he should have judged Watson’s speech which didn’t involve mental imagery as follows:
“it is nevertheless the possibility of a memory image in the child and an imagination image in the hearer that makes the essence of the ‘meaning’ of the words. In so far as this is absent, the words are mere counters, capable of meaning, but not at the moment possessing it.” (ibid p. 302)

If Russell was true to his word he should have judged a non-imager like Watson’s speech as a meaningless counter. That Russell didn’t adopt this extreme approach in interpreting the verbal utterances of Watson is evidence that Russell didn’t fully endorse his theory in practice.
But things get much worse for Russell’s theory when consider how he uses it to deal with the problem of the unity of the proposition. On Russell’s view our mental imagery of a state of affairs is true when the imagery is isomorphic with the state of affairs it depicts and false when it isn’t. Russell thinks this isomorphism between our mental images and the reality it depicts solves the unity of the proposition problem. But we saw earlier that Russell argued that Watson was incapable of forming mental imagery. Based on this fact Russell would have to conclude that Watson was incapable of judging statements such as ‘the cat is on the mat’. This is a fantastical result.
However things get even more fantastical when we consider what Russell said about his own mental imagery. In his ‘Philosophy of Logical Atomism’ written the year before his paper on Propositions Russell made the following claim:
“That, of course, is especially likely in very abstract studies such as philosophical logic, because the subject-matter that you are supposed to be thinking of is so exceedingly difficult and elusive that any person who has ever tried to think about it knows you do not think about it except perhaps once in six months for half a minute. The rest of the time you think about the symbols, because they are tangible, but the thing you are supposed to be thinking about is fearfully difficult and one does not often manage to think about it. The really good philosopher does once in six months think about it for a minute. Bad philosophers never do.” (Bertrand Russell ‘The Philosophy of Logical Atomism’ p. 185)
Above Russell argues that the philosopher (including Russell) when thinking about abstract matters in logic can rarely get to the thing itself and instead has to think purely in symbols. Yet in his ‘On Propositions’ written less than a year later Russell argues that when a person is thinking in symbols they are using counters, capable of meaning but without imagery devoid of meaning. This would mean that Russell only thinks meaningfully only once every six months for a half a minute.
Of course a defender of Russell could argue that he is in the above quote talking about one of the most difficult subjects known to man and that when it comes to more prosaic subjects he can think in images for a much longer period. But even this defence of Russell doesn’t fit with the facts we know about his intellectual capacities.
In his masterful biography of Russell Ray Monk noted that Russell read Williams James’s Principles of Psychology and found James’s account of people having different capacities to form Mental Imagery fascinating. In a section where James discussed Galton’s breakfast table test; James noted:
An exceptionally intelligent friend informs me that he can frame no image whatever of the appearance of his breakfast-table…the ‘mind-stuff of which this ‘knowing’ is made seems to be verbal images exclusively.
In the margins where Russell wrote about the above example, he noted ‘this is almost my own case’ (ibid p. 86). Furthermore when Russell was psychologically tested by Crawshay-Williams, the results indicated that Russell was primarily a verbal thinker.
It is both fascinating and strange that Russell who by his own acknowledgement (and according to a psychologist who tested him), was a poor Visual Imager; would construct a theory of judgement that meant he was virtually incapable of forming any judgements. The examples, Russell gave of using imagery that were isomorphic to mind independent facts seem to be the very facts he would be incapable of forming. So, for example, if he were incapable of forming an image of his breakfast table he couldn’t judge whether the salt was to the right of his cup of tea. He could use verbal reasoning to make this judgement but according to Russell this type of judgment wouldn’t be an example of real thinking.
The question which now needs to be asked is why would Russell sketch a theory of thinking that implied that he could rarely think? Russell’s reliance on imagery becomes even more strange when one thinks on the philosopher who Russell spent a lot of his time criticising; Henry Bergson. After meeting and talking with Bergson; Russell made the following comment:
“I didn’t find out anything except, even more strongly, what one gathers from his books, that he is a very vivid visualiser, but has little auditory or tactile imagination-his whole philosophy is dominated by a sense of sight” (ibid p. 239)
It is surprising to find that Russell’s theory in ‘On Propositions’ results in a state of affairs where Russell is arguing that Bergson, (who he strongly disagrees with) is capable of meaning; whereas Russell most of the time is just using counters and not thinking.
It is difficult to understand what could lead Russell to advocate a view made meaning so parasitic on something Russell claimed to be incapable of. Could Russell really think that he wasn’t meaning anything when he spoke?
Strangely enough; there is a sense in which Russell did take his public utterances to be somewhat meaningless. Russell’s childhood was tragic. At the age of two his mother and sister died. His father died a year later. Russell was then brought up by his grandparents. After his grandfather died and his brother was sent off to boarding school Russell was brought up at home by his puritanical grandmother. Russell learned early to form polite chat at home with his grandmother and to keep his real thoughts and feelings to himself. Later on in life when he was informed about psychosis being a major factor in his family he came to associate these thoughts and feelings with fears of going insane.
Nonetheless, despite outward appearances Russell had a rich fantasy life and felt resentful at his grandmother and his life in general. Throughout his life he would swing between finding people he believed could penetrate to the true him; artists like Joseph Conrad and D. H. Lawrence or his lover Lady Ottoline, to disenchantment when relations strained with these muses. When this happened Russell would throw himself into his mathematical logic and regressed to the way of acting he maintained with his grandmother:
“Everything vital and important to him was kept hidden behind a polite, stiff and priggish exterior” (Monk ‘Russell p. 158)
“And, just as he had then, so he now tried desperately ‘to avoid all deep emotion’ and to live once more on the surface” (ibid p.305)
But what is the nature of this self, he believed, these artists grasped about him? When discussing his bond with Conrad; Russell noted:
“It had to do with our shared ‘Satanic Mysticism’, the truth of which he had never been convinced about, but ‘in moments of intense emotion it overwhelms me.’ It consists in thinking there are two levels: ‘one of science and common sense, and another, terrifying, subterranean and periodic, which in some sense held more truth than the everyday view.” (ibid p. 317)
So for Russell at some level; the subterranean depths were more real than the level of science and common sense. But what did these subterranean depths consist of? A further discussion of Russell’s capacity for mental imagery is revealing.
Despite what we discussed above; Russell could form some mental imagery. While living abroad and away from his girlfriend he commented in a letter that:
“I have lost the power of visualising thee, which I only keep a few days of parting” (ibid p. 87)
So when it came to someone he was emotionally and sexually attached to he could form mental images. Furthermore, and more disturbingly, when noting his violent tendencies he also noted that these violent thoughts gave him burning hot imagery:
“I remember when I wanted to commit murder, the beginning was a sudden picture (I hardly have pictures at ordinary times) of a certain way of doing it, quite vivid, with the act visible before my eyes. It lived with me then for ever so long, always haunting me; I took to reading about murders and thinking about them, and always with that picture before me. It was only hard thinking that kept me straight at the time- the impulse was not amenable to morals, but it was amenable to reasoning that this was madness.” (ibid p. 256)
So with someone he cared deeply about he could form some mental imagery even if he couldn’t maintain the imagery for long without contact with that person. His imagery improved as we saw when he was highly emotional; for example in the disturbing example of his fantasies of murder. However, he didn’t always need to be strongly aroused by love or hate to form mental images:
“Bates bores me while I am reading him, but leaves pictures in my mind which I am glad of afterwards.” (ibid p.529)
But in again his forming images is related to reading fiction. For Russell, mental imagery is restricted to lovers, hated people he thinks violent thoughts about, and to the realm of fiction. In other words, for Russell mental images are primarily the realm of fantasy. Yet in his paper in propositions he thinks that the essence of thinking is in images. He goes as far as to say that true thinking cannot occur without such images.
It is a reasonable conjecture that Russell growing up in Victorian society with a puritanical grandmother taking care of him, resulted in him hiding all his passions and bad thoughts from public consumption; even from himself at times. But at an unconscious level he always thought that this hidden self was his real self (as opposed to a just another aspect of his evolving personality). Russell was terrified that at his core he had inherited insanity from his family. His sometimes violent fantasies became associated with this madness and these fantasies, typically accompanied by imagery, became associated with what he unconsciously believed was the truth. His public philosophy which he constructed without mental imagery became unconsciously associated with a meaningless facade.
It is my contention that Russell’s ‘On Propositions’ was an unconscious confession that his public philosophy was a meaningless facade, and that his true nature was the self revealed in his intense imagery of lovers, enemies, fiction and madness itself.