2.4 Gaut's analysis of imagination
Berys Gaut's main concern in his paper is to provide an account of the relationship between imagination and creativity. But in section 2 of his paper he offers an analysis of the notion of imagination, which we will look at here.
Read the introduction to Gaut's paper and then section 2, entitled ‘Imagination’. You should read the whole section at least once through first, and then consider each paragraph more carefully as you answer the following questions, which are partly intended to guide you through a more detailed reading. The penultimate paragraph, in particular, packs in a number of different issues. You should concentrate, at least initially, on picking out the main point or points.
Click on the 'View document' link below to read Berys Gaut on 'Creativity and imagination'.
In the first three paragraphs of section 2, Gaut distinguishes four uses of the term ‘imagination’. What are these four uses, and how do they relate to what I have already said, and to Stevenson's conceptions?
In the fourth and fifth paragraphs, Gaut argues against identifying imagining with imaging. What is his argument? Do you find it convincing?
In the sixth paragraph, Gaut presents his basic conception of what he calls (at the beginning of the seventh paragraph) ‘propositional imagining’. How does he articulate this conception? Gaut offers several formulations, which he claims are equivalent. Are they equivalent? (There are a number of technical terms used here. You should focus on Gaut's basic conception.)
In the seventh paragraph, Gaut explains what he calls ‘objectual imagining’. What is this further kind of imagining, and how does Gaut see his account of propositional imagining being extended smoothly to cover it? Is he right that ‘imagining some object x is a matter of entertaining the concept of x’?
The eighth paragraph is the most difficult of all. Gaut distinguishes a third kind of imagining, which he calls ‘experiential imagining’, covering both ‘sensory imagining’ and ‘phenomenal imagining’. What are these? Experiential imagining, he says, involves imagery. But since not all imagery implies imagining, what does Gaut suggest is the difference between imagery that is, and imagery that is not, a form of imagining? How does he see the relation between experiential and propositional imagining?
In the final paragraph, Gaut considers a final kind of imagining, ‘dramatic imagining’. What does he mean by this, and how is it related to the other kinds?
In the first use that Gaut distinguishes, ‘imagining’ means ‘falsely believing’ or ‘misperceiving’. This (derogatory) use is related to the sense of ‘imagining’ as ‘fantasising’ mentioned above, reflecting the connotations of ‘imaginary’, and to Stevenson's third and seventh conceptions. In the second use that Gaut distinguishes, ‘imagining’ is more or less synonymous with ‘creating’. This is the more appreciative sense of ‘imagining’, reflected in the connotations of ‘imaginative’, and in the third group of Stevenson's conceptions (numbered 9 to 12). In Gaut's third use, ‘imagining’ means ‘imaging’, which is Stevenson's fifth conception. In the first three paragraphs, Gaut does not say exactly what the fourth use is. He merely says what it is not: ‘imagining’ is here to be distinguished from ‘imaging’. The first paragraph also suggests that he sees this as the core sense.
Gaut argues that there are cases of imaging – such as we find in memory, dreams and perception – that are not cases of imagining; and conversely, that there are cases of imagining – such as imagining an infinite row of numerals – that are not cases of imaging. We will return to the relationship between imagination and imagery later in this course. But with regard to this last example, you might have wanted to object that while no one can form an accurate mental image of an infinite row of numerals, they might well have some image in mind, such as a row of numerals going off into the distance. So even if having mental images cannot be all there is to imagining, the possibility has not yet been ruled out that imagery may be a necessary condition of imagination.
‘Propositional imagining’ is imagining that something is the case – imagining that p, or ‘entertaining the proposition that p’, as Gaut puts it. What is meant by ‘entertaining’ a proposition is thinking of it without commitment to its truth or falsity. Gaut offers two other formulations – thinking of the state of affairs that p, without commitment to the existence of that state of affairs, and thinking of p without ‘asserting’ that p. Gaut raises a doubt about the latter himself, and one might have doubts also about the appeal to ‘states of affairs’. Could there be a ‘state of affairs’ involving an infinite row of numerals, for example? It is not clear what this might mean. Yet Gaut would presumably admit that we can entertain propositions about infinite rows of numerals. So ‘entertaining propositions’ is arguably not the same as ‘thinking of states of affairs’. (Involved here, again, are questions about whether – or how – we can imagine ‘impossible’ things. There can be no state of affairs involving round squares, but can we imagine them? Have we not, at least, just entertained a proposition about them? More controversially, it might be argued that there can be no state of affairs in which a human is an insect, since humans and insects are essentially different. But did Kafka not write a story about this? Could we appeal to fictional states of affairs to get round the difficulty? Or does this just cover up the difficulty?) However, Gaut's main point is clear: as he conceives it, imagining that p is thinking of p without commitment to its truth or falsity (without ‘alethic’ commitment, as he puts it in more technical language).
‘Objectual imagining’ is imagining an object. Gaut suggests that, just as propositional imagining is a matter of entertaining a proposition, so objectual imagining is a matter of entertaining a concept. But what exactly is the relevant concept? Take Gaut's example of a wet cat. Can I not ‘entertain’ the concept of a wet cat without thinking of any particular cat (real or imagined)? Of course, Gaut defines ‘entertaining the concept of x’ as ‘thinking of x without commitment to the existence (or non-existence) of x’, so that by ‘concept of x’ he means ‘concept of a particular x’. Nevertheless, we should still distinguish between imagining a particular wet cat and thinking (without existential commitment) of wet cats in general. And perhaps we might want to go further, and draw a distinction too between imagining a particular wet cat and ‘merely’ entertaining the concept of that particular wet cat. Imagining, it might be objected, has more of a sensory quality than talk of entertaining concepts does justice to. (In effect, this objection is recognised in what Gaut goes on to say about ‘experiential imagining’.)
‘Experiential imagining’ is a richer kind of imagining, imagining with a ‘distinctive experiential aspect’, as Gaut describes it tautologically. More specifically, he suggests, it covers both sensory imagining and phenomenal imagining. As an example of the former, he gives visually imagining something, and as an example of the latter, imagining what it is like to feel something. In both cases, he says, imagery is involved. But as he argued in the fourth paragraph, imagery does not in itself imply imagining, since imagery can occur in remembering, dreaming and perceiving. So what distinguishes imagery that is, and imagery that is not, a form of imagining? According to Gaut, to have an image is to think of something, and the content of that thought can either be ‘asserted’ or ‘unasserted’. Having an image is a kind of imagining, he says, when the thought-content is unasserted. (Talking of a thought-content having ‘intentional inexistence’ is just a technical way of saying that what we think of on a given occasion exists in thought, even if not in reality.) So experiential imagining is like propositional imagining in having an unasserted thought-content. The difference between experiential and propositional imagining, according to Gaut, lies in the way in which the thought-content is presented. In propositional imagining, that content is merely ‘entertained’; in experiential imagining, it is visualised or otherwise represented with imagery. In visually imagining a wet cat, for example, I conjure up an image of a wet cat: this is experiential (sensory) imagining. Experiential imagining may well involve propositional and objectual imagining, but imagery is what gives it its experiential character.
‘Dramatic imagining’ is imagining what it is like to be someone else or to be in someone else's position. Gaut sees it as even richer than experiential imagining, but as ultimately reducible to the first three kinds of imagining.
Central to Gaut's account is the idea that mental acts have a ‘thought-content’ that can be thought of in different ways. The idea is often expressed by saying that mental acts involve the adoption of a certain attitude to a proposition – a ‘propositional attitude’ – or, in the case of objects, involve thinking of an object under a certain ‘mode of presentation’. Take the thought or proposition that Immanuel (my cat) is wet. I can adopt various propositional attitudes to this. I can believe it, I can desire it, I can fear it, or I can imagine it, to mention just four. On Gaut's conception, to imagine it is to think of it without commitment to its truth or falsity. To believe it, on the other hand, as Gaut himself notes, is to think of it with commitment to its truth. But in both cases, the ‘thought-content’ of the propositional attitude is the same – the proposition that Immanuel is wet. The basic idea suggests, too, that there is a kind of bare thinking of this proposition lying at the core of the mental acts, which one might even be tempted to identify as ‘imagining’ in the most general sense possible. (This would give us Stevenson's sixth conception.) Certainly, imagining comes out as a form of thinking, for it involves thinking of a proposition or object (but without alethic or existential commitment, i.e. without commitment to its truth or falsity, or its existence or non-existence, respectively).
The appeal to ‘propositions’, however, is more controversial than Gaut's account implies. What are ‘propositions’? I have already raised doubts as to whether they can be construed as ‘states of affairs’. Even understanding them simply as ‘thought-contents’ is problematic. For it has proved notoriously difficult to provide criteria of identity for such contents. Can one really identify one and the same ‘thought-content’ across the various mental acts? If there is at least some sense in which imagination plays a role in perception itself, then specifying a ‘thought-content’ independent of the act of imagining may be more difficult than one might assume. However, let us leave this can of worms for the moment. Let us just accept the tautological point that, when I imagine or think of something, there is something that I am imagining or thinking of. And let us take Gaut's example of a wet cat. There is a whole range of imaginings that might, in some sense, have a particular wet cat, say Immanuel – or the ‘state of affairs’ in which Immanuel is wet – as their ‘thought-content’. I can imagine that Immanuel is wet, I can imagine a wet Immanuel, I can conjure up an image of a wet Immanuel, I can visually imagine Immanuel as wet, I can imagine Immanuel (who is standing dry before me) as wet, I can imagine what it is like to be as wet as Immanuel, I can imagine (perhaps) what it is like to be wet Immanuel, I can imagine (perhaps) what it is like to be in wet Immanuel's paws, and so on. Obviously, I have set it up this way, but there is clearly something that relates these various imaginings, and the natural thing to say is that they all concern a wet Immanuel, and furthermore involve thinking of a wet Immanuel in some way, a way that does not commit me to taking Immanuel as actually wet. It seems plausible to regard imagining, then, in a wide variety of cases, as thinking of something without alethic or existential commitment. At any rate, this is what Gaut offers as the core sense of ‘imagining’, and which we can take as constituting his main definition.
How does Gaut's definition compare with the general definition suggested by Stevenson's twelve conceptions? Are they equivalent? Does one follow from the other?
Do any of the examples I gave to illustrate Stevenson's twelve conceptions provide counter-examples to Gaut's definition? If so, how might Gaut respond?
I considered possible counter-examples to the general definition in the last section (Section 2.3). Are they also counter-examples to Gaut's definition?
Is Gaut's definition an improvement on the general definition?
In both cases, imagining is defined as a form of thinking. But while on the general definition imagining is thinking of something that is not present to the senses, on Gaut's definition imagining is thinking of something without commitment to its truth or falsity, existence or non-existence. The two are not equivalent, since Gaut's definition does not follow from the general definition. To take the case of objects, if I am thinking of something that is not present to the senses, then it does not follow that I am not committed to its existence or non-existence. (Nor does it follow that I am committed.) However, the general definition does follow from Gaut's definition. For if I am thinking of an object without commitment to its existence or non-existence, then it follows that it cannot be present to the senses, since otherwise, presumably, I would be committed to its existence. Being acknowledged to exist is part of what is meant by being ‘present to the senses’, as that phrase is intended to be understood here.
Stevenson's very first conception provides a counter-example to Gaut's definition. For in imagining how my daughter looks as I speak to her on the phone, I am certainly committed to her existence. Gaut might reply that I am not committed to the particular way I imagine she looks. Might I not be wrong? But perhaps I do know how she looks, and can imagine it precisely. Stevenson's third conception provides a further counter-example. For when Macbeth imagines that there is a dagger in front of him and reaches out towards it, he is committed (at least at that point) to its existence (even if he is wrong). Gaut might reply here that he has admitted that imagining can sometimes mean ‘believing falsely’ or ‘misperceiving’ (this was the first use he distinguished, in the first paragraph of section 2 of the reading), but that this is not its core sense. But why should we not try to reflect it in our main definition, if we can? Stevenson's fourth, seventh and eighth conceptions provide further counter-examples. In ‘imagining’ fictional characters, I may be well aware that they are not real. In ‘imagining’ that smoking is good for me, I may actually believe it. And in ‘imagining’ that the whole of something exists when I can only see a part of it, I am certainly committed to its existence. Perhaps Gaut would see these too as derivative or non-standard senses, but the counter-examples seem to be extensive.
The example from Activity 3 only adds to the counter-examples to Gaut's definition that I have already mentioned. For in imagining a tree as a person, I may be perfectly aware that no such person exists, i.e. I may well be committed to the non-existence of what I imagine. However, Gaut's definition does suggest an answer to the problem raised in the discussion of Activity 3. On Gaut's conception, in imagining something I am not committed to the truth or existence of what I imagine, whereas in remembering something, he would presumably argue, I am so committed. So the requisite distinction between imagining and remembering can be drawn.
The definition suggested by Stevenson's twelve conceptions was too general in including things that do not count as imagining, such as remembering. In indicating the relevance of issues of truth and existence, Gaut's definition is an improvement. However, the requirement that, in imagining something, I am not alethically or existentially committed seems too strong. It rules out too many cases in which it seems perfectly legitimate to talk of ‘imagining’. In all these cases, I can imagine something while being firmly committed to either its existence or non-existence, truth or falsity. There are too many counter-examples, in other words, for Gaut's definition to be acceptable as it stands, at least if we are trying to capture a basic sense that underlies all the main uses of ‘imagine’.
Can a definition be offered that is better than both the general definition and Gaut's definition? What seems to be needed is a definition that is more specific than the general definition, yet less specific than Gaut's definition. Gaut is right that issues of truth and existence are relevant, but they are relevant, I think, in a different way to the one he supposes. At the end of the last section, it was suggested that, in imagining something, I am thinking of something that need not be true or existent (in either the past, present or future) in the sense that, were it not true or existent, it would still be legitimate to talk of my ‘imagining’ it. I may well be alethically or existentially committed myself, but my thinking would still count as imagining even if I were wrong in my alethic or existential commitment. The actual truth or existence of what I am imagining, in other words, is not essential to the imagining itself in the way that it is essential to perceiving or remembering by contrast.
If a definition of ‘imagining’ can be offered at all, then it is something along these lines that I think is needed. But the suggestion will not do as it stands. For we now seem to have lost the distinction between imagining and believing. In believing something, am I not also thinking of something that need not be true or existent, in the sense indicated? So what is the difference between imagining and believing? Gaut characterised this as the difference between thinking of something without alethic or existential commitment and thinking of something with such commitment. But if what I have said is right, then imagining too can occur with alethic or existential commitment. An obvious response is to say that while in the case of believing, I cannot believe something without commitment to its truth or existence, in the case of imagining, I can imagine something without such commitment, or indeed with commitment to its falsity or non-existence. This is true, of course. But it does not enable us to distinguish cases of believing something with commitment to its truth or existence from cases of imagining something with commitment to its truth or existence. How can we draw this distinction? Perhaps all we need is a minor refinement to my current suggestion, on the basis of which we can then draw the following contrast. In the case of imagining, if what I imagine were not true or existent, and I were to realise this, then it would still be legitimate for me to talk of ‘imagining’. But in the case of believing, if what I believe were not true or existent, and I were to realise this, then I could no longer talk of ‘believing’ it.
This is correct as far as it goes, but it obscures what I think is the crucial point here. Consider the case of Macbeth's imagining that there is a dagger before him, as he feverishly prepares to kill Duncan at the start of Act II of Shakespeare's play. Does he not also, as he reaches out to take it, believe that there is a dagger before him? And is this believing not part of his imagining? In the core sense of ‘imagine’ as Gaut defines it, the answer to this latter question would have to be ‘No’. But as we have seen, Gaut does allow that ‘imagine’ can sometimes mean ‘believe falsely’. So his considered answer would presumably be: ‘Yes, but not in the core sense of “imagine”’. However, a different response is possible that does not force us to distinguish two incompatible senses of ‘imagine’. The crucial point, I think, is this. To say that Macbeth ‘believes’ there is a dagger before him implies that he is committed to the existence of a dagger before him. But to say that he ‘imagines’ that there is a dagger before him implies not that he is not committed to the existence of a dagger before him (though he may be), but that we are not committed to its existence in describing him as ‘imagining’ it. The lack of commitment to the truth or existence of something, in other words, lies not so much on the side of the person who is being described as ‘imagining’ something, as on the side of the person who is doing the describing. In talking of ‘imagining’, it is we who are indicating that lack of commitment. Of course, someone can themselves talk of ‘imagining’ something, as Macbeth in effect goes on to do (after he fails to take hold of the dagger he initially believes to be before him), but this would indicate their own lack of commitment to the truth or existence of what they are thinking of. On this alternative account, then, Gaut is right about the relevance of considerations of truth and existence, but he has brought them in at the wrong place.