What a happy day to have read this work! I haven't encountered any thinker yet who so exactly puts his finger on the phenomenon I've been caring about, and who provides a compelling positive account of it, even if this account is quite general or preliminary. It is a wonderful surprise. Let me summarize Sartre's ideas and then ramble a bit about them.
In chapter 1 "Description," Sartre lays out his method and motivates its importance. It is broadly a phenomenological method, by which he asks us to attend to our everyday experiences and notice details in them. Sartre argues that it would be pertinent if psychologists did this more; they too quickly leap to theories, and so miss out on all the further empirical evidence they could have for inspiring theory, by failing to do phenomenology. Sartre begins describing imaginative states (what he calls "the image" or "the imaged object.") Whenever we talk about objects, of which we're consciously aware, we're in the habit of connoting that they lie in space and time of reality. Imagined objects violate that. Moreover, imagined objects always differ from perceived objects with respect to that in apprehending the former, we are aware of how the object consists in our consciousness (or conscious activity/awareness) itself; in contrast in apprehending the latter, we are aware of how the object comes from the world, and so is material/worldly. The relationships between knowledge and perception, on the one hand, and knowledge and imagination, on the other hand, also differs. We gain knowledge through bodily interaction with the world, when we perceive it. In contrast, it is knowledge we already have that is necessarily drawn upon and governs the appearances of the imagined object. So in apprehending an imagined object, knowledge is 'immediate'; it is manifest in the imaged object itself, and needs not be acquired (and in fact we can't acquire new knowledge from imagined objects, according to Sartre, a point I'll return to below.)
Whenever we perceive the world, there is "an infinity of relations" that stand between things of the world. So there are many different possible states of affairs that we can alight upon, in interacting with and perceiving the world. A metaphor is that there are many different constellations that could be drawn, given the extreme density of stars. In contrast, imaginative states have "a kind of essential poverty"; the constitutive elements of an imaginative state lack relationships to the rest of the world, and have finite relations to one another, namely the relationships of which we already know, and are now primed to let govern the act of imagining. Nevertheless, given sensorily/visually minimal or truncated imagined objects can impact us profoundly, and evoke a wealth of thoughts and memories. This is because any imagining flows from past knowledge, whose active role in partially constituting the imagining at this moment implies that there is an intention we have, to imagine something or another, in a certain way or another; this intention may be a mental state that extends past the imagining itself, unifying and impacting different sorts of mental attitudes (Sartre doesn't go into this point, but this is my reconstruction of what he gestures at. I think it is fruitful to explore this relationship between intention and imagination further, for the sake of analyzing what role imagined objects can have in our lives; I'll go into this below.)
Any imagined object is apprehended as "a nothingness." We grasp anything we imagine as not actually being present, but rather as nonexistent, absent or elsewhere, or as neutral regarding its existential status but as at least not in the distal here-and-now. This follows from the fact, according to Sartre, that imagination is always creative and active; that we are imagining, as an intentional act, implies that we are aware of our intention and thus of the fact that this is an imagining. So we will be aware of the imaginative, non-real character of the object. Of course imagined objects may stand in for or represent actual objects in the world, but when they do, whenever we encounter such imagined objects, we never confuse the imagining itself with those objects out there in the world, for the reasons just explained. At the same time, however, we never perceive a mental image as a mental image; whenever we imagine, we encounter the object imagined, of which we're also aware is not a perceived object before us, but is delivered by imagination.
In chapter 2 "The image family," Sartre explores how many different phenomena interactions with which may yield imagined objects. It's not just whenever we lay in bed and imagine something, we encounter imagined objects. There are also photographs, portraits, verbal descriptions, and caricatures, for example. Sartre does an interesting phenomenological analysis of the different sort of experiential states interactions with each of these types yields. Verbal descriptions of a person have no visual resemblance to the person. Instead, the association is purely conventional. In contrast, photographs and portraits may bring about similar psychological effects upon the viewer as a perception of the object might (e.g., the elicitation of affect or emotion.) This may partly explain superstitions and rituals certain peoples have had towards pictures. These psychological effects continue to happen to us, even when we know that the person isn't really there; this obstinacy of effects may be compared to that of the effects of certain visual illusions (e.g., the stick in water still appearing as bent, when we know it is not.) Sartre explores how schematic drawings, like caricatures, can have especially profound emotional effects on us. Why is that? He doesn't offer an explanation, but this fact is certainly fascinating, and I'd like to explore it further (it is related to rhetorical devices and how those work, in general.)
Chapter 3 "The nature of the analogon in the mental image" was my favorite. Here, Sartre offers a positive account for how imagined objects can affect us emotionally at all, as if they were actually there impacting our bodies and raising some stakes. Sartre starts off with analyzing further how knowledge and intention is responsible for any imagining. Whatever we encounter in imagination will be "degraded," with respect to the full grade of relevant objects as perceived. If we can transform or anticipate new unfoldings in an imaginative episode, this is made possible by what we already know about the imagined objects. Sartre proposes that we can understand the basic structure of objects of imagination in terms of "symbolic schemata" or "affective-motor analogons," which are roughly synonymous on his terms. A nice example to get us started in understanding this is Hume's missing shade of blue. Hume asked whether we can picture a particular shade of blue that stands between other shades that we perceive. Sartre's answer is that we may not be able to visually imagine that shade itself, but we do imagine some overall form of that shade, where this form partially consists in our knowledge and intentions pertaining to the relationship this shade has to all the others, which we do perceive.
In perceiving the world, we attend to something outside of ourselves, and grasp that it is indeed external. In contrast, when we imagine the world, our consciousness implicitly aims to determine, or to make more clear/precise, the object that is imagined, which is apprehended as something that is not really there, and so which can't be interacted with; interaction is no longer available as a possible source for increased understanding or determination.
Sartre visits the theory of affect available in his day. Psychologists broadly define affect in terms of a purely subjective feeling, which has no relation to objects in the world. Sartre criticizes this; affective states must be linked to the world. He proposes that we understand affect as a modification of the mode of consciousness awareness as a whole, and thereby as having the functional status of modifying all the particulars of which we become aware, rather than a particular mental stage that fills one's conscious awareness at a certain time. Once we are affected, or have affect, we often aim to discover more about our situation, to figure out what is affecting us. In imagining an object, when it gives us affect, we then are compelled to attend to our knowledge and background intentions, which are responsible for the contents of this imagining. This knowledge and intention will be tethered to the world, or to reality. There are things in the world that we've perceived in the past and acquired knowledge from, which now informs the present imagining. Remembrance of those things, in turn, further affect us, changing our affective state into different emotions.
Sartre insists that we should avoid the temptation to think about affect as something that's added on top of knowledge states. He claims that instead, affect itself is constitutive of the knowledge. He doesn't elaborate on what this means exactly; I find it wonderfully suggestive, and will elaborate upon this idea below.
Sartre visits the idea of cross-modal perception. People blindfolded, when touching an object from all sides, can come to visually imagine the shape of the object accurately. Husserl writes on retension and protesion: certain happenings trigger in us awareness of the next happenings to come, which we anticipate. Sartre proposes that imagined objects are multi-modal and have this dynamically unfolding character; an imagined object is "a form in the making" and activates in our bodies kinesthetic sensations, which amount to a sort of action inclination or motor compulsion, which coincides with the the form. If we imagine a sphere, for example, this will activate a kinesthetic sensation of the possibility of turning it over, rolling it, or doing other things that are consistent with our knowledge and present intention regarding the sphere thus imagined.
Sartre calls the succession of kinaesthetic sensations and their corresponding affects which partially constitute an imagined object "an affective-motor analogon." This is always a schematic form of an object we have knowledge and intentions of, where this form develops over time, on the basis of what appears, how our bodies respond (both kinaesthetically and affectively), and what in turn unfolds in appearance. The analogon functions as a proxy of an object in the world, which we could perceive and act upon. In perception, when we interact with an object, more of its aspects unfold, which are given by the world itself; but in imagination, the imagined object will seem to spontaneously unfold and transform according to the dynamics that define the analogon (i.e., the interaction of kinaesthetic and affective sensations, orchestrated by background knowledge and intention).
Sartre has a fascinating analysis of the relationship between linguistic expression and "pure thought," which just is imagination itself. We can have a pure thought, which is nonlinguistic, and consists in affective-motor analogons. Then, we could try to describe or narrativize what is going on in our minds. In doing so, we will now newly present into consciousness specific words. These words are apprehended differently than analogons. Words are signs for something that has a sensory-visual presence. The words themselves lack any sensory-visual resemblance to that thing. So it is possible to apprehend linguistic expressions in the absence of any analogons; in contrast, in pure thought or in imagination, that is impossible. Analogons constitute pure thought itself. Sometimes when we apprehend linguistic expressions, the forcefulness or presence of the relevant analogons are still present. The words are like delicate signs laid over the pure thought. At other times, the distinctive function of language takes over, and the analogons are lost.
Once we articulate our pure thoughts in the form of linguistic expression, these expressions themselves will carry with them further knowledge/intentions, which go beyond those which were initially constitutive of the pure thought. So while we cannot learn anything new from imagination itself, often in imaginative episodes we employ language, or we describe/conceptualize what's going on in our minds to ourselves—it is this linguistic activity that presents new materials, which makes learning possible during imaginative episodes. Here's another way to think about what the introduction of language does to thought/imagination. A linguistic expression is a new form that's external to our pure thought. We now can interact with this form, as something "outside" of our initial conscious state. This allows for all sorts of new associations, ideas, and discoveries. I really love this point. I will muse upon it below.
In chapter 4 "The role of the image in psychic life," Sartre examines aesthetic experience and dream experience in light of his concept of the analogon. In chapter 5 "The irreal object," Sartre focuses on the temporal aspect of imaginative episodes. The sense of time within the imaginary world of imagined objects differs from that of the real world. It can feel like no time has passed at all, if you're absorbed in a fantasy, when in fact an hour has passed. The feelings we imagine ourselves as having, when we imagine ourselves interacting with imagined objects, are not actual feelings. Consider the difference between a memory or imagination of a pain, and actually having a pain. In chapter 6 "Conclusion" Sartre examines the implications that these essential features of the imagination have upon the nature of consciousness. Whenever we reflect upon experience, and so conceptualize what is going on, Sartre argues that this must be an imaginative act; and as imaginative, we are aware that whatever we have posited as real need not be exactly the case. Whatever we're aware of explicitly or reflectively (i.e., imaginatively) is "emptiness" or "nothingness" in the sense that it is presented by our mind, and is not reality in itself—and importantly, we are aware of all of this. One can see how this theory of mind paves the way towards Sartre's theory of radical freedom, which he will write in his later work Being and Nothingness.
It is totally worth reading chapters 1-3 in careful detail. This work as a whole is greatly underrated. It gets at more profound issues pertaining to the imagination than most contemporary works on the imagination (e.g., Greg Currie, Ian Ravenscroft, Peter Langland.) Sartre points out that not all sorts of conscious experiences are created equal. Many contemporary works presuppose that imaginative experience is just like perceptual experience but represented in a certain way; Sartre shows that the difference is beyond that, or he goes into detail regarding the exact differences.
Now here are some ramblings. I've been thinking about how in imagination/thought, objects can be presented as either part of reality, or as unreal (e.g., make-believe, false, illusory.) Reading Sartre sensitizes me to a couple of distinctions and makes me wonder about the relations between them: (1) spontaneity v. reflection; (2) manifest reality v. unreality, (3) no language use v. language use. For example, it seems that reflective language use severs us more fully from the affective-kinaesthetic effects of the analogon, but spontaneous language use can preserve the functional powers of the analogon. Sometimes, it seems that if experience is more spontaneous, it is likely to show up as real, whereas if the contents of experience require much deliberation or care, it is more likely to show up as unreal. This may be because once more deliberation or care is involved, this usually means we're using language reflectively; or, we're actually making things up, if we're not drawing inferences using language reflectively. There are many open questions in this vicinity.
Sartre also helped me realize how concerned I should be about my pet hypothesis that the primary modifers of objects of experience are tracked by the concepts of our sense of reality v. make-believe. There are various different ways by which an imagining relates to reality. For example, maybe the category of things that show up as real but as uncertain can be on par with straight-up make-believe at least with respect to certain key dimensions. The two would be unified by awareness of how what we come up with can reflect reality in different ways. For example, compare an artist’s expression, on the one hand, and my imagining something that might or could happen on the other hand. I could do the latter for the sake of predicting the future, for my own amusement, etc. The purpose of this imaginative project would make a difference as to how it ‘symbolizes’ the world, and as to what sort of genuine emotion it evokes. If I imagine something that could happen for the sake of my own amusement, it’s ever more close to straight-up make-believe. I'm not sure whether this means that manifest reality v. unreality/make-believe is no longer a primary distinction. At least it means that when we consciously focus on a certain object, that object could be make-believe, while at the same time, it anchors us onto some part of manifest reality, by virtue of our grasp of some 'symbolic' or 'expressive' relationship.