Samantha Matherne on Intuitions of Sense, Intuitions of Imagination, and Full-Blown Experience

Abstract

In her book Seeing more,¹ Samantha Matherne first gives a characterization of imagination in general and then explains how we use imagination in theoretical, aesthetic and practical contexts. Seeing More is an extraordinarily valuable and helpful contribution to Kant scholarship, offering a remarkably clear and comprehensive account of the faculty of imagination. It is rare to encounter philosophical writing that is so lucid and accessible. Particularly impressive is the way she develops and applies her conception of imagination across a wide range of philosophical domains, including theoretical philosophy, aesthetics, and practical philosophy.

Matherne starts her book with examining how imagination relates to the two cognitive capacities of sensibility and understanding and argues for the view that imagination belongs to the faculty of sensibility. Her main reason is that Kant characterizes sensibility as the general capacity to bring about intuitions and imagination as the capacity to bring about specific intuitions. On her view, sensibility has two parts, namely sense and imagination, which are capacities to bring about two different kinds of intuitions, which she calls “intuitions of sense” and “intuitions of imagination” (SM, 74). Whereas intuitions of sense are the direct result of the senses being affected by objects, intuitions of imagination require an act of synthesis that is performed by imagination. Furthermore, she claims that in order for imagination to synthesize a sensible manifold it has to be guided by concepts. Intuitions of sense, by contrast, neither require synthesis nor the use of concepts.

Matherne begins the second part of Seeing More, in which she analyzes the use of imagination in theoretical contexts, with discussing a specific kind of intuitions of imagination, namely perceptions. She distinguishes perceptions not only from intuitions of sense, but also from what she calls “full-blown experience”. Whereas in her view both perceptions and full-blown experience require the use of empirical concepts, only full-blown experience involves the application of concepts in judgments. The way in which concepts guide acts of the imagination does not consist in being applied in a judgment.

In my comment, I will only discuss topics from the first and second part of the book. More specifically, what interests me is how Matherne conceives of perceptions and how she distinguishes them from intuitions of sense on the one hand and from full-blown experience on the other hand. In the first part of my comment, I will treat the relation between perceptions and intuitions of sense. In the second and third part, I will examine how perceptions relate to concepts. This will include a discussion of the relation between perceptions and full-blown experience which is the topic of the third part of my comment. From now on, like Matherne, I will refer to intuitions of sense as “intuitions_s” and to intuitions of imagination as “intuitions_i”.

Matherne distinguishes between three stages of experience: the sense stage, the imagination stage and the understanding stage (cf. SM, 5.2). On the sense stage of experience, empirical intuitions_s are generated, on the imagination stage a specific kind of intuitions_i, namely perceptions, are generated and on the understanding stage judgments about objects of experience are formed. So, from the quotation from p. 160 we learn that perceptions differ from (empirical) intuitions_s in two respects: (i) in the way in which they represent a certain spatio-temporal form and (ii) in representing objects as qualitatively determinate.

I take it to be uncontroversial that, on Kant's view, representing an object as qualitatively determinate involves employing a concept. This is because, according to him, concepts are general representations, that is representations of properties that are shared by objects of a certain kind. So, one cannot represent an object as sharing a property with objects of a certain kind without using a concept. With respect to this point Matherne and I agree.

Now consider the following situation: When Anne and Miriam first see the cow, it is slightly foggy and the cow is extremely far away from them, so that even Anne is not sure about what it is that she sees. They walk for some minutes, get much closer to the cow and the fog disappears. We might characterize this situation by saying that first Anne and Miriam perceive or intuit the cow not as a cow, but just as a brown spatial area, and only perceive it as a cow when they get closer and the fog has disappeared. By saying this, we do not mean that they classify or identify the cow either as a brown spatial area or as a cow. Instead, by saying that the cow is represented in these ways we specify how the cow looks to Anne and Miriam. First, the cow looks like a brown spatial region, later it looks like a cow. I refer to perceiving or intuiting an object x as F in this sense as “phenomenally perceiving or intuiting x as an F".² And I take it to be controversial whether phenomenally intuiting/perceiving an x as an F requires possessing the concept F. If one believes that perceptions have conceptual content, then one assumes that phenomenally representing x as F requires possessing the concept F³; if one believes that perceptions have non-conceptual content, then one denies that phenomenally representing x as F requires possessing the concept F.

The point of this example is to describe how your surrounding looks to you after you have left the movie theater. In a situation like this, the parking lot and the cars look different than if you haven't come out of a dark movie theater. They look blurred, and so look different than when you perceive them as having a determinate shape. Thus, Matherne characterizes the first difference between intuitions_s and intuitions_i as a difference in the way in which they phenomenally represent objects.

To sum up, when I have a perception, (i) I phenomenally intuit its object as having a determinate spatio-temporal form, and (ii) represent it as qualitatively determinate (as sharing a property with other objects of a certain kind). By contrast, when I have an intuition_s, (i) I phenomenally intuit an object as having an indeterminate spatio-temporal form, and (ii) I do not represent it as qualitatively determinate. For example, when I leave the movie theatre and have an intuition_s of my surrounding, the objects that I intuit look fuzzy shaped, and, say, red to me, but I do not represent them as being qualitatively identical to other red objects with respect to color, that is, I do not classify or identify them as being red. Neither do I classify them as having certain indeterminate shapes. By contrast, after my sensible representations have been synthesized and I have a perception of my surrounding, I phenomenally intuit my surrounding as consisting of red objects with determinate shapes, for example car-shapes and I represent these car-shaped objects as sharing the property of being red with red things and the property of being a car with cars, and thus classify them as red and as being cars.

In this part of my comment, I will only treat the first respect in which intuitions_s and perceptions differ.⁴ Specifically, I will discuss the question whether we should – as Matherne suggests – interpret Kant as claiming that we have two kinds of intuitions, the first of which does not require synthesis and phenomenally represents objects as having an undetermined spatio-temporal form, whereas the second does require synthesis and phenomenally represents objects as having a determinate spatio-temporal form.

Yet, the claim that every intuition_i, and thus every perception, has to be preceded by an intuition_s is phenomenologically unconvincing. Since, as Matherne claims herself, intuitions_s for Kant are conscious representations, she has to assume that in our stream of consciousness intuitions that represent objects as having fuzzy boundaries constantly alternate with intuitions of the same objects as having determinate shapes. Put differently: On Matherne's interpretation, Kant is forced to claim that the world looks alternately blurred and sharp. But in fact, this is not the way we experience the world. Secondly, to me at least it is unclear what could be an example of an intuition_s of other sensory modalities such as an auditory, olfactory, or tactile intuition_s. What could it mean to hear a melody and to represent it as temporally indeterminate? Thirdly, according to Kant, we form empirical concepts by comparing empirical intuitions (or by comparing their objects). Since according to Matherne, the synthesis that results in the generation of perceptions is guided by empirical concepts and thus presupposes that we already possess empirical concepts, I assume that in her view, empirical concepts are formed by comparing intuitions_s. Yet, the empirical concepts we would form on the basis of intuitions that represent objects as spatio-temporally indeterminate would be other empirical concepts than the ones that we indeed possess. Our concept of a car for example is not the concept of an object that has an indeterminate shape. So, if Kant distinguished between intuitions_s and intuitions_i in the way Matherne proposes, he would not be able to explain how we form the empirical concepts that we in fact possess.

In this passage, Matherne claims that in order for an intuition to (phenomenally) represent an object as having some shape, synthesis has to take place. Let us apply this to Matherne's example of the intuition_s you have after having left the movie theatre: This intuition_s (phenomenally) represents the objects you see as having shapes. As Matherne claims herself: “[Y]ou are […] conscious of something like an array of colors […] in space […].” (SM, 85). An array of colors in space clearly has some shape. So, after you have left the movie theatre, the objects you see (phenomenally) look shaped in some way. You do not (phenomenally) intuit them as having determinate shapes; you (phenomenally) intuit them as having fuzzy shapes. Still, (phenomenally) intuiting something as having fuzzy shapes is (phenomenally) intuiting it as being shaped in some way and for this, according to the passage quoted above, synthesis is required. So, if we take Matherne's claim that, in for an intuition to represent an object as being shaped in some way, synthesis has to take place, seriously, then it follows that also intuitions_s require synthesis.

What Kant stresses in this passage is that the generation of any intuition of a spatial or temporal region requires that I synthesize the representations of the parts of the region with each other. This is a general view about the way in which intuitions of spatial or temporal regions are formed, and I don't see why, for Kant, the fact that an intuition represents a spatial or temporal region as having fuzzy instead of determinate boundaries should have the consequence that the formation of such an intuition does not require synthesis.⁵

If this last objection against Matherne's interpretation of Kant's conception of intuition is correct and Kant indeed assumes that not only intuitions that represent objects as having a determinate spatio-temporal form, but also intuitions that represent objects as having an indeterminate spatio-temporal form require synthesis, then it turns out that there are no intuitions that do not require synthesis. Now, since of the two capacities sense and imagination, only imagination is a capacity to synthesize representations, accepting my objection has the consequence that all intuitions are what Matherne calls “intuitions of imagination”.

If I understand this passage correctly, it shows that according to Matherne, synthesis has to be guided by concepts not only in order for the resulting intuitions to represent objects as exhibiting concepts and thus as qualitatively determinate,⁶ but also in order for them to phenomenally represent objects as spatio-temporally determinate. Still, to me at least, it is unclear why Matherne attributes this position to Kant. In her view, the claim that acts of the imagination are guided by concepts follows from the Schematism chapter. As she understands this chapter, Kant answers the question of how we are able to subsume intuitions under empirical concepts in the following way: Empirical schemata are rules that guide the synthesis of imagination responsible for the generation of perceptions. Since empirical schemata depend on empirical concepts, the perception that results from an act of synthesis guided by schemata exhibits those concepts on which the relevant schemata depend. And since perceptions exhibit concepts, we are able to subsume them under concepts. Thus, on Matherne's interpretation, Kant introduces concepts as rules for acts of the imagination to explain how intuitions represent objects as qualitatively determinate. Yet, since the topic of the Schematism chapter is not how we bring it about that we phenomenally represent objects of intuition as spatio-temporally determinate, in Seeing More we do not find an explanation for why acts of the imagination have to be guided by concepts in order for intuitions to phenomenally represent objects as spatio-temporally determinate.⁷

I myself do not believe that we can phenomenally intuit objects as having a determinate spatio-temporal form only if we possess empirical concepts. This claim seems to be empirically false. It is not the case that before I acquired the concept of a hare and the concept of a bunny, say, hares and bunnies looked fuzzy to me. More generally, to acquire a concept F is (at least) to acquire the capacity to classify Fs as being F. Now, I don't see any reason for the view that in order to phenomenally intuit an F as having a determinate shape we have to be able to classify the object as an F. As I have already said, to me quite the contrary seems to be true. As long as we do not intuit objects as having determinate shapes, we are not able to acquire the concepts they fall under. If hares and bunnies looked like fuzzy brown regions to me, I would never form the concept of a bunny and a hare.

So, since the claim that having an intuition that represents its objects as spatio-temporally determinate requires a concept-guided act of synthesis is systematically not convincing, and since in Seeing More I do not find an account of why Kant adopts this view, I prefer not to ascribe it to him. In the next part of my comment, I will explain why I furthermore do not believe that, for Kant, in order for an intuition to represent its object as qualitatively determinate, it has to be generated by a concept-guided synthesis. In order to do so, I will examine how Matherne conceives of the relation between perceptions and full-blown experience.

According to these passages, we can use empirical concepts in two different ways: We can either apply them in judgments about objects of intuition or we can use them as guiding acts of imagination. In the first case, what is generated is a perception. In the second case, what is generated is a full-blown experience. In both cases, we identify the intuited object as being F, but we identify it in different ways. In the first case, we intuit the object as being F, in the second case, we judge it to be F.

As I understand Kant, he does not distinguish between two ways in which we can use concepts or two ways in which we can identify objects of intuition as being some way. What I object to is Matherne's claim that Kant takes it to be possible to identify an object as being F without applying a concept in a judgment. I agree that, on Kant's view, we can perceive or intuit an object x as being F, but as I understand him, this amounts to intuiting x and judging that x is F. My first reason for this view is Kant's claim that “the understanding can make no other use of these concepts than that of judging by means of them” (CpR, A 68/B93). Kant here explicitly denies that, apart from applying concepts in judgments, there is a second way of using them, and thus he denies that we can represent objects as being qualitatively determinate by using concepts as guiding acts of the imagination.

Similarly, in the Critique of Pure Reason, Kant writes that “one first judges something problematically, [and] then assumes it assertorically as true […].” (CpR, A 76/B101). According to both passages, the stage that comes before asserting, accepting, or taking to be true is not intuiting, but judging problematically. As I see it, the reason why Kant characterizes the mental state of representing an object of intuition as qualitatively determinate without committing oneself to the correctness of this representation not as a perception or empirical intuition, but as a problematic judgment, is that representing an object as qualitatively determinate requires applying a concept, and – as we have just seen – concepts, in his view can only be applied in judgments. For these reasons, I think it is highly likely that according to Kant, we cannot distinguish between perceiving or intuiting an object as being F and judging that an intuited or perceived object is F. Instead, perceiving or intuiting an object as being F always involves making a judgment about the intuited object, which might either be a problematic or an assertoric one.

Intuitions_s a) phenomenally represent objects as spatio-temporally indeterminate and b) do not represent objects as qualitatively determinate. Intuitions_i a) phenomenally represent objects as spatio-temporally determinate, and b) represent objects as qualitatively determinate. Inuitions_s are the direct result of causal affection of the senses. Intuitions_i require a synthesis of sensible representations and therefore require imagination as the capacity to synthesize sensible representations.

According to Matherne, acts of the imagination have to be guided by concepts for two reasons: (i) Otherwise perceptions would not phenomenally represent their objects as spatio-temporally determinate. (ii) Otherwise, perceptions would not represent their objects as qualitatively determinate.

From my objection to Matherne's claim that an intuition that represents its object as qualitatively determinate is brought about by a concept-guided act of synthesis it follows that for Kant having a perception that represents its object as being qualitatively determinate involves judging that it has this property. Thus, since perceptions on their own do not represent objects as qualitatively determinate at all, Kant does not assume that the way in which a perceived object is represented as qualitatively determinate by a perception is different from the way in which it so is represented by a judgment. At least that is what I have argued.

If we take my objections to Matherne's claims together, we arrive at a conception of the relationship between sense, imagination and understanding that differs significantly from her position. According to this alternative picture, the generation of all of our intuitions requires that imagination synthesizes sensible representations. Yet, this synthesis is never guided by concepts. Intuitions phenomenally represent objects as a having a determinate or indeterminate spatio-temporal form and phenomenally represent objects as having qualities, but they do not involve identifying their objects as having a property shared by objects of a certain kind. For example, in having an intuition of a red ball, the ball looks spherical and red to me, but I do not identify it as being spherical, as being red or as being a ball. In order to identify an object of intuition in this way, I have to apply a concept to it either in a problematic or an assertoric judgment.