B. Noninferential nonconceptual content
The logical version of nonconceptual content maintains that states with nonconceptual content are not inferentially related to each other or states with conceptual content. Perception is the paradigm mental state in exemplifying this aspect of nonconceptual content as well.
Tim Crane (1992) elucidates nonconceptual content in terms of those logical properties of conceptual content that content of this kind lacks. Crane distinguishes between three kinds of inferential relations: logical, semantic, and evidential. The logical and semantic relations come down to the idea that "if a thinker has a belief, then he or she must also have many others... [because] the content of any intentional state depends, to some extent, on the contents of the others" (Crane 1992, p. 145). Thus if I believe that p while also believing that q, then I also believe the obvious logical consequences of these beliefs, such as p and q, not (p and not-p), and not (q and not-q). Furthermore, I believe propositions that are semantically related to p. For instance, if the content of p is that cheese is nutritious, then by believing p I also believe that cheese is edible because anything that is nutritious is also edible. Finally, evidential relations refer to the sensitivity of beliefs to perceptual evidence as well as to other pieces of conceptual evidence. If I believe that it is raining outside then I believe that I will get wet if I go out without an umbrella or waterproof clothing. If my perception does not support this belief, I revise my belief accordingly.
Crane argues that beliefs are holistically related to one another by all three kinds of inferential relations - logical, semantic, and evidential - whereas perceptions lack all these relations. First of all, "there is no such thing as deductive inference between perceptions. If I perceive that a is F, and I perceive that a is G, there is no such thing as inferring the perception that a is F and G" (Crane 1992, p. 152). It is also possible to have perceptions with explicitly contradictory contents whereas this is impossible in the case of beliefs. The Waterfall Illusion in which water seems to be both moving and not moving is a case in point. No semantic inferences from perceptual content are possible either. For instance, if one believes that a table is brown, one can also infer from this belief that the table also has shape because anything that has color also has some shape. In contrast, "the content of the perception that the table is brown already contains the perception of its [actual] shape" (ibid., p. 153). Therefore, unlike believing that p which entails that "there are other beliefs that you ought to have if that belief is to have the content p... to perceive that p, there are no other perceptions that you ought to have. ...You simply perceive what the world and your perceptual systems let you perceive" (ibid., p. 154, original italics). A visual perception of any object gives rise to a series of anticipations of how the object would look from different angles. Yet this does not entail that one ought to perceive the object from another angle in the anticipated way, nor does a failure to see the object in the anticipated way preclude the perception from having the content it has. The object we take to be a barn may turn out to be a barn façade when we see it from another angle. Finally, perceptions are resilient to conclusive counterevidence, which shows that they are not evidentially related to other states. The Müller-Lyer visual illusion is a well-known example: conclusive evidence that the two lines are the same length does not adjust our perception accordingly. All these differences in the properties of belief and perception indicate that perceptual content is not inferentially related to the contents of conceptual representations and therefore not subject to the constitutive ideal of rationality.
Noninferential structure is Sabine Dörings main argument for the nonconceptual content of emotion and perception. She illustrates this claim with the phenomenon of "conflict without contradiction". Döring (2003, p. 223) claims that the emotional content "resembles the content of sense-perception in that both kinds of representational content need not be revised in the light of belief and better knowledge". She compares emotions to the Müller-Lyer visual illusion and remarks that emotions may persist in the face of belief or knowledge in the same way as our visual perception that the two lines differ in length survives our belief that the lines are the same length. Döring argues that this kind of recalcitrance is not restricted to evolutionarily basic emotions and pathological emotions but applies to ordinary human emotions as well. Love for an unfit partner may linger on in spite of the partner's abusive behaviour, or one may feel guilty for an accident even if one knows that the accident could not have been prevented. These examples indicate that a subject can coherently have a sense perception or experience an emotion of the content that p, while consciously believing that ~p, whereas this is impossible in the case of explicitly contradictory beliefs. In spite of a rational conflict in content, no logical contradiction is involved, because there are no inferential relations between the contents of emotion or perception and belief. Instead, perceptions and emotions noninferentially justify perceptual beliefs and evaluative judgments, respectively. Neither are there inferential relations between emotions and other states, including other emotions. There is no such thing as inferring an emotion or a sense perception with a conjunctive content from the contents of distinct emotions or sense perceptions.
The question whether emotional content is noninferential boils down to the question whether or not it is subject to the constitutive ideal of rationality. Donald Davidson (1980) famously argued that a person's individual beliefs, desires, intentions, and actions as well as many emotions are intelligible only insofar as they fit into a coherent and holistic pattern that also determines their rationality. Perception is not subject to the constitutive ideal of rationality due to its noninferential structure, and insofar as emotion and perception resemble each other in this respect, emotion should not be subject to this ideal either. However, Bennett Helm has presented an elaborate theory of the nature and rationality of emotions on the basis of an opposite view, arguing that emotion participates in the holistic ideal of the mental.
In his Emotional Reason (2001), Helm proposes that emotions are evaluative feelings of import that both constitute and display the value of their focus, a particular object. Moreover, emotions with a common focus impose rational constraints on each other and on the subject's other states, including desires and evaluative judgments. Thus, individual emotions are warranted by virtue of their being elements of a projectable pattern of rationality, formed by the subject's felt evaluations and evaluative judgments. This means that one is prima facie committed both actually and counterfactually to experiencing rationally required emotions if the focus of one's emotion is or would be affected favorably or adversely. For instance, if Sam fears that his prize Ming vase is about to be destroyed, he is committed to hoping that the vase will remain intact, feeling relieved if the vase escapes unscathed, and becoming sad or angry if the fear of its breaking is borne out. Helm admits that there may be occasional and isolated gaps or anomalies in the overall pattern of rationality. Sam may for instance be too exhausted or depressed to actively care about his Ming vase. But if Sam generally does not feel relieved when he sees the vase escape undamaged from dangers or shocked when he witnesses its breaking into pieces, and so on, we find it hard to believe that Sam was afraid of its breaking in the first place, because he ought to have felt one of those ways if the vase has import for him. Paraphrasing Crane, 'to feel that p there are other emotions that you ought to feel about p, where this ought is not reducible to mere anticipation of other emotions but represents a requirement of rationality.
Obviously, Sam's failure to feel relieved when he sees the vase escape undamaged does not rule out the possibility of his having felt afraid of the vase's breaking; the fact of experiencing an emotion is distinct from its rationality. Indeed, only a wider pattern of Sam's felt evaluations and evaluative judgments is capable of determining which of these responses are subjectively rational for him. To some extent, the situation is similar for perception. Other perceptions lead us to conclude that some perceptions are illusory. The perception of a bent stick in water counts as a visual illusion because the same stick is consistently seen as straight in other surroundings. However, an important disanalogy between emotion and perception emerges here. There can be no 'rational ought' to perceive the stick as straight, because we cannot bring the perception in line with our other perceptions and beliefs in any case. We renounce the belief that the stick is bent in the face of conclusive counterevidence, but the content of perception is not revisable on the basis of this evidence, which shows that perceptions are not evidentially related to other states.
Döring claims that ordinary emotions of ordinary people are similar to perception in the sense that they persist in spite of the subject's better knowledge. Indeed, evolutionarily primitive emotions, such as fear of snakes or heights, as well as pathological emotions are resilient to rational counterevidence. However, ordinary emotions of ordinary people are not like this. They are evidentially related to other states even if they do not immediately vanish in the face of better knowledge. Yet the fact that we regard many recalcitrant emotions as well as pathological emotions as irrational rather than arational, and try to get rid of them, implies that the problem with recalcitrant emotions is not so much whether they need to be revised in the light of better knowledge, but rather whether they can be so revised.
Since ordinary emotions of ordinary people are weakly or moderately rather than strongly modular, as I have argued above, the phenomenon of conflict without contradiction is an untypical case which does not exempt emotions categorically from the constitutive ideal of rationality. In fact, Döring admits this point in maintaining that emotion, perception, and belief are different cognitive systems that provide us with information about the world. More importantly, she argues that conflicts between perception and belief as well as emotion and belief are rational conflicts because they are "conflicts in content about how the world actually is" (Döring 2009, p. 240; original italics). However, the fact that the content of most human emotions, unlike perceptual content, is both semantically and evidentially related to other states suggests that only conflicts between emotion and belief are rational conflicts in a more robust, Davidsonian sense of being conflicts within a coherent and holistic pattern of mental states and attitudes. Conflicts between perception and belief are rational conflicts in a thinner sense, or rather psychological conflicts because perception unlike emotion is not subject to the constitutive ideal of rationality. These considerations show that typical human emotions do not have noninferential nonconceptual content.
-  A person who detects a barn façade may think that the content of his previous perception of the same object from another angle cannot be a barn and should be modified accordingly. However, this thought does not create an ought to perceive the object differently but only a new expectation that the object's façade character may be detectable also from the other angle by taking a closer look. Yet this expectation cannot rationally influence the content of perception: the façade either is perceivable also from that angle, or it isn't.