The notion of conceptual content can be characterized by means of mental states and linguistic utterances that possess content of this kind. A paradigm case for a state with conceptual content is a propositional attitude - an attitude towards a proposition or a thought whose content remains the same in spite of a change in the attitudinal mode. Thus I can believe, desire, hope, fear etc. that the soccer team of United States will win the 2014 FIFA World Cup in Brazil. York Gunther (2003) calls this property of conceptual content its force independence: the content of a propositional attitude can be individuated independently of its illocutionary force. The fact that conceptual content is constituted of concepts allows us to form and grasp an infinite number of thoughts and expressions by combining concepts in different ways. This compositionality of conceptual content explains how language can be both systematic and productive and why states with conceptual content can enter into inferential relations with each other. For instance, if I believe that "Tom is a pianist" while also believing that "if Tom is a pianist, then Alice was his teacher", I can infer from the content of these premises that "Alice was Tom's teacher". The content of a propositional attitude specifies its determinate reference, which can be either an actual, conditional, historical, or a fictitious state of affairs. Each type of attitude also has its own condition of success. Thus beliefs succeed when their content is true, whereas desires succeed when their content is satisfied or desirable (Green, 1992; de Sousa, 2002).
Emotions can have conceptual content in many ways. First, some emotions qualify as propositional attitudes in an ordinary sense by having propositionally specified content. For instance, investors fear that stock prices will plummet or diplomats are relieved that an international conflict was avoided. Sometimes propositional content can be fully expressed without a that-clause. Such emotion is, for instance, pride in graduating from a prestigious university. Yet some emotions resist propositional specification of their content. "A description of someone as pitying, hating, resenting, or loving that p would be syntactically deviant", as O.H. Green (1992, p. 39) remarks. Instead, these emotions take direct objects, as when Mike pities refugees or Lisa despises a treacherous friend. Even so, a cognitive theorist may suggest that it is possible to specify the propositional content of such emotions in terms of a thought or belief that describes the respect in which the object is comprehended in the emotional state (ibid., p. 40). Thus Mike's pity for refugees involves a belief that the refugees have suffered a loss, or Lisa's despising of her friend involves a belief that the friend has betrayed her confidence. However, this response is not entirely satisfactory as the emotion is still felt towards the whole person rather than towards some aspect of him or her that figures as a reason or motive for the emotion. Therefore, it may not be possible to spell out specific propositions that define emotions with direct objects but only the type of thoughts or beliefs that essentially rather than contingently belong to these emotions (e.g. Goldie, 2000). Finally, Green (1992) suggests that evaluative judgments or thoughts can be analyzed further into semantically interrelated beliefs and desires. In this approach, fear is a combination of an uncertain belief that p and the desire that not-p; sadness a compound of a certain belief that p and the desire that not-p; hope is a synthesis of an uncertain belief that p and the desire that p, whereas happiness comes out as a certain belief that p together with the desire that p. Other emotions differ from these basic emotions by involving more elaborate beliefs.
The main problem with conceptual accounts of emotional content is that we sometimes feel emotions even if we do not believe or judge or think that our goals are affected favorably or adversely by external events. Fear of flying, fear of spiders, and pathological fears such as a fear of crossing a bridge are examples of rationally recalcitrant emotions that we can feel in spite of our judgments of being safe. The conflict between emotion and judgment is a problem for theories that analyze the content of emotions in terms of propositional attitudes. If emotions are or involve ordinary beliefs or judgments, the subject of a recalcitrant emotion logically contradicts himself in the same way as a person who attempts to believe contradictory statements p and not-p. The fact that emotional recalcitrance is both logically and psychologically possible, whereas it is impossible in both respects to believe that p while consciously believing that not-p, indicates that the content of all human emotions cannot be fully analyzed in terms of propositional attitudes (see e.g. Oeigh, 1994).
Cognitive theories have tried to adjudicate the problem of recalcitrant emotions in various ways. Strong cognitivists who analyze emotions in terms of evaluative judgments with propositional content are most vulnerable to this problem. They suggest that the extra-semantic, attitudinal qualities of emotional judgments such as their freshness, haste, self-involvement, and intensity allow these judgments to conflict with nonemotional evaluative judgments (e.g. Solomon, 1988; Armon-Jones, 1991; Nussbaum, 2001). However, these proposals are metaphorical at best and incapable of offering a satisfying response to the problem of recalcitrance. Other cognitivists, sometimes referred to as "quasi-judgmentalists" (D'Arms & Jacobson, 2003), have suggested that the propositional content of emotions should be analyzed in terms of evaluative thoughts that - unlike beliefs and judgments - do not involve an epistemic commitment to the truth of the content (e.g. Neu, 1977; Greenspan, 1988). The appeal of this solution is reduced by its reliance on defining-propositions methodology that unites both traditional versions of cognitivism because it is not obvious that the contents of recalcitrant emotions are inferentially related to the contents of evaluative judgments (D'Arms & Jacobson, 2003; Döring, 2004). If the content of a recalcitrant emotion were analyzable in terms of a propositional attitude short of belief, the emotion would still be in rational conflict with the subject's assented belief, which should dispel the emotional thought or construal in the long run. However, the persistence of recalcitrance of some, especially phylogenetically ancient, emotions shows that this need not happen, which in turn suggests that those recalcitrant emotions do not have conceptual content. Finally, perceptual cognitivists invoke the idea of gestalt perception in which feeling-infused thoughts or concern-based construals hold the mind in more or less temporary grip (Goldie, 2002; Roberts, 2003). The affect explains the salience and compellingness of an emotional thought or construal and its capacity to persist in spite of conflicting belief or evaluative judgment. Perceptual cognitivists offer the best solution to the problem of emotional recalcitrance. However, their main challenge is to explain how emotions come to have this kind of affective-cum-evaluative content. The claim that emotions are sui generis in this respect amounts to little more than giving up the challenge of explanation in favor of a dogmatic view that hybrid states of this kind exist.
In spite of these problems with rationally recalcitrant emotions, cognitive theories are capable of accommodating the lag it normally takes from disconfirming evidence to the propositional content of emotion to "sink in", so to speak. An emotion whose conceptual content has been disproved does not immediately go away because the other noncognitive components of an emotion such as strong physiological arousal, bodily sensations, and action tendencies are capable of keeping it going for a while. A lingering emotion may also skew the subject's epistemic perspective, urging him or her to come up with other reasons for the emotion, as Peter Goldie (2004) has pointed out. Yet this kind of lag is common to all cognitive phenomena and should not be mixed up with the kind of persistent rational recalcitrance that is characteristic to phylogenetically ancient affect programme emotions on the one hand and to ontogenetically specific pathological emotions on the other hand. The fact that emotions of the latter kind exist indicates that the content of all human emotions is not conceptual.
-  This Fregean understanding of conceptual content is not the only game in town, however. In The Representational Theory of Mind, concepts are understood as mental representations that function as constituents of structured mental representations whose type is identified by the functional role of the representation in the subject's mind. Accordingly, the representational theory of conceptual content can be associated with the functional account of cognition discussed above. Moreover, some philosophers maintain that propositional structure is not limited to conceptual representations but applies to nonconceptual representations as well. Peacock (1992) for instance presents such a view about the content of perceptual experiences, and Döring (2004) applies it to affective perceptions (emotions). Therefore, propositional and conceptual content are not synonymous or even coextensive even if conceptual content typically has propositional structure.
-  Satisfaction and desirability are different kinds of success conditions for desires, as Green and de Sousa emphasize. My desire that p is satisfied if I get p, but the desire is not successful if the object of my desire is not desirable.