According to a classical version of emotional cognitivism, emotions are preceded by a form of value awareness (e.g., Lyons 1980, chap. 3; DeSousa 1987, 122, 133; Kenny 2003, 143ff; Mulligan 2010, 485ff; Müller 2017, 288ff; 2019, chap. 4). For example, on this view, fear of an impending recession is preceded by awareness of this prospect as dangerous, anger towards someone by awareness of her as in some way provocative or offensive.

In a recent paper, Mitchell (2019) has challenged this account, which he calls the “content-priority view” (I will follow him in using this label). Mitchell takes issue with the view’s commitment to a pre-emotional state with evaluative content.¹ According to him, the various candidates that have been proposed for this state are all problematic. Mitchell thus claims that there is at present no persuasive formulation of the view. Unless the problems he raises can be overcome, he proposes, the view does not represent a plausible competitor to rivalling cognitivist views. As he argues, the content-priority view is by no means mandatory since the main considerations in support of it can be accommodated by its rivals, too.

Mitchell’s paper is a significant contribution to an ongoing debate between opposing strands of emotional cognitivism and extends an important invitation to proponents of the content-priority view to clarify their core commitments. As I argue in this discussion, I do not think that Mitchell’s case against the view is persuasive, though. Most importantly, Mitchell misconstrues the principal motivation for the view. Properly reconstructed, this motivation provides a strong case for the indispensability of the view to an adequate cognitivist treatment of emotion. That is, it shows the view to be entailed by any account that recognizes emotions as directed towards objects.

Moreover, Mitchell’s critical survey of possible candidates for pre-emotional evaluative states rests on contestable premises. More specifically, his chief objection to what is often considered the most promising candidate depends on a questionable phenomenological constraint. I should stress, though, that, in contrast to Mitchell, I do not think that a successful case for the view must include a substantive account of pre-emotional evaluations. If the content-priority view alone can account for the intentionality of emotion, this itself makes it a rather strong contender among current cognitivist accounts.

In what follows, I develop these points. Before I do this, though, I shall explicate the view and note one respect in which Mitchell’s own explication can seem misleading (section 1). I then consider Mitchell’s discussion of the view’s main motivation (section 2) and argue that, properly reconstructed, this motivation provides strong grounds to consider it indispensable (section 3). Finally, I turn to his considerations on specific candidates for pre-emotional value awareness (section 4).

1 The Content-Priority View and the Evaluative Content View

According to proponents of the content-priority view, emotions are preceded by awareness of value. This construal of the temporal relation between emotion and value awareness is a consequence of a specific conception of the connection between emotion and value properties themselves: emotions are conceived as responses to value properties of their intentional object. On the relevant use of “response”, for someone’s emotion to be a response to \(x\) is for her emotion to be felt in light or on occasion of \(x\) or, equivalently, for \(x\) to be a reason for which she feels it.² Since reasons for which someone feels, thinks or acts some way (motivating reasons) are made psychologically available by mental states that are temporally prior to and distinct from the attitude or action they motivate, emotional responses to value are preceded by distinct states of value awareness. Compare: If Mary believes that it will rain for the reason that the sky is grey, the perception which makes this fact psychologically available as a reason for her belief is temporally prior to and distinct from this belief.

To be precise about how the link between emotions and values is conceived on this view, it is worth making explicit that the relevant use of “response” is one of at least two familiar uses. Consider the following examples:

In each of these cases, “response” (or one of its grammatical variants) serves to ascribe a reason for which a certain action or attitude is performed or held. Accordingly, there is a cognitive requirement for the reason to have registered with its subject prior to her acting (feeling) for this reason.³ On a different use, which has its home in scientific contexts, the term serves to ascribe a mere cause, rather than a motivating reason. Consider, for example, the statements:

One might, in principle, formulate the view that emotions are responses to value in accordance with this second use of “response”. This formulation does not entail that emotions are preceded by value awareness since causes are not subject to the cognitive requirement on motivating reasons: one’s skin burns in response to high exposure to sunlight irrespective of whether one is aware of the sunlight. However, this is not how proponents of the content-priority view think of emotions. Their view is motivated by considerations on the ascription of motivating reasons to emotion.⁴ (I elaborate on the view’s principal motivation in the next section.)

Since the evaluative properties to which, on the content-priority view, emotions respond feature in the intentional content of a state distinct from the emotion, the view can seem unorthodox. That is, it can seem to contrast with the popular view that emotions themselves have evaluative content.⁵ According to Mitchell’s (2019, 772) explication, it in fact involves the explicit denial of this view and thus qualifies as its rival.

It seems to me that this explication has to be treated with care, though. This is because proponents of the content-priority view recognize emotions as directed at their objects under a specific evaluative aspect (cf. Lyons 1980, 77ff, 99ff; DeSousa 1987, 122; Kenny 2003, 143ff; Müller 2017, 288ff; 2019, 69ff).⁶ As they suppose, to fear something is to fear it as a danger, to be angry with someone is to be angry with her qua offensive. According to an influential account of intentionality, this is precisely to accord evaluative content to emotions. For a state to have intentional content is, on this account, for it to be directed at something under a specific aspect (which is commonly called, following Searle 1992, 131, its “aspectual shape”).

The fact that, on the content-priority view, evaluative aspects are supplied by a distinct state should here not be taken to conflict with the claim that they qualify the intentional content of the emotion rather than exclusively of the preceding state. While, arguably, in the case of some intentional states, the aspect under which those states are directed at something is supplied by that very state, there is no presumption that this has to be so.⁷ That an intentional state is directed at something under a certain aspect does not imply that it is this same state which presents that thing under this aspect. Indeed, it seems if the view required this, it would end up mischaracterizing paradigm cases of intentional attitudes as lacking content. On a common conception of belief and desire, what one believes or desires is believed or desired under a certain aspect which is supplied by another, preceding cognitive state. Consider my desire that the government acquire more covid vaccine. In desiring this, I desire something (a certain prospect) presented in a specific way, which is specified by the proposition that the government acquire more covid vaccine. I may not desire that same thing as characterized by the proposition that the government acquire more tozinameran. Here, the relevant aspectual shape is not made available by the desire itself, but by the state of entertaining or grasping this proposition, which is a necessary precondition of having this desire.⁸ (The same goes mutatis mutandis for doxastic attitudes.) As my later considerations on the evaluative aspect qualifying objects of emotion suggest (section 2), the fact that the aspectual shapes of some intentional states are supplied by distinct states might plausibly be due to the specific way in which these states are directed.

If we suppose, then, that there is a plausible sense in which emotions have evaluative content on the content-priority view, Mitchell’s way of contrasting the content-priority view with rivalling cognitivist accounts thus can seem a little puzzling. For Mitchell’s explication to be intelligible as pointing to a genuine disagreement between cognitivists, it is therefore important to note that, for him, the attribution of evaluative content to emotion entails that emotions constitute awareness of their objects as having an evaluative property (cf. 2019, 771). This understanding, too, recognizes emotional objects as presented under evaluative aspects. But in conceiving of emotions as forms of awareness of value, it is committed to a different conception of the relation between emotion and value awareness than the content-priority view. While, according to the latter, we are already aware of the value of an emotion’s object prior to the emotion, the former takes this awareness to be supplied by the emotion itself.

Admittedly, from what Mitchell writes, it is not entirely transparent to me why he supposes that the attribution of evaluative content to an emotion entails that it constitutes a form of value awareness. Perhaps Mitchell implicitly assumes that aspectual shapes are always supplied by the very state to which they are ascribed. Given this restriction, for an emotion to have evaluative content implies that the emotion itself (rather than some prior state) presents its object as (dis)valuable. It then seems that one comes to be aware of the value of an emotion’s object only in having the emotion. If what I just said about aspectual shapes is correct, this requirement is by no means trivial, though, and there are reasons to resist it. Whether or not Mitchell is ultimately making this assumption, his understanding of evaluative content, in any case, strikes me as somewhat restrictive in that it precludes its attribution to states other than forms of value awareness. Although this, in turn, has certain ramifications for how we are to chart the territory of extant cognitivist views, I shall here not quibble further over Mitchell’s take on emotional content, but instead focus on whether his attack on the content-priority view, as explicated here, is successful.

2 The Intelligibility of Emotion

The content-priority view is motivated by the thought that evaluative properties make emotions intelligible.⁹ More specifically, its motivating thought is that something intelligibly qualifies as the intentional object of an emotion only under a specific evaluative aspect. Properly spelled out, this is taken to imply that emotions are responses to value.

In discussing the motivations of the content-priority view, Mitchell clearly pays heed to considerations on the intelligibility of emotion. He explicitly states that the view aims to account for the observation that emotions “make sense” as responses to specific values (2019, 774). However, Mitchell does not get the relevant notion of intelligibility into focus.

One problem is that Mitchell’s phrasing of this observation is ambiguous. On one reading, to say that an emotion makes sense is to claim that it is appropriate or justified. On this reading, Mary’s fear of a meandering dog makes sense as a response to the dog’s dangerousness insofar as what she responds to is a reason for her to be afraid (a normative reason for fear). On a different reading, to say this is to affirm the very cogency of its ascription to someone. On this further reading, Mary’s fear of the dog makes sense as a response to danger insofar we can coherently conceive of her as being afraid of it given that her fear is motivated by danger. Unfortunately, Mitchell does not recognize the content-priority view as concerned specifically with the latter notion of intelligibility. Yet, as proponents of the view have variously stressed, they are interested in basic conceptual constraints on the proper ascription of emotions. As they argue, we can coherently conceive of someone as having a certain emotion directed at \(x\) only if we presume that her emotion is directed at \(x\) in response to the (real or apparent) value of \(x\) (cf. Lyons 1980, 78, 99ff; DeSousa 1987, 122, 133; Kenny 2003, 143ff, 51–52; Müller 2017, 288ff).¹⁰

A further problem is that Mitchell takes it that intelligibility in the primarily relevant sense is first-personal: it is about what makes sense for the subject of the emotion to feel. First-person intelligibility is moreover qualified by Mitchell as “experienced intelligibility”, that is, a kind of intelligibility that is typically, though not always, conferred by emotional experiences themselves and does not rely on accompanying mental states (2019, 792). This is misleading since proponents of the content-priority view are explicitly concerned with conceptual constraints on ascriptions of emotion. Their focus is thus on canonical ways of attributing emotions in thought and language rather than emotional experience. Also, as I read Mitchell (2019, 792), it is perfectly coherent to ascribe emotions to people that are not experientially intelligible to them. Experiential intelligibility thus clearly differs from intelligibility in the sense of coherent conceivability. Note further that, since adherents of the content-priority view are proposing a view of emotion in general, there is also good reason why their focus is not on the specific class of experientially intelligible emotions.¹¹

The main problem with this failure to delineate the appropriate notion of intelligibility is that Mitchell ignores why the content-priority view has been considered indispensable to an adequate account of emotion. That is, Mitchell seems to be unaware of the main argument in its favour. To show this, let me consider Mitchell’s take on the claim that the content-priority view alone can adequately account for the intelligibility of emotion.

To assess whether this claim is warranted, Mitchell considers a common form of reason explanation discussed by Mulligan (2010, 485–486). Consider, for example:

Mulligan proposes that such explanations provide strong grounds for thinking that value awareness is, as he puts it, “outside of” emotion (2010, 486). In response, Mitchell makes little effort to reconstruct Mulligan’s reasoning, but gives the proposal short shrift. As he comments, “such third-person reports are surely not decisive with respect to philosophical theories, or indeed how we frame the intelligibility of the relevant emotional episode as experienced” (2019, 792).

Now, in line with my above remarks, Mitchell’s complaint seems unjustified inasmuch as he criticizes Mulligan’s considerations as failing to speak to the experienced intelligibility of emotion. This is not Mulligan’s concern, which is with constraints on the explicit ascription of emotions. And insofar as Mulligan is defending a view of emotion in general, it makes sense that he chooses a different focus.

It is also not clear why we should agree with Mitchell that such explanations are not decisive with respect to theories of emotion in the first place. While he considers this to be obvious, I think it requires explanation. After all, the content-priority view is a theory that explicitly assigns an explanatory role to values. According to this view, emotions are responses to evaluative properties and, as such, explained by motivating reasons provided by them. One should thus think that the view stands and falls with whether it is borne out by the way we do explain emotions. Looking more closely at the above form of explanation, we can see that it in fact confirms the view. To say that Mary is afraid of the dog because it is dangerous is to say that she is afraid of it for the reason that it is dangerous. Since motivating reasons are made available by states prior to and distinct from the motivated response, her emotion is preceded by a distinct state of value awareness.

I also do not think it is appropriate here to dismiss appeal to this type of reason explanation as a case of overemphasizing the philosophical significance of “folk reports”. If this is what Mitchell has in mind, I think it is worth stressing that we are here concerned with reason explanation. The appropriate way to get clear on the structure and content of reason explanations of a given attitude or action is to look at common explanatory practice. It is not as though we will find a more “theoretical” form of reason explanation by consulting affective science. Nor it is obvious that we can read such explanations off first-person experience alone.¹² Inasmuch as the content-priority view is, fundamentally, a claim about motivating reasons for emotions, it thus seems perfectly warranted to consider reports on the lines investigated by Mulligan in assessing the view.¹³

As far as I can see, the most plausible way to understand Mitchell’s scepticism about the dialectical import of such explanations is to think of him as calling attention to the fact that they are by no means mandatory. Note that we also often cite non-evaluative features as reasons for emotion:

Hence, it can seem that, as far as common explanation goes, we are not committed to regarding emotions as responses to value (cf. Teroni 2007, 411; Deonna and Teroni 2012, 96ff).

However, while this may seem a discerning objection, it actually rests on a misunderstanding. It misses that reason explanations in terms of value contribute to the very coherence of emotion ascriptions. That is, since Mitchell does not get the relevant notion of emotional intelligibility into focus, he fails to appreciate that, far from being optional, such explanations constrain our very grasp of emotions qua directed. Yet, to ignore this is in fact to ignore the main consideration in favour of the content-priority view and, accordingly, what makes this view seem mandatory.¹⁴

To make this consideration explicit and show what Mitchell ignores, we must look more closely at the idea that emotions are directed. It is in fact uncontroversial among cognitivists that emotions rely for their objects on some prior awareness of them (known as their “cognitive base,” see Mulligan 2010, 476; Mitchell 2019, 791; Milona and Naar 2020, among many others). For example, unless we suppose that Mary has perceived the dog, we cannot properly conceive of her fear as being about the dog. Similarly, it does not seem coherent to suppose that someone who is glad or angry that p has not apprehended that p. Consider how strange it would be to say that Sam is glad that he has won the race but that he has no idea that he has.

A straightforward way to explain this requirement is by noting that emotions are responses to their object (cf. Kenny 2003, 51–52; Dietz 2018; Müller 2017, 287–288; 2019, 68–69). If Mary is afraid of the dog, it follows that she is afraid in light of or on occasion of the dog’s presence. It sounds just as bizarre to say that Mary fears the dog but that she is not afraid in light of its presence. What Mary fears is what she responds to with fear, i.e., the reason for which she is afraid. Thus, she must have registered the dog.

If this is account is accurate, then ascriptions of emotions imply that they are felt because of their object (where “because” specifies a motivating reason). Although this explanation does not explicitly refer to value properties, it is crucial to note that it would not work if the reason specified were the emotion’s object simpliciter. We fail to comprehend the dog as something that should upset Mary unless we suppose that she responds to a specific feature of it. For there to be a genuine explanatory relation Mary must respond to the dog qua danger. It is only if the dog is apprehended as being of concern to Mary in this respect that we can understand its presence to be a reason for which she is afraid. This is why emotion ascriptions support the content-priority view. Their coherence depends on the implied explanatory relation between the emotion and its intentional object. And this relation depends on the latter exemplifying a certain evaluative property.

To further support this account, note that it relies on a widely accepted general constraint on reason explanations. This constraint states that something qualifies as someone’s motivating reason for an action or attitude only if it is taken by her as a reason to perform the action or hold the attitude. Thus, even if the reason for which someone feels an emotion is not actually a reason to feel it, the cogency of explanations in terms of this reason requires conceiving of her as responding to something that, to her, presents itself as a corresponding normative reason.¹⁵ In this respect, the intelligibility conferred by such explanations can be conceived as a kind of rational intelligibility.

This constraint on motivating reasons nicely explains the connection between emotion ascriptions and explanations in terms of value. The evaluative property to which an emotion is a response, according to the content-priority view, is a reason to feel it: danger is a reason to be afraid; a genuine offence speaks in favour of anger. Accordingly, if Mary fears the dog under the aspect of danger, she responds to (what she apprehends as) a reason to be afraid. In conceiving of her in these terms we secure the cogency of the explanation implicit in thinking of her as fearing the dog.

At this point, one might want to reply that this requirement is satisfied also if we think of emotions as responses to non-evaluative features of their object. After all, such features can be normative reasons, too. Thus, plausibly, the dog’s being aggressive and sharp-toothed are reasons for Mary to be afraid of it (cf. Deonna and Teroni 2012, 96ff).

However, this reply won’t do. While such features can clearly be reasons to feel emotions, too, their status as normative reasons depends on their connection to value. It is only in their capacity as grounds of danger that the dog’s aggressiveness and sharp teeth speak in favour of fear. To appreciate this, note that these features may in principle be reasons for different emotions. For example, for someone who relies on the dog to protect her property, they might be reasons for contentment. In this case, they favour a different emotion in virtue of their relation to a different value (the importance of keeping the property safe). This suggests that, for non-evaluative features to qualify as normative reasons for emotions, they must be suitably linked to a specific value. The cogency of “Mary is afraid because of the dog’s aggressiveness and sharp teeth” requires thinking of her as responding to these features as grounds of danger and thus as having apprehended the dog in evaluative terms.

In light of this argument, it seems that for us to so much as coherently think of emotions as taking objects, we are committed to the content-priority view. Since nowadays most philosophers, and certainly all cognitivists, recognize emotions as directed, Mitchell seems thus wrong to consider the view dispensable in favour of competing accounts.

In order to properly situate this view of emotional directedness, it is worth adding that it also sits well with further observations about the relevant type of reason explanation and that there are clear analogues in the case of other, familiar attitudes and actions. As DeSousa (2011, 72) notes, there is something trivial about reason explanations of emotion citing corresponding value properties. They seem vacuous, yet perfectly appropriate.¹⁶ If the very coherence of such ascriptions depends on our conceiving of the emotion as a response to the corresponding value, this is just what is to be expected. Such explanations are parallel to explanations that instantiate schemes like

Such explanations seem vacuous because there is a conceptual requirement for the respective action to be motivated by a particular (dis)value.¹⁷ Like these actions, emotions can be understood as conceptually committed to a specific reason. They form part of a familiar and rather large class of psychological phenomena that are united by their responsive character.¹⁸ In this respect, the fact that emotion ascriptions are subject to a conceptual constraint involving values as reasons should not make them seem in any way peculiar. Note, further, that there are analogues to this kind of conceptual commitment in the case of responses in the mere causal sense. Consider, e.g., a property like sunburn, whose ascription is subject to conceptual constraints concerning its cause: we cannot coherently conceive of an inflammation of the skin as a sunburn unless we think of it as caused by exposure to sunlight (cf. Dardis 2008, 115).¹⁹ Attitudes and actions that are conceptually committed to certain reasons can plausibly be thought of as the reason-theoretic analogue of reactions that are committed to causes in this way.

3 The Indispensability of the Content-Priority View

Supposing these remarks are on the right track, I do not think that the credentials of the content-priority view as a plausible contender among cognitivist theories of emotion are contingent on further characterizing the prior value awareness. If emotions are intelligible as directed only as responses to value, this should in fact be sufficient to accept it.

This consideration in support of the content-priority view can be framed as an argument of a classical transcendental form: If we are to avoid making it unintelligible how emotions can be directed at something we must maintain that they are responses to this thing as (dis)valuable. This argument takes it as agreed that emotions are directed and proceeds to its conclusion via a constraint on the intelligibility of their directedness. Transcendental arguments of this sort are not uncommon in the theory of intentionality.²⁰ They differ from their traditional anti-sceptical cousins which start from a psychological premise and proceed to a conclusion about the mind-independent world. The type of argument in question, in contrast, both starts from a psychological premise and has a psychological conclusion. In accordance with a common dialectical purpose of transcendental arguments, the point of this transcendental argument for the content-priority view is to show that the value responsive-character of emotion is a condition on something that all cognitivists, including Mitchell, take for granted.

Maybe there will be some resistance to this inference in light of Mitchell’s scepticism about pre-emotional value awareness. If a survey of the various candidates that one can think of in this context turns out not to yield a single viable option, one might think this should give us pause. The appropriate thing to do given Mitchell’s charge, one might think, is to revisit the considerations offered in favour of the content-priority view and to try and find fault with the constraint on intelligibility which underwrites this inference.

Though this may seem a natural response, I don’t think it is successful. To see why, it is important to be clear on the status of this constraint. Considering the close analogy between emotions and reason-committed actions made explicit at the end of section 2, this constraint seems no less integral to our understanding of emotions than the idea that in criticizing someone we respond to some purported wrongdoing or the idea that in thanking someone we respond to some purported benefit is to our understanding of these actions. If this is right, letting go of this constraint is not a palatable option. Consider, for instance, what would become of our grasp of criticizing someone if tokens of this action were no longer conceived as responses to a purported mistake? I don’t think we need to be committed to a purely descriptive approach to the metaphysics of mind and action to find this prospect deeply dissatisfying.

To back up this point, let me elaborate somewhat on the place of this conceptual commitment within our understanding of the corresponding phenomena. As some further reflection suggests, in the case of many reason-committed actions as well as the emotions, the intelligibility constraint on their directedness is at the same time a constraint on their intelligibility as the kinds of psychological phenomena they are. A helpful way to see this is by considering joint explanations of emotions and actions such as e.g., “\(S\) responded with indignation and harsh criticism to \(x\)’s wrongdoing” or “\(S\) was grateful to \(x\) and thanked \(x\) for \(x\)’s beneficence.” Such explanations are perfectly cogent. Moreover, it seems that part of their cogency is a matter of the respective emotion and action making sense as cognate responses to a specific (dis)value. To refer to them as cognate responses is not, or not merely, to refer to their common motivating reason, but also to call attention to the fact that they are similarly valenced: both action and emotion are negative (positive) responses.²¹ I take it that this aspect is fundamental to our conception of each of these actions and attitudes. Intuitively, criticizing or thanking someone is essentially a valenced response. These actions are a form of sanction (in a suitably broad sense of “sanction” that includes forms of positive acknowledgment or appreciation). The same goes for the negative (positive) character of indignation (gratitude). While we may wish to reserve the term “sanction” for certain valenced actions, we can highlight this similarity by speaking of indignation (gratitude) as a case of taking a positive or negative stand or position on something or, perhaps more colloquially, as form of (dis)approval.²² However, and crucially, our appreciation of these phenomena as sanctions or position-takings is contingent on their recognition as value responses. Criticism is coherently conceivable as a negative response only in virtue of the (purported) disvalue of its target. It seems very puzzling to suppose that this target is negatively sanctioned without being sanctioned for its (purported) negative import. (Mutatis mutandis for the act of thanking someone.) Similarly, indignation’s character as a negative response depends on its character as a response to disvalue: it makes little sense to conceive of indignation as the taking of a negative stand on something without conceiving of this stand as being taken because of this object’s (purported) negative import. (Mutatis mutandis for gratitude.) Now, if the intelligibility of emotions as position-takings hinges on our ability to conceive of them as responses to value, then we should be wary of dispensing with the intelligibility constraint I have made explicit. That is, what is at stake here is not the “mere” fact that emotions take objects, but their character as the specific attitudes they are. To ignore their commitment to particular values as motivating reasons is to crucially impoverish our grasp of them in the same way that ignoring this commitment in the case of valenced actions makes them unrecognizable as the actions they are.

I take this to provide strong warrant for thinking that this constraint is not really up for dispute. Accordingly, I doubt that the suggested response to Mitchell’s survey is feasible.²³ In assessing this response, it is moreover worth pointing out that it also loses much of its initial force if we look to similar dialectical contexts outside the philosophy of emotion. Once we somewhat broaden the perspective, it is not at all evident that the apparent shortage of plausible candidates for a form of awareness to which we are committed by virtue of our core understanding of a certain phenomenon requires revisiting this understanding. Consider, for instance, the kinds of implicit cognition that have been posited by major contemporary accounts of conceptual or semantic competence in order to explain how someone may possess a word or concept without having a fully articulate understanding of the principles governing its use (cf. e.g., Peacocke 1998; Toribio 1998). These forms of cognition are introduced in response to the recognition that extant philosophical conceptions of propositional knowledge do not account for our cognitive relation to these principles in accordance with this plausible restriction. In fact, one of the main explanatory purposes of introducing them is very much in line with the role played by pre-emotional value awareness on the content-priority view, i.e., to acknowledge certain types or linguistic or concept-involving behaviour as rationally intelligible (cf. Peacocke 1998; Toribio 1998).²⁴ This is not to say, of course, that they are posited without any concern for whether they actually form part of our psychology. But their point is first and foremost that of sustaining what is independently recognized as an adequate characterization of semantic or conceptual competence. This is normally not in and of itself seen as compromising the status of the respective account as a worthwhile contender in the field. And for a good reason: It seems misleading to be sceptical against an account that is founded on core features of our grasp of some phenomenon for the sole reason that it fails to fully align its psychological commitments with those of extant philosophical psychology. It may well turn out that, in the end, the apparent lack of candidates for the required cognition proves due to implicit theoretical strictures on our ontology of the mind that need loosening, rather to any fault with the account itself.²⁵

While I thus remain unconvinced that the content-priority view should be deemed unpersuasive unless a more substantive account of pre-emotional value awareness is supplied, I still think it is worth finally taking a look also at Mitchell’s main objection to what some authors see as the most promising way of elaborating the view. In this way, I hope to moreover show that Mitchell’s survey of suggestions for this form of awareness does in fact not succeed in demonstrating that we are short of plausible candidates.

4 The Phenomenology of Responding to Value

As the content-priority view is sometimes elaborated, emotions are preceded by states of evaluative “seeing-as” or axiological perceptions of their objects as (dis)valuable (Müller 2019, chap. 5).²⁶ Mitchell’s objection to this proposal is based on the constraint that, in the case of paradigmatic emotional experiences, pre-emotional evaluations must be discernible phenomenologically from emotion (cf. 2019, 783).²⁷ For Mitchell this requires that, when undergoing such experiences, the evaluation be phenomenologically conspicuous as preceding the emotion. As he rightly notes, this requirement is not met. Consider common “quick-fire” emotions, such as a bout of terror felt in response to a suddenly approaching car when intending to cross the road (cf. Mitchell 2019, 785). Such emotions clearly purport to be immediate reactions rather than consequences of a prior perception of value. Mitchell takes this to undermine the proposal’s phenomenological credentials.

While it is fair to enquire about the phenomenological plausibility of this formulation of the view, this objection is too quick, though. It is not obvious why we should follow Mitchell in requiring that evaluative perceptions be discernible from emotion in the throes of experience. A straightforward reason to reject this requirement might be that paradigmatic emotional reactions are often simply too quick for us to notice that they are preceded by prior evaluations at the time. That is to say that while the emotion follows upon the perception of a value property, this very fact may itself not be conspicuous to us. It may take some post hoc enquiry to properly discern its psychological antecedents and, accordingly, that it is motivated by a specific value property.

While my earlier remarks on the intelligibility of emotion might be adduced in support of this reply, I do not think that it is in fact necessary to advert to them. In keeping with Mitchell’s concern with emotional phenomenology, we can also provide first-person grounds to take this line.

Note, first, that it is not uncommon for responses to occur too fast for us to be able to fully discern their mental antecedents and appreciate what they are motivated by. Consider reflex-like actions, such as automatically hitting the brakes when spotting an obstacle on the road or unreflectively backing away from a close talker. As with quick-fire emotions, here we seem to respond without conscious prior perception of the situation in evaluative terms. As things seem first-personally there and then, we immediately act on the bare perception of its basic spatial layout. However, we can retrospectively check this impression. That is, we can probe the source of these responses, e.g., by recalling different features of the situation or by imaginatively modifying it, in order to find out what, as Pugmire (2006, 17) nicely puts it, “clicks”. Such tests are likely to confirm, for example, that my inclination to back away from a close talker was motivated by her proximity. By imaginatively varying the talker’s relative distance, I may even succeed in further specifying the motive: there is a certain invisible yet significant boundary (surrounding, perhaps, my peri-personal space) relative to which she was too close. To think of the action in these terms is to apprehend it as a response to an intrusion and hence to the situation construed in evaluative terms.

It seems that these same tests are applicable also in the case of quick-fire emotions. I can similarly probe, for example, the source of my terror at a suddenly approaching car. Thinking back to the incident and focusing on what emotionally resonates with me, I may find out that I was responding specifically to the suddenness of the car’s appearance and its speed. Plausibly, there is room for even further precision. By imagining counterfactual variations concerning the car’s relative distance, speed and direction, as well as by picturing the respective consequences, I may even come to see that I was frightened specifically by the car’s being too fast for me to be confident in my ability to avoid collision and the corresponding anticipated injury. In realizing this, I apprehend my terror as a response to an impending adversity or threat. Crucially, since I apprehend the car qua threat as motivating my terror, I also understand the situation to be one in which the car was apprehended in evaluative terms prior to emotion.²⁸

There is, perhaps, a worry here that this type of procedure is prone to post hoc rationalization and self-deception. Thus, one might wonder how we are to be sure that what we determine to be a motivating reason in this way did actually motivate our response. Given this concern, one might then come to question its use in answering Mitchell’s objection.

However, while this type of mnemonic-imaginative reconstruction of responses, like other forms of post hoc enquiry into motivating reasons, is clearly not immune to mistakes, liability to error is reduced in this case by the procedure deployed for validating candidate reasons. To say that the features to which I selectively attend must “click” is to hold hypotheses and judgments about my reasons subject to a specific kind of experiential confirmation. Whether some aspect of the situation plausibly constitutes a reason for which I feel an emotion is a matter of whether it palpably resonates with the emotion (or my memory of it).²⁹ This form of validation is not subject to direct voluntary control and also surprisingly impervious to attempts at rationally persuading myself of the (purported) plausibility of certain candidates. In this respect, the procedure is considerably more reliable than common ratiocinative forms of post hoc explanation which rely exclusively on cognitive forms of validation, such as considerations of coherence with other beliefs I may hold about the circumstances (or counterfactual variations of them). In consequence, I don’t think we have strong grounds to think it too undependable to be of much help in defending the proposed formulation of the content-priority view against Mitchell’s charge.

If we, then, assume that this procedure can aid us in probing the source of responses in these cases, we should insist that there is room to question whether phenomenologically immediate emotions threaten this proposal. That is, if my examples illustrating this procedure are cogent and we can retrospectively detect evaluative states that precede these reactions, it seems plausible that quick-fire emotions happen too quickly for us to discern the prior state in responding. This in turn casts doubt on Mitchell’s requirement to this effect and hence his principal objection to this proposal.