Apprehending Value,Position-Taking and the Manifest Image of Emotion: Responses to Commentators

I am very grateful to Ronnie de Sousa, Rick Furtak, Agnes Moors, Kevin Mulligan, Rainer Reisenzein and Philipp Schmidt for their thorough engagement with my book. Their generous and sensitive responses have prompted me to think more carefully about both my criticism of the axiological receptivity view (AR, for short) as well as my proposed alternative, the position-taking view (PT, for short). While putting pressure on the central arguments and claims of the book, some of their queries also target my methodological commitments. I am thankful for the opportunity to clarify and defend my views on these matters.

Methodology

de Sousa and Moors take issue with my decision to focus on what I call, following Sellars (1962), the “manifest image” of emotion. As I use this term, it denotes the way we pre-theoretically conceive of and experience emotions. Both are skeptical of the conspicuity and theoretical import of this image. Moors supports her concern by calling attention to what she thinks of as a plausible rivalling view of what emotions are pre-theoretically like and by canvassing alternative, purportedly more robust views of emotion, which focus on psychological mechanisms. Though ultimately indicative of a similar methodological outlook, de Sousa's worry is instead based on purely philosophical grounds. He argues that my project is threatened by externalism about linguistic meaning (Millikan, 1993; Putnam, 1973). If the meaning of a term is determined by external factors beyond the speaker's ken (e.g., the term's causal history), then we cannot assume that it is transparent from the speaker's first-person point of view. (Millikan calls this assumption “meaning rationalism”.) Yet, my method of testing intuitions about the correct application of emotion terms presupposes first-person access to meaning.

I will respond to Moors's concern about my interpretation of the manifest image in the section “Emotional feeling as position-taking” and here focus on Sousa's charge. It is worth stressing, though, that what I say below speaks to a methodological commitment which de Sousa and Moors share.

A first thing to note is that de Sousa assumes a fairly radical form of semantic externalism. My semantic considerations focus on terms for intentional attitudes. By contrast, externalism is traditionally concerned with uncontroversial natural kind terms such as “water” and “tiger” (Putnam, 1973). de Sousa thus seems to assume that externalism generalizes to attitude terms. However, whether it does is quite controversial (e. g. Sehon, 1997).

Further to this, one might actually doubt that a suitably generalized version of externalism would pose a threat to my semantic inquiries. Intuitions about correct usage are not put in jeopardy by externalist theories of meaning. By contrast, these theories usually accommodate for most intuitively correct applications of the relevant terms (e.g., “this is water”). Moreover, and crucially, although externalism opposes the idea that meaning is exclusively a matter of first-personally accessible features of use, one might see some room for externalists to allow that intuitions about usage make a significant contribution to determining meaning after all (for details, see Jackman, 2005).

I here lack the space to properly address Millikan’s (1993) notable attack on meaning rationalism. I should point out, however, that my project requires a version of meaning rationalism that is weaker than the one attacked by her. ¹ Meaning rationalism centrally involves the claim that rational speakers are able to detect sameness and difference in meaning from the first-person perspective. For Millikan, this involves the ability to reidentify meanings and contents of thought across time, which, she argues, rational speakers lack if externalism is true. Now, my methodology no doubt presupposes the ability to recognize sameness and difference in meaning. For example, my central semantic observations rely on judgments about whether ascriptions of emotions admit of the cancellation of certain contents (Müller, 2019, pp. 66–71). And making such judgments is impossible without this ability. However, it does not require reidentifying meanings or contents across time. What is needed is that we be able to determine relations between meanings we grasp at the same time. There is thus is plenty of room for failures of reidentification of meaning. Hence, for this semantic inquiry to be viable meaning need not be transparent in the strong sense Millikan takes issue with.

As far as I can see, de Sousa's externalism-inspired critique is indicative of a broader outlook, according to which philosophy of mind is supposed to be continuous with cognitive science in tracking a kind of “underlying essence” of psychological phenomena. This essence is typically conceived in terms of mechanisms, which explain the “superficial” properties reflected in ordinary psychological concepts. ²

I believe that this outlook embodies a questionable prejudice in favour of the primacy of mechanisms in accounting for the nature of psychological phenomena. Since qua underlying essences mechanisms need not be transparent to ordinary psychological subjects, according to this outlook it should in principle be possible that our common beliefs about psychological phenomena are radically mistaken. But, at least as regards intentional attitudes, this does not actually seem to be coherent possibility (Sehon, 1997). For this reason, among others (see, e.g., McDowell, 1994, 2013), I doubt that the specification of mechanisms may help us understand what it is to have attitudes. This is by no means to deny it may provide insight into what enables us to form and sustain them.

My Case Against the Axiological-Receptivity View

de Sousa is not convinced by my argument for the incoherence of AR. This argument takes its cue from the observation that emotional feeling is preceded by the very apprehension of value which AR takes it to consist in.

de Sousa first targets my assumption that one cannot apprehend what one has already apprehended. He claims that this assumption relies on a mistaken view of apprehension as an instantaneous event which marks the onset of acquaintance. According to de Sousa, apprehension is more plausibly conceived as a process that results in a state of acquaintance, which can vary along dimensions such as clarity, detail and valence. We ordinarily self-ascribe apprehension in ways that, de Sousa supposes, attest to its processual character: “As I took in the scene, I realized that it was scary.” He moreover suspects that if apprehension is conceived as the instantaneous onset of acquaintance, my assumption amounts to a tautology, in which case it does not advance my argument.

I disagree with de Sousa that apprehension is a process. The fact that I may realize that a scene is scary in taking in that scene does not entail that the process of contemplating the scene is part of my realization. Contemplating the scene is what puts me in a position to realize this. As such, it is distinct from my realization. More importantly, apprehension does not possess the characteristic features of processes. For example, it does not seem to be temporally extended in the way processes are: the question “For how long did you apprehend / realize / detect the scariness of the scene?” is infelicitous. ³ Similarly, unlike the humming of my fridge (a process), my apprehension that it is humming cannot persist or be continuous. ⁴ Given the temporal profile of apprehension, I find it much more natural to think of it as an event that marks the onset of acquaintance (see also, Mulligan, 2007). ⁵

Moreover, while I agree that, with apprehension conceived in this instantaneous way, my assumption is tautological, I am not sure why this should render it useless. It is part of our concept of apprehension that what is within our ken can no longer be brought within it. This spells bad news for AR: AR is in tension with a conceptual truth.

de Sousa's proposal that apprehension is a process also informs his broader concern with my argument. As I read him, his central objection is that I fail to consider the option that, rather than being instantaneous, emotional feeling involves a process that in its entirety constitutes a way of apprehending value and, at the same time, accounts for its value responsive character. He suggests that this idea might be fleshed out and assessed by drawing on three supplementary considerations: i) It is plausible to think of emotional feeling, like perceptual experience, as caused by a subpersonal processes. Zajonc (1980, 2000) provides evidence for thinking that no personal-level state of prior apprehension is needed for an emotional response to occur. Despite often seeming instantaneous, emotions are caused by (and, I take it, therefore include) a process that takes time. ii) Psychological appraisal theory might provide resources to further characterize the process of value apprehension. In particular, it can help adjudicate the question of whether emotion is constituted by or a response to the apprehension of value on empirical (rather than phenomenological) grounds. iii) Kriegel (2015, p. 147) provides resources for thinking that feelings that are built up from familiar experiential resources may still appear primitive and irreducible from a purely phenomenological perspective. My claim that temporally distinct episodes must be involved in having emotional feelings ignores this possibility. What I say seems compatible with emotional feeling first-personally presenting itself as a primitive form of apprehension.

Since (as noted above) I doubt that apprehension is a process, I do not think the option put forward by de Sousa can help vindicate AR. But it also does not seem to me to speak to the relevant notion of a value response. This notion qualifies value as a reason for which we feel some way. Thus conceived, value responses require an act or state that makes available a reason for the subject to respond to and which, as such, qualifies as a person-level act or state. This tells against the suggestion, which I take to be part of de Sousa's first supplementary consideration, that the responsive character of emotional feeling can be made intelligible in terms of subpersonal processes. ⁶ Moreover, as I argue elsewhere, pre-emotional evaluation in fact features in pre-reflective experience—and thus at the personal level—even in those cases where the emotion appears instantaneous from the first-person perspective. This prior evaluation can be ascertained on phenomenological grounds by probing into the reasons for the way one feels (for details, see Müller, forthcoming a).

Although I believe that there are interesting parallels between PT and appraisal theory (Müller, 2017), I doubt that the relation between emotion and value apprehension can be determined on largely empirical grounds. There are definite conceptual constraints on this relation which we must heed if our object of inquiry is to bear even a basic resemblance to emotions as we pre-theoretically know them. In particular, these constraints are essential to recognizing emotions as intentional (Müller, 2017, forthcoming a).

I admit that I am not sure I fully understand the dialectical force of Sousa's reference to Kriegel's account of emotional feeling. But it seems to me that this account does not speak to the dialectic at issue. Unlike Kriegel's proposal, my claim that emotional feelings are preceded by a distinct apprehension of value does not aspire to be a reductive account of their nature, but part of a description of the way they present themselves first-personally. What I say in chapter 3 of the book gives reason for thinking that this description is not compatible with emotional feelings presenting themselves as cases of apprehension. Position-taking and apprehension are very much opposite forms of intentionality. ⁷

Now, Schmidt's commentary might be thought to question whether my take on position-taking and apprehension is accurate in this respect. He presents several examples of emotional feelings which purport to show that they can make us aware of value whilst also being position-takings. As Schmidt clarifies, his aim is not necessarily to vindicate AR, however, but to prompt me to clarify whether I wish to deny any epistemological import to emotional feeling at all.

In response, let me first stress that I do not wish to deny this. My concern is with the claim that emotional feeling in and of itself (or, more specifically, qua intentional) is a way of apprehending value. Rejecting this claim is compatible with maintaining that emotional feeling is epistemically important in other respects. ⁸ Indeed, Schmidt's own considerations illustrate what those respects might be. ⁹

As he points out, emotional feelings may direct our attention to features of the situation which bear on our concerns. Moreover, they may indicate the need to inquire about the reasons for which we have them. ¹⁰ In this way, we can come to be aware of exemplifications of particular evaluative properties. Similarly, the feelings involved in recalcitrant emotion may signal conflicts between our evaluative attitudes and thereby help us figure out what we really value. In addition, we sometimes deliberately try to arouse emotional feelings in order to find out what matters to us or reflects our values. As Schmidt puts it, in such cases it is by taking an emotional position that we gain awareness of value.

Looking more closely at these observations, it turns out that none of them imply that emotional feeling as such is a case of value apprehension. The direction of attention in emotional feeling is plausibly a causal effect of it: the way I feel makes me attend to certain features of my situation. In this context, registering evaluative properties is not, as AR would have it, constitutive of the way we feel, but a consequence of it. The same plausibly goes for evaluative insights we gain as a result of emotional feelings prompting inquiry into their reasons or signaling attitudinal conflicts. And when we try to determine how we feel about something order to figure out what we value, we usually interpret the feeling as a sign or evidence of our values. This is not the same as apprehending value simply by way of feeling some way (von Hildebrand, 1969a, p. 78). Accordingly, I do not think Schmidt's considerations give us grounds to think that emotional feelings are in and of themselves both position-takings and apprehensions of value.

A similar reply is available in response to Furtak's objection to my verdict on AR. Furtak insists that AR makes best sense of pre-theoretically central cases of emotional experience. One such case is illustrated by Sartre's Moore-paradoxical remark upon the death of his friend Camus: “He is dead and I don't believe it”. As Furtak suggests, although Sartre is aware of his friend's death, it is only in responding with grief that he registers the loss it constitutes for him. A further case discussed by Furtak involves recalcitrant emotions: A person chronically afraid of spiders knows in some sense that they pose no real danger, but, Furtak suggests, there is another sense in which she does not know this. Intuitively, her awareness of spiders is equivocal: her emotion constitutes a kind of awareness as of danger, which makes us hesitant to say that she knows “perfectly well” that spiders pose no threat.

In line with my response to Schmidt, I doubt that Furtak's examples require any concession to AR. As I argue elsewhere (Müller, 2017, forthcoming b), there are good reasons to think of grief as a way of acknowledging a loss as such. Correspondingly, Sartre's statement can be read as reporting a mismatch between what he knows and what he acknowledges. This interpretation does not conceive of grief as a way of registering loss. One emotionally acknowledges a loss as such by appropriately responding to what one has already registered. Similarly, I conceive of Furtak's example of recalcitrant fear in terms of the phobic's emotional acknowledgment of herself as being in danger despite her knowing she is not. Furtak may be right that the phobic does not know perfectly well that she is safe. Presumably, as Furtak himself suggests, she lacks a sufficiently firm belief in her safety. But we can grant this without thinking of her fear as awareness as of danger. Her lack of firm belief might be due to the fact that there is an impression of threat to which she responds with fear and which she emotionally acknowledges as authoritative in so responding. This account makes good sense of the tension characteristic of such cases: the appearance of danger remains authoritative and is acknowledged as such despite the phobic's awareness of the facts. ¹¹

Emotional Feeling as Position-Taking

Reisenzein, Moors and Furtak question my understanding of the pre-theoretical grounds for PT. They doubt that PT adequately reflects the manifest image of emotion.

Reisenzein objects to my take on ordinary emotional experience. He denies that introspection shows that emotional feelings are in and of themselves world-directed or intentional. Rather than being phenomenologically conspicuous, he claims, intentional emotional feelings are a theoretical postulate. ¹² Reisenzein seems to suppose that all that introspection suggests is that emotional feelings are “connected to” objects. He believes this does not require thinking of them as intrinsically intentional. He proposes that we might instead think of the impression of connectedness as an artefact of the fact that emotional feeling is experienced simultaneously with the focusing of attention on its elicitors. Alternatively, drawing on the resources of his own computational belief-desire theory of emotion, this impression might be explained in terms of an “affect-tinged thought”, which results from a cognitive process in which the feeling is bound together with a doxastic representation of its elicitors.

Reisenzein's verdict regarding the theoretical status of intentional emotional feeling strikes me as too quick. Even if Reisenzein were right that the intentionality of emotional feeling is not a datum of introspection, it would not follow that it is a theoretical postulate. After all, my case for their intentionality is also based on the locutions we use to ascribe them (Müller, 2019, pp. 61–63). This source of support is just as pre-theoretical as first-person observations on emotional experience.

I do not think that Reisenzein is right about what introspection reveals, though. There is a battery of first-person considerations in support of the claim that emotional feelings are experienced not merely as connected to aspects of our circumstances, but as our response to them (e.g., Mitchell, 2021, pp. 100–103; Müller, 2019, pp. 58–60, p. 80). This impression cannot be explained in terms of an experienced simultaneity of feeling and attentional focus. Even if we suppose that a causal link between the feeling and its elicitors is part of the overall experience, we do not end up with the right kind of responsive phenomenology. In particular, there is a normative aspect to experienced emotional responsiveness which this proposal fails to capture: the feeling purports to be called for by some object qua (dis)valuable (Müller, 2019, p. 80, pp. 124ff.). ¹³ This responsive phenomenology also goes missing on Reisenzein's second proposal: I do not see how unifying feeling and thought helps capture the impression that it is at the thought (in light of what I am thinking) that we feel as we do.

Reisenzein might reply that responsiveness in the relevant sense falls short of genuine intentionality. However, as a response to something as (dis)valuable, an emotional feeling clearly displays what intentionality is standardly taken to involve (e.g., Crane, 2001; Searle, 1992): it is directed at that thing under a particular aspect (provided by its formal object). ¹⁴

Moors and Furtak both worry that PT is in tension with the manifest image of emotion insofar as the view conflicts with the common thought that emotions are passive. After all, position-taking is plausibly a kind of activity. Furtak elaborates the idea of emotional passivity by noting that our cares and concerns make us vulnerable in a particular way that leaves us with no control over whether we respond emotionally to things that bear on them. The emotions we feel, and even the underlying cares, are thus not up to us.

I agree with Moors and Furtak that the passivity of emotion is a crucial aspect of their manifest image. However, PT does not deny their passivity. As Furtak acknowledges, I have argued elsewhere that this passivity is a matter of their character as responses and so—seemingly paradoxically—of the very respect in which they are active (Müller, 2021). In emotionally responding to formal objects we give in to their authority as normative reasons to feel a certain way. Being grounded in our concerns, this authority is personally binding so that, when the stakes are sufficiently high, it is not within our discretion whether to feel accordingly. Furtak does not seem entirely satisfied with this account, however. At any rate, he wonders whether I end up with a view somewhat like Solomon’s (1993), who notoriously claims that having an emotion is a matter of choice. Yet my account in fact supports the view that emotions are typically not under our voluntary control: as in the case of belief, we are hostage to the authority of the reasons to which we respond (Moran, 2012; Müller, 2021). This is what being vulnerable qua subjects of particular concerns amounts to. ¹⁵

This account of emotional passivity strikes me as preferable to a rivalling one, which recognizes emotions as active, and attempts to show the impression of passivity to be illusory. Moors opts for a view on those lines in arguing that emotions are essentially tied to our agential capacities and yet appear passive since much of the relevant cognitive processing is not conscious. I believe that the need for this sort of revisionary account becomes much less pressing once we look closely to the phenomena and are careful to distinguish between different readings of the active/passive distinction (Müller, 2021).

According to a further line of criticism, there are problems with the way I elaborate PT. These objections are based on considerations on the plausibility of my understanding of several core aspects of PT, which arise largely independently from the issue of its adequacy to the manifest image of emotion. I address two objections on these lines, raised by Reisenzein and Moors, in the remainder of this section. Three further queries, which revolve around my understanding of pre-emotional evaluation, are addressed in the final section.

Reisenzein's objection focuses on causal considerations. He argues that the way I elaborate PT is deficient since I suspend judgment on whether the notion of a value response is to be understood causally. Reisenzein maintains that in order to understand why an emotion follows upon the apprehension of value, we must think of value responses as involving a causal connection between emotions and the preceding mental act or state which apprehends their formal object. He takes it that, as a corollary of this, exemplifications of formal objects themselves become intelligible as partial causes of emotion.

Since causal analyses of motivating reasons are controversial for a number of reasons (some of which Reisenzein alludes to), I prefer to remain neutral on whether responding to value should be conceived in the way Reisenzein proposes. I do not think this means that I fail to explain the temporal sequence between emotion and evaluation, however. It is part of the very idea of a motivating reason that what is to move me to feel or act in some way must be within my ken. This in and of itself explains why emotions are subsequent to apprehensions of value. To insist that this explanation is unintelligible unless the latter are conceived as causing the former is to disregard the option that the link between motivating reasons and what they motivate might be explanatorily just as fundamental as the relation between causes and their effects (e.g., Bittner, 2001, chap. 5; Dancy, 2000, chap. 8).

Moors objects that I elaborate PT in a way that betrays the very idea it is meant to substantiate. The position-taking aspect of emotional feeling threatens to be lost on my account since, as she supposes, I conceive of the attitude of (dis)approval as reducible to concern-based construals.

Moors is right that if (dis)approving came down to concern-based aspect perception, PT would not live up to its promise: it would be a badly disguised version of AR. However, I do not think (dis)approval can be reduced in this way. Concern-based construals are distinct from attitudes of (dis)approval: the former serve as cognitive precondition for and precede the latter. Perhaps Moors's worry is based on the supposition that I appeal to the idea of (dis)approval in order to cash out the notion of appraisal that is central to emotional appraisal theories. However, although I argue in other places that emotions themselves are a form of evaluation, this evaluation must be sharply distinguished from the evaluation which they presuppose as position-takings and which I identify with concern-based construals (Müller, 2017, forthcoming b).

Pre-Emotional Evaluation and the Nature of Concerns

Reisenzein argues that my view of pre-emotional evaluation is overly demanding. He supposes that my proposed candidate for pre-emotional evaluations (concern-based construals) involves conceptual awareness of concern-(in)-congruence. Since this implies that having an emotion requires the possession of mental concepts (i.e., concepts for concerns, such as desires or needs), it seems to set the cognitive bar too high even for adult humans. Although Reisenzein acknowledges that one might think of concern-based construals as having non-conceptual content, he claims that this proposal can only be made good on by specifying how they are psychologically implemented. The best way to do so, he goes on to argue, is by conceiving of them in accordance with his own computational belief-desire theory, i.e., as non-conceptual output signals of a hardwired belief-desire comparator.

In the book, I actually go at some length to show that concern-based construals do not outrun the cognitive capacities of creatures capable of emotion (Müller, 2019, pp. 127–132). Admittedly, in arguing for this, I do not specify a psychological mechanism of the sort indicated by Reisenzein. However, I do not think that this is required. To meet this charge, we need an account of the capacities exercised in concern-based construals. To suppose that the only satisfactory account of such capacities is mechanistic is to commit to a prejudice on the lines I criticized in the section “Methodology”.

Since what I say in response to the charge of over-demandingness focuses on evaluative concepts, rather than on mental concepts, I would still like to expand somewhat on this response. Reisenzein assumes that apprehending something as (in)congruent with a concern amounts to its representation as (in)congruent with a particular mental state. However, this is not the most plausible reading. More plausibly, “concern” in “congruence with a concern” denotes a concern qua content of a mental state. For example, in the case of my desire to get a certain amount of writing done today, my writing said amount by tonight is congruent with my desire in the sense of being congruent with what I desire. Thus understood, apprehending the congruence between the fact of my having done the writing and this concern does not require apprehending a relation between this fact and a mental state, but between the fact and a specific goal at which this concern is directed. Crucially, it is possible to apprehend this goal without apprehending it as the content of my desire. For example, I may apprehend it as something to be attained or as something of import. This bears on the force of Reisenzein's charge: as conceived here, apprehending something as (in)congruent with a concern does not require representing mental states. ¹⁶

Pushing a somewhat more fundamental, value-theoretical worry, de Sousa puts some pressure on the idea that what I think of as a prior apprehension of value is aptly so conceived to begin with. He argues that formal objects are not evaluative properties, but purely natural properties. Value is not given prior to emotion, but rather conferred on things by emotionally responding to their exemplifying the corresponding formal object. The properties thus conferred are response-dependent (e.g., being fearsome, being admirable), which supervene on the respective formal object. de Sousa further suggest that we experience them as pertaining to the target of emotion because its possession of the corresponding formal object is apprehended as an adequate basis for the emotion.

As I understand de Sousa's objection, it is fueled by the thought that formal objects must not be conceived as a kind of brute axiological given. Instead, we should opt for a more anthropocentric conception of value: “independently of our “concerns”, the world is devoid of both values and value” (de Sousa, 2022, p. 3). Yet this is by and large what I do: I propose that the axiological character of formal objects consists in relations of (in)congruence with our concerns (Müller, 2019, pp. 79–82). Now, there is no doubt a legitimate question as to why we should think of concern (in)congruence as an evaluative property. But, as I will explain in my response to a cognate concern by Mulligan below, this seems a fairly natural view to take.

Although de Sousa's alternative, response-dependent view of value clearly makes for an interesting and noteworthy competitor ¹⁷ , I doubt that it is coherent. On the standard conception, a response-dependent property like being fearsome is the property of meriting or giving reason for the corresponding emotion. Given this conception, it makes little sense to suppose that this property is projected on things by that very emotion. After all, as a deontic property, its point is to guide us on how to feel. This is possible only if things possess and are apprehended as possessing this property prior to our responding with the corresponding emotion (Müller, 2020).

Likewise coming from a value-theoretical angle, Mulligan raises several queries about my specific relational understanding of formal objects. Since I write that the axiological character of formal objects “consists in” the positive or negative relevance of something for a concern (Müller, 2019, p. 43), Mulligan suspects that I am committed to a reductive view of value. On this assumption, he goes on to raise several issues. To begin with, aiming in similar direction as de Sousa, Mulligan wonders why we should think of relevance for a concern as an axiological or evaluative relation (rather than a non-evaluative relation which itself possesses (dis)value)? A further, related query concerns the scope of my account: does the view intend to do justice also (i) to comparative value, (ii) to thin and thick positive and negative value, (iii) to axiological indifference, to (iv) Husserl’s (1988) axiological principle of the Excluded Fourth and to (v) value ranking?

In response, I would first like to clarify that I do not offer a reductive account of value. My use of “consists in”, which Mulligan picks up on, is intended to specify the axiological character of formal objects, not to reduce it. (As an example of this specificatory use, consider the sentence “The core of the thesis consists in a series of arguments to the effect that reductionism is false.”) As I conceive of positive (negative) relevance for a concern, it is a matter of bearing (un)favorably on the satisfaction of this concern (where “concern” signifies the content of a mental state). To return to the above example, my having done a particular amount of writing is conducive to the satisfaction of my corresponding desire insofar as it is congruent with the content of this desire. Thus understood, relevance for a concern is a case of instrumental (dis)value. On my account, formal objects qualify as evaluative properties since what bears (un)favorably on the satisfaction of a concern thereby has (dis)value for me qua subject of this concern. ¹⁸

In my book, I am explicitly concerned only with the formal objects of emotions. Accordingly, I do not offer anything like a comprehensive theory of value. To properly address Mulligan's second query, let me nonetheless make explicit some consequences of my account of formal objects for the further axiological issues he raises. As for (i), my view does account for comparative value. More specifically, it captures relations of comparative instrumental value. On this view, x is instrumentally better for me than y in being more conducive to the satisfaction of my concerns than y. And x is more dangerous for me than y insofar as x bears more unfavorably on (by posing a greater risk of harm to) something I care about than y. Regarding (ii), since there are plausible candidates for final (non-instrumental) thick (e.g., being sublime, being prosaic) and thin (e.g., the (dis)value of (in)equality) (dis)values, my account of formal objects plausibly does not generalize to positive and negative value across the board. But—turning to (iii)—it makes sense of axiological indifference in respect of instrumental (dis-)value: something that bears neither favorably nor unfavorably, i.e., makes no difference to, the satisfaction of a concern is neutral for me qua subject of this concern. Moreover, in the specific evaluative domain at issue, there are only three ways things can be: either they bear on the satisfaction of a concern favorably or they bear on it unfavorably, or they are irrelevant for its satisfaction. Hence, as regards (iv), my account satisfies Husserl's axiological principle of the Excluded Fourth with respect to the relevant evaluative domain. Finally, concerning (v), what I say allows for formal objects to stand in relations of relative importance. Arguably, such a ranking is constrained (in part) by a ranking of the concerns that feature as their relata. For example, while the formal objects of sadness and grief are both forms of loss, they might be ranked relative to each other in terms of the relative depth or quality of one's attachment to what is lost. And if a prominent tradition in value theory is right, then relevance to concerns generally ranks lower than concern-independent value.

Mulligan also raises some queries concerning my understanding of concerns, focusing on their normative properties and origins. Regarding the former, Mulligan wonders whether I can do justice to the plausible view that concerns themselves admit of justification or are, in my terminology, responses. ¹⁹ He worries that I may not be able to account for our awareness of the reasons to which concerns are responsive. As he supposes in light of my remarks on feeling value in chapter 5 of the book, I would not want to conceive of this awareness as a type of acquaintance. He considers a further option, which introduces higher-order concerns (concerns about concerns), but claims that it would lead to a regress.

I agree with Mulligan that, at least in many cases, cares and concerns are plausibly responses and thus admit of justification. Thus, we can clearly desire, take an interest in or value something for certain reasons. Mulligan is also right that I do not want to think of our access to reasons for concerns as a type of acquaintance. That is, at least if we suppose that the reasons in question are evaluative properties and conceive of acquaintance with them as fundamentally different from any type of epistemic contact we enjoy with the non-evaluative world. However, I do not think the further option that Mulligan considers commits me to a problematic regress. To show why, let me briefly elaborate on this alternative.

As I understand the proposal in question, concerns are often higher-order states in the sense of being responses to something's (in)congruence with other concerns, where this (in)congruence is registered in concern-based construals. For example, Maria's long-standing desire for structural reforms in higher education is a response to the present state of academia qua being incongruent with one or more of her concerns (she values equal opportunity and flat hierarchies, say). As in the case of pre-emotional evaluations, she registers this incongruence by construing the present state of academia in terms of those concerns. Now, what about Maria's concerns for equal opportunity and flat hierarchies? Are they responses to concern (in)congruence, too?

Clearly, the view that all concerns are responses to (in)congruence with further concerns leads to a vicious regress. But I do not endorse this view. Since I do not claim that all (dis)value is instrumental (dis)value relative to concerns, my view allows that some concerns are responses to concern-independent value. Moreover, as Mulligan himself recognizes, plausibly some concerns are not responses. One obvious candidate are basic biological concerns, such as, for example, a primitive aversion to physical harm. Moreover, taking inspiration from Rohracher (1971), one might argue that there are ethical and cultural drives, which qua urge do not admit of reasons. In a similar vein, von Hildebrand (1969b) makes a case for fundamental attitudes in the ethical sphere, which are not responses. ²⁰ If there are plausible candidates for non-responsive concerns, these can in principle serve as regress-stoppers. Of course, to fully answer Mulligan's worry more would need saying about these candidates for non-responsive concerns and their relation to the types of concerns that can feature in pre-emotional evaluations. But I think it is safe to say that the idea of non-responsive concerns provides a further avenue worth exploring in responding to Mulligan's worry. ²¹

Considering the origins of concerns, Mulligan also worries that I cannot allow that concerns (in particular, attachments) may come into being through emotion. This is because emotions which mark the beginning of an attachment cannot themselves be responses to (in)congruence with the very attachment to which they give rise. After all, an emotion cannot both give rise to concern and be a response to a property whose instantiation requires that this concern already exists.

In surveying possible responses, Mulligan notes that I might deny that attachments ever originate in emotion. However, there is room for me to allow for this possibility whilst maintaining that emotions are responses to concern (in-)congruence. We can think of the emotional origins of attachments as responses to (in)congruence with other concerns. Consider Mulligan's example of an episode of falling in love with someone, which marks the beginning of an ongoing attachment to this person. This emotional episode cannot be a response to a relation in which this person stands to the very attachment originating in this episode. But we can think of it as motivated by congruence with other concerns. Relevant candidates are those conformity with which underwrites romantic attraction (common candidates include the desire for a specific type of attachment, sexual preferences, likings of specific personality and character traits, shared (dis)values). This proposal sits well with independent considerations on emotional (dis)pleasure. If the affective episodes that mark the beginning of certain concerns are paradigm emotions, they are forms of (dis)pleasure and as such have a phenomenology of felt concern satisfaction. The concerns which are satisfied in these cases must exist prior to the emotion and plausibly provide a standard of congruence in terms of which its intentional object is pre-emotionally evaluated.