1 Introduction

It is a familiar feature of our affective psychology that our moods ‘crystalize’ into emotions, and that our emotions ‘diffuse’ into moods. Anxiety provides a central example: an occurrent emotion of anxiety about an upcoming medical test might diffuse into a generalized anxious mood which concerns ‘everything and nothing’; contrastingly, a generalized anxiety of the latter kind might become more ‘focused’, as expressed in an occurrent emotion of anxiety which targets a particular object, for example an upcoming event.

Providing a detailed philosophical account of these affective shifts, as I will call them, is the central aim of this paper. Drawing on contemporary philosophy of emotion and mood, alongside distinctive ideas about affective shifts from the phenomenologically-inspired writer Robert Musil (which were introduced into the contemporary discussion by Peter Goldie), a broadly ‘intentional’ and ‘evaluativist’ account will be defended.Footnote 1 I argue that we do best to understand important features of these affective shifts in terms of intentional and evaluative aspects of the respective states of moods and emotion. At same the time, the account is pitched at the phenomenological level, as dealing with affective shifts primarily in terms of moods and emotions as experiential states, with respect to which it feels-like-something to be undergoing the relevant experience.Footnote 2 The paper also applies the intentional-evaluative model of affective shifts to anxiety in detail, developing the idea that the way our moods ‘develop’ into emotions can, in certain cases at least, allow for a kind of ‘release from affect’ which is arguably a central feature of psychological well-being.

The structure of the paper is as follows. Section 1 details aspects of affective shifts and outlines constraints on a plausible account thereof. Section 2 explicates and defends the intentional-evaluative account. In doing so it also provides reasons for not accepting an account of affective shifts which characterises them in terms of a shift from a non-intentional mood to an intentionally-directed emotion. Finally, Sect. 3 applies the account in more detail to anxiety (as both a mood and an emotion).

I now speak to the broader implications of the discussion here and connected issues. First, an account of affective shifts promises to shed light on emotions and moods in tandem, and so as standing in important relations to each other. This is welcome within philosophy of emotion, and in affective psychology and science, since more often than not accounts of these states are provided without detailed consideration of how such accounts might explain affective shifts, and more broadly the connections between emotion and mood.Footnote 3 Secondly, the success of the intentional-evaluative account of affective shifts should be taken as a point in favour of broadly Evaluativist and Intentionalist approaches to emotion and mood; put otherwise, the ability of such accounts to make sense of them speaks in their favour. Finally, the reflections provided on the connection between affective shifts and well-being opens up avenues for further research concerning how the proper functioning of our affective psychology is tied to our ability to shift between emotions and moods.

Let me now say something in favour of the phenomenological approach that is taken here–the pitching of the account of affective shifts at the ‘phenomenological level’. First, it bears noting that emotions (and arguably moods) understood as conscious experiences with a phenomenology are an integral part of everyday life and engagement with the world, as is clearly reflected in folk psychological discourse of feelings of fear, jealousy, love, shame, regret, admiration, depression, melancholy, joy, elation (etc.), with such talk of ‘feelings’, being a placeholder for a more detailed account of the relevant affective phenomenology. As such, a phenomenological analysis promises to elucidate something we pre-theoretically take to be critical to an understanding of affectivity, namely its felt or experiential character. Indeed, if–as seems independently plausible–emotions and moods are paradigmatically felt states, then a phenomenological approach seems especially appropriate. Connected to this, such an approach to the affective domain promises to contribute to an understanding of subjects at the personal level, what Daniel Dennett calls ‘the explanatory level of people and their sensations and activities’, rather than in terms of psychological constructs and subpersonal processes.Footnote 4 In a similar vein, Peter Goldie describes the personal point of view as ‘the point of view of a conscious person, capable of thoughts and feelings...’.Footnote 5 Not only does a phenomenological analysis of emotions and moods clearly contribute to such an understanding of subjects at the personal level, but within this context questions concerning the link between a person’s emotions, moods and well-being arise and can be discussed in a way which is sensitive to their ‘lived experience’.

2 Clarifying the phenomena

To get clearer on affective shifts, it is helpful to begin by distinguishing them from similar phenomena. A subject might ‘shift’ to a number of different emotions across a given period. Consider the following vignette:

Oscillating emotions: Waiting patiently at the traffic lights someone bumps my car from behind. I first feel anger and indignation, and exit my vehicle ready to remonstrate with the individual. I then see they are unconscious and as I approach their vehicle, I feel a rising sense of panic. As they appear to be rousing, I feel some relief as it appears they are fine.

We often oscillate between a range of different emotions across a period of time. One might call these ‘affective shifts’ since we are ‘shifting’ between different affective states. Yet this is not the phenomena I am interested in here. The kinds of affective shifts that are my focus, as outlined in the introduction, involve shifts from emotion to mood (or vice versa).

However, even once this restriction is made there are other kinds of affective (emotion-mood) shifts with which our primary phenomena may be confused. There are cases in which sufficiently contrasting affective states can ‘pull us out of’ a given affective state or ‘push us into one’. For example, a piece of good news which precipitates an emotion of happiness, say, can have the effect of ‘pulling us out of’ a bad mood, say a generalised depression. Contrastingly, say we were enjoying a period of joie de vivre, and we are confronted with a piece of bad news which precipitates anger. This bad news can have the effect of ‘pulling us out of’ that good mood. In such cases it is the contrasting nature of the affective state which (in part) leads to this distinctive ‘pushing into/pulling out of’ effect. While there is more to be said about these kinds of affective shifts, they are not my focus here.

The kinds of affective shifts that will be my focus are those in which, put in metaphorical terms for now, a mood crystalises into an emotion and an emotion diffuses into a mood. Here are some examples:

Irascibility ‘crystalising’ into anger/indignation/rage.

Joie de Vivre ‘crystalising’ into admiration/pride/happiness/elation.

Affability ‘crystalising’ into empathy/fellow-feeling/happiness.

Amusement/gratitude ‘diffusing’ into cheerfulness.

Sadness/grief/regret ‘diffusing’ into melancholy.

Fear/terror ‘diffusing’ in anxiety.

Here is vignette which embeds a number of these kind of affective shifts:

Affective Shift: My alarm goes off and I am roused from sleep. I feel somewhat drowsy and hungry, but after those feelings fade away – say I manage to eat some breakfast and have a quick shower – there remains a general feeling of irritation (Mood). Sat at the breakfast table I check my emails only to see yet another administrative error has occurred and I will be required to spend the first part of my day ‘putting out fires’. I feel a rising indignation and anger at the University administrators who seem almost wilfully incompetent at their job, such that their failings create no end of extra work for me (Mood ‘crystalising’ into Emotion). Later that day, I receive news that a long-awaited verdict on a co-authored paper has ‘gone through’. I feel a sense of joy (Emotion) at that achievement, quickly informing my co-author of our success, and as I proceed with a range of others tasks there is diffuse feeling of optimism and positivity, ‘the world seems like a good place again’ (Emotion ‘diffusing’ into Mood).

This description is supposed to be reasonably familiar (give or take the details). Consider any given day and we will have undergone a number of affective shifts in this sense, of mood ‘crystalizing’ into emotion and emotion ‘diffusing’ into mood. Based on this vignette, and the list above, a number of aspects of affective shifts, so understood, can be outlined.

First, note that there isn’t a simple one-to-one correspondence between the emotion-type and mood-type. A mood can ‘crystalize’ into a number of different emotions, as reflected in the way the list represents a number of emotions on the right-hand side of the first three examples. It also seems possible to have a one-to-many correspondence from emotion to mood, for example anger might diffuse in either melancholy or irascibility.

There also seems to be a type-constraint on these affective shifts such that the respective mood and emotion must be ‘in the same ball park’ or ‘family grouping’. One natural way to cash this out at first is in terms of the valences of the respective affective states, that is their being in some sense ‘positive’ or ‘negative, ‘good or ‘bad’, such that there will be continuity of valence across the affective shift. We might put this by saying that in any given affective shift the valence of the respective mood and emotion will be same. This is reflected in the way we would think it odd to talk in terms of an affective shift–remember, specifically of a mood crystalizing into an emotion or an emotion diffusing into a mood–say, from the mood of joy to the emotions of rage or disgust, that is where there is an obvious difference in valence. Consider how the following explanations strike an odd note: ‘Don’t worry about Bill getting angry with you, he’s just in a good mood’; ‘I’m not surprised she’s skipping round the corridors cheerfully, after all this morning she got some terrible news;’ and finally in the first person: ‘I angrily snapped at my partner, and I wasn’t surprised, I’d been in a great mood all day’. Again, one explanation for why these examples strike us as out of keeping with the kinds of affective shifts we are familiar with, is the contrasting valences of the respective affective states. Naturally, it is a further question–and one we will come to in the next section–what more precisely the valenced aspects are which remain in some sense the same across affective shifts (marking out a kind of continuity) and the extent to which they may be reflected in the phenomenal characters of the respective states.

Building on the above though, there seems to be a further aspect to this type-constraint on affective shifts which goes beyond valence continuity. Consider, for example, that anger/indignation/rage seem like more fitting candidates for a way in which a mood of irascibility might ‘crystalize’ than do fear/terror/panic.Footnote 6 It seems like fear/terror/panic share something in common with the mood of ‘anxiety’ that is to be distinguished from whatever it is that anger/indignation/rage share with the mood of irascibility. There looks, therefore, to be a continuity in affective shifts that is not merely a matter of valence continuity (I will provide an account of what this amounts to in the next section). Connected to this, there is a kind of intelligibility in the background which seems importantly tied to this further aspect of the type-constraint, which might be expressed as follows: it makes sense for a subject to be ‘in this emotion’ given they were in ‘that mood’ (and vice versa). Making sense of these links, and the type-constraint on affective shifts adumbrated above, will be central to providing a detailed account of affective shifts.Footnote 7 Note, however, the way in which certain pairings ‘make sense’ will not be taken as merely causal (although of course one might offer an account of certain emotion-mood pairings by appealing to the way in which the relevant states have similar causes), but will be seen to turn on first-person phenomenal aspects of the relevant states.

Let me, however, say something more on this type-constraint and its scope. Consider the following case. It might be claimed that an individual could feel a sense of satisfaction in some sense ‘from’ an emotional episode of anger or sadness. Yet the ‘mood’ in this instance doesn’t obviously (if at all) ‘fit’ into the family grouping which includes either anger or sadness. Such examples problematize the idea that in affective shifts the respective mood and emotion must be in the same ‘family grouping’ and raises the question of whether the type-constraint is a conceptual constraint on affective shifts (delimiting what is to count as an affective shift) or rather an empirical constraint (reflecting something we discover as being the case for the affective shifts we experience). Let me say how one might respond.

On one reading of the relevant ‘sense of satisfaction’ it might be thought to be an occurrent emotion which is about the prior emotional state, such that I have a sense of satisfaction about having been angry or sad. If that is the case though we don’t obviously have an affective shift from a mood to an emotion which would undermine the type-constraint. Secondly, we might contest whether the relevant sense of satisfaction is a genuinely affective state, and might prefer to construe it along the lines of some kind of ‘cognitive feeling’, perhaps associated with thoughts of the kind ‘I needed that [the emotion]’, or ‘I feel better now that I’ve cried it out’. But if the relevant sense of satisfaction isn’t (or at least isn’t obviously) an affective state, then while there might be a shift in such a case it would not be an affective shift.

However, on the broader issue, regarding the scope of the type-constraint, we can say the following. The idea that there seems to be a type-constraint on affective shifts such that the respective mood and emotion must be in the same ‘family grouping’ can be taken to cover paradigm cases of affective shifts (see the list above). As such, while there might be specific affective shifts–say the case of moving from anger to satisfaction if we are convinced that this is a shift from mood to emotion–that don’t require the satisfaction of this constraint (or at least don’t require it in such an obvious way), the paradigm cases of affective shifts do, and they are the cases that a theory or account of affective shifts must (first and foremost) aim to capture. In that sense, the type-constraint can be taken as empirical, at least in the sense that it reflects something that is the case for a wide range of cases of affective shifts.

As a final point on this issue, note that a similar notion to our type-constraint is found in affective psychology in terms of the idea of mood(emotion)-congruence. The operative idea is that certain types of mood will typically lower the threshold for the occurrence of relevant types of emotion. A key example is that irritability or irascibility has been shown to lower the threshold for the occurrence of the relevant ‘offense-related’ emotions, of which the paradigms are anger, rage, and indignation. And furthermore–somewhat supporting the idea that the respective mood and emotion must be in the same ‘family grouping’–the relevant studies found that irritability does not lower the threshold for the occurrence of emotions plausibly classified as belonging to different ‘family groupings’, such as fear and sadness.Footnote 8 In that sense the type-constraint on affective shifts does not merely reflect something which is phenomenologically and philosophically prima facie plausible about these shifts (which it does), but something which we find in the relevant studies in affective psychology. With these reflections in mind, we can now turn to providing an account of affective shifts.

3 An account of affective shifts

3.1 Emotions and moods

Now that we have a better sense of what affective shifts are, and constraints on making sense of them, we can begin to offer a philosophical account. Firstly, it should be clear that to make progress on understanding affective shifts we have to operate with definitions of the respective states of emotion and mood. The least controversial claim we might make in this respect is the following:

Phenomenology claim: Both emotions and moods have a distinctive what-it-is-likeness. There is something it is like to be enjoying one of these affective experiences as occurrent mental states.

In the case of episodic emotions, the phenomenology claim is widely accepted. Emotions are typically occurrent conscious episodes where there is something-it-is-like to be the subject of episodic or occurrent fear, jealousy, love, shame, regret, or admiration, to experience feelings of fear, jealousy love, shame, regret, and admiration. Emotional experiences are, therefore, occurrent conscious episodes with a felt phenomenology, although talk of felt phenomenology is merely a placeholder for a more systematic and detailed account of the what-it-is-likeness of emotional experience.Footnote 9

In the case of moods, however, matters are more controversial. Perhaps if one thought of moods as no more than mere dispositions to enter into occurrent emotions then one might think that they lack phenomenology.Footnote 10 However, a little reflection shows that moods (at least can) have a distinctive phenomenology, that there is something-it-is-like to be currently depressed, joyous, elated, listless, melancholic, anxious, morose, irritable, and calm, such that those moods ‘feel’ a specific way for those who are subject to them. Folk psychology latches onto moods in this sense, as typically occurrent conscious experiential states, when we report of ourselves, and others, that we are ‘in a mood’. Note, I am not ruling out that in virtue of being in a mood–understood as an occurrent experiential state with a distinctive phenomenology–one might be more disposed to certain emotions than one otherwise would be; indeed, some version of this idea is central (as we shall see) to understanding those type-constraints on affective shifts we mentioned in the previous section. Yet, this can be correct without moods being, at least fundamentally or essentially, non-phenomenal dispositional states. As in the case of emotions, recognising the phenomenology claim as (typically) true of moods is merely a starting point, a descriptively adequate account of that phenomenology is a substantive undertaking.Footnote 11

Accepting the phenomenology claim provides a further elucidation of affective shifts as we are interested in them here:

Affective phenomenal shifts: An affective shift involves some change and also some continuity in phenomenology.

Let me explain the above claim with an example. When one’s mood of irritation ‘crystalises’ into anger there will be a phenomenal difference between the state that one is in now and the previous one; irritation as a mood feels different from anger as an emotion. Yet insofar as we are dealing with an affective shift from irritation to anger, in this case (and not merely a disconnected or entirely separate affective state), then there will be a phenomenal similarity (or continuity). Put somewhat metaphorically for now, it will be as if one’s anger ‘preserves’ something of one’s irritation, that the irritation somehow ‘infuses’ the anger. Cashing out these metaphors relating to change and continuity in phenomenology will be one of the central tasks when providing a more detailed account of affective shifts.

In addition to the phenomenology claim we might think that there is a further respect in which emotions and moods share something important in common, and which might be enlisted to understand affective shifts between them, namely their intentionality. Here is a starting point on that score:

Intentionality claim: Both emotions and moods are kinds of intentional experiences. Neither are merely qualitative states but rather enjoy a kind of ‘directedness’ or ‘aboutness’.

In the case of emotions, the intentionality claim is relatively non-contentious. One point of consensus in contemporary philosophy of emotion is that emotional experiences are intentional states, or at least involve intentional states. At its most minimal this amounts to the claim that emotional experiences are directed toward particular objects; so not just concrete or physical particulars, but also persons, animals, events and states of affairs involving these things.Footnote 12 Consider some examples. Walking around an art gallery, I enjoy aesthetic admiration towards a brass sculpture of Julius Cesare. Out on a stroll in the woods, I experience fear when confronted with an Alsatian which bounds towards me. Dining in a restaurant, I am disgusted when the entrée arrives and it is covered in mould. In all these cases, the emotional experience is about something, and so has a target, focus or theme. If in these cases someone were to ask ‘what are you admiring’, ‘what is it that you feel disgust towards’, I would respond by informing them about the particular object of my emotional experience.

The idea that emotional experiences are evaluative phenomena is another point of agreement in contemporary philosophy of emotion, although theorists significantly diverge on the best way of developing the emotion-value connection in relation to the content of emotional experience. On one prominent view, paradigmatic emotional experiences represent their particular objects as possessing the relevant thick evaluative properties. As such their intentional content would be, at least in part, evaluative. Simply put: Emotional experiences have evaluative content.Footnote 13 The idea is that fear represents its particular object as fearsome, amusement as amusing, contempt as contemptible, terror as terrifying, and so on. For the sake of this paper, I will assume something in the vicinity of this view is correct.

Matters are significantly more controversial when it comes to the question of whether moods are intentional states, and indeed how, if they are to be thought of as intentional, to characterize their intentionality. Before offering an account of the intentionality of moods which can help us better understand affective shifts, I want to consider an account of affective shifts which appeals to the idea that while emotional experiences are intentional states, moods are not.

3.2 The intentional-non-intentional view

Consider a view which would have it that moods are non-intentional states, in broad terms they would be ‘undirected’ qualitative feels, and that emotional experiences are intentional states at least in the sense that they are directed towards particular objects.Footnote 14 Drawing on such a view of emotions and moods we get the following view of affective shifts:

Intentional-non-intentional view of affective shifts: In an affective shift we are moving from the intentional to the non-intentional (or vice versa). An ‘undirected feeling’ (a mood) ‘crystallises’ into emotion in such a way as to ‘find an object’, and an ‘object-directed’ emotion ‘diffuses’ into an undirected objectless qualitative state, that is a mood.

However, this view encounters numerous problems. Let me document the most serious.

In the previous section (Sect. 1) we observed that there is a type-constraint on affective shifts, such that there is a kind of ‘family grouping’ between the respective moods and emotions. What could such a view appeal to in order to make sense of this dimension of affective shifts? Perhaps it might be said that valenced phenomenology is that which is shared between the respective mood and the emotion. Such that while the mood, say, of irascibility has no object (it is non-intentional), and while the instance of object-directed anger (or indigitation, rage etc.) which it develops into does, they nonetheless share a non-intentional valenced phenomenology. Perhaps they both ‘feel bad’ in a sufficiently similar way to capture the sense of ‘this mood’ making sense as a ‘development’ of ‘this emotion’ (and vice versa).

The problem, however, is that as we saw in the previous section valence continuity, even if it is expressed by a way of a shared valenced phenomenology, is insufficient to explain the way in which certain mood-emotion pairings seem to ‘make sense’ in way that mood-emotion pairings that nonetheless share the same valence do not (or at least do not to the same extent). Consider that an affective shift from depression (mood) to sadness (emotion), seems to make more sense to us–it seems to be in some sense more ‘fitting’–than an affective shift from depression (mood) to fear (emotion). In this case, all the relevant states share the putative valenced phenomenology, they all ‘feel bad’ in some way, but nonetheless there seems to be something like ‘family groupings’ of affective states where the relevant states seem to ‘fit’ each other in a more obvious way. And it is just not clear what the ‘intentional-non-intentional’ view could appeal to in order to make sense of this.

One potential route might be to posit further shared putatively non-intentional phenomenologies across these mood-emotion ‘family groupings’. For example, one might say the reason depression and sadness occupy the same mood-emotion grouping in a way that depression and fear do not, is because in the former case there is some shared bodily-phenomenology. However, one immediate problem with this kind of view is that it is committed to the necessity of bodily phenomenology for affective experiences. Yet, if, as a growing number of philosophers and psychologists maintain, it is possible to experience a range of ‘cool’ or ‘calm’ emotions (and perhaps also moods), such as admiration, reverence, and regret, in the absence of bodily phenomenology,Footnote 15 then this account of affective shifts will have to be limited to those cases in which there definitely is the relevant bodily phenomenology in play.

Furthermore, note that even if one were to appeal to shared bodily phenomenology it is not clear that a bodily view of moods is a non-intentional view. After all, one might plausibly argue that moods are constituted, at least in part, by an intentional awareness of the relevant somatic states. This view is close to that of Michael Tye, who theorises moods as ‘sensory representations’ of our ‘bodily landscape’, as an extension of the Jamesian theory of emotion to moods.Footnote 16 However, note that this version of an intentional view of affective shifts looks in any case to be phenomenologically implausible, that in cases of affective shift from mood to emotion we move from an intentional state which is about the body to one which is about particular objects in the world (or vice versa). When my mood of ‘irritability’ crystalises in anger or indignation it does not seem descriptively faithful to the experience of such a shift that what happens is a shifting of focus from a way my body is to a way a particular object in the world is (I provide a more plausible intentionalist view in the following section).

However, even if one were to resist this kind of intentional-somatic view of moods–insisting that they enjoy a genuinely non-intentional bodily phenomenology–it is simply not clear that the relevant somatic states across affective states are a point of continuity which could support our type-constraint on affective shifts. Does one’s body really feel sufficiently similar in depression (mood) and sadness (emotion) in a way which is sufficiently distinctive from the way one’s body feels in other negative moods, say irritability or anxiety? If we think not, this is suggestive that our type-constraint is unlikely to be generally satisfied by appeal to specifically shared non-intentional bodily phenomenology across mood and emotion.

Note, as a more general point, it is an open question whether the fine-grained distinctions we make between different types of moods, such as, joy and ecstasy, or melancholy and depression, are based on a personal level experience of a distinct somatic profile for each mood.Footnote 17 This is reflected in first-personal reports of moods, which often highlight the same somatic aspects as consciously present, for example, sluggish or heavy body in depression, boredom, and apathy, and a tense body in irritable mood, anxiety, and nervousness.Footnote 18 Connected to this, one might seek to accommodate the different intuition that there are (fairly) distinct non-intentional bodily phenomenological differences across moods in the following way. Arguably the appeal of such a claim might be the effect of a generalization from pathological cases (which are often the focus of empirical psychological research on moods), which skews the importance of somatic states. For example, anxiety disorders are frequently thought, by both psychologists and clinicians, to be caused by hyper-analytic attention to somatic states.Footnote 19 However, when we look away from such cases towards non-pathological moods, especially those which are ‘cool’ rather than ‘hot’–for example, serenity, boredom, or listlessness–arguably we find a phenomenology that need not essentially involve bodily feelings, and whatever bodily feelings are present are not obviously sufficient to individuate those moods from a range of other moods.

Taking stock, perhaps there are further ways of finessing the ‘intentional-non-intentional’ view of affective shifts, for example one might seek to posit further non-intentional phenomenologies common to the respective affective states that go beyond putative non-intentional valenced or bodily phenomenology.Footnote 20 Note, there is no claim here to have provided decisive reasons to reject the ‘intentional-non-intentional’ view, although as noted it does encounter problems, and so it is legitimate set it aside here to pursue a different (and I hope to show more promising) account. This different account promises to not only deal with our type-constraint in a neater way, but provides a more substantial explication of what it amounts to. It is to that view that I now turn.

3.3 Intentional-evaluative view of affective shifts

If we put to one side the ‘intentional-non-intentional’ view of affective shifts a natural alternative can be expressed in the following claim:

Affective intentional shifts: affective shifts are fundamentally developments which turn on the intentional aspects of the relevant affective states.

It is then a further question what the relevant ‘intentional aspects’ are and how to better understand affective shifts in intentional terms. Before that though, it should be clear that this kind of view is going to require an understanding of moods as intentional states.

Here is the view of moods intentionality I prefer: In moods, the “world as a whole” is presented in a certain way; the intentionality of moods is therefore “global” and “diffuse”. Philosophers and psychologists express similar ideas as follows. In irritability, as Jesse Prinz claims, we are presented with the “general offensiveness of the world”. Folk psychological reports also employ similar terms. For example, in Joel R. Davitz (1969) study, 42% of subjects reported depression as involving a sense that “everything seems useless, absurd, meaningless” and 34% reported anxiety as involving an experience that “everything seems out of proportion.” On the positive side, 66% of subjects reported cheerfulness and contentment as involving a sense that “the world seems basically good and beautiful” and 62% reported serenity as involving “peace with the world.”Footnote 21 On one more precise account, which I have developed in other work, we can frame much of this talk in the following thesis:

Mood‐Intentionality claim: The object of a mood for S is their total environment presented in an evaluative light through a felt valenced intentional attitude towards it (as a felt response with the characteristic of globality).Footnote 22

For the moment I just focus on the ‘object side’ of this view, namely the idea that the intentionality of mood at least partly consists in the subject’s total environment being presented in an evaluative light. I will have something to say on the ‘attitudinal’ components later.Footnote 23

First, this view has it that moods enjoy a kind of ‘exteroceptive intentionality’, as directed toward the ‘world’ in contrast with being (solely or exclusively) about states of the subject. It also tells us how that ‘total environment’ is presented, namely, in a specific evaluative light. As such, the intentional content of moods is, as we are assuming is true for emotions, evaluative (this is a theoretical reflection of the way evaluative terms figure in the above self-reports of moods). Broadly then, moods on this kind of view are intentional affective-evaluative experiences. But what, more precisely, does it mean to talk of a mood being directed towards a subject’s ‘total environment’. According to this account, the sense of one’s total environment in the Mood‐Intentionality thesis does not just concern the perceptual environment of spatiotemporal particulars but also encompasses both present and modal relations, and also one’s current projects, values, aims (one’s evaluative horizons). In this sense, it is not of a particular object, and certainly not a highly determinate state of affairs, with specific determinations of different aspects. Rather, it is one’s total environment, understood as encompassing the broadest set of relations (and potential relations) between self and world, and so as necessarily open‐ended, which is presented in an evaluative light. This view is a plausible theorizing of contexts in which to questions like “what are you anxious about,” a reasonable and unsurprising response might be “everything” and/or “nothing in particular.” Moreover, it can capture familiar modal dimensions to moods, such as the sense in depression of one’s future as hopeless or empty, or in joy of one’s future as full of possibility. It would be in this sense that moods could be described, in Nico Frijda’s words, as ‘experiences of situational meaning with the characteristic of globality’.Footnote 24

Importantly this understanding of the intentionality of moods has the resources to respond to a stock argument given in arguing that they are non-intentional. The argument usually takes the following form:

(P1) For any candidate intentional state we should be able to say what it is an experience about or of

(P2) Moods are not about or of anything in particular

(C) Therefore, moods are not intentional states

This kind of argument is found in John Searle, who writes ‘my nervousness and undirected anxiety need not…be about anything’, and therefore ‘there are forms of nervousness, elation and undirected anxiety that are not intentional’, and is also echoed in discussions of moods as non-intentional states in Julien Deonna, Fabrice Teroni and Eric Lormard.Footnote 25

The unargued for assumption here is that intentionality should be thought of as exclusively directedness towards particular objects (i.e., physical particulars, persons, events or state of affairs), which as we have seen is indeed characteristic of (part of) the intentionality of emotion. Yet, while we can agree that moods may not be about anything in particular it does not follow that they cannot fruitfully be characterised as enjoying a different general form of intentionality. And the Mood-Intentionality claim here is helpful since it provides significant detail on what form that more general kind of intentionality could take.Footnote 26 This explication of the Mood-Intentionality claim will suffice for my purposes here.Footnote 27

Let me now return to our main thread. First, if we accept something in the vicinity of the mood-intentionality claim, then we can begin explicating the idea that affective shifts are developments which turn on the intentional aspects of the relevant affective states. The first thing we can say in this regard concerns the ‘object’ of the respective affective states. Robert Musil, explains this as follows: ‘in every emotion [in every affective experience] there is a distinction to be made between a development toward specificity and a development toward non-specificity….’Footnote 28 Musil’s idea is that in an affective shift from a mood to an emotion we have a development towards specificity, and in an affective shift from emotion to mood we have a development towards non-specificity. Given our analysis so far, we can provide more detail about this idea.

The idea of a ‘development toward specificity and a development toward non-specificity’ can be more concretely specified as follows: (i) a development towards specificity is to be thought of as a development towards a specific intentional object, that is a particular object (i.e., physical particular, person, event, or state of affairs), that is the minimal intentionality of emotion, (ii) a development towards non-specificity is to be thought of as a development towards a ‘general’ intentionality which focuses on the subject’s ‘total environment, that is the intentionality of mood. This kind of view has the benefit that is provides a philosophical articulation of our pre-theoretical description of affective shifts as cases in which a mood crystalises into an emotion and an emotion diffuses into a mood–this talk of ‘crystalizing’ and ‘diffusing’ is explicated in terms of the changing intentional aspects of the relevant states.

We can now provide a refined version of the intentional view:

Affective intentional shifts (refined): affective shifts are fundamentally developments which turn on the intentional aspects of the relevant affective states, one of which concerns the contrast between the way an emotion targets a particular object and the way a mood targets the subject’s ‘total environment’.

Let me, however, note a potential worry with this account. It might be argued that while we can make good sense of the idea of a development towards specificity, such that a mood becomes ‘more focused’ as expressed in an occurrent emotion, that the change in intentionality in the opposite direction, that is a ‘diffusion’ from an emotion to a mood, in terms of a kind of ‘loss of specificity’ is more problematic or at least in need of further clarification.

However, consider the following cases which purport to capture the relevant sense of ‘diffusion’. Say a subject dreams that their partner has been unfaithful. Upon waking they feel an occurrent ‘intense’ anger directed at their partner. Yet, as they ‘come around’ and progress through the day this anger gives way, as it were, to a broader ‘ill feeling’ and ‘annoyance’, which we might label as the mood of irascibility. Alternatively, consider receiving some great news about a job promotion. One is overcome with the emotion of joy, yet as attention to that specific event recedes something nonetheless persists, namely a more ‘diffuse’ sense of optimism and a ‘positive glow’ which seems to concern ‘everything’. The emotion of joy has diffused into the mood of elation or joie de vivre. And finally, consider how an emotion of fear which targets a specific intentional object, say an immediate threat, might, when that threat is no longer present, ‘give way’ to a more generalised ‘anxiety’ which pervades one’s relation to the world around one; one is ‘on the lookout’ for (further) potential threats, and one’s fear has plausibly diffused into a kind of generalised anxiety. Such cases support the idea that a reasonably familiar feature of affective shifts is not just the way a mood ‘crystalizes’ into an emotion, but that the relevant form of intentionality can become more ‘diffuse’ in a shift towards a mood, in which we move from ‘targeting’ a specific intentional object to a form of intentionality which targets the subject’s ‘total environment’.

Now it might be questioned what, more precisely, the psychological mechanisms are by which a mood might ‘gain’ the specificity of a particular intentional object (in the shift from mood to emotion) and indeed ‘lose’ the specificity of a particular intentional object (in the shift from emotion to mood). While providing a detailed account of these will not be possible here, it seems plausible to say that such modulations in the specificity of the relevant form of intentionality are undoubtedly connected to relevant changes in attention.Footnote 29 In the move from occurrent joy about the great news to ‘diffuse’ joie de vivre a certain intentionally-relevant aspect of that shift is the way in which the relevant form of attention characteristic of one’s affective experience itself becomes more diffuse–as my mind ‘wanders’ away from the specifics of the job promotion, my affective experience nonetheless persists in a more ‘diffuse’ form in which a ‘positive glow’ seems to spread over ‘everything I encounter’ (the intentionality of which is characterised more precisely above in the discussion of the mood-intentionality claim). In the opposite direction, a mood of depression which seems at once about ‘everything and nothing’ which then ‘crystalises’ and shifts to an occurrent sadness about an event depicted in the humanitarian aid appeal I am watching, involves a (paradigmatically non-voluntarily or ‘passive’) ‘focusing of attention’ in my affective experience which targets a particular object.

We can further build on this view of affective shifts by drawing in the evaluative dimension of the respective states (which is reflected in some of the above descriptions and cases). Here is the claim which I explicate in what follows:

Evaluative themes across Affective shifts: In shifting from emotion to mood and mood to emotion, the evaluative aspect/property in which the relevant ‘object’ is presented develops along the lines of an ‘evaluative theme’ which is fitting for that affective shift.

The subsequent examples bear out the above point.

Take the following affective shift. A subject is in a mood of irritation and this develops into a specific episode of anger. If what we have said thus far about the intentionality of affective shifts is along the right lines, then in this shift there is development from a sense of the ‘world as irritating’ to a specific particular object (say a particular person) as determinately ‘offensive’ (as ‘offensive’ in a particular way). There is clearly a shared evaluative theme here, say something approximating to ‘wrong-doing’, specifically that which impinges on the subject of the affective experiences. And in the affective shift we are considering, it seems clear that this evaluative theme runs through the development, such that the ‘general offensiveness of the world’–which might be said to characterise our irritability–takes on a more determinate character as the offensiveness of the particular kind instantiated by this person, say, in anger at the way someone spoke to one. Put otherwise, the development in the shift to emotion from mood is not merely the focusing in on a particular object–moving from ‘the world’ to something specific in it–but also a development in terms of the way in which that value is instantiated, which often leads to a more specific determination of it (the specific kind of offensiveness of the rude comment).

Contrastingly, in the ‘development towards non-specificity’–in an affective shift from emotion to mood–the relevant evaluative theme becomes ‘less determinate’. This might be reflected in both the relevant value becoming a less determinate type, say offensiveness in general rather than specifically rudeness, moral wrongdoing, or the personally insulting. But there also will no longer be a ‘specific way’ the evaluative theme is instantiated, so we would not have the rudeness or offensive of the comment, but offensiveness ‘of the world’ in general, seemingly categorizing ‘everything’. Such reflections also further clarify the sense in which–in the move from emotion to mood–there is a loss of specificity, along the lines of the relevant form of evaluative intentionality.

Given the above analysis, the development towards specificity/non-specificity in affective shifts concerns not merely the way the object-intentionality of the respective affective experiences become more or less specific (from ‘world’ to ‘particular object’ and vice versa), but also involves a development towards specificity/non-specificity concerning the relevant ‘evaluative theme’, and as such plausibly concerns the evaluative intentionality of the relevant states.

Evaluative Themes in Affective Shifts (refined): In affective shifts, both the emotion and the mood have the same family-grouping of ‘evaluative theme’, and that evaluative theme becomes more determinate (more specific) in a transition to emotion, less specific in the transition to a mood (the world as irritating becomes that man as offensive; that event as brilliant becomes ‘the world’ as a joyous place).

Recognising the way evaluative themes ‘crystalize’ and ‘diffuse’ in affective shifts provides us with a plausible explanation of that central motivation between our type-constraint (see Sect. 1), namely that there is a kind of ‘fit’ between the mood and the emotion in an affective shift, and one which goes beyond a minimal kind of valence continuity, or indeed any appeal to non-intentional phenomenologies.

Building on the above though, this ‘fit’ is not merely a matter of judging it to be the case that irritability and anger/rage/indignation ‘hang together’ by way of shared evaluative theme, in a way that similar emotions would not with different moods, say ‘melancholy’ or ‘cheerfulness. Rather it is plausible that this ‘fit’ is something which is reflected in the phenomenology of the relevant states, as a kind of phenomenal continuity.

With the above account in hand, we can be more precise about such a claim. Pre-theoretically we might describe this by saying that the subject’s anger preserves something that was felt in the mood of irritability, and likewise in a move in the opposite direction, the subject’s mood of irritability preserves something that was felt in anger. And in a way which clearly goes beyond mere valence continuity, the kind of phenomenal continuity in affective shifts can partly be accounted for in terms of the relevant evaluative phenomenology of the ‘evaluative theme’. Indeed, this is one way of buttressing what Musil has to say about the way different affective experiences ‘persist’ and ‘infuse’ the respective affective experiences in affective shifts:

There is no “mood” that does not also include specific emotions that form and dissolve again; and there is no specific emotion that, at least where it can be said to ‘radiate’, ‘seize’, ‘operate out of itself’, ‘extend itself’, or operate on the world directly’, without an external emotion, does not allow the characteristics of the non-specific emotion to peer through’ (Musil, 1961: p. 985).

Indeed, what ‘peers through’ is the shared evaluative theme.

Let me, however, note an important caveat. A central claim of Musil’s in the above quoted passage seems to be that there is necessarily a corresponding mood for an emotion, and vice versa. However, one might think this problematic: after all, we might question what mood corresponds or ‘peers through’ the emotions of jealousy, embarrassment or disgust. Likewise, is it not the case that moods like nostalgia, or a mood of hopefulness don’t have obviously corresponding ‘specific emotions’. Note, we might appeal to a sufficiently broad ‘evaluative theme’ so as to accommodate such cases; perhaps there is an evaluative theme of ‘the repulsive’ which can fairly be claimed to characterise both occurrent emotional disgust and some kind of mood of disgust. In this context it is worth noting that we should not expect there to always be emotion or mood natural language terms corresponding to the relevant experiences; that we have no distinct term for a mood in which ‘the world’ is presented as somehow repulsive in character is not, in and of itself, a decisive objection to such an experiential possibility. Nevertheless, for my purposes here I don’t commit to there necessarily being–as Musil seems to–a mood corresponding to every emotion and an emotion to every mood. It will be sufficient for the analysis that follows, and to develop the idea of shared evaluative theme which ‘peers through’ in affective shifts, that in a range of paradigm cases there is a corresponding mood or emotion (or at least a ‘family grouping’ to which the relevant mood and emotion both belong).

Returning to the main thread, at the outset of this section it was also said affective shifts will involve a change in phenomenology–so just as something important is felt to remain the same, or persist, something important is felt to change. The intentional-evaluative account provided has the means to make sense of this, insofar as there is a change in the respective intentional aspects, that is both the relative specificity of the intentional object and the development towards specificity/non-specificity of the evaluative theme. Given this, it follows that the phenomenology will also be different in relevant respects.

As a final point, let me note a further aspect of affective shifts, as understood on the kind of intentional-evaluative picture I have developed. Affective experiences, both moods and emotions, are not plausibly understood exclusively in terms of being presented with an ‘object’ (in terms of an object making itself manifest), whether that be a ‘particular object’ or the subject’s ‘total environment’, and indeed whether or not the ‘aspects’ under which those ‘objects’ are presented are manifestly evaluative. Affective experiences also involve ‘responding’ ‘reacting’ or ‘being moved’ by the relevant object and its evaluative aspects. Put in different terms, it is plausible that there is what we can call an attendant attitudinal phenomenology (as what has often been characterised in the literature as a kind of ‘felt aboutness’ or ‘feeling towards’).Footnote 30 One relatively uncontroversial way of capturing this ‘attitudinal phenomenology’ characteristic of affective experiences is in terms of felt attitudes of favour and disfavour.

Think of an episode of anger in which I seem to register the offensive character of a remark; in virtue of what aspect of my experience do I register such offensiveness? Or take an episode of amusement in which I seem to register the humorous quality of a joke; in virtue of what aspect of my experience do I register the joke as funny? Alternatively, take a mood of depression in which I seem to register the ‘hopeleness of the world’, or a mood of elation in which I seem to register that ‘everything seems somehow brilliant and joyous; in virtue of what aspect of my experience do I register such evaluative themes as characterising ‘the world’? There is a variety of cognate, paired, (bi) valenced attitude terms which capture, across a range of cases, this felt aspect and arguably involve affectively registering the relevant evaluative properties–that is being moved by value. Here is a by no means exhaustive list: approval/disapproval, reject/accept, attraction/repulsion, like/dislike, approach/avoid, toward/away. Some of these attitude terms fit the affective response dimension of certain affective experiences better than others. Nonetheless, what is arguably common across cases is an occurrent, phenomenologically salient attitude, as part of affective experience. As such, affective experiences involve positively or negatively charged intentional feelings as felt favourings or disfavourings, as a primitive ‘yes’ or ‘no’ which ‘comments’ on object and its value. These felt valenced attitudes are both intentional and monadic, targeting the relevant object under an evaluative specification, rather than being dyadic preferences expressible by way of comparative attitudes (i.e. favouring x over y).

There is significantly more to be said about this way of cashing out the ‘attitudinal phenomenology’ of affective experience, but let me explain how it figures in the case of affective shifts. Again, Musil provides clarity on this point, writing that ‘a specific attitude toward something corresponds to the specific emotion, and a general attitude toward everything corresponds to the nonspecific emotion: the one draws us into action, while the other merely allows us to participate from behind a colourful window’.Footnote 31 The idea would be that in an affective shift our attitudinal relation to the relevant object (particular object vs ‘the world’) itself become ‘more focused’ or indeed ‘more general/diffuse’ as we shift between affective states. So, it is not merely that ‘things out there’ seem somehow ‘more or less specific’, turning on manifest changes in the relevant intentional aspects, but also our affective-attitudinal relation to ‘things out there’ become more or less specific.

For Musil this difference in attitudinal relation is manifest in the idea of motivation and participation. The kind of ‘general attitude’ which is characteristic of mood, a general felt (dis)favour, which inherits its ‘generality’ from its intentionality, targeting at once ‘everything and nothing’ (to put it metaphorically; less metaphysically the subjects ‘total environment’) will lack the often more concrete motivational profile and action-tendencies of ‘specific emotions’.Footnote 32 For example, a mood of irascibility and its general felt disfavour towards ‘the world’ as an irritating place needn’t manifest any felt action-tendencies to do anything in particular. Contrastingly in the case of ‘specific emotions’, it is more likely that there will be relatively determinate motivations and actions associated with the emotion (although of course there will be exceptions; I can be jealous of my partner and yet that jealousy needn’t carry any relatively specific motivation). In anger I may typically feel ready to definitively strike out against the offending person or object, or in elation about having passed a difficult test feel ready to jump up and down. As we will see this contrast is an important consideration in the following section concerning ‘affective release’.

Let me now take stock. I have presented a detailed intentional-evaluative account of what is involved in affective shifts and how to understand the respective states involved in them. The account is able to provide clarity on a number of important features of affective shifts, such as the type-constraint concerning ‘family groupings’ of mood-emotion pairings, and the sense of phenomenal continuity and difference across affective shifts. In addition, and more broadly, the intentional-evaluative view provides a philosophically plausible way of cashing out our original metaphors of moods and emotions ‘crystalizing’ and ‘diffusing’ in affective experience by appealing to the intentional and evaluative aspects of emotions and moods. In the next section I apply the account in more detail to a specific case and explicate the above-mentioned feature of ‘motivation’ across affective shifts.Footnote 33

4 Shifting anxiety and well-being

In this section I take the account of affective shifts provided and apply it in detail to a specific shift which is arguably one of the most familiar, namely that of moving from a general mood of ‘anxiety’ to an emotion of what I will call targeted-anxiety. I also suggest that a shift from generalized-anxiety to targeted-anxiety can be psychologically beneficial for the subject insofar as it provides a kind of ‘release from affect’ which ties to the notion of psychological well-being.

Let me first note an important caveat. My interest here will not exclusively be on what might be more properly called clinical anxiety (or ‘anxiety disorders’) as what is in medical contexts labelled a ‘mood disorder’.Footnote 34 Although it is difficult to offer definitive criteria to distinguish so-called ‘normal-moods’ from purported pathological cases, there is nonetheless an important distinction. For example, one can be depressed without meeting (at least some of) the diagnostic criteria for clinical depression or be elated without meeting the diagnostic criteria for mania. Likewise, one can be in a mood of generalised-anxiety (as I am calling it), without meeting the diagnostic criteria for clinical anxiety. As such, I want to discuss generalised-anxiety in a folk psychological sense of the term rather than as a definitive ‘mood disorder’, although I do comment on how when a mood of anxiety fails to shift to the emotion of anxiety, we begin to see more ‘pathological features’, and important contrasts qua ‘release from affect’ and psychological well-being, when compared with non-pathological cases. Likewise, my interest will not specifically be in the emotion of targeted-anxiety of a sort that might fall on some more obvious ‘disorder’ spectrum. Say a compulsive anxiety about public speaking which is crippling and psychologically destructive.Footnote 35

If we apply the intentional-evaluative account of the previous section to this specific case, we get the following account. A subject is experiencing a mood of generalised-anxiety. This is both an intentional and phenomenological state. More specifically, the subject’s ‘total environment’ is presented, by way of general attitude of disfavour, under an ‘evaluative theme’–‘the world’ seems unfriendly, hostile, uncertain and oppressive. Indeed, in Davitz study subjects reported anxiety as involving an experience characterized by ‘everything seeming out of proportion’. Our subject then undergoes an affective shift toward the emotion of targeted-anxiety. Say there is an upcoming social engagement which becomes the ‘focus’ of their targeted-anxiety. As such they move from that generalised-anxiety to this specific emotion which targets a particular object. As they do so, that evaluative theme also takes on a more determinate character. The social engagement is experienced as ‘hostile’ and ‘oppressive’ and ‘uncertain’ in a specific way, say due to the sheer amount of unknown people there.Footnote 36

In this case, there is both phenomenal continuity and change. Due to the overarching evaluative theme–say ‘the oppressively uncertain’–which serves as a common thread through the affective shift, there will be a shared evaluative phenomenology. Nonetheless through the ‘crystalizing’ into targeted-anxiety there is also a central phenomenal change, which at least in significant part, turns on that development towards specificity, that is both the way the intentional target has become more focused as a particular object (in this case an event), and the evaluative aspect has become more determinate, a specific ‘mode of presentation’ of the uncertain or hostile. This is a phenomenologically faithful, and at the same time theoretically informed, description of this affective shift. That the intentional-evaluative account is able to neatly capture this familiar affective shift is a point in its favour.

Developing the account, I now want to explore a further important aspect. One interesting dimension, touched on at the end of the previous section, is the difference in motivational profile. In the case under consideration, when our subject is merely experiencing generalised-anxiety they may not form any specific action plans or be particularly motivated to act in line with their mood (as we might put it). Although their ‘interactions with the world’ might be ‘coloured’ by their generalised-anxiety, ‘participating in the world from behind a coloured window’ as Musil puts it. Charlie Kurth provides the following gloss on the motivational aspect of generalised-anxiety: ‘this general unease will not–at least not in the first place–lead to specific risk-minimization and epistemic behaviours of the sort characteristic of anxiety the emotion. Rather, someone in an anxious mood will display certain biases. For instance, she will be more likely to see features of her situation as potentially threatening and so will be more likely to feel anxious (the emotion)’.Footnote 37

But once our subject shifts to targeted-anxiety there is a much more specific motivational profile, and connected action-tendencies come in play. Here is how Musil expresses this idea:

whenever an emotion develops toward specificity, it focuses itself, so to speak, it constricts its purposiveness, and it finally ends up both internally and externally in something of a blind alley; it leads to an action or a resolve, and even if it should not cease to exist in one or the other, it continues on, as changed as water leaving a mill. (1961: p. 982)

Applied to our case we can say the following. Targeted-anxiety will typically ‘lead to an action or resolve’. Our subject, given their affective experience, may be motivated to take steps to avoid attending the social gathering, or make plans such as to stay incognito when present. Or more specifically, they may ask a friend for details about the event with the hope of quelling worries that she may know no one at the party. These are what we might, following Kurth above, call ‘risk minimizations’ strategies or behaviours; the aim of the subject is to in some sense ‘resolve’ or otherwise negate the instance of ‘uncertain hostility’ which the particular object presents. As such targeted-anxiety might be characterised as involving what Kurth describes as a ‘motivational tendency to be cautious toward the uncertain threat or challenge one faces’.Footnote 38

Importantly this aspect of affective shifts, such that the motivational profile in the case of shifting from mood to emotion becomes more ‘constricted in purpose’, can serve a function for its subject, which connects to the final ideas I touch on here, namely ‘release from affect’ and its connection to psychological well-being. The affective shift we are considering allows for what we can identify as a kind ‘release from affect’: By developing along the lines of specificity, and acquiring a (possible) expression in a more distinct and concrete set of actions and behaviours, our subject is potentially provided with a way of at least in some sense ‘resolving’ their anxiety.Footnote 39 In what follows I explicate this idea as an extension of the intentional-evaluative account of affective shifts.

Let me, however, be clear about the scope of the claim that will be explicated in what follows. The idea that in an anxiety-case of a development towards specificity (from mood to emotion) we get a kind of ‘affective release’ is intended to capture something that can happen as a result of this affective shift, not something that necessarily happens. In that sense, it remains a live possibility that, for whatever reason, the relevant ‘harmful’ mood might resurface or, to put it another way, that despite a concrete development towards specificity that in some sense the mood might persist or ‘linger on’, such that the ‘release from affect’ is either fleeting or that for all intents and purposes there is no real such release. I discuss such possibilities at the end of this section, and provide some reasons why in certain cases we might get a ‘release from affect’ and not so in other cases. Nonetheless, since this kind of ‘affective release’ is importantly connected to some instances of a development towards specificity in the case of the move from an anxious mood to the emotion of targeted-anxiety, and connects to relevant issues of psychological well-being, it is worth explicating in detail.

First it bears emphasising, as is well-accepted, that emotional experiences are typically short-lived episodes which come to a conclusion or end. Sometimes this happens through action or inaction, and sometimes just by running their course (and no doubt by other means as well). When emotions come to an end this can provide a kind of closure or conclusion of that broader affective experience (including mood and emotion), and with it the evaluative theme that characterises it. In one sense, and drawing on our example, if one chooses, as motivated by one’s targeted-anxiety, to refuse to attend the social gathering, that evaluative theme which characterized one’s mood and emotion, namely ‘the oppressively uncertain’, ceases to be present to one. As I put it above, the subject is able–for a short while at least–to have a certain kind of ‘release from affect’.

Note, so framed the claim is not that the mere development towards specificity in such a case somehow ‘soothes’ one’s generalized-anxiety, indeed it seems plausible that one’s ‘anxious feelings’ might, and certainly at first, intensify in the development towards specificity. Rather the relevant claim is that with a development towards specificity we get a ‘constriction in purpose’ and a more concrete motivational profile which is part and parcel of an emotional episode that, by some means or others, comes to an end. In this sense, the ‘release from affect’ is to be thought of as a possible consequence or entailment of what is involved in a development towards specificity rather than turning merely on such a development. In addition to this, it should be noted that I am not claiming that such a release from affect is a ‘once and for all matter’. Such releases may in certain cases be relatively short-lived: an apt analogy here is thinking of the release from affect as akin to a from building pressure, and the ‘emotional pressure’ (in our case) may start to build again fairly shortly after the release (see the discussion below for more on such cases).Footnote 40

Having clarified the basic phenomena, let me now say more about the benefits of the ‘release from affect’ that can occur when generalised-anxiety ‘crystallises’ into the emotion of anxiety. First, as is well-documented, and as was noted in Sect. 1, typical affective states have a valenced phenomenology, which arguably includes a kind of ‘feels good’ or ‘feels bad’ component. Regardless of precisely how we cash this out, if we accept that generalised-anxiety and targeted-anxiety both (broadly) ‘feel bad’, then insofar as our anxious subject experiences a ‘release from anxious affect’, then their psychological state is no longer pervaded by some such ‘feels bad’ component. More concretely, this might be a case of no longer undergoing distinctly unpleasant bodily feelings associated with anxiety (e.g., tight chest, increased respiration, palpitations), but also might involve ‘the world no longer seeming like a hostile place’. As such we might say there are obvious first-person phenomenological benefits to a release from affect. Connected to this, it is worth noting that it is plausible that psychological well-being in the sense of ‘one’s life going well’, is arguably incompatible with persistently ‘feeling bad’ or being in a persistent ‘state of suffering’ (as is attested by subjects of chronic pain). As such the kinds of phenomenological benefits we get with a release from anxious affect clearly tie into the overall psychological well-being of the subject.

Secondly, consider that both the mood of generalised-anxiety and targeted-anxiety (the emotion) have distinctive effects on attention and arguably drain the attentional resources of their subjects. The attentional profile of anxiety is broadly that of being ‘hyper-vigilant’–one is constantly on the lookout for potential harms or threats.Footnote 41 This kind of vigilance, or even ‘hyper-vigilance’ in some cases, can be a drain on the psychological resources of the subject; put simply, the anxious subject will likely struggle to attend to features of their environment not directly related to potential harms or threats. While this form of attention might serve an important function in certain contexts–think of the importance of the vigilance characteristic of targeted-anxiety when waking down a dark alley–a pervasive or relatively ‘unending’ kind of ‘threat-vigilance’ has the potential to radically undermine the subject’s ability to attend to other relevant, potentially ‘positive’ aspects of their environment or situation. Insofar as a ‘release from anxious affect’ can include the cessation of such ‘threat-vigilance’ then at least for a time we might expect the subject’s cognitive and affective psychology to be able to ‘focus on’ more positive or non-threatening matters, as attentional resources are freed up. As such there are clearly attentional benefits connected to a kind of release from anxious affect. More concretely, and connected to psychological well-being, certain forms of self-care activities or non-anxiety provoking social activities, which are essential to the psychological well-being of subjects, might only seem ‘live possibilities’ to a subject who is not currently anxious, and so in the case of a ‘release from anxious affect’ relevant positive aspects of what we might call the ‘evaluative landscape’ become open to the subject.

By way of contrast, consider cases in which moods fail to develop towards ‘specificity’ in this way and so in which we get no release from affect. To use a folk psychological description, we often talk of bad moods which ‘fester’ or ‘linger on’. At least one dimension of this ‘festering’ of our moods is that they, for whatever reason, fail to develop towards specificity–that is they fail to undergo an affective shift towards emotion in which they take a particular object and the evaluative theme also becomes more determinate.Footnote 42 Indeed, in the case of generalised-anxiety (and perhaps other broadly ‘negative’ or ‘bad’ moods) this seems especially true. Think, for example, of a long-lasting anxious mood that seems to go on for an entire day. As attested to by people who suffer from something more akin to clinical anxiety, which tends to have this ‘persistent’ or long-lasting character, the effects can be psychologically and physically devastating (in some case leading to widespread motivational inertia) and undermine their more general psychological well-being and functioning. And to take one further example, a persistent generalised-anxiety may lead its subject to switch from the mood’s ‘focus on the world’ to focus on the affective state itself. For example, anxiety disorders are frequently thought, by both psychologists and clinicians, to be caused by hyper‐analytic attention to the states themselves, especially their somatic aspects (e.g., changed patterns of respiration), which in certain cases leads to anxiety or panic attacks.Footnote 43

Given the discussion so far it should be clear that at least in the case of generalised-anxiety, a failure to develop towards specificity, and so to affectively shift in this specific way, can be deleterious to the psychological well-being of the subject. Put simply: they ‘miss out’ on the opportunity affective shifts towards emotion, and that motivational ‘constriction in purpose’, can have for providing a kind of ‘release from affect’.

Of course, even in those cases where subjects find ‘affective release’ through a development towards a specific emotional episode which has a certain duration and finality to it, connected to a more specific ‘motivational profile’, it always remains a possibility that the generalised-anxiety will resurface–indeed this is surely typical for those who are ‘persistently anxious’.Footnote 44 And in that sense the release from anxious affect will be relatively fleeting, such that we might think it unlikely that the subject will enjoy any of those benefits (detailed above) of a more genuine or ‘long-lasting’ release from affect.

Consider, for example, the subject whose ‘anxiety’ alternates between that more generalised-anxiety we have discussing and instances of targeted-anxiety. While such a subject certainly undergoes a number of affective shifts from ‘mood to emotion’ (and back again), and in the mood-emotion direction ‘moves towards specificity’, nonetheless we might say that there is a broader persistence of their ‘anxiety’–they enjoy no real release from affect by moving towards specificity. Arguably what we have in such a case is a subject who could be fairly described as having trait-anxiety, the relevant anxious affects (moods and emotions) are a pervasive and recurrent feature of their affective psychology.

In fact, consideration of these kinds of cases bolsters the analysis and reflections of this section. Let me explain how. We often distinguish between subjects who suffer from anxiety and related affective states and subjects who while occasionally experiencing the relevant generalised-anxiety and targeted-anxiety (and indeed shifts between them), are not usually said to suffer from anxiety (or at least not in the same way). Arguably, a key part of this idea of suffering from anxiety can be connected to the way the subjects do not, or at least not for any significant duration, attain any genuine release from anxious affect. Indeed, it is surely for this reason that for the subject whose affective psychology involves occasional ‘anxiety’, no therapeutic or medical intervention is required–the affect runs its course, often by way of the kinds of developments towards specificity and ‘release from affect’ we have been discussing (affording the benefits detailed previously).

Contrastingly, in those cases where the subject simply is unable to secure such release from anxious affect the psychological disbenifits are often adjudged to be sufficiently detrimental to their psychological well-being that some form of intervention is required, whether that be in the form of ‘anxiety-supressing’ medication (essentially achieving release from anxious affect in a way which bypasses the self-regulating functions of the subject’s own affective psychology) or forms of therapy such as CBT or ‘talking therapy’ which may include learning to ‘control’ or ‘live with’ anxious affect given that no genuine form of release from it is possible or more than fleeting. In this sense, we can see the potential importance of ‘release from affect’ by consideration not just of cases where it is secured but in a range of cases in which it fails to be secured, or has to be secured in a way which bypasses the psychological mechanisms of a ‘development towards specificity’ and a ‘constriction in motivational purpose’ detailed here.

Naturally, there is more work to be done to further clarify the themes discussed in the final section, however I hope to have demonstrated how the account of affective shifts provided in the previous sections both applies to a concrete case of anxiety, and ties into important notions of psychological well-being and ‘release from affect’.

5 Conclusion

This paper has offered an intentional-evaluative view of affective shifts. In doing so it has elucidated the way in which mood and emotion are importantly connected in our affective psychology and has suggested that an account which appeals to intentional and evaluative aspects of the relative states is best placed to make sense of much of what seems intuitive about affective shifts. The account was also couched primarily at the phenomenological level, as elucidating these affective shifts qua the experiences of mood and emotion. Finally, remarks were offered concerning the more specific case of anxiety, and how a development towards specificity, especially with respect to motivational profile–moving from generalised-anxiety to targeted-anxiety–is often (although by no means always) connected to a kind of ‘release from affect’ that is importantly tied to the psychological well-being of the subject.