2. Multiple Aspects of Simple Passions

As I have indicated, my account of Humean self-knowledge appeals to various qualitative aspects of the passions identifiable by introspection. But this appeal may appear vitiated by Hume's claim that the passions are simple. He writes that simple impressions ‘admit of no distinction nor separation’ (T 1.1.1.2) and that ‘whatever objects are different are distinguishable, and whatever objects are distinguishable are separable by thought and imagination’ (the ‘Separability Principle’) (T 1.1.7.3). If we can distinguish between qualitative aspects of passions, then those aspects appear to be separable, making passions complex.

Yet Hume says that we cannot define pride and humility ‘or indeed any of the passions’ because they are ‘simple and uniform impressions’. The extension of indefinability to ‘any of the passions’ indicates that this is a problem for all Humean passions, at least when they are experienced alone.Footnote ⁴ (As we will see, Hume describes compound experiences that constitute the whole ‘character’ of some passions.) Space precludes a full defense of this claim, but Hume construes even ‘mixed’ passions, such as hope and fear, as simple impressions. Mixed passions should not be confused with ‘complex’ impressions, which ‘may be distinguish'd into parts’ (T 1.1.1.2). As far as I see, Hume never refers to passions as ‘complex’. He instead uses the language of ‘compounds’ or ‘mixtures’.Footnote ⁵ Hume compares impressions, ‘especially reflective ones, to colours, tastes, smells’.Footnote ⁶ Whereas ideas ‘never admit of a total union, and are capable of forming a compound by their conjunction, not by their mixture’, impressions are capable of both. The mixture creates a new impression whose elements are not experienced as distinct. ‘Like colours’, the passions ‘may be blended so perfectly together, that each of them may lose itself, and contribute only to vary that uniform impression, which arises from the whole’ (T 2.2.6.1). Hope is not like an apple, whose color we can distinguish from its smell, but more like its particular color. The color of a uniformly orange apple may be a mixture of red and yellow, but we do not see it as composed of red and yellow parts.

Given this simplicity, how can passions have multiple qualitative aspects? I will not respond to this question at length because other commentators have already done so well. I will, however, outline two different, but related, treatments of Humean simple impressions that explain how those impressions can have multiple aspects, from Daniel A. Schmicking and Hsueh Qu. Each takes its cue from Hume's discussion of distinctions of reasons at T 1.1.7.17–18, with the famous example of a white marble globe. He says that we can view the ‘perfectly inseparable’ qualities of color and form ‘in different aspects, according to the resemblances, of which they are susceptible’.

Schmicking reads these ‘aspects’ as Husserlian ‘abstract moments’ (Schmicking, Reference Schmicking2004). Abstract moments are not temporal moments but reciprocally dependent parts of some whole, where these parts cannot exist apart from that whole. Although these moments cannot be sensed or imagined as distinct impressions, we can comprehend them by a distinction of reason. Hume's separability principle applies only to wholes with independent or proper parts, not to abstract moments or the ‘different aspects’ mentioned at T 1.1.7.18. After considering various sensory impressions, Schmicking argues that Humean passions can have abstract moments as well. This reading provides a theoretical apparatus for understanding passions as simple – viz., not separable into parts that could be imagined existing on their own – without denying that we can identify various aspects of those passions.

Unlike Schmicking, Qu frames his account as responding to a problem specific to Hume's theory of the passions. Qu is concerned with the ‘duality’ of Humean passions − their having both qualitative character and intentionality or directedness. Pride, for example, is both a pleasant feeling and has an object (the self). These two aspects, Qu argues, are distinguishable and separable as abstract ideas but not distinguishable in any particular instance of a passion. The intentionality of pride is not representational but consists in a directedness that is actually ‘constitutively determined’ by the qualitative character of the passion – an ‘inseparable aspect of its feeling’ (Qu, Reference Qu2012, p. 110).

To explain how simple passions can possess dual aspects, Qu draws on Don Garrett's explication of distinctions of reason. Garrett explains Humean abstract ideas as particular ideas that tend to bring to mind resembling ideas (the ‘revival set’) (Garrett, Reference Don Garrett1997, pp. 62–64). This tendency allows them to signify generally. We can distinguish the set of white things from the set of spherical things. But we can make no such distinction in a particular instance of sensing or imagining a white sphere, because we cannot imagine whiteness without shape or shape without color. Applying this analysis to the passions, Qu concludes that ‘the abstract ideas of intrinsic intentionality and qualitative character are both distinguishable and separable, since these sets are not coextensive’, but ‘we can neither distinguish nor separate the intrinsic intentionality of a particular passion from its qualitative character’ (Qu, Reference Qu2012, p. 110). Again, intentionality is an inseparable aspect of how a particular passion feels.

Both Schmicking and Qu provide ways of articulating the distinction of reason between aspects of simple passions, and the outlines of their views may be compatible with one another. We have at least two plausible accounts of how simple impressions can have multiple aspects. What we do not yet have, I believe, is a thorough exploration of the aspects that Hume suggests contribute to passions’ qualitative feel. The following sections begin that exploration.Footnote ⁷

3. Pleasure/Pain

In searching for different qualities of simple impressions, we are looking for ways that those impressions resemble one another. Hume says that we recognize distinctions of reason because we see that even in simple impressions ‘there might be contain'd many different resemblances and relations’. We notice that a white sphere might resemble a square's color but a ball's shape. Thus ‘we find two separate resemblances, in what formerly seem'd, and really is, perfectly inseparable’ (T 1.1.7.18). What resemblances does he mention among the passions?

Most salient is pleasantness or painfulness. Sometimes Hume seems even to identify them with the full phenomenological experience of the passions.Footnote ⁸ The pleasing and displeasing qualities of pride and humility are central to his theory that a double relation of impressions and ideas produces these passions. Pride arises when we associate the idea of some attribute with the idea of ourselves – e.g., the idea of some virtue of our character, beauty we possess, riches at our command. These ideas would not produce pride without a corresponding relation of impressions. This relation is resemblance between the pleasure arising from the attribute and the passion of pride itself, which is also pleasing. A contrary resemblance produces humility: ‘The sensation of humility is uneasy, as that of pride is agreeable’, so we feel humility when qualities associated with us displease – vice instead of virtue, deformity instead of beauty, poverty instead of riches (T 2.1.5.9). Likewise, despite complications to account for the displeasure that others feel when they realize that we think well of ourselves, Hume insists that pride is a virtue in part because ‘’tis always agreeable to ourselves’ (T 3.3.2.9). Humility, always disagreeable, is likewise a vice.

Pleasing passions resemble one another as well as other pleasant sensations. Thus, in explaining ‘the amorous passion’, Hume notes that the ‘appetite of generation, when confined to a certain degree, is evidently of the pleasant kind, and has a strong connexion with all the agreeable emotions. Joy, mirth, vanity, and kindness are all incentives to this desire’. When we feel one pleasant emotion, others follow because they resemble the first. Unpleasant emotions have the contrary effect; ‘sorrow, melancholy, poverty, [and] humility are destructive’ of the same appetite (T 2.2.11.2).

Hume appears to think that all passions are either pleasant or painful. But does pleasantness/painfulness exhaust their qualitative character? After explaining that pride and humility have self as object ‘by an original and natural instinct’, he identifies a ‘second quality’ of these passions – ‘their sensation, or the peculiar emotions they excite in the soul, and which constitute their very being and essence. Thus pride is a pleasant sensation, and humility a painful; and upon the removal of the pleasure and pain, there is in reality no pride nor humility’ (T 2.1.5.4). Likewise, he seems to identify ‘momentary pain or pleasure’ with ‘the present sensation’ of a passion (T 2.2.9.2). To feel a passion seems to be to feel some kind of pleasure or pain, or perhaps a confusing mixture of the two.Footnote ⁹

Despite the centrality of pleasure and pain to the qualitative experience of a passion, there are strong reasons for acknowledging other aspects of that experience. If there were no other aspects, the difference between the feel of agreeable passions would be a brute, ineffable fact. The only alternative to this unsatisfying conclusion would be the suggestion that all agreeable (or disagreeable) passions feel the same. There would be no phenomenological difference between pride and love. But Hume knows that these two sensations feel quite different from one another. The ‘sensation’ of prideFootnote ¹⁰ does not have ‘any thing in common with that tender emotion, which is excited by a friend or mistress’ (T 2.2.1.2). To say it has nothing in common must be an overstatement, since Hume identifies both emotions as pleasant. Nonetheless, they feel different from one another. There are other aspects of their qualitative character. When he says that removal of pleasure and pain constitutes a removal of the passions, he therefore means that pleasure is an essential but not the sole attribute of pride.Footnote ¹¹

4. Calmness/Violence

Hume highlights another aspect of the qualitative feel of passions in his introduction to Treatise 2 – their calmness or violence. But this introduction has proved misleading. It appears to be a taxonomic divide akin to his distinction between direct and indirect passions; indeed, he begins his ‘Division of the subject’ by situating the passions in his taxonomy of perceptions from Treatise 1. Passions are not thoughts (ideas) but feelings (impressions), and they are not sensations or bodily pains/pleasures (original impressions) but secondary or reflective impressions. He then adds:

Is Hume following, as James Fieser claims, the long tradition of distinguishing between a class of calm affections – variously understood as arising from adequate ideas, instincts, or rational apprehension – and a class of violent passions, arising from inadequate ideas, bodily disturbances, or confused sensations? (Reference Fieser1992, p. 6). On the contrary, Hume actually subverts this tradition in important ways, as I have argued elsewhere (Watkins, Reference Watkins, Ruys, Champion and Essary2019b). These traditional distinctions come with a general approval of calm affections and disapproval of violent passions. Although Hume notes that the virtue of strength of mind ‘implies the prevalence of the calm passions above the violent’, he does not identify calm passions with virtuous ones or violent passions with vicious ones (T 2.3.3.10).Footnote ¹² The contrast is striking between his presentation of this distinction and that of his own contribution to the taxonomy of the passions – the indirect/direct distinction.Footnote ¹³ The former is full of conditionals and passive voice: the impressions ‘may be divided’ and ‘have been commonly distinguish'd’; hence this ‘vulgar and specious division’, ‘far from exact’, for the sake of order. The latter distinction between direct and indirect ‘occurs’ ‘when we take a survey of the passions’, and it structures the intricate discussion that follows (T 2.1.1.4). Despite the reference to order, the calm/violent distinction does not structure Book 2 at all. The direct/indirect distinction captures, he believes, an important truth about the differences between passion-types. The calm/violent distinction, on the other hand, bears no such weight. As Hume tells us immediately after introducing it, any type of passion can be calm or violent. We can experience calm anger and violent appreciation of beauty.

As the last possibility indicates, this distinction has nothing to do with violence as aggression. Nor does it indicate a passion's efficacy in motivating action. ‘’Tis evident’, Hume writes, ‘passions influence not the will in proportion to their violence, or the disorder they occasion in the temper; but on the contrary, that when a passion has once become a settled principle of action, and is the predominant inclination of the soul, it commonly produces no longer any sensible agitation’ (T 2.3.4.1). Hume calls the efficacy with which a passion motivates ‘strength’ (opposed to weakness). Violence, on the other hand, is the feeling of a passion as disturbing, agitating, or perhaps exciting or arousing.

Rather than being a distinction between kinds of passions, calmness/violence is another way that passions can resemble one another (or fail to do so). Evidence that Hume accepts this view comes from his discussion of some puzzles about love and hatred. He reminds the reader that, unlike ideas, passions ‘are connected only by their resemblance, and . . . where any two passions place the mind in the same or in similar dispositions, it very naturally passes from the one to the other’. Dissimilar passions have the contrary effect. Among the dissimilarities that can produce this ‘repugnance’ is a difference in level of calmness/violence: ‘A man, when calm or only moderately agitated, is so different, in every respect, from himself, when disturb'd with a violent passion, that no two persons can be more unlike; nor is it easy to pass from the one extreme to the other, without a considerable interval betwixt them’ (T 2.2.2.22).

If ‘disturb’ or ‘agitate’ suggest unpleasantness, they are misleading. Hume describes many pleasing passions as violent (e.g., love, joy, and pride). In fact, he believes that excitement (to a point) is in itself pleasant. In explaining how custom strengthens passions while also making them calmer, he observes that the difficulty of trying to do or understand anything new creates arousal. ‘As this difficulty excites the spirits, ’tis the source of wonder, surprize, and of all the emotions, which arise from novelty; and is in itself very agreeable, like every thing, which enlivens the mind to a moderate degree’.Footnote ¹⁴ But this enlivening is not identical with pleasure, as the reference to moderation suggests. Calmness/violence cannot be reduced to the pleasure/painfulness of a passion. Beyond a certain degree, arousal becomes painful. Moreover, any agitation can become unpleasant if it accompanies an unpleasant passion. ‘Tho’ surprize be agreeable in itself, yet as it puts the spirits in agitation, it not only augments our agreeable affections, but also our painful’ (T 2.3.5.2). The idea here is not that the concomitance of surprise – a typically pleasant passion – with distinct, painful passions produces an overall feeling of pain. Instead, the agitation of surprise gets transferred to the painful passion. Although this interpretation is controversial,Footnote ¹⁵ I take this to be what Hume means by his ‘conversion principle’, summarized in this same paragraph: ‘every emotion, which precedes or attends a passion, is easily converted into it’. The emotion that is converted is of a subordinate passion; its violence is transferred to the ‘predominant’ passion (T 2.3.4.2).

Is a passion's violence then a separable impression, which can be detached and transferred from one passion to another? I do not think so. Imagine you are feeling pride from an accomplishment, and as you bask in that agreeable situation, meet your lover for a romantic evening. Let us stipulate that your love is greater than your pride at this moment; if so, Hume would predict that the ‘agitation’ of pride would get transferred to the love, making the latter passion more ‘violent’. But this does not mean that at any particular moment you could feel love – a simple impression – without any violence at all. We always experience the passion with a particular level of violence, even if that violence comes from a distinct passion through a causal process so instantaneous that we are unaware of it. The violence of a particular passion is thus inseparable from our experience of it; we cannot imagine experiencing a passion with no level of agitation whatsoever.

But do we always experience passions with a particular violence? Calmness and violence are poles of a continuum, and Hume sometimes seems to indicate that at one extreme of that continuum, we cease to feel a passion although we are still having it. Again, passions ‘may decay into so soft an emotion, as to become, in a manner imperceptible’ (T 2.1.1.3). Likewise, when he explains the confusion of reason with calm passion, he identifies their resemblance as a lack of feeling: ‘certain calm desires and tendencies, which, tho’ they be real passions, produce little emotion in the mind, and are more known by their effects than by the immediate feeling or sensation’ are mistaken for reason ‘because their sensations are not evidently different’ (T 2.3.3.8).

I do not think we can take these passages to imply that Hume believes that we can have entirely unconscious passions, of which we have no sensation whatsoever. This view would conflict with his description of the impressions as feelings, his faith in the power of reflection to illuminate consciousness, and his insistence that the sensations of passions constitute their being and essence. There are helpful qualifiers in these passages: the calm passions become imperceptible in a manner;Footnote ¹⁶ they are more known by their effects than their sensation. And as Qu points out, to say that the sensation of calm passions resembles that of reason is not to imply an absence of all sensation, as Hume describes the operations of reason as having a distinct phenomenology as well (Qu, Reference Qu2018, pp. 451−54). But we can say more about how we feel such passions on Hume's theory – namely, we feel their aiming at ends. I explain this point in section 6.

5. Elevation/Softening

When considering a puzzle about respect and contempt, Hume explicitly distinguishes another quality of passions from their pleasantness or painfulness: they can be either elevating or softening. Respect and contempt are mixtures of love and humility, on the one hand, and pride and hatred, on the other. The pleasing qualities that produce love when observed in another also produce pride when we recognize them in ourselves. But as we compare ourselves to another person, we experience pain if we judge ourselves lacking in those same qualities. The resulting humility, mixed with love, produces respect. Likewise, recognizing pleasing qualities in ourselves, in comparison with others who lack them, produces the mixture of pride and hatred that constitutes contempt. The puzzle, then, is: why do we ever experience love without respect, or hatred without contempt?

Hume responds by arguing that some pleasing qualities are more consonant with pride than others, and some displeasing qualities more consonant with humility. After reminding the reader that love and pride are ‘always agreeable’ and humility and hatred always ‘painful’, he adds:

More elaboration here would be helpful: why does Hume claim that love or tenderness weaken or enfeeble the soul? And what does ‘enfeebling the soul’ mean anyway?

Note that this passage indicates that Hume distinguishes the invigorating quality of a passion from its pleasure: pride and love are both pleasant passions, but one is invigorating and the other enfeebling. Nor can we map this distinction onto the violent/calm distinction, although violence might seem akin to invigoration. Hume explicitly categorizes all of the passions listed here as typically violent, including enfeebling love and humility. Although any passion can be calm or violent, he would not identify a passion as both typically violent and typically not invigorating if these terms referred to the same quality. Moreover, a feeling of humility that is very violent in Hume's sense – viz., involving an intense agitation – can also be very discouraging or dejecting. In fact, the enfeebling aspect of humility would increase with its violence. So not only can we not identify violence with invigoration; in some cases, there will be an inverse relationship between them.

Hume also says that the invigorating passions of anger and hatred ‘bestow a new force on all our thoughts and actions’, which might suggest the passions becoming stronger in his technical sense – that is, more likely to motivate action. But he claims that pride is not a motive at all. It is among the ‘pure emotions in the soul, unattended with any desire, and not immediately exciting us to action’. The quintessentially invigorating passion motivates no specific action at all. Love and hatred, in contrast, ‘are not completed within themselves, nor rest in that emotion, which they produce, but carry the mind to something farther’ (T 2.2.6.3). Hatred reliably produces anger, or a desire for the misery of our enemies. The production of this desire does not make hatred intrinsically motivating: hatred and anger are distinct passions, ‘only conjoin'd . . . by the original constitution of the mind’ (T 2.2.6.6). But as I will explain, we can think of anger as part of the ‘character’ of hatred. In this sense, hatred might be a strong passion leading to a specific action or actions. But an invigorating passion ‘bestows a new force on all our thoughts and actions’ (emphasis added). To be an invigorating passion is something other than being a strong one.

Nonetheless, the distinction Hume makes here does relate to action and motivation. To see this, let us examine another passage referring to the invigorating quality. In ‘Of the Will and Direct Passions’, he considers various factors that affect the passions, including that ‘a very great distance [in time or space] increases our esteem and admiration for an object’ (T 2.3.8.1). Central to his explanation of this effect is the idea that opposition can produce invigoration or elevation of soul: he uses the same language as when talking about pride and anger's common quality: ‘Any opposition which does not entirely discourage and intimidate us, has rather a contrary effect, and inspires us with a more than ordinary grandeur and magnanimity. In collecting our force to overcome the opposition, we invigorate the soul, and give it an elevation with which otherwise it wou'd never have been acquainted’ (T 2.3.8.4, italics added). Conversely, ‘the soul, when elevated with joy and courage, in a manner seeks opposition, and throws itself with alacrity into any scene of thought or action, where its courage meets with matter to nourish and employ it’ (T 2.3.8.9, italics added).

Hume thus recognizes that certain passions, even if they do not directly motivate a particular action, prime us to perform actions in general, especially ones that appear challenging or difficult. The claim is plausible: these passions provide not only energy, but a sense that one's actions can be effective. Some recent psychological research supports this idea, at least with respect to pride.Footnote ¹⁷ These emotions make it more likely for us to act on desires that occur to us while feeling them. They may even make us likely to seek out desires – to look about us for something to work on, perform, or overcome. They create a sense of agency rather than passivity. This effect is more than the mere production of energy. Fear or anxiety might produce energy but be paralyzing, so that the energy cannot be spent in action.

‘Elevating the soul’ suggests the metaphor of ‘greatness of soul’, long associated with a kind of pride. Thus Hume writes in the section on greatness of mind that ‘a well-regulated pride . . . secretly animates our conduct’ (T 3.3.2.13). Is the quality of being invigorating just a quality of pride, rather than of multiple passions? Hume does say in this same paragraph that ‘whatever we call heroic virtue, and admire under the character of greatness and elevation of mind, is either nothing but a steady and well-established pride and self-esteem, or partakes largely of that passion’ and then counts courage among the heroic virtues. But while pride is the quintessentially invigorating passion, it is not the only one.

We have seen that Hume identifies other passions that have the quality. Even if courageFootnote ¹⁸ is a form of pride or a mixed passion including pride, joy and anger are not, and Hume says that they are invigorating. It may seem that anger always carries a tincture of pride, if we conceive anger as a response to feeling slighted or wronged. We can only feel wronged if we conceive ourselves as worthy of better treatment, which is to possess some pride. This conception of anger has deep roots in traditional theories − for instance, in Aristotle, Cicero, and Aquinas.Footnote ¹⁹ It remains popular in contemporary emotion theory.Footnote ²⁰ It is not, however, Hume's view. Again, he believes that hatred reliably produces anger. But any enduring quality of a person that produces pain or aversion in others can produce hatred. ‘One that is disagreeable by his deformity or folly is the object of our aversion’, he writes, ‘tho’ nothing be more certain, than that he has not the least intention of displeasing us by these qualities’ (T 2.2.3.4). We might try to justify hatred or its consequent anger by imputing vice to people who annoy us for morally innocent reasons, but we need not actually feel slighted, since ‘independent of the opinion of iniquity, any harm or uneasiness has a natural tendency to excite our hatred’ (T 2.2.3.9). Humean anger, then, as the natural result of Humean hatred, implies no necessary sense of injury and therefore no necessary underlying pride.

Does anger have an elevating effect? It is difficult to know how to begin answering this question, although the feminist reclaiming of anger in the service of power is suggestive.Footnote ²¹ I find some evidence, however, in the greatest novel of the great nineteenth-century psychologist, George Eliot. The heroine of Middlemarch has discovered the man she loves and respects in a compromising position with another woman. She hastily leaves, and the narrator describes the ‘fire of Dorothea's anger’ thus:

Being ‘animated by a more self-possessed energy’ and needing ‘something active to turn [one's] excitement out upon’ strikes me as an apt description of what Hume means when he says that such passions invigorate the soul.

What of ‘enfeebling’ passions? Hume names humility and love; it is safe to add joy's opposite, sorrow, to this list. Not all forms of sorrow have this effect. After all, there is much sorrow in Dorothea's anger. And Hume acknowledges that grief can lead to anger.Footnote ²² But then there is the sorrow that remains when anger is gone, leaving Dorothea crying all night on a cold floor. A gentler melancholy enfeebles too.

When we turn to love, we confront a problem with Hume's terminology. First, he must have in mind only specific kinds of love. Some love rivals pride for invigoration and elevation. He hints at the variety of love he has in mind, referring to ‘love or tenderness’ (italics added). Later, listing qualities that he thinks well-suited to produce love but not pride, he includes ‘good nature, good humour, facility,Footnote ²³ generosity, [and] beauty’ (T 2.2.10.8). The relevant form of love would be that generated by these gentle traits. I imagine the love one feels for sweet children, or the woman in my mother's hometown whom everyone called ‘Aunt Billie’, regardless of relation. But I can also imagine that Hume has in mind something like the friendship love he describes in ‘Of Polygamy and Divorces’ –‘a calm and sedate affection, conducted by reason and cemented by habit’ (E 189). It would be difficult to maintain a spirit of invigorated pride if suddenly confronted with the object of such loves.Footnote ²⁴

Second, even if we understand love to refer only to these tender feelings, ‘enfeebling’ connotes a negativity Hume cannot intend for a passion so pleasurable and with such good effects. It is calming or tempering, not debilitating. I therefore propose the term ‘softening’ to refer to this quality of the passions. If the invigorating passions prime us to perform actions in general, the softening ones coax us to slow down.

6. Direction

Finally, Hume identifies one other quality in which passions can resemble one another – their ‘impulses or directions’. Surprisingly, he treats the tendency of passions to motivate actions or even produce desires that motivate actions as part of their qualitative character.Footnote ²⁵ That such tendencies are part of some passions’ feel helps him explain why pity or communicated sorrow tends to produce something like love rather than something like hate. This is a problem for Hume because pity arises when we feel someone else's sorrow by sympathy, or the adoption by contagion of others’ passions.Footnote ²⁶ According to Hume's theory of the double relation of impressions and ideas, pain received from another person should generally produce hatred. Some quality of the other person causes one pain, and pain resembles hatred in its painfulness. Yet ‘there is always a mixture of love or tenderness with pity’ (T 2.2.9.1). To explain this apparent anomaly, Hume writes that ‘’tis not the present sensation alone or momentary pain or pleasure, which determines the character of any passion, but the whole bent or tendency of it from the beginning to the end. One impression may be related to another, not only when their sensations are resembling, as we have all along supposed in the preceding cases; but also when their impulses or directions are similar and correspondent’ (T 2.2.9.2).

The chain here is somewhat tangled: pity, like benevolence, ‘is a desire of happiness to another, and aversion to his misery’ (T 2.2.9.3). It therefore resembles benevolence in its ‘bent and tendency’. But benevolence, Hume believes, has a natural tie to love: love reliably produces benevolence, although they are distinct passions. Thus pity tends to produce love as well.

The ‘bent or tendency’ of passions is another way in which passions can resemble one another. It is a quality of those passions that either are desires (like anger, a desire for the misery or another) or reliably produce desires (like hatred, which generates anger).Footnote ²⁷ For this resemblance to obtain, the ‘character’ of love or hatred must be determined partly by what are strictly speaking effects of these passions. Hume insists that benevolence is distinct from love because he believes that one may feel love without the desire for well-being arising until ‘the ideas of the happiness . . . of our friend’ occur to us (T 2.2.6.5). Once we consider that happiness for some reason, we desire it, but meanwhile love might not express itself this way.

Yet when Hume introduces the relation between love/hatred and benevolence/anger, he writes that the former passions ‘are always follow'd by, or rather conjoin'd with benevolence and anger’ (T 2.2.6.3, italics added). He continues to refer to the relation as a conjunction in the following.Footnote ²⁸ What he seems to be suggesting (inchoately) here is that we commonly experience love and hatred as an anticipation – a reaching toward another feeling, itself a motive that goes beyond the immediate sensation of these passions in isolation. Love is naturally the beginning of benevolence: when the requisite ideas occur to the lover, desire for the well-being of the beloved becomes part of her experience of love. This suggestion helps explain why people believe such a desire to be an essential part of love, although on Hume's analysis this belief is mistaken.

If the above account is correct, passions that reliably generate desires share the quality of directedness toward some object and, though they are not desires, resemble desires in this respect. A desire, for Hume, is a motive to action: to experience a passion with this quality, then, is to experience oneself as motivated to a particular action. This feeling is part of the phenomenological experience of the relevant passions. In cases of love/hatred and benevolence/anger, this experience, unlike that of a single passion, is not simple. It is the experience of a compound passion. But just as we rarely dissect our impression of an apple into distinct impressions of color, taste, and smell, we do not typically dissect our experience of love into the pleasant feeling directed toward another person and our desire to benefit her. Both are part of the ‘character’ of our typical experience of love.Footnote ²⁹

The feeling of directedness is not identical to the will, which Hume defines in the Treatise as ‘the internal impression we feel and are conscious of, when we knowingly give rise to any new motion of our body, or new perception of our mind’ (T 2.3.1.2). Directedness, in contrast, can be felt before acting or without the motive leading to action at all. Although love and hatred have the quality of directedness, they do not always lead to action. And even desires can fail to do so; they may be frustrated by external circumstances, awareness of one's incapacities, or conflicting motives. They nonetheless possess the character of directedness.

We can now return to the puzzling claims Hume makes about calm passions being ‘more known by their effects than by the immediate feeling or sensation’ (T 2.3.3.8). Given his definition of passions as reflective impressions, the suggestion that passions might be known by their effects rather than by their sensations is problematic. Impressions are feelings,Footnote ³⁰ and he claims that the sensations of pride and humility ‘constitute their very being and essence’. It cannot be, therefore, that we have passions without feeling them at all. But Hume does not say that we feel nothing in experiencing calm passions; they are only ‘more known’ by their effects.Footnote ³¹ These effects might be the impression of the will, actions themselves, or perhaps desires in the case of a compound of love and benevolence, for instance. In this case, the feeling of the effect would count as part of the character of the passion, though not an ‘immediate feeling’, as it is mediated by another passion we can separate in the abstract.

Introducing the possibility of calm passions, Hume names a set of ‘desires and tendencies’ that are ‘readily taken for the determinations of reason’–‘benevolence and resentment, the love of life, and kindness to children; or the general appetite to good, and aversion to evil consider'd merely as such’ (T 2.3.3.8). But he later suggests that any passion that becomes ‘a settled principle of action’ can be calm (T 2.3.4.1). (‘Principle’ refers here to a source, not a rule.) Any desire experienced often, through the force of custom, can become a more efficacious motive at the same time as it becomes a less agitating emotion. There is a sense, then, in which we feel such passions less. But we still feel them, in part because to feel ourselves motivated is to feel the motivating passion, and sometimes the passion it is typically compounded with as well.

What would such a feeling be like? It cannot be a compound of a passion-feeling with an idea of some desire's object. The idea of such an object would be separable from the feeling (since we could have the same idea without feeling any passion about it, or while feeling a different passion about it). According to Hume's separability principle, then, desiring passions would all be complex because they would be made up of two separable components. But as a single passion, desire is simple. The feeling of love plus benevolence may be a compound, but Hume cannot say that the feeling of benevolence alone is.

Instead, such passions would include what Peter Goldie calls ‘feeling towards’– an ‘unreflective emotional engagement with the world beyond the body’ (2002, p. 241).Footnote ³² (It is ‘unreflective’ because it does not require consciousness of the emotion, although humans acquire the capacity for such consciousness as they move away from infancy.) Goldie thinks that the nature of this essentially personal feeling cannot be adequately characterized from an impersonal, theoretical perspective. But he uses the example of Irene, ‘an icy-cool ice scientist’, to illustrate the difference between a judgment without feeling about a danger and a fearful feeling-toward the object of that danger. Irene, cognizant of the dangers of falling on ice but never having experienced such a fall, will have a transformed phenomenological experience of ice after falling. She now thinks of the ice ‘with fear’, and this new way of thinking has effects throughout her mental economy (Goldie, Reference Goldie2002, p. 245). Before she recognized that she should try to avoid falling on the ice; now she desires to avoid the ice-as-fearful-object. She can empathize with others’ fear, imagine fearful experiences she has not yet had, and re-imagine her own past experiences of danger.

To bring this analysis to Hume's theory of the passions, consider a standard instance of Humean benevolence, a father caring for his child. The father habitually prepares dinner for his daughter every evening. This habit grows out of his love for her, which produces a desire for her well-being, including an aversion to her hunger. We can consider a Humean desire as a feeling-toward an object, which involves attraction to the object's attainment and aversion to its not being attained. Assuming that father and daughter are food secure and in affluent enough circumstances to rarely encounter obstacles to meals, the father's feelings will be calm. His desire to feed his daughter will not agitate him positively or negatively, yet he will feel himself desiring that she eat insofar as he experiences himself as motivated to prepare dinner.

Suppose circumstances were to change and he struggled to find food for his daughter. He would then experience his desire as more agitating: the prospect of his daughter's going hungry confronts him with horror. Because it is a prospect for a future event, it presupposes an idea: he must imagine her experiencing the pain of progressively stronger hunger.Footnote ³³ But the desire does not contain the idea; it is not itself a representation of the hunger. It is instead a way of feeling-toward that hunger, much as Irene feels-toward the ice after her fall. In other words, the desire is not an idea; it is a way of experiencing the idea.

If there is a real danger that the father will be unable to feed his daughter, he experiences the idea of her hunger with fear or, in more extreme cases, grief. The pain of this emotion may be the most salient aspect of his experience, but if the situation is not hopeless, he may feel the direction of his benevolence most strongly, as he tries harder to find her sustenance. In the everyday situation where he has no trouble providing this sustenance, however, all he may feel is his directedness toward the end of providing for her welfare. But this feeling too is an aspect of his experience of benevolence.