Contemporary compatibilists remain unable to provide a plausible account of self-determination, i.e., of what makes certain motivationally effective psychic elements "mine" and what makes others "external" to me (also described as the problem of identification). I attempt to develop a solution to this problem by exposing and explaining the special relation between motivation and caring. The implications of this phenomenologically-informed picture, however, are that both freedom and identification are matters of degree and consist in a particular type of necessitation about which we care deeply.
CARING, IDENTIFICATION, AND AGENCY
David W. Shoemaker
California State University, Northridge
Department of Philosophy
18111 Nordhoff St.
Northridge, CA 91330-8253
CARING, IDENTIFICATION, AND AGENCY* Any robust theory of free agency must account for its two central features, namely, the availability of alternative possibilities and self-determination (i.e., autonomy).1 I here wish to focus on this latter feature, specifically with respect to its treatment by compatibilists. For both classical (e.g., Hobbes/Hume) and contemporary compatibilists, free agency involves a dependence relation between a person's actions and, roughly, that person's wants. Where the two differ, however, is in their explication of the genuine wants of a person and the relation of those wants to self-determination. For the classical compatibilist, one's will – one's effective desire – just is one's self. Thus, to be self-determined is to be determined by one's will. The only possible impediments, then, to free action are external impediments, obstacles to the will of the self from forces outside the physical boundaries of the self. But this conception of free agency is recognized by contemporary compatibilists to be incomplete, for even though one's externally unimpeded actions may depend on one's will (one's effective desire) in one sense, there is another crucial sense in which one's actions may conflict with, or be independent of, what one really wants, a state of affairs illustrated most starkly by examples of compulsion or addiction. These are cases in which what moves one to action is not what one genuinely wants to be moved by, cases in which one's autonomy is impeded not by anything external, but by one's own effective desires to the contrary.
Of course, these desires remain external in a sense as well. Specifically, they are external to the self – external not to the physical boundaries of the self, but to the psychic elements with which the self identifies. But if self-determination is not always equivalent to determination by one's effective desires, then what precisely does it involve? What does it mean to regard a particular motivating psychic element as external, as an element with which one does not identify? This is the problem faced by contemporary compatibilists, and their response has generally been to construct complex divisional models of the self in which freedom is a matter of a special kind of harmony – a harmony stemming from critical self-reflection – between the psychic elements of the various divisions. Thus Harry Frankfurt, for example, lays out a model of a self divided horizontally, according to which certain of one's desires have lower-order desires as their objects, and a person's will is free to the extent that it depends on her second-order volitions (second-order desires about which first-order desire is to be effective).2 In contrast, Gary Watson has sketched a model of the self divided vertically, according to which a person has both a set of considerations moving him to action and a set of considerations yielding judgments about the relative worth/value of various possible actions, and a person is free to the extent that these motivational and valuational systems coincide.3
Unfortunately, both approaches are problematic, for well-known reasons. Specifically, neither model can plausibly account for the notion of identification, for what precisely it is that provides the special authority required in self-determination (specifically with respect to externality). Frankfurt's model, for example, yields the view that what makes a certain motivationally-efficacious first-order desire mine (i.e., one with which I identify) is that I form a higher-order volition to have it be my will. And what renders a desire external to me is that I have a contrary higher-order volition about it. But of course higher-order volitions are just desires themselves, so what could it be about their higher-order status that lends them any special authority whatsoever with respect to self-determination and externality?4 And Watson's model suffers from a similar problem. On his view, it is only evaluations (judgments about worth/value) that provide one with reasons to render certain desires external, and it is "only when agents' behaviour expresses their evaluations [that they are] sources and 'authors' of (because they 'authorized') their behaviour."5 Nevertheless, as Watson himself now recognizes, clear-thinking, rational persons may not always identify, in a particular case, with those things they generally judge best. One may, after all, upon due deliberation and reflection, judge that in certain circumstances one course of action would be best and yet still pursue and embrace as one's own a different course of action when one finds oneself in those circumstances, a course of action that is, perhaps, simply more fun or stress-free.6 Things I judge to be external to my self from the evaluational standpoint may nonetheless be fully embraced by me as my own in particular circumstances without, it seems, undermining my freedom in any way. Consequently, contemporary compatibilists "are left with a rather elusive notion of identification and thereby an elusive notion of self-determination."7
What I wish to do is offer a way out of this morass for the interested compatibilist. In so doing, I intend to defend five theses, culminating in the conclusion that identification is, for the most part, a passive process, garnering its authority for self-determination from one’s nexus of cares. I will offer these theses as inductively-derived phenomenological claims that I hope will resonate powerfully with the reader and then I will attempt to defend them from various possible objections. In proceeding in this way I intend to co-opt what libertarians often employ as a powerful methodological tool for their own cause, viz., gathering evidence about agency "from the inside."8 If this methodology actually yields significant evidence for compatibilism, and if by the end we have articulated a compatibilist view that can dissolve the crucial problem of identification, then the plausibility of compatibilism itself should be greatly increased. I begin with the thesis requiring the most exposition and defense.
Thesis 1: What we typically, upon reflection, are motivated to do, in any given situation, depends ultimately on what we care most about, with respect to that situation. There is much to explain here, and I begin by phenomenological reportage. When I enter most intimately into what I call myself, it seems nearly everything I find that has, or has had, some motivational effect on me is dependent in some way or other on things I care, or have cared, about. Consider some of the variety of psychic elements that may be motivationally efficacious. It seems that, when they move me to action, my desires, aversions, and evaluational judgments are all typically dependent in some way on cares that I have. To speak rather loosely, my motivationally efficacious desire to is typically derived from my caring about something which my -ing would help promote/preserve. My motivationally efficacious aversion to -ing is typically derived from my caring about something which my -ing would betray/undermine. And my motivationally efficacious evaluational judgments are, when they make reference to things I in fact value, simply judgments about a proper subset of things I care about. Furthermore, when I turn outward from myself to try and motivate others, I do so by tapping into things they care about, e.g., "If you truly care about X, you ought to ." Caring seems to be the great motivator.
But what does it mean, precisely, to care about something, as opposed to desiring it or judging it valuable? What exactly are the objects of care? What is the nature of the dependence relation between desiring, evaluating, and caring? Furthermore, what does it feel like to care, as opposed to wanting, or to judging valuable? And finally, just how sweeping is this thesis intended to be: is it in fact supposed to be true for all motivated actions? These are essential questions, of course, and I need to take some time to address them.
First of all, what is caring? Caring involves emotional, desiderative, and evaluative elements, but the relation between the three is quite complex. First of all, in caring for X, I am rendered vulnerable to gains and losses – to emotional ups and downs – corresponding to the up-and-down fortunes of X. I have an emotional investment in what happens to the cared-for object, so that there is a cost I incur, such as disappointment or sadness, when what I care about is negatively affected (when I am aware of this negative effect, of course).9 When I care about something, I am tied to it, as it were, by an emotional tether, such that my feelings are tugged this way and that in accordance with the way – negatively or positively – it is affected by various events. Caring about something thus often involves a kind of nervous (and occasionally exhilarating) vulnerability: I am always the potential subject of hurt, or disappointment, or frustration, and these feelings are dependent on forces affecting the things I care about that are, for the most part, beyond my control. But of course I am also the potential subject of joy, depending on whether or not those forces beyond my control affect the cared-for object in a positive way.10 Second, and as a direct result of such vulnerability, certain dependent desires crop up, desires to act on behalf of the cared-for object, e.g., to protect it and/or contribute positively to its maintenance or development. Third, in many cases my caring for something will also involve a judgment of its value: I will judge it to be something worth protecting, maintaining, or developing. In what follows I hope to flesh out these remarks in some detail.
Consider first the relation between caring and emotional dispositions. Suppose I care a great deal for my dog. Such caring will dispose me to a variety of affective states. For example, when I come home from work and see his tail wagging and his mouth open in what I take to be a smile, my affection for him will come rushing to the forefront (producing dependent desires as a result: I will want to pet and play with him and give him a treat for not tearing up the house while I was away). And when I see my dog limping or listless, I will be distressed (wanting, as a result, to get him to the vet and fix whatever the problem is). Indeed, caring requires such emotional vulnerability; I would, after all, reject any attempt to block the feelings of distress attached to seeing my dog in pain. Suppose I took a drug that caused me to have no negative emotions upon seeing my dog limping and the like, while I still could experience the positive emotions upon seeing him healthy and happy. How could I genuinely be said to care anymore about my dog? I would be like the fair-weather fan, in this case, coming out to cheer on the local team only when they are winning. The complaint of the true fan, that these bandwagon-jumpers don't really care about the team, then, carries significant weight. Genuine caring about something involves a package deal: one must, along with the possibility of joy (and other positive emotions), accept the possibility of distress (and other negative emotions) when things are not going well with the cared-for object in order for one truly to be said to care for it in the first place. To claim to love something only when things are going well for it seems to be no love at all.11
The relation between cares and affective states is extremely tight. We have just seen how the disposition to experience a range of emotions seems necessary for care. But are emotional reactions sufficient to indicate the presence of a care, i.e., are all experienced emotions enough to guarantee that one cares about something related, or can one have certain affective reactions without their being tied to anything one cares about? I believe that, for mature adults, emotional reactions are sufficient in this way, and the reason is that our experiencing of emotions makes sense only under an interpretation involving reference to our cares. In other words, it is constitutive of emotional reactions that there ought to be – in the sense that there is expected to be – some care thereby reflected or revealed. We articulate cares in light of our emotional reactions – it is the only way to "rationalize" them – and so of course emotional reactions are sufficient as indicators of care. What makes them emotional reactions – as opposed to mere physical reactions – is just that they are sensibly interpretable (and typically interpreted) within a matrix of cares. An emotional reaction, in other words, must be a reaction to events affecting something that matters; otherwise, it just cannot be an emotional reaction.12
I take it to be an analytic truth, then, that creatures like us who undergo emotional reactions care in ways that make sense of those reactions. If one has emotional reaction X, the only way to render X intelligible as an emotion is by reference to care Y. Furthermore (as we have already seen), there is a tight expectability relation going in the other direction as well: if one actually cares about Y, one is expected to have emotional reaction X, depending on what happens to Y. If one does not, then it seems that one's claims of caring are disingenuous.13
What, then, does it feel like to care? If what I have suggested is correct – that caring is a kind of conceptual/linguistic structural framework we overlay upon emotional reactions to render them intelligible – then the question itself is a category mistake. Talk of caring is simply a way of referring to the range of emotional reactions one is expected to have with respect to the fortunes of the cared-for object. To care is thus to have the disposition to feel, but it involves no – and could involve no – feeling in and of itself.
What, then, are the objects of care? Given the connection between caring and emotions, and given the very wide and diverse range of events to which we may have emotional reactions, the range of potential objects of care is itself extremely wide and diverse. One may care about oneself, other people (and/or the plight of other people), animals (and/or the plight of animals), the environment, how one is treated, peace on earth, works of art (and beautiful things in general), God, fictional characters, morality, one's nation, one's ideals, and even other of one's cares. And this is just the tip of the iceberg. We have emotional reactions to a manifold of events, and the wide-ranging diversity of the things we care about reflects our thoroughgoing emotionality. As I will explain in detail later, the emotions we have make us the agents we are.
In the variety of examples just mentioned, however, it may seem that an important distinction – between caring for and caring about – has been ignored. After all, caring for certain others, as many feminist thinkers have shown, and as my dog example reveals, is a rather thick concept, involving as it does a personal concern for the object of care for its own sake, whereas caring about things (like peace on earth, say) is a more general, or much thinner, concept, and does not seem to involve such personal concern. And so, one might say, while caring for someone will give rise to the emotional reactions I have discussed, caring about something does not necessarily have such a close tie to emotions. After all, I may care a lot about wearing clean shirts, but that certainly won't involve any concern for anything for its own sake, and it thus may seem difficult to imagine it giving rise to any real emotional reactions.14
I am inclined to view this distinction as somewhat overdrawn. What is crucial to caring – of any sort – is precisely the investment involved. And while I can obviously be invested in the fortunes of other people (or sentient creatures), where this is typically articulated as a personal concern for them for their own sake, I can also be invested in the fortunes of things which, we might say, have no sakes, i.e., there is no coherent notion of well-being applicable to them. I may find myself invested in either sort of object, and while the effects on these objects may be assessed in different ways, the fact that those effects may in turn affect me emotionally is what matters here, and it constitutes the common thread running throughout the variety of cared-for/about objects. I see no reason to believe, therefore, that if I really care about wearing clean shirts, I won't be upset when I realize I have forgotten to do the laundry.
Nevertheless, these considerations do raise questions about the distinction between caring, desiring, and evaluative judgments. Consider first the relation between desires and cares. While there is an entailment relation from cares to desires, the converse clearly does not hold. For me to care about X entails that I will have certain desires with respect to X, desires whose frustration or satisfaction is what gives rise to the type of emotional reactions I have already discussed. But to desire X does not entail that I care about X. To use an example I will explore in more detail later, I may now desire some vanilla ice cream more than I desire some chocolate, but I may very well care about neither.15 Now as we shall see, there may yet be cares at work in my decision about which flavor to order, but the desires involved in and of themselves entail no cares. To put this another way, these are desires about things I simply do not regard as important; I am not committed to their objects in such a way that the frustration of my desires for them will affect me emotionally.16 Finding out the shop is out of vanilla may, in such a case, simply involve my shrugging and picking chocolate.
The structural relation between cares and desires is thus asymmetrical. When I care about something, this caring produces a web of desires, e.g., I will, as a result of my caring, want the cared-for object preserved or protected or enhanced, etc. But desiring something, in and of itself, does not necessarily mean I care about it, and this is true even if I want it a lot.17 Surely though, one might say, the frustration of such a strong desire will produce an emotional reaction, in which case the earlier analysis implies one cares about it. But this claim does not follow, for even if an emotional reaction is produced, that does not mean one cared about the desired object; rather, it could mean one cared only about having such a strong desire not be frustrated, a state of affairs that could have been produced not only by a satisfaction of the desire but also by anelimination of the desire itself. This point will be crucial in my later discussion of addiction.
What, though, of evaluative judgments? Here, too, there is an important distinction between judging valuable and caring. On the one hand, to judge X valuable is not necessarily to care about X. For example, I may recognize the considerable intrinsic value of a possible project or way of life, while failing in any way to be drawn to it myself.18 I simply find myself without any commitment to it – it's just not important to me – although I can easily see how it could indeed be a good pursuit.19 On the other hand, to care about X is not necessarily to judge X valuable. I may, for instance, care here and now about enjoying myself by bungee jumping, even though I now judge and have consistently in the past judged such an activity to be without value.20
This is not to say, however, that there is not a significant connection between caring, desiring, and evaluative judging. As we will see developed throughout the present work, I take the ordinary structure of the will to be as follows: one first comes to care about X, i.e., one becomes subject to certain emotional reactions dependent on the fortunes of X, where this caring produces certain desires with respect to X (having to do, e.g., with promoting/preserving X), and one will then often, but not always, judge X to be valuable in light of one's caring, depending on the overall role X plays in one's life. One or more of these desires or judgments will typically then serve as, or set parameters on the range of, one's will, and action follows. And it is with this claim we are finally brought round again to my first general thesis: what we are typically moved to do depends ultimately on our cares.
What precisely is involved, then, in the relation between motivated action and care? Knowing what we now know, we can see that it is simply false that all my motivated actions depend on things I care about. We all do a variety of things each day that seem to bear no dependence relation at all to our cares, e.g., we get out of bed, we scratch itches, we reach for the milk, we change the TV channels, etc. These are all intentional, motivated actions, explained (rendered intelligible) by reference to certain of our desires, one might say, without any necessary reference to things we regard as important, things whose changing fortunes would tug on our emotional tethers. These are instances in which we act as wantons, then, as unreflective humans who simply do not care what our wills are to be.21 In such cases we are moved by various impulses, with no real reflection on whether these are the impulses that we want to move us, or on whether these impulses flow from what we care about, and the reason is usually that the situation just doesn't warrant that kind of reflection – the situation itself just doesn't matter.
I am not, then, claiming that all intentional, motivated actions are dependent on cares. Instead, I want to defend a much more restricted claim, namely, that when we find ourselves in situations that matter, where we do care about what our wills are to be and where we are usually thereby driven to self-reflection on the nature of our wills, our actions are typically (in a sense I will explain shortly) dependent on our cares. Now this may have the ring of a merely trivial truth: when we care about what we do, what we do reflects what we care about. But this reading is mistaken. I am claiming that our capacity for self-reflection is called into play by situations in which there is something at stake in the outcome of our action (or inaction). These are situations in which I care that my will should perhaps not be constituted by a mere unreflective impulse, or momentary desire. And it is in these cases, I am suggesting, that what I ultimately am moved to do is typically dependent on other things I care about. It is the motivation for non-wanton action – the action of Frankfurtian persons – that is grounded in cares, and this is not a trivial claim.
Before defending this claim, however, a few qualifications are in order. I have in mind the dependence relation that typically holds between motivation and care, and all I mean by this is that such a relation virtually always holds. There are a couple of implications of this qualification, then. First, what I will be discussing is merely a contingent relation, not a conceptual, or analytic, one. As it so happens, given our construction as emotional creatures and the way emotions ordinarily color our decision-making processes, the motivations for our (non-wanton) actions ordinarily – when things are functioning properly – flow from our cares. But this means there will be non-standard, exceptional cases, and I will discuss them below. Second, in light of what I have just said, I am not explicitly defending an internalist account of motivational reasons here. Both externalists and internalists can at least agree that such "reasons very often have subjective conditions,"22 and this is all I need to move forward in the argument. It is the nature of the purported counterexamples to this claim about subjective conditions in general that is the source of their debate, whereas I am merely concerned to show what those subjective conditions involve, so while what I say may have some bearing on that debate, I don't believe that my remarks will, in the end, carry the day for either side.