Acting Contrary to Our (Professed) Beliefs

Acting Contrary to Our Professed Beliefs

Eric Schwitzgebel

Department of Philosophy

University of California at Riverside

Riverside, CA 92521-0201

eschwitz at domain- ucr.edu

951 827 4288

January 9, 2007

Abstract:

There appears to be no ordinary language occurrent use of “believing”. Further, using the phrase “occurrent belief” to describe instances of judgment can prevent us from recognizing that judgment is often insufficient for belief. For example, someone may sincerely judge (even know?) that all races are of equal intelligence (or that death is not bad, or that the bridge is closed) without undergoing the broad dispositional transformation necessary to make us comfortable ascribing her the corresponding belief. Our dispositions do not always fall neatly into line when we reach a judgment. Often it takes work to fully, dispositionally believe something we occurrently judge to be the case.

Acting Contrary to Our Professed Beliefs

Eric Schwitzgebel

October 13, 2006

Contemporary philosophers of mind generally distinguish between what they call “dispositional” and “occurrent” belief – roughly, between believing as being, in general, prone to do and say certain things in certain situations and believing as entertaining and holding true a particular proposition at a particular moment. Typically, this distinction is invoked with little discussion, as though it were clear and uncontroversial (for instance, in Burge 1979; Fodor 1987; Kim 1998; Williamson 2000; just to mention a few).[1] I aim in this paper to suggest the naturalness and advisability of treating belief wholly dispositionally and abandoning talk of “occurrent belief”. Of particular interest are cases in which one’s overall pattern of behavior and reaction fail to line up neatly with one’s occurrent judgments – cases in which, one might say, we act contrary to our professed beliefs.[2]

First: What is the difference between disposition and occurrence that appears to underwrite talk about dispositional and occurrent belief? One useful way of thinking about it is the following. An occurrence is an event – a particular event that transpires at a particular time and place. A disposition is a proneness or tendency to be involved, in a particular way, under particular conditions, in events of a particular type. A vial of Morton’s salt dissolves in Lake Erie on July 3. That is an occurrence. Of course, also, salt is generally prone to dissolve in water (under particular conditions). Solubility in water is a disposition that salt possesses, a dispositional property of salt. Here’s another example: In college, Jamie read the Bible. In saying this, we can intend to refer to a single, particular occurrence – a single (protracted) event of reading through the Bible – or we can intend, instead, to impute to Jamie a certain dispositional property, the property of having been a Bible-reader, someone with a proneness or tendency to read the Bible.

We can, if we like, distinguish pronenesses from tendencies. Jamie did not in fact have the dispositional property of being a Bible reader if he did not on several occasions read (some portions of) the Bible, regardless of how prone he may have been to read the Bible, had circumstances allowed it. In saying he was a Bible reader, we attribute to him a (past) tendency or habit, requiring multiple instances of fulfillment. In contrast, a vase can be fragile – can have the dispositional property of being prone to break – even if it has never in fact broken. The mere proneness to break, alone, unactualized, is enough to underwrite our dispositional ascription. Some dispositional ascriptions don’t require multiple instances of fulfillment but still require one instance: She bench-presses 200 pounds. Philosophers have not always been careful about distinctions between these types of dispositional ascription.[3]

I put my claim about Jamie – that he read the Bible – in the past tense because the ambiguity between dispositional and occurrent trait attributions tends to dissolve in the present tense. We say, dispositionally, that Jamie reads the Bible, or occurrently that he is reading the Bible. We say, dispositionally, that Corina runs a six-minute mile, or, occurrently, that she is running a six-minute mile. In the present tense, English marks the dispositional/occurrent distinction fairly well, and better than many other languages.

Returning to belief, then, we say, dispositionally, that Armando believes that New York City may not exist in 200 years and, occurrently, that Armando is believing that New York City may not exist in 200 years. Wait – no we don’t! Why does my word processor mark that last italicized phrase as a grammatical error? My computer marks no error – and there is no violation of ordinary usage – when I write that Armando is guessing that New York City may not exist in 200 years. Does Microsoft know what my colleagues mostly deny – that “believe” has no occurrent use? Let’s try Google, perhaps a more beneficent company. When I enter “is believing” I find instance after instance of “seeing is believing” – but of course that phrase is not in the present progressive characteristic of ongoing occurrences. If I exclude pages containing the word “seeing” I still find only phrases with the same structure: “hearing is believing”, “skiing is believing”, “cc-ing is believing”, not a present progressive in sight.

But we philosophers needn’t be(-ing?) chained by grammar (especially if it’s defined operationally by appeal to Microsoft’s grammar checker). The argument from Google doesn’t yield apodictic metaphysics. Let’s look at the sense. Jamie does not constantly read the Bible. He ceases sometimes. He is a Bible reader because he has a tendency, occasionally, to read the Bible for a while, after which he stops reading. Likewise for Corina: When breakfast calls, she ceases to run entirely. If Armando has the dispositional property of believing in the impermanence of New York City because he is disposed to occasional fits of occurrent belief in its impermanence, then what are we to say when his fits are over? That he ceases to believe (but still believes in the dispositional sense), as Corina ceases to run (but still runs in the dispositional sense), and Jamie ceases to read (but still reads in the dispositional sense)? Now we butt up not just against the grammar checker but against robust folk-psychological intuition: Armando does not in fact cease to believe in the impermanence of New York City when his mind turns to other things (when he is no longer “occurrently believing” it). He continues to believe, without recess. So there’s a structural difference here – a difference between Armando’s believing and Corina’s running. Structurally, it’s not a typical case of a term with both a dispositional and occurrent sense.[4]

This suggests, perhaps, that if we are not to overthrow common sense and grammar – which despite my attraction to iconoclasm I think we should hesitate to do lightly – we should embrace the idea that believing is not an event of short duration, something that transpires briefly and is then over, like morning runs and Bible-reading sessions. Being a believer is not a matter of being disposed toward short bursts of belief that quickly expire in the same way that being a runner means being disposed to go running for a while. If believing is an occurrence or event at all, it is one of long duration, less like the event of running than like the event of being a runner.

ii.

It may seem that I am quibbling. When philosophers say that someone “occurrently believes” something, I think I know more or less what they mean. There are nearby words with occurrent uses: “is thinking”, “is judging”; we could take occurrent belief as roughly synonymous with those. (But note: The occurrent use of “think” diverges considerably from the dispositional.) Why not forgive them the word, allow “believe” to be given an occurrent sense in technical philosophy of mind?

Here’s the problem. “Occurrent belief” sounds like a species of actual belief – a species of belief in the ordinary (dispositional) sense of the term. The phrase invites one to suppose that the person who “occurrently believes” that such-and-such also, thereby, believes that such-and-such in the ordinary, proper, and preferred sense of “believes”. But occurrences of the type of mental event most naturally interpreted as the referent of phrase “occurrent belief” – what I would call judgments – do not entail the presence of belief proper. The usage of the phrase “occurrent belief” thus occludes an interesting set of phenomena, cases in which – as I would put it – judgment is insufficient for belief. I will dedicate the remainder of this paper to exploring such cases.

One model I wish particularly to resist is the following. To believe (dispositionally) that some proposition “P” is the case is to have a representation with the content “P” stored somewhere in the mind or brain (in some “belief box” or “file folder” or “memory store” perhaps); and to believe P, occurrently, is somehow to bring that representation forward into view – to bring it into “working memory” or a “central workspace” – for some brief period.[5] Philosophers and psychologists sometimes forget how metaphorical such talk is. I’m sure it’s a useful metaphor in some contexts, but it inhibits understanding the types of phenomena that interest me most in belief – in-between cases of believing, self-deception, habit, gradual learning and forgetting, know-how, indeterminacy, vacillation, muddlement. It’s too clean to serve as a good model for such things.

iii.

Many Caucasians in academia sincerely profess that all races are of equal intelligence. Yet I suppose many of those same people would also be less quick to credit the intelligence of a black student than a white or Asian student, feel some (perhaps suppressible) twinge of reluctance before hiring a black person for a managerial job requiring mental acuity, expect slightly less from a conversation with a black custodian than a white one – and, in short, reveal through their actions a pervasive if subtle racism. Such people, you will perhaps agree, don’t fully and completely believe in the intellectual equality of the races, as genuine and unreserved as their rebukes of racism may be. Such cases needn’t involve self-deception. One may be fully aware of a need to reform, to overcome the racism implicit in one’s everyday reactions.

Or: Having been won over by the Stoics, or by pessimists, or by believers in eternal glory, one may quite sincerely judge – either on one particular occasion or repeatedly – that death is not bad. Nonetheless, one may tremble before the executioner (and not just at the anticipation of pain); one may regret the death of a good person (and not entirely on behalf of those who have lost the benefit of her company); one may attempt to forestall one’s own death (and not wholly from a sense of duty). In this case, as in the racism case, and indeed I think quite widely, one may intellectually accept a proposition, take it unreservedly as true, and yet fail to possess the general dispositional structure we would expect from someone who believes the proposition endorsed.

Such cases can be described in several ways. We might say something like this: Juliet says she believes in the intellectual equality of the races, but she doesn’t really – not deep down. Kaipeng says he believes that death isn’t bad, but he doesn’t believe it in his “heart of hearts”. In a way, I agree with these attributions: Juliet and Kaipeng, as I’m imagining them, do avow belief in propositions they don’t actually fully believe. But these characterizations also suggest that there is some self- or other-deception involved, some imperfection of sincerity, some respect in which Juliet and Kaipeng are not fully convinced. Certainly we can, if we wish, imagine them to be less than fully convinced; but I don’t think we need to regard them that way. One may be absolutely, completely persuaded of the truth of a proposition, in the sense of reaching a sincere, unequivocal, unmitigated, unqualified, unhesitant judgment, and yet that judgment may fail to penetrate one’s entire dispositional structure. One may find oneself, against one’s will, unable to shuck old habits of thinking and reacting. One may even recognize in advance that these habits will persist, to some extent, despite one’s current sincere judgment, which rationally requires their alteration.

So maybe we should say that Juliet and Kaipeng do fully believe, but they can’t bring themselves consistently to act in accord with their beliefs?[6] But then are we elevating the occurrent sense of “believe” (if there is one) over the dispositional sense? That yields only a temporary, evaporating belief: Juliet “is believing” in the intellectual equality of the races, but the moment this virtuous thought passes, she no longer believes. What does that leave us to say about her more enduring state, which is probably of greater interest? Or maybe her dispositions to avow the equality of races trump all other dispositions in the ascription of belief? This is a dispositional approach to belief, but one built on a very narrow base. It may appeal to certain language-oriented analytic philosophers looking for a simple criterion.[7] But shouldn’t belief be seen as what animates my limbs as well as my mouth, what shows itself diversely in my actions and my reasoning and my emotional responses, not just in my affirmation or denial of a particular sentence?

Does Juliet simultaneously somehow believe both propositions – the egalitarian one and its negation – each driving the behavior that accords with it? Such a view might appeal if one is drawn to a theory on which actions and responses must always be underwritten by beliefs fully possessed. But barring that dubious theory, I see little to recommend this approach. Certainly in everyday life we do not say: Juliet fully and completely believes both that the races are intellectually equal and that they are not intellectually equal. It’s hard to know what to make of such an attribution. Maybe we can say that part of her believes one and part believes the other, but there are serious problems with taking such a division literally.

Does Juliet shift between belief in one proposition and belief in the other? Then what do we say when she is engaged in some neutral task to which race is irrelevant? How about in a single moment, when she is both having a racist reaction and thinking to herself that the races are intellectually equal?

In cases like Juliet’s, we should resist the temptation to make punctate, yes-or-no attributions. While she does not fully believe, neither will it do simply to say that she fails to believe. She is somewhere in between. Her dispositions are divided, and our attribution must be nuanced. Why should we expect, anyway, simple, punctate models of cognition always to work smoothly, to be anything other than fallible simplifications of the richly complex structure of the mind?

iv.

In conversation, John Campbell once challenged the broad-based, dispositional approach to belief I am now recommending by questioning the plausibility of the vast number of dispositions associated with any one belief suddenly changing, “all in a twinkling”, upon receipt of a small bit of new information. I responded that such change was not at all implausible: After all, a vast number of dispositions change, all in a twinkling, on any physical transformation. Pour water into a glass. Suddenly the glass reflects light differently, is less prone to tip, will more readily dent and stain the papers it sits on, will extinguish fire, will react differently to cold, will attract a thirsty person, will not safely hold a paper airplane – cut your dispositions finely enough and you’ll discover that an infinite number have changed. No law requires a separate expenditure of energy for each dispositional shift.

However, I now think that response was too simple. Sometimes all, or practically all, of the dispositions appropriate to a belief arise at once, upon formation of the corresponding judgment. When I learn that Georgia has succeeded Andy as department chair, I seem instantly to acquire the appropriate raft of dispositions. When I hear that my favorite hat company now has a website, I seem to have no trouble (dispositionally) believing this. The knowledge sticks with me, informs all my relevant actions and reactions from here forward. The dispositional shift is complete “in a twinkling”. But whether this easy relationship between judgment and belief counts as the rule or the exception may depend on the kinds of belief at issue.

Suppose I sincerely and whole-heartedly endorse a new (to me) philosophical view – that moral claims cannot literally be true or false (to take an example of a proposition I have in fact tried to believe). The arguments of Ayer and Hare win my unreserved assent; the opposing view seems mere hokum. Later, I find myself implicitly assuming that moral claims do have truth-value, both in everyday interactions and in my immediate responses to philosophical arguments. It’s not that I doubt my new position – I fully endorse it, still, when it’s recalled to me – it’s just that old, deep habits of thinking are not overthrown in a day. Do I believe that moral claims have no truth-value, then, when listening to an ethics lecture I innerly inveigh against the speaker’s “false” moral claims? People won’t misunderstand too badly if we say I do still believe. However, I think the more careful observer will refrain from the punctate attribution. My case is not so different from Juliet’s.

A homier example: Ben learns (and occurrently judges) that a bridge he normally takes to work is closed for the month and he’ll have to take a different route. The next day he finds himself on the old route, or he neglects to consider that he will be driving past the dry cleaners, or doesn’t take into account the extra commute time. In these neglectful moments, does he believe the bridge is closed? Here are two things it would be natural for him to say, in retrospect: “I knew the bridge was closed” and “I forgot the bridge was closed”. (He may lower his brow with the first statement, raise it with the second.) This juxtaposition may seem strange if you were inclined to think that once someone forgets something she no longer knows it. However we sort that issue out, though (perhaps Ben didn’t “really” forget?), I find no unequivocal ordinary-language impulse either toward ascribing or denying Ben the belief that the bridge is closed, in his moments of forgetfulness. Perhaps this sets me free to say, or even supports my saying, that he neither quite believes nor quite fails to believe. Pre-existing dispositions aligned with the old state of affairs were so pervasive and ingrained that a single act of judgment was insufficient to unseat them completely.

Many philosophers say that “S knows that P” implies “S believes that P”.[8] Yes, I have just denied that. Attachment to that hoary implication may impel you to insist that Ben believes the bridge is closed. Why, I wonder, don’t I myself feel troubled by this? Maybe the key here is in seeing knowledge as a capacity and belief as a tendency.[9] One can have the capacity without the tendency.

How do cases of swift dispositional alignment differ from the others? That’s an empirical question, but I’d guess that judgments narrow in their application, comfortable and expected, important, striking, well matched with pre-existing habits and predispositions, may be more swiftly accommodated into one’s dispositional structure than judgments broad in application, uncomfortable, unexpected, unimportant, inconsiderable, and misaligned with pre-existing habits and inclinations. This seems to me eminently worth exploring. If you want to pursue this and have some undergraduates to run, drop me a line!

But now have I committed myself to saying that people don’t believe – or don’t fully believe – things that they’re prone quickly to forget? At a party, I am introduced to Jerry. I will forget his name in a minute. But for at least a moment after we are introduced, as his “Hi, I’m Jerry!” echoes through my brain, don’t I fully and completely believe that his name is Jerry? Doesn’t Ben, for that matter, fully and completely believe that the bridge is closed the moment he learns of its closure (and then again multiple times later, when he reminds himself)? Don’t we all, indeed, form slews of micro-beliefs, forgotten in an instant, in our everyday coping with the world – about (for example) the positions of cars and hazards around us as we drive down the road?

Dispositions can be fleeting, can come and go. Twigs are fragile when frozen. Francie is prone to snap at her husband when drinking her morning coffee. For just a few seconds, my computer is in a state such that it would crash if I pressed the space bar. We may distinguish between such passing pronenesses, and the corresponding occurrences: The twig needn’t actually break, Francie needn’t snap, the computer needn’t crash. Perhaps we can say that I likewise momentarily believe that that guy’s name is Jerry, and that Julia momentarily believes that there’s a red car to the right. For a minute, our dispositions are all in line. They just don’t stay that way. What we have in these cases, perhaps, is not tardy dispositional alignment but temporary alignment: the dispositions change quickly enough, but they won’t stay put. In the case of the red car to the right, such ephemerality might be a good thing. Maybe my problem with retaining Jerry’s name is that I’m negotiating the party a bit too much like I negotiate a roadway, coping with near-term hazards and opportunities rather than cultivating long-term knowledge.

But should we dignify such momentary micro-beliefs with the term ‘belief’? Consider the “Jerry” case. Maybe we should break to some extent from (what I take to be) folk psychology and ordinary speech and deny me full belief even as I say “Hi, Jerry!” In some sense, it seems, I am not even in that moment disposed to respond to him as Jerry, attribute him that name, in a relevantly broad range of situations.

A dispositional approach to belief can allow for – indeed explain – both the belief-attributing and the belief-denying inclinations here. The key is in the interpretation of the relevant counterfactual conditionals. Consider the following disposition (proneness): If down in the hotel lobby, someone were to ask me Jerry’s name (pointing to him across the room), the right answer would strike me. Now consider: Do I have that disposition the moment I say “Hi, Jerry!” at the party? It depends on what one loads into the antecedent of the conditional. If we assume my state of mind in the lobby to be very much as it is now, then yes. “Jerry” is big in my mind now; it is big in my mind in the counterfactual situation. If, on the other hand, if we don’t hold the centrality of Jerry constant across the counterfactual situations – if we assume that my mind has returned to a more neutral, less Jerry-ish state, then no. I am not disposed correctly to recall his name in the lobby.

Since we have considerable leeway in evaluating counterfactual conditionals, we have corresponding leeway in the ascription of pronenesses.[10] I hesitate to say, then, that there is a definite fact of the matter, at the moment I greet Jerry, whether I have the panoply of dispositions necessary to underwrite genuine belief.

Let it be a choice of temperament, then, or a practical decision guided by the interests driving the ascription or denial of belief in the particular situation. For example, if this so-called “Jerry” is an impostor (Dan Dennett, say, pretending to be Jerry Fodor), and you are wondering with a colleague whether I’m onto Dennett’s tricks – if the central issue is one of truthfulness – then I see no great infelicity in simply saying that the possibility of malefaction had not crossed my mind: I believed him when he said he name was “Jerry”. On the other hand, if it’s the real Jerry in the flesh and you are principally concerned not about deception but about my long-term dispositions – about whether the belief has “sunk in” – you may wish to describe my cognitive situation in a more nuanced way.

Looking back, it seems we can similarly weigh matters of relevance and nuance, simplification versus articulation, in describing Ben’s forgetfulness of course, but also in all the cases I have sketched. Since I have mostly characterized those cases as not full belief, you will guess which way my own temperament lies. As a general matter, I think the more cautious and complete approach to belief ascription involves standing back from the moment and describing the arc of pronenesses, and the splintering reactions, as fully as is practical.

Furthermore, too ready an ascription of belief can, in some cases, be hazardous moral psychology, encouraging too sanguine a view of our virtue and self-knowledge. A simplistic approach to belief may invite Juliet to the conclusion that she is not a racist, or the unconsciously sadistic professor to the conclusion that he believes what he says he believes about the respect due to graduate students. It risks obscuring the complex relation between belief, reflection, self-knowledge, and behavior in just those morally important domains where we’re prone to sustain illusions about ourselves.

vi.

In any case, it is evident that many of our most important beliefs change only incompletely, transiently, or gradually. Sometimes, we have to work to bring our overall dispositional structure in line with our judgments. This isn’t a kind of work that many of us like, and it’s a kind of work that it may be harder to see the need of on a representationalist drop-the-belief-into-the-box model of judgment and understanding. It’s very easy, indeed very pleasant, to say “I believe in God” or “I think my marriage is worth the effort of preserving” but to live these beliefs, to shape one’s tendencies and pronenesses so that it is accurate to say that one (in any steady, meaningful way) really believes these things, is no simple, effortless matter. Genuine belief does not always flow passively from sincere judgment. Most English speakers, and most English-speaking philosophers, do not I think fully appreciate the force of this, in part due to our too-linguistic, too-avowal-oriented view of belief – a myopia reflected in and reinforced by philosophers’ unhappy tendency to refer to judgments as “occurrent beliefs”.

To profess belief in God, or the value of one’s marriage, or the unobjectionableness of death, is not entirely – perhaps not even primarily – a matter of reporting on some pre-existing inner state or expressing some fully formed belief about the world. It’s commissive, entailing a certain amount of forward-looking self-regulation. It’s partly prospective, something a speaker must work to make true, contingent in part on the speaker’s ongoing commitment to corral contrary inclinations. This commissive, prospective element can tinge the utterances with anxiety: You have to live up to them.[11]

[1] More nuanced treatments of the distinction include Price 1969; Armstrong 1973; Audi 1994. For some arguments, of course, a relatively rough and unreflective use of this distinction may be good enough.

[2] Kent Bach (1981) also rejects the idea of “occurrent belief”, contrasting belief and thinking much as I contrast belief and judging. See also, recently, Frankish 2004; Shah and Velleman 2005; and Lawlor 2006. Despite appearances, the position advocated in this paper may also be consistent with Adler (2002), if one reads Adler as meaning by “belief” more or less what I mean by “judgment” or “disposition to judge”. In fact, clearly distinguishing broad-based dispositional beliefs from occurrent judgments may make an Adler-like view more attractive.

[3] My discussion here owes much to Ryle’s (1949) nuanced treatment of the distinction between dispositions and occurrences. He emphasizes, among other things, a distinction between “capacities” and “tendencies” which is similar to, though not identical with, the distinction I draw here between pronenesses and tendencies. Fara (2005) is also helpfully nuanced on such matters.

[4] For a psychological example, we might compare fear. Consider: “Julia fears snakes”. Though it is awkward to say that Julia “is fearing” the snake she now presently sees, we can naturally say, occurrently, that she “is feeling fear” or “is afraid”. The dispositional attribution of fear of snakes to Julia may be tantamount to attributing a tendency toward short episodes of occurrent fearing. Thus, it may structurally parallel the dispositional attribution of being a runner or Bible reader. Once the snake is gone, Julia ceases feeling fear, ceases being afraid, but still fears snakes in the dispositional sense. Other attributions of fear, however, such as “Armando fears that a nuclear weapon may be detonated in a large American city within the next fifty years” may be relatively unconnected to particular episodes of feeling afraid, may be more revealed in Armando’s answers to certain questions, his decisions about where to live, his attitudes toward homeland security, etc. Attributions of this latter sort more closely parallel attributions of belief, as I see it.

[5] Few leading philosophers would accept exactly this schematic characterization – it is overly simple – but quite a few would accept something in the near vicinity. For example: Lycan 1986; Fodor 1987; Nichols and Stich 2003.

[6] I hope it is evident that I do not intend the phrase “fully believe” in this essay to imply having degree of confidence p = 1 in the Bayesian sense. I intend full belief rather as a contrast to being in an “in-between” state of belief, a state in which is it is neither entirely accurate to say one believes nor entirely accurate to say one fails to believe. Having an intermediate degree of confidence is only one way of being in an in-between state of belief; and conversely full belief may be compatible with something less that absolute certainty in one’s betting behavior. (For this reason, I think the common Bayesian practice is misleading in using the phrase “degree of belief” to refer to degree of confidence.) I discuss in-between cases of belief at some length in Schwitzgebel (2001, 2002).

[7] Carnap (1947/1956), Sellars (1969), and de Sousa (1971) are among those who express temptation toward such a view – though whether they entirely yield to the temptation is another question.

[8] For discussion of this issue, see Armstrong (1973), Lehrer (2000), and Williamson (2000). I am also tempted to describe Juliet as knowing but not believing P if, for example, P has been proven to her by well-founded studies whose results she endorses and is willing to defend. However, I agree that in neither of these cases does ordinary intuition clearly favor ascribing knowledge without belief.

[10] My emphasis on counterfactuals here shouldn’t be taken necessarily as an endorsement of a counterfactual analysis of the metaphysics of dispositions. On that issue see, for example, Prior (1985), Martin (1994), Mumford (1998), and Fara (2005).

[11] My thoughts on this topic have been much informed by conversations with Tori McGeer and Ted Preston. The views expressed (committed to?) here are not, I hope, entirely unlike the views found in some of their work (esp. McGeer 1996; McGeer and Pettit 2002; Preston submitted). I also recommend H.H. Price (1969). Thanks also to helpful conversations with Jonathan Adler, Peter Graham, David Hunter, Krista Lawlor, and Coleen Macnamara. I have explored related themes in Schwitzgebel (2002) and McGeer and Schwitzgebel (2006).