How to cite

Reply to Noordhof

by Paul M. Pietroski e-mail

Dept. of Philosophy
Skinner Building
University of Maryland
College Park, MD 20742 - USA



In his thorough and illuminating remarks, Noordhof rightly highlights several points at which Causing Actions is, at best, ‘programmatic’ and ‘suggestive rather than demonstrative’. Readers will have to judge how well the suggestions compare with the considerations others have advanced in favor of mind-brain identity theories. Here, I can only offer some clarifications concerning the details and motivations for my kind of dualism.

            I’ll structure my reply around four issues where disagreement (about something) seems to emerge: the relation between semantics and metaphysics; whether human actions include some nondeviant processes that include bodily motions; the argument from “differential vagueness;” and the relation of ceteris paribus laws to explanation. In my view, these are very good places for critics to press. Noordhof has nicely located—and initiated discussion of—topics that need more attention. But let me start with a few comments, which will help set the stage for what follows, about his opening claim that I am “being as Dualist as you can be without being tarred with Descartes’ brush.

            Part of me really likes this characterization. But alas, I’m less radical than it might suggest. For example, I think the mental globally supervenes on the physical; and I am prepared to accept other more local supervenience theses. I have a rather physicalistic conception of metaphysical possibility. I think at least many mental causes and effects are covered by (ceteris paribus) laws, and that our actions are literally caused—not merely influenced—by prior mental events, which can in turn be caused by external events. (Also, I say nothing about consciousness.) On the other hand, I do think that the particulars located in what Sellars called the “Manifest Image” are distinct from the particulars located in the “Scientific Image.” And some of what Noordhof says made me wonder if he is willing to countenance this “moderately radical” dualism as it is intended, without reinterpreting it as a hypothesized (but undermotivated) distinction among particulars located (even if we know not how) in the Scientific Image.

            To use some different terminology, my view is not that Intentional Psychology is one among many Special Sciences, where it turns out that the events covered by the laws of Intentional Psychology are distinct from the events covered by the physical-chemical-biological sciences. My view is that while we advert to mental causes in giving rationalizing explanations, these mental causes are not to be identified with causes we advert to in giving scientific explanations. Particulars in the “space of reasons” differ from particulars in the space of “arational causes” (with impersonal natures) that can be described in impersonal terms, even though all the causes in question have spatiotemporal location. This is enough of a dualism to engender reasonable scepticism. But as Noordhof notes, part of my aim is to get the proffered form of dualism onto the table as a view to be taken seriously. I also find the view plausible, all things considered. But I wouldn’t (and didn’t) pretend to have a compelling argument for that conclusion; “all things considered” assessments are hard, and our philosophical “intuitions” are often influenced by which views we have learned to take seriously. In the end, Noordhof and I may just differ with respect to some deeply held hunches about rationalizing causes. But even if adjudication is not in the offing, I welcome his forthright assessments of the complications anyone dealing with this terrain is going to face. For if it turns out that a “moderately radical” dualism is not less plausible than identity theories, that is something; and more importantly, that may in turn affect the next stage of the discussion. Jerry Fodor once said that the form of a philosophical argument is often “Let’s look over here.” He might have added that subsequent arguments can take the form “Things look OK from over here.

1.  No Metaphysical Rabbits from Semantic Hats

Since my exchange with Stout focussed mainly on actions and the semantics of action sentences, I won’t dwell again on those details. Besides, Noordhof’s real concern has nothing to do with the details. [1]

He thinks that “rarely, if ever, have semantic analyses shown the nature of worldly entities.” I agree. Given a semantics according to which an apparently true sentence has a tendentious consequence, one can (i) revise the semantic theory, or (ii) conclude that the sentence isn’t true after all. In the case of sentences like ‘Nora raised the glass’ and ‘Nora raised her hand’, I assume that (ii) is an option of last resort. But as Noordhof notes, I explicitly showed how one could revise traditional semantic analyses of such sentences so that the truth of ‘Nora raised her hand’ does not ensure that some action by Nora caused a rising of her hand. So Noordhof wants to know why the semantic considerations are relevant at all to the question of whether actions are (typically) inner causes of bodily motions.

            The reason is that in this case, I think the relevant revisions are implausible, at least if we also adopt (following Davidson and others) an “identificationist” conception of action—according to which Nora performed one action if she raised the glass by raising her hand. For me, the issue concerns the best overall fit between: an independently motivated semantics that imposes constraints on actions and the complex events they ground (if our ordinary sentences are ever true); and a cluster of Davidsonian intuitions about the actions of persons. If you don’t share the identificationist intuitions, this kind of argument won’t move you. Noordhof also notes, correctly, that neither the identificationist intuitions nor the ancillary arguments settle the matter; there is room for an alternative semantics-plus-metaphysics. This neither surprises nor bothers me. But many philosophers seem to think that traditional “volitionist” conceptions of action are nonstarters. My aim was to show that far from being incoherent, something like a traditional conception (shorn of the regressive idea that every action is caused by a prior volition) has a lot going for it. Indeed, it is relatively easy to show how this view coheres with our best semantics of action sentences and the identificationist intuitions (which I find compelling).

            I share Noordhof’s desire for arguments that establish more definitive conclusions. But I don’t think we have a prayer of settling metaphysical debates if we think about metaphysics as independent of semantics. Without at least the constraints that an explicit semantics impose on the discussion, I despair. (One needn’t be a verificationist to worry that certain “philosophical theses” have become detached from any defensible strategy for assessing their plausibility.) And the more we learn about the semantics of natural language, the more interesting—and closely connected with traditional Kantian projects—it seems to be. That’s not anything I argued for in Causing Actions; but I think it’s true. If natural language semantics was simply a matter of explicitly codifying some arbitrary conventions about how linguistic signals are related to meanings, it would (in my view) be philosophically uninteresting. Of course, even if the study of natural language tells us something about how humans think about and categorize the world we encounter, one might deny that this has any bearing on “real” metaphysics. On some days of the week, I feel like endorsing this “realist” sensibility with regard to mental causation; on other days of the week, which is when Causing Actions was written, it makes me lose track of how one is supposed to do metaphysics (if not just by doing theoretical science).[2]

            These are obviously very big questions that aren’t going to be settled here. But for better or worse, I assumed from the outset that semantics is a good prolegomenon to metaphysics. Indeed, I suspect it may be the only prolegomenon we’ve got. Though I tried to temper this view by granting allowing that a semantic theory won’t (by itself) settle the questions of interest to philosophers of mind; “all things considered” judgements require taking all things into consideration—a big job. But the idea was that an independently defensible semantics might impose some constraints that are easily (and perhaps “naturally”) satisfied by the proffered view of actions.[3]

2.  What’s Mine is Mine

A good chunk of Noordhof’s discussion is devoted to (rebutting) five ancillary reasons for adopting a Hornsby-style view of actions as inner events that can cause bodily motions—as opposed to, say, O’Shaughnessy-style processes that start with tryings and include bodily motions as parts. This is, as it seems, an intramural dispute. And there is ample opportunity for terminological confusion. But there is also an interesting issue here: when I raise my hand, is my contribution to the causal order a process that includes a motion of my hand; or is my causal contribution limited to some cause of the hand-rising? Consider, by contrast, a third kind of view according to which actions include some processes that extend beyond my body—say, to any effects within six inches of my left hand. At least some such views would simply label certain processes as actions even though they were not, in any interesting sense, the causal contributions of persons. Indeed, even the motion of my glass is not obviously part of my causal contribution—a part of what I did, as opposed to a mere occurrence—when I raise my glass. So one can wonder whether when I raise my hand, thereby effecting some change in the world, the motion of my hand is part of my distinctively personal contribution to the causal order. One can, no doubt, be too glib here: there is a sense in which raising my glass is something I did, and so no mere occurrence. But even if forms of words like ‘what I did’ admit of multiple interpretations, it hardly follows that nothing is left of the original volitionist intuitions.

            Noordhof’s question here is closely related to the bigger question of whether the “grounders” of the relevant processes are themselves biochemical events. For if all inner causes of bodily motions are biochemical events, there is less reason to focus on some particular inner cause C of a bodily motion B, as opposed to a process that includes both C and B. But if C is not a biochemical event, one might wonder if any “process” that includes both C and B isn’t really a hodgepodge—a mixture of some personal causal contribution with an impersonal effect. For just this reason, Noordhof is right to press the issue. But it’s not that my views in chapter one were settled first, before moving on to chapter five. The aim was to provide a plausible semantics-plus-metaphysics of actions-and-action-sentences.                                                                   

            In chapter five, I said that my actions are mine in a way that my heartbeats are not, and that bodily motions seemed too much like my heartbeats for comfort. I didn’t speak in terms of control; but I should have said more to distance myself from arguments about control, if only because there are (as Noordhof rightly notes) multiple senses of ‘control’. [4] I would, however, note that Noordhof is led to say that “When I try, I am in the hands of my brain.” To my ear, this sounds like an admission that tryings aren’t the (personal) causal contributions of persons. Agents qua agents seem to vanish. I agree that “I couldn’t try unless my brain were a certain way.” But other things equal, I’d like a metaphysical picture that makes room for a distinction between the correct second claim and the objectionable first claim.

            Noordhof then ascribes to me the view that “actions are things we explain completely in terms of an agent’s beliefs and desires.” But this isn’t quite my claim. (In fact, I try to avoid talk of ‘complete’ explanations.) I said that the distinctively psychological component of explanation remains the same, whether we explain why someone lifted a rock or merely why they tried to lift the rock. This is a prima facie reason for thinking that that the primary object of intentional explanation is something—an action/trying—that falls short of a bodily motion.

            Noordhof is absolutely right, however, that I tout the following fact: Hornsby and I can describe certain familiar cases without the usual appeals to an unanalyzed notion of “deviant causal chains.” Suppose Bloggs wants to spill the soup on a customer, in order to create a distraction, and gets so nervous that he stumbles and spills the soup on the customer. There is a long history of attempts to say what it is for a bodily motion B, which causes the transfer of soup from bowl to customer, to be causally related “in the right way” to Bloggs’ prior mental events. But if actions are inner events of trying, one doesn’t have to play that game (which, to the best of my knowledge, nobody has won). Since Bloggs never got around to trying to spill the soup, the transfer of soup from bowl to customer is simply a causal consequence of carrying the soup (clumsily). Bloggs successfully tried to carry the soup towards the customer; his action had a further effect that we can use to describe the action; and there was a process, grounded by his action, that ended with soup-motion. But an event of spilling the soup need not be grounded by an event of trying to spill the soup, anymore than an event of waking the baby has to be grounded by an event of trying to wake the baby. If you slammed the door, not knowing that the baby sleeping in the next room, you may well have woken the baby. In Bloggs’ case, the relevant consequence was an event of an intended type; but the consequence was not caused by a trying with the right intentional content. Or so Hornsby and I can say, without saying that causal chain connecting Bloggs’ bodily motion to his trying (to carry the soup) was somehow deviant compared with other causal chains.

            Noordhof thinks that philosophers will need an account of “deviance” for independent reasons. But I think he is begging the question by speaking of “the problem” of deviant causal chains. There are facts to be accounted for—including facts about Bloggs-type cases—and talk of deviant/nondeviant causal chains is one way of gesturing at those facts. But one can’t tell in advance what the best way of describing and explaining the facts will be. Here are three possibilities concerning Bloggs: actions are bodily motions causally related in the right way to belief and desires, but Bloggs’ bodily motions were not caused in the right way; actions are inner tryings, and Bloggs didn’t try to spill the soup (so in that sense, his bodily motions were not caused in the intended way); actions are processes that include bodily motions causally related to inner tryings in the right way, but Bloggs’ bodily motions were not caused by any inner trying in the right way. I favor the second view. My primary target in giving the argument was the first view, but the third view also posits a distinction (between kinds of causal chains) that has proven very hard to make out. [5]

            Noordhof’s counterclaim seems to be as follows: a certain range of facts concerning perception and action will require, for their explanation, a substantive account of deviance that will also apply to Bloggs-type cases. But this is an assumption. Perhaps Noordhof thinks that a “no satisfactory solution has been found [despite serious efforts, in linguistics and philosophy, by clever motivated folks]” style of argument is so weak that it does not even motivate alternative approaches to a problem. If so, then I disagree. Success in philosophy often takes the form of finding a way around (rather than through) a stumbling block. The style of argument is familiar enough in science: no satisfactory solution to the phlogiston problem has been found; so let’s try to find another way of describing and explaining the facts noticed by phlogiston-theorists.

            The discussion of successful/unsuccessful actions left me puzzled. I’m not denying that you can amuse someone by speaking in a strangulated staccato (without trying to amuse them in that way), or that you can unintentionally speak in a strangulated staccato. So I’m not sure what the problem for me is supposed to be. The Hornsby quote inserted at that point is not one I would—or need to—endorse. And what Noordhof calls my “fourth reason” I would view as another way of getting at the first point: what distinguishes my contribution to the causal order isn’t just causal history. I agree that this remains an intuition in need of further theoretical defense, not a datum, and that it once again raises questions about whether some processes count as actions.

            Noordhof has a quite interesting discussion of the muscle-contraction puzzle. It seems like the right sort of thing to say if you want to resist the Hornsby-style view; whether it works or not is a hard question I propose to set aside for present purposes. The bottom line, from my point of view, is the one Noordhof suggests: “Hornsby and Pietroski might argue that their theory is simpler and cleaner than the suggestion I have put forward.” Right. It’s a prima facie consideration in favor of the Hornsby-style view that it makes the fist-clenching “puzzle” disappear so easily: if some theory plus the fact that humans have muscles makes it look like there is backwards causation, check the theory; the view I endorse doesn’t require sophisticated elaboration to avoid the unwanted consequence

3. Avoiding Arbitrariness

I turn now to what Noordhof calls my “official justification and rationale” for event dualism—the argument from differential vagueness. But let me say at once: to the best of my knowledge, I have never had an “official” justification or rationale for anything. With respect to the philosophy of mind, I find myself with a cluster of broadly (though not entirely) Cartesian intuitions that seem to hang together and point in a certain direction. Of course, one can and should support intuitions with as many arguments as one can muster. Though one should also remember that philosophy (like science) is often less about establishing that something is true and more about explaining how some puzzling truths could be true. And as I said in chapter five, the point of offering the arguments there was largely to indicate the position that interests me, rather than to command assent from those not antecedently disposed towards the conclusion. I discussed the Hornsby/Benaceraff-inspired argument at some length, because it strikes me as interesting (and not easily dismissed as a reductio of its premises). And I welcome the opportunity to discuss it further. But Causing Actions is more an attempt to show how there could be mental causation, given event dualism, than an attempt to convince skeptics of a dualistic metaphysics; although the latter surely requires the former.[6]

            In its baldest form, the argument from differential vagueness is as follows: the scale at which mental events exhibit vagueness of spatiotemporal location is different than the scale at which biochemical events exhibit vagueness of spatiotemporal location; so mental events are not biochemical events. The idea is that a mental event like that of Nora’s trying to raise her right hand has a spatiotemporal location; it is (somewhere) where Nora is. But there will be spatiotemporal regions such that it is indeterminate whether the event of Nora’s trying occurred (wholly) in those regions. While similar remarks apply to biochemical events occurring inside of Nora, it seems that for any given biochemical event B, there will be at least some spatiotemporal regions such that: B determinately occurred in those regions, while Nora’s trying did not determinately occur in those regions.

            Perhaps offering this argument commits me, one way or another, to substantive views about vagueness. But it wasn’t my intention to presuppose anything in particular about where vagueness comes from—if only because I think the relations between logic, natural language semantics, and human concepts remain obscure. [7] The idea was to highlight the following thought: just as one can give a “too specific” answer to the question of where a hurricane occurred, so one can give a too specific answer to the question of where a mental event occurred. And I suggested that one could always (correctly) describe the location of a biochemical event in a way that would be too specific as an answer to the question of where Nora’s trying occurred. This suggestion is, of course, hardly decisive. Perhaps unbeknownst to us, each mental event is a biochemical event. But it would be question-begging to assume this in reply to an argument against neuralism—the very “token identity” thesis in question. The issue is not merely whether mental events have more precise boundaries than we can currently specify. It is whether (it is plausible, given current evidence, that) each mental event has the precise boundaries of some biochemical event. Reasonable people can differ in how they assess this plausibility. But it strikes me as a worthwhile consideration to have in the mix; and my own view is that it tells, prima facie, against neuralism.

            On the other hand, one might say that (so long as the spatiotemporal properties of a mental event are not intrinsically vague) there are sure to be biochemical events whose spatiotemporal properties are at least “in the right ballpark” for purposes of an identity theory. But the worry is that too many biochemical events will be “in the right ballpark;” so all of these equally good candidates will be equally bad candidates. It is tempting to think that this is a perfectly general issue that arises for all “macro” events. There will be many equally good (and so equally bad) candidates for the event, described in physical terms, that is some hurricane. But with regard to hurricanes, a little arbitrary precisification makes no difference to our explanatory interests. It doesn’t matter which event, within a certain range of events, is the hurricane; one can say, without loss, that there is no fact of the matter as to exactly which physical event the hurricane is. I claimed that this move isn’t available with respect to actions, at least if these are taken to be the (personal) causal contributions of persons. The guiding intuition here is that the relevant question—which event is my causal contribution?—can’t be settled by a third-person arbitrary decision. [8]

            Noordhof comes close to putting it this way at the end of his section three. But it’s not just the personal perspective I want. Davidson provides that. I’m interested in a picture according to which human actions are (by their nature) personal contributions—and not just causes that interpreters can “describe as” the contributions of persons. I’m interested in a picture according to which human actions are exercises of freedom—and not just events that observers can “think of” as exercises of freedom. I don’t deny that one can build the personal perspective into an interpretationalist conception of the mental. But since this invites the worry that some events are just labelled as mental, either for certain pragmatic purposes or simply as a sop to our self-conception, I want to explore a position that clearly isn’t just labelling certain physical events as mental—even at the cost of needing an alternative account of mental causation. Perhaps the fact that some event is my action is, after all, a fact about how my behavior can be interpreted. But surely there is intuitive motivation for thinking about things the other way round: my behavior is sometimes (correctly) explicable in rationalizing terms, because the causes of my bodily motions include actions that are mine in some independent sense. I agree that this sense wants more spelling out; and perhaps, as Noordhof says, Davidson is “clearer in his spelling out of the commitments of the personal perspective.” But one wants plausibility as well as clarity. The worry is that Davidson provides too much perspective and not enough mental reality.

            By way of contrast, Fodor offers a view according to which a person has (or fails to have) mental events quite independently of how her body moves—and how she is interpreted by others. But I am hardly the first to worry that Fodor’s naturalization of the mental threatens the personal perspective; although as I briefly discussed in the appendix, these issues are bound up with hard questions about mental content. Maybe this all shows that you can’t have the kind of personal perspective I want without Davidsonian interpretationalism or Cartesian dualism. But the idea is that one can have a dualism of Manifest and Scientific Images, even at the level of particulars (pace Davidson and Fodor), without lapsing into Descartes’ view. Once this position is in view, I think its initial attractions are sort of obvious. The question is whether the obvious difficulties are fatal—and if not, how plausible the view is, all things considered.

4. Covering Laws, Causal Explanations

Finally, let me turn to Noordhof’s remarks about ceteris paribus laws (henceforth, ‘cp-laws’).

            One could, I suppose, just take as given that explanation via cp-laws is legitimate. Instead, I offered an argument of sorts: lots of apparently legitimate explanations seem to work via subsumption under cp-laws; and the prima facie concerns about such laws can be dispelled, given a plausible sufficient condition for nonvacuity. But as explicitly stated, my aim was not to reduce cp-laws to (or analyze ‘cp-laws’ in terms of) anything else. In the end, I think that ‘cp(F —> G)’ is true iff it is a cp-law that if F then G. Unfortunately, this doesn’t address the concerns that appeals to cp-laws invites; so one says what one can to alleviate the concerns. But when Noordhof asks “What stops there being a ceteris paribus law relating introductions of drag with failure of the length to be different,” I’m afraid my answer must be boring. The world is what’s responsible for which cp-laws there are. (Sources of drag are just not, so far as we can tell, nomically related to failures of pendulum-lengths to change.) My account wasn’t intended as a metaphysical constraint that rules out false putative laws. Nor was I trying to provide a rule for determining which cp-laws obtain, given the totality of facts described without any reference to nomic relations. I was trying to show how cp-laws can be nonvacuous and explanatory. Noordhof says that “To avoid a vicious regress, Pietroski must deny that we should understand the explanatory character of at least one relevant interferer in terms of ceteris paribus laws.” But I don’t think that’s quite right. To avoid a vicious regress, I need the regress (if there is any) not to be vicious; and whether a regress counts as vicious depends on what you’re trying to accomplish. I don’t see how Noordhof’s remarks undermine my attempt to show that (despite the worries) at least some cp-laws are OK after all.

            Noordhof also thinks that “the fact that there is more than one way in which the parents may be caused to have blue eyes should not be part of the basis for a decision about whether the parents having blue eyes is a cause of their children having blue eyes.” I’m not sure what to say about this, since I don’t know what ‘should’ means in this context. But in the trivial case I offered, concerning cosmetic eye-color change, it would be crazy to treat being blue-eyed as a cause of the kids being (born) blue-eyed; and I would have thought that other cases have to be assessed as they come up. Noordhof doesn’t, however, offer what he takes to be a counterexample to any sufficient condition I advanced.[9]

            For me, the major point of concern lies with Noordhof’s claim that “We seem happy to allow ceteris paribus laws without causality between antecedent an consequent.” One can, of course, define ‘cp-law’ as one likes. But I don’t know how to avoid charges of vacuity if someone says that (i) Fs are accompanied by Gs, as a general pattern, but (ii) sometimes you get Fs without Gs. This isn’t to say that the charges can’t be rebutted; but I don’t know how to do it. More importantly, talk of patterns (without corresponding talk of causality that is not reducible to patterns) invites the worry—a worry I tried to acknowledge and dispel in chapter seven—that my proposed account of mental causation relies on an objectionably thin “regularity” conception of causation. There is surely something right about the idea that we can explain events by locating them in patterns of regularity. But if one wants to talk about nonvacuous cp-laws, and say that subsumption under cp-laws is explanatory, and say that events related by mentalistic cp-laws are related as cause to effect, then one had better be working with a relatively demanding notion of cp-law according to which not every pattern counts as a cp-law.

            Noordhof seems to think that by restricting attention to cp-laws that are plausible causal laws, “Philosophical concerns are forcing us to distort scientific practice by providing an account of ceteris paribus laws that has little plausibility outside of its capacity to assuage philosophical concerns.” But surely it is part of scientific practice to care about causal explanations—and not just “noting certain patterns either for the purposes of rough prediction or because we want to investigate them further.” If there are scientific generalizations that don’t fit my characterization of cp-laws, so be it. It wouldn’t follow that no scientific generalizations do fit the characterization. I discussed cases from physics, chemistry, and biology; but Noordhof doesn’t say how the proposal distorts those examples. I also discussed cases where scientists apparently appeal to cp-laws that are almost always violated—and necessarily so, given physical laws—making it hard to see how talk of “patterns” gets a grip (and cases where the violation of a cp-law is due to a “one time only” occurrence like the Big Bang).

            Noordhof thinks the problem is that my approach “ignores the vital question of the relationship between the events cited in the antecedent of a ceteris paribus law and the events mentioned in lower level laws governing the micro level.” In one sense, this seems a little unfair given my discussion of supervenience (and briefer discussion of mechanism) in chapter six. But if we are focussing on the application to mentalistic ceteris paribus laws, which was the point of introducing a discussion of laws into Causing Actions, then I wonder if Noordhof is (illicitly from my point of view) assuming that mentalistic cp-laws are just one more species of cp-law that we discover within the Scientific Image. I’m not saying that mental events are events related to “underlying causes” in much the same way that being born with blue eyes is related to underlying biochemical causes—and then adding the claim that mental events are not identical with their biochemical realizers. I’m suggesting that at least some mental explanations are not Scientific Image explanations, even though they are causal explanations. Chapter seven is, in part, about how the notion of a “ceteris paribus law” can still figure in the Manifest Image in a way that provides a kind of unity—letting us begin to see how we can have these two Images of one world—without forcing identifications at the level of particular events. Do I pull it off? Maybe not. Surely not in any detailed and fully satisfying way. But one can’t object by noting that given my kind of dualism, mental events are not realized by neurophysiological events in the same way that (say) neurophysiological events are realized by biochemical events. That’s more or less the view.

            Perhaps event dualists should be embarrassed by the fact that the scope of the Manifest Image keeps shrinking the more we learn about the world in Scientific Image terms. But that’s a different concern, which I also tried to address—though not in any detailed and fully satisfying way. Most of the book is devoted to providing an account of mental causation that does not rely on token physicalism. The positive arguments for adopting this account are indeed ‘programmatic’ and ‘suggestive rather than demonstrative’. (Is the situation with regard to identity theories any different?) Building on the work of many other philosophers, including Davidson and Fodor, I went in a certain direction. And I’m saying that things look OK from over here.


1_ There are some technical issues concerning Noordhof’s (1A) and (2A). Verbs like ‘shot’ raise additional complications, since there is a passive reading of ‘He was shot’. And it is important for me that ‘Booth shot lincoln’ is true only if some complex event of shooting has both an Agent and a Patient.

2_ While Strawson might not care for the details of my semantics or metaphysics, or my way of connecting them, my views have been influenced (as Chapters 5 and 7 suggest) by his studies in ‘descriptive metaphysics’. Saying just how the semantics of natural language connects up with descriptive metaphysics—apart from saying that philosophers still have much to learn from Chomsky—is a project for another day.

3_ Noordhof also suggests that my discussion of how Fregeans can respect “semantic innocence” is unneeded. But at least initially, it’s hard to see how the semantic value of ‘that Fido barked’ could be the Fregean sense of ‘Fido barked’ if the semantic value of ‘Fido’ in ‘that Fido barked’ is not itself a sense; it’s implausible that the semantic value of ‘Fido’ in ‘that Fido barked’ is a sense; and at least initially, it’s hard to see how my claims about explanation could be correct if the semantic value of ‘that Fido barked’ is not (something like) a Fregean sense. So I offered an account that showed how the semantic value of ‘that Fido barked’ could be the Fregean sense of ‘Fido barked’ even if the semantic value of ‘Fido’ is always the dog Fido. Perhaps Noordhof thinks that one doesn’t have to worry about such matters when doing philosophy of mind. But one has to worry about them eventually; and in any case, I can hardly dismiss certain puzzles as “merely” semantic, given my views about semantics and metaphysics.

4_ I don’t agree that his second kind of control is “just as plausible” for purposes of understanding action as the first; and I think there may be some confusion about content externalism lurking in the passage from Hornsby he cites. But I agree that this just restates the initial intuitions. The idea that tryings are causally dependent on brain states bothers me, since I don’t think of supervenience as a species of causal dependence. But that’s another issue.

5_ I wonder if Noordhof is thinking of a related—but in my view distinct—issue: when a person acts, we can describe their action in terms of (and ascribe responsibility for) some but not all of its effects. One can unintentionally break a vase; but there is a difference between breaking a vase (unintentionally or not) and merely causing the vase to break. So at some point, we need to say what distinguishes a genuine process of breaking a vase from a mere “event fusion” whose parts include both an inner action and a vase-breaking. Talk of deviance might find a place here; though I would bet that this has as much to do with semantics and/or morality than with mind-independent causal relations. In any case, this distinction is not the one relevant to Bloggs, who did spill the soup; the transfer of soup from bowl to customer was, in this sense, a nondeviant effect of his mental events. And I see no reason to think that some parametizable all-purpose notion of deviance will cover all these issues.

6_ What makes me take event monism seriously is not the metaphysical arguments for it, but that it provides an account of how there could be mental causation. If mental events just are biochemical events, then it’s not hard to see how they can be causes of bodily motions. Absent an alternative account, of the sort event dualists can embrace, there’s little point in trying to assess the relative plausibility of event dualism; event monists have an “only game in town” argument that would, for me, swamp all others. But if there are two possible accounts of how there could be mental causation, we can start assessing for relative (overall) plausibility. Noordhof makes some suggestive remarks about “emergence;” though for better or worse, I have never understood how talk of “emergence” provides a middle way between token identity theories (which face Leibniz Law objections) and event dualism (which faces overdetermination objections). It would be nice if we could avoid both the difficulties that face monists and those that face dualists; but I don’t see how.

7_ I am indeed suspicious of supervaluationist semantic theories. But I don’t reject such theories because they lead to arbitrary precisification that would “detract from the goal of personal ascriptions.” My suspicions, which I won’t go into here, stem from facts about natural languages. (I think Mark Sainsbury’s talk of “concepts without boundaries” provides a better starting point for discussion.)

8_ Which is why I set the discussion up in terms of subjectivity (and references to Nagel). This kind of circularity makes the case for event dualism weaker than a case consisting of several compelling arguments based on independent considerations. But how many independent arguments are there for event monism?

9_ Saying why “having the wrong genetic material” doesn’t count as an interfering factor would require restating a large part of chapter four—and in particular the discussion of why, given a coin that came up heads, it need not be a cp-law that the coin comes up heads when tossed. But the underlying idea is that appeals to interfering factors come along with assumptions, which can turn out to be wrong, about what counts as a “Normal” case. (If I understand Noordhof’s question correctly, it is parallel to the question of why “having too little mass” doesn’t count as an interfering factor with regard to explaining why charged particles don’t comform to the law of gravity. My hunch is that one can’t sustain the idea that there is some “Normal” amount of mass for “things” to have.)

top | back to symposium index | Noordhof's paper