Leiter Reports: A Philosophy Blog

News and views about philosophy, the academic profession, academic freedom, intellectual culture, and other topics. The world’s most popular philosophy blog, since 2003.

  1. Daniel Kaufman's avatar
  2. AG Tanyi's avatar
  3. Noah Gordon's avatar
  4. Roger of Invisible America's avatar
  5. Luis Marxuach's avatar
  6. David Velleman's avatar
  7. Will Schofield's avatar

“Two different pictures of how we are related to the world”: Davidson vs. Anscombe

Brian Leiter

Frederick Stoutland (St. Olaf) concludes his very illuminating introduction to this new volume of essays on Anscombe by contrasting the "two pictures" of Davidson and Anscombe. He writes (p. 19):

In the Davidsonian picture, the fundamental relations human beings as such–as knowing the world and acting intentionally in it–have to the world are causal.  We know the world because it affects our senses and thus causes us to have various perceptual beliefs that constitute knowledge of the world if they are internally coherent.  We know the objects and events in the world, not directly but through their causaing beliefs in us that are about them.  Our acting in the world is similarly indirect.  We act when our beliefs and desires cause bodily movements that cause events outside our body  The movement of our fingers causes the switch to flip, which causes the light to go on, and so on.

By contrast, on Anscombe's picture,

The knowledge we gain through percetual experience is of the world directly without the mediation of beliefs:  the causal relations that underlie such experience enable it to be direct rather than constitute it as indirect.  Action is also direct.  To act is not to have one's bodily movements caused by one's beliefs and desires; it is to exercise the power to move one's body directly and intentionally.  Further, to exercise that power is not primarily to cause events outside one's body; it is to perform actions that extend beyond one's body and its movements.  Walking, running, eating, drinking, pounding, skiing, greeting, writing–ordinary bodily activities all–do not consist of bodily movements plus events they cause; they are our moving our bodies in ways that extend beyond them.  We can run or walk only on a surface, that is, only in a world outside ourselves that also acts on us….To us a hammer is not to cause it to move, and to ski is not to cause skis to move:  those are extended bodily movements.  All these bodily activities require that the bodily movements occurs as constituents of a structured activity that is more than the sum of the movements.

Professor Stoutland then sums this all up, quite aptly it seems, as follows (pp. 19, 20):

The Davidsonian picture has its roots in the Cartesian revolution, which conceived of the physical world as consisting only of what plays a role in the new physics, a physics purified of the teleological, intentional, and normative terms of Aristotelian physics….

Anscombe's conception of the physical world is not physicalist, and hence her conception of agents, which she took to be physical beings, is not simply physicalist either.  This allows the relations of agents to their world to be much closer and richer than causal relations….

 What Stoutland does not say directly, but certainly implies, is that Anscombe's world is the world of "Aristotelian physics," one rich in teleologies. 

Thoughts from readers?

Leave a Reply

Your email address will not be published. Required fields are marked *

30 responses to ““Two different pictures of how we are related to the world”: Davidson vs. Anscombe”

  1. J.S.

    The excerpts from Stoutland would seem to suggest that he endorses the suggestion that direct knowledge/action is incompatible with causation; for example, that "to exercise the power to move one's body directly and intentionally" is "not to have one's bodily movements caused by one's beliefs and desires."

    What is the sense of "direct" employed here, other than simply "not causally mediated"? Why is this notion of directness valuable? In what does that value consist, cashed out in some way other than by appeal to other hand-waving notions like "richness," "closeness," etc.?

    One would have thought that the point of, say, Davidson's anomalous monism was to explode the idea that there is in fact a valuable distinction between direct knowledge/action and causally mediated knowledge/action. (And that the full richness of the relations among agents and between agents and the world could in fact consist in the complex causal interactions among agents and between agents and the world.)

    Reply
  2. John Schwenkler

    This seems basically right, though obviously it would be a mistake to saddle Anscombe with any of the *details* of Aristotelian physics – form and matter, natural place, etc. Here it is worth noting her deep skepticism of immaterialist views of the soul, articulated in "The First Person" as well as a number of lesser-known papers. I also doubt whether Anscombe would have denied the *truth* of modern physics, except when coupled with a certain sort of claim to explanatory completeness. But she clearly believes that there is a level of explanation applicable to living things that is irreducible to and incommensurate with merely physical explanation.

    Reply
  3. Daniel Lindquist

    I've been puzzled for a while by the Anscombeans' picture of Davidson. When Stoutland says something like this: "In the Davidsonian picture…. We know the objects and events in the world, not directly but through their causing beliefs in us that are about them" I want to know what makes him call this "indirect" knowledge.

    Here's Davidson, from the preface to "Subjective, Intersubjective, Objective" p. xvi, discussing how his paper "A Coherence Theory of Truth and Knowledge" has been misunderstood: "I was so eager to get across the idea (for which I should have given credit to Wilfred Sellars) that epistemic intermediaries between the world and our beliefs are a mistake that I made it sound to many readers as though I were repudiating all serious commerce between mind and world. In truth my thesis then as now is that the connection is causal and, in the case of perception, direct."

    So, Davidson doesn't think there's anything indirect about our perceptual knowledge of the world: that knowledge is direct, with no "epistemic intermediaries" between our knowledge and its objects. There are causal intermediaries between the cube I see and the knowledge I have of it, to be sure, but there are also intermediary _stops_ if one takes a "direct" bus route across town (at stoplights and stop signs, for instance). The existence of "intermediaries" in *some* sense doesn't rule out something being "direct" in *another* sense. Our knowledge is physically indirect (like all macro-level physical happenings) and perceptually direct, says Davidson. This should've been one of the points taken away from his paper "Mental Events": we sometimes describe the physical world using non-physical vocabularies. I can't see what the problem with Davidson here is supposed to be, for the Anscombean. (There are other issues an Anscombean could raise about Davidson on perception, and McDowell has raised most of them, but _this_ part seems like it should be fine by their lights.) Why is Davidson supposed to be unable to say that "the causal relations that underlie such experience enable it to be direct rather than constitute it as indirect"?

    The case of action isn't any clearer to me. If the Anscombeans want to say that skiing is simply a bodily movement one can make in certain conditions, and not a bodily movement which causes movements in skis: Well, fine, if you want to talk that way. Suppose we rephrase the Anscombean point by saying "It's a bodily movement which can only be made while one is standing on a pair of skis" — where did Davidson ever say otherwise? What texts of Davidson are they objecting to when they talk like this? What's the point of the vocabulary shift they propose?

    I suspect that the Anscombeans are actually arguing against something/someone else, and using "Davidson" as a proxy name for the line of thought they dislike. But I can't get a good grip on either what their real target is, or what it has to do with Donald Davidson.

    On the broader issue of "world pictures": I can't see why Davidsonians can't already say the things the Anscombeans want to say, and so I don't see why they think they need to re-Aristotelianize "the physical world" to say them. And none of the Anscombeans actually wants to re-Aristotelianize _physics_, to my knowledge — surely the folk at Fermilab can keep on manipulating equations instead of talking about _kinesis_ and hylomorphism? What does having a "non-physicalist" conception of "the physical world" come to, if one isn't demanding THAT, but one is also going further than Davidson?

    Reply
  4. Dave Maier

    No doubt Daniel has beaten me to it, but since I have this written already:

    What Stoutland writes isn't exactly wrong, but I think it's misleading to see the matter as one of Cartesian physicalism vs. Aristotelian teleology, or even to see Davidson and Anscombe (or their heirs anyway) as that far apart.

    In current parlance, the key issue is that of normativity. Both Davidson and Anscombe believe that normativity is irreducible to physics. One way to make this point, seemingly following Wittgenstein (long story there!) is to distinguish sharply between reasons (norms) and causes, without at all denying the reality of the former realm. This can look like a return to dualism; so the early Davidson's "anomalous monism" tried to bridge the gap by saying that reasons can themselves be causes. My desire to eat an apple – the reason I ate it, along with the relevant beliefs – is neither physical nor mental in itself, but can be described in physical terms as an spatiotemporal event which causes another such event, e.g. my picking it from the tree and eating it.

    Stoutland is right that Davidson's philosophy "has its roots in the Cartesian tradition," and anomalous monism is indeed a bit dated in that respect. On one reading at least, however, Davidson's later thoughts about semantic interpretation and belief attribution (especially after 1986) help to overcome the residual dualism of norm and nature, again without slighting the reality of either.

    But Anscombe too has contributed to this picture, and it is surely the post-Cartesian Wittgenstein rather than the pre-Cartesian Aristotle who is more relevant here (re: the point about walking, running, etc. being "extended bodily movements" for example). On McDowell's view of Davidson's legacy (as in some respects tweaked for improvement over even the later Davidson's actual writings), nothing prevents us moderns from accepting a "direct" relation to the world, one which Anscombe rightly saw to be threatened or obscured by the outdated Cartesian picture.

    McDowell has acknowledged a co-editor of this book, Jennifer Hornsby, as a big influence in the application of these ideas to action theory, so this book looks like it should have more about that. Thanks for the tip!

    Reply
  5. Jimmy Doyle

    Certainly Anscombe denied, and Davidson asserted, that the relation between intentions, desires etc and intentional actions posited by standard (“reason-giving”) explanations of intentional actions is a causal one. But I don’t know about the rest of what Prof Stoutland says. Suppose someone is skiing across level country in the middle of a moonless night on luminous skis, and an observer who has never seen anyone skiing before says, “what’s making those glowing things move?” or (what means the same thing) “what’s causing the movement of those glowing things?” If Prof Stoutland is right about Anscombe, she would have rejected as inadmissible the answers “the man skiing, whom you can’t see,” or “the leg-movements of a man you can’t see.” This is bizarre in itself, and I don’t know of anything Anscombe wrote that implies the kind of “extended body” thesis Prof Stoutland attributes to her, which would have required her to reject these answers. Perhaps he has in mind the last paragraph of Intention §30, in which she discussed the person using a pumping mechanism to keep something at a distance level; but there her point there is psychological-epistemic: one’s attention is focused on what’s happening at a distance rather than on one’s bodily movements. There’s no suggestion that those movements don’t count as causing the thing to stay level.

    Furthermore, in her Cambridge inaugural lecture “Causality and Determination,” she wrote:

    “The word ‘cause’ itself is highly general. How does someone show he has the concept *cause*? We may wish to say: only by having such as word in his vocabulary. If so, then the manifest possession of the concept presupposes the mastery of much else in language. I mean: the word ‘cause’ can be *added* to a language in which are already represented many causal concepts. A small selection: scrape, push, wet, carry, eat, burn, knock over, keep off, squash, make (eg noises, paper boats), hurt.”

    She was very suspicious of the basically Humean account of causation that is accepted by many philosophers, but it’s clear that in calling these concepts ‘causal’ she had in mind a manifest-image notion that amounts roughly to ‘efficient’ rather than ‘final’ causality. Equally clearly, many of the concepts retain their causal character when deployed in the description of intentional action.

    Reply
  6. John

    To the extent that I grasp the contrast, it seems to more or less mirror the central contrast animating Gilbert Ryle's classic _The Concept of Mind_, which may be usefully viewed as a systematic investigation of the broadly causal — perhaps more aptly, in Ryle's phrase, "para-mechanical" — picture (as well as an attempt to advance a workable alternative). I wonder if others agree that those interested in these issues might profit from revisiting Ryle's discussion.

    Reply
  7. John Schwenkler

    Jimmy Doyle writes that on Stoutland's reading, Anscombe "would have rejected as inadmissible the answers 'the man skiing, whom you can’t see,' or 'the leg-movements of a man you can’t see.'" But that's not right. Stoutland rightly says that according to Anscombe exercising the power to act "is not *primarily* to cause events outside one's body", and that actions "do not *consist of* bodily movements [he might have added: or mental events] plus events they cause" (emphases mine). But it does not follow that there is never any truth in saying that a skier, or the movement of his legs, causes his skis to move.

    To Dave Maier's point that Davidson also rejected the reducibility of psychological concepts to physical ones: he did, but he still held that all token psychological event could be identified with token physical events. Anscombe explicitly rejects such a position: in "The Causation of Action", I believe, her objection has to do with considerations of content externalism, and as Stoutland suggests her overall metaphysics of action runs entirely counter to Davidson's. In any case, she *certainly* did not think that an action consists of a bodily movement plus a distinct mental event that is its cause.

    Reply
  8. Daniel Lindquist

    JD: "Certainly Anscombe denied, and Davidson asserted, that the relation between intentions, desires etc and intentional actions posited by standard (“reason-giving”) explanations of intentional actions is a causal one."

    Interestingly enough, I don't believe that Anscombe explicitly denied this until after Davidson had explicitly affirmed it (which was of course well after Anscombe's book had influenced him). In "Intention" she merely says that cause-talk is not helpful when trying to delineate which actions are the ones done intentionally. Here's a quote from section 5: "It will hardly be enlightening to say: in the case of the sudden start [which was the reason I knocked over a glass] the 'reason' is a cause; the topic of causality is in a state of too great confusion; all we know is that this is one of the places where we do use the word 'cause'." Here Anscombe doesn't deny that reasons can be causes; she says that we do talk that way sometimes, but that the area is too obscure for it to do philosophical heavy lifting. Very different from explicitly denying that reasons can be causes. (And presumably very relevant for what she thinks of a "Cartesian" picture of the world where all agent-world relations are "causal", which Stoutland's quote in the post has her denying. "Denial" seems to be the wrong word for how she stands to this picture: what can be denied is clear enough to be affirmed, but Anscombe seems to think it unclear what affirming this picture would mean.)

    (As a sidenote, some people have been mislead by Anscombe's discussion of "mental causes" in sections 10 and 11 of that book. A "mental cause" in Anscombe's sense is not a cause which is mental, but is "what someone would describe if he were asked the specific question: what produced this action or thought or feeling on your part: what did you see or hear or feel, or what ideas or images cropped up in your mind, and led up to it?" — This sort of phenomenological trivia was never anything Davidson cared about, and Anscombe's discussion of "mental causes" has no bearing on the his question of whether reasons can be causes. Not Anscombe's finest choice of words.)

    Reply
  9. Jimmy Doyle

    John Schwenkler:

    Suppose we are both right that Anscombe can admit as true the answers "The man skiing…" etc to the questions “What’s making those glowing things move?” etc. And suppose Stoutland and you are both right that she holds that “to ski is not to cause skis to move” (that's a quote from Stoutland). Then she would have to concede that there are at least two things the skier is doing, viz skiing and causing the skis to move (in this way). But this would be contrary to the doctrine of Intention, which implies in this case that the skier's (or his movements') causing the skis to move in this way is identical with his skiing, but conceived under a different description. Her treatment of the "concertina" case in Intention, where the pumping movements of the arm and the causing of the water to flow and the poisoning of the inhabitants are all the same action considered under different descriptions, seems to me to make it clear that she would say that the causing of the skis to move, and the skiing, are also one and the same action. Positing multiple actions here is the Goldman view she rejects in "Under a Description." (Davidson in “Actions, Reasons and Causes” agrees with Anscombe in this, given what he says about flipping the switch, turning on the light, alerting the prowler etc.) So if you and Stoutland want to give Anscombe a position according to which our answers (“The man skiing” etc) are admissible, you cannot give her a position according to which “to ski is not to cause skis to move.”

    Perhaps you and/or Stoutland meant that Anscombe’s view is that “causing skis to move” is not the *privileged* description of what the skier is doing – one that gives us what’s really going on, in a way that the simple “skiing” doesn’t. That would be right (about Anscombe, at least). But it isn’t what Stoutland said – he very much seemed to be asserting a straight nonidentity. And it would still be a bit misleading, since on Anscombe's view there usually isn't a "privileged" description at all — there are simply the descriptions (nearly always plural) under which an action is intentional, and the other descriptions (very many indeed) under which it is not. Both our descriptions here will often belong to the former class.

    Reply
  10. John Schwenkler

    It is true that the target of Anscombe's rejection of "mental causality" in /Intention/ is far narrower than the Davidsonian view, but this is likely because "Actions, Reasons, and Causes" wasn't written until after the first edition of /Intention/ had already been published. (Davidson was responding to Anscombe, not the other way around.) In Anscombe's later papers it is clear enough that she thinks Davidson's view is deeply wrong, and even in /Intention/ she takes frequent aim at views that attempt to factor intentional action into bodily movements plus some further mental component that accompanies them or brings them about.

    Reply
  11. Anne Jacobson

    The project of naming the difference between Davidson and Anscombe reminds me uncomfortably of "pinning the tale on the donkey." There's too much luck involved, and winning depends merely on how off target others are.

    Still, here goes. The venerable distinction between direct and indirect looks to be less than useful. Aristotle and Aquinas thought that perception, and indeed knowledge, depended on getting features of the world realized in one. These forms – sensory or intellectual – were, as Aquinas remarks, not what one knows, but that by which one knows. A number of scholars, perhaps starting with Yolton, have argued that philosophers who treat ideas not as introspectible items but rather something like contents are also entitled to claim their theories allow them direct knowledge. To put it in more modern language, mental representations are not what we see or know, but that by which we know. On this view, Fodor and Davidson are direct theorists.

    What distinguishes Anscombe from such theorists, I suggest, is much more a Wittgensteinian mistrust of ordinary language as providing us with an illuminating mental ontology. We can't read ontology off of logical form.

    As I remember many of her classes, Elizabeth would start with an intriguing anecdote about what she or Tamsin, one of her daughters, or indeed one of us said. The point was often enough that while the locutions were perfectly respectable, they would contravene any general theory that one might have been tempted to draw up.

    One example of hers I've used in some recent work and continue to use is this: One runs up the stairs and then stumbles. One says, "I thought there was another step." Of course, the remark is correct, but it might alse be correct to say that one had no such thought.

    Reply
  12. Jimmy Doyle

    John Schwenkler: I agree that Intention contains at least the materials for an argument against the 'causalist' picture, although assembling these materials is rather difficult. Daniel Lindquist: her remarks about the sudden start don't show that she would concede that the reasons for intentional action could generally be causes of it, because the sudden start is not an intentional action: the 'reasons' for it, whether causes or not, could not be given in answer to the special sense of the question "Why?" whose applicability is the litmus test for intentionality of action. She may seem to be denying this in §5, but her point is really that the sudden start cannot be ruled out as an intentional action at this stage of the inquiry without circularity. It becomes clear later that we can only speak indifferently of reasons and causes in borderline cases — see her remarks in §15 ("'Why did you wiggle your two forefingers by your temples?' — 'Because *he* was doing it'").

    Reply
  13. Daniel Lindquist

    JS: "even in /Intention/ she takes frequent aim at views that attempt to factor intentional action into bodily movements plus some further mental component that accompanies them or brings them about."

    This is a point fairly made, but how much does it actually hit Davidson? Davidson never wished to *factor* intentional action out into a bodily movement /plus/ something mental; this would require that we can get a grip on the "something mental" independently of understanding what someone is doing intentionally. Everything I can see in /Intention/ is arguing against an attempt to analyze intentional action as a bodily movement plus something-mental-and-independently-graspable, but this was never part of Davidson's action-theory. Part of his holism is that we can't learn what someone's beliefs and desires are without also learning under what descriptions her actions are intentional. And when he characterizes an intentional action as caused by a belief and a desire, this has to be "caused in the right way", and he despaired of clarifying what "in the right way" meant if not "in the way required for these attitudes to rationalize the action done". If you want to give an illuminating account of intentional action, you have to relate it to things like the A-E order and the accordion effect, just as Anscombe did, and not try to reduce it something "better understood" like physics. Davidson never held otherwise, as far as I can see.

    I see /Intention/ as arguing against attempts to reduce agent-explanation to a (pre-existing) species of causal-explanation, but reduction was never part of Davidson's project. Davidson wants to take on board the vast majority of what Anscombe says in /Intention/, and then add that there's no reason (given all of this) to deny that these irreducible rationalizing explanations are a species of causal explanation. And I don't see what the Anscombean reason for denying this move is.

    Reply
  14. R. Clarke

    Consider: Generally when I deliberate, make a choice, and carry out my choice, my deliberation is among the causes of my subsequent behavior.

    Would Anscombe, or Anscombians, deny this? I gather the disagreement about action is supposed to be about something further, perhaps something about the analysis or the constitution of action.

    Reply
  15. Eric Wiland

    I'm not sure why many think Anscombe is an anti-causalist. She explicitly endorsed Aquinas's view that a person's practical knowledge is the cause of the action it understands. And it's quite compatible with everything she says to think that some reasons for action are causes. She clearly thinks that there's a certain kind of mental causality at work in cases of intentional action.

    But Lindquist is also correct to note that Anscombe thought that cause-talk is not very helpful in thinking about the subject matter of Intention, and that intentional action is best understood via other entry points.

    Here's one way to think about the contrast: Davidson understood intentional action primarily in terms of what causes it, while Anscombe understand intentional action primarily in terms of the way an agent knows how the various true descriptions of her action are related to each other.

    Reply
  16. John Schwenkler

    It is pretty clear in /Intention/ that Anscombe does not think the causal relationship of practical knowledge to action is the sort of "efficient" causality that Davidson presumably takes to hold between primary reasons and actions.

    Re: Davidson's holism, compare the case of a representationalist about perception who doesn't think that the content of an experience can be specified except in terms of the things in the world that give rise to it. Here there is a denial of a certain sort of explanatory independence of mind from world. Nevertheless the representationalist endorses a thesis that a relational theorist denies: namely, that there is something "mental" that is present both in a veridical experience and an illusory one, and in the former case is generated ("in the right way") in the mind by corresponding events in the world. In the same sort of way, Davidson believes that intentional action consists in interior mental events giving rise to bodily movements. Anscombe clearly rejects this analysis.

    But no, Anscombe wouldn't deny that *when one deliberates* there is a sense in which one's deliberation causes one's behavior. She does, however, deny that we always deliberate before acting, and she would also say that the way deliberation causes behavior is nothing like the way a doctor's tapping my knee causes my leg to go up.

    Reply
  17. Anne Jacobson

    Eric Wiland, you say, 'while Anscombe understand intentional action primarily in terms of the way an agent knows how the various true descriptions of her action are related to each other.'

    That is a surprising thought since an agent might be quite ignorant of many of the true descriptions. Perhaps you meant 'some' instead of 'the various', but then the thesis is too weak.

    Perhaps you could explain whar qualification you want?

    Reply
  18. Jimmy Doyle

    Eric Wiland: I said that the argument(s) in Intention against construing reasons for action as causes is/are obscure; but the thesis is certainly there:

    "Intentional actions, then, are the ones to which the question "Why?" is given application, in a special sense which is so far explained as follows: the question has not that sense if the answer is evidence or states a cause…" (§16)

    True answers to the question "Why?" in the relevant sense are simply what reasons for action are. Anscombe cannot admit that putative reasons might ipso facto be causes because it would follow from this, along with the above quotation, that there are never any true answers to the question "Why?" in the relevant sense (in fact, the question could simply never have that sense), so that there could be no such thing as a reason for action at all, a position she certainly thought absurd:

    "People do in fact give accounts of future events in which they are some sort of agents; they do not justify these accounts by producing reasons why they should be believed but, if at all, by a different sort of reason; and these accounts are very often correct. This sort of account is called an expression of intention. It just does occur in human language" (§3).

    As I mentioned, she does think there are a lot of borderline cases; but in the non-borderline cases she is quite clear:

    "The full-blown cases are the right ones to consider in order to see the distinction between reason and cause. But it is worth noticing that what is so commonly said, that reasons and causes are everywhere sharply distinct notions, is not true" (§15).

    She is equally clear on the matter, if not more so, in "The Causation of Action."

    Reply
  19. Jimmy Doyle

    (To Eric Wiland again): Sorry, I should acknowledge that you said "it's compatible with everything she says to think that *some* reasons for action are causes." And this is technically correct, but I think it might mislead. I think it's clear from Intention that she thought the concepts of reason and cause were utterly distinct. "Some reasons for action are (also) causes" only because there are borderline cases where a factor might fall under both concepts; but insofar as it falls under the one, it doesn't fall under the other. So reasons are never causes *qua* reasons. Davidson thought that all reasons were causes *qua* reasons, because he thought a reason's status as the reason why an action was done simply is its being the reason (always construed as a mental state) that caused the action (in the right way). Anscombe's position is still quite hard-line "anti-causalist," and very far from Davidson's on the issue. Even where a reason is a cause, it doesn't have to be a mental state; eg a "mental cause" might be a knock at the door(§11)!

    Reply
  20. R. Clarke

    I don't think we'll find any action theorist–Davidsonian or not–saying that we always deliberate before acting, or that the way that deliberation causes behavior (when it does) is exactly like the way that a doctor's tapping my knee causes my leg to go up. But "nothing like"? Similarities are plentiful.

    Reply
  21. Eric Wiland

    @Anne: Yes, "some of the various" is preferable to "the various". The man in Anscombe's example knows immediately that he's moving his arm up and down because he's pumping water, and this knowledge partially constitutes the intentionality of his pumping water. The man does not necessarily also know he is thereby, e.g. casting a peculiar shadow.

    @Jimmy: Yes, I like your characterization of Anscombe's view as "So reasons are never causes *qua* reasons." But for her it's clear some reasons are indeed causes; see "Why is it that in revenge and gratitude, pity and remorse, the past event (or present situation) is a reason for action, not just a mental cause?" (Intention, sec. 14). This is not just a borderline case. But I mostly want only to dispel the notion that Anscombe thought that reasons are not causes. For her, at least sometimes they are.

    Reply
  22. walto

    What would make perception indirect are not causal, but epistemic intermediaries. Nobody ought to doubt that there are causal connections between chairs and retinae (though, admittedly, some disjunctivists do)–but the fighting can nevertheless arise over any claimed interposisition of "sense-data." It's thus confused to call a Davidson a supporter of indirect perception because he recognizes the importance of causal connections in perception.

    Reply
  23. John Schwenkler

    @ R. Clarke: Davidson thinks that the primary reason for an action is token-identical with a neural state in the brain, which causes a movement of the body in very much the same way as a doctor's tapping one's knee (or at least the firing of the nerve she stimulates) causes one's leg to move. Of course everything is similar in *some* way to everything else, but it seems fair to say that according to Anscombe the relationship between reasons and actions is indeed nothing like this.

    Reply
  24. Fred Stoutland

    On my reading of Davidson, there are no epistemic intermediaries (e.g., sense data) between our beliefs and the world. But Davidson holds that beliefs are epistemic intermediaries between our perception and the world: we perceive the world only in having perceptual beliefs about it,, and in that sense, our perception is not direct. John McDowell has made this point about Davidson a number of times, and he takes it to be a serious criticism.

    An Anscombian view would allow that there must be causal relations (of some kind) between the world and perceivers, but would deny that perception itself is constituted by having beliefs caused by the world.

    Reply
  25. R. Clarke

    @ John Schwenkler: I don't suppose the disagreement is about how neural states of the brain cause movements of one's body, is it? Perhaps it's about whether beliefs, desires, and intentions are neural states?

    By the way, causalists needn't follow Davidson with regard to what reasons are. It can be said that normative reasons are facts, such as that it's now cloudy, and proximate causes of actions are representations of these things, such as the belief that it's now cloudy. Disagreement about whether reasons are proximate causes could be disagreement about whether psychological states are causes, or it could be instead disagreement about whether reasons are psychological states.

    Reply
  26. Jimmy Doyle

    Eric: Good point. Certain motive cases will be counterexamples to my claim that on the Intention view, the more it's a reason the less it's a cause and vice versa. It can't be right to say that the more prominent the thought "he killed my father" is among the conscious mental antecedents of my killing the killer, the less it's operating as a reason! Nevertheless the *fact* that he killed my father can still be the reason even when revenge is served cold, without my having any such thoughts as I kill him.

    Reply
  27. Daniel Lindquist

    FS: "But Davidson holds that beliefs are epistemic intermediaries between our perception and the world: we perceive the world only in having perceptual beliefs about it,, and in that sense, our perception is not direct. John McDowell has made this point about Davidson a number of times, and he takes it to be a serious criticism."

    I don't think this is quite the way to put McDowell's criticism of Davidson. Davidson doesn't think that beliefs *mediate between* perception and the world; he thinks there is _nothing to_ perception but having certain sorts of beliefs (ones which are caused by the objects they are about). And McDowell's objection to this isn't that it leads to epistemic intermediaries (which would give rise to skepticism and the other ills of representationalism), but that it e.g. makes Davidson unable to distinguish between chicken-sexing and perception. Davidson's reply to this was simply to grasp the nettle: he said he couldn't see what was supposed to be the important difference between our ordinary perception and what the chicken-sexers can do. McDowell was never able to find a criticism to make of Davidson that actually convinced Davidson there was something wrong with his view of perception (so he told me when I asked him about this) — but if McDowell had charged Davidson with trafficking in epistemic intermediaries, this would have been a criticism Davidson would be sensitive to. After all, "Coherence Theory" was largely about why we must avoid all epistemic intermediaries in our conception of empirical knowledge; it's Davidson's term.

    So I don't think the way you put it can be quite right. Davidson doesn't have beliefs as intermediaries between our perceptions and the objects we perceive; "perception" as McDowell understands it simply isn't part of Davidson's picture. (He makes no allowances for a seeing which is not itself already a state to which one attaches a subjective probability.) It isn't the case that Davidson has perception in view, but mistakenly depicts it as indirect; he just doesn't have it in view. It was a blind spot.

    There is a related issue in "Mind & World" about whether Davidsonian perceptual beliefs can count as knowledge of the sort we want (where McDowell complains that Davidson's arguments from "Coherence Theory" "come too late" to help us avoid skepticism (as he says on p.17 and p.68)). Rereading Leiter's original quotations, this is probably what you had in mind all along: Davidson's account of perceptual knowledge is an account of perceptual belief, and then an additional story about how that belief can count as knowledge. Perceptual knowledge for Davidson is thus "indirect" because it's primarily belief, only secondarily knowledge; whereas for Anscombe it's primarily knowledge, and only derivatively belief. But I think it's important that McDowell grants Davidson that his arguments work insofar as they go; he just thinks they can't get us out of the oscillation M&W is directed against, because they don't attack the picture of the realm of nature as the realm of law. It's a subtle criticism of Davidson's strategy. I don't think the direct/indirect talk gets at it. (But now I suspect I'm quibbling over labels.)

    Also, Nth-ing that JD's "So reasons are never causes *qua* reasons." is a helpful way to put Anscombe's point. I think Davidson would also want to _say_ this (since he allows that there can be reasons which are *not* causes), but it's certainly not a point he thematized much, and he was not nearly as clear about it as Anscombe — and perhaps he really did miss how deep the point goes.

    Reply
  28. Jimmy Doyle

    Daniel: If we're talking about the reasons why a given action was done (what some philosophers call 'motivating' reasons), Davidson would not "want to say" that reasons are never causes qua reasons; on the contrary, he says they always are: he does not "allow that there can be [such] reasons that are not causes." For Davidson, such a reason's being a cause is *essential* to its being such a reason — its operating as a cause (in the right way) is what *makes* it a reason that 'motivates' as opposed to one that's just 'lying around'. Whereas for Anscombe, such a reason's also being a cause will be (i) relatively rare and (ii) totally irrelevant to its status as a reason why the action in question was done (its status as a 'motivating' reason). There really is a gulf here.

    Reply
  29. Daniel Lindquist

    JD: "For Davidson, such a reason's being a cause is *essential* to its being such a reason — its operating as a cause (in the right way) is what *makes* it a reason that 'motivates' as opposed to one that's just 'lying around'."

    He *does* think there are reasons that "lay around", though — and they're epistemically relevant for him. A "non-motivating reason" isn't a square circle for Davidson: They're reasons as good as any. Sometimes you want to cite a reason for a belief which wasn't a cause of you coming to hold that belief, thinks Davidson, and that's perfectly all right: it's a reason that provides support for the belief in question, even if it wasn't one that motivated the change in belief which caused you hold the belief you're currently trying to justify. This is how his picture has perceptual beliefs working: you're caused to have one by epistemically irrelevant stuff, and then you can cite various beliefs that it coheres with to give you reasons to keep holding it, and as belief is by nature veridical, holding coherent beliefs is justified, and they counts as knowledge. (I rush a bit here — I don't want to defend his account, but just get clearer on what it is.)

    I don't know offhand of any place where he discusses reasons for action which aren't causes, but (looking at his discussion of perceptual beliefs in "Coherence Theory") I don't see why he wouldn't tell a parallel story there: you can give reasons for your action which weren't among the causes of that action, and they're still perfectly good reasons, if what you want to do is make your action look like it was the right thing to do at the time. It's *not* essential to a reason being a reason that it also be a cause, for Davidson, and I don't see that he says otherwise even in the case of action.

    You're right that there is a gulf between him and Anscombe, though; I didn't mean to deny that. Anscombe doesn't think a reason's being motivating has anything to do with what it causes, but Davidson thinks it does. That's a difference. But it's not a difference where you said it was. Only *motivating* reasons are essentially causes, for Davidson.

    Reply
  30. Jimmy Doyle

    Daniel: The issue between Davidson and Anscombe is the role of reasons in the explanation of why an action was actually performed. No-one would claim that actual causation has any bearing "if what you want to do is make your action look like it was the right thing to do at the time," but now there is no longer any connection with the idea of a reason why you actually did what you did. I don't know why you say "only *motivating* reasons are essentially causes, for Davidson," as if you were pointing something out to me. That's exactly what I said. See the first sentence of my previous post, and my subsequent qualifications of the word "reason" with "such."

    Reply

Designed with WordPress