onegoodmove

The primary reason philosophers have been interested in folk psychology is the prospect of it becoming integrated with mature psychology. Philosophers such as Jerry Fodor (1987) think that folk psychology works so well that it must be for the most part true. On the other hand philosopher's such as Paul Churchland (1981) think that folk psychology is a degenerating theory which so radically false that its ontology will be eliminated in the same way phlogiston, caloric, and crystal spheres were. Ramsey, Stich, and Garon (1990) argue that the prospects of folk psychology don't look good in the face of the advent of connectionism. Both eliminativist arguments depend on folk psychology being construed as a theory. Though this construal has been contested (Wilkes 1991a; 1991b) I will assume that folk psychology is properly viewed as a theory. Both sides of this debate also share many common assumptions about the commitments of folk psychology.

There are four commitments that are generally attributed to folk psychology. I have found that all sides generally suppose the following

1. Realism- Beliefs, desires and other intentional states are real. They are not merely instrumental terms for explanation.

2. Propositional Attitude Analysis- It has been noted that sentences containing our folk psychology verbs are often constructed with an embedded that clause. This clause is claimed always to be a proposition. For example, he believes

that the movie is at 7:00 pm and he desires that the movie is a good one. On this analysis our folk psychological verbs are attitudes one can take to various propositions.

3. Discreteness- Folk psychological mental states (FP-states) are discrete causal entities. According to discreteness it makes sense that we could add or subtract one FP-state without effecting any of our other FP-states.

4. Causal efficacy- FP-states are entities in the head that engage in physical interactions and so produce other FP-states and external behaviours.

The last three of these commitments combine to form what Ramsey, Stich, and Garon (1990) have termed "propositional modularity." They rely on folk psychology's purported commitment to propositional modularity to reach their eliminativist conclusion. Their conclusion is conditional. "If connectionist hypotheses of the sort we will sketch turn out to be right, so too will eliminativism about propositional attitudes" (Ramsey, Stich and Garon 1990, p. 316).^[1] The reason they come to this conclusion is that representations in connectionist models of the sort they have in mind are highly distributed. Connectionist networks are made up of a number of abstract nodes that have weighted connections. The same mode may be a part of more than one representation in such a network. If folk psychology is committed to discrete causal propositional attitudes, it is argued, then it is inconsistent with such a networks.

My purpose in this paper will be to argue that folk psychology is not committed to propositional modularity. I will begin in section two by saying more about what folk psychology is by way of investigating just what should be included under the label and the scope of folk psychological explanation. In section three I will argue that folk psychology is not committed to discreteness, that it is only weakly committed to causality and that it is not committed to any particular causal mechanism. Finally, in section four I will argue that the propositional attitude analysis of FP-states is overly restrictive and would thus eliminate from the label folk psychology many types of explanation that the folk in fact use. A general theme of the paper will be that if there are a number of viable empirical alternatives for any feature of folk psychology, then folk psychology will generally not be committed to any particular alternative.

2. The Scope of Folk Psychology

In this section I will argue for a broad picture of folk psychology. That is to say that since folk psychology has been around for thousands of years it is a very rich and diverse body of knowledge and practices. It has traditionally been the case that philosophers have focused vary narrowly on just beliefs. At the very most they talk about "belief/desire" psychology with the very slightest nod to any other psychological verb by way of providing a list. This has been to our detriment.

Besides beliefs and desires, perceptual reports are a part of folk psychology. The use of the five basic perceptual verbs, see, hear, smell, taste, and feel are central to the laymen's understanding of his fellow human being. Suppose a postman suddenly runs and climbs up a fence. We might explain this in the following way. The postman saw a big dog with foam at its mouth, believed that the dog was dangerous, and climbed the fence because he wanted

to avoid getting bitten. Perceptual verbs are used all the time in folk psychology explanations and it is clear that folk psychology must have them.

Another essential part of folk psychology is emotions. It is astounding that there has been so little investigation of emotions in the philosophical literature considering what a central role they take in our mental lives. Not a days goes by that we don't explain something using terms such as happy, sad, lonely, etc. In order to avoid my own criticism about merely paying lip service to these terms by listing them I should say something more. I will do this in section four by considering whether emotion terms can be analysed properly in the propositional attitude format.

Philosopher's failure to consider the fact that perception and emotions, not to mention know-how which I will discuss in section four, are a part of folk psychology has misled them into thinking folk psychology has various commitments which it does not. It is this narrow focus on belief that has led philosophers to think that FP-verbs must be individuated in terms of the propositions that they are an attitude towards. This issue is discussed in more detail in section four.

It is also important to consider what the explanatory scope of folk psychology. Paul Churchland (1981) has argued that folk psychology is a degenerating research programme because it suffers from massive explanatory failures. Churchland points out that folk psychology fails to explain mental illness, the faculty of creative imagination, intelligence differences between individuals, the psychological functions of sleep, the ability to catch a fly ball on the run, the internal construction of a 3D visual image, visual illusions, the nature of memory, and the nature of learning itself. As Botterill and Carruthers (1999, p.42) remark, "a failure to explain - where there is no serious attempt at explanation - is not the same as an explanatory failure." The question becomes which of these features of psychology does folk psychology attempt to explain.

It is quite plausible that folk psychology attempts to explain mental illness, at least in part. Generally folk psychology tries to explain such things in terms of inaccurate beliefs and failure to meet certain normative standards of rationality. Though these explanations are rather shallow and unsatisfactory, they are not failures to explain for all that. Intelligence differences between individuals is often explained in terms of having more knowledge. The more intelligent person knows more thinking techniques. Perhaps he has taken a class in critical thinking for example. According to folk psychology a person is more intelligent than another because has a greater knowledge base from which to make well thought out decisions. Again, this seems unsatisfactory, but it is an attempt to explain. On the other hand it does appear to be the case that folk psychology does not even attempt to explain any of the other cognitive features on Churchland's list. Though folk psychology does seem to have some significant explanatory failures it seems a bit excessive to say that it suffers from "massive" explanatory failure.

What has become clear from this discussion of the scope of folk psychology is that it is a much broader base of knowledge than the narrow conceptions philosophers have worked with hitherto. Folk psychology is not just belief/desire psychology and its explanatory scope is quite broad.

3. Discreteness and Causal Efficacy

In this section I will consider the purported commitments to the discreteness and the causal efficacy of folk psychological mental states. This will become apparent later in the discussion. Functional discreteness is most famously formulated by Ramsey, Stich and Garon (1990). There are two interesting reasons given by Ramsey et al to suggest that FP-states must be discrete. One is that it seems plausible that one could gain or lose one belief without effecting any of one's other beliefs. The second is the idea that it should be possible to ascertain which particular beliefs and desires are responsible for a particular action. This second reason pulls double duty as an argument for folk psychology's commitment to causal efficacy as well as discreteness. I will discuss both reasons in turn and conclude this section by arguing that folk psychology is not committed to any particular internal causal mechanism.

3.1 Discrete addition and subtraction

Ramsey et al write "Perhaps the most obvious way to bring out folk psychology's commitment to the thesis that propositional attitudes are functionally discrete states is to note that it typically makes perfectly good sense to claim that a person has acquired (or lost) a single memory or belief." For example I might forget that my book is on the nightstand in my bedroom and this would not effect any of my other beliefs or desires. Here discreteness means something like unattached to other beliefs, desires, and other FP-states. Another way Ramsey et al try to show that folk psychology is committed to functional discreteness is to show how one proposition could be removed from an old style semantic network such as Collins & Quillian (1972). The semantic network they have encodes the following propositions: 1. Dogs have fur 2. Dogs have paws 3. Cats have fur 4. Cats have paws. There are a set of 5 nodes in this network: Dogs, Paws, Fur, Cats, Have. These nodes are connected to one another by units that take two objects and a relation. In this simple network there is only one relation node, namely "have", and 4 objects. Ramsey et al show that you can simply remove the unit that takes "cats" and "paws" as its objects and "have" as its relation to remove the proposition "cats have paws" from the network. This is done while leaving the other propositions untouched. The idea behind this example is to show before the advent of connectionist networks various ways of modelling memory shared the intuition that you can gain or lose a FP state one at a time without effecting the rest of the system's beliefs and desires. The question here is whether such a network adequately models FP intuitions about beliefs and desires. The question is not whether one can add beliefs and desires one at a time, which is surely true, but whether one can do so without affecting system wide changes.

Daniel Dennett (1991) argues that FP-states must be treated more holistically. Dennett asks us to suppose that we explain Mary's running up the stairs by citing her belief that she left her purse on the bed and her desire to take her purse with her when she goes out that day. He points out that at any one instant Mary might believe that her purse is on a flat surface that you sleep on, or that he purse is in the room that she slept in every night for a weak etc. If folk psychology is committed to discreteness then there most be a fact of the matter about which of these beliefs have to light up and team up in order to cause Mary's action. Dennett argues, "Folk psychology recognizes, if you like, the holism of belief attribution in everyday life, and in fact boggles at the suggestion that somebody could believe that her handbag was on the bed and not believe any of these other propositions." (Dennett 1991, p. 142) This is a very important point. The idea is that somebody cannot just acquire the belief that there handbag is on the bed without making changes to the rest of their beliefs. If this is the case, then FP-states are not unattached to one another and thus are not discrete in this sense.

Let us take a look at Ramsey et al's own example. They ask us to consider a man who has forgotten that his keys are in the refrigerator and is thus looking for them elsewhere. They argue, "it is perfectly possible for a person to have a long standing belief that the keys are in the refrigerator, and to continue searching for them in the bedroom." (Ramsey, Stich & Garon 1990, p. 320) The idea here is that a person could be absent minded. This is a strange example. Leaving your keys in the refrigerator is not something one normally has long standing beliefs about. Suppose for the sake of argument that this person regularly puts their keys in the refrigerator and did so the previous week. Now they believe the keys are in the refrigerator for a significant amount of time. One day in a rush this person looks frantically in their bedroom for the keys despite the fact that they have a long standing belief that the keys are in the refrigerator. What this is supposed to show is that the person has a belief that their keys are in the refrigerator, but that this need not effect any of their other beliefs. Thus, they conclude that beliefs are discrete and separable. This does not meet Dennett's objection. It may well be the case that this absent minded person could search for there keys in the bedroom despite believing they are in the refrigerator. However, this person could not believe that their keys are in the refrigerator and not also believe that the keys are in the device that keeps the food cold, or that they keys are in the kitchen because the refrigerator is in the kitchen. Beliefs tend to travel together in groups.

Ramsey et al have a response of sorts to the above criticisms.

In saying that folk psychology views beliefs as the sorts of things that can be acquired or lost one at a time, we do not mean to be denying that having any particular belief may not presuppose a substantial network of related beliefs. The belief that the car key are in the refrigerator is not one that could be acquired by a primitive tribesman who knew nothing about cars, keys or refrigerators. But once the relevant background is in place, as we may suppose it is for us and Henry, it seems that folk psychology is entirely comfortable with the possibility that a person may acquire (or lose) the belief that the car keys are in the refrigerator, while the remainder of his beliefs remain unchanged. (Ramsey, Stich & Garon 1990, p. 320)

Does this notion of having the relevant "background" beliefs help there case? It is clear that it doesn't. Once Henry loses the belief that his car keys are in the refrigerator he will lose the belief that he ought to search for them there or the belief that they might be on his nightstand. In short, Henry's background beliefs will change. The point is that it is not just that Henry has beliefs about cars etc. He has more short term beliefs about his immediate circumstances and this is what Ramsey et al have overlooked. It need not be the case that all our beliefs are changed, it is enough that some are to show that beliefs, desires, and the like cannot be added and subtracted one at a time without effecting other beliefs. Ramsey et al have provided a convincing case that folk psychology is committed to discreteness based on this argument.

3.2 Causal Efficacy and Discreteness

As I see it there are two main issues regarding the causal commitments of folk psychology. First, is folk psychology is committed to the causal efficacy of mental states. Second, if folk psychology is committed to causality does this mean that the mental states must be discrete entities in the head in order to be causal. I will argue below that folk psychology has a weak commitment to causality and that mental states need not be discrete entities in the head in order to be causally efficacious.

Donald Davidson (1963) and Ramsey, Stich, & Garon (1990) argue that an agent can have a reason for taking an action, take that action, and yet it could be the case that this reason is not the reason the action was taken. Botterill & Carruthers (1999, p. 35) couch this in terms of the need to make a distinction between a possible reason and an operative reason. It looks as though the only way one can make this distinction is in terms of the causal involvement of intentional states. To see how this works consider the following case. Suppose Fred stole some candy at the supermarket. Fred had lots of possible reasons for stealing the candy. He didn't have any money, he wanted something sweet to eat at the movies, and it turns out that Fred used to be employed at the this particular supermarket and the manager treated him badly. Now it might be the case that had the manager treated Fred with more respect Fred would not have stolen the candy even though he had the other two reasons for stealing it. In this case Fred stole the candy because he was treated badly by the manager. The operative reason is Fred's maltreatment by the manager while the other two reasons were merely possible reasons.

Another reason folk psychological states must be causally efficacious according to Ramsey et al is the role these states play in inference. They argue that a person may have a number of belief clusters, any one of which might lead a person to infer some new belief. Once again we need the distinction between the operative belief in the inference and beliefs that could possibly have accounted for the inference. I will paraphrase Ramsey et al's (1990, p.321) example. Suppose Inspector Clouseau believes the butler said he spent the night in a village hotel and that he said he came on the morning train. Clouseau also believes that the hotel is closed and that the morning train is out of service. Given these beliefs and a set of relevant background beliefs Clouseau will probably infer that the butler is lying. According to Ramsey et al, from the perspective of folk psychology it could be the case that it was because of Clouseau's belief about the train only that he made the inference that the Butler was lying. Equally he could have made this inference because of his belief about the hotel or both. It is an empirical matter which belief was the operative belief.

Ramsey et al make the preceding arguments do double duty. As we saw above they provide compelling reasons to think that folk psychology is committed to the causal efficacy of FP-states. However, the authors argue that the only reason there can be an operative reason is that each reason is discrete. In each case amongst all the possible reasons Ramsey et al suggest that there can be one reason that is causally active while the rest are causally inert. It might be the case that Clouseau's belief about the train is causally active while his belief about the hotel is causally inert. Likewise it could be the case that Fred's belief that the manager treated him badly is causally active while his belief that he did not have any money or his desire for something sweet were causally inert. "Once again we see common sense psychology invoking a pair of distinct propositional attitudes, one of which is causally active on a particular occasion while the other is causally inert." (Ramsey, Stich, and Garon 1991, p. 321) This is not to say that all of these possible beliefs cannot be simultaneously causal or that various combinations of them cannot be so, but just that they are separable in this way. The next step of the argument is to infer that since beliefs are separable in this way they must be discrete causal entities. Each belief is a separate billiard ball, as it were, which can bounce around in all sorts of causal relations with other beliefs.

One reason to be weary of this is though the operative belief or desire accounts for a counter-factual like, "Fred would not have stolen the candy if the manager had treated him better," it does not follow from this that the other reasons were completely "causally inert." Note that this is not the same as saying all three of Fred's beliefs were operative. In this case the counter-factual would have to include all three beliefs. If we imagine each of the beliefs as providing some quantum of reason for Fred's action, it may well be the case that the operative reason provided the most reason for the action and that it was this reason that put Fred over a causal threshold. Furthermore, it may be the case that had Fred not had those other beliefs, the operative reason would still have put him over the causal threshold for the action. A stronger point against Ramsey et al's argument is that it is likely that many of Fred's background beliefs were absolutely necessary in order for him to steal the candy. This is to say if Fred did not believe what he stole was candy he probably would not have stolen the candy. Indeed, if Fred did not believe he was at the supermarket he would not believe there was candy to be stolen. It looks as though there is a long and wide causal chain of inner events that lead up to Fred's action. It certainly cannot be the case that the operative reason alone caused his action while all these other beliefs remained "causally inert." The argument I have just given applies equally well, mutatis mutandis to the discreteness argument regarding inference.

I would like to provide two reasons that explains what motivates intuitions for discreteness. Daniel Dennett (1991) argues that the illusion of discreteness is created by the fact that we are language speakers. The idea is that when we talk about our reasons for doing something, the demands of efficient communication prevent us from naming all of the beliefs and desires that we had and were necessary for our action. Only a philosopher would try to name them all. By naming a reason or a few reasons we highlight the operative reason and this makes it appear that it is the only reason we take to be causally efficacious.

The second point is to note how contrastive causal explanation privileges one cause out of many in explanation. To see how this works in detail see Lipton (1991). The basic idea is that causal why-questions make implicit contrasts. The easiest way to see how this works is to consider what John Stuart Mill calls the "real cause." He characterizes the whole cause in the following way. “It is seldom, if ever, between a consequent and a single antecedent, that this invariable sequence subsists. It is usually between a consequent and the sum of several antecedents; the concurrence of all of them being requisite to produce, that is, to be certain of being followed by, the consequent.” (Mill, Book III Ch. V §5). In our current discussion the whole cause of an action or inference might include many of our background beliefs as well as reasons that generally are cited in explaining a behaviour. Since we do not want to give the "real cause" every time we give a causal explanation, we need a way to focus in on and privilege certain causes over others in our explanation. The way this is done is by asking a contrastive question like "Why did Fred steal the candy rather than buying it?" We then apply Lipton's difference condition: “To explain why P rather than Q, we must cite a causal difference between P and not-Q, consisting of a cause of P and the absence of a corresponding event in the history of not-Q.” (Lipton, p.217) In the example I gave above P is Fred's stealing the candy and Q is Fred's buying the candy. Both of these events will share similar causal histories. All of Fred's background beliefs will be essentially the same in either case. The difference condition selects the relevant difference between these two causal histories and privileges that part of the causal history in the explanation. What I am saying is that folk psychological explanations make implicit contrasts and it is for this reason that we have the intuition that the operative cause is discrete and separable from the rest of the causal history.

3.3 Causal Mechanisms

In the previous two sections I have argued that while folk psychology is committed to causation in some form or another it is not committed to discrete entities in the held. I now want to broaden the scope of the discussion a little to argue that folk psychology is not committed to any particular internal mechanism. Botterill and Carruthers (1999) offer two arguments in favour of folk psychology being committed to a specific internal mechanism. Dennett (1987) and Davidson (1970) would argue that the folk are not committed to any internal mechanism whatsoever. I will argue that there is a third alternative. My view will be closer to Botterill and Carruthers than it will be to Dennett or Davidson.

The first argument Botterill and Carruthers offer is to consider the Turing test. Turing (1950) suggested a test for intelligence in a computer. If a machine could be programmed into fooling people who ask it questions into thinking the its response were from a human being then the computer could be considered intelligent. It seems this would be enough for intelligence for people inclined to agree with the behaviourists. Botterill and Carruthers argue "in order to engage in anything like human thinking is not sufficient to imitate the responses a human being would make- one would also need to imitate the processes by which such responses are produced. They point out that though a chess playing computer might produce moves that we might attribute to a human being, when we (as folk psychologists) find out that the computer uses a brute force algorithm to look at thousands of positions per second we would no longer attribute human intelligence to such a machine. There is a sort of essentialism the folk have with regard to our inner processes. Botterill and Carruthers also point that we as folk psychologists are readily willing to say that someone who is paralysed or who has cerebral palsy with little control of motor skills has a rich inner life. We do this despite the fact that there is no behaviour to interpret from. Both theses cases are meant to show that folk psychology is committed to a realism about mental states and that there are specific inner processes that must occur in order for something to be intelligent in the same way as human beings.

I take the point I am about to offer to be entirely consistent with Botterill and Carruthers line. I take it that I am just adding an additional comment to their story. Though folk psychology is committed to a somewhat specific internal story. Namely, it cannot be the case that we think like desk top computers. The commitments end here. The folk are not committed to any particular mechanism. They are only committed to something like "whatever scientists find out to be the actual mechanism we have." Folk psychology just does not say much about internal mechanisms beyond that simple fact that there are some and that they are distinctly human. Before alternatives to the classical program in AI and cognitive science it appeared that the folk were committed to something like mind is like a computer metaphor. Now that there are empirical alternatives such as connectionism. The folk are not committed to any particular alternative. Folk psychological causal stories are "giving a job description, which some internal process must satisfy." (Bechtel and Abrahamsen 1993, p. 359.)

4. Propositional Attitude Analysis

In this section I consider folk psychology's purported commitment to a propositional attitude analysis of mental states. I will not being using the term "propositional attitude" just as a name for a group of mental states such as beliefs, desires, and hopes. Rather, a propositional attitude is a particular philosophical analysis of the grammar and logic of terms like "belief" and "desire." A propositional attitude account of folk psychological states notes that our English idioms using terms like belief and desire have a particular syntactic structure. These words are followed by "that" and a proposition. A proposition is an expression that can be evaluated as true or false. I might say, "Fred believes that it is windy outside." Certain verbs can be taken as an attitude someone can have towards a proposition. This has been a useful analysis in many ways. One determines a belief to be a false belief if the proposition in its that-clause is false. We can compare and contrast propositions which we each might take an attitude towards. For example, I might desire a proposition to be true and you might desire that the very same proposition be false. I will argue that folk psychology is not entirely committed to this sort of analysis. It is not my goal to argue against the intentionality of thought or even to suggest that many Folk psychological states are not usefully analysed in this way. The point I want to make is that the propositional attitude analysis does not apply to many types of folk psychological states that I take to be an important part of folk psychology.

There are three types of cases that seem not to be committed to this sort of analysis. They are perception, know-how, and emotional moods. With respect to perception, it is clearly the case that perceptual verbs such as sees, hears, feels, tastes, and smells are a part of folk psychological explanations. For example I might say, "I don't eat anchovies because they taste bad." Bechtel and Abrahamsen (1993) point out that these perceptual verbs don't generally represent an attitude toward a proposition. They generally just take a direct object. I taste the ice-cream, see the blue sky, hear the music, feel the rough sandpaper, and smell the flower. I am not claiming that perceptual statements never represent an attitude toward a proposition. One can say, "I see that the door is open" or I can say, "I see the open door." However, these expressions arguably don't mean the same thing. The first expression means something like I notice the state of the door being open, and the second expression means something like I notice the open door itself not the fact that the door is open. Bechtel and Abrahamsen argue that perceptual verbs report how a whole person relates to their environment. "Perceptual verbs such as see, hear, and notice specify the object from which the organism is receiving information in a particular modality. Thus they specify an informational contact between an organism and some aspect of its environment." (Bechtel and Abrahamsen 1993, p. 357) The propositional attitude analysis does not apply well to perceptual verbs.

The second type of cases are know-how. Gilbert Ryle (1949) famously made a distinction between knowing how and knowing that. When we say somebody knows how to ride a bike we are not saying that this consists in a collection of propositions. Knowing how is having a capability whereas knowing that is being in possession of a piece of information. An example of knowing how is knowing how to play chess. Now it is true that one may store certain propositions such as ceteris paribus knights on the side of the board are badly placed. One might store some of the rules of the game as propositions as well. The king cannot castle through check for example. The majority of understanding how to play chess is not propositional however. It is common to talk about developing a strategic intuition and a skill at seeing combinations. These things are learned through practice and observing the play of others. Know-how is a part of folk psychology. We learn to recognize the various facial expressions as expressions of emotion just as we learn how to recognize the presence of combinational possibilities in a chess game. Perhaps our ability to recognize facial expressions is innate, but the point is that recognizing emotions is a capacity we have rather than a set of stored propositions. This capacity is a part of our understanding of the behaviour of our fellow human beings. Folk psychological know-how is a problem for the propositional attitude only account of folk psychology.

At least as important as beliefs in our understanding of our fellow human beings are emotional moods. Many emotional terms are not easily analysed in terms of attitudes to propositions if at all. Moods simply do not admit of a propositional attitude analysis. I might make the following prediction. "Fred will not go to the cinema because he is depressed." Moods are states that we are in that explain certain behaviours. They do not bear a relation to a proposition. Mood explanations use the following construction. "Agent Y behaves in this way because he is X," where X stands for any mood term. Examples are depressed, happy, sad, angry, etc. There are cases where terms like this are used in that-clause constructions. "He is sad that he won't get to go on the holiday with us." What I want to distinguish here is being sad about a particular event and just being sad in general. This depression or melancholy may just be a way a person is feeling without being about anything in particular. It is this latter case that does not admit of a propositional attitude analysis. Paul Griffiths (1997)has argued that the propositional attitude analysis is mistaken for all emotions, not just the moods that I have been discussing here. If his arguments are sound then there is a large class of mental states that are central to folk psychology which do not admit of the propositional attitude analysis. It is beyond the scope of this paper to consider all of these arguments. I am merely trying to show that not all folk psychological states can be analysed in this way. It has been my purpose to show that when you consider folk psychology in all its richness, the propositional attitude analysis places too strong a constraint on what will count as folk psychology. If it were a commitment of folk psychology that all folk psychological states must be given a propositional attitude analysis one would eliminate important features of folk psychology. When these features are removed we are no longer talking about the lay persons understanding of his fellow humans behaviour. This overly constrained folk psychology is a philosophical construction. Since folk psychology has been around in one form or another since arguably the time humans became a language speaking animals one should expect there to be more to it than beliefs and desires.

I want to end this discussion of the propositional attitude analysis by suggesting that folk psychology is not committed to any particular theory of internal representation. That is to say that folk psychology is neither committed to classic symbolic form of representation, nor is it committed to distributed representations such as those in a connectionist network. Given that connectionism is a possibility it is possible that not even the content of beliefs will take a propositional form. If this is indeed a possibility and folk psychology is not committed one way or the other on the form our representation takes, then folk psychology is not committed to content being represented in propositional form. Now a distinction needs to made here. I am not arguing that folk psychological states are not committed to be described with propositions. I am arguing that folk psychology is not committed to its states being represented in a propositional format. So long as whatever form folk psychological states are represented in can properly be described in a propositional format (except of course in the case of perception, know-how, moods, and possible all emotions which aren't even described in propositional form) folk psychology is not committed to propositional representation.

5. Conclusion

I have argued that folk psychology is a broad and rich body of knowledge and practices. Philosophers narrow conception of folk psychology as merely consisting in beliefs and desires has led to an overly constrained view. Folk psychology is certainly not committed to "propositional modularity." It is not committed to discreteness because our beliefs and desire take on a necessarily holistic character. Though folk psychology is committed to some form of causation, it is not committed to discrete causal entities in the head and it is not committed to any particular causal mechanism. Finally, folk psychology is not committed to the propositional attitude analysis because this would leave out perceptual verbs, know-how, and moods. The consequence of this is that though the eliminativist arguments may well be good arguments against a narrow philosophically constructed construal of folk psychology, they do not show us that the ordinary understanding of the lay person ought to be eliminated.

References

Bechtel, W. and Abrahamsen, A. A. 1993. "Connectionism and the Future of Folk Psychology." in Christensen & Turner, 1993.

Botterill, G. and Carruthers, P. 1999. The Philosophy of Psychology. Cambridge University Press.

Christensen, S. and Turner, D. 1993. Folk Psychology and the Philosophy of Mind. Lawrence Erlbaum Associates, Publishers.

Churchland, Paul. 1981. "Eliminative Materialism and the Propositional Attitudes." Journal of Philosophy, 78. Reprinted in his A Neurocomputational Perspective, MIT Press, 1989; and in Christensen and Turner, 1993.

Collins, A. and Quillian, M. 1972. "Experiments on Semantic Memory and Language Comprehension." in L. Gregg ed., Cognition in Learning and Memory. Wiley.

Cummins, Robert. 1991. "Methodological Reflections on Belief" in Mind and Commonsense Psychology, 1991, Bogdan, R. ed. Cambridge Press.

Davidson, Donald. 1963. "Actions, Reasons, and Causes." Journal of Philosophy, 60.

Reprinted in Davidson 1980.

Davidson, Donald. 1970. "Mental Events. In L. Foster and J. Swanson eds., Experience and Theory, Duckworth. Reprinted in Davidson 1980.

Davidson, Donald 1980. Essays on Actions and Events. Oxford University Press.

Dennett, Daniel C. 1987. The Intentional Stance. MIT Press.

Dennett, Daniel C. 1991. "Two contrasts: folk craft versus folk science, and belief versus opinion." in The Future of Folk Psychology: Intentionality and Cognitive Science. Greenwood ed. 1991. Cambridge University Press.

Fodor, Jerry. 1987. Psychosemantics. MIT Press.

Griffiths, Paul. 1997. What Emotions Really Are. University of Chicago Press

Lipton, P. 1991. "Contrastive Explanation." In D. Knowles ed., Explanation and Its Limits, Cambridge: Cambridge University Press. Reprinted in D-H. Ruben ed, Explanation, Oxford University Press, 1993, pp. 207-227. Pagination is from Ruben ed.

Mill, J.S. 1973. A System of Logic, Books I-III, Collected Works of John Stuart Mill, Toronto and London: University of Toronto Press and Routledge. (First ed. 1843)

Ramsey, W., Stich, S., and Garon, J. 1990. "Connectionism, Eliminativism, and the Future of Folk Psychology" Philosophical Perspectives, 4. Reprinted in Greenwood, 1991; and in Christensen & Turner, 1993. Pagination is from Christensen & Turner.

Ryle, Gilbert. 1949. The Concept of Mind. Hutchinson.

Turing, Alan. 1950. " Computing Machinery and intelligence. Mind, 59.

Wilkes, K. 1991a. "The long past and the short history." In R. Bogdan ed., Mind and Commonsense, Cambridge University Press.

Wilkes, K. 1991b. "The Relationship between Scientific Psychology and Common-sense Psychology." Synthese, 89. Reprinted in Christensen and Turner 1993.

^[1] Cited pagination for this article is from the reprint in Christensen and Turner 1993.