The Dreyfus Model of Skill Acquisition: How to Develop True Expertise

An Interview with Hubert and Stuart Dreyfus by Bent Flyvbjerg

HD: Hubert Dreyfus

SD: Stuart Dreyfus

BF: Bent Flyvbjerg

Where does a chess player feel that a move is right? What is the sensation and where is it located if it is located at all?

In examining a variety of skills as disparate as driving a car and playing chess the Dreyfuses find that there is a common pattern associated with improvement in ability which can be schematized into a developmental sequence. There are five distinguishable stages whose order of appearance is invariant: novice, advanced beginner, competence, proficiency, and expertise. Novices look for context-independent features which they interpret on the basis of rules. Beginning drivers look at their dashboard for velocity or RPM and consider rules for when to change gears. Beginning chess players evaluate their pieces by numerical value and apply the rule “Always exchange if the total value of pieces captured exceeds the value of pieces lost.” With the attainment of experience the advanced beginner can begin to pick out situational cues such as the sound a car makes when it needs to change gears, or the appearance of a “weakened king’s side” or a “strong pawn structure.” Maxims predicated on the ability to recognize such situated cues then serve to instruct one to shift up in response to a “racing sound” or attack a weakened king’s side. At the level of competence, one can organize a great deal of data about a situation through categorization. In so doing, one can hone in on particulars requiring attention by making them the salient points taken against a background understanding of the situation as a whole. Action is the product of concentrated deliberation over alternatives in the light of the salient particulars and an ensuing emotional involvement with the outcome is the by-product of a course of action which is felt to elicit individual judgment. A competent driver who recognizes herself taking a curve on a slippery road, at an unsafe speed, deliberately chooses whether to hit the brake, or just take her foot off the accelerator. She experiences the outcome of her choice in a more personal vein than would a beginner. A competent chess player recognizes a weakened king as the salient situation at hand and plans his attack without being distracted by any number of other attributes of the board or secondary consequences of his plan of action. At the level of proficiency there is no moment of detachment at which point one’s situation is sized up and categorized. The “spell of involvement” is not broken but rather one is just struck by the kind of situation one is in. The spell is broken however at the point of selecting a response, which becomes subject to deliberation. Finally, when one achieves expertise one can immediately both perceive the nature of the situation and “know” what must be done. The expert driver feels when slowing down is required and “knows” how to perform the action without calculating and comparing alternatives. The expert chess player, classed as an international master or grandmaster, can generally play at a rate which disallows the possibility of analysis and comparison of alternatives. The ability of the chess expert to play by intuition is illustrated by an experiment conducted by the Dreyfuses in which they had an international master play five second chess while rapidly adding numbers. He was still able to “hold his own” against a master level player although clearly deprived of any opportunity for deliberation.

BF: Concepts like the body, power and ethics have occupied a central place in Hubert Dreyfus’ work, for instance on Foucault, Heidegger and Merleau- Ponty. Yet, I’ve been struck by the virtual absence of these concepts in your work on learning. Let’s talk about the body first. You have given the most comprehensive account of your five-stage model of learning in Mind Over Machine. In the title of the book and in the book’s general statements about learning ‘mind’ is emphasized. It seems to me, however, that many of the examples you and Stuart provide depend on bodily learning, even if you don’t explicitly state it as such. There is a budding discussion these years about the role of the body in philosophy and social thinking. How does the body enter into your model of learning and how does this relate to the old split between body and mind in Western thinking?

HD: Mind Over Machine wasn’t our title since body is just as important as mind. Our idea cuts across this distinction. That is why we talk about chess and driving a lot. Chess is a mind skill whereas driving is a body skill. And, according to us, you go through the same five stages in either kind of skill. One might think that the holistic ability to do things intuitively is in the body, basically. But I don’t think that is right because chess is a very disembodied skill. So skill just seems to cut across the body/mind distinction and the misleading title of the book is just unfortunate. It should have been instead something like Body-Mind Over Machine.

BF: Intuition is central to your model. Where does intuition reside in the body- mind?

HD: One would like to think in the body. My favorite philosopher is Merleau-Ponty who writes about body skills and the holistic body gestalting the world. But it seems obvious to me that you couldn‘t possibly think that chess intuition has anything to do with the body. Intuition does not reside in the body or the mind. It resides in the fact that the brain can store whole gestalts. Chess is a perfect example because you don‘t even need perception. A blind person can become a chess master. It‘s just a totally mental activity. The distinction between mind and body is important in that the mind can be analytical and then the mind is opposed to the body because I don’t think the body can be analytical. So it is always the mind that causes the trouble if there is trouble. If you don’t become intuitive, it must be because the mind got in the way. But if you do become intuitive, it isn’t because the mind got out of the way and let the body do its thing. It seems to me, it’s because the analytic mind got out of the way and let the intuitive mind do its thing.

BF: Where does a chess player feel that a move is right? What is the sensation and where is it located if it is located at all?

SD: In the whole body. In the pit of the stomach. It is similar to asking where do you feel you are hungry when you are hungry. You can’t say that your brain thinks it is hungry. You experience your whole body as craving and the chess player has the same type of experience. So you see I disagree with Hubert on this point. When a chess player plays one-second-a-move chess – that’s really fast chess – I have heard them describe this funny sense that their hand is playing and they aren’t. They talk that way. Their hand is just moving pieces as fast as it can and they almost feel as if their detached brain looks down at their hand playing chess. So the whole body is even in that picture.

Ultimately, my work is leading me to deny the split between mind and body. The split has caused a lot of trouble for Western culture

 HD: What comes into chess is emotions. You‘ve got to have involvement in the chess game. You‘ve got to care. But I don‘t know exactly where emotions are embodied. They can‘t be in the mind understood as a detached objective analytical mind which is the usual Cartesian way of understanding the mind.

SD: Ultimately, my work is leading me to deny the split between mind and body. The split has caused a lot of trouble for Western culture and gets us into believing in rationality and the like. My work with neural networks and connectionism is leading me to see much of skilled behavior as an activity in which there isn’t any role for the mind as the mind is usually defined. It’s all “body” if you think of the neurons as being part of the body and not part of the mind.

BF: What does that mean, more specifically, “neurons being part of the body and not part of the mind?”

SD: The tradition thinks of the mind as being a detached thing. It is at a level of viewing the world which is detached from the world and thinks about the world. And it thinks of the body as being always in the world. You can‘t detach your body from the world, but you can detach your mind from the world and so the traditional mind-body point of view requires seeing a detached mind manipulating a body in the world. Opposed to this, the connectionist point of view sees the involved mind of an expert performer as just as involved as the body. They both operate on stimuli coming in from the world and without any level of interpretation. Another way to say this is that the usual view of the mind sees what is coming in as brute data and sees the mind as interpreting this data by manipulating its representations. The connectionist point of view is that the mind of a skilled performer isn‘t doing that and if you do that you won‘t be a skilled performer. You have got to just respond. The neurons in the brain produce a response when the stimuli are a certain way. It‘s a much more physical interpretation.

Foucault doesn‘t talk about the important point that disciplined bodies are not going to be as good at doing what they do, as intuitive bodies.

 BF: In addition to the body, power and ethics are concepts that seem to me essential for an understanding of skills and the process of de-skilling, i.e. people losing craftsmanship and intuitive abilities to mechanized work processes. How come you place so little emphasis on these concepts in your work on learning?

HD: We were constantly being pushed to talk about the ethics and politics of tacit knowledge by Tom Athanasiou who was helping us write Mind Over Machine. Our line constantly was, we talk about what we know about and at least one thing is clear: if you don’t understand what expertise is and what tacit knowledge is, you can’t even discuss the social, political, labor movement issues, power issues, etc. intelligently. You spend your time worrying about what to do if computers and expert systems come along and replace experts and workers when that is not the real problem, because they can’t. So we didn’t talk about the social issues. I presume there is going to be a kind of Taylorism similar to Foucault‘s account of docile bodies that would be the science of efficient behavior. One would try to discipline expert skills. As far as I know, Foucault doesn‘t talk about the important point that disciplined bodies are not going to be as good at doing what they do, as intuitive bodies. They may be good for other things, being more regimented, more reliable, more controllable, even more enhanceable in a welfare sense, but they just won‘t be as skillful. Labor unions know this. There is power in the tacit knowledge workers have or that skilled people have that power-knowledge can‘t get any grip on.

BF: Would it be possible to have a science of these skills?

HD: No, that‘s exactly why power-knowledge can‘t control them. You can‘t have a science of these skills and therefore they are a kind of counter power. It‘s what Foucault talks about when he talks about marginalized knowledge, the knowledge that dispersed marginal groups have. I guess there is a knowledge that anybody who is skilled at anything has that is outside of the whole science-power complex. The Swedish labor unions think about what tacit knowledge is and how to preserve it and enhance it.

BF: If we broaden this question of skills to encompass human practices in general, it seems to me what you’ve just described for labor closely parallels a concept you’ve touched briefly upon in a number of contexts, namely the concept of so-called marginal or non-rationalized practices. Could you expand on this concept?

HD: The favorite example these days seems to be friendship. People have a lot of practices for supporting their friends and choosing their friends and abandoning their friends, but philosophy hasn‘t worked on it or tried to make principles about it and it hasn‘t become part of the efficient technological ordering of things. People have not gone around saying we should make friendship more efficient, more productive and so forth, so it exists as a kind of non-rationalized practice.

BF: Could you mention other domains of marginal or non-rationalized practices?

HD: My two other favorite examples are nursing and teaching. There is a book called The Primacy of Caring by a nurse who used to come to our Heidegger and Kirkegaard courses. In it she shows that caring about people just because they are people is a practice that nurses have and that we all have because it comes down to us from Christianity. But it is not a rationalized practice in the sense that we don’t do it on principle and we haven’t tried to organize it for the sake of efficiency. It is a sort of practice she thinks you couldn’t organize for the sake of efficiency. She thinks nurses have it and doctors tend not to have it. They want to just make people healthy and productive but they don’t want to have to care about them. She thinks, in fact, that unless you care about people you can’t treat them very well medically. In our culture there are people who become high school teachers, some of them because it is the only job they can get, but some of them, I gather, because they feel that there is nothing more important than nurturing the youth, mentoring people. That is another kind of practice; in this case it comes to us from the Greeks. I think of caring as a Christian marginal practice that was once central but now isn’t central anymore because Christianity is no longer a living part of what Hegel called our Sittlichkeit, our customs and shared social practices — at least not a central part of it. The mentoring practices were a very important thing for the Greeks, at least the people around Plato and Socrates, but now they don‘t have the institutionalized importance they did then. I mean, you don‘t get any credit for being a high school teacher, you don‘t get much money and you don‘t get much prestige. So why does anybody do it who is intelligent enough to do something else? If there really are people who do it even though they are intelligent enough and hard working enough to do something else, it must be because it is important to them and its importance must be of this kind of marginal sort.

Heidegger says you should preserve the saving power of the humble things, but how you are supposed to do it he doesn't say. Bob Dylan says strengthen the things that remain. T. S. Eliot says, shore up fragments against our ruin.

BF: Couldn‘t one say that what is happening with friendship, caring, etc. in modern society is exactly that they are being rationalized? For instance, if you want success in the corporate or academic world it may be important to have the right friends and people may choose friends primarily on this basis.

HD: Marginal practices always risk being taken over by technological rational understanding and made efficient and productive. I think it was my former student, Jane Rubin, who told me that she heard a radio announcement that it is healthy to have friends. As soon as you have friends for your health or for your career you‘ve got some new kind of friendship which is of a technological-rational kind. That always could happen and it could also happen, I suppose, that this new kind of friendship could replace the other kind of friendship. People wouldn‘t even know anymore what real friendship was. That is a danger and that‘s why you have to, I think, strengthen what I call endangered species of practices. That‘s what Heidegger and Foucault believe, too, I think. Romantic love, which Kierkegaard was trying to save is another of these practices. So is animal rights, by the way. There were, recently, many animal rights activists picketing a new building and fighting the university police here [at University of California, Berkeley]. At first I thought, “Oh, more crazy fanatics,” but then I realized, no, these people represent a version of another kind of marginal practice, one that feels a community with animals and doesn’t think that they are so different from human beings that they should be sacrificed and made to suffer in order to make us healthier and longer living. All these practices should be preserved, I think, and constantly around to call into question the hegemony, as people say these days, of human efficiency, productivity, health and welfare.

BF: How can one defend these marginal practices, strengthen them, and create more of them to counteract rationalization?

HD: I really don‘t know. Heidegger says you should preserve the saving power of the humble things, but how you are supposed to do it he doesn‘t say. Bob Dylan says strengthen the things that remain. T. S. Eliot says, shore up fragments against our ruin. Poets have talked about these things, but, as far as I know, nobody has done anything about it. I think all one can say is that one should not try to rationalize and technologize them, nor should one judge them as trivial and insignificant just because they can‘t be rationalized. Rather you should actually practice as many of them as seem natural to you and not let it worry you that from the reigning understanding of reality they seem at best trivial and at worst counter-productive. By that I mean it may seem that you could have been spending this time doing something to improve yourself and here you are instead only playing with your children, or talking to your friends or going around looking in the eyes of and petting animals. Some might say you are wasting your time, that you are not getting the most out of your life, but you should resist that.

BF: You are referring to poets in making this point. Is there a role for the professional philosopher and social thinker in this kind of issue?

HD: I suppose not, if by professional philosophers you mean people in the line from Plato via Kant to Habermas. But people in philosophy departments could write books like The Primacy of Caring which happens to have been written by a professor of nursing but very influenced by philosophers. I presume one could write a book on friendship which was the wrong sort of book asking questions like, “what are the necessary and sufficient conditions for counting someone as a friend?” or, “what are the universal principles that dictate what you should do with friends?”, that would be terrible; but you could write a philosophy book describing the practice of friendship either through history or as it is now. Philosophers don’t tend to write these books. Zen and the Art of Motorcycle Maintenance is such a book but wasn’t written by a philosopher. The nearest thing to a book like that written by a philosopher in a philosophy department is Technology and The Character of Contemporary Life by Albert Borgmann. He talks about focal practices, which is his name for a sub-class of marginal practices. He sees how technology disperses and destroys these practices because technology makes it easy to get results without any effort and focus. For instance, for music you can just switch on the compact disc player. Borgmann contrasts this with learning to play an instrument, getting people together, practicing, and then finally playing a quartet. That was a focal achievement pulling all these people and things together. It is like Heidegger talking about people sitting around with a jug of wine. Borgmann says you have to resist the way technology gives you the result without any of the effort, because you won‘t have the focal experience if you do it that way. It is very interesting. Another practice for him is the family meal. The family all sitting together at a dinner that they have actually helped prepare, not each one of them eating his tv- dinner in his separate room. We are losing this focal practice. Thanks to technology and tv-dinners you don‘t have to prepare your meals, nor eat them at the same time anymore.

BF: You’re giving examples on how we individually can counter rational- ization. In Scandinavia we have become used to asking, in addition, what can be done at the societal level, collectively and politically?

HD: My examples do not concern what we can do as individuals only but also as a group: two friends, a family or a quartet. But you want something bigger. What you really want to know as a Scandinavian is how saving practices can be legislated. Your question is typically Scandinavian, it is fascinating and intriguing: if it is good for people, and we want to preserve it, then we ought to have some kind of political movement and write laws to preserve it. I can‘t imagine what that would look like, laws to preserve friendship, to preserve caring, to preserve mentoring, to preserve family meals, and people playing music together, or even laws to prevent the ways that these things are being undermined. At least we can outlaw what is endangering these practices, would be a Scandinavian idea. I don’t see how, since what’s endangering these practices isn’t something you can outlaw, it is rationality, efficiency. I mean, why should we make music together if each of us can put on a CD? Why should you waste your time with friends — with what we used to call friends — if you can go play golf with the head of the company? It is fascinating. I don‘t know which side you are on when you ask this question as a Scandinavian but I would say that‘s where the real problem lies in the Scandinavian approach. It is part of the attitude of rationalization to think you can legislate whatever is valuable.

Habermas wouldn't like this at all. And Heidegger would say about people like Habermas that they think of technology as a practice that we should control and use for good ends. But that is a technological way of thinking about it, Heidegger would say.

 BF: On the other hand many people may find it rather hopeless that you can’t do anything major, collectively, about what you see as a major problem, namely rationalization

HD: That‘s right. It does sound hopeless and people accuse Heidegger and Foucault of being different versions of a kind of pacifism. Heidegger was explicitly passive. He said, you can‘t do anything organized about rationalization at all. Only in your own life can you preserve the humble things and hope that a new paradigm will come that will make them more powerful. Foucault was much more active and he tried to fight on the side of marginal knowledge. You can, for instance, be involved in what he was involved in, gay rights and other activities like that. If you find some kind of marginal practice that the law is working against, for instance by making it impossible for antivivisectionists to picket laboratories, then fight it. If you were Foucault you would certainly want to say that if you felt like it you should join the animal rights people. You shouldn‘t scoff at them because they can‘t give any efficient rational basis for their thinking animals are as important as people, and you should certainly not have laws that work against them saying it is built into the Constitution that people have rights and animals don‘t, therefore you have no right to fight for animal rights. But that‘s pretty minimal. It does look like it is not the sort of thing that you can politically pursue. Heidegger is very explicit about this. He says, as soon as you try to make the preservation of marginal practices a positive, political goal, you bring in technology and undermine exactly what you are trying to do. I never understood that until I told you the story about Scandinavia, and now I see why he says that. You want to legislate the quality of life and you get this funny problem that the receptive, spontaneous aspects of the quality of life would be lost if you legislated it. Luckily there is a kind of middle area where there is a lot to do. That keeps the Scandinavians happy and makes them interesting. I mean, you could certainly legislate against deskilling, for instance. People in Sweden are worrying about that. That is an area to do something in.

BF: Some people, probably including Jürgen Habermas, would think of this minimalist attitude as conservative and individualistic.

HD: Yes, Habermas wouldn‘t like this at all. And Heidegger would say about people like Habermas that they think of technology as a practice that we should control and use for good ends. But that is a technological way of thinking about it, Heidegger would say. For Foucault, it is the universality of the principles underlying such an attempt that would be repugnant and dangerous, while for Heidegger it is the attempt to gain political control of our lives that would be the problem. But there is a kind of conservatism that is necessary on this view, namely that you will be conserving what are ancient, marginal practices in our culture. But I wouldn’t call it conservative since it can be quite radical and you won’t look conservative when you fight for some of these practices.

BF: Now, if we step back and look at this problematic of rationalized versus non-rationalized practices along a slightly different dimension than politics and legislation, namely that of ethics, I’m wondering what kind of ethic is implied by your position. It seems to me that your position, including the five- stage model of learning, is antithetical to the search for theory, rules and universals which is at the heart of most works on ethics.

SD: I teach a course called The Social Applications of Technology. It is co-taught by several professors. The first professor who lectured to the students taught them traditional Western philosophical views of ethical decision-making. She talked about utilitarianism, rights, rules and principles and told the class, basically, that in making social decisions concerning technology, one has to get values straight and reason from first principles. Her thesis was that you cannot decide rationally whether first principle A or first principle B is right but you can reason from any of those perspectives. She taught rationality in the sense that at least one should be able to defend one‘s conclusions based on one‘s perspective. Based on the five-stage skill model I spent the next three lectures in the course trying to tell the students that that was a waste of time. First of all, it is polarizing and leaves no room for discussion and argument if one person says my principles say A and therefore I conclude B and another says my principles start from C and therefore I conclude D. Secondly, that isn‘t the way people really make decisions when they are skilled in an area. Decisions are made based on prior experience with what has proved to be good or bad, successful or unsuccessful. I argued that one shouldn‘t try to objectify the question but should instead discuss historical precedent in a non-polarizing productive way. That is a total transformation of the issue of what is moral or ethical behavior from a view based on universal principles into one based on history and narrative.

BF: Could you give an example of this?

SD: Yes, just the framing of the question as to what is the problem is already half the answer, and nobody thinks you can rationally define what the problem is. In this country, for example, you can see our major problem as crime, or you can see it as poverty. If you see the problem as poverty you say that if you can deal with poverty, crime will go away. You see crime as a result of the fact that there is an impoverished class. If you see the problem as crime, on the other hand, you don‘t worry about taking care of the poverty problem. You try to build jails and hire more policemen. Your experience dictates what you see as the problem. It is only my lifetime of experience that makes me want to see the problem as poverty and not crime. I am sure someone else who has experienced crime against his family may see it as crime and not poverty. So you see ethics is experience-based and not rational and principle-based.

Read full interview and see notes and sources here: http://bit.ly/1BGhsQo