I find it hard to accept the notion that an expert professional can’t explain their actions. I’m an expert musician, and I’ve taken lessons from world leading players. I have my own introspection, and all of expert players have been highly articulate in their domain, although how they frame their knowledge may not be immediately helpful to lower levels of player. What I mean by this is what is instantly retrievable in their consciousness subsumes so many proceduralised skills that at the time of expressing the ‘why’ of an action so many other components are subconsciously present that only those who are aware of those immersed components can make sense of the ultimate verbal account of the action. Another way of putting this is that one verbal account may comprise dozens of earlier threshold concepts. If the interlocutor hasn’t passed those thresholds, the account will make no sense.
However, once the expert’s mind is focussed onto the components of each higher level skill, they are able to address the components. In other words, speaking to an expert requires a skill in itself, rather like trying to get information about language from a native speaker who is not trained as a language teacher. In the studies we read this week, the expert’s inability to articulate their ideas was not presented at all convincingly. Eraut (1994) noted that Schon focussed on exceptional cases “searching out examples of professional work which demonstrate artistry and thereby most clearly refute the technical rationality model” (p. 143). Eraut also presented the concept of deliberation time and why having or doing that will not be normal for professionals.
Rolfe (1997) attempted to present expert’s self-presentation of their knowledge as a reflection on their intuition partly supports my thesis. Rolfe argues that fuzzy logic is a abduction, and using this, expert practice can be articulated. Fuzzy logic and her sister Bayesian probability are rational models. Rationality does not depend on certainty. It refers to the thought processes of considering all components in a situation and making a decision. Some components will be vague, emotional, or unverifiable. The point is that these logic methods are rational and can be verbalised. Fuzzy logic questions the amount of a property present in a set (Watkins, n.d.), whereas Bayesian schemes question how probable that the property is present. See Kosko, (1990) for a mathematical description of the difference. Again, I return to the notion of speaking to experts in an expert way. Rolfe wants experts to recognise their fuzziness, I want them to recognise the schemas they use.
My friend, Dr K., is a psychiatrist of 20 years standing: the 2nd in charge of a large local specialist hospital (also an excellent viola da gamba player). I asked him about his typical clinical procedure. He explained what can be described as Gestaltian; he listens to a patient for about five minutes and suddenly he ‘knows’ what to do. I asked how do you know? He wasn’t sure of the how. This account matches Dreyfus^2‘s account of proficient expert (2005). I asked him about one particular patient, and firstly asked him to describe (without divulging personal details) the general situation of that patient. I tried to activate his top-level script. K-sensei could easily recall that information. I asked about the general features of that typical situation, what usually happens in cases like these. Again, he was able to talk (in length!) about this sort of case. I asked about what kind of medical knowledge is useful that is specific to these cases. No problems there; and so on down to the particular medicine and treatment protocols that should be used. Going back up the schemata again, our conversation focused on the individual differences in this case. He could locate many. In the end, he realised that his ‘a-ha’ moment was the result of decades of submerged but available knowledge. And none (or very little) was actually obscured from his reflection—once his mind was pointed in the right direction.
I have no doubt that most humans will forget the what and the why of what we do most of the time. Our available conscious memory is limited, severely at times, and demands placed on it push out less important information. Expertise is known to develop as the combination of lower-level knowledge. Without these combinations, humans would need to begin every conversation (assuming conversation were possible) from first principles. Asking an expert to describe what they do without giving them a framework in which to place that description leads to all sorts of problems regarding the interpretation of the data. The Dreyfus^2 model feels intuitively right at first glance, but I am aiming at the doctoral level, not at pop-psychology. I need to question base assumptions and look for supportable evidence.
The question this week asks what expert knowledge can tell us. My answer is a lot if we look into how their schemas are built. And then we need to develop a language for speaking with experts. But I’ll finish this now by quoting from my learning log from last week. Educators may need to look at novice knowledge systems more than at expert knowledge.
This evening, I asked my 12-year-old if she thought kids talking together about a maths problem would be better than hearing the teacher talk about it. She replied that the kids would be better because they understood each other more than the (her) teacher did, and if one kid was weaker, the other kids could know better what to say to make the weak kid stronger. This answer amazed me as a father, but it’s intuitively right. Of course, kids can’t know the full implications of any concept, but they don’t need to. All they need to do is know how to function with that concept at a level appropriate to getting the right answer in class. Social interaction, peer-to-peer may be seen as a social constructivist technique, but there is a very strong argument that this interaction provides the optimal input for individual cognition.
Dreyfus, H. L., Dreyfus, S. E., & Dreyfus, S. E. (2005). Peripheral Vision: Expertise in Real World Contexts. Organization Studies, 26(5), 779–792. http://doi.org/10.1177/0170840605053102
Eraut, M. (1994). “Theories of Professional Expertise.” Developing Professional Knowledge and Competence.
Kosko, B. (1990). Fuzziness_Vs_Probability. International Journal of General Systems, 17, 211–240.
Rolfe, G. (1997). Science, abduction and the fuzzy nurse: an exploration of expertise. Journal of Advanced Nursing, 25(5), 1070–1075. http://doi.org/10.1046/j.1365-2648.1997.19970251070.x
Watkins, T. (n.d.) The logic of fuzzy sets. Retrieved on July 17 from http://www.sjsu.edu/faculty/watkins/fuzzysets.htm