Margolis (2000) has suggested that subjects solve Wason's selection task badly because they misinterpret it. This explanation seems correct but trivial. I propose that to understand why the task is so difficult it is necessary to reject the conventional model of problem-solving underpinning most hypotheses in relation to Wason's task. An explanation based on the idea of self- organisation is considered.
2. The conventional model suggests that two orthogonal variables determine performance. One variable is the wording of a problem. This variable determines whether an individual is able to comprehend the problem, define its initial condition and goal, and load the problem onto working memory without errors and omissions. The second variable is familiarity with the domain from which the problem is extracted. Numerous findings demonstrate that experts activate many more operators than do novices, and that they apply the operators simultaneously rather than successively for solving the same problem (Erickson & Lehmann 1996). It seems obvious that where there is a high enough level of familiarity, it is only wording that determines performance because the necessary operators can always be activated effectively.
3. Two consequences can be inferred from the model. The first is that performance must be proportional to the complexity of wording and/or familiarity with the problem. Indeed, everyday experience teaches that experts solve problems better than novices, and that it is easier to solve clear problems than vague and ill-defined ones. The second consequence, interconnected with the first one, is that weak changes in wording and familiarity may result in corresponding changes in performance [NOTE 1]. This idea underpins the training procedures in most domains.
4. The Wason selection task (Wason 1966) seems inconsistent with the first consequence. The formal wording of the task is unusual but very clear; most people in an industrial society are able to solve similar problems and are familiar with the reuisite logical rules. Nevertheless, it is very difficult to find the right solution to the task (Johnson-Laird & Wason 1977). On the other hand, the Margolis version of the task (Margolis 2000) violates the second consequence. Changes in the wording, which, however, keep constant the sense of the task, substantially improve performance. I agree with Margolis that subjects cannot solve the problem because they misinterpret it. But this explanation raises another question: why do the subjects misinterpret the task if their knowledge and skills allow the correct interpretation to be constructed? (Margolis's version confirms this.) The reference to the pragmatics of ordinary language seems insufficient to answer this question because of the innumerable problems and puzzles which are either difficult to solve or whose solutions are usually wrong, notwithstanding the apparent "simplicity" of such problems and the appropriate knowledge and skills. Given that in all such cases the required knowledge and skills are available, misinterpretation can be considered as a correct but trivial assumption. A more profound explanation appears necessary.
5. In order to answer the above question it is necessary to abandon the conventional model of problem solving which implicitly underlies the diverse responses to the above question [NOTE 2]. I have suggested elsewhere (Prudkov & Rodina 1999) that as an individual constructs a goal-directed process (of which problem solving is an example), both goal and means are synthesised in a self-organising activity. In the present context this means that it is impossible to separate the comprehension of a problem and the activation of operators from long-term memory. This is a common process in which the brain simultaneously constructs an ongoing representation of the problem and the means to solve it. Integrating all the available information, the brain synthesises the initial condition, goal, and operators; the criterion for such a synthesis is to minimise the brain activity which is necessary to construct them. When the parsimonious synthesis goes astray, the solver fails to handle the problem adequately, irrespective of their level of competence in similar problems. A new iteration of the problem solving process is not the simple selection of another operator (or operators); this only becomes possible after synthesising other representations and means.
6. I suggest that subjects perform the Wason selection task badly because the brain constructs an inadequate representation and operators for this. With the abstract and unusual wording of the task, logical routines based on pragmatic experience cannot be activated effectively. Then, on the basis of the criterion of minimal activity, and in the process of synthesising the representation and operators, the brain biases us toward the content of the task, especially toward the rule which should be verified. As a result, the subject does not realize that an example falsifying the rule should be selected and gives a wrong solution. In order to avoid this it is necessary to provide the interaction between the pragmatic routines and the content. Indeed, the "concrete" versions of the task result in substantial improvement of performance (Johnson-Laird & Wason 1977). On the other hand, the Margolis version of the task reduces the number of incorrect alternatives toward which the process of synthesis can be biased.
7. Why is psychology unable to explain how people solve problems? Perhaps it is because psychologists have not yet identified the problem that needs to be solved in order to answer this question.
[1] Of course, it is sometimes enough to alter two words in a simple problem in order to convert it into an insoluble one. However, in most situations, small changes in wording do not influence the essence of problems.
[2] The task was devised by Wason in 1966. The fact that this task, whose formulation requires only several lines, is still current in 2000, reflects the fundamental difficulties of contemporary psychology in understanding the mechanism of human thinking.
Ericsson, K. A., & Lehmann, A. C. (1996). Expert and exceptional performance: Evidence of maximal adaptation to task constraints. Annual Review of Psychology, 47, 273-305.
Johnson-Laird, P.N. and Wason, P.C. (1977). A theoretical analysis of insight into a reasoning task. In Thinking. Readings in Cognitive Science, Eds. Johnson-Laird, P.N. and Wason, P.C, Cambridge University Press.
Margolis, H. (2000) Wason's Selection Task With Reduced Array. PSYCOLOQUY 11(005) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/2000.volume.11/ psyc.00.11.005.reduced-wason-task.1.margolis http://www.cogsci.soton.ac.uk/psyc-bin/newpsy?11.005
Newell, A. (1990), Unified theories of cognition, Cambridge, MA: Harvard University Press
Newell, A. and Simon, H.A. (1972). Human problem solving. Prentice- Hall.
Prudkov, P. N. & Rodina, O.N.(1999). Synthesis of Purposeful Processes. PSYCOLOQUY 10(070). ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.070.purposeful-processes.1.prudkov http://www.cogsci.soton.ac.uk/psyc-bin/newspy?10.070
Wason, P.C. (1966) Reasoning. In B. M. Foss (Ed.) New Horizons in Psychology I. Penguin