Jack R. Adams-Webber (1995) A Pragmatic Constructivist Gambit for Cognitive Scientists. Psycoloquy: 6(34) Human Choice (2)

Volume: 6 (next, prev) Issue: 34 (next, prev) Article: 2 (next prev first) Alternate versions: ASCII Summary
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 6(34): A Pragmatic Constructivist Gambit for Cognitive Scientists

Commentary on Lefebvre on Human-Choice

Jack R. Adams-Webber
Department of Psychology
Brock University
St. Catharines, Ontario
Canada L2S 3A1



It is argued that it makes epistemological sense for cognitive scientists to adopt the pragmatic constructivist strategy pursued by Lefebvre (1995). The intent is to identify formal principles which can impose logical constraints on the construction of computational models applicable to specific psychological problems such as, in the case of Lefebvre's own research, the prediction of human choice behavior.


choice; computation; decision theory; ethical cognition; mathematical psychology; model building; parameter estimation; probability; rationality.
1. Lefebvre (1995) illustrates how an epistemological strategy that has proved useful in contemporary theoretical physics can also be employed in the construction of computational models in cognitive science. Specifically, he formulates an a priori assumption which cannot be falsified empirically, but allows us to generate testable hypotheses concerning possible values of parameters and boundary conditions for existing models of cognitive processes.

2. As clearly explained in ordinary English for the benefit of non-physicists by several leading cosmologists, including Hawking (1988) and Penrose (1989), the anthropic principle entails an explicit criterion for excluding from further consideration a certain subset of possible mathematical models of the evolution of the universe that do not allow for the appearance of a human observer at a certain point in its development. As Penrose (1989, p. 434) also notes, it "could provide a reason that consciousness is here without it having to be favoured by natural selection".

3. Thus, the anthropic principle serves an important epistemological function in cosmology, which possibly also has implications for the "physics of mind". Lefebvre (1995) offers an "abstract principle of freedom" that might serve an analogous epistemological function in contemporary cognitive science by imposing logical restrictions on the selection of basic parameters in computational models of human behavior.

4. His work provides an example of a methodological gambit that is novel from the standpoint of traditional experimental psychology, but has become almost commonplace in physics and computer science, that is, employing abstract a priori postulates, which are not in themselves empirically falsifiable, to guide the selection of initial parameters and/or the location of boundary conditions for theoretical models without recourse to experiments (e.g., the "temporal projection" problem in artificial intelligence). Once such parameters are inserted into specific models, at least some of their predictive implications can be deduced and tested empirically; however, their initial selection or exclusion from consideration may be guided only by logical discourse.

5. For instance, I can hardly use the second law of thermodynamics to specifically predict how fast the coffee in my cup will cool off, although it implies generally that, if I were to place it next to a cup of relatively cold coffee, I would thereby irreversibly increase the total amount of entropy in the universe by some unspecified amount. More to the point, the same law also has specific implications for cognitive science in that it tells us that any model of the organization of information in memory, either human or computer, must entail a net increase in entropy when we take into account the amount of energy dissipated in its implementation. That is, as Hawking (1988, p. 147) puts it, "this increase in disorder is always greater than the increase in order of the memory itself." It follows that the direction of time in which the past is remembered must be that in which entropy increases. Therefore, in constructing models of memory, we can rule out the possibility of symmetry with respect to time (say we were thinking in terms of a quantum computer); and accordingly, all of such models will be, in principle, logically subject to the multifaceted temporal projection problem (cf. Adams-Webber, 1993).

6. Within the explicit context of modeling binary choice, Lefebvre (1995) shows that his own "principle of freedom" has some empirical implications that can be tested against experimental data. Indeed, he adduces several experimental observations to illustrate this point. He refers specifically to the fact, as he has demonstrated elsewhere (Lefebvre, 1985, 1992), several "golden section" findings can be deduced directly from his computational model of binary choice; and in the current paper, he shows that specific parameters in this model can be derived from a formal principle of much broader generality, that is, an a priori assumption of freedom.

7. Another interesting aspect of the anthropic principle in cosmology is that it is reflexive in the sense that it applies directly to the fact that there currently exist cosmologists who are developing models of the evolution of the universe which may be judged to be either consistent or inconsistent with the anthropic principle itself. Lefebvre shows that his "principle of freedom" has the same kind of reflexive property in that it can be applied to our own choice of parameters in the construction of cognitive models.

8. Perhaps the deepest implication of his argument is that the "principle of freedom" imposes a restriction on the range of convenience of the anthropic principle itself. That is, the two principles, in combination, rule out all models of the evolution of the universe that do not include the conditions for the appearance, at a certain point, of an observer similar to a human being who enjoys free will, and under some circumstances, freedom of choice as well. As Lefebvre, himself points out, this position allows for the epistemological possibility that such observers might themselves construct the phenomena which they observe.

9. An important premise underlying Lefebvre's thesis is that, as in theoretical physics, empirical findings are not necessarily either constitutive of, or regulative of, all of the principles applicable to the choice of parameters for computational models in cognitive science. The wider implications of this assumption need to be explored more fully. For example, Miles (1986, p.172) speculates, "there is implicit in empirical science an a priori structure that provides a reference frame without which empirical investigation would itself be impossible." From a strictly pragmatic standpoint, it is difficult to see how we can ever develop fully specified formal models of any generality, such as Lefebvre's (1992) own algebraic model of reflexion, without using certain a priori postulates to guide the initial choice of assumptions. Nonetheless, this process of construction should not be free-floating: that is, explicit logical criteria should govern the use of formal postulates in the development of cognitive models, such as, for example, mathematical consistency.

10. Lefebvre could possibly strengthen his general case for our using the methodological gambit exemplified by the anthropic principle, if he were to more fully explicate other assumptions underlying his argument that we can construct useful functional descriptions of cognitive processes at relatively high levels of abstraction in which the selection of certain parameters is guided by a priori formal principles of wide generality (of which his principle of freedom is but one example) without direct recourse to experimental data. For instance, how might we initially delimit the potential search space for finding relevant parameters? Perhaps we need to devise some clear epistemological criteria for determining the range of relevance of any empirical generalization (cf. Von Wright, 1966). Further elaboration of his thesis might also require our distinguishing between different levels of logical discourse with a view to avoiding category errors, while, hopefully, also eschewing metaphysics (cf. Hayes, Ford & Adams-Webber, 1994).


Adams-Webber, J. (1993). The robot's designer's dilemma. American Journal of Psychology, 106, 300-303.

Hawking, S.W. (1988). A brief history of time: From the big bang to black holes. New York: Bantam.

Hayes, P.J., Ford, K.M. & Adams-Webber, J. (1994). Human Reasoning About Artificial Intelligence. In E. Dietrich (Ed.), Thinking computers and virtual persons: Essays on the intentionality of machines (pp. 331-353). San Diego: Academic Press.

Lefebvre, V. A. (1985). The golden section and an algebraic model of ethical cognition. Journal of Mathematical Psychology, 29, 289-310.

Lefebvre, V.A. (1992). A rational equation of attractive proportions. Journal of Mathematical Psychology, 36, 100-128.

Lefebvre, V.A. (1995). The Anthropic Principle in Psychology and Human Choice. PSYCOLOQUY 6(29) human-choice.1.lefebvre.

Miles, M. (1986). Kant and the synthetic a priori. University of Toronto Quarterly, 55, 172-184.

Penrose, R. (1989). The emperor's new mind. New York: Oxford.

Von Wright, G.H. (1966). The paradoxes of confirmation. In J. Hintikka & P. Suppes (Eds.), Aspects of inductive logic (pp. 208-218). Amsterdam: North Holland.

Volume: 6 (next, prev) Issue: 34 (next, prev) Article: 2 (next prev first) Alternate versions: ASCII Summary