American Scientist

The Things We Do: Using the Lessons of Bernard and Darwin to Understand the What, How, and Why of Our Behavior. Gary Cziko. xi + 290 pp. The MIT Press, 2000. $37.95.

Gary Cziko thinks that many currently trendy fields in the study of brain and behavior are in need of a history lesson. In The Things We Do, which is an engaging blend of historical review and scientific critique, he notes the complementarity between Charles Darwin's logic of natural selection and Claude Bernard's logic of homeostatic regulation and shows how they point to an alternative view of some otherwise unquestioned assumptions of current behavioral biology. Unfortunately, the contents are poorly represented by the dull and unimaginative title, which might better (and less humbly) have been something like Beyond the Clockwork Mind: Using Darwin and Bernard to Untangle Nature/Nurture.

Cziko's goal is to reintroduce "purpose without spirit" into the behavioral sciences. Following the lead of William James, who described behavior as the application of "varying means to a fixed end," he reasons that no behavioral description is adequate that does not make reference to some internally represented goal; otherwise, we become lost in trying to specify an interminable maze of if-then input-output instructions. But he believes that current cognitive science has accepted assumptions that undermine efforts to understand purposive behavior.

His villain in this battle of ideas is the linear one-way cause-effect conception of material processes that has been the cornerstone of modern materialism since Isaac Newton. He argues that misapplication of this paradigm to living systems is the central confusion still plaguing behavioral science. His alternative is the concept of "circular causality," exemplified by the logic of a thermostat or the guidance of a homing device. Physiological processes maintaining homeostasis and behaviors for achieving some perceived goal can both be described in these terms.

Cziko's mentor in this arena, the behavioral scientist William T. Powers, used these ideas to characterize behavior as being organized to control perceptions rather than the other way around. According to Powers, purpose is concretely represented as a perceptual reference signal in some internal feedback loop embedded in other, higher-order feedback loops. Inherited predispositions are, as a result, more likely to be implemented by perceptual biases and not by behavioral or cognitive instructions. This becomes one cornerstone of Cziko's purposive behavioral logic.

The other major influence on Cziko is the social scientist Donald T. Campbell, who articulated an extended view of Darwinian processes based on "vicarious blind variation and selective retention." This "universal Darwinism" provides a theoretical framework that can be extended to apply to evolutionlike processes within the organism?from antibody production to neural development to creative thinking?as well as to the social transmission of information that constitutes culture and social habits. Cziko implies (although he never fully elaborates this possibility) that the circular causality that controls purposive behavior is not merely the logic of servomechanisms but is rather a Darwinian (Campbellian) process itself. Cybernetics meets universal Darwinism!

Unfortunately, this is roughly the point at which Cziko leaves off. He ends the book by contrasting his own purposive organism-centered perspective with both gene-centered and environment-centered perspectives, in which the notion of goal-directed organization is absent. Contemporary evolutionary psychologists are given particularly harsh scrutiny. They fully exploit the basic Darwinian insight that past natural selection can predispose certain behavioral, perceptual and cognitive biases affecting an organism's current behavior. Yet they still insist that cognition can be reduced to some function instantiated in a brain characterized (by Leda Cosmides and John Tooby) as "information processing machinery that maps informational input onto behavioral output." Their failure to consider the crucial role of middle levels of physiological and cognitive adaptation, interposed between genes and behaviors, forces them to postulate all manner of unnecessary internal complexity to compensate. The current ascendancy of a "computational" theory of mind and a gene-centered evolutionary psychology is strong evidence that we are still caught in the spell of the Newtonian metaphor and that the basic biological wisdom of circular causality and Darwinian information processing has not been fully integrated into our thinking about behavior.

The generic title will likely turn away many readers who would otherwise have much to gain from this book. This is a shame, because it is a terrific book whose critiques should be taken seriously by scientists in the field and whose alternative perspective should be debated in introductory classes preparing the next generation of behavioral scientists.