Steel, Daniel P.: Across the Boundaries: Extrapolation in Biology and Social Science.
Steel's focus in Across the Boundaries is to deal with the methodological problems associated with making extrapolations in the social sciences via a comparison with the manner in which extrapolation works in the biological sciences. Extrapolation is understood as inferential reasoning from one set of causal relationships into another distinct set, thus from one environment and situation into another.
Steel looks at various strategies for handling problems in extrapolation and the limitations of such a form of inductive reasoning. He focuses upon a mechanisms approach which "rests on the intuition that knowing how a cause produces its effect ... can provide a basis for extrapolation." Steel defines the mechanisms under investigation as "sets of entities and activities organized so as to produce a regular series of changes from a beginning state to an ending one." Identifying and evaluating mechanisms is the crucial first step to building a sustainable extrapolation.
Extrapolations are actually ubiquitous in the biological and social sciences, although rather unremarked upon as problematic. It has likely taken the relaxing of deeply habitual materialist reductivism in epistemology to allow the importance of inductive extrapolation to be noted. Questions such as how a test drug which showed positive results in a controlled test group of mice will work on a larger heterogeneous population of people or whether a welfare program in one locale can be effective in another dominate the applied practice of biological and policy fields. Clearly a Lockean demand for justified true belief and modern thought's infatuation with finding and accepting only universal laws as reasonable fail to comprehend the rational necessity of informed extrapolation. Modern thought's seduction by the successes of physics left it ill-equipped to recognize the conditions present for reason in practical and largely inductive fields.
Steel steps into the void of neglect given to the logic of extrapolation by putting forward a well-designed theory to explain how it can be successfully done. The theory rests upon what he calls comparative process tracing (CPT) and is offered in opposition to analogical approaches. Steel explains it thusly: first, use experimentation to determine the stages of causal mechanisms involved in your model organism or test group. Second, compare stages discerned in the target mechanism to those found in the model with a focus on those points where they are most likely to have significant differences. Finally, evaluate whether these differences are decisive enough to make extrapolation impossible or if the extrapolation can be justified.
The two primary challenges to extrapolation are the "extrapolator's circle" and the "problem of difference." The first challenge is the problem that since a model is being used as a source for further understanding of a separate mechanism, this entails limited and partial information about the target mechanism. But under such a limitation it would seem impossible to determine the suitability of the extrapolation since it would require more than the partial and limited knowledge that has forced one to attempt to extrapolate in the first place. The problem of difference is the fact of heterogeneity in different mechanisms and groups. The challenge is determining the causal relevance or lack of in the differences between the two mechanisms. Steel shows how CPT can establish the positive and negative causal relevance of qualitative differences between the model and target mechanisms. CPT is a labile approach which does respond adequately to both of the main challenges to extrapolation.
The great value in this book is in dealing with the sciences as lived practices, as containing as much intuition and induction as the more rationally satisfying inferences and deductions. Extrapolation is a rationally necessary and justifiable component of the success of the biological and social sciences and Steel has gone far in examining just what is entailed in secure and reliable extrapolation.--Brian J. Fox, Suffolk County Community College.
|Printer friendly Cite/link Email Feedback|
|Author:||Fox, Brian J.|
|Publication:||The Review of Metaphysics|
|Article Type:||Book review|
|Date:||Sep 1, 2012|
|Previous Article:||Schweber, Silvan. Nuclear Forces: The Making of the Physicist Hans Bethe.|
|Next Article:||Taylor, Robert. Reconstructing Rawls: The Kantian Foundations of Justice As Fairness.|