Cardozo Law Review
This Article develops the concept of a "theory of uncertainty" as a tool for warranting non-monotonic inferences. It proposes that theories of uncertainty have an explanatory function in inference that is roughly parallel to the role played by theories of causation in the empirical sciences. An adequate theory of uncertainty provides a framework for explaining the warrant for non-monotonic inferences by explaining the possible sources of error in such inferences. A theory of uncertainty should explain the extent to which the available evidence warrants a particular conclusion and the kinds and degrees. of uncertainty associated with that conclusion. Such a theory may also help predict how additional evidence might affect the warrant for that inference. The proposal is that fact finders develop theories of uncertainty for particular inferences from particular bodies of evidence, and that the adequacy of those theories plays an important role in warranting the inference itself.
This analysis is presented in three parts. Part I discusses theories of uncertainty generally, and some of the characteristics such theories have in common. Part II examines a theory of uncertainty for conclusions warranted by scientific data. Such a theory is particularly useful in evaluating conclusions about generic causation and in assessing whether proffered scientific opinions are admissible under Daubert and General Electric Co. v. Joiner. Part III discusses the evidence charting theory of David Schum, which is especially promising for inferences about unique historical events. These two extended examples illustrate how the concept of a theory of uncertainty is useful in analyzing the non-monotonic inferences important in legal fact-finding.
Vern R. Walker,
Theories of Uncertainty: Explaining the Possible Sources of Error in Inferences, 22 Cardozo L. Rev. 1523
Available at: http://scholarlycommons.law.hofstra.edu/faculty_scholarship/166