British Academy: The UK's National Academy for the Humanities and Social Sciences
Enquiry, Evidence and Facts: An Interdisciplinary Conference
Generalizations and Evidential Reasoning
Professor Terence J. Anderson
University of Miami School of Law, 1311 Miller Drive,
Coral Gables, Florida 33124, U.S.
An abstract presented to the conference
‘Enquiry, Evidence and Facts: An Interdisciplinary conference’
at the British Academy, London, on 14 December 2007
Biography
Terence J. Anderson, Professor of Law, received a B.A. from Wabash College in 1961 and a J.D. from the University of Chicago Law School in 1964. Before joining the University of Miami faculty in 1976, he served for two years as a regional local courts commissioner in Malawi, Africa, practiced for seven years with a law firm in Chicago, and taught for three years and served as academic dean for one at the Antioch School of Law in Washington, D.C. Since joining the UM Law faculty, Professor Anderson has continued to participate in public interest litigation. Most notably, from 1982 until 1991, he represented former U.S. District Judge, now Congressman, Alcee Hastings in all litigation relating to his office, including his impeachment trial before the United States Senate in 1989. During the 1994-95 academic year, Professor Anderson was a fellow at the Netherlands Institute for Advanced Studies in Wassenaar. He is the author of The Battles of Hastings: Four Stories in Search of a Meaning (NIAS 1996); On Generalizations I: A Preliminary Exploration, 40 So. Tex. L. Rev. 455 (1999), “Wigmore Meets the Last Wedge” in Evidence and Inference in History and Law (William Twining and Ian-Hampshire Monk, eds. 2003), and co-author (with David Schum and William Twining) of Analysis of Evidence (2d ed. 2005).
Abstract
Every argument must be based upon a generalization that can be stated as a major premise. In deductive logic, the generalization is a universal proposition and is ordinarily stated – to use a classical example: All men are mortal; Socrates is a man; therefore, Socrates is a mortal. In (informal) inductive logic, the generalization is less than universal and ordinarily is implicit – to paraphrase the Pollock [1] example: In most instances an object that appears to be red is an object that is in fact red; object x appears to red; it is highly probable that object x is in fact red. In abductive logic, the conclusion is speculative and the generalization is the hypothesis that would need further testing before it could be used to support an inference – to use a 9/11 example: Foreign nationals from the middle east, who had no prior flight training or experience, enrolled in American flight schools seeking to learn how to fly large commercial airplanes; if Mideastern terrorists were planning to hijack American commercial airliners and use them as weapons, they might send terrorists to flight schools in America to learn how to fly commercial airliners; Mideastern terrorist might be planning to hijack American commercial airliners and use them as weapons.
Evidential reasoning is reasoning grounded in evidential data that is based upon rationally acceptable generalizations. Evidential reasoning employs all three forms of logic, but is primarily inductive. The relationship between the supporting proposition or propositions and the inferred proposition can be restated in a quasi-deductive form by identifying the generalization upon which the inference depends. The degree of certainty or the probability of the inferred proposition depends upon degree of certainty of the generalization. Disagreement about the strength of an inferential argument represents a disagreement about the certainty of the generalization that supports it. A claim that an inference is unjustified is a claim that there is no acceptable generalization that would justify it.
Evidence is and only can be a datum or a proposition that alters the probability, positively or negatively, of a proposition to be inferred. In order to demonstrate that an evidential proposition is relevant, an analyst must be able to identify and articulate a generalization that justifies the claim that the evidential proposition alters the probability of an inferred proposition. An inferential proposition becomes an evidential proposition if it is offered as “evidence” tending to support or negate another inferred proposition. For example, a policy based upon evidence becomes an evidential proposition when it is used in conjunction with other evidential propositions to support a conclusion or another inferential proposition.
Notes
1 John Pollock Cognitive Carpentry (MIT Press 1995)
2 I recognize that the second clause introduces an element of circularity into the argument. It is necessary, however, to exclude generalizations that are unacceptable even though their adherents might claim that they could be grounded in evidential data. For example, an argument based upon a generalization about the superiority of one race or ethnic group over another would not be acceptable to those engaged in a discipline which required that conclusions be rationally justified. Other examples might be drawn from astrology or religion.
3 Value-based generalizations can also be classified along a scale. At one end are untestable generalizations based upon acceptable beliefs such as political or constitutional generalizations. At the other, are generalizations based upon unacceptable beliefs such as generalizations based upon the alleged superiority or inferiority of a race or an ethnic group or upon gender. Generalizations based upon religious beliefs would fall in the mid-range. A value based generalization grounded in acceptable beliefs may be used in a hybrid argument based in part on evidential reasoning and in part on value based generalizations.
4 See, e.g., Bradwell v. Illinois, 83 U.S. 130 (1872) (affirming decision by Illinois Supreme Court that women could not be admitted to the bar), esp. 141-42 ("The paramount destiny and mission of women are to fulfil the noble and benign offices of wife and mother. This is the law of the Creator. And the rules of civil society must be adapted to the general constitution of things, and cannot be based upon exceptional cases.") (Bradley, J., concurring).