- Argument from Analogy in Law, the Classical Tradition, and Recent Theories
Argument from analogy is a common and formidable form of reasoning in law and in everyday conversation. Although there is substantial literature on the subject, according to a recent survey (Juthe 2005) there is little fundamental agreement on what form the argument should take, or on how it should be evaluated. The lack of conformity, no doubt, stems from the complexity and multiplicity of forms taken by arguments that fall under the umbrella of analogical reasoning in argumentation, dialectical studies, and law. Modeling arguments with argumentation schemes has proven useful in attempts to refine the analyst’s understanding of not only the logical structures that shape the backbone of the argument itself, but also the logical underpinning of strategies for evaluating it, strategies based on the semantic categories of genus and relevance. By clarifying the distinction between argument from example and argument from analogy, it is possible to advance a useful proposal for the treatment of argument from analogy in law.
Analogy in Legal Reasoning
Analogy is one of the most common forms of reasoning in law (see Hage 2005). Through analogical reasoning, legal inference is drawn from one case [End Page 154] that has already been classified and is assessed to another case on the basis of similarity or dissimilarity. This form of argument is widely used to fill “the gap between facts and rule” (Weinreb 2005, 92). In other words, analogy is used to apply general legal rules to cases not directly falling under the classifications of the rule.
The approaches examined in this section show that there are different types of analogy based on different logical structures, and these differences are key to evaluative strategies. While Klug distinguishes between different kinds of analogy, Alexy and Weinreb focus their studies on the species of analogical reasoning most common in law, namely the imperfect inference leading to presumptive conclusions. This kind of reasoning, according to Brewer (1996) for instance, is often formalized as a syllogistic argument; however, several other authors maintain that analogy cannot be reduced to a deductive form. This section shows that a defeasible approach to analogy in legal arguments is stronger than a deductive or inductive one. Some authors have identified all kinds of defeasible reasoning, that is, reasoning leading to merely plausible conclusions, with inductive reasoning (see for instance Grennan 1997), or with deductive reasoning (see for instance Groarke 1999). On our view (see Walton 1996), defeasible reasoning should be clearly distinguished from induction and deduction. Defeasible reasoning can be described as an alternative to (apodeictic) deductive reasoning, which stems from a universally quantified premise (like “All men are mortal”). While apodeictic deductive reasoning is based on the passage from the universal to the particular, in default reasoning the link between premises and conclusion is not guaranteed by quantification. The premises are simply endoxical (acceptable), and the reasoning is grounded on patterns of reasoning called maxims, and represented in modern argumentation theories as argumentation schemes. For instance, consider the following reasoning:
Tweety is a bird.
Therefore Tweety flies.
In this case, the conclusion is supported by the implicit premise that “birds (usually) fly.” However, this premise is not an absolute truth: it is only endoxical, and it warrants only a defeasible conclusion, that is, a conclusion that might be wrong. The link between premises and conclusion is warranted by some commonly shared principles of inference, like “the property [End Page 155] is predicated convertibly of the subject” (Aristotle, Topics, I, 5). These rules of inference have been developed in modern argumentation theories in the so-called argument schemes, abstract argumentation patterns combining a semantic principle with a logical axiom (see Walton, Reed, and Macagno 2008). For instance, the argument mentioned above can be analyzed using the argument from verbal classification (Walton, Reed, and Macagno 2008, chapter 9; for further discussions of this scheme see Hastings 1963, 36–52; Kienpointner 1992, 250–52; Walton 2006, 129):
|Definition Premise||Classification Premise||Conclusion|
|a has property F.||For all x, if x has property F, then x can be classified as having property G (if G is a semantic property of F...|