Document Type

Book Chapter

Publication Date



The process of normalizing a legal rule requires a drafter to indicate where the intent is to be precise and where it is to be imprecise in expressing both the between-sentence and within-sentence logical structure of that rule. Three different versions of a legal rule are constructed in the process of normalizing it: (1) the logical structure of the present version, (2) the detailed marker version, and (3) the logical structure of the normalized version. In order to construct the third version the analyst must formulate and answer specific questions about the terms that are used to express the logical structure of the first version that relates the constituent sentences marked in the second version. Questions about the two types of logical structure may be of two different kinds: (1) direct questions about the interpretation of terms that express each type of structure, and (2) indirect questions by means of hypothetical situations that indicate how the terms that express structure are intended to be interpreted. Direct questions are generated from natural language terms that are used to express structure by a series of transformations that use progressively more detailed defined structural terms and that culminate in structure that is expressed entirely in the defined structural terms of the basic normalized form. Arrow diagrams accompany these direct questions to help teach normalization to those unfamiliar with it. Examples of such direct questions, as well as examples of indirect ones, are provided with respect to normalization of section 2-207 of the Uniform Commercial Code. Indirect questions are generated about hypothetical situations that involve various appropriate combinations of conditions expressed in the rule that lead to the various mentioned results. This kind of question may be easier for an expert to respond to and thus be a better vehicle for eliciting the expertise of such a person. It is possible that some computer assistance can be provided in generating direct questions, but less likely for indirect questions. Furthermore the number of indirect questions generated my be unmanageably large and require too much human assistance to be practical. In this chapter the feasibility of such computer-aided question generation will be explored to determine to what extent it can facilitate the normalizing of legal rules.


Computer Power and Legal Language: The Use of Computational Linguistics, Artificial Intelligence, and Expert Systems in the Law by Charles Walter, ed. Copyright © 1988 by Charles Walter. All rights reserved. Reproduced with permission of ABC-CLIO, LLC, Santa Barbara, CA.