Varieties of Ordinary Meaning: Comments on Kevin P. Tobia, 'Testing Ordinary Meaning'

Abstract

This paper critically examines Kevin Tobia’s forthcoming paper Testing Ordinary Meaning: An Experimental Assessment of What Dictionary Definitions and Linguistic Usage Data Tell Legal Interpreters. Please note that this is a work in progress and that I plan on posting a revised version shortly (for some value of “shortly”). Although I believe that Tobia’s analysis is problematic in multiple respects, I will focus here on the criticism that I think is most important. That criticism challenges not only Tobia’s conclusions, but also his paper’s central premise. In particular, I dispute Tobia’s conclusion that the results from his Concept-Condition experiments establish that corpus linguistics is an inaccurate tool. (Those experiments were the ones in which test subjects were asked, e.g., whether a golf cart is a vehicle.) As I will explain, Tobia’s analysis is based on an unexpressed assumption, and if that assumption is invalid, Tobia’s conclusions are invalid, too. The assumption is that in the context of legal interpretation, “ordinary meaning” means only one thing. But that assumption is unfounded. “Ordinary meaning” is not a technical term in linguistics and to the extent that it has a technical meaning in philosophy of language, I’m unaware of that meaning having had influence on ordinary meaning as a legal concept (which dates back at least to Blackstone). So in its use in connection with the practice of legal interpretation, “ordinary meaning” is a purely legal term. Within that practice, therefore, the meaning of “ordinary meaning” is determined by what the courts say it means and by how they apply the concept in particular cases. And when we look at the caselaw, we can see that there are multiple conceptions of what constitutes ordinary meaning. I will focus here on two of those conceptions (one of which has two subcategories). The differences between these conceptions of ordinary meaning have practical consequences. The outcome in a given case can depend in part on which conception is invoked (either explicitly or implicitly). Similarly — and crucially — the choice of an appropriate interpretive methodology will vary depending on which conception of ordinary meaning one is assuming. Tobia’s Concept-Condition experiments are suited for use with respect to one of the conceptions that I will discuss but not with respect to the other. And the suitability of corpus linguistics presents essentially the opposite situation: it is suitable with respect to the conception of ordinary meaning for which Tobia’s Concept-Condition methodology is unsuitable, and for the most part unsuitable for the conception for which his methodology is suitable. This means that to use Tobia’s experimental methodology to evaluate the accuracy of corpus linguistics is to fall prey to a category error. Tobia’s comparison of the corpus-condition results against those for the concept condition proves nothing about the accuracy of corpus linguistics.