- Purchase/rental options available:
American Speech 77.3 (2002) 327-328
[Access article in PDF]
Intuition, Data, and Meaning
Ronald R. Butters
Lawrence's (2002) note raises a number of important issues about the nature of linguistic data, especially data that underlie lexicographical and lexicological research. My comment (Butters 2000) that linguists must rely upon "what native speakers tell us they mean" was certainly not intended to limit methodology to questionnaires (or linguists' intuitions): speakers also "tell" us a great deal through speaking and writing. What I meant to convey was rather that prescriptive arguments about the aesthetic value, etymology, or "logic" of a linguistic form are not valid data. Certainly, questionnaires that revealed only gross facts about English plurals and Japanese iru would be inadequate—nor would any native speaker tell us that Lawrence's observations about the subtleties of English plurals are anything but correct.
Lawrence's third example is interesting because it points up a paradox in my position—a paradox that I did not mention. Though "logic" is not a methodologically sound basis for the linguist to use in reaching linguistic conclusions, logical arguments may nonetheless play an important role in [End Page 327] convincing segments of the public at large that a word or phrase "really" means one thing or another, or that a particular linguistic construction (e.g., multiple negation, split infinitive) is incorrect. The duty of the linguist is to report whatever meanings and rules people have—and, for that matter, to inform readers of dictionaries and scholarly articles if native speakers disagree with each other (as they do, for example, about the meaning of such words as century and millennium, or the phrase to beg the question, or about the preferred past participle of drink).
Lawrence's assertion that "the meaning of a word can only be arrived at inductively" is, however, far too strong. One cannot get very far performing inductive analyses of actual usage without drawing upon linguistic intuitions. For example, did Lawrence learn that the English plural is used in expressions like 0.8 pounds only through empirical data, or did he also consult his own intuitions—if not by thinking long and hard about it, at least by confirming that his empirical data were not merely performance errors, stylistic variations, or dialect forms? Moreover, direct questioning of subjects is sometimes the only practical way of ascertaining what is really in their heads, for example, when dealing with rare syntactic forms. (See, e.g., Butters 1973; Butters and Stettler 1986).
In the end, practicing linguists must make use of all three kinds of data—the linguists' intuitions, the informants' intuitions, and objective empirical data—in arriving at the most accurate conclusions about the meanings that exist "in the human brain." Where there is variability in the data, as in the case of millennium, that variability must be reported as well, regardless of what reasons people may have for believing what they do.
Butters, Ronald R. 1973. "Acceptability Judgments for Double Modals in Southern Dialects." In New Ways of Analyzing Variation in English, ed. Charles-James N. Bailey and Roger W. Shuy, 276-86. Washington, D.C.: Georgetown Univ. Press.
———. 2000. "The 'Real' Meaning of Millennium." American Speech 75: 111-12.
Butters, Ronald R., and Kristin Stettler. 1986. "Existential and Causative have . . . to." American Speech 61: 184-90.
Lawrence, Wayne P. 2002. "Do Native Speakers Know What Words Mean?" American Speech 77: 325-27.