Regular readers of this blog will have realized, at least subcutaneously, that the author has a prescriptivist bias and does not shrink from proscribing certain contemporary usage as incorrect when it violates the traditional norm. This bias still rubs almost all professional linguists the wrong way, owing to the fact that in their hearts there lurks the old behaviorist antipathy to prescribing––as against faithfully describing––whatever usage is extant. In post-war American behaviorist linguistic circles, the idea that language usage should be viewed through the prism of correctness was rejected as unscientific, an attitude epitomized by the Romanist Robert A. Hall Jr.’s 1950 book, Leave Your Language Alone!
But the social science approach to language norms, no matter how it is couched or what terminology it recognizes, ultimately comes a cropper when confronted with the undeniable presence of the criterion of correctness in every language user’s Sprachgefühl, or what a prominent contemporary theoretician of historical linguistics (nomina sunt odiosa) calls “metagrammar.” This includes the knowledge every speaker possesses of what constitutes infractions of the linguistic norm, whether or not a given language has a codified standard. This metagrammatical superstructure, as it concerns correctness, is necessarily present in every act of language use––most often in the null mode–-irrespective of school learning or modern-day usage manuals.
A useful illustrative comparison is with music. When my father taught me as a child the rudiments of chorale writing, he instructed me inter alia to “avoid parallel fifths.” Writing such sequences was simply an error. It violated the norms of chorale writing as codified in books on harmony and composition. This was not a matter of scalar values, deontic logic, or “norms of appropriateness.” Even less so was it dependent on taste or preferential behavior.
To continue in the same vein, a cellist who plays a wrong note in a Bach suite cannot justify it by appealing to creative freedom: the note is either right or wrong, either what Bach wrote or not what Bach wrote. No interpretation of Bach licenses wrong notes. Any minimally musically-literate listener would know when the cellist played a B flat instead of a B natural.
Returning to language, naturally, attitudes toward norms and correctness vary in strength across the speech community depending on speakers’ education and personal preferences. Contemporary dictionaries like The American Heritage Dictionary of the English Language (AHD) routinely reflect this spectrum explicitly by citing the results of polling data from “Usage Panels.” For example, the verb err has a traditional pronunciation which is rapidly disappearing in American speech, as registered in AHD‘s report immediately following the entry:
“Usage Note: The pronunciation (ûr) for the word err is traditional, but the pronunciation (er) has gained ground in recent years, perhaps owing to influence from errant and error, and must now be regarded as an acceptable variant. The Usage Panel was split on the matter: 56 percent preferred (ûr), 34 percent preferred (er), and 10 percent accepted both pronunciations.”
When my gym trainer (a native speaker of American English in his thirties) pronounces err in the common phrase “to err on the side of caution” to rhyme with air, I immediately register it as incorrect, a violation of the (traditional) norm, even though I know full well that the pronunciation is not idiosyncratic. His phonetic trait does not lower him in my estimation of him as a trainer, but for better or worse it does automatically align him with those who are ignorant of (or knowingly ignore) the norm that characterizes my own speech.
MICHAEL SHAPIRO
I switched from ‘er’ to ‘ûr’ around a year ago because:
1. My mother corrected me,
2. The ‘ûr’ pronunciation distinguishes “to err” from “to air,” and
3. Let English preserve at least two full-fledged words (‘awe’ and ‘err’) each of which consists in a single pure (non-dipthongal) sound. Well, at least the way that I speak English.
Objection There’s a standard whereby the root in ‘err’ and ‘errant’ should be pronounced the same.
Reply So it seems but, if so, then an exception should be made in this case for the above-stated reasons, or at least the linguistic reasons among them. If anything, the pronunciation of ‘errant’ should be changed from ‘er?nt’ to ûr?nt’.
Peirce the fallibilist had, of course, a special relationship with both the ‘ûr’ sound and the idea of ‘to err’, though I wonder whether he pronounced any of the r’s there or in his names – somewhere he hinted that he dropped his g’s from “-ing” too; I think it was where he discussed physically recovering, from Aristotle’s environs the sound of Aristotle’s voice and reputed r-dropping.
Back before the Internet took off, a linguist wrote a NYT column that had me ranting, when he argued that linguistic standards are socially illegitimate because they’re value-based rather than “scientific.” That kind of argument might find science itself illegitimate. As if Chevy Chase had said to him, “Be the strawman, Danny, be the strawman.” Not that the linguist’s name was Danny, I can’t remember it.’ Of course it’s fun to remember extreme examples, probably not all descriptivists are so far out.