Eggcorn Forum

Did eggcorns exist before they had a name? More precisely, were the slips we call “eggcorns” ever tracked under a synonymous name before Pullum and Lieberman launched the modern discussion on Language Log?

If we were to conduct a systematic search for an earlier name, there are two bodies of academic literature to which we might turn: works in psychology and works in linguistics. A scan of current psychology texts, however, does not turn up useful discussions. For modern psychologists, semantic/phonetic confusion is not an object of study in itself; it is a way to illustrate other workings of the mind. The nearest that psychology comes to the eggcorn concept is the Freudian slip, which, as we have noted, is sometimes an eggcorn, sometimes not.

Modern linguistics textbooks, with their attention to language as an end in itself, come closer to describing and naming eggcorns. But the authors of these texts approach this task in an offhand manner, often tucking away relevant discussions in minor chapters on “semantic change” or “analogy.” The great victories of linguistics, the successes on which it rests its claim to be a science, lie not in semantics, but in phonology and morphology. The semantic behavior of language – the behavior that gives us eggcorns – does not lend itself, say these texts, to broad generalizations. Without laws, there can be no science.

The term in linguistics textbooks that most resembles what we call “eggcorns” on this Forum is “folk etymologies.” We spent some time on this forum discussing A Smythe Palmer’s 1882 Folk Etymology, A Dictionary. and I think that many of us emerged from that discussion with the feeling that “eggcorns” and “folk etymologies” are not exact synonyms. The two concepts are overlapping but noncongruent circles of a Venn diagram. Knowing where the circles overlap and where they don’t is difficult, however. Authors of linguistic texts, working under the assumption that reliable regularities cannot emerge from semantic analysis, don’t make much of an effort to formulate clear definitions of “folk etymologies.”

The slighting of semantics is a characteristic of modern linguistics texts; there was a time, however, when linguists, working in partnership with early psychologists, thought that they might be able to frame a comprehensive theory of semantic change. A pioneer of this semantic project was Wilhelm Wundt, often credited as the founder of modern psychology. Wundt adhered to an approach to psychology that focused on big-C Consciousness. He and his followers assumed that the basic elements of transcendental Mind could be discovered and systematically described. The field of psychology, though it started with Wundt’s assumptions, didn’t stay with his program. Competing perspectives – linguistic structuralism and behavioralism – gave psychologists and linguists new places to stand, and as the decades passed the approach taken by Wundt’s school began to look too much like the German idealism that had inspired it. The collapse of nineteenth century German idealism in the years between the two World Wars undermined whatever was left of Wundt’s authority.

In discarding Wundt, psychologists also banished a set of assumptions that gave semantics a significant role in psychological theory. Wundt had called attention, in the first decade of the twentieth century, to how a semantic classification might emerge from psychological insights when he published his two-volume Die Sprache. ( Language ). In the three decades following the publication of Die Sprache, researchers tried to extend Wundt’s early work on language into a full-blown taxonomy of semantic change.

Most of the Wundt-inspired work on semantics was done in German. The pinnacle of work on a comprehensive theory of semantics, however, appeared in English in 1931, authored by one Gustav Stern, a Swedish professor of English studies at Göteborg. Meaning and Change of Meaning (electronic copy available from the Internet Archive) was an ambitious attempt to shoehorn all types of semantic change into a single conceptual framework. The section of the book that is of interest to visitors to this Forum begins on page 230, in Stern’s section on analogical changes, where he discusses what he calls “phonetic associative interference.”

PAI (my abbreviation) happens, says Stern, when words are associated into groups that are “not historically justified,” when speakers group words by phonetic criteria rather than by proper semantic criteria. Though such phonetic groupings are not justified, they can appear so “serviceable” that they catch on and become standard speech.

Meaning and Change of Meaning makes connections, in the tradition of Wundtian psycholinguistics, between semantic change and psychological states. A speaker, Stern says, may revert to serviceable PAIs for several reasons:
.

1. Inattention. When the speaker’s attention “is deflected owing to fatigue, carelessness, hurry, etc., similarities of sound are apt to take the place of meaning….”
2. Ignorance. PAI can also be triggered when the speaker does not know the exact meaning of a word. Archaic words “are especially exposed to associative interference,” as are foreign words, whose strangeness may make a speaker “involuntarily attempt to associate it with some known word or word group in order to procure a support for its meaning.”
3. Similarity. Though speakers show a strong ability to keep homonyms separate, when “the meanings of two words approach each other, from different starting points,” they can also fall prey to PAI.

Stern acknowledges in his discussion, by the way, the overlap between PAI and popular (i.e., folk) etymology, without, however, adopting the received terminology. He probably wrestled with the same imprecisions that we noted above. Popular etymology, he says with a touch of irony, is “one of the most popular subjects of writers of popular books on philological questions.” Its very popularity was probably the reason Stern felt the need for more precision.

Stern goes on to cite specific examples of PAI, appealing to many of the same words that populate our Database and Forum: sandblind << sam-blind, standard (no “stand”), andiron << andier, shamefaced << shamefast, curry favor << curry Favel, the two sense of watershed, pester (no “pest”). He also adds a few examples that have not, as yet, been part of our discussion. “Bully,” he notes, has no “bull” in it, and “welcome” no “well.” The German word for graveyard, “Friedhof,” comes from “vrithof,” an enclosed court, even though it is almost universally associated with the German word for peace, “Friede.” “Sündflut,” the German term for the Noahic deluge, is from “sin-vluot,” a general flooding, and has no etymological connection to Sünd, sin. The French word for privation, “soufraite,” derives from Latin “suffracta” and does not take its form from “souffrir.”

Stern’s Phonetic Associative Interference, I submit, is a close synonym to the term “eggcorn.” Had Wundt and his metaphysics not been left behind when the Weltgeist turned a corner, his followers might have gone on to have the some of the same discussions that we have had on this Forum about what is and is not PAI material. Most of us, in this alternate-world scenario, would have come to the topic through psychology rather than through linguistics. We could even imagine Psychology Today having a “Bumbleberry PAI” columnist.