At the eggcorn’s edge

Two cartoons from the 11th, passed on to me by Benita Bendon Campbell, both turning on (mis)perceptions and (mis)interpretations:

(#1) Family Circusguest towel / guess towel (cf. brand-new / bran-new)

(#2) Luanncake stand / keg stand (cf. acorn / egg corn)

Family Circus. The conventional compound is guest towel ‘towel for guests (as opposed to members of the household)’. Apparently not in shorter dictionaries, but it is in OED2:

guest-towel n. a small hand-towel intended for visitors’ use. [1st cite 1921]

A child — Dolly in the strip is 5 — might have encountered  the compound, but might not be particularly familiar with the use of guest in it; so for her it would be an opaque expression with a mystery first element. On the other hand, the verb and noun guess will be everyday words for her.

Then, although guest towel and guess towel are phonologically (and phonetically) distinct if pronounced carefully as a sequence of two words separated by a pause —

/gɛst taw(ǝ)l/ guest towel vs. /gɛs taw(ǝ)l/ guess towel

essentially no one would ever pronounce them that way; in connected speech, both are pronounced:

[gɛs.taw(ǝ)l] or [gɛ.staw(ǝ)l]

(where the period indicates syllable division). Guest loses its final t, by the much-studied process of “final t/d-deletion” in English (to give it one of its common names); there is a Page of postings about the process on this blog.

The way is then open for Dolly to analyze such a pronunciation as a realization of a compound guess towel — which might not make a whole lot of sense, but at least has familiar parts (and maybe you could make up a story that makes sense of it).

This would all be a commonplace story of eggcorning, except for the fact that Dolly in the strip is represented as saying guess towel rather than guest towel, even though the two expressions are homophonous: Dolly is treated as speaking in spellings, producing spellings rather than sounds (in the air, in speech balloons). This is a not uncommon practice in cartooning — and it builds on the folk belief in literate cultures that orthographic representations are the real, true language, while spoken language is only an imperfect approximation of this — but, still, it smells of paradox.

Spelling does have a significant role in all of this: eggcorns are usually detected in spellings that evidence a non-standard analysis. If Dolly had written “Billy’s usin’ the guess towel”, there would be no whiff of paradox at all, just evidence that she took [gɛs] to be an occurrence of the word guess rather than the word guest.

Then her misunderstanding would be seen as similar to this example from the Eggcorn Database:

brand-new » bran-new

Classification: English – nearly mainstream – final d/t-deletion

Analyzed or reported by:Ben Zimmer (Word Routes, Visual Thesaurus, Dec. 5, 2008)

Brand-new dates to 1570, but the variant bran-new was already appearing less than a century later. See the Word Routes article for a full analysis, including this eggcornic “etymythology” given by a Wiktionary contributor:

The term ‘brand new’ or ‘bran new’ was when new items were packaged up with unwanted bran grain in the 18th Century to protect the object during transit. When the item was unpacked, the owner would often find traces of bran in the item. Hence the term.

| Comments Off link | entered by Ben Zimmer, 2008/12/08 |

If the two compound adjectives are pronounced carefully as a sequence of two words separated by a pause, then they are distinct:

/brænd nu/ brand-new vs. /bræn nu/ bran-new

Essentially no one would ever pronounce them that way; in connected speech, both are pronounced:


Crucially, brand loses its final d, via final t/d-deletion.

Luann. Here, standard spellings are used to convey the lexical intentions of the two speakers, but this is less jarring than in the Family Circus, for two reasons: the speakers are literate adults, so what’s given is at least the way they’d spell things (while Dolly is presumably not yet up to distinguishing guest and guess orthographically); and keg in keg stand and cake in cake stand are surely not fully homophonous for them, but only phonetically similar.

Standing alone, the words are phonemically distinct:

/kɛg/ keg vs. /kek/ cake

Syllable-final /g/ is very weakly voiced in American English, but it’s still distinct from syllable-final /k/, which is voiceless and glottalized and is very often realized as a glottal stop [ɂ]. Meanwhile, the /ɛ/ of keg picks up an offglide (schwa-like in my speech) before syllable-final /g/, and for some American speakers tenses and raises towards [e]; for some of these speakers, the raising, tensing, and offgliding take the vocalic nucleus all the way to [eɪ] (the primary allophone of the vowel phoneme /e/, as in cake), or at least very close to it. The nouns keg and cake are phonetically — articulatorily and acoustically — similar for everyone, very close for some people, though remaining distinct for almost everyone.

But phonetic similarity is the basis for a wide range of mishearings, which appears to be what’s happened in #2: one guy visualizes using the high chair as a keg stand for keggers, and the other hears him as having said he wants to use it as a cake stand.

So: not an eggcorn. But not unrelated to matters eggcornic, since the keg > cake mishearing is phonetically almost exactly the reverse of the acorn > eggcorn reanalysis that gave us the name eggcorn for the phenomenon in the first place. That reanalysis might have originated in a mishearing — a mishearing that would then have struck people as semantically felicitous. (No such extra value accrues from mishearing keg stand as cake stand.)


Leave a Reply

%d bloggers like this: