I have been meaning to write about homonyms for quite some time, and now this time has come. Some clarification of terms is needed. English is full of homographs, or words that are spelled the same but pronounced differently. They are the bane of a foreigner’s life: bass, bow, sow, lead, and so forth (however, not only foreigners suffer from such traps: one has to be a teacher to know how many times even college students believe that the past tense of the verb lead is also lead, like read ~ read). Then there are homophones: words like by ~ bye ~ buy; sew ~ so, two ~ too ~ to (add tutu), and so forth. They are spelled differently but sound the same. Finally, we find perfect matches: match¹ (a box of matches), match² (I am not his match), and match³, as in boxing match. The tenuous connection between match² and match³ in our language intuition is a problem to which we’ll briefly return below. Words like match^{1, 2, 3} are called homonyms. As long as we are dealing with oral speech, homophones are indistinguishable from homonyms.

Some homonyms are more perfect than others. For example, ash (a tree) and ash (the residue left after burning) sound the same in the singular and in the plural. By contrast, lie (tell falsehoods) is a weak verb (lie, lied, lied), while lie (as in lie awake) is strong (lie, lay, lain); they are homonyms only in some of their forms. I will skip the niceties of classification and bypass the question whether words like love (noun) and love (verb) should be called homonyms. (Hardly so!) Here we are interested in one question only, to wit—why so many obviously different words are not distinguished in pronunciation, or, to change the focus of the enquiry, why language, constantly striving for the most economical and most perfect means of expression (or so it seems), has not done enough to get rid of those countless ambiguities.

Match girls. (Image by Maria Kirilenko via Flickr)

The answer was given long ago. In most cases, homonyms occur in non-overlapping contexts, so that misunderstanding never occurs, or the situation disambiguates the clash. Even when homonyms meet and we say that the ash tree was burned to ashes, the message remains clear. We may appreciate or wince at the play of words (I prefer the first option) but have no trouble deciphering the message. Homographs are a terrible nuisance (and this is where a moderate spelling reform might improve the situation), but homophones are the joy of wits and poets. One can lie in bed and lie to the doctor about a nonexistent illness. Some literary periods revel in this type of language. Shakespeare lived in one of them. His puns fill a volume, and not all of them are "bawdy." In the sonnets about the woman whom he both loved and despised, we find such a couplet: "Therefore I lie with her, and she with me. / And in our faults by lies we flatter’d be" (Sonnet 138). In Old English, the two forms were quite different, namely licgan and lēogan. But phonetic change, that great leveler, gradually effaced the differences.

Let us look at some examples from a historical point of view. Even in the best popular books on the development of the vocabulary of Modern English, such as Greenough and Kittredge’s Words and Their Ways in English Speech and Bernard Groom’s A Short History of English Words, one won’t find a chapter on the rise of homonyms, though the subject is quite worthy of discussion. Why is English so full of homonyms? First, by pure chance. For example, ash (the tree name) and ash (as in sackcloth and ashes) have, from a historical point of view, nothing in common, but they were close even in Old English (æsc and æsce). When the second word lost its ending, the harm was done. Nothing is more noticeable in the history of the Germanic languages than the loss of postradical syllables. The oldest runic inscriptions have preserved very long words. Even in the fourth-century Gothic their cognates are sometimes a syllable shorter. More shrinking happened in Old English, then apocope chopped off most endings in the Middle period, and today we are left with a storehouse of monosyllables: be, come, do, put, man, child, calf, good, and the rest. The Gothic for they would have sought was the unwieldy form so-kei-de-dei-na (five syllables). We need four monosyllables to express the same idea. And of course, we keep shortening even our recent words. That is why doc "doctor" has become a homophone of dock, ad "advertisement" is indistinguishable in pronunciation from add, and Mike can freeze in front of a mike.

The musicians are bowing. They will bow later. (Image via pxhere)

Monosyllables cause the greatest trouble to etymologists. For example, take bob, as in bobbed hair, and bob "move up and down." No one knows for sure where such short expressive verbs come from (especially such as begin with and end in the same consonant: tit, tat, gig, kick, and so forth). Are bob¹ and bob² different words or two senses of one word? After all, any impulsive movement can be described by using the syllable bob! If I say bob along! you will probably understand me, though you have never heard such a phrase. Cob is "round head; testicle; a mixture of unburned clay and straw; a short-legged riding horse," and many more things. They are "symbolic formations": say cob and conjure up a picture of something round, never mind what! Dictionary makers wonder whether to list them under one entry, with 1, 2, 3, etc., or separate them. Did a protoform exist that split into many fragments, or are they unrelated products of language creativity, with people constantly saying bob, bug, box, cob, pop, etc., and endowing them with the senses that suit them at the moment?

But sometimes we can follow the process of disintegration. It would probably be counterproductive to treat love (noun) and love (verb) as homonyms. But what about fall "autumn" and spring as opposed to the verbs "to fall" and "to spring"? Clearly, the fall is the time of "falling" and spring is the time of "springing," but this connection is rather vague, and again one wonders whether today we have homonyms or different words. And remember match! An etymologist treats match¹ (a perfect match) and match² (a boxing match) as two senses of the same word but what about "naïve" speakers?

Lying in bed. (Image via piqsels)

I am returning to the verb to lie. Gothic has liugan "to lie" (that is, "to tell falsehoods") and liugan "to marry a man" (not just "to marry"!). This incompatible pair has troubled philologists for a long time, though the great and nearly infallible Jacob Grimm offered a good explanation that covered both senses. Some people accepted his idea, but most did not. Grimm believed that in the oldest languages, homonyms should in general be analyzed with caution and that in most cases an attempt should be made to trace them to the same protomeaning. Gothic interests few of our readers, and I’ll not address the problem, the more so as I have written an article about this enigmatic pair in which I tried to defend Grimm’s idea, but to agree with Jacob Grimm, we must understand what the speakers of Old Germanic meant when they referred to lying and what the ceremony of an ancient wedding entailed.

Alas, the vocabulary of a language like Modern English is so rich and so multifaceted that Grimm’s dictum can be applied to it in exceptional cases. "In our faults we flatter’d be."

Feature image: Chandos portrait, National Portrait Gallery, via Wikimedia Commons (public domain)