Is there a trend toward more homophones over time? What can counteract that trend?

Question

It is my understanding (correct me if I am wrong) that many homophones develop as a result of phonemic mergers.

For instance, I, like many Americans, have a "cot-caught" merger where I do not make a distinction between the "au" and the "short o". This means that for me "caught" and "cot" are homophones, as are "Don" and "Dawn".

A non-English example would be a "Kaf-Qof" merger in Hebrew where כּל ("kol" meaning "all") and קוֹל ("qol" meaning "voice") are both pronounced "kol".

My question is: Is there a tendency for the number of homophones in a language to increase over time? It seems like once a group of people stop making a distinction between two phonemes, it is very hard for them to start making a distinction again. That would imply that older versions of a language had fewer homophones and more recent versions more.

I am reluctant to buy into the hypothesis stated above because many trends in languages are cyclical with a frequency in the thousands of years. For instance agglutinating vs. isolating; it is likely that languages are one, then the other, then back again. If homophone frequency only increased over time, we would all only be grunting identical sounds after several thousand years.

The most obvious counterweight to the mergers is comprehensibility. If too many mergers occurred, the language would become more and more difficult to use to communicate. This implies a limiting factor, but it does not really explain why we haven't inherited mergers and sets of homophones from (say) three thousand years ago. Is there a counteracting force where people make (possibly new and different) distinctions between previously identical phonemes to undo some of the mergers and make what were previously homophones become unique? If so, what is that phenomenon called?

Answer 1

Phonological change is not composed solely of mergers and losses. You can have chain shifts, where the net number of phonemes stays more or less the same. It is also possible that the number of phonemes increases, in phonemic splits and additions.

Phonemic additions are usually via contact with another language. A good example in English is the phoneme /ʒ/, which was imported from Middle French loanwords; the basis for this was that the Middle English /z/ was already palatalising into [ʒ], so the "foreign" phoneme was already phonetically present in the language.

Phonemic splits are usually a result of interaction between sounds in the language, and often branch out from allophones. In British English, the FOOT-STRUT /ʊ/-/ʌ/ split of the mid-17th century and the BATH-TRAP /ɑː/-/æ/ split of 17th-/18th-century London are the most well-known. Taking the BATH-TRAP split: it was the result of phonetic lengthening before fricatives like /f/, /s/, /θ/, later even /ns/ that brought it about. However, it had incomplete diffusion through the language, preventing it from becoming a complete chain shift.

At the moment in southern British English, what has been called the WHOLLY-HOLY split is emerging, where /əʊ/ of the GOAT set has an allophone [oʊ] before the dark, almost vocalised final /l/ when in the same syllable.

A reversal of phonological merger can occur, but a net reversal and a retardation/resistance to change are different. The case of meat and mate in the English of 16th to 17th-century southern England (specifically London) is one of the better known cases of the former, where meat had been /ɛː/ at the end of Middle English, maintained that phoneme through the 16th century, and had closed to /iː/ by the 17th-century; whereas mate had been /aː/, merged into /ɛː/, and then re-emerged as /eː/ (as summarised in this 2004 paper), although it has been analysed as 1) contact between sociolects as reintroducing the phonemic split, or 2) only a near-merger occurred.

Localised resistance to change can also result in a form of reversal, although this is unlikely to be phonemic. The STRUT vowel in colloquial London English has moved from quite near cardinal /ʌ/ to a more central /ɐ/ and back towards /ʌ/ with the rise of "Estuary" English over the 20th century (see this 2006 paper).

Answer 2

Andy Wedel has done some really interesting work on factors in the maintenance of lexical contrasts (syntactic category, frequency, functional load).

Check the top three papers here.

Answer 3

Well, there are more homophones in Mandarin than in its ancestor Middle Chinese, but this is compensated for by increased use of compounding. Still, other Sinitic languages like Cantonese did not lose as many distinctions and (as far as I know) still have an increased frequency of compounds, so it’s debatable whether this is a case of compensation.

In Haudricourt’s theory of tonogenesis, which applies to Chinese, Vietnamese and many other Southeast Asian languages, final consonant mergers led to contour tones, and voiced/unvoiced initial mergers led to low/high tones, generally. So new distinctions in the rhyme arose (or were already present but became phonetic) to replace the consonantal distinctions that were lost.

Answer 4

Splitters

There are actually splitters. In the Cantonese language (my native language), there was a historical splitter around a century ago, in which the high flat tone was split into two tones (53 and 55), with unpredictable rules. This caused some homophones to lose their status as homophones and helped to distinguish the meanings. The splitter has re-merged in Hong Kong Cantonese (my dialect) so I do not even know which words were pronounced 53 and which words 55.

Adding extra information

So how do languages deal with the increase in number of homophones? In Chinese, many thousand years ago, most morphemes consisted of a single syllable. However, as the language evolves, homophones arise, and the Chinese formed compound words from the monosyllabic words, solving the problem of homophones.

In Spanish, which evolved from Latin, the word for "heart" is "corazón" while that of Latin is "cor". Such a simple word is lengthened in Proto-Romance already, by adding "-tione".

cor -> *coratione -> corazón