Is there a finite number of pronunciations for anything?

Applied Discrete Structures 2024-09-13

Below is a guest post by Corey Miller.


Azi Paybarah of the Washington Post quoted Trump as follows:

“There’s about 19 different ways of pronouncing it, right,” Trump said falsely, during a speech in Michigan on Thursday. “But Kamala is, at least it’s a name you sort of remember.”

The most interesting part of this to me is the assertion that it was a false claim. I suppose the intuition is that there are two common ways to stress Kamala, either initially/antepenultimately or medially/penultimately, so that Trump's "nineteen" is clearly hyperbolic.

What do we mean when we speak of a number of pronunciations for a word? One interpretation might be “how many ways can we represent the pronunciation of a word, as spoken by a fluent speaker interpreting the phonemes of a relevant variety of the language in question?”. Under this interpretation, three American English pronunciations of Kamala come to mind:

1. ˈkɑmələ 2. ˈkæmələ 3. kəˈmɑlə

Pronunciation 1 seems to be how the Vice President pronounces her own name and the preference of a majority of younger speakers at the Democratic National Convention. Pronunciation 2 is one I hadn’t considered, but noticed it was very popular among older speakers at the DNC; it seems to be on analogy with Pamela, which to my knowledge has only 1 pronunciation under the definition above. I assume Pronunciation 2 wouldn’t be considered an affront in the way Pronunciation 3 is, but this could be investigated further. As a final note on Prounciation 2, it is related to an interesting phenomenon I first read about in an article by Geoff Lindsey and that was further developed in my classmate Charles Boberg’s dissertation and discussed more recently by him here.

Pronunciation 3 seems to be the preferred pronunciation used by those seeking to needle the Vice President, but it seems like it can be used “innocently enough” given the predilection for penultimate stress in such words as suggested by the English Stress Rule as formulated in Liberman & Prince and elsewhere. For example, Malala (Yousufzai) seems to be a name that we hear uniquely with something like Pronunciation 3.

There is another pronunciation noted occasionally in the press for Kamala that is more “native” to the Sanskrit origins of the name, meaning “lotus flower”. Using standard American English phonemes and their IPA labels, this might be something like Pronunciation 4:

4. ˈkʌmələ

The first syllable could just as easily have been transcribed with a stressed schwa by those who admit such things. [ɐ] is used in the Sanskrit etymon for all vowels in the word in Wiktionary. This phenomenon of the “Indic short a” is also encountered in words like pundit and Punjab which are sometimes written as pandit and Panjab.

So, are there only four pronunciations? There are certainly other possibilities using IPA interpretations for various varieties of English, American and otherwise. The letter 'a' can also of course be pronounced as [ej], but perhaps using such a pronunciation in Kamala would be considered particularly outrageous. But maybe it could occur in the speech of someone less familiar with English, or someone learning to read?

Of course,  there are indefinitely many pronunciations, if we consider "pronunciations" as the articulatory and acoustic signals implementing a word, rather than IPA-ish symbols. But I assume the lay view of what it means to be a pronunciation is more along the lines of the IPA alternatives I gave above, and this is reflected in a long line of pronunciation dictionaries like Kenyon & Knott or indeed the curious symbols used in American dictionaries.

In summary, I think Mr. Paybarah was right to call Mr. Trump’s claim of 19 pronunciations false; but I think it could be litigated…


Above is a guest post by Corey Miller.