Has anyone ever ranked the prevalence of phones by number of speakers worldwide?

Question

I'm interested in knowing the most-used and least-used phones worldwide. According to Wikipedia, the IPA charts about 140 pulmonic consonants, 80 non-pulmonic consonants, 30 co-articulated consonants, and 33 vowels. I would like to know how common these different phones are, ranked by number of people in the world who speak any language using a given phone.

For example, if there were a phone that appeared exclusively in English and Mandarin Chinese, it would have about 2.468 billion speakers, while a phone that appeared in both those languages plus Hindi would have about 3.068 billion speakers, and so on.

At the cost of making the whole thing even more complex to calculate, it might be interesting to also weight the rankings by number of times a given phone appears in the, say, one thousand most common words in each language.

Answer 1

This has not been done. There is a basic informational limit on what questions you can ask – you cannot get frequency of use data for phonemes for most languages of the world. Indeed, getting reliable phoneme lists for most languages is challenging. You can get information for some languages (English, Spanish, Chinese...) but not Tiriki. However, there are only about 100,000 speakers of Tiriki, so lack of information on Tiriki would have a negligible impact on the count of "most frequent". You might then assemble phoneme lists for the 100 languages with the most speakers and get a ranking that is about those languages. The one concern would be that there might be sounds that are frequent in widespread low-population language phyla. For example, about half the world speaks Austronesian or Niger-Congo languages. But while there are thousand(s) of languages in those phyla, few of the languages in those phyla have "many speakers" (Javanese, Indonesian, Yoruba, Igbo and Sundanese are the members of those phyla in the top 50, population-wise).

Another approach is driven by specific phonemes. Here is a map of languages thought to have [ɓ], from PHOIBLE (they report 300 languages). One problem is that they missed quite a number of languages (e.g. Matumbi, Pare, Zina and other languages that they don't know about – plus, "Shona" is really a language family, not a single language). One reason is that they cull information from very many written sources and people completely disagree on "phonemes". Information on "phones" is even less reliable (the vast majority of sources don't tell you what the phones of the language are, in fact it's hard to get a list of phones for English – is there a phone [ɬ] in English, as in "play"? What about [ɹ] versus [ɹʷ]).

Phones in fact encompass not just letters like [ɪ], it includes all of the variants with myriad diacritics, which are reported in some specialized publications, but are usually omitted when people talk about the sounds of language ("phones"). You should note that the phones of Hindi only appear in Hindi and the phones of Bengali only appear in Bengali.