Earth to Jean-Marc Phillippe

August 8, 2006 9:16 PM

Are /k/ , /'e/ , and /o/ really the only phonemes "common to every language?"

So I'm finally getting caught up on some reading, and the May 2006 issue of Wired has a brief article about KEO, a satellite that's going to be launched sometime next year.

The Wired article stated, "...the vehicle [is] (named for the phonemes /k/, /'e/, and /o/, according to Phillippe the only ones common to every human language)"

Is that true? My Google-fu is failing - I'm finding the most common phonemes for language groups, but not across the board.

It also seems kind of suspicious that mere "supporters" of the satellite project were doing the research on this - "To name their satellite project, supporters researched the most frequent phonemes in major world languages and came up with "K," "E" and "O." A phoneme is the smallest phonetic unit in a language that is capable of conveying a distinction in meaning."

posted by Liosliath to writing & language (11 comments total)

"WordIQ" is sort of a Wiki about linguistics.

In this article it says:

The most common vowel system consists of five vowels: /i/, /e/, /a/, /o/, /u/. The most common consonants are /p/, /t/, /k/. Not all languages have these; the Hawai'ian language lacks /t/, and the Mohawk language lacks /p/, and Hupa lacks both /p/ and /k/. If one of the three is missing, the language will have /?/ (glottal stop).

Which means that KEO is blowing it out of their ear.
posted by Steven C. Den Beste at 9:52 PM on August 8, 2006

It sounds like it was just a misunderstanding or misinterpretation.

It's not out of the question that "the most common phonemes among all the world's languages" got reinterpreted as "the only common phonemes among all the world's languages" during the writing, subbing, rewriting and editing process, known to be somewhat fraught at WIRED...
posted by AmbroseChapel at 10:58 PM on August 8, 2006

More on the prevalence of various phonemes in a group of 25 languages that has pretty good global coverage.

[?] can result from eliding any oral stop enough, so its presence, in a language without [p] [t] or [k], does cast doubt on the claim that the language lacks /p/ /t/ or /k/. However, I am in no place to dig up a professional argument that this is or isn't the case in, e.g., Hawai'ian, so I'll just mention the possibility and leave it at that.

Languagehat?
posted by eritain at 3:52 AM on August 9, 2006

A bit of an answer from left field, but if you're talking about "every human language," it must be noted that signed languages would certainly be lacking all of those phonemes.
posted by Eldritch at 4:22 AM on August 9, 2006

There are no phonemes common to every human language (even limiting it to spoken ones). Either AmbroseChapel is right (the generous interpretation) or they're just talking crap.
posted by languagehat at 5:50 AM on August 9, 2006

It actually doesn't even make sense to talk about "phonemes common to every human language," because a phoneme by definition has meaning only within the phonemic system of that particular language—the phoneme /s/ is not the same in English and Russian, even if it sounds the same, because it takes part in a different system of oppositions.
posted by languagehat at 5:52 AM on August 9, 2006

It seems obvious that nobody could make a statement about phonemes appearing or not appearing in “every” spoken language because not all of them have been inventoried and it borders on impossible for one person to check all of them. Of course a computer could do the checking, but then somebody would have to input the data in a consistent and computer-searchable format.

Then of course there is the pesky case of sign languages.

And, Languagehat, if I have a native bilingual speaker produce one-syllable words beginning with [s] in Russian and English and run them through a spectrograph, what happens if the [s] waveforms are identical? Are they really two different [s]s? I know you are trying to make a point about definitions, but it is a point that falls down in your example.
posted by joeclark at 6:38 AM on August 9, 2006

But Languagehat didn't mention tokens of [s] in Russian and English. He said "the phoneme /s/ is not the same in English and Russian" [emphasis added]. You can't record a phoneme; it's a mental abstraction.
posted by Utilitaritron at 7:25 AM on August 9, 2006

Utilitaritron, you are again positing a definition.
posted by joeclark at 8:50 AM on August 9, 2006

"Positing a definition"? WTF are you talking about? There is a definition of "phoneme," and it's an abstraction, not a sound. (Look it up.) When I said "even if it sounds the same," that was shorthand for "even if you hear a physical realization of the English /s/ that sounds like a realization of the Russian /s/." I wasn't "trying to make a point about definitions," I was trying to make a point about phonemes.
posted by languagehat at 2:26 PM on August 9, 2006

AmbroseChapel, maybe you're right...

"... one partner came up with the idea to research the sounds (phonemes) common to the most widely spoken languages today and to choose the ones that are used most frequently: [k], [e], and [o]... and hence was born KEO, a name that is pronounceable by all cultures."

I like how they somehow make the jump from most widely spoken languages to "all cultures."

In a nutshell, they're new-agey kooks with a lack of understanding about phonemics, correct?

Also appreciated the point about sign language - surely there must be symbols common across cultures?
= The Spaceship Formerly Known As KEO.
posted by Liosliath at 6:13 PM on August 9, 2006