underlying phoneme

Speech sound identifying for instance geographical origin

An underlying phoneme is an abstract category of speech sound which can have different phonetic realizations according to context, For example, the phoneme /k/ in English has different realizations (articulation) as the first consonant in the words "key" and "car". A phoneme is related to semantic differentiation, thus /l/and /r/ are different phonemes in English, but constitute a single phoneme in some Asian languages. By contrast differing articulations of the English phoneme /l/ have no semantic significance in English, but they do in Russian ("clear L" vs "dark L"). Also note that there are "phonological rules" which switch phonemes according to context. For example the articulation of the letter "s" shifts from phoneme /s/ to/z/ in many English plurals.

This is very interesting, Fresco. Would it be possible to give an example of the Russian Ls?



I wonder, can we even say that they "constitute a single phoneme" when 'L' doesn't exist in a language. As regards Japanese, the reason English speakers sometimes perceive an 'L' sound is because Japanese speakers produce their 'Rs' with the tongue put in the position where we produce our 'Ls'.

Say

ra - ri - ru - re - ro

Where does your tongue sit/lie in the production of those five?

Anyone and everyone, please feel free to answer this.

bookmark

Say

ra - ri - ru - re - ro

Where does your tongue sit/lie in the production of those five, MJ?

are you talking about the vowels or the consonant? It sounds like you were unaware that different languages split up the phonetic spectrum in different ways. For example in Spanish the two vowels in "bebe" sound like the same sound to Spanish speakers, but in English they sound very similar to "bet-bay" except English "long a" is a diphthong but the Spanish is not" (I didn't do that well: it's the vowel sound in "bet" but without the "t"). Rhe Spanish phoneme covers a wider span of the phonetic spectrum than the English, which splits it into two. Similarly in Hindu and Urdu "b" and an aspirated "bh" are two different phonemes, but English speakers find it very very difficult to distinguish between the two because the difference is not meaningful in English.

If you're talking about the "r", that too is different in different versions of English. I grew up in Michigan and my "r"s had most of the time a tongue flip. Having spent decades in Boston, a lot of my "r"s have become "ah"s (it's more complicated than that, but that's kind of a good starting point. People with Boston accents hear both "r" and "ah" as the equivalent of an "r" sound}/ And the remaining "r"s that are still "r"s don't seem to have the tongue flip most of the time any more (the tongue flip kind of gave me an advantage in Spanish, where "r" and "rr" are different phonemes, the "r" with a flip and the "rr" with several.

Is that like 'Brenda' being "Brender"?

No. That's a different case. It's "car" and "cah".

Examples for clear and dark L are given on Wiki for Russian and other languages. I seem to remember that the final "l" in the Russian"shtol" changes the meaning of the word according to whether it is clear or dark
Note also that the phrase "single phoneme" refers neither to /l/ or /r/ specifically, but to a range of articulation which is semantically insignificant in one language (Japanese say) but significant in another.
"Existence" of a sound category like "phoneme" is cultural not physical.

Yes, it's called Peklo.

The two together. In Japanese a consonant with the exception of 'n' doesn't exist as a sound without a vowel accompanying it.

ra - ri - ru - re - ro

It does if you're a Japanese phonetician.

Even in a completely innocuous situation, you seem to be afraid of making a fool of yourself. Why are you so afraid, MJ?

The question is, rather, why you always jump right in and make a fool of yourself. The question is not syllabic structure. "r" is clearly a phoneme in Japanese, as witness the minimal pair ma-ra. That the predominant Japanese syllabic form is CV (there are of course others too)is irrelevant to whether or not it exists as a distinct sound in Japanese. It does.

One thing that always got me about the idea of underlying phonemes, or underlying representations, is the assumption that one phoneme underlies the morpheme and not another.

Like in the example of the English plural morpheme /s/, /s/ is taken to be the underlying phomeme. What's to say that the underlying phomeme isn't /z/?

This isn't in IPA:
/socks/
/bedz/
The phones change because of assimilation with the previous phones, /k/ which is unvoiced and /s/(unvoiced), whereas /d/(voiced) and /z/(voiced).

According to Wikipedia, underlying phoneme or underlying representation is a theoretical construct.

The traditional phonological rule in English is to add /s/ in the plural or third person singular of verbs, not /z/. i.e. /s/ is the default mode and the transformation of this underlying form to /z/ comes about by its succession to "voiced" consonants or vowels. No doubt the specified "underlying form" is related to the textual representation of grammatical rules which reflects the perception of speech as a sequence of items. The speech waveform does not reflect such segmentation. (Note there are other phonological theories which do not rely on such segmentation, such that the phrase "underlying phoneme" would inappropriate).

But you're talking about the orthographic representation of "s" as plural, not phonemics.

Yes, we add "s" to nouns to show the plural when we're writing. But when spoken we pronounce /s/ to mean plural only after words that end in unvoiced phones. It's not so much a "rule" as it is a natural phenomenon of speech. Phonemics attempts to describe the phenomena of speech.

My question regards the phonemics of spoken language, not the orthography.

You miss the port that the orthography acts as a theoretical substrate for the phonology. We need orthographic representations of the abstract phonemes which reflect the segmental syntax. Since the orthographic "s" remains syntactically unchanged, that is why phoneme /s/ it is utilized as "underlying".

Remember that phonology has to relate to semantics and syntax and not merely to phonetics. There was a celebrated debate in linguistics entitled "How abstract is phonology ?" (instigated by Paul Kiparski, I think.)