I don't know where it came from, but the "west" at least as I have learned, came up with the idea of "vowels" and "consonants" at some point, and we just go with that classification. But besides linguistic analysis revealing that these categories are in fact fuzzy, and there being all kinds of features behind the production of each sound (voice, airflow/restriction, nasality, etc.), in just thinking about it now, here's how I might break it down if I never learned about this stuff through school (somewhat):

  • voice, no voice (s vs. z)
  • continuous vs. discrete (s vs. p)
  • nose vs. mouth (m vs. s)

No where to be found is the consonant/vowel distinction.

In addition, following question Why r, h, and w aren't vowels, some things could be considered vowels which are consonants, the classification grouping of English, for example, seems arbitrary. So I wouldn't have the "vowel vs. consonant" distinction, how did that even get made in the first place? It seems so arbitrary.

The only way I can think of the reason why vowels might be their own class, is they are voiced, continuous sounds with an open mouth. But z and ʒ are also voiced, continuous sounds, but perhaps with only difference between them is that the airflow is sort of restricted by the tongue. Most vowels and consonants use the tongue in some way to make their sound it seems, so it's not obvious at first thought that categorizing based on this tongue feature would be, well, important or useful. In fact, "i" (ee sound) kind of is like "z", but using the middle/back of the tongue (restricted airflow) instead of the front, so there's that too.

So how did the consonant/vowel distinction arise (tangential question), and what are other classification systems used by other cultures from around the world (main question)?

Maybe y and w are consonants because you treat them as discrete sounds. So y is like i and w is like u, but saying "uuu-one" is not the same as "one" (with a w sound), and "iii-es" is not the same as "yes" (with a y sound). I can't pinpoint the difference here really, but it seems like maybe the length of how I say one vs. the uuu-one, or no, maybe it is that I don't have a slight glottal-stop at the beginning when it's a w, but I sort of have a break/glottal-stop when I start the uuu in uuu-one. It's like my tongue releases from the top of my mouth (closed airflow) when I say "uuu", but doesn't close airflow and hit the top of my mouth at all when I say "w". So it's like the vowels then are things which start with a closed tongue/pipe, and is continuous voice! That is maybe why "errrr" is spelled that way (with an e before r), because it starts with the closed pipe, but "rrr" doesn't have a closed pipe to start. So if it is voiced, continuous, and starts with a closed pipe when stated on its own, it's a vowel? Everything else then is a consonant.

  • 1
    Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Linguistics Meta, or in Linguistics Chat. Comments continuing discussion may be removed.
    – prash
    Mar 20 at 5:34

3 Answers 3


The discovery of the distinction vowel vs. consonant is lost in ancient unrecorded history. Over 3,000 years ago, a set of Indian rituals was canonicalized into the Vedas, where exact performance was a desideratum, thus was born the discipline of linguistics. We only know the result of those studies, not the internal history e.g. "what was the first linguistic technical term?". Sanskrit phonetics (śikṣā, transmited via the prātiśākhyas) presents a systematic conceptual hierarchy that states all of the phonologically-relevant properties of Sanskrit segments, where syllables (akṣara) are decomposed into vowels (svara) and consonants (vyañjana), etymologically 'sounding' and 'ornaments', then further subdividing into stops, approximants and sibilants, and so on.

It is well-established that there was ancient contact between Greek and Indic civilization, meaning that it is plausible that the Greek concepts φωνῆεν (vowel), σύμφωνον (consonant) were modeled after the corresponding Sanskrit concepts. The unity on these systems of classification, which is substantially carried over to modern acoustic theory (e.g. Fant) is that vowels provide the power source, and consonants are modulating ornaments (obvious in speech, really obvious in singing). Latin-based "vowel, consonant" are simply calqued from Ancient Greek.

Other (known) ancient grammatical traditions were derived from contact with India (Chinese and Arab grammarians, and further developments radiating from those sources. It is always possible that there were ancient grammarians in Egypt, but they left no trace. It might be worth investigating the history of Babylonian Grammatical Theory to see what kind of account they had of phonetics and phonology.

In other parts of the world, being free from the influence of Indic grammarians, the question would naturally arise "Why do we need this science of linguistics and why should we devise obscure grammatical terms like 'vowel'?". In other cultures, people just talk and it isn't important to distinguish "consonant" from "vowel", or "labial" from "velar". As the ancient terminology indicates, vowels are the power, voice, or sound of speech, and what's left over is adornments that "go with" that power source. Vowels are easy to characterize phonetically and phonologically, consonants are not independently identifiable, other than saying "not vowel".


This could be tackled from many points of view, but since you seem to be thinking mainly of the phonetic (as opposed to phonological) one, I'll first mention that fricatives and approximants like /z/, for example, have a turbulent airflow (more so fricatives, while approximants are, eh, indeed closer to vowels), which is something that has a physics definition and can be seen in an airflow visualization. Vowels don't have that, as the constriction created by the tongue or whatever isn't strong enough to cause it. Compare the need to play a flute in a specific way to get "sound" rather than "noise" (it's all sound, but I hope it's clear what I mean here).

The thing with /w/ and /j/ is mostly phonological, not phonetic. They are very close to /u/ and /j/ although they vary by language and are usually a little more constricted, but the main distinction in the IPA is simply that /j/ and /w/ don't form the nucleus of a syllable and they pattern with consonants in the language in question and the analysis in question. It's no longer a matter of "what it sounds like", but "how it is used in the language". And, really, this distinction is very important: your question can be about phonetics (the study of speech sounds in themselves), or phonology (phonemes, the sounds as conceptualized in a given language and a specific analysis of it), and have different answers.

For example, /h/ is trivially not a vowel because it's not voiced, phonetically speaking (the vocal cords are not moving), although whenever something seems "trivial", you have examples like Japanese devoiced vowels. In Japanese, those still form the nucleus of a syllable (or a mora at least), so even though they are phonetically voiceless, they can count as phonemic vowels.

By the way, you mentioned the vowel/consonant distinction as arising in the "west", but unless India is west to you, well, their scripts are abugidas which imply a distinction between consonants and vowels, and the Indian grammarian Panini certainly knew of a distinction between vowels and consonants.

The sound of "err" is an R-colored vowel, which is quite definitely a vowel phonetically ([ɚ]), and sometimes also treated as a vowel phonemically in English varieties that use it, although sometimes it's treated as the sequence /ər/ instead. Note here that /r/ has nothing to do with the sound of [r] as typically used in the IPA phonetically (that sound is more like the sound of Italian "rolled" R), but it's still used in a lot of English IPA when doing phonemic transcriptions, because the IPA symbols chosen are arbitrary to a degree, and many times linguists will pick the "more familiar" shape, in this case /r/, even though the consonant sound of English R (as in "rate", not in "err") is generally closer to [ɹ]. I point this out to again underline the difference between a phonetic analysis and a phonological analysis of phonemes.

(When you talk about a "closed pipe", that sounds to me like you have a glottal stop in mind, but that is not really used to define vowels; however, there are languages, such as German, where words starting with phonemic vowels systematically get a phonetic glottal stop before them.)


Vowels and consonants are fuzzy in English linguistic tradition. In Russian linguistics, for instance, it is absolutely not fuzzy. Only a vowel produces a syllable or can be stressed. There are no "semi-vowels" or stressed consonants, and such.

In Russian consonants can be soft and hard (some consonants are only soft or hard, but one can imagine and produce them in another quality, it just does not exist in used words).

Each consonant can be paired or not paired (that is, has a pair of opposete voicedness). Thus a squad of voiced hard, voiced soft, voiceless hard and voiceess soft form the basic structure of consonants.

There squads can be further paired into plosive-fricative (what you called continuous/discrete) pairs or in other way.

This is studied in Russian schools, every pupil shoul be able to do phonetic analysis of a word.

  • 2
    Downvoting this post because of the misleading presumption that English phonemic complexity is due to its "tradition" rather than to features of the language itself, such as semivowels actually patterning with consonants in many significant ways. In addition, nobody asked for a tirade on what should be taught in schools.
    – LjL
    Mar 16 at 17:45
  • @LjL so-called "semivowels" ARE consonants, not just patterning with them.
    – Anixx
    Mar 16 at 18:01
  • @LjL there is no statement on what should be taught in schools, just description of facts because the OP asked about cultural differences and mentioned his school.
    – Anixx
    Mar 16 at 18:02
  • The semivowels /j/ and /w/ are defined as identical to the vowels /i/ and /u/ (at least in idealized IPA, then languages tend to make them differ a bit), the only difference being that in a given language's phonotactics, they act (pattern) as consonants. So phonologically, they are consonants, but phonetically they can be indistinguishable from the respective vowels. See Wikipedia too: "The palatal approximant can often be considered the semivocalic equivalent of the close front unrounded vowel [i]"
    – LjL
    Mar 17 at 19:41
  • It also says: "The term glide emphasizes the characteristic of movement (or 'glide') of [j] from the [i] vowel position to a following vowel position. The term semivowel emphasizes that, although the sound is vocalic in nature, it is not 'syllabic' (it does not form the nucleus of a syllable)." So in other words, it's phonetically a vowel, [i], but when in a language it doesn't act as the nucleus of a syllable and serves as a transition to it, [j] is used to indicate that. But "the sound is vocalic in nature".
    – LjL
    Mar 17 at 19:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.