Semivowel

[sem-ee-vou-uhl]

Semivowels— also known as glides or non-syllabic vowels —are vowels that form diphthongs with full syllabic vowels. That is, they are vowel-like sounds that do not form the nucleus of a syllable or mora; they are not the most prominent part of the syllable. They are normally written by adding the IPA non-syllabicity mark [ ̯ ] to a vowel letter, but often for simplicity the vowel letter alone is written.

Semivowels may contrast with approximants, which are similar to but closer than vowels or semivowels and behave as consonants.

To illustrate, the English word wow may be transcribed as [waʊ̯] (often approximated as [waʊ]). Even though both the [w] and the [ʊ̯] are similar sounds to the vowel [u], the transcription [waʊ̯] indicates that the initial segment is considered to be a consonant by the transcriber, while the final segment is considered to form a diphthong with the preceding vowel. The approximant is more constricted and therefore more consonant-like than the semivowel [ʊ̯] or the vowel [u].

Because they are so similar phonetically, the concepts of semivowel and approximant are often used interchangeably. In this conflated usage, semivowels are defined as those approximants that correspond phonetically to specific close vowels. These are , corresponding to ; for ; for ; and for . In American English, there is also rhotic [ɹ] for [ɝ]. (See approximant for details.) However, languages such as Nepali, Romanian and Samoan have additional semivowels such as [e̯] and [o̯] that correspond to mid vowels, and which other than being non-syllabic are not at all like consonants.

Examples

A number of languages contrast between the semivowel element of a diphthong and a similar approximant. For example, a number of Spanish dialects make phonemic contrasts between a palatal approximant and a palatal semivowel. Though the approximant is more constricted (having a lower F2 amplitude), longer, and unspecified for rounding (e.g. viuda 'widow' vs ayuda 'help'), the distributional overlap is limited. The approximant can only appear in the syllable onset (including word-initially, where the semivowel never appears). The two overlap in distribution after /l/ and /n/: enyesar ('to plaster') aniego ('flood') and, although there is dialectal and ideolectal variation, speakers may also exhibit other near-minimal pairs like abyecto ('abject') vs abierto ('opened'). One potential minimal pair (depending on dialect) is ya visto ('I already dress') vs y ha visto ('and he has seen'). Again, this is not present in all dialects. Other dialects differ in either merging the two or in enhancing the contrast by moving the former to another place of articulation (e.g. [ʒ]).

Romanian, a related language, contrasts the diphthong /e̯a/ with /ja/, a perceptually similar approximant-vowel sequence. The diphthong is analyzed as a single segment while the approximant-vowel sequence is analyzed as two separate segments. In addition to phonological justifications for the distinction (such as the diphthong alternating with /e/ in singular-plural pairs), there are phonetic differences between the pair:

/ja/ has a greater duration than /e̯a/
The transition between the two elements is longer and faster for /ja/ than /e̯a/ with the the former having a higher F2 onset (i.e. greater constriction of the articulators).

Although a phonological parallel exists between /o̯a/ and /wa/, the production and perception of phonetic contrasts between the two is much weaker, likely due to a lower lexical load for /wa/ (which is limited largely to loanwords from French) and a difficulty in maintaining contrasts between two back rounded glides in comparison to front ones.

Samoan contrasts close semivowels with mid ones:

Samoan ’ai [ʔai̯] ('probably')
Samoan ’ae [ʔae̯] ('but')
Samoan ’auro [ʔau̯ɾo] ('gold')
Samoan ao [ao̯] ('a cloud')Semivowel schwas

Non-rhotic dialects of English have a non-syllabic schwa immediately after the vowel nucleus, as in RP [ˈfɛə̯] fair. Many dialects of German do something similar, as in Tor [ˈtʰoːɐ̯] 'gate' and Würde [ˈvʏɐ̯də] 'dignity'. In rhotic dialects of English, the final r may be considered a rhotic semivowel rather than a consonant; the decision whether to transcribe fair as [ˈfɛɚ̯] or [ˈfɛɹ] is similar to the choice of [ˈbaɪ̯] vs. [ˈbaj] for buy (see below).

Transcription

Diphthongs are variously transcribed in English. The simplest method, typographically, is to write eye as [aj] and cow as [kaw]. However, phoneticians often object that the final segments of these words do not have the constriction that characterizes the consonants [j] and [w] in yes [jɛs] and wall [wɔɫ], but rather are purely vocalic, and that therefore the symbols and are inappropriate. In languages that contrast [ao̯] with [au̯], such as Samoan, the symbol obviously cannot be used for both. Transcribing them with vowel symbols not only enables that contrast, but it allows a more precise transcription of other diphthongs. For example, the diphthong in English bay is often transcribed with a near-high semivowel, [beɪ̯], as being more accurate than a fully high semivowel, [bei̯].