Voiced velar approximant
A voiced velar approximant is a type of consonantal sound used in some spoken languages. In the International Phonetic Alphabet, it is written ⟨ɰ⟩. To avoid suggesting that it is spread like its vowel equivalent [ɯ], it may instead be written ⟨ɣ̞⟩ or ⟨ɣ⟩ (or ⟨ɣ˕⟩, which is now dated). This symbol is the voiced velar fricative ⟨ɣ⟩ with a lowering diacritic.
This consonant is not found in English. It can be approximated by making [ɡ] with the tongue body lowered, or [w] with the lips apart. In many cases, it can be seen as the semivocalic partner of the close back unrounded vowel [ɯ]. Some transcription systems use ⟨ɰ⟩ and ⟨ɯ̯⟩ with the non-syllabic diacritic for the same sound, though others make a difference between them.
In some languages, such as Spanish, this sound is an allophone of /ɡ/, as explained below.
For a voiced post-palatal approximant, sometimes also called a voiced pre-velar approximant, which is further forward in the place of articulation than a typical velar approximant, see Voiced palatal approximant § Post-palatal.
Phonetic ambiguity and transcription usage
Further information: Spirant approximant
Some languages have a voiced velar approximant that is not rounded, so it cannot be treated as the "glide" version of [ɯ] or of its rounded partner [u]. Examples are Catalan, Galician, and Spanish, where this sound appears as a version of /ɡ/.
The expert Eugenio Martínez Celdrán says that the symbol ⟨ɰ⟩ is a poor fit for the Spanish sound. Many writers note that [ɰ] is not rounded, and it is rounded before rounded vowels in Spanish. He prefers ⟨ɣ̞⟩, which matches the other soft sounds in Spanish, [β̞] and [ð̞]. Ball and Rahilly also point out that the difference between this sound and the voiced velar fricative [ɣ] is that the [ɰ] sound needs spread lips and becomes [ɯ] when it is held longer.
There is a similar problem when writing voiced palatal approximants. In broad transcription, the lowering diacritic may be left out, so the symbol is written ⟨ɣ⟩, the same as the fricative.
Features
The manner of articulation of this sound is approximant, which means the tongue narrows the vocal tract but not enough to create a turbulent airstream. Its most common type is a glide or semivowel. A glide moves from the ɯ vowel position to the next vowel, and a semivowel is vowel-like but does not form the center of a syllable. For the spirant approximant variant used in Spanish, see § Phonetic ambiguity and transcription usage.
The place of articulation is velar, made with the back of the tongue (the dorsum) near the soft palate. The phonation is voiced, meaning the vocal cords vibrate. It is an oral consonant, so air cannot escape through the nose, and a median consonant, so air flows down the middle of the tongue rather than the sides. Its airstream mechanism is pulmonic, meaning air is pushed out only by the intercostal muscles and abdominal muscles, as in most speech sounds.
Occurrence
The list below includes languages with variable pronunciations.
A nasalized voiced velar approximant is a type of consonantal sound used in some spoken languages. In the International Phonetic Alphabet, it is usually written as ⟨ɰ̃⟩ or ⟨ɣ̞̃⟩.
| Language | Word | IPA | Meaning | |
|---|---|---|---|---|
| Aragonese | caixigo | [kajˈʃiɣ̞o̞] | 'oak tree' | |
| Astur-Leonese | Asturian | |||
| Extremaduran | ||||
| Leonese | ||||
| Mirandese | ||||
| Basque | hego | [eɣ̞o] | 'wing' | |
| Catalan | aigua | [ˈajɣ̞wə]ⓘ | 'water' | |
| Cherokee | ᏩᏥ wa-tsi | [ɰad͡ʒi] | 'watch' | |
| Danish | Older speakers | talg | [ˈtsʰalˀɣ̞] | 'tallow' |
| Dutch | Western East Flemish | |||
| French | Belgian | ara | [aɣ̞a] | 'macaw' |
| Galician | órgano | [ˈɔrɣ̞ɑ̟nʊ]ⓘ | 'organ' | |
| Greek | Cypriot | μαγαζί | [maɰazˈzi] | 'shop' |
| Guarani | gotyo | [ɰoˈtɨo] | 'near, close to' | |
| Ñandewa Paulista-Paranaense | [adʒaˈɰa] | 'I cut' | ||
| Hiw | wr̄og | [wɡ͡ʟɔɣ̞] | 'through' | |
| Ibibio | ufokọ | [úfʌ̟̀ɰɔ̞] | ||
| Icelandic | Skagabyggð | [ˈskɐːɣ̞ɐˌpɪɣ̞θ]ⓘ | 'Skagabyggð' | |
| Irish | naoi | [n̪ˠɰiː]ⓘ | 'nine' | |
| Judeo-Spanish | gato | [ˈɣ̞at̪o] | 'cat' | |
| Korean | 의사 / uisa | [ɰisɐ] | 'doctor' | |
| Mwotlap | haghag | [haɣ̞haɣ̞] | 'sit' | |
| Occitan | Gascon | digoc | [diˈɣ̞uk] | 'said' (3rd pers. sg.) |
| Shipibo | igi | [i̞ɣ̞i̞] | ||
| Spanish | Most dialects | pagar | [päˈɣ̞äɾ]ⓘ | 'to pay' |
| Standard European | blog | [ˈbloɣ̞̊] | 'blog' | |
| Swedish | Central Standard | temadagarna | [ˈtʰîɛ̠mɐˌdɒ̟ːɣæɳɐ̞]ⓘ | 'themed day' |
| Tagalog | igriega | [iːɡɾɪˈje̞ɣ̞ɐ] | 'y (letter)' | |
| Tiwi | ngaga | [ˈŋaɰa] | 'we (inclusive)' | |
| Vietnamese | Southern | gặp | [ɣ̞ap̚˨˩˨]ⓘ | 'to meet, to see' |
Bunched or molar r
Some languages have a velar approximant made with the body of the tongue bunched up at the velum, along with pharyngealization at the same time. This creates a retroflex-like sound that is almost the same as ɻ. Catford (1982) described it precisely as an "apico-postalveolarized, advanced velar approximant". The extensions to the IPA call it a "bunched r" or "molar r" and suggest the centralized mark, ⟨ɹ̈⟩, to tell it apart from the apical version ⟨ɹ̺⟩. Usually this mark is left out, so the sound is written simply as ⟨ɹ⟩ or ⟨ɻ⟩, as if it were a coronal consonant.
In Dutch, this type of r is called Gooise r ("Gooi r"), after het Gooi, a region of the Netherlands where Hilversum, the main centre for television and radio broadcasting, is located. The sound is an approximant in manner of articulation, which means the vocal tract is narrowed but not enough to create a turbulent airstream. Its place of articulation is velar, made with the back of the tongue at the soft palate. It is voiced, it is an oral consonant (air does not escape through the nose), and it is a median consonant (air flows down the middle of the tongue). Its airstream mechanism is pulmonic, made by moving air with the intercostal muscles and abdominal muscles.

Quiz time!
5 questions about Voiced velar approximant
Which part of the mouth does the back of the tongue come close to when making this sound?
Keep exploring
This article is a child-friendly adaptation of the Wikipedia article on Voiced velar approximant, available under CC BY-SA 4.0.