A labialized velar or labiovelar is a velar consonant that is labialized, with a -like secondary articulation. Examples are , which are pronounced like a , with rounded lips, such as the labialized voiceless velar plosive and labialized voiced velar plosive , obstruents being common among the sounds that undergo labialization.
The most common labiovelar consonant is the voiced approximant . It is normally a labialized velar, as is its vocalic equivalent . (Labialization is called rounding in vowels, and a velar place is called back).
and its voiceless equivalent are the only labialized velars with dedicated IPA symbols:
Labialized velars frequently derive from a plain velar followed by a rounded (labialized) vowel, such as or . In turn, they may sometimes develop into simple bilabial consonants. An example of this is the development of Proto-Indo-European *k÷, *g÷ before *a or *o into Greek /p, b/, producing cognates as different as English come and basis. The full sequence is demonstrated by the Satsuma dialect of Japanese: in northern Satsuma, Standard Japanese 'eat!' has contracted to ; in southern Satsuma, it has proceeded further to .
A notable development is the initial *k÷ in Proto-Indo-European interrogative words. In English, it developed into wh or h (how), pronounced /w/ in most dialects and /h/, respectively, via Grimm's law followed by wh-cluster reductions. By contrast, in Latin and its descendants, the Romance languages, that developed into qu (later Spanish cu (cuando) and c (como)), pronounced as /k÷/ in Latin and variously as /kw/ or /k/ in the Romance languages. See etymology of English interrogative words for details. The English phonemic spelling kw for qu (as in ') echoes its origin.