The presence of the voiceless labial fricative phoneme /f/ in a system presupposes the simultaneous presence of the voiced labial fricative phoneme /w-v/.
Standardized
IF there is the voiceless labial fricative phoneme /f/, THEN there will be the voiced labial fricative phoneme /w-v/.
1. In Gamkrelidze’s definition of a paradigmatic stop, positions are limited to bilabial, dental/alveolar, and velar. Palatal and uvular are excluded. Only stop systems with voicing contrast are considered in his analysis. In his terms of which segments can be considered legitimate members of the paradigm for stops, aspirated, glottalized and plain are allowed to fill the respective positions. Only phonemes are considered. The paradigm for fricatives includes only labial and velar positions, alveolar, palatal, interdental, and uvular are excluded.2. There is a statistical correlation in the labial series, whereby the relative frequency of the voiced fricative /w-v/, as a rule, exceeds that of the corresponding voiceless fricative /f/. (Gamkrelidze 1975: 251)3. Cf. a similar statement of Terentjev’s (#1681). 4. This universal implies the opposite of Nartey’s less restrictive claim (#799): If there is a voiced obstruent, then most likely its voiceless cognate is present as well.
1. In Gamkrelidze’s definition of a paradigmatic stop, positions are limited to bilabial, dental/alveolar, and velar. Palatal and uvular are excluded. Only stop systems with voicing contrast are considered in his analysis. In his terms of which segments can be considered legitimate members of the paradigm for stops, aspirated, glottalized and plain are allowed to fill the respective positions. Only phonemes are considered. The paradigm for fricatives includes only labial and velar positions, alveolar, palatal, interdental, and uvular are excluded.2. There is a statistical correlation in the labial series, whereby the relative frequency of the voiced fricative /w-v/, as a rule, exceeds that of the corresponding voiceless fricative /f/. (Gamkrelidze 1975: 251)3. Cf. a similar statement of Terentjev’s (#1681). 4. This universal implies the opposite of Nartey’s less restrictive claim (#799): If there is a voiced obstruent, then most likely its voiceless cognate is present as well.