SEMIVOICED

Semivoiced is a term used in phonetics and descriptive linguistics to describe sounds that are between fully voiced and fully voiceless in their phonation. It is typically applied to consonants (such as stops, fricatives, or affricates) that show partial voicing during at least part of their articulation. In practice, a semivoiced consonant may have little or no vocal fold vibration during the constriction but exhibit some periodic vibration during the release, or it may begin with brief voicing (prevoicing) that does not persist throughout the entire segment. The exact realization can vary across languages, speakers, and contexts.

Semivoiced sounds are not a standard category in the International Phonetic Alphabet. Rather, they are a descriptive label used when a language or speaker does not produce a clear, phonemic contrast between fully voiced and fully voiceless variants. The phenomenon can arise from coarticulation, speech rate, phonotactic constraints, or phonological history, and it often requires careful acoustic analysis to distinguish from neighboring phonation types such as breathy or creaky voice.

Acoustic correlates of semivoicing typically include irregular or reduced amplitude of voicing, occasional low-frequency energy corresponding to vocal fold vibration, and a voice bar that may be faint or intermittent on spectrograms. Because the category is not universally standardized, researchers may differ in whether they label a given segment as semivoiced or as simply partially voiced.

See also: voicing, voice onset time, glottalization, semivowel.