The boukólos rule is a phonological rule of the Proto-Indo-European language (PIE). It states that a labiovelar stop (*kʷ, *gʷ, *gʷʰ) dissimilates to an ordinary velar stop (*k, *g, *gʰ) next to the vowel *u or its corresponding glide *w. The rule is named after an example, the Ancient Greek word βουκόλος (boukólos; from Mycenaean Greek qo-u-ko-ro /gʷou̯kolos/) "cowherd", ultimately from PIE *gʷou-kolos < *gʷou-kʷolos. The second constituent of this word was originally *-kʷolos, which can be seen from the analogously constructed αἰπόλος (aipólos) "goatherd" < *ai(ǵ)-kʷolos. The same dissimilated form *gʷou-kolos is the ancestor of Proto-Celtic *bou-koli-, the source of Welsh bugail (which would have had -b- rather than -g- if it had come from a form with *-kʷ-).
Another example could be the Greek negation οὐκ[ί] (ouk[í]), which Warren Cowgill has interpreted as coming from pre-Greek *ojukid < *(ne) oju kʷid, meaning approximately "not on your life". Without the boukólos rule, the result would have been *οὐτ[ί] (out[í]).
The rule is also found in Germanic, mainly in verbs, where labiovelars are delabialised by the epenthetic -u- inserted before syllabic resonants:
- Old High German queman ("to come"), past participle cuman ("come"), from Proto-Germanic *kwemaną and *kumanaz
- Gothic saiƕan, Old High German sehan ("to see"), past plural OHG sāgun ("saw"), from Proto-Germanic *sehwaną and *sēgun (-g- results from earlier -h- through Verner's law)
- Beekes, Robert Stephen Paul (2011). Comparative Indo-European Linguistics: An Introduction. John Benjamins Publishing. ISBN 902721185X. Retrieved 12 August 2017.
- Fortson, Benjamin W., IV (2004). Indo-European Language and Culture. Blackwell Publishing. p. 64. ISBN 1-4051-0316-7.
- Matasović, Ranko (2009). Etymological Dictionary of Proto-Celtic. Leiden: Brill. p. 72. ISBN 978-90-04-17336-1.
- Fortson (2004:133)