Why do sonorants not voice in Hungarian? And why do they voice in Slovak?
This paper investigates the acoustic aspects of voicing assimilation in Hungarian and Slovak, paying special attention to the position before the sonorant consonants /m/ and /l/. With the help of acoustic experiments, we enumerate and compare the acoustic phonetic correlates of laryngeal contrast for the obstruents /t d s z/ in various contexts in the two languages, presenting acoustic evidence of why voiceless obstruents are voiced before sonorants in Slovak but remain voiceless in Hungarian. We show that regressive voicing assimilation — including pre-sonorant voicing and final devoicing in Slovak — is close to be a completely phonologised process in both languages.