(The non-existence of) secondary stress in Hungarian

Secondary stress in Hungarian has been subject to considerable debate. Proposals include a basic quantity-insensitive trochaic pattern (Kerek 1971), a quantity-sensitive pattern (Szinnyei 1912), and a distinction between secondary and tertiary stress (Hammond 1987). These claims have been made on the basis of impressionistic transcriptions, without acoustic measurements, and they are unsupported by phonological evidence (Kálmán & Nádasdy 1994). We examine a phonological argument for the existence of secondary stress by Varga (2000), and argue that his ndings are inconclusive. We also present the results of our pilot acoustic studies, showing no correlation between non-primary stress and vowel duration, pitch or intensity. We conclude that evidence for Hungarian secondary stress is lacking.