Sonority profile and temporal organization of clusters: evidence from Russian

It is well-known that consonant clusters obeying the sonority sequencing principle are universally preferred over clusters violating it; what is less clear is the status of sonority violating clusters in languages that have them. These clusters could betray their typological " markedness " by differing from sonority-obeying clusters in the same language. However, it may actually be the case that once they are part of a language's grammar, sonority-violating clusters are indistinguishable from sonority-obeying ones. In the current research, we address this issue by comparing onset clusters with different sonority profiles in Russian – sonority-raising /bl-/, /gl-/, sonority falling /lb-/, /lg-/, and sonority-flat /kt-/, /tk-/. Because the articulatory timing of onsets has been previously shown to be sensitive to various factors affecting syllable and cluster composition (as we describe below), we compare the different sonority profile clusters in terms of their temporal organization. Research on several languages and cluster types has shown that as onset complexity increases, the timing lag between the vowel-adjacent consonant and a fixed anchor point decreases, indicative of an increased overlap with the vowel (cf. Marin & Pouplier, 2014 for a review). This effect is assumed to be the manifestation at the production level of a so-called c-center organization of onsets (Browman & Goldstein, 2000). Exceptions to a c-center organization have also been reported on stop-sibilant or stop-stop clusters, with patterns reflecting additional constraints (Pastätter & Pouplier, 2015). The timing between the consonants themselves in the cluster (henceforth intra-cluster timing) has also been shown to be affected by cluster composition (cf. Marin & Pouplier, 2014 for a review). On the basis of articulatory (EMA) data from 6 Russian native speakers, we investigate the extent to which the sonority profile of an onset cluster affects the expected onset (c-center) organization, as well as its intra-cluster timing. We recorded four repetitions of clusters and matching singletons (e.g. set BL: blag – lag). The target words were preceded by a similar context in the singleton and cluster condition (see Table 1), and they were embedded in a carrier sentence. To assess the c-center organization of these sets, the timing lags between the vowel-adjacent consonant (e.g. /l/ in set BL) and the constant consonantal anchor (e.g. /g/ in set BL) were computed and compared across singleton and cluster words (the peak velocity of the constriction formation was used as temporal landmark for both adjacent consonant and anchor). Ratios between cluster …