the focus is on Vnt vs Vnd sequences NC constraint ---------------------------------------- resolving the incompatibility articulatory phonology and autonomous gestures ---------------------------------------- Vnd words wie Ende, Bande, Bunde VnC words with Ente, bannte, Bunte für verschiedene Vokalqualitäte 35 native speakers of German (22 female), aged between 19 243 and 35 years materials A total of 1365 251 items from 43 words containing post-vocalic alveolar NC sequences were analyzed in the 252 current study: 366 /Vnd/ items and 999 /Vnt/ items. In all items, the target V was one 253 of the vowels /a, a:, E, e:, I, i:, O, o, œ, y:, ø:, aU, aI/. All items were either monosyllabic 254 words with (C)CVNC structure, or disyllabic words with CVNCV2 structure with stress on 255 the CVNC syllable and V2 being one of the unstressed centralized vowels /@, 5, a/. During the scanning sessions, the words appeared on a computer screen, as project- 258 ed/reflected onto a mirror placed inside the MRI scanner. Beddor and model and articulatory overlap and parsing of strategies i.e. inverse relationship here are variable perceptual strate- 42 gies for identifying nasalization in ṼN sequences (where Ṽ signifies a vowel with some 43 coarticulatory nasalization): some listeners base their judgements mostly on information 44 in N, others associate nasalization with ṼN (without parsing such judgements with either 45 just Ṽ or just N), while yet others’ perceptions of nasalization are swayed to a greater 46 extent by information in Ṽ alone ---------------------------------------------------------- we analyse German because we want to get at the earliest stages of sound change. NB the focus is on Vnt vs Vnd sequences ---------------------------------------------------------- accurately tracking the time course and magnitude of velum raising 77 and lowering from speech acoustics both across speakers and within individual tokens is 78 notoriously difficult (Barlaz et al., 2019; Carignan, 2018b; Feng and Castelli, 1996; Saxon 79 et al., 2019; Styler, 2017). While aerodynamic mea- 83 surements of nasalization (e.g., the magnitude of nasal air flow as recorded from pressure 84 transducers attached to a vented mask) can be used to characterize nasalization in both 85 vowels and consonants, these measurements are inherently indirect estimates of velum 86 lowering (