Introduction Stressed or pragmatically prominent syllables are typically produced with longer durations and a more careful articulation. For instance, stressed vowels in English are not reduced to schwas, while unstressed vowels often are. In Dutch, unstressed vowels are not always reduced to a schwa, but still produced with a different spectral tilt, i.e. relatively less energy in the higher frequency bands [1,2,3]. Articulatory strategies to mark pragmatic prominence on vowels have been identified, but seem to vary between speakers [4]. One way is to move the tongue to a more extreme position, the other is to increase the sonority contrast in more prominent syllables, by making consonants less and vowel more sonorous [5]. However, vowel and consonants might not be subject to prominence marking in the same way. Consonants could be shortened to increase the sonority contrast within a syllable, but they could also be articulated at more extreme positions to protect their identifiability. If a consonant has to be articulated in a shorter time with the articulators still reaching the target position, movement velocity has to increase. For lingual coda consonants, there might be a need to compensate for more extreme tongue positions reached during the preceding vowel articulation. If a speaker does not speed up the articulator movements or lengthens the consonant, she should accept loss of quality, such as lenition or coarticulation. Hence, multiple predictions on the effect of prominence on consonants can be made and different speakers might apply different strategies. In addition, prominence derived from lexical stress and prominence from pragmatics have effects on articulation [6,7], but these two ‘prominences’ might be marked differently. Given the myriad of possible effects of prominence on consonant articulation, the present study features an in-depth exploratory analysis of a subset (4 pp) of Electromagnetic Articulography (EMA) recordings of 40 speakers of Dutch. Our goal is to formulate hypotheses on possible strategies employed to mark prominence in consonant articulation that can be tested on the remaining data in a later stage. Method The study focussed on coda consonants, because 1) they might be more susceptible to prosody-driven lengthening than onset consonants [5], 2) they are susceptible to coarticulation in Dutch and 3) they allowed a relatively elegant elicitation of lexical and pragmatic prominence. We analysed /n/ and /t/ in coda position of morphemes ‘in’ (/ɪn/) and ‘uit’ (/œyt/) before a labial. Stress was manipulated by placing the morpheme in a compound verb (/'ɪnpɑkə/, to pack: /'ɪnbɑkə/, to season: ) or as preposition in a phonemically identical phrase (/ɪn 'pɑkə/, in packages). The /n/ is more sensitive to coarticulation with the labial to [m] than the /t/. The morpheme was made pragmatically prominent to three degrees by embedding it in a given sentence that served as an answer to a given question. Pragmatic prominence levels varied between low (after a different phrase with narrow focus), neutral (the whole sentence or phrase contains new information), and high (the morpheme alone corrects an assumption). Tongue tip (for /n,t/) and labial articulatory movements (distance upper-lower lip, for /b,p/) were segmented using Mview, courtesy Mark Tiede. Syllable and vowel durations were extracted from the acoustic signal. A linear mixed model was fitted for all dependent variables, with lexical stress, pragmatic prominence as independent variables and participant and item as random effects. Results Due to The tongue was less front with low focus (p = 0.03); for the /t/, the tongue was lower for stressed syllables (p < 0.05). The distance the articulator travels depends on prominence, interacting with the consonant (p < 0.05*). For /n/, the tongue moved less for high focus compared to low (B = -4.4mm, p = 0.009**). For the /t/, stress and focus interact (p = 0.03*). For stressed /t/, the tongue moves less (B = -3.9mm, p = 0.004**). Focus prominence now had a positive influence, with more distance for high focus than low focus consonants (B = 3.9mm, p = 0.019*), but only on the stressed syllables (B = 6.9mm, p = 0.004**). For onset velocity, there was a significant interaction between phoneme and stress (p = 0.04*); only for /t/ the unstressed consonants are articulated significantly slower (B = -3.5 cm/s, p = 0.02*). No effect for entry velocity peak or focus was found, neither were overlap, lags, exit velocities or articulatory durations significantly affected by any prominence level. Stressed vowels seem slightly longer acoustically, but not significantly. Discussion As no evidence was found that prominence affects articulatory or acoustic duration of consonants, we remain agnostic on the possibility that sonority is increased by reducing consonant duration as well as the possibility that stress or focus are expressed by syllable lengthening. However, consonants might be (only) proportionally shorter, if the vowels are longer when more prominent; we did not find an effect on vowel length now, but it might found in the full data set. If so, phoneme identifiability can be preserved, while the sonority contrast is still enhanced. Consonants were found to be produced at more extreme positions when pragmatic prominence was higher, but the speakers did not consistently make larger articulatory movements to achieve this. We hypothesise the following general strategy exists: when a prominent consonant is realised, the tongue might be positioned for it during the preceding vowel, suggesting that carryover coarticulation can follow from anticipatory planning. The resulting sacrifice of precision of the vowel articulation is, however, only made if needed; we hypothesise the /t/ needs less protection (probably because it is a stop, but lexical neighbourhood size happens to be a confound). If the consonant needs more protection, vowel and consonant are articulated at more extreme positions, causing an increase in movement. We cannot yet identify one consistent strategy that is completely supported by the data. However, our exploration does allow for a focussed use of the larger dataset for hypothesis testing. In addition, we plan to not only analyse the group-wide strategy for prominence articulation in coda consonants, but also to identify participant-specific strategies. References [1] Rietveld, A. C. M. & Koopmans-van Beinum, F. J. 1987. Vowel reduction and stress. Speech Communication, 6, 217–229. [2] Sluijter, A. M. C. & van Heuven, V. J. 1995. Effects of focus distribution, pitch accent and lexical stress on the temporal organization of syllables in Dutch. Phonetica, 52,71–89. [3] Sluijter, A. M. C., van Heuven, V. J. & Pacilly, J. J. A 1997. Spectral balance as a cue in the perception of linguistic stress. Journal of the Acoustical Society of America, 101(1) [4] Harrington, J., Fletcher, J., & Beckman, M. 2000. Manner and place conflicts in the articulation of accent in Australian English. In Broe, M. B. & Pierrehumbert, J. B. (Eds.), Papers in Laboratory Phonology V: Acquisition and the lexicon, 40–51. Cambridge, UK: Cambridge University Press. [5] Beckman, M., Edwards, J., & Fletcher, J. 1992. Prosodic structure and tempo in a sonority model of articulatory dynamics. In Docherty, G. J. & Ladd, D. R. (Eds)., Papers in Laboratory Phonology II: Gesture, segment, prosody, 68–89. Cambridge, UK: Cambridge University Press. [6] Turk, A. 2012. The temporal implementation of prosodic structure. In Cohn, A.C, Fougeron, C., & Huffman, M. K. (Eds.), The Oxford Handbook of Laboratory Phonology, 242–253. Oxford, UK: Oxford University Press. [7] Beckman, M. E. & Edwards, J. 1994. Articulatory evidence for differentiating stress categories. In Keating, P.A. (Ed.), Phonological Structure and Phonetic Form: Papers in Laboratory Phonology III, 7-33. Cambridge, UK: Cambridge University Press.