***CALL FOR PAPERS mfm25 FRINGE MEETING / GDRI PTA Dataset Workshop***

We are pleased to announce the mfm25 Fringe Meeting/GDRI Phonological Theory Agora Dataset Workshop on ternarity in English (deadline for abstract submission 31st March 2017).

The goal of the PTA Dataset Workshop is to promote discussion and theory-oriented debate in an original way. The idea is to collect a dataset and to ask participants to resolve the specific problems that it poses. We think that this workshop format (unprecedented in linguistics) is an interesting way to challenge phonologists working within different frameworks to talk about the same empirical problems and directly confront the successes and underpinnings of their formal analyses.

Facebooktwittergoogle_plusredditpinterestlinkedinmail
Posted in Calls for papers
5 comments on “***CALL FOR PAPERS mfm25 FRINGE MEETING / GDRI PTA Dataset Workshop***
  1. Joe Pater says:

    Thanks to the organizers for putting this together! It’s gotten me thinking about data that I had mulled over quite a bit in the past, and it’ll be fun to see what people have to say about it. One piece of this dataset that was new to me, and a bit surprising, is the pronunciation of Monongahela with a coda alveolar [n] in the stressed “non” syllable. Thanks Ricardo for bringing this to our attention. I can’t have the [n] there myself, so I did a poll of members of the “ling-phonology” mailing list at UMass:

    “In looking over their dataset, I noticed a datum that I hadn’t considered before: the place of articulation of the coda “n” in Monongahela. Could I ask native speakers of American English to tell me whether the velar nasal, the alveolar, or both, are OK for you there? And please do it before you look at the dataset.”

    To my surprise, the majority judgment was that [n] is possible there, in contrast to conga. For all of them, the engma was also possible. There was also one respondent who found like me that [n] was impossible. Interestingly, one of the [n]/[ŋ] respondents, John Kingston, reported that he grew up near the Monongahela river, so it’s not just unfamiliarity that leads to [n] being relatively OK.

  2. I’m delighted that were intrigued by the Monò[n]gahéla datum. Thank you very much for eliciting those native-speaker judgements! It’s good to know that Kenyon & Knott’s (1944) transcriptions remain valid nearly three quarters of a century later.

    My students and I have been looking hard for other words with nasal+plosive clusters in similar pretonic environments, but we have so far drawn a blank. Still, my feeling is that Monò[n]gahéla accords so perfectly with the pattern of rightward syllable adjunction in Medi[tʰ]erranean and Tara[h]umara that it simply cannot be dismissed as an accident. See the following set of class notes:

    http://www.bermudez-otero.com/Monongahela.pdf

  3. That’s an extremely interesting question, Joe. If you adopt the theoretical approach pursued in my class notes (linked above), there are at least two possible sources of variation: (i) the prosodic conditions on nasal place assimilation, and (ii) the direction of adjunction of pretonic light syllables.

    As regards (i), there is some suggestive evidence that nasal place assimilation may be undergoing what I called ‘rule generalization’ in Bermúdez-Otero (2015: §3.1). In trochaic bipodal words like cónquèst, for example, Wells’s Longman Pronunciation Dictionary (3rd edn) reports variation between [n] and engma in American English, but obligatory engma in British English. In my class notes, I argue that this observation, if true, would suggest that nasal place assimilation is bounded by the maximal foot projection in conservative varieties, and by the minimal prosodic word projection in innovative varieties. At the same time, both dialect types remain subject to the trochaicity requirement.

    As regards (ii), it appears that different pretonic sequences, such as LLL, LHL, and HL, differ in respect of the metrical behaviour of the pretonic syllable. In the case of LLL (e.g. Àbracadábra), all the strands of evidence converge on rightward adjunction: i.e. Àbra|[kʰ]adábra. Conversely, in the case of HL, the evidence for leftward adjuction is strong, at least when morphological boundaries do not intervene: this is shown, for example, by the flapped /t/s in words like tòr[ɾ]e|llíni or àr[ɾ]e|mísia (I am very grateful to Quentin Dabouis for these examples). It is conceivable that pretonic LHL sequences might lie somewhere between LLL (which strongly favours rightward adjunction of the last L) and HL (which strongly favours leftward adjuction of the last L).

    That’s just my tuppence worth. The matter deserves to be investigated more systematically. Hopefully the fringe workshop will cast light on this!

    Bermúdez-Otero, Ricardo. 2015. Amphichronic explanation and the life cycle of phonological processes. In Patrick Honeybone & Joseph C. Salmons (eds.), The Oxford handbook of historical phonology, 374-99. Oxford: Oxford University Press. Available at http://ling.auf.net/lingbuzz/001679

  4. I sometimes think that we should look into words from fiction books/films to see how the regularities we find in the “normal” vocabulary hold up. I believe one particular example neatly fits into the current discussion: the formula “Alohomora” from Harry Potter (normally pronounced [ˌæləhəˈmɔːrə]).

    This word sims to fit Ricardo’s proposal for words like Mediterranean or Monongahela: the presence of [h] before schwa suggests rightward syllable adjunction to the rightmost foot.

    Interestingly, I’ve come across a Youtube video of a kid pronouncing that word differently from the way it is pronounced in the films: https://www.youtube.com/watch?v=Ds1svonbFYI

    I believe it’s something like [əˌloʊhəˈmɔːrə], which is interesting in two regards:
    – the fact that [h] remains suggests that rightward syllable adjunction seems to be favoured in LHL structures (like Monongahela).
    – the fact that secondary stress is on the second syllable (and not on the first one like the “standard” pronunciation) suggests that onsetless initial syllables prefer to be unstressed.

Leave a Reply

Your email address will not be published. Required fields are marked *

*