Colloque Didier Demolin 2023 - John Kingston, "Does order matter, and if so, how much and when?"

Does order matter, and if so, how much and when? The English words cat, tack, and act comprise all three of the phonotactically possible orders of the three phonemes, /k/, /æ/, and /t/. But it turns that such examples are exceeding rare. Inspection of the 8141 one-syllable English words in IPhOD2 consisting of 2-5 phonemes (Vaden, Halpin, & Hickok, 2009) showed that while the number of permutations of a word’s phonemes that make phonotactically possible words was reliably more than one, the number that were actual words was most often just one, i.e., the word itself. And the difference between the number of possible and actual words for any given set of phonemes was most often just one or two, which shows that although English phonotactics are permissive compared to those of many other languages, they still strictly constrain the phoneme orders that actually occur in the English lexicon. Recent eye tracking studies show that hearing one order of phonemes evokes words consisting of some other orders, e.g., hearing bus /b2s/ evokes sub /s2b/ (Toscano, Anderson, & McMurray, 2013), but hearing leaf /lif/ does not evoke flea /fli/ (Gregg, Inhoff, & Connine, 2019); see also Dufour & Grainger (2019) for similar results from French using long-distance priming in a lexical decision task. Taken together, these findings suggest a division in two of labor both between and within the lexicon and the phonotactics.

1. Lexicon. Irregular properties of words (morphemes):
• Lexicon 1 – out-of-time, underlying representations,
• Lexicon 2 – real-time, word-recognition or word-production, fine-grained phonetic detail;
2. Phonotactics (grammar). Regular properties:
• Phonotactics 1 – out-of-time, abstract mental grammar,
• Phonotactics 2 – real-time parsing grammar, which governs the performance of a speaker
or listener.

In the strongest hypothesis, phonemes can be unordered in many lexical 2 entries, because phono- tactics 2 regulate their order. Lexicon 2 underuses the orders permitted by phonotactics 2. Phono- tactics 2 are learned as words are learned, but acquire independence (McQueen, 1998). Using more of the orders permitted by a permissive language’s phonotactics would require listeners to attend to order information earlier in processing.

References
– Dufour, S., & Grainger, J. (2019). Phoneme-order encoding during spoken word recognition: A priming investigation. Cognitive Science, 43(10), e12785.
– Gregg, J., Inhoff, A. W., & Connine, C. M. (2019). Re-reconsidering the role in temporal order in spoken word recognition. Quarterly Journal of Experimental Psychology, 72(11), 2574-2583.
– McQueen, J. M. (1998). Segmentation of continuous speech using phonotactics. Journal of Memory
and Language, 39(1), 21-46.
– Toscano, J. C., Anderson, N. D., & McMurray, B. (2013). Reconsidering the role of temporal order
in spoken word recognition. Psychonomic Bulletin & Review, 20(5), 981-987.
– Vaden, K. I., Halpin, H. R., & Hickok, G. S. (2009). Irvine phonotactic online dictionary, version
2.0.