Illuminating the Structure of Syllables through Experimentation Using the Play Language (Verlan)

The syllable is the most salient prosodic unit any speaker of a natural language perceives. It is made up of smaller entities called segments, e.g. three segments [kæt] comprise the monosyllabic word cat. Since the 1940s, the mainstream belief is that the segments are abstract timing units organized into constituents such as onsets and rimes, the latter in turn consisting of nucleus and coda, hence [onset k[rime [nucleusæ][codat]]. In the 1980s, this Onset-Rime Model (OR- Model) gave way to the Moraic Model (u-Model) of the syllable where segments are identified by their relevance for contributing to syllable weight. In the u-Model, segments are no longer thought of as timing units and do not form constituents. Though less intuitive, the u-Model is argued to be superior to the OR-Model in its ability to explain why prosodic patterns do not make reference to fine distinctions of syllable length but instead apply coarsely to a two- (sometimes three) way distinction of heavy and light syllables. The u-Model and the OR-Model make different predictions about what manipulation speakers can do to syllables. The OR-Model by virtue of the constituencies, predicts that manipulation can be done to groups of segments while preserving linear order in ways not possible with the u-Model.

This project studies the internal structure of syllables through experimentation on speakers’ ability to manipulate sub-syllabic entities. Speakers are trained to use a play language known as Verlan where words are deliberately pronounced in reverse, thus cat would be pronounced [tæk]. To avoid biases, training will be done with stimuli from a non-English source carefully checked for structure-neutrality during reversals. Participants will extrapolate their manipulative strategies to English words. That way, results obtained would be free from any undesirable priming effects from the training.

Pilot study by the applicant has shown promising results, with a variety of patterns reflecting fairly sophisticated organization of sub- and supra-syllabic elements hitherto not considered by both the OR- and u-Models. At present, more comprehensive experimentation is needed to establish a more viable theory of the syllable that would capture the insights of the two earlier models while simultaneously breaking new ground. Such a study offers a window into the how the human mind organizes sounds, with important implications for phonological theory and also a general understanding of the human mind.