The Power of Pondering Spatchcock

Previous research has suggested that, compared to the definite determiner the, demonstrative determiners in English (this/that) are more difficult to process. Listeners are slower to identify the nouns that follow demonstratives, and they are less sensitive to fine phonetic detail in those nouns. This project presents a subset of a larger study designed to answer two questions: 1) what other types of processing can demonstrative determiners interfere with? 2) What other types of ‘difficult’ linguistic features behave like demonstratives? Today, I present an experiment that tests three types of processing: a) a phoneme monitoring task examines phonological processing structures with ‘difficult’ linguistic features; b) a recognition memory task examines lexical encoding of those structures; and c), a same/different voice manipulation in the memory task examines phonetic encoding of those structures. The ‘difficult’ features at issue here include demonstratives like this/that (compared to easier definite the); low lexical frequency (compared to easier, high frequency); and high phonological neighbourhood density (high PND, compared to easier, low PND).

Participants were presented with audio recordings of sentences containing target phonemes at the start of ‘difficult’ words and ‘easy’ words, with instructions to press a button when they detected the target phoneme. During each sentence, listeners’ pupil size was recorded, as a measure of cognitive effort. After completing each block of sentences, participants then completed a recognition memory task for the words that contained the target sound in each of the sentences in the previous block. Results confirm that demonstrative determiners interfere with phonological processing. Compared with definite determiners, RTs in the phoneme monitoring were slower slower with demonstratives, and pupil sizes were larger, with a delayed peak dilation. Low-frequency words also elicited larger peak pupil dilation in the phoneme monitoring task, compared to high-frequency words. However, there was no effect of frequency on RT, and unexpectedly, accuracy was higher with ‘harder’ low-frequency words. Both determiner and frequency stimuli also showed an unexpected late effect of difficulty on pupil size, which emerged well after the initial peak dilation. PND, however, had no effect on phoneme monitoring performance or pupillometric responses. For all three stimulus types, the recognition memory task was mostly inconclusive. In sum, these results suggest that the difficulty of certain lexical properties (frequency but not PND) and phrase-level syntactic properties (determiner + noun sequences) affect participants’ phonological processing, but are of less consequence for lexical or phonetic encoding. Further research is ongoing to investigate the role of ‘difficult’ syntactic properties at the clause level, and the effects of all types of ‘difficult’ features on semantic processing and cognitive effort in passive listening.

Link to session