Project 4: Perceptual salience of FPD: Cross-linguistic comparison of contextual sensitivity
Identification of segments and words often requires FPD to be evaluated and potentially re-evaluated in the light of following speech. E.g. bad girl may sound more like bag girl due to regressive place assimilation. The listener must evaluate the /g/ in bag girl relative to the place of articulation of the following consonant. Previous research has shown an exquisite sensitivity to the FPD of assimilation and has highlighted a context-sensitive recognition process for assimilated speech. However, current debate focuses on whether the perceptual system becomes tuned to the assimilation processes specific to a listener’s native language. We will address this question by manipulating both the language of the stimuli and the listener’s native language.
This project will combine phonetic and psycholinguistic expertise across a range of European languages. We will use pairs and triplets of languages varying in the presence of key assimilatory processes. E.g. place assimilation (English [present], French [absent]); voice assimilation (Czech [present], Dutch [present], English [absent]). Language triplets will be particularly important because we can then use stimuli for which two different sets of listeners are matched in terms of their experience of the language (i.e. it is a foreign language to both), but experience of the particular assimilation type varies. We can then examine the extent to which perception of the foreign language is shaped by language-specific experience. The influence of FPD on contextual sensitivity will be examined by developing continua with varying degrees of assimilation and examining how native language experience shapes perception across the assimilation spectrum. Computational analyses, developed alongside the behavioural work, predict that perception of mild assimilation will be dominated by language-universal perceptual mechanisms, whereas perceiving stronger assimilations will rely on language-specific compensation.
Materials Interdisciplinary collaboration in the generation of suitable materials representing the assimilation continua in the respective languages will be crucial. This will require a good phonetician to do detailed phonetic analyses to identify dimensions of variation within each assimilation process and for each language. This will allow identification, for each type of assimilation studied, of both the acoustic cues that are shared between languages and those that are particular to each language. It will also allow generation of two types of continua: those that are plausible examples of assimilation in more than one language, and those that are idiosyncratic.
Methods Psycholinguistic methods will range from ‘low-level’ tasks (e.g. phoneme monitoring, mismatch negativity) to more lexical procedures (e.g. cross-modal priming, word monitoring). The language-dependency of the recognition system may vary with the processing level, with low-level processes being language-general and higher-level processing being more language specific. The range of tasks will ensure that variation of this type is captured.
Young researchers One ER (Clayards) at York, and one ER (Niebuhr) at Aix (also working with Geneva). They will most likely visits Nijmegen and Prague. This project involves phoneticians and psychologists.
Links Project 4 is informed by Project 2, and is related to Project 6. It may also inform Theme IV projects.
Working on this project: » Prof Gareth Gaskell » Prof Noël Nguyen » Prof Dr Ulrich Frauenfelder » Prof Dr Oliver Niebuhr » Dr Meghan Clayards » Dr Christine Meunier
