Mobilex Mobility Grant
Paving the Way for a More Reliable Linguistics
The Danish Council for Independent Research, with support from the Marie Curie Program under FP7.
Speech transcriptions, the decoding of sounds in letters or other visual symbols, are important means for investigating pronunciation. For instance, sounds of a particular dialect can only be statistically analysed if they have been transcribed. Transcriptions are also used in foreign language education, speech therapy, and language counselling (like in asylum requests). Transcriptions are currently often made by speech processing software, but have to be corrected manually: it cannot fully replace human codings. Some sounds cannot be analysed by speech processing software at all. But are human codings reliable? The answer is probably no. Recently, it turned out that linguistically subjects are easily biased towards an overall accent they perceive in a speaker, such that coders exposed to a certain accent may expect and therefore actually hear prototypical characteristics of this accent—even if these are not present in the phonetic signal. Even professional coders turned out to be subject to coder bias. The scope of this deviation in formal transcription tasks is still unknown, neither has a standard procedure been developed in order to arrive at better results. Since human codings should be as accurate and as reliable as possible, this research aims at the development of a “gold standard” for linguistic transcriptions.
30 September 2013
January 2014-December 2015
Seed Grant Interacting Minds Center
Does Musical Perception Influence Segmental Perception in Language?
Interacting Minds Center, Aarhus University
Musical skills are known to facilitate acquisition of prosody in second language acquisition. But does it also facilitate perception of vowels and consonants? Human perception is subjective and segmental perception is easily biased by an overall accent of the speaker. Can musical perception lead to a more objective way of listening?
We investigate this for consonant perception in Chinese. For the same set of data, one group of listeners will be informed they listen to Mandarin data and an other group will be informed it is Wu dialect. The crucial difference is that Mandarin has a two-way plosive contrast (tai-thai), whereas Wu has a three-way plosive contrast (tai-thai-dai). We expect perception is influenced by the known contrast. We will repeat the experiment with the same subjects but in this case with music fragments in between. A control group will just repeat the first experiment.
If musical perception makes segmental processing more objective, this might be a way to lead to more objective transcriptions of language.
Phonological Grammar and Frequency Effects
Deutsche Forschungsgemeinschaft, Graduiertenkolleg 1624/1, Freiburg University
In language change, we usually find that words with a relatively low word frequency change at a faster rate than words with a relatively high word frequency. Reversely, in reduction processes, we observe that high-frequency words tend to change first and low-frequency words change later. How do these facts relate to variable processes in which neither analogical change nor reduction is involved? Does frequency also play a role, for instance, in stable patterns of variation or in loanword adaptation?
This dissertation investigates phonological variation and change in three case studies: the pronunciation of the long vowel <ä> in Standard German, Dutch loanword integration in Indonesian, and sequential voicing in Japanese. It shows that frequency effects occur such that relatively high-frequency words adapt to a general phonological rule and relatively low-frequency words behave differently. This exceptional behaviour of low-frequency words may be related to their relatively unstable or opaque lexical representation or their opaque morphophonological structure.
This thesis also investigates the relation between frequency effects and grammar. It is shown that these two factors are not independent, as suggested by earlier literature, but, on the contrary, intimately related. Frequency effects are sensitive to grammatical structure. This calls for an amalgamation of phonological models which were previously regarded as disconnected; therefore this thesis proposes a combined Exemplar-Prototype-Optimality theoretical model (EPOT).
October 2009-September 2010