Visiting Lecture Course: Signal processing, phonetic analysis automation and Praat scripting
- Education level
- External lecturers
- Francesco Cangemi
Short CV: Francesco Cangemi
Francesco Cangemi (Cologne) pursued his PhD in linguistics at the Laboratoire Parole et Langage in Aix-en-Provence. His published dissertation on prosodic detail in Italian (Cangemi 2013) focuses on whether and how abstract canonical forms in intonation could be enriched in order to account for various sources of phonetic variability. He is currently exploring the limits of classical structuralist phonological categories by focusing on speaker and listener-specific behaviors.
The aim of this course is to provide a brief overview on how actual research in experimental linguistics is planned and executed.
We will retrace the steps behind an interdisciplinary study which uses experimental synchronic evidence (a nonce word reading task) in order to support mechanisms invoked in diachronic reconstructions (lexical contamination and phonological analogy). The findings have implications on the storage of linguistic knowledge (the nature of the mental lexicon) and on the notion of categoriality as used in linguistics.
We will focus on the planning of the study and the collection, processing and analysis of data. In the process, we will introduce the basic aspects of various tools for phonetic research, including Perceval (presentation of stimuli), Praat (phonetic analyses), ASSI (forced alignment), R (plotting and statistical analyses).
Day 1 - lecture: theoretical background (romance linguistics)
Day 2 - handson: stimulus preparation (experimental setup; Perceval)
Day 3 - handson: data collection (stimulus management; Praat)
Day 4 - handson: feature extraction (formant analysis; ASSI and Praat)
Day 5 - handson: data analysis (plotting and statistics; R)
Day 6 - handson: data analysis (plotting and statistics; R)
Day 7 - lecture: theoretical implications (psycholinguistics)
Time and Dates:
27.4., 29.4., 30.4., 4.5., 6.5., 7.5., 8.5. from 16.00 til 17.30
The course is tailored for students of electrical engineering, informatics, speech communication and linguistics ; no specific training in programming or linguistics is required. The hands-on sessions (Days 2-6) will involve practical work in small groups.
Dyads of participants will be asked to choose one of the four hands-on activities and produce by June 30th a folder containing a brief description of the activity, the relevant scripts and, where applicable, the output files. Ideally, dyads will be composed by one student in engineering and one student in linguistics. Upon request, individual students might be allowed to produce instead a short essay on one of the remaining course topics (from analogy in romance phonology and exemplar-based approaches to the mental lexicon to formant extraction and mixed-effects modelling).