Psycholinguistics is concerned with how people understand (perceive and comprehend) spoken and written language, how people produce language, and how languages are learned (and lost).
Phonetics is concerned with describing the speech sounds that occur in human languages. Articulatory phonetics describes how speakers produce speech sounds (speech production), speech perception is concerned with how listeners process speech sounds, and acoustic phonetics studies the physical properties of the sound wave that travels from the speaker's mouth to the listener's ear.
I started to become actively involved in infant speech perception research when Linda Polka (http://www.mcgill.ca/scsd/faculty/polka/) asked me to collaborate with her on cross-language infant vowel perception. The focus of our research has been on
My work in this area started when I became JamesFlege's (http://jimflege.com/) first post-doc (1987-1989). Part of our work on Flege's Speech Learning Model involved cross-language studies, but the main focus was on the effect of language experience on adult second language learners' ability to perceive and produce nonnative speech sounds. Most of our studies were on the perception and production of English vowels by nonnatives (Bohn & Flege 1990, 1992, 1997; Flege, Bohn & Jang 1997). We also did a study of VOT perception by native English and native Spanish listeners (Bohn & Flege 1993).
The collaboration with Winifred Strange ( http://web.gc.cuny.edu/Speechandhearing/labs/sapl/indexsapl.htm) started when we discovered a common interest in vowels of North German. In addition to studies on the acoustic correlates of the (perceived) identity of North German vowels in infants and adults (see Acoustic Correlates of Perceived Vowel Identity), this interest resulted in studies of cross-language acoustic and perceptual similarity of vowels (Strange et al. 2004, 2005).
Catherine T. Best (http://marcs.uws.edu.au/people/best/index.htm) and http://www.haskins.yale.edu/staff/best.html) and I have collaborated on studies which test the predictions of her Perceptual Assimilation Model. We conducted cross-language studies on the perception of approximants (Best & Bohn 2002) and vowels (Best, Halle, Bohn & Faber 2003), and we are currently examining the perception of Nthlakampx (http://sail.usc.edu/~dbyrd/listening/thompsonlist.html) consonants and of English labial, labiodental, and dental consonants by nonnative listeners.
Terry Gottfried (http://www.lawrence.edu/fast/gottfrit/) spent a Fulbright-sponsored sabbatical here in Arhus. We studied the effects of speaking rate on cross-language vowel perception (i.e., perception of American English vowels by Danish listeners; Gottfried & Bohn 2002).
Diane Kewley-Port (http://www.cogs.indiana.edu/people/homepages/kewley.html) spent part of a sabbatical here in Arhus. I contributed to the Danish part of her studies on how different vowel systems affect vowel discrimination and identification (Kewley-Port, Bohn, & Nishi, K. 2005).
Anja Steinlen ( http://univis.uni-kiel.de/prg?show=info&key=867/persons/2007w:philos/zentru_5/itundm/steinl) was one of my first PhD students. Her PhD thesis (Steinlen, Anja K., 2005, The influence of consonants on non-native vowel production: A cross-linguistic study. G. Narr: Tübingen) provided the basis for a number of acoustic studies on which we collaborated (see Comparative Phonetics) and for a study of phonetic context effects on cross-language vowel perception (Bohn & Steinlen 2003).
I have also worked on various aspects of foreign accented speech with Anja Steinlen (on the acoustics of Danish accented vowels; Steinlen & Bohn 1999), and with my former students Leila Trapp (http://personprofil.aau.dk/Kontakt/112846) on the effects of training a nonnative consonant contrast on perception and production, and with Rikke Bundgaard-Nielsen (http://marcs.uws.edu.au/people/Nielson/index.htm) on the intelligibility of Danish accented vowels (Bohn & Bundgaard-Nielsen 2008).
I have just started on a project which examines the comprehensibility and the psychological cost involved in listening to the languages that are typically used at Danish Universities: Apart from Danish, these are Norwegian, Swedish, and foreign accented and native English.
A great source for anyone interested in foreign accented English (as well as in varieties of English) is Steve Weinberger's speech accent archive (http://accent.gmu.edu/), to which I have contributed with a few samples.
The study of phonetic similarity is of broad theoretical interest in general phonetics and, of course, basic for the study of cross-language speech perception and second language speech. Almost all of the studies that I have conducted are in some way concerned with phonetic similarity. For an overview, see Bohn 2002 and my keynote at the New Sounds 2007 conference.
Comparative phonetics is the basis of the study of phonetic universals and typology. Cross-language speech perception is part of comparative phonetics, but traditionally, acoustic comparisons of speech sounds across languages have been the core area of comparative phonetics. Here are some studies in which acoustic comparisons have been used to address issues in comparative phonetics:
Bohn & Flege 1990, 1992, Strange, Bohn, Trent & Nishi 2004, Strange, Bohn, Nishi & Trent 2005, Bohn & Steinlen 2003, Bohn 2004.
For a long time, speech scientists used to believe that vowels are pretty well specified by the articulatory target configuration of the vocal tract in vowel articulation, and by the corresponding acoustic target frequencies. This traditional view of vowel identity was challenged by Winifred Strange and her colleagues. Strange's Dynamic Specification Theory (DST, Strange 1989, Journal of the Acoustical Society of America 85, 2135-2153) postulates that important information on vowel identity is specified over the dynamic portions of syllable onsets and offsets in consonant-vowel-consonant syllables. A study on North German vowels confirmed the predictions of the DST (Strange & Bohn 1998), and one of my studies of infant vowel perception (Bohn & Polka 2001) showed that, for prelinguistic infants, the dynamic portions of syllable onsets and offsets (in consonant-vowel-consonant syllables) specify vowel identity as least as reliably as target formant frequencies.
Fering is a dialect of the West Germanic language North Frisian. Of the three Frisian languages, West Frisian (ca. 300,000 speakers in the northwest Netherlands) is least endangered, East Frisian (less than 200 speakers in three villages in Lower Saxony) is moribund, and North Frisian (ca. 7,000 speakers) is acutely endangered. North Frisian itself consist of several mutually unintelligible dialects, the most important division being between Mainland North Frisian (with at least five dialects), spoken in the extreme northwest of the mainland of the German state Schleswig-Holstein, and Island North Frisian spoken on the North Sea islands Sylt, Föhr, Amrum, and Heligoland. Of these, only the dialects spoken on Föhr (Fering) and the neighboring island Amrum (Öömrang) are mutually intelligible. All dialects of North Frisian are acutely endangered: There are no adult monolingual speakers of any dialect, many speakers are fluent in 3 or more languages, and the number of speakers of the dialects of North Frisian ranges between less than 100 and ca. 2000.
It appears that Fering stands strongest among the North Frisian dialects. It has the largest number of native speakers (1500 on the island plus an unknown number of emigrants especially in New York and Northern California), and it differs from most other dialects in that it is used not just at home but also publicly, especially in the Western part of the island (Westerland Föhr with the Weesdring dialect of Fering). The Westerland Föhr villages of Söleraanj, Olersam, and Taftem (German: Süderende, Oldsum, Toftum) are considered a stronghold of Fering, and the variety of Fering spoken in these villages is reasonably homogenous. It is the form of Fering as spoken in these villages that is documented in this project.
The first stage in the documentation of Fering had two aims:
1. As is customary in dialectological and sociolinguistic studies which aim to document conservative and/or endangered speech forms, the speaker group in the first stage of the documentation consisted of older male speakers of Fering. Consultations of dictionaries, textbooks, and native speaking informants led to the construction of word and sentence lists, which were used to elicit the segmental inventory of Fering in different phonetic contexts (for vowels and consonants), at two speaking rates (for selected vowels), and at two speaking styles (for selected consonants). Additionally, the talkers were recorded reading the fable "the North Wind and the Sun" and the Swadesh word list. The complete corpus (on CD-ROM) can be obtained by sending a request to firstname.lastname@example.org (For samples, click the links below.)
2. The fairly comprehensive acoustic documentation of the sounds of Fering is used to address questions of general interest in phonetic typology. For instance, how does a language with a large consonant and vowel inventory organize its sound system? Fering has been described as having maximally 29 consonant phonemes with typologically unusual contrasts (dental vs. alveolar vs. postalveolar place of articulation for stops, nasals, and laterals). With respect to the vowels of Fering (15 stressed monophthongs, 7 diphthongs, and 3 triphthongs), the general questions that have been addressed concern how Fering vowels are distributed in the acoustic vowel space, how a language with a large vowel inventory like Fering differentiates vowels that are close in the F1/F2 space, whether consonant-vowel coarticulation affects the acoustic differentiation of vowel categories, and to what extent the implementation of the long-short vowel contrasts is affected by speaking rate. The results from this study were published as Bohn 2004.
The files that can be accessed here have all been digitized at 44.1 kHz after low-pass filtering at 20 kHz. The speaker is Ocke Danklef Bohn. (For detailed information on the recording and digitization procedure, as well as sound files exemplifying vowels and consonants of Fering as produced by 10 older native speakers, consult the CD-ROM.)