Introduction to Phonotactics
What is phonotactics?
The cross-linguistic variation of syllable structure has attracted much interest from linguists, but this interest has been restricted by the difficulties of accessing a large amount of data. With the World Phonotactics Database you can examine phonotactic patterns for a large number of languages over large areas.
Phonotactics at its simplest is concerned with the freedoms and restrictions that languages allow in terms of syllable structure. Which sounds can precede and follow which other sounds; whether consonant clusters are allowed, and what sorts are allowed; whether a language has syllabic consonants, and if so which ones; whether length is contrastive in vowels; which sounds can occur in a syllable coda: these are all examples of phonotactic restrictions that can appear in a language.
For instance, in English we can characterise the following words (all of which are one syllable long) according to their phonotactic patterns:
|see||CV:||simple onset, no coda|
|seat||CV:C||simple onset, simple coda|
|tree||CCV:||complex onset, no coda|
|street||CCCV:C||very complex onset, simple coda|
|treats||CCV:CC||complex onset, complex coda|
We can see that at least two of the variables in English are the maximum size of the onset (three consonants, in the examples above), and the maximum size of the coda (only two consonants in the examples given here, though three are possible). It is relevant to examine the kinds of consonants that can occur in different positions: these examples show the privileged status that liquids enjoy as the second element in clusters, and the privileges that are associated with /s/ in the 'outermost' (furthest from the vowel) member of clusters, both in the onset and in the coda.
Similarly, there is great variation in the permissability of certain kinds of segments in different positions: many languages restrict plosives in codas to unaspirated voiceless stops, for instance, or disallow velar nasals in onset position. Both of these are variables that can be applied as filters to the database.
We have included information of a non-phonotactic nature, such as the presence (and number) of tonal contrasts, and the number of contrastive vowels qualities present in the languages, to allow users to check for correlations between phonotactic variables and variables elsewhere in the phonology.