Models and transcription (next meeting to be fixed)

At the present time, there are four important models for the relation between linguistic and phonetic levels of sound structure: phonemic, generative, prosodic, finite state. Find definitions for the basic terms in each of these approaches, and illustrate them using the characteristic examples given for each:

1. morphophonemic / phonemic / allophonic (structuralist): list the distinctive features for each phoneme in English and German as a matrix; then - regarding the features as `dimensions' in `phonological space' (called `correlations' in the Prague school, with individual phonemes in `opposition' based on distictive feature contrasts) - draw diagrammes for the consonant systems, e.g.:

                 p ---- t ---- ....
                /|     /|
               b ---- d ---- ....
               | f ---| s ---- ....
               |/     |/
               v ---- z ---- ....

   What are the main differences between the two languages?

2. underlying lexical / systematic phonetic (generative, incl. lexical phonology): What is the rule for `final devoicing' (Auslautverhärtung) in German? What are the rules for word stress in English?
3. `non-linear', `multi-tier', `hierarchical' (prosodic, including autosegmental phonology and metrical phonology): How is nasal assimilation (e.g. im Bielefeld, im Birmingham described in the autosegmental approach? How is word stress described in metrical phonology?
4. finite state phonology: Describe the syntagmatic structure of English or German syllables using a `finite state transition network' (or `finite state automaton'). Hint: you will find details on my web links on the Reference Materials page.

Try the further exercises on distinctive features below.

And: what is the difference between a theory, a model, a domain/universe (of discourse)?

Would you still say that the IPA is a theory (or maybe a model) of speech sound segmentation and classification?

And if you have questions, don't hesitate to send me an email.

Further exercises on distinctive features

Distinctive features, the relations between them, and rules which explain variations in distinctive feature specifications, are the most important descriptive devices in modern phonology of all kinds.

The following tasks will help you to understand distinctive features thoroughly.

1. How many suggestions for different sets of distinctive features can you find? Hint: Look for Jakobson & Halle; Chomsky & Halle; Ladefoged; ...
2. What is a natural class of segments (based on distinctive features)?
3. Distinctive features are (in phonemic and generative phonologies) arranged in a regular matrix. This matrix can also be thought of as a relation in a database.
4. Try out this example of a distinctive feature database relation, which uses the Jakobson & Halle set of distinctive features for consonant phonemes, and try out different natural classes by clicking `Query' and then clicking input feature(s) (lefthand button, query attribute) and phoneme (righthand button, output attribute(s)). (Check the DB help function if you have a problem, or send me an email.) You can of course make your own distinctive feature tables using other database systems.
5. In autosegmental and related prosodic phonologies distinctive features are arranged in various kinds of graph structure. Illustrate the kinds of structure can you find in the literature? Hint: trees, `book' structures, parallel tramlines ...
6. What kinds of rule can you find which relate different values of distinctive features in different contexts? Hint: How would you relate the `ich-Laut' and the `ach-Laut' in German? How would you relate the clear and dark /l/ in English?

If you prepare your distinctive features matrixes for the task above as a plain text (ASCII) file in the format explained below, and send it to me by email, I can incorporate it into the web database system for you to select and experiment with.

In fact, you can also enter your own database in the format below using the database construction function.

And as an exercise for students interested in programming, download the DB system (which is programmed in HTML and JavaScript), examine the file format for the databases, and create your own databases locally on your own computer.

The format is the following:

1. First line: Names of features separated by one space.
2. Other lines: Specifications of features, separated by one space.
3. (SAMPA notation should be used for the phonemes.)

Example (corresponding to database in the demo):

Phoneme Vocalic Consonantal Compact Tense Voiced Nasal Discontinuous Strident Checked Grave Flat Sharp
p - + + + - - + - 0 + 0 0
t - + - + - - + + 0 - 0 0
k - + + + - - + - 0 + 0 0
b - + + - + - + - 0 + 0 0
d - + - - + - + + 0 - 0 0
g - + + - + - + - 0 + 0 0
f - + + + - - - - 0 + 0 0
T - + - + - - - - 0 - 0 0
s - + - + - - - + 0 - 0 0
S - + - + - - - + 0 - 0 0
v - + + - + - - - 0 + 0 0
D - + - - + - - - 0 - 0 0
z - + - - + - - + 0 - 0 0
Z - + - - + - - + 0 - 0 0
tS - + - + - - + + 0 - 0 0
dZ - + - - + - + + 0 - 0 0
m - - + - + + + - 0 + 0 0
n - - - - + + + - 0 - 0 0
N - - + - + + + - 0 + 0 0
w - - + - + - - - 0 + 0 0
j - - - - + - - - 0 - 0 0
h - - 0 0 - - - - 0 0 0 0
l - - - - - - - - 0 - 0 0
r - - - - - - - - 0 - 0 0