Syllable structure

  We assume that each lexeme has a phonological structure consisting of a sequence of syllables, each syllable having the conventional structure shown below:

 As in Cahill (1990b) and Bleiching (1992), we define syllabic structures by means of simple context-free phrase structure rules:

syllable $\rightarrow$ onset rhyme
rhyme $\rightarrow$ stress peak coda
coda $\rightarrow$ body tail
disyllable $\rightarrow$ syllable syllable
trisyllable $\rightarrow$ syllable syllable syllable

A syllable consists of an onset and a rhyme; a rhyme consists of a stress value, a peak and a coda; and a coda consists of a body and a tail. A disyllable consists of two syllables, and a trisyllable of three. We can use DATR to express these context-free phrase structure rules. Bleiching (1992) does not address the issue of polysyllabic roots explicitly. She provides only for monosyllables and disyllables with a final schwa syllable. She does not class this latter as a separate syllable at all, but rather as a ``coda extension''.

Syllable:
    <phn $yll form>  ==        "<phn $yll onset>" "<phn $yll rhyme>"
    <phn $yll rhyme> == Stress "<phn $yll peak>"  "<phn $yll coda>"
    <phn $yll coda>  ==        "<phn $yll body>"  "<phn $yll tail>"
    <phn root>  == <phn syl1>
    <> == Null.

Disyllable:
    <> == Syllable
    <phn root>  == <phn syl2> <phn syl1>.

Trisyllable:
    <> == Syllable
    <phn root>  == <phn syl3> <phn syl2> <phn syl1>.

Strictly speaking, the definition of Syllable given above is a rule schema, rather than a rule, since it makes crucial use of a variable $yll that ranges over attributes (syl1, syl2, ..) that denote syllable positions. Note also that the maximally unspecified path (<>) at the Syllable node is defined by reference to Null which always returns the empty sequence as its value. An <onset>, <peak> or <coda> which is left undefined at lower levels of the hierarchy will, as a consequence, end up as null.

The definitions of di- and trisyllables number the syllables from the right. This is a language-specific aspect of our analysis and reflects the fact that German morphology primarily involves suffixation. Reference to final syllables is thus more frequent than reference to the initial syllables and it is technically convenient to have a constant identifier (syl1 here) for final syllables.

Given this set of axioms for syllabic structure, we can now use them to help define particular concrete (poly)syllables. Here, for example, are possible definitions for a couple of monosyllabic suffixes -es and -en, realised phonologically as /@s/ and /@n/.

Suffix_es:
    <> == Syllable
    <phn syl1 peak> == @
    <phn syl1 coda> == s.

Suffix_en:
    <> == Syllable
    <phn syl1 peak> == @
    <phn syl1 coda> == n.

Likewise, a disyllabic word root such as Tutor or irgend can be specified in terms of the individual components of its two syllables.

Tutor:
    <> == Disyllable
    <phn syl2 onset> == t
    <phn syl2 peak> == u:
    <phn syl1 onset> == t
    <phn syl1 peak> == O
    <phn syl1 coda> == r.

Irgend:
    <> == Disyllable
    <phn syl2 peak> == i:
    <phn syl2 coda> == r
    <phn syl1 onset> == g
    <phn syl1 peak> == @
    <phn syl1 coda> == n t.

From these node definitions, taken together with the axioms for syllable structure given above, we can now infer that:

Suffix_es:
    <phn root form> = @ s.

Suffix_en:
    <phn root form> = @ n.

Tutor:
    <phn root form> = t u: t O r.

Irgend:
    <phn root form> = i: r g @ n t.

In the case of Tutor and Irgend, the phonological root form emerges as the result of concatenating the <phn syl2 form> value and the <phn syl1 form> value as determined by the equations given at the Syllable and Disyllable nodes.

Crucially for our approach to inflection, the realisation of any component of the phonological structure can be determined by morphosyntactic features. The idiosyncratic noun Medikus, for example, which has the plural form Meditsi, could be represented thus:

Medikus:
    <> == Disyllable
    <phn syl2 onset> == m
    <phn syl2 peak>  == e:
    <phn syl1 onset> == d
    <phn syl1 peak>  == i:
    <phn syl1 coda>  == k
    <phn syl1 coda  plur> == t s.

This immediately yields the distinction between singular and plural roots that is required for this word:

Medikus:
    <phn root form sing> = m e: d i: k
    <phn root form plur> = m e: d i: t s.

This aspect of our approach can also be usefully illustrated by the core component of our treatment of the inflectional phonology of the definite article and relative pronoun forms:

Der:
    <> == Strong_2
    <phn syl1 onset> == d
    <phn syl1 peak> == e
    <phn syl1 peak nom sing femn> == i:
    <phn syl1 peak nom sing neut> == a
    <phn syl1 peak gen sing masc> == E.

This says that the phonology of Der is determined (by default) by inheritance from the Strong_2 declension; that Der's onset is d; that its (default) peak is e; but that the nominative singular feminine and neuter peaks are i: and a respectively and the genitive singular masculine peak is E. The details and consequences of this treatment of Der are spelt out in much greater detail in the section on the definite article.