[Author Login]
Title:First Steps in the Acquisition of German Phonology: A Case Study
Authors:Janet Grijzenhout, Sandra Joppen
Abstract:First Steps in the Acquisition of German Phonology: A Case Study

Janet Grijzenhout & Sandra Joppen

The first words in child speech supposedly have a CV-structure

(e.g., Jakobson 1941/68) and they are thus characterised by the

presence of an onset. In this paper, we will present evidence

against this assumption. In the speech of Naomi - the German child

that we examined - we found that between the ages of 1;2;6 and 1,5;01

each word contains at least and at most one consonantal place of

articulation. It is striking that the consonant in question may but

must not be initial. We argue that previous accounts of the

acquisition of syllable structure (e.g. Fikkert 1994) cannot be right

and that the facts are best captured by an analysis that uses the

correspondence version of Optimality Theory (McCarthy & Prince 1995).

We will formulate a constraint that says that every word should have

a consonantal place of articulation (C-Place). The constraint called

ONSET, which says that syllables must have a consonant in onset

position, never plays a prominent role in German child speech.

We will also show that Naomi does not realise fricatives at the

earliest stage of acquisition. Later, she replaces word-initial

fricatives by voiced stops and we will formulte constraints that

help to explain that the best output for a word which has an initial

fricative in the adult form, is a word which has an initial voiced

stop. These constraints are satisfied in words which begin in stops,

nasals, or /l/. A condition against inserting material (formulated

as 'DEP I-O' by McCarthy & Prince 1995) seems to be inviolable.

From 1;6;05, Naomi realises stops and approximants instead of

fricatives (e.g. fertig --> 'jatig, 'datig (1;7:16, 1;7:27) 'ready').

This suggests to us that faithfulness conditions such as 'realise the

input feature [+continuant]' (IDENTITY [+cont]) begin to play a more

prominent role. We aim to show that at the initial word-stage, each

position in the word has its own preferences for particular

phonological features and that gradually, preference laws begin to be

outranked by faithfulness conditions on separate features.

Type:Paper/tech report
Article:Version 1