Lieb, Hans-Heinrich. 2008. "The case for two-level phonology"
Lieb, Hans-Heinrich. 2008. "The case for two-level phonology: German Obstruent Tensing and Nasal Alternation in French".

In: Robin Sackmann (ed). Explorations in Integrational Linguistics: four essays on German, French, and Guaraní. (Studies in Integrational Linguistics, 1). Amsterdam; Philadelphia: Benjamins. (= Current Issues in Linguistic Theory 285). 21–96.


A. Table of contents

  1. Aims and procedure
  2. Structured sound sequences in a two-level phonology
    2.1 Two-level phonology
    2.2 Sound sequences and structured sound sequences
    2.3 Constituent structures
    2.4 Syllables and syllable sequences
    2.5 Intonation structures
    2.6 Notation
  3. Connection, variant, and type of structured sequence
    3.1 Phonological connection
    3.2 Phonetic connection
    3.3 The variant relation
    3.4 Basic classes of structured sound sequences
    3.5 Second-level and third-level classes
  4. The framework. Word-final tensing
    4.1 Introduction
    4.2 Two aspects of GOT
    4.3 Further clarifications
    4.4 Representation. Three notions of variant
    4.5 Word-final tensing
  5. Preconsonantal tensing
    5.1 Motivation
    5.2 The Obstruent Tensing Theorem
    5.3 Comments on the antecedent
    5.4 Comments on the consequent
    5.5 Objections
    5.6 Tensing and schwa: example
    5.7 The interplay of tensing and schwa
  6. The Alternation Theorem
    6.1 Overview
    6.2 The Nasal Alternation Theorem
    6.3 Paraphrase and examples
    6.4 Examples
    6.5 Further explanations
    6.6 Remarks on the connection functions
    6.7 On the logical form of the Alternation Theorem
  7. Defending the Alternation Theorem
    7.1 The status of nasal vowels
    7.2 Nasal consonants: stable and unstable
    7.3 Dealing with liaison
    7.4 The problem of phonological schwa
    7.5 Empty-Vc sequences: a treatment of schwa
    7.6 Empty-C sequences: a treatment of h-aspiré
    7.7 Scope of the theorem: suppletion
  8. Generalizations
    8.1 The Consonant Alternation Theorem
    8.2 A critical case for the Consonant Alternation Theorem
    8.3 Approaches to consonant alternation
    8.4 Is the phonological level dispensable?
    8.5 Summary and conclusions

B. Abstract

A fundamental feature of Standard German, obstruent tensing, and an equally fundamental feature of Standard French, nasal alternation, are analysed within the framework of Integrational Phonology, a recent two-level phonology. As already suggested in part of the literature, nasal alternation is found to be a special case of consonant alternation (the phenomenon of 'unstable consonants'). Obstruent tensing is therefore compared, in French, not only to nasal alternation but, more generally, to consonant alternation.
It is argued for German that word-final tensing should be kept apart from all other obstruent tensing, whose true nature as preconsonantal tensing is not recognized in the literature. Correspondingly, word-final tensing is shown to be just one of several conditions identifying phonetic words, whereas preconsonantal tensing is covered by the Obstruent Tensing Theorem. This theorem shows interrelations between phonological connection, phonetic connection, and the variant relation. The crucial point is the fact that a phonetic variant of a certain phonological sequence ends in a tense or a lax obstruent depending on the beginning of another phonological sequence with which the first is connected: tense, if the second sequence is consonant-initial, lax, if it is vowel-initial. In short, the occurrence of phonetic sounds may be conditioned by phonological environment — a statement impossible by definition in a one-level phonology.
It is then argued for French that there is a largely analogous situation in the case of nasal alternation or, more generally, consonant alternation. Leaving the word-identification aspect undiscussed, the Nasal Alternation Theorem is formulated, and subsequently, the Consonant Alternation Theorem, to account for the appearance vs. non-appearance of certain consonants at the end of a variant when a phonological sequence is connected with a vowel-initial vs. a consonant-initial phonological sequence. Differently from German, liaison phenomena have to be taken into account; but just as in the case of German, it proves necessary, in defending the theorem, to confront other basic phonological problems of the language, which are solved by applying the tools of a two-level phonology.
Both the Obstruent Tensing Theorem for Standard German and the Nasal and Consonant Alternation Theorems for Standard French are eventually upheld. The essay thus provides solutions to two fundamental, much-discussed phonological problems of, respectively, German and French phonology that eventually draw on the means made available by a two-level as opposed to a one-level phonology.
Moreover, it is shown — apparently for the first time — that German Obstruent Tensing and French Consonant Alternation (in particular, Nasal Alternation), two seemingly unrelated phenomena that lack any historical connection, each exemplify precisely the same phonological-phonetic mechanism, to the extent that the theorem on German and the theorems on French have identical logical structure (with a single qualification) and are obtained from each other by interchanging different linguistic terms that have the same, or very similar, theoretical status.
In summary, nobody so far appears to have given a serious account in a strictly one-level phonology either of German Obstruent Tensing or of Nasal or Consonant Alternation in French. The present analysis may claim the following merits:
  1. in either language, it completely covers all phenomena
  2. in either language, it would easily become part of a more comprehensive theory of the sound system
  3. it exhibits a single phonological-phonetic mechanism at work in both languages, a mechanism for which even greater generality across languages may be suspected
The analysis uses the means of Integrational Phonology, a 'declarative', two-level phonology of considerable power, in essential ways, thus making a strong case for such phonologies.