Academic journal article Journal of Slavic Linguistics

A Conspiracy of Gliding Processes in Polish

Academic journal article Journal of Slavic Linguistics

A Conspiracy of Gliding Processes in Polish

Article excerpt

Abstract. One of the significant consequences of the autosegmental theory of representations is a different way of drawing the distinction between glides and vowels. The distinction is made in terms of syllable structure rather than in terms of the feature [[+ or -]syllabic], as was the case in SPE phonology. This article pursues the problem of the glide-vowel distinction for Polish and shows that with few exceptions this distinction is derivable from distributional generalizations. The generalizations are first stated in terms of rules and then reanalyzed in terms of OT constraints. It is argued that the OT-based analysis is superior to the rule-based analysis.

1. Introduction

The goal of this article is to compare the analysis of gliding processes in Polish in two frameworks, the rule framework and the Optimality Theoretic framework. It is argued that the latter is superior to the former.

The distribution of high vowels and glides in Polish has been studied before, notably by Gussmann (1980), Rubach (1984), Rubach and Booij (1990a), and Bethin (1992). It became a central issue in autosegmental phonology, which recognizes a distinction of three tiers of representation: the melodic tier, the skeletal tier, and the syllable tier (Clements and Keyser 1983, Levin 1985, Hayes 1989, and others). The adoption of this theory of phonological representation is tantamount to eliminating the feature [[+ or -]syllabic] from the repertoire of phonological features. The distinction between [i] and [j] as well as [u] and [w] is delegated to syllable structure. At the melodic tier, the rep resentation of vowels and glides is identical. The difference is a matter of syllabification: vowels are syllable nuclei while glides are syllable margins (onsets or codas). The details of how this difference is represented depend on the type of skeletal theory employed. In the moraic theory (Hayes 1989), which I adopt here, syllable nuclei carry a mora while syllable margins do not. (1) Since vowels and glides are non-distinct prior to syllabification, their underlying representation is identical at both the melodic tier and the moraic tier (1a). In order to be syllabified into an onset or a coda, vocalic segments must lose their moras (1b). Glide derivation is therefore a process of mora deletion.

(1) jej [jej] 'her'


As is well known, syllable structure is non-distinctive at the underlying level and therefore is always derived. If, as noted above, the distinction between vowels and glides is a matter of syllabification, the prediction is that this distinction should be derivable because all syllable structure is derivable. I show that with few exceptions this prediction is borne out in Polish.

Section 2 introduces the data (3) and shows how gliding processes can be analyzed in the rule framework. Section 3 looks at the same facts from the perspective of Optimality Theory (Prince and Smolensky 1993/2004, McCarthy and Prince 1995). Section 4 is a summary of the conclusions.

2. Rule-Based Analysis

The basic generalization regarding the distribution of high vowels and glides is that //i// and //u// erect syllables in the three contexts in (2); (4) otherwise we have glides. Dots denote syllable boundaries in polysyllabic words.

(2) a. word-initially before a consonant: (5)

    iskr+a [i.skra] 'spark', Iren+a [] 'Irene'
    urwis [ur.vis] 'urchin', u+raz [u.ras] 'injury'

    b. between consonants:

    mit [mit] 'myth', komitet [ko.mi.tet] 'committee'
    mur [mur] 'wall', profes+ur+a [] 'professorship'

c.  word-finally after a consonant:

    grub+i [] 'fat' (masc nom pl), alibi [] 'alibi'
    mit+u [mi.tu] 'myth' (gen sg), tabu [ta.bu] 'taboo'

Postvocalic and prevocalic contexts induce gliding.

(3) Gliding after a vowel

   a. maj [maj] 'May', klejnot [klej.not] 'gem', slajd [slajt] 'slide',
   sejm [sejm] 'parliament'

(3) b. … 
Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.