LINGUIST List 23.4233

Wed Oct 10 2012

Review: Phonology: Operstein (2010)

Editor for this issue: Rajiv Rao <rajivlinguistlist.org>



Date: 10-Oct-2012
From: Thomas Purnell <tcpurnellwisc.edu>
Subject: Consonant Structure and Prevocalization
E-mail this message to a friend

Discuss this message

Announced at http://linguistlist.org/issues/21/21-2492.html
AUTHOR: Natalie OpersteinTITLE: Consonant Structure and PrevocalizationSERIES TITLE: Current Issues in Lingustic Theory 312PUBLISHER: John BenjaminsYEAR: 2010

Thomas C. Purnell, Department of English, University of Wisconsin-Madison

SUMMARY

Natalie Operstein’s work has two stated goals: first, to provide a typology ofconsonant prevocalization (CP) through cross-linguistic documentation andclassification; and second, to provide an autosegmental analysis of CP phenomenausing a vocalic node for consonants. Operstein defines CP as “phonologicalprocesses which have in common the development of a vocalic prearticulation byconsonants” (p. 3), not to be confused with diphthongization of the precedingvowel. The first two paragraphs of the preface provide an indication of the typeof analysis Operstein will provide (p. 1), namely, one drawing on AP, dependencyphonology and feature geometry. Specifically, overt distinctive features under aconsonant’s vocalic node are realized temporally, prior to other consonantalfeatures, using Articulatory Phonology (AP) tiers (Browman & Goldstein 1986).Operstein includes a wide range of phonological processes under CP, with a fewexamples being: excrescent vowels (e.g., English); prepalatalization, regardlessof whether palatalization is induced by a following /i, j/ or is a contrastiveaspect of the consonant (e.g., Estonian); and vocalization of consonants (e.g.,Catalan). In short, over 150 varieties of dialects/languages (those listed inAppendix II) are examined for various effects that can fit the CP pattern.

Structurally, the book is divided into two parts plus a conclusion. The threechapters comprising Part 1 (“The theory”) cover Operstein’s theory andjustification for her treatment of the data that makes up the two chapters ofPart 2 (“The data”); this introduction of a theoretical treatment of CP,followed by the relevant data, gives the reader Operstein’s perspective up frontand allows all data to be contextualized within the theory. There is aconclusion chapter (Chapter 6) with no overt structural break from the Part 2chapters; the conclusion’s inclusion with Part 2 is likely for balance in termsof numbers of chapters. The end material includes references, two appendices(“Rosapelly’s vocaloid” and “Languages in the survey”), an index of languages inthe text and a two-page index of subjects and terms. The Rosapelly appendix is atwo-page summary of Rosapelly’s focus on fixed articulatory positions for vowelsand consonants. The useful nine-page summary of the languages in the study goesbeyond just listing the languages and the page number on which each one occurs,as in other works; here, each language is listed with references Operstein usedfor that language and additional notes (i.e. sound changes and relevantenvironment). Unfortunately, no author references are provided either in theappendix or indices.

After a two-page preface, Chapter 1 (“Consonant prevocalization”) introduces thereader to the topic of consonant prevocalization, Operstein’s approach to thetopic, and the organizational structure of the work. In the first half of thechapter (sect.1.1 to sect.1.4), Operstein clarifies for the reader some termsrelated to the theory of prevocalization (sect.1.5 and sect.1.6). Perhaps themost crucial definition is of the “prevowel” -- the nonsegmental, vocalicprearticulation of a consonant -- as a cover term for a host of other termsappearing over the years, such as epenthetic vowel, parasitic vowel, etc. (p.7). The greatest benefit of taking a phonological view of CP is that it allowsfor a strong generalization to be made about a wide set of processes. As notedin section 1.3, CP is focused on the production of a vocalic prearticulation andnot on the wide range of conditioning factors that include the phonological(i.e., lexical or phrasal stress, adjacent distinctive features), themorphophonological and the sociolinguistic factors. Throughout this book,crucial themes reappear, such as the range of data, the concept of thenon-segmental prevowel, the mechanics of prevocalization via the persistence ofan oral V-place tier in consonants, and the distinction between prevocalizationand diphthongization. It is this last topic that is discussed in section 1.4.Here, Operstein provides evidence for CP from perception (quoting Reinhardt1970), phonology (e.g., Mascaró’s 1985 analysis of Catalan), and phonetics(e.g., portions of prevowels are shorter than offglides of diphthongs;differences in intra-elemental ‘binding’; optional loss of prevowels in fastspeech, etc.). Operstein notes that concrete phonetic analyses are lacking,other than Lehiste (1965) (although references to phonetic analyses are pepperedappropriately throughout the book). The second important piece of Chapter 1 isOperstein’s situating of her analysis against other theoretical explanations ofthe data. Section1.6 is an important section to understanding Operstein’sposition because it is here that she goes through past analyses, pointing outsome shortcomings, and laying out some basic data. She begins first withcomments on Andersen’s (1972) analysis of dialectal Polish. Then, Opersteinbuilds on Wetzel & Sluyter’s (1995) analysis of Maxakali, however, she arguesthat the analysis falls out from general properties and not just as a means fora specific language analysis. Finally, Operstein rejects Gussenhoven & Weijer’s(1990) diachronic analysis of CP in English as vowel diphthongization. Thereason for spending time to walk through differences between CP anddiphthongization is that the focus remains on CP emanating from the consonantand not the vowel. As such, CP and its related vowel lengthening and vowelepenthesis processes all strengthen the vowel and lenite the consonant. Notethat an explanation of Operstein’s analysis does not necessarily appear inChapter 1.

Where the reader finds Operstein’s actual theory is in Chapter 2(“Intrasegmental consonant structure”). This chapter has two main thrusts,arguing for bigestural consonants (i.e., comprising a consonantal and vocalicdistinctive feature node), and a realignment of the vocalic portion of theconsonant being articulated, consistent with sonority and other prosodicallydriven factors (e.g., prevocalization not appearing word initially in Estonian).After first laying out the theoretical framework (sect.2.2), in which it isargued that consonants are bigestural, and arguing for the specificity ofdistinctive features in the vocalic node of the consonant that motivatesprevocalization, Operstein then explains intrasegmental gestures (sect.2.3.1),followed by ensuing prevocalization of the consonant’s vocalic node with respectto modified (sect.2.3.2) and plain consonants (sect.2.3.3). The chapter relieson the overlapping tiers and degrees of closure specified in AP (Browman &Goldstein 1986 et seq.). The important piece of Operstein’s argument withrespect to the tiers is that there are two tongue body tiers, one for consonantfeatures and one for vowel features; specifications on these tiers can besimultaneous or sequentially ordered. Because they are autosegmental gesturaltiers, they have some degree of independence; thus, one can start a gesture,such as labial closure, either before, during, or after the tongue root assumesa pharyngeal position. A helpful example is provided in Figure 11 (p. 53), wherea fully articulated dark [l] is distinguished from a prevocalized dark [l] and avocalized variant. The tongue tip tier, specified for “alveolar/closed,” isalmost entirely aligned with the tongue body tier, specified for“pharyngeal/wide.” The prevocalized variant occurs because the tongue body orpharyngeal part of the consonantal gesture occurs prior to another oral closure;without closure, vocalization ensues on the portion of the consonant that doesnot overlap with closure. Consequently, for the vocalized variant, there is noclosure tier specification such that the entire pharyngeal specification assumesan open vocal tract. The next important distinction between Operstein’s analysisand others is that she argues for a persistent V-place node or set of features,arguing from an articulatory observation that labial and tongue tip consonantshave a wide range of tongue body positions arising from coarticulation. Thus,Operstein argues, the vocalic tier for tongue body must be present in allconsonants, not just in secondarily articulated ones.

Rounding out the first part on theory, Chapter 3 (“Related processes”) firstdiscusses consonant postvocalization (e.g., palatal offglides arising frompalatal consonants), and then addresses syllabic consonants (i.e., the ensuingvowel quality derives from the V-place features of the consonant). Opersteinargues that the persistent V-place tier, and not the C-place tier, results inpostvocalization and vocalization of syllabic consonants. The chapter ends withan overview of alternative approaches to the data. The two alternatives coveredare vowel intrusion and compensatory vowel lengthening. Operstein shows that thetiming of specified C-place and V-place tiers accounts for these phenomena as well.

The second part of the book on data is oriented toward the quality of prevowels.The front prevowels are covered in Chapter 4 (“Front prevowels”) and all otherprevowels in the subsequent chapter. The front prevowel is the most common kindof prevowel, triggered by laminal-coronal type consonants (i.e., alveolar,palatals and other palatalized consonants; dentals are excluded becauseapicality permits wide variance in tongue body position). The order ofpresentation is palatalized consonants (sect.4.2), palatals (sect.4.3) andalveolars (sect.4.4). The contrastively palatalized consonant section(sect.4.2.2) includes separate subsections on Finno-Ugric, Slavic, Greek,Romance, Gallo-Romance and Celtic. Likewise, the contextually palatalizedconsonant section (sect.4.2.3) includes Iranian, Archaic Latin and NorthernItalo-Romance, Tocharian, English and other languages. The role of diachronicanalysis of CP is often used to explain modern patterns, as in the case of theEnglish discussion. The last section of palatalized consonants (sect.4.2.4)suggests that CP had a hand in Germanic i-umlaut (although note that there areno references on the topic after 1997). The palatal section (sect.4.3) coverspalatals, palatoalveolars and alveopalatals in Catalan, French, Portuguese,English and other languages. The alveolar section is laid out slightly differentin that it covers sound classes, liquids (sect.4.4.1), /s/ (sect.4.4.2) and /n/(sect.4.4.3). A very brief portion on word-final weakening rounds out the chapter.

The second chapter in the data portion of the book, Chapter 5 (“Otherprevowels”), is divided into three main subsections focusing on the consonanttype that generates the prevowels: prevelarization and prelabialization(sect.5.2); postvelars (sect.5.3); and retroflexion (sect.5.4). Most of thediscussion of this chapter lies in the first section on prevelarization andprelabialization, with some discussion by sounds (e.g., dentals, sect.5.2.1;dark [l], sect.5.2.2; velar nasals, sect.5.2.3; labial and labializedconsonants, sect.5.2.6), as well as by language (e.g., the broad consonants ofGaelic, sect.5.2.4) and by process (e.g., back umlaut and breaking, sect.5.2.5).Note that these two data chapters are devoid of the mechanics of prevocalizationfrom Chapter 2; what is demonstrated here is Operstein’s command ofcross-linguistic patterns. Nevertheless, throughout, references are made to C-and V-gestures, as it is assumed that Chapter 2 has been understood prior toexamining the data chapters.

The brief Chapter 6 (“Conclusions and outlook”) consists of four pages of text.The two main conclusions are that plain nonlaryngeal consonants lenite byprevocalizing, and that prevocalization arising from secondary articulation inconsonants is an enhancement of the consonant, the redundancy of which oftenleads to a loss of the consonant’s secondary articulation. A number of futuredirections are listed on the last two pages.

EVALUATION

Overall, Natalie Operstein provides the field with a comprehensive compilationof phenomena that result in the appearance of vocalic subsegments prior to theconsonants that are the source of those prevowels. Out of her command of data,Operstein also provides a generalization, namely, that V-place features arepersistent in consonants, not just in those with secondary specification. Thetheoretical machinery affording her this claim is AP, with its autosegmentalgestural tiers. In AP, onsets of specified tiers (e.g., the beginning of agestural specification) can be temporally aligned with other specified tiers.The result is either that the V-place and C-place tongue body gestures cancompletely overlap with each other, or that the gestures overlap to some varyingdegree. Prevowels, then, are the consequence of a V-place gesture on a consonantstarting before the rest of coda C-place gestures.

Operstein is to be commended simply for gathering a wide range of data andgrouping them with an eye towards instances of prevowel formation, where thedata had not previously been classified as such. Given the way data ispresented, this book may appear to be a throwback phonological treatise withoutall the anacronism one might expect. First, sidestepping level ordering, etc.,Operstein tackles the subphonemic process that lends itself to an autosegmentalexplanation. The data is replete with underlying, feeding and bleedingenvironments, etc., all of which remind one of analyses in vogue from the earlydays of 1970s autosegmental phonology up through Archangli & Pulleyblank (1994),with phonology at the fore and phonetics playing a supporting, less phonetic''ground hugging'' role (Halle 1962/1972). This is not to say that there is nophonetic discussion or that other types of analyses cannot handle such data(e.g., alignment constraints); CP is a collection of observed patterns and anytheorist can take a shot at explaining the data. Operstein, to her credit,offers an analysis of a wide range of data so that the explanations given forboth specific instances and the general case of CP afford the reader someinsight into human language. Nevertheless, theoretical phonologists may havepreferred more representations or situating the data in theoretical constructs.

Although Operstein makes good use of the phonological approach in order tocapture generalizations and group together a broad set of data, those interestedin “lab phon” methods or the interaction of phonetics and phonology may beslightly frustrated with what appears to be an old-school approach to phonology(e.g., no spectrographic and/or waveform images). Throughout, Operstein does tryto relate the data to acoustic representations, albeit via citing relevantliterature. This reader, for one, needed more hands on reassurances and ended upaccessing the Estonian data from the JIPA website (Asu & Teres 2009) and alsoexamined x-ray microbeam data for English variation in dark [l] production inorder to examine phonetic evidence of CP. Operstein uses several occasions todistinguish between diphthongs and prevowels, leading one to wonder whether thisis the time for a new articulatory analysis of much of the CP data.

Phonological data is present throughout the book and the variety of data, comingfrom diachronic and synchronic analyses of languages and dialects, will beextremely useful to the field. In spite of this, the reader might get theimpression that Operstein was drawing from the ‘oldies-but-goodies’ list by thesources used. What is not meant here is that she failed to find relevant data;instead, it appears that she was working with non-contemporary data andanalyses. This may be most apparent to readers working on American Englishdialects. While having great respect for work by early-to-mid 20th centurydialectologists (e.g., Stanley, Sledd), this reader, for one, felt that workslike the Atlas of North American English (Labov, Ash & Boberg 2006), Thomas(2001) and other works with greater nuanced analyses on vowel qualities inSouthern American English would be referenced, but were not. One might arguethat they deal with vowels and not consonants, but since works like these arerife with discussion on diphthongs, and since diphthongs are often confused withprevowels, Operstein could have taken a look at them.

This leads to an additional question arising from Operstein’s work. In spite ofthe vast number of languages and dialects included, one wonders whetherOperstein sold herself short on a couple of issues. The first has to do withvowel diphthongization in American English, just to use one example. Is thediphthongization involved with many of the vowel shifts stimulated by thisprocess? What comes to mind are diphthongized vowels, specifically /æ/ beforenasals and voiced obstruents, and their prevelar raising in the Upper Midwestbefore the voiced velar plosive (e.g., BAG > BEG). It would have been reallyinteresting if Operstein had broached such cases in her section on breaking(sect.5.2.5) and/or in the long footnote on /æ/ in English (p. 145), where themost contemporary reference is Wells (1982). The second issue where Opersteinmay have shorted herself is that she does not bring laryngeal specificationsinto the analysis, which are important given the articulatory relationshipsbetween laryngeal and supralaryngeal gestures. Omitting this tier prevented herfrom making an even broader generalization, connecting CP to pre- andpost-consonantal aspiration. In fact, if one wants to make the broadestgeneralization, then prevocalization and preaspiration should be connected,presumably distinguished only by specification of vibrating vocal folds orcessation of vocal fold vibration, yet motivated by fortition of the vowel orlenition of the consonant.

Rosapelly’s appendix is an interesting feature of the book, but one that mighthave been more relevant to contemporary phonetics if it informed modernphonetics rather than being a historically interesting note. Rosapelly’sobservations were that consonants and vowels generally have three comparableparts such that the steady state of a vowel and fricative are comparable to theclosure of a plosive; thus, sounds are either vowels or vocaloids. It is usefulto place these comments in the context of x-ray studies premised on held steadystates that begin appearing just after Rosapelly’s writings (e.g., Russell 1928,Perkell 1969) [NB, this is the only place in the text where English translationsare not provided for the reader.]. Upon reflection, two questions come to mind.First, since the time of Rosapelly’s analysis, articulatory analyses now makeuse of fairly high sample rates (e.g., Westbury’s 1994 x-ray microbeam studieshave a sample rate of one sample every 6.866 ms). These higher sampled analysesdemonstrate that, while articulators slow down, there is constant movement inthe mouth, often reflecting high degrees of co-articulation. Why did Opersteinnot cast Rosapelly against more contemporary studies? Such a cast would havebeen generally more insightful than just noting an observation that vowels andconsonants share a steady state. Second, regarding this appendix, Opersteinwould have contributed more to her argument with a comment on the phonetics ofcoarticulatory effects not occurring at the juncture in which they occur (e.g.,CP occurs at the beginning of the consonant, not where the vowel or offglideoccurs).

Lastly, I would like to mention a few minor shortcomings. Chapter 2 could haveused more Browman & Goldstein-esque diagrams showing how the distinction worksbetween the subtypes of plain consonants. For example, representations of clear[l] of English onsets and German codas are missing from Figure 11 and the plain[p] is missing from Figure 12. Also, Operstein could pull from sonority clines(again, perhaps a phonetic picture) to better explain the lenition of theconsonant and strengthening of the vowel. Specifically, prevocalizationincreases the time and declension of energy, thereby using energy changes toenhance the syllable peak. Then there is the problem of AP tiers and, while thisis not her problem directly, she should explain how to constrain tiers. Perhapsshe might need a more traditional representation, with paths between tiers,similar to what was proposed in Archangeli & Pulleyblank (1994)? Finally, thereare just a few minor typos, though none cause confusion. The one notable erroris in the table of contents and in the text, where section 2.2 is mislabeled assection 1.2.

In sum, this is a worthy addition to the Current Issues in Linguistic Theorybook list.

REFERENCES

Archangeli, D. & D. Pulleyblank 1994. Grounded Phonology. Cambridge, MA: MIT Press.

Asu, E. & P. Teras. 2009. Estonian. Journal of the International PhoneticAssociation 39: 367­72.

Browman, C. & L. Goldstein. 1986. Towards an articulatory phonology. PhonologyYearbook 3: 219-52.

Gussenhoven, C. & J. Van De Weijer. 1990. On V-place spreading vs. featurespreading in English historical phonology. The Linguistic Review 7: 311-32.

Halle, M. 1962/1972. Phonology in generative grammar. In V. Makkai (ed.),Phonological theory: Evolution and current practice, 380-92. New York: Holt,Rinehardt and Winston.

Labov, W., S. Ash & C. Boberg. 2006. The Atlas of North American English.Berlin: Mouton.

Lehiste, I. 1965. Palatalization in Estonian: Some acoustic observations. In V.Koressaar & A. Rannit (eds.), Estonian Poetry and Language: Studies in honor ofAnts Oras, 136-62. Stockholm: Tryckeri AB Esto.

Mascaró, J. 1985. Compensatory diphthongization in Majorcan Catalan. In L.Wetzels & E. Sezer (eds.), Studies in Compensatory Lengthening, 133-46.Dordrecht: Foris.

Perkell, J. 1969. Physiology of speech production: Results and implications of aquantitative cineradiographic study. Cambridge, MA: MIT Press.

Reinhardt, K. 1970. Intrusive [i] before /S/ in Brazilian Portuguese. Word 26:101-6.

Thomas, E. 2001. An Acoustic Analysis of Vowel Variation in New World English.PADS 85. Durham, NC: Duke University Press.

Wells, J. 1982. Accents of English. Cambridge: Cambridge University Press.

Wetzel, L. & W. Sluyter. 1995. Formação de raiz, formação de glide e‘decrowding’ fonético em Maxacalí. In L. Wetzels (ed.), Estudos fonológicos daslínguas indígenas brasileiras, 103-49. Rio de Janeiro: UFRJ.

Westbury, J. 1994. X-ray microbeam speech production database user's handbook(version 1.0). Unpublished manuscript, Madison, WI.

ABOUT THE REVIEWER

Thomas Purnell (Associate Professor, English, at the University of Wisconsin-Madison) is interested in cross-language and cross-dialect variation in vowels and consonants from low-level phonetic cues to phonological features. He publishes specifically on sociophonetics and phonological theory.


Page Updated: 10-Oct-2012