Siptar, Peter and Miklos Torkenczy (2000). The phonology of Hungarian. Oxford University Press, hardback ISBN: 0-19-823841-X, $99.00, xiv+319pp (The Phonology of the World's Languages).
Reviewed by Stefan Frisch, Department of Communication Sciences and Disorders, University of South Florida
Synopsis:
S&T provide a comprehensive introduction to the segmental phonology of Hungarian using a rule-based formalism. Hungarian (native name _magyar_) is a Finno-Ugric (Uralic) language that is typologically unlike the other members of Uralic family. The book is divided into three parts. Part I (Background) provides an introduction to the contents of the book and an overview of Hungarian grammar. Part II (Systems) discusses the vowel inventory, the consonant inventory, and phonotactic constraints. Part 3 (Processes) examines morphophonological processes, including vowel harmony, the most well-known feature of Hungarian phonology, and post-lexical (surface) processes.
Chapter 1 (Introduction, 12 pages) gives a brief overview of previous work on Hungarian phonology and the authors' theoretical assumptions. The authors choose a rule-based account of Hungarian phonology, despite the recent shift among formal phonologists toward constraint-based models. Their choice of framework was determined by a desire to make their work accessible to the largest number of readers and a desire to avoid the potential pitfalls of committing to a new theoretical framework when its basic assumptions are not yet agreed upon. The morphophonological analysis in S&T uses a blend of the frameworks of Lexical Phonology and Government Phonology. The authors divide the suffixes in Hungarian into two classes: synthetic and analytic (equivalently, those that introduce + and # boundaries, or Level 1 and Level 2 suffixes). Phonological processes are divided into two corresponding blocks, with Block 1 applying within analytic domains and Block 2 applying to the whole word. For representation, S&T uses fairly standard Feature Geometry, underspecification, and onset-rime syllable structure (with a syllable appendix available for Block 2 rules).
In addition to their own (native speaker) judgments the authors make occasional use of a dictionary corpus of Hungarian lexical items. They also provide a great number of references to previous work on Hungarian phonology, which in addition to being sprinkled throughout the text are conveniently grouped into a section in the introduction. Among these references, those interested in this book as an introduction to Hungarian phonology might also wish to examine Hall (1944); Kenesei, Vago, and Fenyvesi (1998); or Vago (1980).
Chapter 2 (Preliminaries, 35 pages) provides an overview of Hungarian and Hungarian grammar. S&T analyze Educated Colloquial Hungarian, the standard language for educated Hungarians, modulo some influence from regional dialects. Hungarian is an agglutinating language with syllable-timed prosody and a rich vowel system. Hungarian also has a large number of consonant contrasts, including a full series of palatals. Brief treatments of intonation, morphosyntax, and word order are also given, and the level of detail is greater than what is typically found in a typological overview (e.g. the chapter by Daniel Abondolo in Comrie, 1990).
Chapter 3 (The vowel system, 24 pages) introduces the surface phonetic vowel system, provides a featural representation of the vowels, and sketches out alternation patterns in vowel length and quality. Hungarian vowels have a front/back contrast, with two levels of height for front vowels and three levels for back vowels. There is no central series, but there is a series of rounded front vowels, providing seven quality contrasts for short vowels. Hungarian also has a length contrast for all vowels, though there are some quality differences between phonologically contrastive short/long pairs. For example, the long /a/ corresponds to the short /open-o/. The phonological nature of the length contrast for vowels of different phonetic quality is demonstrated in three length alternations. First, morpheme-final short low vowels are lengthened (with a resulting change in phonetic quality) before a suffix. Second, in some stems, final long vowels are shortened before some inflectional suffixes. Third, in some stems, internal long vowels are shortened before some derivational suffixes. These alternations are sufficient to demonstrate that the vowels involves are treated as the same phonological category despite phonetic differences in degree of height (and backness).
This chapter also provides a description of Hungarian vowel harmony. Harmony is controlled by the stem of the complex word (Block 2 process), enforcing agreement in backness between all suffix vowels and the stem final vowel. There are, of course, complications. The front unrounded vowels appear to be neutral to backness harmony, so the process is restricted to rounded vowels as triggers and targets. There are a few suffixes with rounded vowels that are unaffected, and several classes of stems with differing behavior. While the stem final vowel is the primary influence in cases of backness harmony, earlier vowels in the stem control harmony if the stem final vowel is unrounded. If the stem consists only of (front) unrounded vowels, front-vowel suffixes are usually selected, suggesting the neutral vowels are not quite neutral, just of lesser influence. In addition to backness harmony, Hungarian also has rounding harmony that is always controlled by the stem final vowel, but is secondary to backness harmony. This leads to a three-way alternation in suffixes when the stem has a final front unrounded vowel but a front rounded vowel earlier in the stem. In this case suffix vowels are front unrounded vowels. If the stem final vowel is unrounded but the preceding vowel is back (rounded) a back vowel surfaces in the suffix.
Chapter 4 (The consonant system, 19 pages) lays out the consonant inventory. Hungarian has labial, dental, palatal, and velar consonants, with stops and fricatives at all places of articulation. Hungarian has dental and palatal affricates, and voicing contrasts for all obstruents except the velar fricative /x/. This fricative is phonetically realized as [h] in prevocalic position. Hungarian also has nasals at all places, and /l, r, j/. All obstruents are involved in a pattern of voicing assimilation within a cluster, that is regressive (right-to-left) and applies across all prosodic boundaries except for a pause. For example, /st#b/ -> [zdb] in a compound. Exceptions include /x/, which does not undergo voicing but does trigger devoicing, and [v], which undergoes devoicing but does not trigger voicing. Sonorant consonants are inert in voice assimilation. Nasal consonants assimilate to the place of a following obstruent in clusters.
Chapter 5 (Phonotactics, 60 pages) gives a thorough treatment of phonotactic constraints in the Hungarian word. Tables are given that provide a complete listing of word initial and word final clusters. Information on possible word final clusters in monomorphemic contexts as well as with synthetic and analytic suffixes is given. Hungarian is rich in consonant clusters, including for example initial [ft, sv, mn] and final [kt], [ps], [vd]. S&T account for the large number of possible clusters in three ways. The most unusual cases involve a few analytic suffixes, and an appendix position is invoked. For the remainder, S&T propose a Hungarian sonority hierarchy that puts stops, fricatives, and affricates at roughly the same level, with nasals higher and liquids higher still. As with other languages, clusters are licensed in cases where there is a large sonority difference. In addition, clusters with a smaller sonority difference are also licensed if both members of the cluster are coronal.
S&T also present cases of trans-syllabic phonotactic constraints. Hungarian has several regularities in vowel hiatus, including cases where a suffix initial vowel deletes or a glide is inserted. Some cases of hiatus do surface, and there are also distributional regularities on possible vowel sequences. There are also distributional constraints on trans-syllabic consonant sequences. For the most part, these are accounted for by the constraints given for possible initial and final syllables. However, there is one odd gap in that combinations where the second consonant is labial are systematically absent.
Chapter 6 (Processes involving vowels, 19 pages) provides a detailed formal analysis of the patterns of vowel harmony and vowel lengthening/shortening that was presented in . S&T treats backness and rounding harmony simultaneously, giving a nice coherence to their analysis. Though this is a rule-based account, it has a similar flavor to an Optimality Theoretic treatment. There are several feature-spreading rules that can all potentially apply, and in cases where there is no direct conflict, they do. In cases of conflict there is priority between rules. Finally, there is a set of impossible feature combinations to constrain overgeneration (that are, in part, phonetically motivated). A formal analysis of lengthening and shortening is also given. The discussion then turns to exceptional cases in vowel lengthening/shortening that related to exceptional cases in vowel harmony. However, no concrete proposals are put forth and the issue of these exceptions is left primarily as an open problem.
Chapter 7 (Processes involving consonants, 38 pages) presents several cases of assimilation in consonants, including palatalization of dentals that precede palatals, affrication of /t/ and /j/, and place assimilation for adjacent strident consonants. As in the previous section, an attempt is made to unify these processes under a single rule of place assimilation. The authors are only partially successful, however. This chapter also presents a straightforward formal analysis of voicing assimilation and nasal place assimilation.
Chapter 8 (Processes conditioned by syllable structure, 84 pages) is primarily concerned with cases of vowel/zero alternation. Such alternations can be found with stem final vowels, suffix initial vowels, or stem-internally in consonant clusters. In the formal analysis the authors present a representational distinction between full vowels and defective vowels. Defective vowels can be filled and surface or be deleted. The quality of the defective vowel is partially phonologically conditioned and there are a large number of intricacies that are not amenable to a brief summary such as this. This chapter also contains a brief treatment of total assimilation of suffix initial [v] to the stem final consonant in some suffixes and the [h]-[x] alternation discussed above.
Chapter 9 (Surface processes, 19 pages) describes cases of variation in phonetic implementation that are primarily related to speech rate, register, or inter-speaker variability. The authors discuss surface variability in vowel length, cases of liquid deletion with compensatory vowel lengthening, fortition in cases of vowel hiatus, and degemination/cluster simplification
Critical Evaluation:
Overall, this is a lovely combination of language description and analysis that is extremely well written. I think the authors fully achieved their goal of producing a work that is accessible to a wide audience. In addition, the later chapters in the book contain detailed formal analysis that make quite clear the strengths (and weaknesses) of the authors' proposals. I think the organization, which begins with a more descriptive treatment and proceeds to deeper levels of formal detail make the book extremely accessible, as people of different interest level/background can get as much detail as they wish. It is also a book of noticeably excellent quality. There were no typographic errors and the figures and tables are quite well done. My one complaint would be the very spare use of phonetic transcription for the vowel system. I found this to be particularly bothersome in tables with feature descriptions or summaries of phonological patterns where it would have been easy to include both orthographic and transcription labels. Most example forms are also given orthographically, putting the burden on the reader to determine the relevant phonological form. More consistent use of transcription was employed in discussing the consonants.
This book also serves its purpose as a member of the phonology of the world's languages series. The overview of the Hungarian phonological system is comprehensive, covering several areas of the phonology in detail. A great deal of useful data and analysis is provided, including some cases where lexical counts are given based on their dictionary corpus. The presentation is complete and detailed enough for me to use the section on phonotactics in my own research. Given my own interest in lexical corpus research, I would like to have seen more data given along these lines (e.g. as in another book in the Oxford series, Hammond 1999, though S&T is less narrowly focused than Hammond 1999). Regardless, this book is a rich resource for researchers; especially those interested in applying the newer frameworks to complicated data. In addition, the patterns presented appear to contain a degree of gradiency that are becoming a growing topic of research for both formal and functional phonologists (see for example, Ringen and Heinamaki 1999 and the references therein for research along these lines with relevant analysis and discussion). Therefore I think this book would also be of interest to functional phonologists, as the authors provide several cases where there are both broad patterns and problematic exceptions that are difficult to handle completely in a formal account.
References: Comrie, Bernard (1980). The world's major languages. Oxford University Press, Oxford. Hall, Robert A. (1944). Hungarian grammar. Linguistic Society of America, Baltimore. Hammond, Michael (1999). The phonology of English. Oxford University Press, Oxford. Kenesei, Istvan, Robert M. Vago, and Anna Fenyvesi (1998). Hungarian. Descriptive Grammar Series. Routledge, London. Ringen, Catherine and Orvokki Heinamaki (1999). Variation in Finnish vowel harmony: An OT account. Natural Language and Linguistic Theory 17, 303-337. Vago, Robert M. (1980). The sound pattern of Hungarian. Georgetown University Press, Washington.
For more information on the reviewer, see http://www.cas.usf.edu/~frisch
|