Printer Friendly

Models of Inflection.

Ray Fabri, Albert Ortmann, and Teresa Parodi, editors: Models of Inflection. Linguistische Arbeiten 388. Tubingen: Niemeyer, 1998. viii + 289 pp.

What the editors of this volume have in common is that they all worked for a number of years together at the University of Dusseldorf, which has one of the most active linguistics departments in Germany, headed by Dieter Wunderlich. Four of the fourteen essays in this collection come from this background (R. Fabri, A. Ortmann, T. Parodi, and B. Gerlach), while the ten others are by authors working in Germany (I. Zimmermann, H. Elsen, K. Lindner, C. Lleo, W. U. Wurzel, M. Neef, D. Nubling) or in neighboring countries (G. Booij, P. Segeral and T. Scheer, D. Brown). The volume thus reflects to some extent the research activities of the Dusseldorf group and their connections to other scholars, but a reader who perhaps hopes that this book would develop and discuss the Dusseldorf-based approach of "minimalist morphology" (Wunderlich and Fabri 1995) will be disappointed, because this is hardly mentioned in the volume. In fact, the book has no clear focus at all, and the reason for this becomes clear from the preface: it is based primarily on a workshop held at the 1996 annual meeting of the Deutsche Gesellschaft far Sprachwissenschaft, and it shares its fate with many a conference volume. For the purpose of this review, I will group the papers into five categories: (i) the nature of inflection (Fabri, Booij, Ortmann, Nubling), (ii) person-number markers (Parodi, Gerlach, Zimmermann), (iii) marker allomorphy (Neef, Wurzel), (iv) stem allomorphy (Brown, Segeral and Scheer), (v) acquisition of inflection (Elsen, Lindner, Lleo). (Three of the papers are in German: Zimmermann, Wurzel, and Nubling.)

i. The nature of inflection

While Fabri's succinct introduction (pp. 1-10) offers a bird's-eye view of current theoretical issues and approaches, the papers by Booij, Nubling, and Ortmann do not really deal with "models" of inflection. The Booij and Ortmann papers (my favorites in this volume, together with Wurzel's) offer excellent overviews of their topic and thus would fit the genre of a handbook article. Ortmann's contribution is particularly noteworthy because his topic ("The role of [[+ or -] animate] in inflection," pp. 60-84) is quite unique, and his exhaustive survey is very useful. I was left wondering what the reason is for the ubiquitous importance of animacy in grammar. Ortmann points to psycholinguistic studies showing the relevance of animacy for categorization, but this does not explain the consistent directionality of the grammatical splits. Booij offers yet another discussion of the inflection-derivation distinction (pp. 11-27). He notes that some authors have given up the clear-cut distinction in favor of a continuum, and while his discussion is very balanced, in the end he comes down on the side of the discretists, arguing for "a specific inflectional submodule" (though he argues against split morphology a la Anderson 1992). However, some of the arguments are not on very firm ground. For instance, his criterion of obligatoriness can never be applied rigorously, because it would always be claimed that one of the inflected forms is zero-marked (e.g. one could theoretically say that English -ness is an inflectional marker of abstractness, and that each adjective is obligatorily inflected as either [+abstract] (-ness) or [-abstract] (-[Phi])). Another not very helpful criterion is whether a paradigm is formed. In the discussion of Hebrew binyanim, Booij asserts that they "form a kind of paradigm" (p. 17), but I know no clear definition of the concept of paradigm that is independent of the concept of inflection. Nubling, too, discusses the nature of inflection (favoring a continuum model), but the real focus of her article is on possible diachronic changes between derivation, inflection, and syntax (pp. 266-289). In particular, she asks under what circumstances the German preposition-article complexes (e.g. im `in the', yom `from the', zur `to the') could be regarded as inflected prepositions. She notes briefly that these complexes constitute a form-meaning mismatch, because the article belongs with the NP, not with the preposition. It is quite possible that this also prevents a full grammaticalization of these constructions. But if only adjacency played a role, we would probably have more tense inflection on nouns, person-number inflection on adverbs, etc. A completely different way of looking at the preposition-article complexes would be not as "inflected prepositions," but as an expansion of the case inventory of definite articles: im would be an inessive case, yom an ablative case, and zur an allative case of the definite article. I have never seen this perspective adopted, although it is clearly less strange than the notion of a preposition that is inflected for definiteness, case, and gender of its complement.

ii. Person-number markers

Gerlach discusses active and mediopassive person-number markers in Modern Greek (pp. 103-118), where we find a potential counterexample to the generalization that voice markers are closer to the verb stem than tense and person-number markers. The partial paradigm in (1) suggests that in the mediopassive, the person markers (m/s/t) precede the cumulated tense/voice markers (-un/-an).

(1) lin- `loosen'
Active Mediopassive
present past present past
lin-o e-lin-a lin-ome lin-omun
lin-is e-lin-es lin-ese lin-osun
lin-i e-lin-e lin-ete lin-otan


Gerlach argues that this is not the right analysis, that the elements -omun/-osun/-otan are unanalyzable elements cumulating voice, tense, and person/number. Since there is so little regularity here, this seems plausible, but strangely Gerlach points to the great dialectal variability of the mediopassive forms as a justification for nonanalyzability. And one wonders what we have learned from this. Explaining away a single counterexample would be important if one assumed that the hierarchy of verbal morphological categories is innately fixed, so that a single counterexample would suffice to refute the theory. However, Bybee's (1985) original generalization was not that the ordering is universally identical, but that there are clear cross-linguistic tendencies that need to be explained. Exceptions to the generalization are not uncommon, as Bybee herself notes (1985: 34-35). So explaining away all the exceptions does not seem to be a viable research program (and Gerlach herself admits that Ancient Greek is a true exception). Moreover, Greek is exceptional anyway: (medio-)passive markers are in most cases affixes attaching directly to the stem, and affixes cumulating voice and person/number are very rare (Haspelmath 1990:31).

Zimmermann and Parodi are primarily concerned with the syntax of pronominal clitics. Zimmermann deals with the little-known NP-internal clitic-doubling construction in Bulgarian (tazi mu kniga na Ivo [this to him book of Ivo] `this book of Ivo's'), focusing on the technicalities of a minimalist description (pp. 119-133). Parodi discusses clitics, clitic clusters, and clitic doubling in two wtrieties of Spanish (pp. 85-102). Her main point is that clitic doubling is a kind of agreement, which seems reasonable, though I am very skeptical of her idea that the restriction to definite direct objects derives from a "matching effect" (because clitics are definite, too). This misses the generalization that definiteness favors both overt object case marking and object agreement, for well-understood functional reasons (Comrie 1089). The connection of all this to inflectional morphology is not obvious. Parodi seems to think if a pronominal element is an agreement marker, that automatically makes it an affix (or "quasi-affix," p. 95), but this is not evident to me.

iii. Marker allomorphy

Wurzel and Neef both deal with German nominal inflection, which is notorious for its intriguing distribution of case and especially number allomorphs. While Neef limits himself to plural marking, Wurzel includes case as well and uses German noun inflection to argue for a general model of inflectional paradigms (pp. 225-243). This consists of three levels: structured paradigms of individual words, paradigm structure conditions (which express general facts about these paradigms that need not be specified in individual lexical entries), and canonical paradigm-structure conditions (which express generalizations about paradigm structure conditions). This is a beautiful paper, but since Wurzel exemplifies his theory with only a single language, one wonders whether this system can be applied insightfully to any other language. Neef is less ambitious in his claims about German plurals (pp. 244-265). His main point is that the generalizations are best captured by declarative "design conditions," which according to him have a strong prosodic component. In contrast to constraints in optimality theory, the design conditions in his approach ("word design") are unviolable, language-specific constraints.

iv. Stem allomorphy

Segeral and Scheer propose an unusual account of German verbal ablaut patterns that relates them to Afro-Asiatic ablaut patterns and surprisingly claims that the ablaut vowel series i-a-u is part of universal grammar (pp. 28-59). The paper is fun to read and deserves a prize for imagination, but that exhausts the positive things I can say about it. Brown's paper on Russian verbal stem allomorphy (pp. 196-224) has the opposite problem: it applies the theory of network morphology (Corbett and Fraser 1993) (plus a particular computational formalism) in a very serious and conscientious way, but it fails to get a more general point across: what are the advantages of this particular approach?

v. Acquisition of inflection

The papers by Elsen (pp. 134-151) and Lindner (pp. 152-174) deal with the same topic (German past participles) and are based on the same data set (a diary study of Elsen's daughter). Elsen focuses on the hypothesis that overgeneralizations occur when a critical mass of regular verbs has been acquired, and Lindner is primarily concerned with the blocking hypothesis. Both authors conclude that the evidence is more compatible with a connectionist-style single-mechanism view than with a Pinkerian dual-processing model. Lleo examines "proto-articles" in the acquisition of Spanish (e.g. the initial vowel in [a'wow??h] (1;10,17) for el globo `the balloon') and finds that these occur much earlier than has been reported so far (pp. 175-195). This is certaintly interesting, but it is not clear to me that the data warrant the conclusion that "functional categories" are available to the child at this stage. Lleo does not seem to consider the possibility that initially the child's structures are quite different from the adult's.

A general problem that I often see with analyses of inflectional morphology (and the papers in this volume are no exception) is that linguists attempt to extract generalizations that may not be the speakers' generalizations. In inflection, it is more difficult than in syntax to decide whether a higher-level generalization is also captured by speakers.

Consider, as an example, Gerlach's generalization that Modern Greek active verb forms have the following specifications for the vowel that immediately follows the stem (p. 110):
(2) [+low] [+ 1st person or +plural]
 [+round] [+ 1st person]/[-past]


This statement works for the active forms in (1) and for the corresponding plural forms. It is more general than a mere list of the vowels (lsg present: -o, lsg past: -a, lpl past: -a, etc.), but there is no way to tell whether the generalization is a significant generalization. The fundamental problem is that the domain to which it applies is a small, closed class, and there is no reason to assume that speakers could not learn such a simple, boring list.

The same problem arises with generalizations across inflectional microclasses. Consider, as a second example, Neef's "Design condition of ban of final obstruents" (p. 251), which applies to German plural forms with a final schwa syllable. From this condition it follows that a sonorant-final word such as Segel `sail(s)' can be a plural form, but an obstruent-final word such as Abend `evening' cannot be a plural form (cf. Abende `evenings'). Again, the class of word types to which this generalization applies is very small: just the nouns in -en, -el and -er. Speakers could easily acquire three distinct rules for each of these rather than generalizing over them. Again, the claim that the generalization is correct is not falsifiable.

In general, we can be sure that generalizations are significant only if they apply to an open (or at least very large) domain, or if there is other evidence that a generalization is productive, for example from language change. Most of the authors here do not consider it necessary to provide such evidence, and in many cases it probably does not exist. A notable exception is Wurzel's paper, where evidence for productivity from recent changes in inflectional behavior is cited repeatedly. This is also the reason I found Wurzel's paper more convincing than most of the others.

One may ask whether it matters so much whether linguists sometimes overshoot the mark in their drive to express observed patterns in their descriptions. The answer to this question is straightforward: it does not matter if we just want to describe languages and have fun with elegant descriptions. But if we see our theoretical work as part of a larger cognitive enterprise, as all the authors in this volume seem to do, then it does matter. I see no way around the conclusion that when dealing with relatively small, unproductive classes, the grammarian's tools are insufficient to distinguish significant from spurious generalizations. This is not a pleasant prospect for the conventional grammatical theorist, but we have to face it.

References

Anderson, Stephen (1992). A-morphous Morphology. Cambridge: Cambridge University Press.

Bybee, Joan L. (1985). Morphology: A Study of the Relation between Meaning and Form. Amsterdam: Benjamins.

Comrie, Bernard (1989). Language Universals and Linguistic Typology. Oxford: Blackwell.

Corbett, Greville G.; and Fraser, Norman M. (1993). Network morphology: a DATR account of Russian nominal inflection. Journal of Linguistics 29, 113-142.

Haspelmath, Martin (1990). The grammaticization of passive morphology. Studies in Language 14(1), 25-71.

Wunderlich, Dieter; and Fabri, Ray (1995). Minimalist morphology: an approach to inflection. Zeitschrift fur Sprachwissenschaft 14(2), 236-294.

MARTIN HASPELMATH

Max-Planck-Institut fur evolutionare Anthropologie, Leipzig
COPYRIGHT 2000 Walter de Gruyter GmbH & Co. KG
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Review
Author:HASPELMATH, MARTIN
Publication:Linguistics: an interdisciplinary journal of the language sciences
Article Type:Book Review
Date:Jul 1, 2000
Words:2277
Previous Article:Function, Selection, and Innateness: The Emergence of Language Universals.
Next Article:Creoles, pidgins, and sundry languages: essays in honor of Pieter Seuren(1).
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters