A typology of defectiveness
Albright, Adam. 2003. A quantitative study of Spanish paradigm gaps. In: G. Garding and M. Tsujimura (eds) West Coast Conference on Formal Linguistics 22 Proceedings. Somerville: Cascadilla Press. 1-14.
Defectiveness in Spanish verbs is attributed to speaker uncertainty. The gaps correspond to an expected stem-vowel alternation, but the phonological characteristics of the verbs (stem vowel, stem-final consonant) are relatively unusual. Albright proposes that the number of exemplars to work from is simply too small for speakers' morphological decisions (apply vowel alternation or not) to be applied with sufficient confidence.
Baronian, Luc. 2005. North of phonology. PhD thesis, Stanford University.
This thesis is an exposition of the Theory of Connected Word Constructions, and chapter 5 is devoted to defectiveness. In this model, stem alternations are divided into different Lexical Building Blocks, which characterize rules along with sufficient phonolological material to identify the lexemes that the rules apply to. The examples involve defective verbs from English (stride ~ ?stridden), French, Spanish and Russian, with defectiveness construed as resulting from the failure to ascribe to a lexeme any appropriate Lexical Building Block for the values in question, either through lack of information or because of conflicting phonological details.
Boyé, Gilles. 2005. Problèmes de morphophonologie verbale en français, en espagnol et en italien. PhD thesis, Université Paris VII.
Treats defective verbs, among other topics, with particular attention to French. Defectiveness is construed as a type of stem suppletion, whereby the stem is null.
Bresnan, Joan. 2000. Explaining morphosyntactic competition. In: Mark Baltin and Chris Collins (eds.) Handbook of contemporary syntactic theory. Oxford: Blackwell.
The non-existence of *amn't in standard English is attributed to a rule which ascribes a null pronunciation to the feature combination for the verb be.
Daland, Robert, Andrea D. Sims and Janet Pierrehumbert. 2007. Much ado about nothing: A social network model of Russian paradigmatic gaps. In: Annie Zaenen and Antal van den Bosch (eds)Proceedings of the 45th Annual Meeting of the Association for Computational Linguistics in Prague, Czech Republic, June 24th-29th, 2007. Prague: Association for Computational Linguistics, 936-943.
The authors construct a model of how gaps may be learned and maintained across generations on the basis of lower-than-expected frequencies that the learner encounters. Defective Russian verbs are the basis for the model.
Fanselow, Gisbert and Caroline Féry. 2002. Ineffability in grammar. In: Gisbert Fanselow and Caroline Féry (eds) Resolving conflicts in grammars: Optimality Theory in syntax, morphology, and phonology. Hamburg: Buske. 265–307.
Addresses the inherent problem that gaps pose for Optimality Theory, based as it is on violable constraints. (Note that the discussion of inflectional defectiveness is part of a wider typology that includes syntactic and derivational gaps.) After reviewing examples discussed elsewhere in the literature, it is concluded that inflectional gaps can be accounted for by Control (Orgun and Sprouse 1999).
Frampton, John. 2002. The amn’t gap, ineffability, and anomalous aren’t: against morphosyntactic competition. In: Mary Andronis, Christopher Ball, Heidi Elston and Sylvain Neuvel (eds.) CLS 37: The Panels. Papers from the 37th Meeting of the Chicago Linguistic Society, Vol. 2. Chicago: Chicago Linguistic Society. 399-412.
A reply to Bresnan (2001) in which the lack of English amn't is accounted for by stipulating in the lexical entry for -n't that it cannot be affixed to am.
Halle, Morris. 1973. Prolegomena to a theory of word formation. Linguistic Inquiry 4. 3-16.
Proposes that the products of word formation rules (i.e. derived and inflected forms), prior to entering the lexicon, are subject to a filter which accounts for exceptions, including defectiveness (Russian verbs are used as an example).
Hansson, Gunnar Ólafur. 1999. ‘When in doubt…’: intraparadigmatic dependencies and gaps in Icelandic. In: P. Tamanji, M. Hirotani and N. Hall (eds) Proceedings of NELS 29. Amherst, MA: GLSA Publications. 105-119.
The suffix -T ~ -Th is found with (i) the singular imperative of all verbs, and (ii) the past tense of weak verbs. The choice between the two allomorphs (differing w.r.t. the feature 'spread glottis') is in general phonologically predictable, but:
For some verbs, the allomorph is lexically specified, contrary to the default rules. The majority of weak verbs in -ll or -nn exceptionally take -Th in place of expected -T in the imperative singular and the past. There are a handful of strong verbs in -nn/-ll which take neither allomorph in the imperative singular; rather, they have a gap. Working within Optimality Theory, Hansson proposes that, by default, the 'spread glottis' feature of the imperative allomorph must match that found with some other coronal obstruent in the paradigm. In the case of weak verbs, they match the past tense allomorph. But strong verbs have a different past tense formation, and so are unaffected. To account for the imperative singualar gap, he proposes constraint in Control (Orgun & Sprouse 1999) called 'DoubleCheck imper' that requires the imperative form to be mutually supported both by grammatical computation and by analogy with similar sounding words. By the former, strong verbs in -ll/-nn should have regular affix, by the latter they should have exceptional affix, and the Control constraint is violated.
Hetzron, Robert. 1975. Where the grammar fails. Language 51. 859-872.
Discusses typology of defectiveness and allied phenomena, distinguishing arbitrary lexical gaps, which are simply stipulated, from other types, which result from more systematic grammatical principles, such as (i) surface constraints, (ii) conflicts between rules, (iii) failure of rules to cover all possible cases, and (iv) indeterminacy of the rule. Examples discussed include:
The absolute superlative in English (e.g. ‘a most interesting book’) is not permitted for adjectives with suppletive comparatives and graded superlatives, e.g. *a most good solution (nor *a best solution). Russian nouns that take the prepositions na 'to' and s 'from' can’t be used with verbs that govern the alternative preposition pair (v 'to', iz 'from'. Russian 1st person singular verbs. Contraction in Arabic: phonological well-formedness constraints conflict with some derived forms for some types of roots. in some modern dialects there still seem to be gaps here, while in others it has been resolved by phonological repair. Hungarian -CCik verbs. Hetzron argues against Halle’s (1973) extreme position that all inflected forms are stored. Neither this nor the opposite extreme appear to be true. Instead, you probably have 2 ordered rules: (i) lexical exceptions, (ii) regular productive rules.
Hudson, Richard. 2000. *I amn’t. Language 76. 297-323.
The lack of a form amn't in English is attributed to conflicts in a default inheritence network. The problems results from two rules that define a whole word form. On the one hand, a negative auxiliary is a whole word form made up of a morphological base + n't. On the other hand, 1st person singular am is itself defined as a whole word form. The conflicting definitions of what the whole word form of the negative 1st person singular present of be lead to system crash.
Iverson, Gregory. 1981. Rules, constraints and paradigmatic lacunae. Glossa 15/1. 136-144.
Discusses Swedish defective adjectives, using data from Eliasson (1975), asserting that defectiveness occurs when some (lexical) item is identified as an exception to an automatic rule (where 'automatic' = there are no surface violations of the rule). Iverson concludes that 'allomorphic uniformity is the excplicit goal of paradigm gap creation in general, while the mechanism which effects it is the specification of lexical exceptionality on automatic, surface-constrained phonological rules.' (Though he notes various examples not predicted by this axiom, suggesting that 'other, non-phonological forces are at work in these isolated instances.')
Joseph, Brian D. 1997. On the linguistics of marginality: the centrality of the periphery. In: G. Anderson et al. (eds) Papers from the 33rd Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistic Society. 197-213.
Among other topics, discusses defectiveness as the result of an incomplete diachronic development (the Ancient Greek verb 'I say', whose paradigm seems to have arisen piecemeal, and was never fully fleshed out).
Karlsson, Fred. 2000. Defectivity. In: Geert Booij, Christian Lehmann and Joachim Mugdan (eds) Morphologie (Morphology) 1: Ein Internationales Handbuch Zur Flexion und Wortbildung (An International Handbook on Inflection and Word-Formation). Berlin: Mouton de Gruyter. 647-654.
Deals primarily with examples of defectiveness (or infrequency) that have a direct semantic motivation or at least a semantic correlation (such as pluralia/singularia tantum).
Kiparsky, P. 1974. Remarks on analogical change. In John M. Anderson & Charles Jones (eds) Historical Linguistics II. 257-75. New York: Elsevier.
Defectiveness is discussed as an aspect of analogical change. Change can be represented by the formula A>B, which presupposes two distinct historical processes: loss of the old form and introduction of the new one. Sometimes introduction of B is delayed or does not occur at all, which he labels a 'copout', namely 'a gap in the paradigm at points of choice in the rule system'. Examples are are drawn from the Russian 1st person singular, Spanish abolir, Finnish *erata, and from English (back formed verbs from compounds of the type bloodsuck or sightsee, whose degree of adoption of normal verbal morphosyntax varies).
Morin, Yves-Charles. 1995. De l’acquisition de la morphologie: le cas des verbes morphologiquement défectifs du français. In: H. Bat-Zeev Shyldkrot and L. Kupferman (eds) Tendances récentes en linguistique française et générale: volume dedié à David Gaatone. Amsterdam: John Benjamins. 295-310.
Discusses the problems that defectiveness poses for grammars in which some default rule will always be available. French defective verbs are evidence for a distinction between productive rules and analogical processes. Not all inflection is rule-based; in particular, some stem alternations must simply be learnt. If the paradigm is construed as made up of different morphological blocks, defectiveness can be acquired through lack of exposure to the appropriate block. The occasional filling in of such paradigmatic cells can be ascribed to irregular analogical processes.
Orgun, Cemil Orhan and Ronald Sprouse. 1999. From MParse to Control: deriving ungrammaticality. Phonology 16. 191-224.
Argues that the notion of the null parse in Optimality Theory (i.e. that no output at all may be the winning candidate) is inadequate to account for some instances of defectiveness. The behaviour of monosyllabic V-final stems in Turkish is used to argue for a separate component called 'Control', which evaluates the well-formedness of the output of the primary set of constraints. If this is rejected, a gap results. (Tagalog and Tiene are also discussed; while these involve the absence of a possibly expected form, they do not, strictly speaking, involve defectiveness.)
Pertsova, Katya. 2005. How lexical conservatism can lead to paradigm gaps. In: J. Heinz, A. Martin and K. Pertsova (eds) UCLA Working Papers in Linguistics 11 (Papers in Phonology 6). Los Angeles: UCLA. 13-38.
Argues that defectiveness in the Russian genitive plural can be accounted for in Optimality Theoretic terms by an Identity constraint which requires the position of stress to match that found somewhere else in the paradigm (the words in question undergo a phonologically automatic stress shift in the genitive plural).
Raffelsiefen, Renate. 2004. Absolute ill-formedness and other morphophonological effects. Phonology 21. 91–142.
Defends the notion of MParse against Control (see Orgun and Sprouse 1999); the examples are drawn largely from (English) derivational morphology.
Rice, Curt. 2002. When nothing is good enough: dialectal variation in Norwegian imperatives. Nordlyd 31/2. 372-384.
In some Norwegian dialects, imperatives cannot be formed from verbs whose stems terminate in consonant clusters of rising sonority. These are given an Optimality Theory analysis.
Rice, Curt. 2007. Gaps and repairs at the phonology–morphology interface. Journal of Linguistics 43/1. 197-221.
Deals with phonologically conditioned defectiveness. Using data from the Norwegian imperative (first presented in Rice 2002). Advances a model within Optimality Theory which avoids the use of 'null parse' or 'null output' (in effect, morphological zeroes) by evaluating whole paradigms rather than individual forms. Gaps are the result of (phonological) markedness constraints dominating faithfulness constraints.
Sims, Andrea. 2006. Minding the gaps: inflectional defectiveness in a paradigmatic theory. PhD thesis, The Ohio State University.
Analyzes defectiveness the Greek genitive plural and the Russian 1st person singular, advancing the idea that defectiveness can be lexically specified (though in such instances there may previoulsy have been a more transparent motivation).
Strunk, Klaus. 1977. Überlegungen zu Defektivität und Suppletion im Griechischen und Indogermanischen. Glotta 55. 2-34.
Primarily concerned with suppletion, but it does take pains to point out that the term 'defective' as used by Indo-Europeanists is often confounded with 'suppletive', and that the notions should be distinuished.
Törkenczy, Miklós. 2002. Absolute phonological ungrammaticality in output-biased Phonology. In: István Kenesei (ed.) Approaches to Hungarian. Volume 8. Papers from the Budapest Conference. Budapest: Akadémiai Kiadó. 311-324.
Data from Hungarian are argued to be problematic for the Optimality Theoretic notion of Control (Orgun and Sprouse 1999), suggesting a non-output biased declarative framework is needed instead.
Wolf, Matthew and John J. McCarthy. 2009. Less than zero: Correspondence and the null output. In: Curt Rice (ed.) Modeling ungrammaticality in Optimality Theory. London: Equinox.
A reassertion of the role of the 'null output' (formerly 'null parse') in Optimality Theory, first advanced by Prince and Smolensky (1993). The null parse -- i.e. a gap -- satisfies every constraint but MParse, which requires that morphological structure in the input has some correspondence in the output. Gaps thus are related to the ranking of MParse with respect to other constraints. This is argued to be preferable to Control (Orgun and Sprouse 1999) on conceptual grounds.
Prof Greville G. Corbett
Dr Dunstan Brown
Dr Matthew Baerman
Period of award:
September 2006 - February 2009
Arts and Humanities Research Council (AHRC)TOP