Where word forms collide: A typology of syncretism

Syncretism: An annotated bibliography

Overview

This annotated bibliography was produced by Matthew Baerman as part of the ESRC-funded project 'Where word forms collide: A typology of syncretism' (Grant number: R000237939). This support is gratefully acknowledged.

Some of the citations have been taken from a bibliography compiled by Frans Plank as part of the EUROTYP project. Thanks also to Peter Arkadiev for additional citations.

Bibliography

Aronoff, Mark 1994. Morphology by itself: stems and inflectional classes (Linguistic Inquiry monograph 22). Cambridge, MA.: MIT Press.

Introduces the notion of the "morphome", namely a purely morphological formative to which morphosyntactic functions may be associated. Syncretism is briefly discussed (pp. 83-84) w.r.t. nominal declension Latin, where it is treated as the result of the disjunctive association of functions with the morphome.

Baldi, Philip. 1983. On some recent claims in morphological theory. General Linguistics 23. 171-190.

A critique of Williams (1981). Numerous counterexamples are given to the predictions implied in Williams’ model of syncretism in Latin noun declension.

Bátori, István. 1990. Die Markierung des Objekts am Verb im Mordwinischen: morphologische Unterbestimmtheit und Homophorphie. Nyelvtudományi Közlemények 91. 15-23.

The syncretic system of object marking on the verb in Mordvin is attributed to underspecification. The claim advanced in previous accounts that this is due to erosion of a previously fully specified system is refuted.

Bavin, Edith L. and Tim Shopen. 1987. Innovations and neutralizations in the Warlpiri pronominal system. Journal of Linguistics 23. 149-175.

Fieldwork by the authors shows that the inclusive-exclusive distinction in 1pl pronouns in Walpiri is being lost. For the independent pronoun and subject clitic the originally inclusive forms prevail, while for the object clitic the originally exclusive form prevails. In all cases the resulting form has the exclusive meaning. Additionally there is a tendency towards loss of a distinct dual, most pronounced in the object clitics but hardly attested for the independent pronouns.

Bayer, Joseph, Markus Bader and Michael Meng. 2001. Morphological underspecification meets oblique case: syntactic and processing effects in German. Lingua 111. 465-514.

It is proposed that, in German, the oblique cases dative and genitive differ from the structural cases nominative and accusative in that they are have an extra structural layer (Kase Phrase -- though note this is not only realized by inflectional case marking), which requires morphological licensing. Structural cases, on the other hand, can remain unmarked.

Bazell, C.E. 1960. A question of syncretism and analogy. Transactions of the Philological Society. 1-12.

It is argued that phonological or phonetic similarity of forms favors diachronic syncretism in certain morphological contexts. Examples mainly from Old English.

Béjar, Susana and Daniel Currie Hall. 1999. Marking markedness: the underlying order of diagonal syncretisms. Paper delivered at the Eastern States Conference on Linguistics (to be published in the conference proceedings).

Instances of "diagonal syncretism" (where the items share no features, e.g. between one case in the singular and a another case in the plural, or between a noun form and a verb form) are claimed to share the same degree of markedness, as represented by feature geometry, and are thus systematic. Examples from Old Church Slavic and Standard Arabic.

Bentley, Delia and Thórhallur Eythórsson. Alternation according to person in Italo-Romance. In: Laurel J. Brinton (ed.) 2001. Historical Linguistics 1999. Amsterdam/Philadelphia: John Benjamins. 63-74.

In some Italian dialects the perfect auxiliary ‘have’ is suppletive in the 2nd and 3rd singular, showing reflexes of the ‘be’ auxiliary. This substitution is attributed to avoidance of the homophony which would have obtained earlier between the 2nd and 3rd singular ‘have’ forms.

Bickel, Balthasar, Walter Bisang and Yogendra P. Yadava. 1999. Face vs. empathy: the social foundations of Maithili verb agreement. Linguistics 37/3. 481-518.

Syncretism (and other morphological phenomena) of person marking in Maithili verbs is attributed to pragmatic factors. Syncretism of 1st person and 2nd person honorific markers is a result of one of the aspects of the face constraint ("Avoid uniquely identifying reference to speaker"). Non-marking of the object in various context is attributed to the sometimes conflicting demands of face and empathy. Face favors the marking of 3rd person honorific arguments and discourages the marking of 1st person and 2nd person honorific. Empathy, however, favors the marking of speech act participants (1st and 2nd person). When there is a conflict, face overrides empathy.

Bierwisch, Manfred. 1967. Syntactic features in morphology: general problems of so-called pronominal inflection in German. In: To honor Roman Jakobson: essays on the occasion of his seventieth birthday. The Hague: Mouton. 239-270.

The paradigm of the German pronominal adjective dieser is analyzed in terms of transformational rules manipulating the ± values of features (such as Oblique and Governed). Case forms are derived by the application of ordered rules, from more to less specific, with syncretism resulting from underspecification within the domain of a given rule. It is suggested that the order of rules reflects the relative markedness of the features they refer to.

Bílý, Milan and Thore Pettersson. 1988. Neutralization in case morphology. Working Papers (Lund University, Dept of Linguistics) 33. 5-21.

A rejection of what the authors term the "analogical principle" in the determination of noun cases, whereby a case form that has been established in one paradigm is assumed to exist for all paradigms, even where no distinct form is found. This leads to the assumption of massive neutralization in syncretic paradigms, which is deemed superfluous. Three preconditions are proposed for recognizing a distinct case:

(a) there must be a distinct form in at least some paradigm

(b) a covert marked case cannot be recognized for paradigms where only the unmarked counterpart is found (e.g. the marked partitive genitive in Russian occurs only in contexts where the unmarked genitive can also be found, so that we should not recognize a distinct partitive genitive for nouns in which it is not formally manifested)

(c) if a single case form in one paradigm corresponds to multiple case forms in another, these can be considered distinct cases in the former only if they represent incompatible functions in the latter. Where the functions are compatible, what we are witnessing in the syncretic paradigms is rather the replacement of one case by another.

Three types of syncretism are recognized:

(a) one case replaces another (conditions (a) and (c) relevant).

(b) unmarked case takes precedence over marked (condition (b))

The approach of Comrie (1986) is criticized; the authors characterize his approach as "functional constrained by formal criteria", whereas theirs is the reverse.

Bittner, Dagmar. 1995. Affixhomonymie in der Naturlichkeitstheorie. In: Norbert Boretzky et al. (eds.) Natürlichkeitstheorie und Sprachwandel: Beitrage zum internationalen Symposium über Natürlichkeitstheorie und Sprachwandel an der Universität Maribor, 1993. Bochum: Brockmeyer. 123-144.

Developing ideas of Leiss (cf. Liess 1997), a radical approach to the Biuniquenss principle of Natural Morphology is taken, one which assumes any apparent morphological homonymy can be analyzed as underlying identity. The element -er- in German occurs in six contexts:

(a) noun derivation, e.g. Bäck-er ‘baker’ < backen ‘bake’

(b) plurals, e.g. Häus-er ‘houses’ < Haus ‘house’;

comparative adjectives, e.g. gemütlich-er ‘pleasanter < gemütlich ‘pleasant’

(d) adjective declension, e.g. grün-er Baum ‘green tree’

(e) pronominal declension, e.g. dein-er ‘your’

These are all united by the meaning ‘definiteness’, and possibly ‘quantification’ as well.

Blevins, James P. 1995. Syncretism and paradigmatic opposition. Linguistics and Philosophy 18. 113-152.

Analysis of English verbal and German pronominal inflection in the framework of HPSG. An inflection paradigm consists of marked forms which are defined against the background of unmarked forms (e.g. marked walks defined against unmarked walk). While the identification of marked forms is unproblematic (walks = 3sg present), unmarked forms lend themselves to different analyses:

(a) negative feature specification; e.g. walk does not = 3sg.

(b) disjunctive feature specification; e.g. walk = 1sg or 2sg or 1pl

(c) morphological blocking; e.g. walk is completely unspecified, whereby the rules stipulate that an unspecified form cannot be used when a semantically more specific alternative exists; thus walk cannot be used for 3sg (cf. Andrews 1990).

(d) paradigmatic blocking; similar to morphological blocking, but specificity is gauged in terms of those features instantiated within a given inflectional paradigm.

The last approach is favored. Syncretism which results from underspecification of features is purely ‘artefactual’; i.e. it is purely a representational notion, resulting from an overarticulated description. Syncretism that cannot be accounted for in this way represents accidental homophony.

Bloch, Bernard. 1966/47. English verb inflection. Language 23. 394-418; Reprinted in Martin Joos (ed.) Readings in linguistics (volume 1). Chicago: Chicago University Press. 243-254.

Notable for the way in which in which the overarticulated forms of the verb ‘to be’ are handled. The verb paradigm is divided into four categories (formed through suffixation):

(a) 3sg

(b) preterit

(d) gerund

The marking of persons other than 3sg in am, are and was, were is construed as allomorphy, and thus is no basis for postulating additional paradigmatic categories.

Bloom, Douglas B. 1999. Case syncretism and word order freezing in the Russian language. Master’s thesis, Stanford University.

It is observed that Russian nouns with nominative-accusative syncretism do not show the same freedom in word order that other nouns do, even if other inflected elements in the clause make its case function unambiguous. This phenomenon is subject to an analysis in terms of Constructive Case in the LFG framework, the case syncretism being attributed to disjunctive marking of both nominative and accusative.

Boeder, W. 1976. Morphologische Kategorien. In: K Braunmüller and W. Kürschner (eds.) Grammatik. Akten des 10. Linguistischen Kolloquiums, Tübingen 1975 (vol. 2). Tübingen: Niemeyer. 117-126.

Syncretism is analyzed in terms of marked/unmarked feature values. 3 rules are offered:

(a) the more marked features a form has, the greater the likelihood of syncretism (e.g. syncretism is more likely in the marked plural than in the unmarked singular).

(b) syntactic features will be affected before semantic ones (e.g. case before number).

Apparent exceptions to these rules are a result either of an incorrect feature analysis, or because the phenomenon in question should not be characterized as syncretism. Examples are drawn from primarily from German; also Old Georgian, West Greenlandic and Romance.

Bradner, Aleš. 1997. K omonimii okoncanij Nsg i Gpl u sušcestvitel’nyx mužskogo roda v russkom jazyke [On the homonymy of the nominative singular and genitive plural endings with masculine nouns in Russian]. Sborník prací filozofické fakulty brnenské univerzity A 45. 149-156.

The syncretism between nominative singular and genitive plural forms in Russian (e.g. soldat) is shown to be restricted to words which occur mostly in the plural.

Calabrese, Andrea. 1995. Syncretism phenomena in the clitic systems of Italian and Sardinian. In: Jill Beckman (ed.) Proceedings of the New England Linguistic Society 25 (vol. 2). Amherst: University of Massachusetts, Amherst. 151-174.

Historical syncretism of clitics in Italian and Sardinian dialects is examined, namely:

(a) replacement of dative by the genitive or locative, or of the dative and genitive by the locative.

(b) replacement of accusative by reflexive when in a cluster of dative + accusative.

The analysis is in terms of Distributed Morphology. Cases are construed as fully specified feature bundles. Syncretism results from filters which block certain feature combinations at the level of syntax; in its place an adjusted feature bundle is formed. Where there is a form available which directly matches this revised feature bundle, it is used. Where not (as in 1pl-locative syncretism), replacement occurs by metaphor extension based on partial overlap of features.

Calabrese, Andrea. 1998. Some remarks on the Latin case system and its development in Romance. In José Lema and Esthela Treviño (eds.) Theoretical analyses on Romance languages. Amsterdam/Philadelphia: John Benjamins. 71-126.

The Latin case system is broken down into features after the fashion of Jakobson (1936, 1958). Two types of syncretism are identified:

(a) Absolute Case syncretism results from Case restrictions, which ban certain combinations of features. Thus the Case which would otherwise be represented by the banned feature combination cannot exist as a morphologically distinct entity in the language.

(b) Contextual Case syncretism results from underspecification of the Case exponent; that is, the Case, as defined by the component features, exists in the language, but there are contexts where an underspecified exponent is used, resulting in syncretism in a class of words.

The reduction in Case exponents in the Romance languages is attributed to the operation of Absolute Case syncretism; Contextual Case syncretism will not have been a contributing factor. The formal analysis is in terms of Distributed Morphology.

Carstairs, Andrew. 1984. Outlines of a constraint on syncretism. Folia Linguistica 18. 73-85.

Inflectional homonymy appears to be constrained by:

(a) they way morphosyntactic properties are realized, either cumulatively (e.g. with a single exponent for case and a single one for number) or simultaneously (e.g. with a single exponent expressing case and number both).

(b) the distinction between syncretism (where one form realizes multiple morphosyntactic properties), and ‘takeover’ (where the form associated with one morphosyntactic property serves as the exponent for another).

Using data from Latin, Ancient Greek, Sanskrit, Russian, Lithuanian, German, Italian, Arabic, Hungarian, Finnish and Old Georgian, it is suggested that syncretism only occurs with simultaneous exponence, because only in such a system does homonymy decrease rather than increase inflectional complexity. Otherwise, inflectional homonymy must represent a ‘takeover’.

Carstairs, Andrew. 1987. Allomorphy in inflection. London: Croom Helm.

Chapter 4 is an extension and revision of Carstairs (1984). There is further discussion of the distinction between systematic and accidental homonymy. Syncretism and takeovers are defined in terms of ‘relevance’ (Bybee 1985): syncretism occurs where the morphosyntactic context is higher on the relevance hierarchy than the neutralized property, while takeovers occur where the morphosyntactic context is lower. There are added examples from Hungarian, Finnish, Italian, Georgian and Arabic.

Carstairs, Andrew. 1998. How lexical semantics constrains inflectional allomorphy. In: Geert Booij and Jaap van Marle (eds.) Yearbook of Morphology 1997. Dordrecht, Boston, London: Kluwer. 1-24.

An attempt to transfer four generalizations from the realm of lexical semantics to the realm of inflectional affixes, namely:

(a) Principle of Contrast: every two affixes contrast in meaning.

(b) Exclusive disjunction bar: the meaning of an affix cannot contain a disjunction of mutually exclusive morphosyntactic properties.

(d) Unmarked property bar: the meaning of an affix cannot contain reference to the least marked value of a feature.

These constraints affect syncretic patterns by limited what information can be encoded in an inflectional rule, and hence how underspecification works. The device of referrals (takeovers) is admitted as well, though it is presumably not affected by these constraints. Examples from Hungarian, Latin and Turkish.

Carstairs, Andrew and J.P. Stemberger 1988. A processing constraint on inflectional homonymy. Linguistics 26/4. 601-17.

Carstairs (1984, 1987) makes the claim that systematic inflectional homonymy will only occur when the syncretized properties are cumulative (i.e. when the syncretized property and its conditioning context are expressed simultaneously). Using a connectionist computer model, it is shown that homonymy decreases the complexity of paradigms with cumulative exponence, but increases the complexity of paradigms with agglutinative (non-simultaneous) exponence.

Chvany, Catherine V. 1986. Jakobson’s fourth and fifth dimensions: on reconciling the cube model of case meanings with the two-dimensional matrices for case forms. In Richard D. Brecht and James S. Levine (eds.) Case in Slavic. Columbus, Ohio: Slavica. 107-129

A critique of Jakobson (1958), examing the features ‘direct’ and ‘definite’. It is argued that, although these features do not figure in the famous cube metaphor of Russian case, they are needed for an adequate account of the patterns of syncretism found in Russian nominals.

Coats, Herbert S. 1973. Old Russian declension: a synchronic analysis. In Ladislav Matejka (ed.) American Contributions to the Seventh International Congress of Slavists: I: Linguistics and Poetics, 67-99. The Hague: Mouton.

Asserts (following Jakobson 1958) that any synchronic account of a declensional system such as that found in Old Russian must account not just for the forms but for syncretism as well. Syncretism is treated as underspecification within rule blocks. Note that a precondition on syncretism in Coats’ definition is that the forms must share some feature; thus e.g. identity of genitive singular and nominative plural is treated as accidental.

Coleman, Robert. 1976. Patterns of Syncretism in Latin. In: Anna M. Davies and Wolfgang Meid (eds.) Studies in Greek, Italic and Indo-European Linguistics Offered to Leonard R. Palmer. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck. 47-56.

A tripartite typology of the scope of diachronic noun case syncretism is offered:

(a) 1st degree: loss of distinction in all members of one (or more) inflectional paradigm, but not in all paradigms.

(b) 2nd degree: loss of distinction in all paradigms, but restricted to a single number.

Diachronic syncretism is ascribed to (i) phonological change, producing 1st degree syncretism; (ii) analogical extension of 1st degree syncretism, producing 2nd and 3rd degree; (iii) replacement of independent forms by prepositional collocations.

Coleman, Robert. 1987. Early Greek Syncretism and the Case of the Disappearing -phi. In: John T. Killen, José L. Melena and Jean-Pierre Olivier (eds.) Studies in Mycenaean and Classical Greek Presented to John Chadwick (=Minos 20-22). Salamanca: Universidad de Salamanca. 113-125.

The diachronic syncretism of dative and locative in Ancient Greek is attributed to purely phonetic convergence, in contradistinction to Kurylowicz (1964), who assumes there was a prior semantic relationship.

Coleman, Robert. 1991. The Assessment of Paradigm Stability: Some Indo-European Case Studies. In Frans Plank (ed.) Paradigms: The Economy of Inflection. Berlin: Mouton de Gruyter. 197-212.

The typology of syncretism in Coleman (1976) is revised and expanded:

(a) 1st degree ("sporadic"): loss of distinction between some (but not all) lexemes in a given paradigm.

(b) 2nd degree ("partial"): = 1st degree syncretism in Coleman (1976).

(d) 4th degree ("total"): = 3rd degree syncretism in Coleman (1976).

The stability of noun paradigms over time is explored, using data from English, Latin Greek, Armenian, Lithuanian and Sanskrit, whereby instability is correlated with paradigmatic asymmetry (i.e. instances of 1st and 2nd degree syncretism).

Comrie, Bernard 1986. On delimiting cases. In Richard D. Brecht and James S. Levine (eds.) Case in Slavic. Columbus, Ohio: Slavica. 86–106.

Case is seen from two vantage points: "distributional case" refers to a syntactic context where, among the set of all nouns, distinct morphological behavior is found; thus even if only one noun shows a distinct form in a particular context, we say that context calls for a distinct distributional case. "Formal case" refers to distinct forms, each of which may map onto multiple distributional cases. It is assumed that those distributional cases which can be represented by a single formal case are united by shared features. Examples from Russian, Latin Latvian and Dyirbal.

Comrie, Bernard 1991. Form and function in identifying cases. In: Frans Plank (ed.) Paradigms: The economy of inflection. Berlin: Mouton de Gruyter. 41-56.

Elaboration of Comrie (1986). Additional examples from Russian, and from a Swiss German dialect (Gurinerdeutsch) and Standard German.

Corbett, Greville G. and Norman M. Fraser (1993) Network Morphology: a DATR account of Russian nominal inflection. Journal of Linguistics 29: 113-42.

In an exposition of the principles of Network Morphology, the syncretism of the accusative with nominative or genitive in Russian nominals is described in terms of Prediction Rules (Perlmutter and Orešnik 1973) or Rules of Referral (Zwicky 1985), which are here given a formal expression.

Creider, Chet and Richard Hudson. 1999. Inflectional phonology and word grammar. Lingua 107/3-4.

In an article which specifically adresses the treatment of morphology within Word Grammar, syncretism in English verbs is treated by reference to an "x-form" (similary to Aronoff’s 1994 morphome); i.e. there is a single form, defined by morphology, which multiple functions are associated with.

Dalrymple, Mary and Ronald M. Kaplan. 2000. Feature indeterminacy and feature resolution. Language 76/4. 759-798.

In the first part, feature resolution with syncretic nominals is explored. Underspecification and disjunction analyses are rejected in favor of a representation composed of sets of atomic values (e.g. the Polish relative kogo ‘whom’ is analyzed as {acc, gen}, and can serve simultaneously as the object of lubi ‘loves’, which takes the accusative, and nienawidzi ‘hates’, which takes the genitive); the discussion focuses largely on free relatives in German. A similar approach is advocated for noun class agreement in Bantu languages. In the second part resolution of person features of coordinated verb subjects is treated. Although (plural) person values are broken up into atomic values (s=speaker, h=hearer, whereby, in languages without a 1st person inclusive/excusive distinction, {s,h} = 1st person, {h} = 2nd person and {} = 3rd person), syncretism per se is not treated.

Daveluy, Michelle. 1989. Neutralisation morphologique en français parle a Montréal: l'exemple des déterminants démonstratifs. Revue québecoise de linguistique théorique et appliquée 8/3-4. 129-146.

The failure of the demonstrative ce ci/la ‘this/that’ to distinguish gender in some varieties of colloquial Montreal French is attributed to the morphological neutralization of gender. The grammatical and sociological factors favoring this neutralization are analyzed via VARBRUL.

Delbrück, Berthold. 1907. Synkretismus: ein Beitrag zur germanischen Kasuslehre, Strassburg: Trübner.

Detailed historical study of the collapse of the instrumental, locative, dative and genitive in the Germanic languages.

Fennell, Trevor G. 1975. Is there an instrumental case in Latvian? Journal of Baltic Studies 6.41-48.

Latvian is traditionally considered to have an instrumental case, used only with the preposition ar. However the instrumental never has a distinct form: in the singular it is identical to the accusative, in the plural to the dative (the case which all prepositions govern in the plural). It is concluded that there is no real basis for postulating an instrumental case at the level of morphosyntax.

Fradkin, Robert A. 1991. Marking, markedness, and person-gender-number patterning in the Arabic tenses and moods. Folia Linguistica 25. 609-664.

The components of Arabic verb conjugation are analyzed in terms of marked feature values. Syncretism results from underspecification of values in marked contexts. Thus the t- prefix, found in the masculine 2nd person and feminine 3rd person, is analyzed as [-central] – i.e., not the speaker, whereby it can be further specified for only one of two possible marked values, either [+participant], and hence 2nd person but not feminine, or [+feminine], and hence feminine but not a participant in the speech act (and so 3rd person). Also, the marked moods (subjunctive and jussive) share many forms with each other while the indicative remains relatively distinct.

Frampton, John. To appear. Syncretism, impoverishment and the structure of person features. Papers from CLS 38

A Distributed Morphology account of the syncretism of 1sg and 3sg past in Germanic. It is attributed to the Impoverishment (deletion) of reference to 1st person features in the underlying person feature represantations [+1, -2] ‘1st person’, [-1, +2] ‘2nd person’ and [-1, -2] ‘3rd person’, yielding a two-way distinction of [-2] ‘not 2nd person’ and [+2] ‘2nd person’. Further evidence for a bifeatural analysis of person is adduced from Berber, where the distribution of the prefix t- over 2sg and 3sg is attributed to the feature specification [-1].

Georgiev, V I. 1973. Interdependenz von Syntax und Morphologie. In: Georges Redard (ed.) Indo-germanische und Allgemeine Sprachwissenschaft. Wiesbaden: Reichert. 59-65.

Acceptable versus unacceptable in the history of Slavic noun declensions is examined: the accusative does not fall together with the dative, nor the instrumental with the locative. Each pair represents a single syntactic category and a single position w.r.t. word order, so that homonymy would produce ambiguity. Where phonological change brings this about, the offending inflectional morphemes are replaced.

Goddard, Cliff. 1982. Case systems and case marking in Australian languages: a new interpretation. Australian Journal of linguistics 2. 167-96.

It is argued that the notions of S(ubject), A(gent) and O(bject), customarily used in Australianist literature to describe core grammatical functions, should be replaced by the traditional notion of case, namely nominative, ergative and accusative, respectively. So-called split ergativity, where both ergative-absolutive and nominative-accusative systems are found in a single language, should then be treated as resulting from syncretism of the nominative and the accusative or the ergative.

Gvozdanovic, Jadranka. 1991. Syncretism and the paradigmatic patterning of grammatical meaning. In: Frans Plank (ed.) Paradigms: The economy of inflection. Berlin: Mouton de Gruyter. 133-160.

In the first part, case syncretism in standard Serbo-Croatian is analyzed in terms of the case features introduced in Jakobson (1958/1984), which is then extended to account for further syncretic patterns found in the dialects and in the aberrant speech of immigrant children in Sweden. The second part deals with person and number syncretism in verbs in the Tibeto-Burman languages Bahing and Bantawa, whereby person is broken down into [±me] [±you], and number (singular, dual and plural) into [±plural] and [±restricted], with syncretism of subject and object agreement markers determined by underspecification of these features.

Halle, Morris. 1992. The Latvian declension. In Geert Booij and Jaap van Marle (eds.) Yearbook of Morphology 1991. 33-48.

In this Distributed Morphology analysis of Latvian declension, instances of case syncretism (accidental cases aside) are treated as resulting from underspecification, except for one pair of referrals: the nominative/genitive syncretism which occurs in the singular of some noun classes both in the masculine declension (nom/gen tirgus vs. nom tevs, gen teva on the one hand and nom/gen zivs vs. nom masa, gen masas on the other) is ascribed to referrals which take the nominative to form the genitive in the former case, and the genitive to form the nominative in the latter.

Hamilton, William S . 1974. Deep and surface changes in four Slavic noun systems. Linguistics 127. 27-73.

Case syncretism in Czech, Polish Russian and Serbo-Croatian is described. 28 patterns are identified in the contemporary languages. There is remarkable continuity between the underlying (or ancestral) system and the suface (or contemporary) system. Changes have entailed primarily (i) disambiguation of nominative-accusative syncretism by the use of the genitive for the animate accusative; (ii) changes in declension class membership, resulting tangentially in the increase or decerease of particular patterns; (iii) extension of particular syncretic patterns from one declension class to another (chiefly involving the genitive and locative). The most complex system, in terms of the number and variety of syncretic patterns, is that of Czech, the simplest is that of Serbo-Croatian.

Heath, Jeffrey 1991. Pragmatic disguise in pronominal-affix paradigms. In F. Plank (ed.) Paradigms: The Economy of Inflection. Berlin: Mouton de Gruyter. 75-90.

Person/number agreement markers on transitive verbs in Australian languages are frequently morphologically aberrant when the arguments are 1st person and 2nd person. This is attributed to a social-pragmatic taboo on a transitive relationship between speech-act participants. Among the strategies employed are the neutralization of number and the neutralization of person, or the substitution of the markers for different persons.

Heath, Jeffrey. 1998. Pragmatic skewing in 1 <--> 2 pronominal combinations in Native American languages. International Journal of American Linguistics 64. 83-104.

The same theme as Heath (1991) is treated, the examples in this study being drawn instead from the native languages of the Americas.

Helmbrecht, Johannes. 1999. The typology of 1st person marking and its cognitive background. In: Masako K. Hiraga, Chris Sinha, and Sherman Wilcox (eds.) Cultural, Psychological and Typological issues in Cognitive Linguistics. Amsterdam: John Benjamins. 285-297.

The opposition between speaker (1st person) and hearer (2nd and 3rd person) in verbal morphology is explored, including examples where the 2nd and 3rd person forms are syncretic (Chitimacha, Tonkawa and Sherpa). This is attributed to the psychological reality of the split between the self-conscious entity (1st person) and others.

Hewson, John. 1989. Motivated Syncretism. Journal of the Atlantic Provinces Linguistic Association/Revue de l'Association de Linguistique des Provinces Atlantique 11. 39-56.

The term "synapsis" is introduced to refer to the phenomenon of (semantically) motivated syncretism, which is identifiable by three criteria:

(a) analogical reshaping occurs (i.e. the formal identity cannot be attributed solely to sound change)

(b) there are shared semantic features

This is illustrated by the formal identity of a marked animate singular form and inanimate plurals in:

(a) Algonkian (Menomoni, Cree and Ojibwa), where the marked obviative singular (an adnominal form which occurs only with animates) is identical to the inanimate plural.

(b) Indo-European languages, where the marked feminine singular is identical to the neuter plural.

The shared semantic notion is said to be "transcendence", i.e. when something "lies outside or transcends an already established unity". Other examples from French (identity of article and pronominal forms) and English (various instantiations of the -s affix).

Hjelmslev. 1943/1961. Prolegomena to a theory of language (translation of Omkring sprogteoriens brundlaeggelse). Madison: University of Wisconsin Press.

Syncretism is defined within the context of glossematics (note that the term is applied not just to grammatical features, but to any features, e.g. phonological). The notion of dominance is introduced: the feature which is a necessary condition for a particular syncretism is said to dominate the overlapping (i.e. syncretism); for example, neuter gender dominates the overlapping of nominative and accusative singular in Latin. The actual manifestation of a syncretism is said to be a fusion or an implication: a fusion is identical to all or none of its component features, while an implication is identical to the realization one (or some) of its components, but not all. Syncretisms are said to be resoluble or irresoluble. A resoluble syncretism is one where, in the given context, one can determine which of the syncretized features is actually being expressed; one cannot determine this for an irresoluble syncretism.

Huddleston, R. 1975. Homonymy in the English verbal paradigm. Lingua 37. 151-76.

"Words" (i.e. a particular morphosyntactic function) are contrasted with "word forms". In analyzing the English verb system, previous researchers identified anywhere from five to 29 distinct words. Given the two paradigms:

a. I finished it. c. I took it.

b. I have finished it. d. I have taken it

the possible interpretations are:

(a) a + b are the same word, realized by the same word form, while c + d are different words, realized by different word forms.

(b) a + b are different words realized by the same word form, while c + d are different words, realized by different word forms.

(c) a + b are the same word, realized by the same word form, while c + d are the same word, realized by different word forms.

The second approach, the only one which entails homonymy, is advanced here.

Jakobson, R.O. 1936/1984. Contribution to the general theory of case: general meanings of the Russian cases [translation of Beitrag zur allgemeinen Kasuslehre: Gesamtbedeutung der russischen Kasus, originally in TCLP 6]. In: Linda R. Waugh and Morris Halle (eds.) Roman Jakobson. Russian and Slavic grammar: Studies 1931-1981. Berlin: Mouton de Gruyter. 59-103.

Case syncretism in the Russian noun is used support the notion that cases are entities with an invariant general meaning (and are not simply an agglomeration of unconnected specific meanings). Case meaning is defined in terms of marking for four semantic features:

(a) directionality: accusative, dative

(b) scope: genitive I, genitive II, locative I, locative II

(d) shaping: genitive II, locative II

with nominative being unmarked. Possible and impossible syncretic patterns are defined in terms of these features.

Jakobson, R.O. 1958/1984. Morphological observations on Slavic declension (the structure of Russian case forms) [translation of Morfologiceskie nabljudenija nad slavjanskim skloneniem (sostav russkix padežnix form, originally in American contributions to the Fourth International Congress of Slavists, Moscow]. In: Linda R. Waugh and Morris Halle (eds.) Roman Jakobson. Russian and Slavic grammar: Studies 1931-1981. Berlin: Mouton de Gruyter. 105-133.

A reworking of Jakobson (1936/1984) in which the cases are modeled as the points of a cube. The feature ‘shaping’ is conflated with directionality, and each case is represented in terms of ± values for each of the remaining three features (directionality, quantification (‘scope’ in 1936/84) and marginality (‘peripherality’ in 1936/84)), which form the dimensions of the cube. Two further features, ‘direct’ (nominative and accusative) and ‘definite’ (the directional and quantificational cases), play a role in the organization of syncretic patterns.

Johnston, Jason. 1997. Systematic homonymy and the structure of morphological categories: some lessons from paradigm geometry. Ph.D. Thesis, University of Sydney.

An extension of the geometric model of syncretism found in McCreight and Chvany (1991) and other works by Chvany – specifically, a linear model, where the potential for syncretism is equated with adjacency. The model presented here is in the form of a branching tree, whereby each branch carries a plus (+) or minus (–) specification for some feature. This is supplemented by "unmarking", whereby the minus branch of a node may take on the minus values of any nodes it dominates, which allows for the definition of overlapping syncretisms (e.g. a form uniting values A + B and form uniting values B + C). This system allows one to identify the natural classes defined within a linear model (namely in terms of the +/- values). Cross-classification (as in Jakobson 1936/84 or 1958/84) and rules of referral are rejected as being too powerful and, in the latter case, syntactically unmotivated. Principles languages investigated are Russian, Greek, German, Latin and Arabic.

Karinš, A. Krišjanis 1994. Functionalism and linguistic change in Latvian verb morphology. Linguistica Baltica 3. 109-120.

An argument against the claim that avoidance of syncretism is motivating factor in linguistic change. In Latvian, the 2sg present tense verb ending -i has been lost in some verb classes, a change which is inexplicable as a result solely of sound change. Endzelins (1922) claimed this was to disambiguate the homophony between this form and the 2sg past tense, also with the ending -i. The loss of -i was then analogically extended to verb classes which, by virtue of stem alternations, lacked the present~past syncretism in the first place. However, the loss of -i leads to syncretism of 2sg and 3rd person (Endzelins 1922 argued that this latter syncretism was preferable, since it could be disambiguated by pronouns). Also, there does not appear to have been a corresponding elimination of tense syncretism in the 1sg, which occurs in some verb classes. It is concluded that avoidance of syncretism cannot be seen as anything more than a general tendency, one which is readily contradicted.

Kibrik, A.E. 1997. Ierarxii, roli, nuli, markirovannost’ i “anomal’naja” upakovka grammaticeskoj semantiki. Voprosy Jazykoznanija 4. 27-57.

Using Dargin, Svan, Yimas, Yukaghir and Aljutor as cases studies, Kibrik shows (among other things) how unexpected syncretism of agreement markers in verbal morphology can be attributed to intersecting hierarchies of person deixis (1st, 2nd and 3rd person), grammatical role (subject vs. object) and animacy.

Kiparsky, Paul. 2001. Structural case in Finnish. Lingua 111. 315-76.

The bulk of the article is devoted to the relationship between syntax and abstract case in Finnish. The last part is devoted to an OT account of the discrepancies between abstract case and morphological case, which varies between pronouns and nouns: where pronouns have a distinct accusative case form, nouns use the nominative or genitive in the singular (under different syntactic conditions) and the nominative in the plural. This is attributed to faithfulness constraints which selectively target pronouns or nouns. [Max-HR] (favoring the morphological expression of object) targets nouns, while [Max-LR] (favoring the expression of transitive subject and ‘higher’ object) targets pronouns. This is compared to split ergativity and similar noun/pronoun splits, which are said to result from various rankings of the two constraints (note though that [Max-LR] is not actually used in the constraint hierarchy for Finnish, being substituted by the markedness constraint *[-HR], which effectively bans the use of the accusative and partitive cases).

Klenin, Emily. 1983. Animacy in Russian: a new interpretation Columbus: Slavica.

A diachronic account of the rise of the syncretic genitive-accusative form in Russian, used as a marker of animacy in nouns. The syncretism appeared first in the paradigm of the interrogative pronoun ‘who’, then spread through the whole pronominal system (irrespective of animacy), and thence into nominal paradigms (where not all declension were affected). A significant factor favoring it was the prior existence of nominative-accusative syncretism -- i.e. genitive-accusative syncretism reestablished a distinction between nominative and accusative. However, contrary to some previous assertions, this was neither a necessary nor sufficient condition.

Klimonov, Vladimir. 1996. Vidovoj sinkretizm i vidovaja differentsiatsija v zaimstvovannoj glagol’noj leksike [Aspectual syncretism and aspectual differentiation in the borrowed verb lexicon]. Studia Rossica Posnaniensia 27. 181-186.

A treatment of biaspectual verbs in Russian in the framework of Natural Morphology. Paired imperfective-perfective verbs in Russian can be morphologically related to each other in four ways:

(a) imperfective transparently derived from perfective (stem alternation or suffixation)

(b) perfective derived from imperfective via prefixation

(d) suppletive

It is observed that there is a tendency in the non-standard language to differentiate biaspectual verbs (which are typically borrowings); i.e. to rectify the non-iconicity of the forms. In theory devices (a), (b) or (c) could be employed, but in practice device (b) prevails, which is attributed to its being maximally iconic (perfective is seen as morphosyntactically complex, which is reflected by the addition of prefixal material).

Kordic, Snježana. 1995. Genitiv/Akkusativ-Synkretismus beim koratisch-serbischen Relativpronomen [Genitive-accusative syncretism in the Serbo-Croatian relative pronoun]. Zeitschrift für Slawistik 40/2. 202-13.

A discussion of the expansion of the use of the genitive case form for the accusative in the relative pronoun koji. According to the normative standard, the only context where this should occur is when the antecedent is animate being of the masculine gender. However, in non-standard speech and writing there has been a growing tendency to extend this to inanimates, even those of the neuter gender. This is partly attributable to the existence of genitive-accusative syncretism in the third person pronoun, where it obtains for masculines and neuters regardless of animacy.

Kortlandt, Frederik. 1985. The syncretism of nominative and accusative in Armenian. Revue des etudes armeniennes. n.s. 19. 19-24.

Nominative and accusative singular fell together in Proto-Armenian everywhere except in personal pronouns. This was the result of three processes:

(a) phonetic merger of the two cases

(b) replacement of the accusative form by the nominative form

Krámský, Jirí. 1976. A typological study of morphological homonymy. In: Jirí Krámský Papers in general linguistics. 156-180.

Syncretism ("idendity of form" in Kramsky’s terms) is surveyed in languages of different morphological type. Syncretism is most frequent in the flexive type (e.g. classical Indo-European, or Slavic), less frequent in the introflexive type (e.g. Semitic) and still less so in the agglutinative (e.g. Turkish) and isolative type (e.g. French). It is asserted that it is a means of simplifying a language’s morphology, in as much as it rarely leads to ambiguity.

Lakämper, Renate and Dieter Wunderlich. 1998. Person marking in Quechua – a constraint-based minimalist analysis. Lingua 105. 113-148.

The history of object marking on verbs in Quechua is examined through the typology of contemporary dialects. In the most archaic the person of the object is only distinguished if it is higher-ranked than the person of the subject, according to the hierarchy 1 > 2 > 3. In more innovative dialects the marking of 2nd person objects has been eliminated completely, while in the most innovative dialects this syncretism has been eliminated, full marking of both subject and object having been instated.

Laskowski, Roman. 1990. The Structure of the Inflectional Paradigm. Scando-Slavica 36. 149-159.

The inflectional paradigm consists of a functional paradigms, made up of ‘flectemes’ (morphosyntactic slots), paired with a formal paradigm, made up of ‘textual forms’. Syncretism is defined as instances where a single textual form may be paired with more than one flecteme. Syncretic paradigms are to be distinguished from ‘defective’ paradigms (e.g. indeclinables in Polish), where there are simply no morphological means for distinguishing the values of the inflectional category.

Leiss, Elisabeth. 1997. Synkretismus und Natürlichkeit. Folia Linguistica 31/1-2. 133-160.

On the biuniqueness principle of Natural Morphology (‘one form, one meaning’), it is proposed that syncretism – understood as the neutralization of features – should be recognized not only within inflectional paradigms, but even across grammatical categories. The example used is the –s affix in English, namely:

(a) –s plural

(b) –s possessive

all of which are said to represent a single underlying meaning, sharing the feature +Relation (possessive and plural relate parts to wholes, while the 3sg ending expresses an anaphoric relationship).

Lumsden, J. S. 1992. Underspecification in grammatical and natural gender. Linguistic Inquiry 23. 469-486.

A reply to Farkas’ (1990) treatment of Romanain gender syncretism. Neuter agreement forms in Romanian are identical to the masculine in the singular and the feminine in the plural. Farkas (1990) accounts for this at the level of underlying gender features: (i) all three gender are represented as values of ‘feminine’ ([+Fem], [-Fem], [ ]); (ii) a feature insertion rules gives [ ] the value [-Fem] in the singular and [+Fem] in the plural. The inflectional markers themselves are fully specified. Lumsden criticizes the derivation of three values from one feature. In his account gender is fully specified by values of [±Fem] and [±Neut]; syncretism is the result of underspecification of the inflectional morphology.

Luraghi, Silvia. 1987. Patterns of case syncretism in Indo-European languages. In: Anna G. Ramat, Onofrio Carruba and Giuliano Bernini (eds.) Papers from the Seventh International Conference on Historical Linguistics. Amsterdam: John Benjamins. 355-371.

Diachronic syncretism is characterized as the functional merging of morphemes which had previously overlapped, either in their semantic value or their syntactic roles. Both types may be found in the development of the Indo-European case system:

(a) Semantic syncretism is described for Ancient Greek and Hittite (genitive-ablative as ‘source’, dative-locative (plus the directive in Hittite) as ‘location’, locative-instrumental as ‘development of action/process’, dative-locative-instrumental as ‘instrument/agent’, and ablative-instrumental as ‘condition’).

(b) Syntactic syncretism is described for Latin (locative-ablative-instrumental) and Germanic (dative-locative-ablative-instrumental). These differing patterns resulted from the different ways in which syntactic roles were parcelled out: Latin distinguished between nominal arguments (nominative, accusative and dative) and satellites, while German distinguished central arguments (nominative and accusative) from all other syntactic roles.

Since syntactic syncretism produces semantically unmotivated morphological merger, it encourages the use of prepositions for semantic disambiguation.

Luraghi, Silvia. 2000. Synkretismus. In: Geerd Booij, Christian Lehmann and Joachim Mugdan (eds.) Morphologie: Ein Handbuch zur Flexion und Wortbildung. Berlin: Mouton de Gruyter.

The notion of syncretism is opposed to that of homophony and polysemy. Homophony is defined as a formal non-distinction of functions which affects only a subset of the words in a language; i.e. the functions are, in principal, distinct in the language. Polysemy is defined as the multifunctionality of forms, where these functions are never distinct in the language. The term syncretism applies to the diachronic process of formal merger of functions which, if complete, can also be applied to the resulting synchronic state of polysemy. Syncretism of grammatical roles and of semantic roles is discussed w.r.t. Indo-European noun declension (cf. Luraghi 1987).

Martinet, Andre. 1968. Neutralisation et synctretisme. Linguistique 1. 1-20.

Neutralization, where features become irrelevant in a given context, is distinguished from syncretism, or accidental homphony, where the distinction between features remains relevant, though unexpressed.

McCreight, Katherine and Catherine V. Chvany. 1991. Geometric representation of paradigms in a modular theory of grammar. In: Frans Plank (ed.) Paradigms: The economy of inflection. Berlin: Mouton de Gruyter. 91-112.

The representation of morphosyntactic properties in terms of the positive or negative values of features is rejected in favor of a geometric model. Paradigms are laid out in the time-honored tabular form, with morphosyntactic properties (e.g. case, number) along the x, y coordinates describing cells which contain the forms. Adjacent cells may then be conflated into a macro-cell. It is argued (on the basis of syntactic conflict resolution) that this method better accounts for possible syncretisms than does feature analysis. Syncretism plays a key role in such a representation of the paradigm, since possible conflations determine the order of morphosyntactic properties w.r.t. each other: the greater the distance reflecting, the less the morphosyntactic affinity. Examples from German, Finnish and Russian.

Meiser, G. 1992. Syncretism in Indo-European languages. Transactions of the Philological Society 90/2. 187-218.

The many definitions of syncretism that are found in the literature, both in the synchronic and diachronic sense, are explored (though the focus here is diachronic). Two major senses are isolated, and it is proposed that two different terms be used:

(a) synemptosis: the merging of forms due to sound change, whereby the distinct paradigmatic categories are still maintained.

(b) syncretism: the merging of two paradigmatic categories, whereby there is no distinction at any level.

Diachronic syncretism comes about by:

(a) analogical extension of change due to synemptosis.

(b) levelling across paradigms, e.g. from one inflectional class or number to another).

(c) merger due to overlapping functions, i.e. where the functional range of two forms overlaps, this overlap may be extended so as to swallow up the distinction between the two.

(d) outside influence.

Syncretism brings about a superfluity of forms; how this is resolved is explored from the perspective of Natural Morphology (cf. Rix 1991). Examples from Indo-European (Italic, Germanic, Slavic, Armenian).

Miller, Raymond. 1990. Form and function of the peripheral cases in some Slovene dialects. Slovene Studies 12/1. 5-22.

An examination of the various patterns of syncretism of genitive, dative and locative plural in Slovenian dialects as the underspecification of Jakobson’s (1936, 1958) features of peripherality, directionality and quantification/scope.

Mignot, X. 1978. Homonymies entre les désinences casuelles du latin. Langages 50. 45-50.

Case syncretism in Latin is compared to that of number and gender. It is noted that case syncretism is frequent and number syncretism rare and exceptional. In the modern Romance languages, this has resulted in the loss of case distinctions (in nouns) alongside the retention of number, suggesting that there has always been a constraint against it. Gender syncretism, however, is common in Latin, yet gender distinctions are retained in modern Romance. This is because gender is an inherent property of nouns, while case is a contingent property of the grammatical context.

Moravcsik, Edith. 1994. Inflectional morphology in the Hungarian noun phrase: a typological assessment. EUROTYP Working Papers (Theme 7, No. 22). 58 pages.

The failure of the 3sg possessive marker in Hungarian is used as evidence for the following proposed universal constraint: "If a language has homonymous inflectional affixes that express a single meaning element, then, if the language has cumulative inflection at all, some of those will also be homonymous" (p. 43).

Noyer, Rolf. 1997. Features, positions and affixes in autonomous morphological structure. New York and London: Garland.

A model of morphology is outlined which distinguishes positions (dedicated slots for inflectional morphology) from features themselves. Syncretism is attributed to operations on both components. At the level of morphosyntax, Feature Cooccurence Restrictions (filters) may ban certain combinations of features. This triggers a readjustment of the morphosyntactic representation -- represented as a hierarchy of features -- through the delinking of the lowest-ranked feature (Impoverishment). At the level of morphology, on the other hand, there may be simple underspecification. In chapter I the prefix conjugation of Afroasiatic is treated, primarily with Semitic examples, where the neutralization of number distinctions in certain persons is treated (note the coincidence of t- ‘2nd person’ and t- ‘feminine’ is treated as accidental). In chapter II (with examples from many languages, with Mam treated in greatest depth), person and number values are discussed. It is claimed that person features, in addition to +1, +2 and +3, include ±participant (1st and 2nd person vs. 3rd) and –I (2nd and 3rd person vs. 1st) and –you (1st and 3rd person vs. 2nd), though these last two are claimed to be highly marked and thus rare. Not all examples of person syncretism are evidence of these features, as simple underspecification operates as well. Chapter III goes on to treat person and number marking in verbs with multiple arguments, characterized by the neutralization of the person features of the agent and in general by neutralization of number; a number of languages are discussed, Nunggubuyu in greatest depth.

Noyer, Rolf. 1998. Impoverishment Theory and morphosyntactic markedness. In Steven G. Lapointe, Diane K. Brentari and Patrick M. Farrell (eds.) Morphology and its relation to phonology and syntax. Stanford: Center for the Study of Language and Information.

An elaboration of the notion of Impoverishment (Noyer 1997), which entails deletion of features at the level of morphosyntactic representation, so that morphology is not even responsible for its expression. Simple deletion however is not sufficiently powerful. Using data from Nimboran, it is proposed that after the feature is deleted, it is replaced by the unmarked value of that feature by persistent redundancy rules. The effect is similar to rules of referral, but is more constrained, since in effect the form referred to must always be less marked.

Perlmutter, David M. and Janez Orešnik. 1973. Language-particular rules and explanation in syntax. In: Stephen R. Anderson and Paul Kiparsky (eds), A Festschrift for Morris Halle. New York: Holt Rinehart. 419-59.

Non-feminine nouns in Slovenian have no distinct accusative form in the singular: for inanimates it is the same as the nominative, for animates it is the same as the genitive. This is likewise true for any modifying adjectives. However, when the head noun is omitted, the adjective assumes the form of the genitive for inanimates as well. This is accounted for by assuming that adjectives in this context have become pronominalized, and that pronouns are characterized by the feature [+animate].

Pike, Kenneth. 1965. Non-linear order and anti-redundancy in German morphological matrices. Zeitschrift für Mundartforschung 32. 193-221.

Identities in inflectional paradigms of German are explored through geometric modelling, where identical forms in adjacent cells of a paradigmatic matrix are conflated. The focus is not whole word forms, but of ‘formatives’, i.e. elements that may be either super- or submorphemic. Both one dimensional and two dimensional models have their place. Syncretism that can be modelled on a two-dimensional matrix (e.g. where two forms share one vector) is termed ‘first degree neutralization’; syncretism that can only portrayed by a one dimensional matrix (i.e. where two forms, on a two dimensional matrix, share no vector) is ‘second degree neutralization’.

Plank, Frans. 1987. Number neutralization in Old English: failure of functionalism? In: Willem Koopman, Frederike van der Leek, Olga Fischer and Roger Eaton (eds.) Explanation and linguistic change. Amsterdam: John Benjamins. 177-238.

The number syncretism found within individual cases in the Old English noun paradigm might be seen as a counterexample to functionalist claims about the encoding of grammatical features, whereby one should expect the absence of number marking to be correlated with features such as inanimacy and the peripherality of the argument. However, in this case any functionally motivated adjustment will also have counterfunctional ramifications, because the cumulative inflectional morphology of Old English does not allow inflectional features to be manipulated separately. Thus this cannot be construed as an argument against functionalism; rather it shows there may be language-specific structural features which take precedence.

Plank, Frans. 1991b. Rasmus Rask's dilemma. In: Frans Plank (ed.) Paradigms: The economy of inflection. Berlin: Mouton de Gruyter. 161-196.

The starting point is Rasmus Rask’s attempts attempt (in the early 19th century) to find a invariant case order for the representation of Indo-European nominal paradigms, one which would reflect potential syncretism through adjacency. Rask allowed only linear order; here, various two-dimensional models are explored in the analysis of various ancient and modern Germanic and Slavic languages, Arabic, Sanskrit, Lithuanian, Classical Greek and Latin. It is concluded that such two-dimensional models are sufficient to characterize most syncretic patterns; those that do not fit are "isolated, superficial and in various respects local."

Plank, Frans. 1994. Homonymy vs. suppletion: a riddle (and how it happens to be solved in…). EUROTYP Working Papers (Theme 7, No. 23). 81-86.

The conflicting requirements of stem suppletion and inflectional homonymy in the genitive and locative dual forms of Slovenian clovek ‘person’ is explored: the dual stem must be the same as the singular, but the ending the same as the plural. This is resolved by:

(a) combining the singular stem with plural endings

(b) borrowing the plural forms

The second is more usual.

Pullum, Geoffrey K. and Arnold Zwicky. 1986. Phonological resolution of syntactic feature conflict. Language 62. 751-73.

Syncretic forms are characterized as being either neutral (the feature values are unspecified as the result of a systematic morphological rule) or ambiguous (as a result of unsystematic homophony between multiple forms with distinct feature values). Neutral forms can readily resolve feature value conflicts in coordination. Ambiguous forms, which result from a purely surface phonological identity, can as well, although under a more restricted set of circumstances. It is proposed that feature conflict resolution can be effected by ambiguous forms where the syntactic features are syntactically imposed by rules of agreement or government, but not where they are intrinsic. It is further suggested that there may be a gradiant constraint on the number of conflicting features which can be resolved by an ambiguous form. Most examples in the text are from English, German and Xhosa.

Ringe, Donald. 1986. Nominative-accusative syncretism and syntactic case. Penn Working Papers in Linguistics 2. 45-81

Instances of nominative-accusative syncretism in the plural in Ancient Greek and West Germanic are discussed which cannot be reduced to the operation of phonological processes. An analysis is proposed in terms of Distributed Morphology. Nominative and accusative are marked [+Case] and [+Case, +Accusative], respectively, while the remaining, syntactic, cases have the additional feature [+Oblique]. Underspecification of the feature [+Accusative] thus results in syncretism with the nominative only.

Rix, Helmut. 1987. Morphologische Konsequenzen des Synkretismus. In: Werner Bahner, J. Schilot and D. Viehweger (eds.) Proceedings of the Fourteenth International Congress of Linguists. Berlin: Akademie Verlag. 1437-1441.

The problem of diachronic syncretism of the ‘local’ cases (instrumental, locative, ablative) in Indo-European is addressed from the perspective of Natural Morphology. The loss of a functional distinction preceded the loss of formal distinction, so that there was a period where there must have been functionally unmotivated formal variation. This was ultimately resolved in favor of a single form – though different forms prevailed in different languages. The choice was presumably based on (system-specific) principles of iconicity. Examples from Latin, Greek and Old Irish.

Seiler, H. 1967. On paradigmatic and syntagmatic similarity. Lingua 18. 35-79.

Section 5 follows Pike’s (1965) analysis of German pronominal declension, and concludes (largely on the comparison of nominative and accusative forms) that "a close similarity expression […] seems to be paralleled by an equally close semanto-syntactic similarity", but this is "not to say that such a parallelism must be there under all circumstances".

Serbat, Guy. 1989. Le syncrétisme des cas. In: G. Calboli (ed.) Subordination and other topics in Latin, ed. Amsterdam: Benjamins. 273-286.

Underlying syntactic or semantic similarity (‘partial synonymy’) is claimed as a necessary precondition for diachronic syncretism. Examples from Latin, Romance and Greek.

Serzisko, Fritz (1982). Numerus/Genus-Kongruenz und das Phänomen der Polarität am Beispiel einiger ostkuschitischer Sprachen. In: Hansjakob Seiler and Franz J. Stachowiak (eds.) Apprehension: Das sprachliche Erfassen von Gegenständen: II: Die Techniken und ihr Zusammenhang in Einzelsprachen, 179-200. Tübingen: Gunter Narr.

An exploration of polarity (the term introduced by Meinhof 1910) in the Cushitic languages. The term is used when a form denoting one gender and number combination is also used to denote the opposite combination, e.g. the definite article in Somali, where kii is used for masculine singular or feminine plural, and tii for feminine singular or masculine plural. Polarity is attributed to the marking of markedness congruence, e.g. Somali kii denotes either the unmarked gender (masc.) in the unmarked number (sg), or the marked gender (fem.) in the marked number, while tii denotes either the marked gender in the unmarked number, or the unmarked gender in the marked number.

Sologub, A.I. 1983. O sinkretizme form v sklonenii sušcestvitel’nyx ženskogo roda edinstvennogo cisla po dialektnym dannym. In: Ruben I. Avanesov (ed.) Russkie narodnye govory. Moscow: Nauka. 82-88.

A survey of the patterns of case syncretism found Russian dialects; most interesting are instances where the form of the syncretic case vacillates, e.g. dialects where the genitive-dative-locative may take either the form of the original genitive (ženy ‘wife’) or of the original dative-locative (žene).

Spencer, Andrew. 2000. Agreement morphology in Chukotkan. In: Wolfgang U. Dressler, Oskar E. Pfeiffer, Markus A. Pöchtrager and John R. Rennison (eds.) Morphological analysis in comparison. Amsterdam and Philadelphia: John Benjamins. 191-222.

In Chukchi and Koryak, transitive verbs with a 1st person object are identical to antipassive forms. This is attributed to a rule of referral whereby the antipassive forms are introduced into the transitive paradigm.

Stump, Gregory T. 1993. On rules of referral. Language 69/3. 449-479.

Rules of referral (Zwicky 1985) are adduced to account for systematic inflectional homonymy (underspecification yields syncretism by the elsewhere condition, whereby identities are merely accidental). Rules of referral may apply broadly (to whole word forms) or narrowly (to individual elements of the word form, e.g. stems, affixes or prosody); the latter captures more generalizations. Bidirectional referral is achieved by an inversion principle: if a rule of referral applies to a subset of forms within a given domain, the inverse applies to the remaining forms within the domain. Examples from Romany, Macedonian, Sanskrit, Old Icelandic and Russian; formal analysis in terms of Stump’s Paradigm Function Morphology.

Stump, Gregory T. 2001. Inflectional morphology: a theory of paradigm structure. Cambridge: Cambridge University Press.

The book outlines the principles of Paradigm Function Morphology. Chapter 7 (pp. 212-241) is on syncretism. Syncretism is divided into four types:

(a) unstipulated syncretism (underspecification)

(b) unidirectional syncretism (a referral from one form to another)

(c) bidirectional syncretism (a referral which goes in one direction for one set of words, and in the opposite direction for another)

(d) symmetrical syncretism (a stipulated identity which has no identifiable source)

A constraint on syncretism, the Feature Ranking Principle, is proposed: in cases of stipulated syncretism, there is a hierarchy of feature dominance; thus if a value of feature ‘b’ is syncretised in the context of a value of feature ‘a’, then you will never find a value of feature ‘a’ being syncretized in the presence of a value of feature ‘b’; apparent counterexamples are due to the presence of unstipulated syncretisms in the paradigm, which are not subject to this constraint. Examples from Romanian, Bulgarian, Sanskrit and Russian.

Trnka, Bohumil. 1958/1992. On some problems of neutralisation; On morphemic homonymy. 1963/1992. A few remarks on homonymy and neutralisation. 1974/1992. All in Bohumil Trnka. 1992. Selected Papers in general linguistics. The Hague: Mouton. 149-155. 336-339. 356-362.

Trnka contrasts morphemic homophony (where an underlying morphemic opposition fails to be realized) with neutralization (where the morphemic opposition is not present). The former is attributed ultimately to phonlogical change, the latter to morphological change. Examples chiefly from English, German and Latin.

Wheeler, Max. 1993. On the hierarchy of naturalness principles in inflectional morphology. Journal of Linguistics 29. 95-110.

A response to the hierarchy of naturalness principles in inflection (in the framework of Natural Morphology) given in Wurzel (1987 in Dressler Leitmotifs…), in which the Principle of Uniformity and Transparency (i.e. ‘one form –one meaning’) is said to dominate the Principle of Constructional Iconicity (i.e. addition of marked feature is reflected by the addition of morphological material). This ranking suggests we should find diachronic examples of countericonic patterns being extended for the sake of greater transparency. But the history of present tense indicative and subjunctive forms in Central Romance shows mood syncretism being extended for the sake of greater constructional iconicity, in violation of the hierarchy. It is proposed that a Principle of Avoidance of Countericonicity be split off from the Principle of Constructional Iconicity; this would dominate the Principle of Uniformity and Transparency, and this would be ranked alongside a new Principle of Markedness in Syncretism, which favors neutralization in more marked contexts. Examples also from the history of Slavic.

Wiese, Bernd. 1996. Iconicity and Syncretism. In: Robin Sackmann (ed.) Theoretical Linguistics and Grammatical Description: papers in honour of Hans-Heinrich Lieb on the occasion of his 60th birthday. Amsterdam: John Benjamins. 323-344.

Syncretism in the paradigm of the German demonstrative adjective dieser is treated as neutralization (underspecification), on the assumption any identity of form is iconic. Seemingly problematic are instances of ‘reciprocal dominance’ (Hjelmslev 1943), e.g. where case dominates number in one part of the paradigm and number dominates case in another. This is solved by introducing 4 hierarchically related features (±Oblique > ±Objective > ±Standard > ±Special) to describe the 4 cases (nominative, accusative, genitive, dative), whereby the syncretic forms can be described as resulting from unidirectional dominance.

Williams, Edwin. 1994. Remarks on lexical knowledge. Lingua 92. 7-34.

In the context of a study of abstract structures in the lexicon, it is proposed that the cells of an inflectional paradigm are structured as the branches of a tree. The terminal nodes represent the maximal paradigm, while syncretism results from forms being inserted at higher levels. Thus the form of the tree constrains possible syncretisms. It is claimed that, within a set of related paradigms, there will always be at least one paradigm in which all the possible formal distinctions that can be made are in fact made.

Wunderlich, Dieter. 1996. Minimalist morphology: the role of paradigms. In Geert Booij and Jaap van Marle (eds.) Yearbook of Morphology 1995. 93-114.

In the context of an exposition of Minimalist Morphology’s approach to paradigms, rules of referral are rejected, illustrated through a revision of Stump’s (1993) analysis of Macedonian verbs. In Stump (1993) the conflict between specifying certain past tense markers as non-3rd person and the presence of 2-3 person syncretism in singular past forms is accounted for by making the 2nd person refer to the 3rd person for its form. Here, the past tense marker is specified disjunctively (1st person + 2nd plural) so that the 2-3sg syncretism results from underspecification of these forms.

Wunderlich, Dieter and Ray Fabri. 1995. Minimalist morphology: an approach to inflection. Zeitschrift für Sprachwissenschaft 14/2. 236-294.

It is proposed that inflectional information is carried by affixes themselves, characterized as the marked values of (non-abstract) features. Inflection is mediated by paradigms, whose unmarked cells are filled in by the default values of features. Syncretism results from underspecification; since features are arranged hierarchically, underspecification can take place at various levels. Identity of affixes which cannot be reduced to underspecification is assumed to be accidental homophony. Examples from German and Arabic verbs.

Zaenen, Annie and Lauri Karttunen. 1984. Morphological non-distinctness and coordination. Proceedings of the first Eastern States Conference on Linguistics. 309-320.

The ability of syncretic forms to resolve syntactic feature conflicts is discussed, particularly with respect to noun forms. However, feature conflict resolution is not possible in all instances where we might expect it. To explain this a semantic constraint is proposed, namely the Anti-Pun Ordinance, whereby a word cannot be used in two different senses simultaneously. Data primarily from Finnish.

Zagórski, Zygmunt. 1999. Kilka uwag o homonimii deklinacyjnej w polszczyznie XVI wieku. Rzeczowniki zenskie i nijakie [Some remarks on declensional homonymy in XVI century Polish. Feminine and neuter nouns]. Prace komisjii jezykoznawczej 30. 81-88.

A detailed numerical survey of the number of distinct forms in the Polish noun paradigm. It is shown that, over time, in spite of a decrease in the number of morphosyntactic slots (due e.g. to the loss of the dual), the ratio of distinct forms to morphosyntactic slots within any one inflectional class remains similar.

Zwicky, Arnold. 1977. Hierarchies of person. In: Papers from the thirteenth regional Meeting. Chicago Linguistic Society. 714-733.

It is proposed that the morphosyntactic representation of person should be maximally expressed as +I, +I+II, +I-II, +II and +III, corresponding to the referential persons 1st, 1st inclusive, 1st exclusive, 2nd and 3rd. An alternative representation in terms of the features ±speaker and ±addressee is considered. Although it would predict certain syncretic patterns, it also predicts a non-occuring pattern (1st exclusive=3rd, opposed to 1st inclusive and 2nd), and fails to account for syncretism of 1st and 3rd.

Zwicky, Arnold. 1985. How to describe inflection. In: Proceedings of the eleventh annual meeting of the Berkeley Linguistics Society. 372-386.

Three devices for describing syncretism are described:

(a) underspecification.

(b) disjunctive clusters of feature values.

German is used to illustrate these devices.

Zwicky, Arnold. 1991. Systematic versus accidental phonological identity. In: Frans Plank (ed.) Paradigms: The Economy of Inflection. Berlin: Mouton de Gruyter. 113-132.

Identity of forms is taken to be the default state, whereby overt inflectional morphology constitutes a scheme of principled exceptions. Within a morphological system containing distinct forms, identities may still occur because of:

(a) the non-application of rules, resulting from: (i) the absence of an applicable rule; e.g. failure of modal verbs in English to distinguish person); or (ii) the blocking of an otherwise applicable rule; e.g. the constraint against the affixation of possessive –s on plurals in –s in English.

(b) the asymmetry of form sets; e.g. the different range of verb forms in past and present in English).

(d) listing; e.g., a particular identity can be simply stipulated by rule.

(Stem alternation may also be analyzed in these terms.) The distinction between systematic and accidental identity plays a role in resolving syntactic feature conflicts. On the other hand, some constraints refer to phonological form alone, regardless of the nature of this identity (e.g. Spanish rules which address clitics of the form se). Examples also from German and Pashto.

Project members

Prof Greville G. Corbett
Dr Dunstan Brown
Dr Matthew Corbett

Period of award:

September 1999 - August 2002

Funder

Economic and Social Research Council (ESRC)

TOP