Acta Linguistica Lithuanica

LIV (2006), 95 – 100

A Remark on Gender

BOHUMIL VYKYPĚL

Institute for the Czech Language of the Czech Academy of Sciences

The present paper deals with the question of whether the category of substantival gender is a grammatical or a lexical category. The following criterion is suggested: if there is at least one declension paradigm ambivalent in gender, then this category is lexical; in the contrary case it is grammatical. Examples are taken from Lithuanian, Latvian, Russian, Old and Modern Czech. In conclusion, the typological implications of such a conception of gender are briefly mentioned.

Within the framework of the so-called Prague typology, the question arises whether gender should be regarded as a grammatical content category and thus as typologically relevant, or more precisely, as relevant to the description of the typological dominant of a language. Vladimír Skalička, the founder of Prague typology, seems to deny gender the status of a grammatical category in nouns. He thus distinguishes between a declension class and gender:

“The ideal state would be one in which genders and declensions merged completely. However, not a single Indo-European language has progressed that far.” (Skalička 1979: 136 = 1941: 12)

Skalička had already asserted earlier:

“In other cases a morpheme is two-semantic, e.g. gen. sg. vir/i expresses 1. the genitive, 2. the singular.” (Skalička 1935: 15 = 1979: 68)

Im Tschechischen postuliert Skalička (1935: 52f. = 1979: 107f.) für das Adjektiv drei grammatische Seme (Inhaltselemente) – Kasus, Numerus und Genus, für das Substantiv dagegen nur zwei – Kasus und Numerus. Er führt allerdings in Klammern die Zahl drei mit Fragezeichen auch für das Substantiv an und bezieht auch das Genus der Substantive in die

95

Bohumil Vykypěl

grammatical semes. These certain fluctuations are also reflected in the fact that Skalička, in a general passage, readily treats gender as a grammatical category (cf. Skalička 1935: 33f. = 1979: 87f.).

Ľudovít Novák is more explicit in his review of Skalička (1935):

“Concretely: in the ending -a in chlap|a only the semes of the genitive and singular are located, and in žen|a only the semes of the nominative and singular, whereas the seme of masculine or feminine gender is located rather in the roots chlap- and žen-. [Concretely: in the ending -a in chlap|a only the semes of the genitive and singular are located, and in žen|a only the semes of the nominative and singular, whereas the seme of masculine or feminine gender is located rather in the roots chlap- and žen-.]” (Novák 1936: 10)

Unter Berufung auf Skalička und Novák sowie auf Zaliznjak (1967: 66, 74) hat Markus Giger (1998: 173, Anm. 1) folglich das Genus als eigenständige grammatische Kategorie lediglich bei den Adjektiven anerkannt, während er es bei den Substantiven als dem lexikalischen Teil des Wortes inhärent betrachtet, wobei es nur durch die Kongruenz eines determinierenden Attributs ausgedrückt wird. Diese Lösung haben wir dann auch in unserem gemeinsamen Artikel zur Typologie des Litauischen und Lettischen geltend gemacht (Giger–Vykypěl 2001: 47).

Ultimately, this solution has probably been accepted by the majority. In any case, it is also found among authors outside the framework of Prague typology. Besides the already-mentioned A. A. Zaliznjak (1967: 62ff.), who regards gender in Russian nouns as a “concord class” (see also Corbett 1991: 147ff.), Christian Lehmann (1982: 204), for example, in his general treatise on the concept of concord, cites gender concord as an example of the kind of concord in which the relevant category is not explicitly marked on the determined element of the concord syntagma (i.e. actually on the noun) (similarly Kaznelson 1974: 29ff.). Nevertheless, matters seem to be somewhat more complicated.

First of all, it is not only individual endings (or, more generally, grammatical markers) that must be taken into account as “candidates” for expressing the elements of the gender category (as Greenberg 1978:

96

Eine Bemerkung zum Genus

53 und nach ihm Heine 1982, 193 bei der Unterscheidung zwischen dem „overten“ und dem „coverten“ Ausdruck des Genus machen), sondern man muss die ganzen Paradigmen betrachten. Das Kriterium für die Entscheidung, ob das Genus am Substantiv (am determinierten Glied des Kongruenzsyntagmas) explizit bezeichnet wird, stellt folglich die Tatsache dar, ob es in einem Sprachsystem wenigstens ein Deklinationsparadigma gibt, das nicht zu einem bestimmten Glied der Genuskategorie gehört, und nicht die Tatsache, ob jedes Glied der Genuskategorie nur eine Deklination besitzt, wie das Skalička vorauszusetzen scheint (ähnlich Corbett 1991: 62). Ist diese Bedingung erfüllt, d.h. gibt es in einer Sprache wenigstens ein Deklinationsparadigma des Substantivs, das mehrere Glieder der Genuskategorie ausdrücken kann, ist das Genus keine grammatische, sondern eine lexikalische Kategorie des Substantivs. Im umgekehrten Fall ist das Genus als eine grammatische Kategorie des Substantivs zu betrachten.

An instructive example of this is provided by the comparison between Lithuanian and Latvian. In Lithuanian, the nominal declension paradigms of the so-called o- and ē stems are ambivalent with respect to gender (the nouns belonging to these paradigms can agree with both feminine and masculine attributes). By contrast, the (historically) corresponding Latvian a- and e declensions, as well as the i declension, are distinguished by gender: this distinction concerns only the singular dative ending (feminine -ai, -ei and -ij vs. masculine -am, -em and -im), but it nevertheless exists. This means, then, that one need not posit a grammatical gender category for nouns in Lithuanian, whereas such a category must be assumed in Latvian, since there is no nominal declension paradigm here that is ambivalent with respect to gender.

Russian can be compared with Czech in a similar way: in Russian, the declension paradigms of the so-called a- and ja stems are ambivalent with respect to gender (they can express both masculine and feminine), whereas their Czech counterparts divide into feminine and masculine declensions which, although they have some endings in common in certain cases, are distinguished as wholes. It should be noted that the Russian situation corresponds to that in Old Czech, in which the a- and ja stems were likewise

97

BOHUMIL VYKY̌PEL

ambivalent with respect to gender. Thus, we can say that, unlike in New Czech, there is no grammatical gender category for nouns in Russian and Old Czech. As far as Russian is concerned, this finding also agrees with the view of Zaliznjak mentioned above.

What is actually peculiar about the content category of gender, and what probably chiefly prevents linguists from recognizing this category as a normal grammatical content category, is, in my view, something else. It is the fact that nouns are not declined for gender (or that traditional grammar does not decline nouns for gender…): a noun can occur in x cases or numbers (which may also be expressed by means other than endings), but it has only one gender (or, it is assumed that it has only one gender). Nevertheless, opposing cases can also be pointed out, such as the case of so-called morphological conversion. For example, one can claim that in Latvian there is a noun latvietis ‘Latvian’ that can be inflected for case, number, and gender: nominative singular masculine latvietis, genitive singular masculine latvieša etc., nominative plural masculine latvieši, genitive plural masculine latviešu etc., nominative singular feminine latviete, genitive singular feminine latvietes etc., nominative plural feminine latvietes, genitive plural feminine latviešu etc¹. Besides these ordinary cases of morphological conversion, one can also cite “updated” cases: for example, Czech feminine nouns such as babička ‘grandmother’ or Czech neuter nouns such as kotě ‘kitten’ can also be declined “masculine,” as can be observed in various highly expressive, strongly marked utterances (for examples from African languages, see Heine 1982, 198; generally, see also Corbett 1991: 67, 181f.).

Nevertheless, the members of the gender category are usually wholly or partly syncretized. Thus, in Latin, besides the masculine equus ‘horse’ there is also the feminine equa, but hardly a neuter equum,

¹ Corbett (1991: 67) behauptet über ein entsprechendes Paar, nämlich spanisch hijo ‘Sohn’ und hija ‘Tochter’: „Here, quite clearly, we have two separate nouns, sharing a similar stem but with different inflections.“ Warum dies „quite clearly“ ist, verrät er jedoch nicht.

98

Eine Bemerkung zum Genus

if one disregards a possible actualization. Likewise, for example, a feminine form of the Latvian word koks ‘tree’ is difficult to imagine. One could, however, object that syncretisms also frequently occur in the categories of case and number, and yet these categories are recognized by tradition.

To conclude these brief observations, let us briefly mention the typological implications. If one allows that, for example, in Latvian and New Czech, unlike in Lithuanian and Old Czech, gender constitutes a grammatical category of the noun, this means that one more content element (case, number, and gender) is cumulatively expressed in noun endings in Latvian and New Czech than in Lithuanian and Old Czech (case and number), i.e. that Latvian and New Czech are more inflectional in this respect than Lithuanian and Old Czech. This, however, contradicts the fact that in other areas of declension Lithuanian and Old Czech are, conversely, more inflectional than Latvian and New Czech (cf. Skalička 1941, Giger–Vykypěl 2001). Whether this contradiction is merely one of the cases in which different types and typological tendencies are combined hierarchically in the structure and development of a language, as Prague typology generally assumes, or whether it speaks to the incorrectness of the view of the gender category presented above, the discerning reader may now decide for themselves.

LITERATURVERZEICHNIS

Corbett, G. G. 1991: Gender. Cambridge etc.: Cambridge University Press.

Giger, M. 1998: K typologickému postavení horní lužické srbštiny mezi západoslovanskými jazyky: nominální systém. In: M. Nábelková, Hrsg., Varia VII. Zborník zo VII. kolokvia mladých jazykovedcov (Modra-Piesok 3.–5. 12. 1997). Bratislava: Slovenská jazykovedná spoločnosť pri SAV, 172–185.

Giger, M. – Vykypěl, B. 2001: Die Typologie des Litauischen und des Lettischen (mit einem allgemeinen Ausblick zu den Perspektiven der Prager Typologie). Sborník prací filozofické fakulty brněnské univerzity A 49, 45–86.

99

Bohumil Vykypěl

Greenberg, J. H. 1978: How does a language acquire gender markers? In: J. H. Greenberg, Hrsg., Universals of Human Language. Volume 3: Word Structure. Stanford: Stanford University Press, 47–82.

Heine, B. 1982: African noun class systems. In: H. Seiler & C. Lehmann, Hrsg., Apprehension. Das sprachliche Erfassen von Gegenständen. I. Tübingen: Narr (Language Universals Series, 1/I), 189–216.

Kaznelson, S. D. 1974: Sprachtypologie und Sprachdenken. Berlin: Akademie-Verlag (Sprache und Gesellschaft, 5) & München: Hueber.

Lehmann, C. 1982: Universal and typological aspects of agreement. In: H. Seiler & F. J. Stachowiak, Hrsg., Apprehension. Das sprachliche Erfassen von Gegenständen. II. Tübingen: Narr (Language Universals Series, 1/II), 201–267.

Novák, Ľ. 1936: Základná jednotka gramatického systému a jazyková typologia. Sborník Matice slovenskej 14, 3–14.

Skalička, V. 1935: Zur ungarischen Grammatik. Praha: Filosofická fakulta University Karlovy. (Facultas Philosophica Universitatis Carolinae Pragensis. Práce z vědeckých ústavů, 39.) [Reprint in Skalička 1979, 59–125.]

Skalička, V. 1941: Vývoj české deklinace. Studie typologická. Praha: Jednota českých matematiků a fysiků. (Studie Pražského linguistického kroužku, 4.) [Deutsche Übersetzung: Die Entwicklung der tschechischen Deklination. In: Skalička 1979, 126–172.]

Skalička, V. 1979: Typologische Studien. Hrsg. von P. Hartmann. Braunschweig – Wiesbaden: Vieweg. (Schriften zur Linguistik, 11.)

Zaliznjak, A. A. [Зализняк, А. А.] 1967: Русское именное словоизменение. Москва: Наука.

Bohumil Vykypěl Ústav pro jazyk český AV ČR etymologické oddělení Veveří 97, CZ-60200 Brno [email protected]

100