Cagliari, Italy  15-18 September 2003

ALT V Conference

Michael Daniel

Towards a Typology of Personal Locatives. Problem setting.
daniel@qub.com

In what follows, personal locative is the name of a nominal morphological form or construction conveying the meaning ‘at (to, from) X’s place of habitation’; see example in (1). Until now, crosslinguistic discussion concerning personal locatives mainly focused on their grammaticalization in Romance, e.g. [Longobardi 2001]. The present paper (I) investigates the semantic structure of the category, (II) discusses its typological correlation with other spatial or abstract categories, and (III) presents a survey of the formal ways of its linguistic expression.

The approach adopted in the paper views spatial categories as consisting of two (roughly) independent components, localization with respect to the spatial reference point and the direction of movement (if any) of the object with respect to the localization area [Svorou 1993, Kibrik 2002, Comrie&Polinsky 1998]. E.g., in from under the table, the table is the reference point, under expresses the localization and from indicates the direction of the movement. Examples of localizations are IN (inside the reference point), APUD (near the reference point), CONT (in contact with the reference point), etc. Direction categories include Essive (the object is at rest in the localization area), Lative (the object enters the area), Translative (the object crosses the area), Elative (the object leaves the area) and some other.

(I) Personal locatives (Personal Essive / Lative / Elative) apparently constitute a distinct localization. The spatial area where X lives is not ‘spatially identical’ to any of the spatial areas described by other localizations - IN, APUD etc; X does not live inside or near him/herself. This localization area is not defined in terms of the actual position of the human reference point, but rather in terms of possession or similar associative relation linking the area to the human reference point. With personal locatives, and unlike other localizations, the reference point is not even necessarily within the localization area at the moment to which the speech act refers; see (2). Personal locative may be called abstract localization (as opposed to spatial localizations like IN or APUD). This abstract flavor is clear e.g. in Agul where the Personal Lative expressed by a case is possible for ‘I brought the baby to their place’, but not for ‘I brought the money to their place’, see (3b); apparently because a non-human object may not enter someone’s personal sphere in the same sense as a person may. Non-spatial character of personal locatives is further evidenced by the fact that they do not normally combine with movements other than Essive, Lative and Elative; apparently because the Translative, for example, is too spatial for Personal Translative (‘through at Peter’s’) to exist.

(II) Probably, it is the non-spatial component of the personal locative meaning that caused or supported the emergence of new usages further away from spatial semantics in some languages. In Russian, Personal Essive preposition u (originally spatial APUD but now gradually replaced in this meaning by other spatial words) is widely used in possessive predications; see (4); similar usage occurs e.g. in Tatar and Selkup. In Polish, the displaced usage of Personal Essive is to denote part-whole relations (also attested in Russian and Selkup). In Bagvalal, Personal Essives are used with locatives of body parts to form a kind of apparently external possessor construction; see (5). French and Swedish Personal Essives have abstract ‘personal sphere’ uses; and Swedish has genitive-like use.

(III) Formally, personal locative may be conveyed by (a) a generalized locative form (Tatar locative, lative, elative), (b) specialized locative form (Akhvakh adessive, adlative, adelative, see (6)), or by (c) a means which is not used for other spatial meanings (French chez). In (a) no difference is made between personal locatives and other spatial meanings. For (b), an interesting question arises as to which spatial meaning is conceptually closer to personal locatives and is used to express it. APUD seems to be a widespread choice, as in Latin (apud), Slavic (u), Akhvakh and some other. Dedicated means in (c) is typically an adposition (Swedish, see (7)), although at least in one language a dedicated case form is attested (Bagvalal, see (5)). The fourth formal option (d) occurs e.g. in Limbu, see (8); English; and probably Georgian – a locative of a headless possessive, as at Peter’s; this, again, points at an affinity between personal locatives and possessives. The fifth, syntactic, pattern (e) which is quite common is an attributive construction with a more or less desemanticized head, as in Agul, see (3a) or English (at my place). Here is where French and Swedish personal locatives belong etymologically (originating from ‘house’ in respective proto-languages). Some of the languages which today use syntactic personal locative constructions probably present earlier stages of a grammaticalization process similar to that which happened in Romance.

In some languages, different personal locative directions are expressed by non-homogeneous means. In Selkup, the Personal Lative is expressed by the dative-lative case, while Personal Essive and Elative, by postpositions (even though appropriate case forms, locative and ablative do exist in the language); see (9). Similarly in Agul, the only case form available for personal locative is the Lative; other personal locatives are expressed by syntactic constructions. In Russian and Swedish, it is the Personal Essive that has dedicated prepositions, while the Personal Lative and Elative are expressed by spatial postpositions with a general locative semantics. Notably, we know of no language which, of the three directions, uses a relatively more grammatical or dedicated way to convey the Personal Elative, which suggests that the Personal Essive and Lative are ‘cognitively’ more relevant than the Elative.

Language Data.

(1) English. Raymond Carver, fiction.

It was when I was with Pete Jensen and we were on our bicycles and we stopped at Dummy's to get a glass of water.

(2) French. Constructed example.

Il n’est pas chez lui.

He is not at home. (=He is not where he lives)

(3) Agul (Daghestanian). Solmaz Merdanova, p.c.

(4) Russian (folktale)

 (5) Bagvalal (Daghestanian). Kibrik A.E. et al. (eds.) Bagvalinskij jazyk. Nasledije. Moskva, 2000. Page 225.

(6) Akhvakh (Daghestanian). Magomedbekova, Z.M. Akhvakhskij jazyk. Tbilisi, 1967. p. 137 (transcription simplified, glossing unavailable)

 (7) Swedish (a title of a poem by Lukas Moodysson)

(8) Limbu (Tibeto-Burman). van Driem, George. A Grammar of Limbu. Mouton de Gruyter. 1987. p. 49.

(9) Selkup (Uralic) (glossing not available). Kuznetsova, Ariadna & Je.A. Kjelimski & Je.V. Grushkina. Ocherki po selkupskomu jazyku. MGU. Moscow, 1980. pp. 178, 181.

(a) imap taraš qontyptä:qynty qoniššy qälytytkinty [Nenets-PL-DAT/ALL]

‘In the time before he married he used to go to Nenets [name of a neighboring Samoyedic people]’

References:

Comrie, Bernard and Maria Polinsky. The great Daghestanian case hoax. In: Siewerska, Anna and Jae Jung Song (eds.) Case, Typology and Grammar. In honor of Barry J. Blake. Benjamins. Amsterdam, 1998.

Kibrik A.E. Nominal inflection galore... In: Plank F. (ed.) Noun Phrase in the Languages of Europe. Mouton, 2002.

Longobardi, Giuseppe. Formal Syntax, Diachronic Minimalism, and Etymology: The History of French Chez. Linguistic Inquiry, Volume 32, Number 2, Spring 2001.

Svorou, Soteria. The grammar of space. Benjamins. Amsterdam, 1993.