INVESTIGADORES
GOLLUSCIO lucia Angela
congresos y reuniones científicas
Título:
Dyad constructions in Mapudungun and their typological implications
Autor/es:
EVANS, NICHOLAS; GOLLUSCIO, LUCÍA; MELLICO, FRESIA
Lugar:
París
Reunión:
Congreso; VII Congreso de la Association for Linguistic Typology; 2007
Institución organizadora:
Association for Linguistic Typology
Resumen:
Dyad constructions provide a way of referring to pairs or groups of people based on shared social relationships, e.g. ‘(pair of) brothers / neighbours’, ‘father and son’. They may be formed bymorphological derivation (1), or may be unanalyseable lexical roots (2). Until recently the phenomenon has escaped typological systematization (Evans 2006), with cross-linguistic comparison made difficult by highly variable terminology, conflation with related categories (e.g. reciprocals, duals, associative duals/plurals, family group classifiers, additive co-compounds), or by a failure to discuss the phenomenon carefully in descriptive grammars. Data published so far suggests a highly skewed distribution of the phenomenon worldwide, with the majority of reported occurrences in the Western Pacific (see Map), and only sporadic instances in Africa, Western Eurasia or the Americas. Is this a real gap or does it conceal failures to detect a phenomenon which has only recently come to linguists’ attention? The first purpose of this paper, then, is to show, on the basis of a single case-study from one part of the world where dyad constructions have not been identified in previous typological surveys – Mapudungun (Mapuche, aka Araucanian)1 from the South American cone – there is a productive and entrenched dyad construction, so that in at least some areas the apparent geographical skewing reflects absence of evidence (or its limited dissemination or terminological calibration) rather than evidence of absence. Dyad terms are formed in Mapudungun by suffixing a dyad morpheme -wen to nominal roots (3). Though perhaps ultimately related to the reciprocal/reflexive marker -w(ü) and in particular its infinitive form -wïn, the dyad suffix -wen is distinct from it formally, semantically and combinatorically, since -wen forms nominals from nominals while -w(ü) derives one-place verbs from two-place verbs. The second purpose of this paper is to examine in detail several typological features of the Mapudungún dyad construction, based on new field data, which either deviate from what has been reported for dyad constructions in other languages, or exhibit interesting semantic effects which have not been investigated elsewhere: (a) the possibility of forming reciprocalized causative denominal verbs from dyads (4) (b) the possibility – previously not reported for any dyad construction – of forming possessive constructions from dyads (5) (c) the ambiguous interaction of numerals with dyad expressions, which quantify either over the cardinality of a single dyad group (e.g. group of three, comprising a mother and daughters) or multiply the cardinality of dyad groups taken one at a time (e.g. three mother-daughter groups) (d) a further unreported for dyad constructions, which is to derive, from a nominal root denoting an exchanged gift object or experience, a dyad term for the pair of people bonded by sharing that experience e.g. koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways.1 from the South American cone – there is a productive and entrenched dyad construction, so that in at least some areas the apparent geographical skewing reflects absence of evidence (or its limited dissemination or terminological calibration) rather than evidence of absence. Dyad terms are formed in Mapudungun by suffixing a dyad morpheme -wen to nominal roots (3). Though perhaps ultimately related to the reciprocal/reflexive marker -w(ü) and in particular its infinitive form -wïn, the dyad suffix -wen is distinct from it formally, semantically and combinatorically, since -wen forms nominals from nominals while -w(ü) derives one-place verbs from two-place verbs. The second purpose of this paper is to examine in detail several typological features of the Mapudungún dyad construction, based on new field data, which either deviate from what has been reported for dyad constructions in other languages, or exhibit interesting semantic effects which have not been investigated elsewhere: (a) the possibility of forming reciprocalized causative denominal verbs from dyads (4) (b) the possibility – previously not reported for any dyad construction – of forming possessive constructions from dyads (5) (c) the ambiguous interaction of numerals with dyad expressions, which quantify either over the cardinality of a single dyad group (e.g. group of three, comprising a mother and daughters) or multiply the cardinality of dyad groups taken one at a time (e.g. three mother-daughter groups) (d) a further unreported for dyad constructions, which is to derive, from a nominal root denoting an exchanged gift object or experience, a dyad term for the pair of people bonded by sharing that experience e.g. koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways.-wen to nominal roots (3). Though perhaps ultimately related to the reciprocal/reflexive marker -w(ü) and in particular its infinitive form -wïn, the dyad suffix -wen is distinct from it formally, semantically and combinatorically, since -wen forms nominals from nominals while -w(ü) derives one-place verbs from two-place verbs. The second purpose of this paper is to examine in detail several typological features of the Mapudungún dyad construction, based on new field data, which either deviate from what has been reported for dyad constructions in other languages, or exhibit interesting semantic effects which have not been investigated elsewhere: (a) the possibility of forming reciprocalized causative denominal verbs from dyads (4) (b) the possibility – previously not reported for any dyad construction – of forming possessive constructions from dyads (5) (c) the ambiguous interaction of numerals with dyad expressions, which quantify either over the cardinality of a single dyad group (e.g. group of three, comprising a mother and daughters) or multiply the cardinality of dyad groups taken one at a time (e.g. three mother-daughter groups) (d) a further unreported for dyad constructions, which is to derive, from a nominal root denoting an exchanged gift object or experience, a dyad term for the pair of people bonded by sharing that experience e.g. koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways.-w(ü) and in particular its infinitive form -wïn, the dyad suffix -wen is distinct from it formally, semantically and combinatorically, since -wen forms nominals from nominals while -w(ü) derives one-place verbs from two-place verbs. The second purpose of this paper is to examine in detail several typological features of the Mapudungún dyad construction, based on new field data, which either deviate from what has been reported for dyad constructions in other languages, or exhibit interesting semantic effects which have not been investigated elsewhere: (a) the possibility of forming reciprocalized causative denominal verbs from dyads (4) (b) the possibility – previously not reported for any dyad construction – of forming possessive constructions from dyads (5) (c) the ambiguous interaction of numerals with dyad expressions, which quantify either over the cardinality of a single dyad group (e.g. group of three, comprising a mother and daughters) or multiply the cardinality of dyad groups taken one at a time (e.g. three mother-daughter groups) (d) a further unreported for dyad constructions, which is to derive, from a nominal root denoting an exchanged gift object or experience, a dyad term for the pair of people bonded by sharing that experience e.g. koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways.-wen forms nominals from nominals while -w(ü) derives one-place verbs from two-place verbs. The second purpose of this paper is to examine in detail several typological features of the Mapudungún dyad construction, based on new field data, which either deviate from what has been reported for dyad constructions in other languages, or exhibit interesting semantic effects which have not been investigated elsewhere: (a) the possibility of forming reciprocalized causative denominal verbs from dyads (4) (b) the possibility – previously not reported for any dyad construction – of forming possessive constructions from dyads (5) (c) the ambiguous interaction of numerals with dyad expressions, which quantify either over the cardinality of a single dyad group (e.g. group of three, comprising a mother and daughters) or multiply the cardinality of dyad groups taken one at a time (e.g. three mother-daughter groups) (d) a further unreported for dyad constructions, which is to derive, from a nominal root denoting an exchanged gift object or experience, a dyad term for the pair of people bonded by sharing that experience e.g. koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways.koncho-wen ‘pair who have given each other an animal to eat’. That such a deeply-entrenched and productive construction can escape previous typological integration even for such a relatively well-described language as Mapudungun suggests that the phenomenon is likely to be much more widespread than has hitherto been suspected. And, following the ‘typological snowballing’ effect where new languages displaying a phenomenon can throw up questions which need to be added to our global typology (including checking back over previously reported languages) Mapudungun extends our typology of how dyad constructions behave, semantically and syntactically, in several new ways. 1 Resumé of previous mentions is on the data page.Resumé of previous mentions is on the data page. Page Two: Data, References, map Data (1) Kayardild (Tangkic, Australian): ngamathu ‘mother’; ngamathu-ngarrba ‘mother and child’ (2) Miyan (Trans New Guinea) lum ‘father and child’. (3) kure ‘wife’, kurewen ‘husband/wife pair’; wenüy ‘friend’, wenüywen ‘(pair of) friends’. (4) peñi-wen-ye-w-i-ngu [brother-DYAD-VBLZR-RR-IND-3duSUB] ‘they made themselves each other’s brothers’ (5) ta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughterngamathu ‘mother’; ngamathu-ngarrba ‘mother and child’ (2) Miyan (Trans New Guinea) lum ‘father and child’. (3) kure ‘wife’, kurewen ‘husband/wife pair’; wenüy ‘friend’, wenüywen ‘(pair of) friends’. (4) peñi-wen-ye-w-i-ngu [brother-DYAD-VBLZR-RR-IND-3duSUB] ‘they made themselves each other’s brothers’ (5) ta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughterlum ‘father and child’. (3) kure ‘wife’, kurewen ‘husband/wife pair’; wenüy ‘friend’, wenüywen ‘(pair of) friends’. (4) peñi-wen-ye-w-i-ngu [brother-DYAD-VBLZR-RR-IND-3duSUB] ‘they made themselves each other’s brothers’ (5) ta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughterkure ‘wife’, kurewen ‘husband/wife pair’; wenüy ‘friend’, wenüywen ‘(pair of) friends’. (4) peñi-wen-ye-w-i-ngu [brother-DYAD-VBLZR-RR-IND-3duSUB] ‘they made themselves each other’s brothers’ (5) ta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughterpeñi-wen-ye-w-i-ngu [brother-DYAD-VBLZR-RR-IND-3duSUB] ‘they made themselves each other’s brothers’ (5) ta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughterta ñi epu koñi-wen ‘my mother-daughter pair’, i.e. the pair of people comprising my wife and our daughter References Augusta, Félix José de. 1903. Gramática araucana. Valdivia: Imprenta Central L. Lampert.Gramática araucana. Valdivia: Imprenta Central L. Lampert. Evans, Nicholas. 2003. An interesting couple: the semantic development of dyad morphemes. Institut für Sprachwissenschaft, Universität zu Köln. Arbeitspapier (Neue Folge). Pp. 105.An interesting couple: the semantic development of dyad morphemes. Institut für Sprachwissenschaft, Universität zu Köln. Arbeitspapier (Neue Folge). Pp. 105. Evans, Nicholas. 2006. Dyad constructions. In Keith Brown (ed.), Encyclopaedia of Language and Linguistics, Vol. 4. Oxford: Elsevier. pp. 24-27.Encyclopaedia of Language and Linguistics, Vol. 4. Oxford: Elsevier. pp. 24-27. Golbert, Perla. 1975. Epu Peñiwen (Los dos hermanos). Cuento tradicional araucano. Transcripcion fonologica, traduccion y analisis. Buenos Aires: Centro de Investigaciones en Ciencias de la Educacion. Salas, Adalberto. 1992. El mapuche o araucano. Fonologia, gramática y antologia. Madrid: Mapfre. Smeets, Catharina I. 1989. A Mapuche Grammar. Ph. D. dissertation. University of Leiden.Epu Peñiwen (Los dos hermanos). Cuento tradicional araucano. Transcripcion fonologica, traduccion y analisis. Buenos Aires: Centro de Investigaciones en Ciencias de la Educacion. Salas, Adalberto. 1992. El mapuche o araucano. Fonologia, gramática y antologia. Madrid: Mapfre. Smeets, Catharina I. 1989. A Mapuche Grammar. Ph. D. dissertation. University of Leiden.. El mapuche o araucano. Fonologia, gramática y antologia. Madrid: Mapfre. Smeets, Catharina I. 1989. A Mapuche Grammar. Ph. D. dissertation. University of Leiden.A Mapuche Grammar. Ph. D. dissertation. University of Leiden. Previous mentions in the Mapudungunist/Araucanist literature: Augusta (1903): discussion of use of -wen as nominal suffix, plus enumeration of different types of relationships under the rubric of degrees of friendship (los grados de amistad) Golbert (1975) mention of -wen as marking ‘relación compartida’ (shared relationship) Smeets (1989:136-6) labels -wen a ‘relative marker’ referring to ‘a group of two or more persons who are related to each other through a relation which is indicated by a noun’los grados de amistad) Golbert (1975) mention of -wen as marking ‘relación compartida’ (shared relationship) Smeets (1989:136-6) labels -wen a ‘relative marker’ referring to ‘a group of two or more persons who are related to each other through a relation which is indicated by a noun’-wen as marking ‘relación compartida’ (shared relationship) Smeets (1989:136-6) labels -wen a ‘relative marker’ referring to ‘a group of two or more persons who are related to each other through a relation which is indicated by a noun’ Map: Global distribution of Dyad constructions according to the data in Evans (2003, 2006)Global distribution of Dyad constructions according to the data in Evans (2003, 2006) QuickTime™ and a TIFF (LZW) decompressor are needed to see this picture. Page Three. Affiliations of authors Nicholas Evans. School of Languages and Linguistics, University of Melbourne, Parkville, Victoria 3010 Lucía Golluscio. Universidad de Buenos Aires / Consejo Nacional de Investigaciones Scientíficas y Técnicas Fresia Mellico, Universidad de Buenos Aires