GB096 - grambank/grambank GitHub Wiki
Are variations in marking strategies of core participants based on verb classes?
Summary
Marking strategies include both marking by indexing and by flagging. The distinction between intransitive and transitive verbs is not included here. Variations in marking strategies include both variations in the alignment patterns and variations in position (e.g. suffixing vs. prefixing) in case of indexes. Mere differences in forms of the indexes triggered by various conjugation classes are not sufficient for 1. For a language to be coded 1 the respective verb classes need to be of a substantial size (several dozen verbs or more) and ideally represent a productive pattern. The presence of small verb classes whose members are often exhaustively listed in a descriptive grammar and are presented as exceptional patterns does not qualify as 1. The patterns of interest are more common with intransitive verbs and include cases of the so-called active-inactive alignment (or simply: active, agentive-patientive or stative-active) in cases of differentially marked S arguments. Some grammars use the terms unergative-unaccusative verbs. The pattern of interest is, however, not restricted to the S argument. Variations in marking of A and P arguments are also included.
Procedure
- Consider the flagging patterns of the language (coded in GB408, GB409, and GB410). Identify whether there is any variation in the flagging of either S, A, or P that is determined by verb classes.
- If there is variation in flagging, consider whether the respective verb classes are of substantial size. If the classes are substantially large, code 1 in the following cases: Code 1 if there are differences in the form of the respective markers which result in different alignment patterns. Code 1 if there is a variation in the position of indexes (prefix/proclitic vs. suffix/clitic), even if the overall alignment pattern remains the same.
- Consider the indexing patterns of the language (coded in GB089–GB094, and GB410). Consider whether there are any differences in the form of the respective indexes. Specifically look for such terms as active(-inactive), agentive-patientive, stative-active alignment or split-S and consider the respective verb classes. Look for cases of non-canonical argument marking and consider the respective verb classes.
- If there is variation in indexing, consider whether the differences in indexing result in different alignment patterns or whether there is a variation in the position of indexes (prefix/proclitic vs. suffix/clitic). If the answer to either of these questions is positive, consider whether the respective verb classes are of substantial size. If the classes are substantially large, code 1.
- Ignore mere variations in the phonological form of indexes conditioned by various conjugation classes.
- Ignore the relatively common situation where some verb classes do not allow any indexing. This is not sufficient for 1.
Examples
Maca (ISO 639-3: wpc, Glottolog: maco1239)
Mako verb roots can be divided into two classes depending on the phonology of the verb: roots ending in a consonant (Class I) and roots ending in a vowel (Class II) (see Rosés Labrada 2015: 294, 297).
Class I: ed- ‘see’, otid- ‘work’, pʰõɾ- ‘put away’, kɨkɨd- ‘dry in the sun’, amat- ‘squeeze’
Class II: hã- ‘make/do’, di- ‘scrape’, wahi- ‘not know’, wi- fell’, wo- ‘stink/smell’
Class I verbs attach the subject (S/A) indexes in a suffix slot immediately following the verb root, as in (a). Class II verbs attach the subject indexes (S/A) in a prefix slot immediately preceding the verb root, as in (b). Maca is coded as 1.
a. Class I:
me-t-obe
fall-1SG-TAM
‘I fall.’ (Rosés Labrada 2015: 296)
b. Class II:
ʧɨ̃-hãmat-obe
1SG-stand.up-TAM
‘I stand up.’ (Rosés Labrada 2015: 295)
Basque (ISO 639-3: eus, Glottolog: basq1248)
Monovalent predicates in Basque are divided into two groups (see Hualde & de Urbana 2003). The so-called unergative predicates mark the S argument with ergative case, as in (a). The so-called unaccusative predicates mark the S argument with the absolutive case, as in (b).
The class of unergative verbs contains quite a few simple verbs, such as dirdiratu ‘glimmer’, botatu ‘bounce’, bozkatu ‘vote’, gogoetatu ‘meditate’, irakin ‘boil’, iraun ‘persist’, bazkaldu ‘have dinner’, afaldu ‘have supper’, and dimititu ‘resign’ (Hualde & de Urbana 2003: 388–389). Moreover, there is a large class of unergative complex predicates (or light verb constructions), which are monovalent, they are composed of a bare noun and a verb indicating action: egin ‘do’, as in (c). Basque is coded as 1.
a. Jon-ek saltatu du. (unergative)
Jon-ERG jump AUX
‘Jon jumped.’ (Hualde & de Urbana 2003: 364)
b. Jon etorri da. (unaccusative)
Jon.ABS come AUX
‘Jon came.’ (Hualde & de Urbana 2003: 364)
c. Jon-ek korri egin du. (unergative complex predicate)
Jon-ERG run do AUX
‘Jon ran.’ (Hualde & de Urbana 2003: 390)
German (ISO 639-3: deu, Glottolog: stan1295)
German has about a dozen verbs that take the patient-like argument in the dative case (Witzlack-Makarevich p.c. 2020). They include danken ‘thank’, folgen ‘follow’, and verzeihen ‘pardon, forgive’. German is coded 0 because this verb class does not qualify as substantially large. Many other languages (if not every language) have small classes of predicates with marking patterns deviating from the one major pattern.
a. Ich danke dir.
I thank you.DAT
‘I thank you.’
b. Bitte folgen Sie mir!
please follow you me.DAT
‘Please, follow me!’
c. Ich kann ihm nicht verzeihen.
I can him.DAT NEG forgive
‘I cannot forgive him.’
Osage (ISO 639-3: osa, Glottolog: osag1243)
Osage transitive verbs index both A and P arguments, as in (a) (see Pustet 2002). Around 70% of intransitive verbs use the same indexes to mark the S argument as the ones used for the A argument, as in (b), whereas the remaining 30% of intransitive verbs (still a substantial amount) use the indexes identical to the ones used for the P arguments. There seems to be a semantic basis for the formal contrast between the indexing patterns with A and P prefixes. This semantic distinction is described in terms of the notion of agency: i.e. 'active' intransitive verbs use A indexes and 'non-active' intransitive verbs use P indexes. Osage is coded as 1.
a. ǫ-tha-nǫk̩ˀǫ
1SG.P-2SG.A-hear
‘You hear/heard me.’ (Wolff 1952: 234, as in Pustet 2002: 391)
b. tha-k̩ˀǫ'çagi (and other active verbs)
2SG.A-very.fast
‘You are very fast.’ (La Flesche 1932: 88, as in Pustet 2002: 391)
c. ǫ-shį'
1SG.P-fat
‘I am fat.’ (La Flesche 1932: 131, as in Pustet 2002: 391)
Makasae-Makalero (ISO 639-3: mkz, Glottolog: maka1316)
In Makasae-Makalero the core arguments are not flagged. This is true for both pronominal (a-c) and nominal (d) arguments. Thus, there is no variation in flagging of core participants based on verb classes. Furthermore, most Makasae-Makalero verbs do not index core arguments, as in (a)-(d). A set of five verbs index the number of S and A arguments via suppletive stem allomorphs (Huber 2011: 130). This set is not sufficient for coding as 1. Furthermore, a small group of verbs overtly index a referential third person P argument with the prefix k- (Huber 2011: 349-350). This set is not sufficient for coding as 1. Makasae-Makalero is coded as 0.
a. Ani hai mu’a-li’an.
1SG NSIT ground-fall
‘I already fell down.’ (Huber 2011: 146)
b. Ani ei pase.
1SG 2SG beat
‘I beat you.’ (Huber 2011: 218)
c. Ei ani pase.
2SG 1SG beat
‘You beat me.’ (Huber 2011: 218)
d. Ina-uai ni-mata uaro
mother-HON REFL-child wash
‘The mother is washing her child.’ (Huber 2011: 391)
(Abbreviations: NSIT new situation)
Further reading
Comrie, Bernard. 2013. Alignment of Case Marking of Full Noun Phrases. In: Dryer, Matthew S. & Haspelmath, Martin (eds.) The World Atlas of Language Structures Online. Leipzig: MPI EVA. (the section on active-inactive alignment)
Dixon, R. M. W. 1994. Ergativity (Cambridge Studies in Linguistics 69). Cambridge, U.K.: Cambridge University Press. (Section 4.1. Split conditioned by the semantic nature of the verb)
Siewierska, Anna. 2013. Alignment of Verbal Person Marking. In: Dryer, Matthew S. & Haspelmath, Martin (eds.) The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. (the section on active alignment)
References
Hualde, José I. & Jon Ortiz de Urbana. 2003. A grammar of Basque. (Mouton Grammar Library, 26.) Berlin: Mouton de Gruyter.
Huber, Juliette. 2011. A grammar of Makalero: A Papuan language of East Timor. Amsterdam: Vrije Universiteit Amsterdam. (Doctoral dissertation.)
La Flesche, Francis. 1932. A dictionary of the Osage language. Washington, D.C.: Government Printing Office.
Pustet, Regina. 2002. Split intransitivity revisited: Comparing Lakota and Osage. International Journal of American Linguistics 68(4). 381–427.
Rosés Labrada, Jorge E. 2015. The Mako language: Vitality, grammar and classification. London, Ontario: University of Western Ontario. (Doctoral dissertation.)
Wolff, Hans. 1952. Osage II: Morphology. International Journal of American Linguistics 18, 231–237.
Related Features
Features on indexing
- GB089 Can the S argument be indexed by a suffix/enclitic on the verb in the simple independent clause?
- GB090 Can the S argument be indexed by a prefix/proclitic on the verb in the simple independent clause?
- GB091 Can the A argument be indexed by a suffix/enclitic on the verb in the simple independent clause?
- GB092 Can the A argument be indexed by a prefix/proclitic on the verb in the simple independent clause?
- GB093 Can the P argument be indexed by a suffix/enclitic on the verb in the simple independent clause?
- GB094 Can the P argument be indexed by a prefix/proclitic on the verb in the simple independent clause?
Features on flagging
- GB408 Is there any accusative alignment of flagging?
- GB409 Is there any ergative alignment of flagging?
- GB410 Is there any neutral alignment of flagging?
Patron
Alena Witzlack-Makarevich