Sociative causation

Authors: Françoise Rose, Marine Vuillermet

Acknowledgements: Natalia Chousou-Polydouri, Oscar Cocaud-Degrève, Kellen Parker Van Dam, David Inman, Raphaël Luffroy, Nora Muheim, David Timothy Perrot

What?

Sociative causation is a particular type of causation where the causer not only makes the causee do an action but also participates in it (Shibatani and Pardeshi 2022; Zúñiga and Kittilä). Shibatani and Pardeshi (2022) distinguish three semantic subtypes of sociative causation, for which Guillaume and Rose (2010) provide the following paraphrases:

  • joint action ‘make someone do something by doing it with him/her’
  • assistive ‘help someone do something’
  • supervision ‘make someone do something, and supervise this action’

The three subtypes are illustrated in (1), where the basic causative in Japanese [nucl1643] (Japonic; Shibatani and Pardeshi 2022:100) can be used to express all three subtypes of sociative causation.1

(1) a. Sociative causation: joint-action
Hahaoya-ga kodomo-o asoba-se-te i-ru.
mother-NOM child-ACC play-CAUS-CONJ be-PRS
‘Mother is making the child play (playing with her).’
      b. Sociative causation: assistive
Hahaoya-ga kodomo-ni osikko-o sa-se-te i-ru.
mother-NOM child-DAT pee-ACC do-CAUS-CONJ be-PRS
‘Mother is making the child pee.’
      c. Sociative causation: supervision
Hahaoya-ga kodomo-ni hon-o yoma-se-te i-ru.
mother-NOM child-DAT book-ACC read-CAUS-CONJ be-PRS
‘Mother is making the child read a book.’

Shibatani and Pardeshi 2022:99 mention the possibility of having markers dedicated to sociative causation, like the sociative causative ha- in Alamblak [alam1246] (Sepik; Papua New Guinea) in (2a), which exists independently from the direct and indirect causatives, ka- in (2b) and hay- in (2c), respectively (adapted from Bruce 1984: 155-156).

(2) a. Sociative causation
ha-fkne-më-r-m
SOC.CAUS-enter-RPST-3SG.M-3PL
‘He caused them to enter (something) by entering with them.’
      b. Direct causation
ka-fkne-më-r-m
DIRECT.CAUS-enter-RPST-3SG.M-3PL
‘He caused them to enter (something) by physically taking them.’
      c. Indirect causation
yima-r hay-noh-më-r-a hay-noh-më-r-a
person-3SG.M INDIRECT.CAUS-unconscious-RPST-3SG.M-1SG
‘A man gave me (something) (causing) me (to become) unconscious.’

The present survey builds on the typology of the expression of sociative causation proposed by Guillaume and Rose (2007). Our current typology draws a primary distinction between

  • a dedicated construction2 (SocCaus-01) – sociative causation is expressed via a construction specific to this function, like ha- in Alamblak (2a);
  • a non-dedicated construction (SocCaus-07) – sociative causation is expressed through the use of a construction that can express other semantics, like the general causative marker -se in Japanese (1a-c).

Each dedicated construction is further coded for 5 morphosyntactic and semantic features:

  • its morphology (SocCaus-02);
  • the semantic subtypes of sociative causation it can express (SocCaus-03);
  • the kind of verbs semantics it occurs with (SocCaus-04);
  • the transitivity status of these verbs (SocCaus-05);
  • the animacy value of the causee (SocCaus-06).

A separate feature (SocCaus-07) deals with non-dedicated constructions to investigate more generally how languages express the concept of sociative causation.

Sociative causation is discussed in the literature under the following alternative names:

Note that the Assistive concept in Shibatani and Pardeshi (2022) requires causative semantics, while the Assistive in Kulikov (1993; 2001) is not a type of causative per se.

Why?

Sociative causation is a relatively rare feature in the world's languages. In their pilot survey on dedicated grammatical morphemes for sociative causation in a world-wide convenience sample, Guillaume and Rose (2010) show that it is fairly rare outside South America, and that most languages with such markers cluster specifically in South-Western Amazonia.

Because the phenomenon is potentially areal and still under-studied, this feature set also collects information on the morphosyntactic and semantic properties of sociative causation and the type of construction(s), both dedicated and non-dedicated, used to express it.

How?

To analyze a construction as dedicated or not dedicated, we adhered to the following rules:

  • If a construction systematically expresses sociative causation only, it was coded as a dedicated construction, even if it is (apparently) attested with only one semantic verb class;3
  • If a marker expresses sociative causation with only some semantic verb classes (e.g. motion verbs), but expresses some other meanings with other semantic verb classes, it was coded as a non-dedicated construction.

Features

SocCaus-01: Does the language have a dedicated construction to express sociative causation?

 { yes | no }

yes: Alamblak [alam1246] (Sepik; Papua New Guinea)

In (2) above, Bruce (1984: 155-156) contrasts the dedicated sociative causative ha- in (2a) with the marker of direct causation ka- in (2b) and the marker of indirect causation hay- in (2c).

no: Japanese [nucl1643] (Japonic)

Example (1) above shows that the marker se- can express the three semantic subtypes of sociative causation identified by Shibatani and Pardeshi (2022: 102). But se- is also a general causative marker: in (3a-b), the causer certainly makes the causee do the action but does not need to participate in it (Shibatani and Pardeshi 2022: 87).

(3) a. Taroo-ga Ziroo-ni/o hasira-se-ta.
Taroo-NOM Jiro-DAT/ACC run-CAUS-PST
‘Taro had/made Jiro run.’
      b. Taroo-ga Ziroo-ni kabin-o wara-se-ta.
Taroo-NOM Jiro-DAT/ACC vase-ACC break-CAUS-PST
‘Taro made Jiro break the vase.’

SocCaus-02: If <yes> to SocCaus-01, what type of construction is sociative causation expressed with?

 { NA | morpheme | special causee | direct-and-indirect-causative | causative-and-applicative | causative-and-other }

SocCaus-02 explores the five formal categories we have distinguished, listed below. The number which follows each type indicates the number of the example the category is illustrated in.

Types of dedicated constructions:

  • morpheme (2a)
  • combination of causative and
    • special case on the causee (4)
    • (another) causative (5)
    • applicative (6)
    • another marker (8)
morpheme: Alamblak [alam1246] (Sepik; Papua New Guinea)

The sociative causative marker -ha only encodes sociative causation, as illustrated in (2a) above.

special causee: Eastern Panjabic [east2727] (Indo-European; Bangladesh, India, Pakistan)

The sociative causative is conveyed by the regular causative marker, but the causee is encoded in a special way. As can be seen in (4b), there is an extra accusative/dative marker on the causee, which is absent in the regular causative meaning illustrated in (4a) (Khokhlova 2003, cited in Guillaume and Rose 2007).

(4) a. ablative case: regular causative meaning
us ne [kulĩ toⁿ] sũṭkes cukvãiã
3OBL ERG porter ABL suitcase lift+CAUS+PP.M.SG
‘He made the porter lift (his) luggage.’
      b. accusative/dative case: sociative causative meaning
to asĩⁿ [us nũ] samãn cukvãiã
then we 3OBL ACC/DAT luggage lift+CAUS+PP.M.SG
‘(The old woman was weak and did not manage to lift her luggage,) so we helped her to lift it.’
direct-and-indirect-causative: Apalai [apal1257] (Cariban; Brazil)

Sociative causation is conveyed by a combination of the direct causative marker with the indirect causative marker in Apalai (Koehn and Koehn 1986: 49-51, cited in Guillaume and Rose 2007).

(5) a. Causative 1: direct causative meaning
poeto otuh-ma-Ṽko mana
child eat-DIR.CAUS-CONT 3+be+PRES
‘He is feeding the child.’
      b. Causative 2: indirect causative meaning
tapyi tae nu-tũta-nohpo-no
house out.of 3S3O-leave-INDIR.CAUS-IP
‘He caused (him) to leave.’
      c. Causative 1 + Causative 2: sociative causative meaning (supervision)
aimo otuh-ma-nohpo-Ṽko mana
boy eat-DIR.CAUS-INDIR.CAUS-CONT 3+be+PRES
‘He is overseeing the child's eating.’
causative-and-applicative: Ese Ejja [esee1248] (Pano-Tacanan; Bolivia, Peru)

Ese Ejja expresses sociative causation with the transparent combination of the causative marker -mee (Vuillermet 2012: 509), illustrated in (6a), and the comitative applicative -sawa, illustrated in (6b).

(6) a. Causative marker only
E-sho'i kojo-mee-kwe!
NPF-child collect-CAUS-IMP
‘Make the children go collect (the banana)!’ (Vuillermet 2012: 509)
      b. Comitative applicative only
Poki-sawa-naje.
go-COM.APPL-PST
'(I) went with you'. (Context: the addressee had decided to go for a walk to try to get her baby asleep
and the speaker had offered to accompany them.) (Vuillermet's fieldnotes)
      c. CAUS + COM.APPL systematically expressing sociative causation
Jja-aja-mee-ki-sawa-ka-naje
VBZ-healthy-CAUS-VBZ-COM.APPL-PST
'(God) helped (her) to get healthy.' (Shoemaker and Shoemaker 1983: 70)

Note that in other languages, a similar combination might not necessarily encode sociative causation. For instance in Yimas, this same combination gives rise to a caused event with an additional causee (in the dative), rather than involving the causer in the caused event.

(7) Yimas (Lower Sepik-Ramu; Papua New Guinea; Foley 1991: 303)
m-n pu-ŋa-taŋ-tar-kwalca-t
NR.DIST-1SG 3PL.OBJ-1SG.DAT-COM.APPL-CAUS-rise-PFV
‘He woke them up along with me.'
causative-and-other: Nande [nand1264] (Bantu; Democratic Republic of the Congo, Uganda)

Sociative causation is conveyed by the combination of a causative marker plus another marker that is not another causative or applicative marker. This state is heterogeneous in comparison to the previous ones, since the second morpheme can have different functions across languages. For instance, Schneider-Zioga and Mutaka (2019: 281) describe a sociative causative construction in Nande (aka Kinande) that combines a (direct) causative marker and a middle marker4 that synchronically conveys a sociative causative meaning exclusively.

(8) eri̹-hek-ek-i̹-a Kámbalɛ́ y' ɔkɔ mbágɔ
5-carry-MID-CAUS-FV 1Kambale 1LK' 17LOC 10plank
To help Kambale carry planks.

SocCaus-03: If <yes> to SocCaus-01, what sociative causation semantic subtypes are attested?

 { NA | ; separated list of [ joint action, assistive, supervision ] }

For every language that is not <NA>, the list of the attested verb pairs (original root / stem derived with SOC.CAUS) on which this analysis is based is available in the remarks.

SocCaus-04: If <yes> to SocCaus-01, what kind of verb semantics is the sociative causative construction attested with?

 { NA | motion | non-motion | both }

This semantic feature is exploratory and only targets the [+/- motion] parameter. It originated from a remark in Guillaume and Rose (2007: 393) noting that the semantic extensions manifested by applicative markers were particularly frequent with action verbs, and more specifically motion verbs. The complete list of verbs attested with the sociative causative construction is available in the remarks.

SocCaus-05: If <yes> to SocCaus-01, what is the transitivity of the verb roots attested with the sociative causative construction?

 { NA | intransitive | transitive | both }

Guillaume and Rose (2010) observe that the sociative causative construction in Teko [emer1243] (Tupian; French Guiana) is only available with intransitive verb roots, while the one in Cavineña [cavi1250] (Pano-Takanan; Bolivia) is available with both intransitive and transitive roots. In our sample, no language has the construction only attested with transitive roots.

SocCaus-06: If <yes> to SocCaus-01, what is the animacy value of the causees attested with the sociative causative construction?

 { NA | animate | inanimate | both }

This question checks whether the construction is attested - with animate causees only, as in Ese Ejja [esee1248] (Pano-Takanan; Bolivia, Peru); - with inanimate causees only (not attested in our sample); - or with both, as in Karo [karo1305] (Tupian, Brazil).

SocCaus-07: What verbal grammatical construction(s) can express sociative causation (apart from a dedicated construction)?

 { ; separated list of [ direct-causative, indirect-causative, reciprocal, applicative, co-participative, other ] }

SocCaus-07 distinguishes 6 formal categories listed below. The number following each category corresponds to the example number illustrating each type.

Types of non-dedicated constructions:

  • direct causative (9)
  • indirect causative (10)
  • reciprocal (11)
  • applicative (12)
  • co-participative (13a-b)
  • other (14)

Note that it is possible for a language (e.g. Yanomami, as discussed in Perri Ferreira 2017: 488-493) to have more than one non-dedicated construction.

direct-causative: Nanti [nant1250] (Arawakan; Peru)

The <DIR.CAUS> state indicates that indicates that sociative causation can be conveyed by a general causative marker or by a marker of direct causation (if the language shows a formal distinction between direct and indirect causation). In Nanti (Michael 2008: 280), the "agent causation" morpheme ogi mostly expresses direct causation (9a), but can also express sociative causation (9b).

(9) a. Direct causation
YogikoNtetanakeri.
i=ogi-koNtet-an-ak-i=ri
3MS=DIR.CAUS-leave-ABL-PERF-REAL.I=3MO
'He made him leave.'
b. Sociative causation
YogimoNtehanakena
i=ogi-moNteh-an-ak-i=na
3MS=DIR.CAUS-cross.river-ABL-PERF-REAL.I=1O
'He helped me across the river (e.g. by taking me across in a boat).'
indirect-causative: Nanti [nant1250] (Arawakan; Peru)

Sociative causation can be conveyed by an indirect causative marker (if the language shows a formal distinction between direct and indirect causation), with no specific marking of the causee.

Nanti (Arawakan) also has an indirect causative marker called “influential causation” (Michael 2008: 284), illustrated in (10). When discussing its semantics, Michael mentions a situation where a causer "bring[s] about [a causee]'s departure by accompanying them", i.e. a typical sociative causative meaning. (No glossed example illustrates the sociative causative semantics.)

(10) Indirect causation in Nanti
Te tsini pakuhakagarime.
te tsini pakuh-akag-∅-a=ri=me
NEG.REAL who drop-INDIR.CAUS-IMPF-REAL.A=3MO=CNTF
Its not as if anyone induced him to abandon (his spouse).'
reciprocal: Bedja [beja1238] (Afro-Asiatic; Egypt, Eritrea and Sudan)

Sociative causation can be conveyed by a reciprocal marker, as in Bedja (Roper 1928: 73, cited in Guillaume and Rose 2007).

(11) a. Reciprocal
amó-dār-na
RECIP-kill-IMPER.PL
Kill each other!
      b. Sociative causation (assistive)
baró amó-dār-heb
3M.SG.AC RECIP-kill.IMPER.M.SG-1SG.AC
Help me to kill him!
applicative: Sikuani [guah1255] (Guahiboan; Colombia and Venezuela)

Sociative causation can be conveyed by an applicative marker. In (12a), the Sikuani applicative marker has an instrumental meaning, while in (12b) the same marker expresses sociative causation (Queixalós 2002: 320-321).

(12) a. Instrumental applicative meaning
mahalu computadora Ø-ka-yakina-Ø baharpaliwaisianü.
Mahalu computer 3OBJ-APPL-carve-3SUBJ those.stories
‘Mahalu wrote these stories with the computer.’
      b. Sociative causative meaning
Kuwainü Ø-ka-nawiata-Ø pihawa
God 3OBJ-APPL-go.back-3SUBJ his.wife
‘God took his wife back home.’
co-participative: Yurakaré [yura1255] (isolate; Bolivia)

Sociative causation can be conveyed by a co-participative marker. Creissels and Nouguier-Voisin define co-participation "as applying to constructions that imply a plurality of participants in the event they refer to without assigning them distinct roles." The cooperative marker in Yurakaré (van Gijn 2006: 150) may express co-participation only (13a) or sociative causation (13b). This morpheme contrasts with the (dedicated) sociative causative construction (called "comitative object" by van Gijn (2006: 149-150)), which systematically expresses sociative causation (13c).

(13) a. Cooperative meaning with co-participative marker ku-
abëssë-∅ 'he is playing' ku-bëssë-∅ 'he is playing with him'
abayla-∅ 'he is dancing' ku-bayla-∅ 'he is dancing with her'
b. Sociative causation meaning with co-participative marker ku-
dula-∅ 'he makes it' ku-dula-∅ 'he helps him make it'
anënë-∅ 'he is cooking' ku-nënë-∅ 'he helps him cook'
c. Cooperative meaning with sociative causative marker ka-
mala-∅ 'he goes' ka-mala-∅ 'he brings, takes it'
abayla-∅ 'he is dancing' ka-bayla-∅ 'he picks her to dance with'
other: Garrwa [gara1269] (Garrwan; Australia)

This state applies if none of the above states is appropriate. The morpheme(s) used to express sociative causation are described in detail in the 'remark' column.

In Garrwa, the suffix -kunu associated with the causative -mba creates caused motion verbs from verbs of motions or other verbs, often but not always with a sociative causative interpretation (14a vs. 14b; Mushin 2012: 178-179).

(14) a. Sociative causative meaning (both causer and causee move)
burrij-kunu-mba
cross(water)-CAUS.MOTION-CAUS
‘take something across water’
b. Caused motion meaning (causer does not move)
ngangkij-kunu-mba
enter-CAUS.MOTION-CAUS
‘put something inside something else’

Results

Our survey has confirmed that dedicated sociative causative constructions are rare outside South America, but not in South America, where they are mostly found in Greater Amazonia. Out of the 325 languages included in the ATLAs sample, 19 show a dedicated construction for sociative causation (SocCaus-01), and 15 of these 19 languages are spoken in South America. The prevalence of dedicated constructions in this macroarea is striking: South America is the only macroarea where sociative causation is frequently grammaticalized (15/105). This is all the more telling, given that non-dedicated constructions (SocCaus-07) are evenly distributed across macroareas, with an average attestation of one in four languages.

Dedicated constructions in South America are present across families, with a higher prevalence in the Tupian family, for which a Proto-Tupi “comitative causative” form has been reconstructed (Rodrigues and Cabral 2012: 509, 531-533). Combined with other arguments, our data favors Guillaume and Rose's (2010) hypothesis that the feature has diffused from Tupian languages through pattern borrowing (and not borrowing of a specific morphological form).

Formally, most dedicated sociative causative constructions in our survey use a sociative causative morpheme rather than a combination of morphemes or a specific causee marker. Non-dedicated constructions fall in the types previously described Guillaume and Rose (2007).

As for the semantic observations, the results remain exploratory as the information was frequently too scarce in the sources. In some grammars, like the one of Teko (Rose 2011) or Yurakaré (van Gijn 2006), many illustrating examples were available, allowing for a reliable account of the productivity of the morpheme with different semantic verb types. In others, like the one of Borum [kren1239] (Nuclear-Macro-Je; Brazil) (Seki 2004) or Ese Ejja (Vuillermet 2012), examples were scarce and did not allow us to be categorical about apparent gaps. Ungrammaticality of specific semantic combinations was very rarely mentioned. Our observations below could still be used as guidelines for future studies.

We observed that motion verbs are well attested with dedicated sociative causative constructions. The data also confirm Shibatani and Pardeshi's (2022: 99) observation that sociative causation expressions are preferentially used with intransitive verbs: all languages display the sociative causative construction with intransitive verbs, and nine of them exclusively with such verbs. In addition, the South American geographic distribution of the transitivity status of the verbs attested in the target construction (SocCaus-05) splits along an East-West divide: the six easternmost languages display sociative causative constructions with intransitive verbs only, while the nine westernmost languages display constructions with both intransitive and transitive verbs. A very similar divide occurs with constructions involving animate causees only vs both animate and inanimate causees. Both animates and inanimates appear in the sociative causative constructions of the six easternmost languages (plus two more western languages), while only animate causees appear in the six westernmost languages

As for the semantic subtypes (joint action, assistive and supervision), we cannot posit strong generalizations because of the lack of systematic criteria to distinguish them. Joint action seems the most present subtype in the languages of our sample with dedicated sociative causative markers. Although assistive is found in only a few languages, it seems the only subtype present in two languages. Sources systematically give fewer examples of the assistive than joint action, while those of supervision are even rarer. The higher frequency of joint action may be a consequence of the fact that sociative causation is less frequent with transitive verbs, where it is more susceptible to take the supervision and assistive interpretations.

Contributions

Conceptualization: Marine Vuillermet, Françoise Rose

Data collection: Natalia Chousou-Polydouri, Oscar Cocaud-Degrève, Kellen Parker Van Dam, David Inman, Raphaël Luffroy, Nora Muheim, David Timothy Perrot, Françoise Rose, Marine Vuillermet

Supervision of data collection: Françoise Rose, Marine Vuillermet

References

  • Adam, Lucien. 1896. Matériaux pour servir à l'établissement d'une grammaire comparée des dialectes de la famille Tupi. (Bibliothèque Linguistique Américaine, XVIII.) Paris: Librairie-Éditeur J. Maisonneuve.

  • Bruce, Les. 1984. The Alamblak Language of Papua New Guinea (East Sepik). (Pacific Linguistics: Series C, 81.) Canberra: Research School of Pacific and Asian Studies, Australian National University. iv+361pp.

  • Creissels, Denis & Sylvie Nouguier-Voisin. 2008. Valency-changing operations in Wolof and the notion of "co-participation". In König, Ekkehard and Gast, Volker (eds.), Reciprocals and Reflexives: Theoretical and Typological Explorations, 289–306. De Gruyter Mouton.

  • Dixon, R. M. W. 2000. A typology of causatives: form, syntax and meaning. In Dixon, R. M. W. and Aikhenvald, Alexandra Y. (eds.), Changing valency: case studies in transitivity, 30–83. Cambridge: Cambridge University Press.

  • Perri Ferreira, Helder. 2017. Yanomama Clause Structure. Radboud Universiteit Nijmegen dissertation. (382+397pp.)

  • Guillaume, Antoine & Françoise Rose. 2007. A typology of sociative causative: between causatives and applicatives.

  • Guillaume, Antoine & Françoise Rose. 2010. Sociative causative markers in South-American languages: a possible areal feature. In Floricic, Frank (ed.), Essais de typologie et de linguistique générale, Mélanges offerts à Denis Creissels, 383–402. Lyon: ENS Editions.

  • Jensen, Cheryl. 1998. Comparative Tupí-Guaraní Morpho-syntax. In Derbyshire, Desmond C and Pullum, Geoffrey K. (eds.), Handbook of Amazonian Languages, 490–603. Berlin: De Gruyter Mouton.

  • Khokhlova, L. 2003. Infringement of morphological and syntactic operations’: Pairing in “Second Causative” formation. Indian Linguistics 64(1). 1–17.

  • Koehn, Edward & Sally Koehn. 1986. Apalai. In Desmond C. Derbyshire and Geoffrey K. Pullum (eds.), Handbook of Amazonian Languages, 33–127. Berlin: De Gruyter Mouton.

  • Kulikov, Leonid I. 1993. The "second causative": A typological sketch. In Comrie, Bernard and Polinsky, Maria (eds.), Causatives and Transitivity, 121–154. Amsterdam: John Benjamins Publishing Company.

  • Kulikov, Leonid I. 2001. Causatives. In Haspelmath, Martin and König, Ekkehard and Oesterreicher, Wulf and Raible, Wolfgang (eds.), Language typology and language universals: an international handbook, 886–898. Berlin: De Gruyter Mouton.

  • Michael, Lev David. 2008. Nanti evidential practice: Language, knowledge, and social action in an Amazonian society. University of Texas at Austin dissertation. (xxv+462pp.)

  • Mushin, Ilana. 2012. A grammar of (Western) Garrwa. (Pacific Linguistics, 637.) Berlin, Boston: De Gruyter Mouton. xv+476pp.

  • Queixalós, Francesc. 2002. The notion of transfer in Sikuani causatives. In The Grammar of Causation and Interpersonal Manipulation, 319–339.

  • Rodrigues, Arion Dall'Igna. 1953. Morfologia do verbo Tupí. Letras 1. 121–152.

  • Rodrigues, Aryon Dall'Igna & Ana Suelly Arruda Câmara Cabral. 2012. Tupían. In Campbell, Lyle (ed.), The Indigenous Languages of South America, 495–574. Berlin: De Gruyter Mouton.

  • Roper, E. M. 1928. Tu Beḍawiɛ an elementary handbook for the use of Sudan government officials. Hertford UK: Stephen Austin & Sons.

  • Rose, Françoise. 2011. Grammaire de l'Émérillon Teko, une langue Tupi-Guarani de Guyane française. (Langues et Sociétés d'Amérique traditionnelle, 10.) Louvain: Peeters. xix+489pp.

  • Schneider-Zioga, Patricia & Philip Ngessimo Mutaka. 2019. The syntax and semantics of helping: Sociative causation in Kinande. Journal of African Languages and Linguistics 40(2). 271–310.

  • Seki, Lucy. 2004. Aspectos da morfossintaxe Krenak: orações independentes. LIAMES 4(4). 131–148.

  • Shibatani, Masayoshi & Prashant Pardeshi. 2002. The causative continuum. In Shibatani, Masayoshi (ed.), The Grammar of Causation and Interpersonal Manipulation, 85–126. Amsterdam: John Benjamins.

  • Gijn, Rik van. 2006. A Grammar of Yurakaré. Nijmegen: Radboud Universiteit Nijmegen dissertation. (xi+361pp.)

  • Vuillermet, Marine. 2012. A Grammar of Ese Ejja, a Takanan language of the Bolivian Amazon. Université Lumière Lyon 2 dissertation. (xvi+736pp.)

  • Zúñiga, Fernando & Seppo Kittilä. 2019. Grammatical Voice. Cambridge: Cambridge University Press.

  1. Examples are from a variety of sources. We have homogenized the glosses using the Leipzig Glossing Rules with the addition of the gloss SOC.CAUS for the markers (re)analyzed as sociative causative. ↩

  2. We consider that a morpheme is a type of construction. ↩

  3. In some languages, the sociative causation semantics is only available in highly constrained conditions, typically only [motion verb + inanimate object], such as 'go+CAUS' systematically resulting into the semantics of 'take'. Such cases are reported in the remarks. ↩

  4. Schneider-Zioga and Mutaka (2019: 281) do not label ek- a middle. According to them, the few existing studies on cognates of this marker consider them as having "a detransitivizing/stativizing use [...] [called] NEUTER, and a transitivizing use [called] the IMPOSITIVE." In addition, Dom (2014) reports for the Nande marker two major functions (detransitivization/passivization and "ability"/middle marker) that lead us to think it could be a general middle marker. ↩