Authors: David Inman, Marine Vuillermet, Natalia Chousou-Polydouri, Balthasar Bickel
Acknowledgements: Selma Hardegger, Nora Muheim, Kellen Parker van Dam, Anna Graff
This feature set investigates monoexponential verbal plurals, which we define as plural (and collective) agreement markers that are formally separate or separable from person markers. The presence of these markers, depending on the language, may be restricted to a single syntactic role or may cover a range: the Single argument (S) of an intransitive verb, the Agent (A) of a transitive verb, and/or the Patient (P) of a transitive verb. In addition to marking core argument plurality, it may also mark event plurality (E). The presence of monoexponential verbal plurals may be optional or obligatory, and they may be marked directly on the verb or elsewhere on the clause (through a second-position clitic, a required auxiliary, reduplication, etc.). In some cases, the same monoexponential verbal plural also occurs on other parts of speech, such as (pro)nouns, adjectives, or adpositions.1
In order for a verbal plural marker to be considered monoexponential, it needs to indicate only plurality (i.e. no portmanteau markers of person and number are considered).2 However, there is a complication around the 3rd person plural, as many languages mark a 3rd person singular with a null morpheme and 1st and 2nd persons (singular and plural) overtly. In such a system, a morpheme that indicates 3rd plural could be analyzed as either a joint person-number marker (3pl) or a monoexponential plural, for which an interpretation of 1pl or 2pl is blocked by the presence of separate person markers. To be consistent across language, we have made the following analytical decision: if a plural marker is completely separate from the person marking paradigm (e.g. by appearing in a different slot) and only has a 3rd person plural interpretation, we have considered these as monoexponential plural markers, regardless of the rest of the system, while documenting the peculiarity in the remarks. Tapieté [tapi1253] (Tupian; Argentina, Bolivia & Paraguay) exemplifies such a case: the markers indexing person and number are prefixes, while -reta, though it marks only 3rd person plural, is a suffix. (See also Ese Ejja [esee1248] (Pano-Tacanan; Bolivia & Peru) further below).
| (1) | Tapieté (Tupian; Gonzáles 2005: 102,34) | |||
|---|---|---|---|---|
| a. | hama | ha’e-mi-re | shɨ-kwa-ma-yɨwe-reta. | |
| then | (s)he-DIM1-PL | 1SG.INACTIVE-know-RES-???-NOM-PL | ||
| ‘And they already know me also.’ | ||||
| b. | a-mbo’e-po-yɨ-reta-kwi. | |
|---|---|---|
| 1SG.ACTIVE-teach-FUT-FRQ-PL-DISTFUT | ||
| ‘I will teach them again.’ |
There are two edge cases where the phenomena of Sg-Pl stem alternation and monoexponential plurals almost overlap. The main difference between the two are their distribution: Sg-Pl stem alternation is definitionally lexically restricted, and monoexponential verbal plurals must apply without lexically-based restrictions. Typically, this means a single form applies to every verb in the lexicon.
The first edge case involves semantic predictability. Magdalena Peñasco Mixtec [magd1235] (Otomanguean; Mexico) has a semantically restricted plural marker koo, which only appears with motion verbs (Erickson de Hollenbach 2013: 48). Such lexical restrictions appear to be rare. Here, the origin of koo (from koyo ‘pour’) is a metaphor for movement in a group (Erickson de Hollenbach 2013: 151). Because it is restricted to a semantic class of verbs, language learners don't have to memorize the target verbs, and we consider that it is a monoexponential plural, even if it does not target the whole lexicon. This morpheme is not treated as a regular pattern in Sg-Pl stem alternation, as such plural markers require the language learners to memorize the (randomly) lexically restricted set of verb(s) where they appear.
The second edge case involves formal predictability. Rather than having a single monoexponential marker that encodes plurality, some languages have several, mutually exclusive markers. Because these are synchronically available to the whole lexicon, the set of plural markers as a whole would be a bonafide monoexponential plural construction. But the fact that the set consists of many different forms (e.g. 3 to 8 for Tsimshian languages, see Tarpent 1983) makes them akin to Sg-Pl alternating pairs: Each plural form is indeed not transparently distributed across the different verb classes, i.e. there are no semantic criteria to understand their synchronic distribution, and the forms must thus be memorized by the speakers. We chose to typologically separate:
Only the latter is treated as Sg-Pl alternation, while the former is considered to be a monoexponential plural category with irregular lexical classes, akin to the conjugation classes in the Romance language family. A very clear example of a language with such “plural conjugation classes” is Nisga'a [nisg1240] (Tsimshian; Canada), where a series of diachronic changes has led to a synchronic system of many possible plural forms which are not transparently distributed across different semantic verb classes (Tarpent 1983). Note that in addition to its monoexponential plural system, Nisga'a has a subset of verbs whose pluralization is made via suppletion, exactly for the semantics usually targeted by Sg-Pl suppletion (some of its motion and posture verbs). This seems to confirm that these are two distinct phenomena, as they can co-exist within a single language. (In our sample, most languages (15/17) with monoexponential plurals of such a “conjugation class” type also have Sg-Pl stem alternation.) This system is discussed in more detail under MonPl-02 below, which deals with irregular monoexponential plurals.
It has been noted by some observers that morphemes marking the plurality of more than one argument, or ambiguously mark the plurality of either argument of a transitive verb, appear to be relatively more common in North America than in the rest of the world (Mithun 1999: 83-94), or that they form a pattern within an area of the Americas (Leer 1991).
This survey does not only investigate the presence vs absence of monoexponential plurals, but also which syntactic role the plural targets, and if, in addition, it also can indicate that the event (E) occurs multiple times. Distinguishing between event and argument plurality can be difficult, since multiple transitive events often lead to an interpretation of multiple A or multiple P arguments. However, if a plural marker can truly indicate both event and argument plurality separately, then it will have two possible, separate interpretations. For argument plurality, at least one of the following interpretations must be possible for transitive verbs:
A single A(gent) repeats the action on multiple P(atient)s of the same type
Ex: One person shoots multiple targets.
Multiple A(gent)s perform the action
Ex: Multiple people shoot (perhaps the same target).
For event plurality, the following interpretation must be possible:
To ensure that a marker indicated argument plurality (in cases where this was in doubt), we looked for contexts where a plural event interpretation was unlikely or impossible. This meant contexts in which multiple arguments participated in a non-distributed shared event that did not repeat (e.g., go to a location as a group). If all examples had an available interpretation of event plurality and no context of strict argument plurality was found, we considered that the marker was not a monoexponential plural but a marker of event repetition, and entered this into the remarks field.
Once argument plurality was established, to ensure that a marker could also indicate event plurality, we looked for contexts where a plural argument interpretation was unlikely or impossible. This meant contexts where the plural marker was used with an intransitive verb and a singular S, or a transitive verb with singular A and singular P arguments. If no such context was found, we considered the marker to only index argument plurality.
Note that in some grammars, authors explicitly mentioned that the two types of plurality were available, depending on the context.
{ yes | no }
For the purposes of this survey, we are interested in morphemes that mark event plurality only if the same morpheme is also used for argument plurality, as described in the How? section. Verbal plural markers that indicate event plurality only are excluded.
The Pipil person indexation system has a suffix -t which accompanies all plural subjects. It thus has a monoexponential plural marker that is used for the S and A roles Campbell 1985.
| ni- | I |
|---|---|
| ti- | you (singular) |
| ∅- | he/she/it |
| ti-...-t | we |
| an-...-t | you (plural) |
| ∅-...-t | they |
The morpheme -reta in Tapieté exemplifies the edge case mentioned in the introduction where we consider as monoexponential a morpheme which is only used for the 3rd person plural but is distinct from the rest of the person-number paradigm. In Tapieté, the markers indexing person and number are prefixes, while -reta, though it marks only 3rd person plural, is a suffix (González 2005: 102,34), as exemplified above in (1a-b).
Central Kanuri has an -a- morpheme which marks the plurality of 1st and 2nd person in the P role, but not 3rd person (Hutchison 1981).
| sg | pl | |
|---|---|---|
| 1st person | s+ | s+a+ |
| 2nd person | nz+ | nz+a+ |
| 3rd person | ∅ | ∅ |
Aleut has a suffix -ngis that indicates argument plurality of 3rd person A or P. However, it is in a paradigm with other person-number combinations (1sg.A+pl.P), so it does not meet our definition of a monoexponential plural (Leer 1991: 170).
| (2) | Pitram | ∅ | kiðu-ku-ŋis. |
|---|---|---|---|
| Peter-SG.REL | 3 | help-PRES-PL | |
| ‘Peter is helping them.’ | |||
| (3) | ∅ | kiðu-ku-niŋ. |
|---|---|---|
| 3 | help-PRES-1SG/PL | |
| ‘I am helping them.’ | ||
{ NA | yes | no }
In the vast majority of cases, a monoexponential plural can only be analyzed as a singular morpheme that is applied across the lexicon (perhaps with predictable morphophonemic alternations). However, in some languages the use of one or another form is conditioned entirely by lexical class, membership in which is unpredictable. As mentioned in the What? section above, this grammatical phenomenon is both similar to Sg-Pl suppletion (in that specific plural verb forms have to be memorized), and to monoexponential plurals (in that the set of mutually exclusive markers target all verbs in the lexicon).
The plural form in Tapiete is always -reta. It has no irregular or lexically-conditioned allomorphy (González 2005, as exemplified in detail in (1a-b) in the introduction above and in (6-8) in MonPl-03 below).
There are eight possible shapes that a verbal plural morpheme can take in Nisga’a: Full stem reduplication, lv- prefixation, irregular reduplication (with additional consonant codas), tx̌a·- prefixation, partial reduplication, full word reduplication, Cix- prefixing, and qa- prefixing (Tarpent 1983). All verbs have one of these pluralizing strategies, but which one they use is lexically conditioned and irregular.
{ NA | no | ; separated list of [ S:_frequency_, A:_frequency_, P:_frequency_ ] }
Teko has three monoexponential verbal plural morphemes. We are going to discuss one of them here, while the other two will be discussed under MonPl-05.
The plural clitic -kom has a variety of uses in Teko: it can pluralize nouns, pronouns, demonstratives, and also possessors, objects of postpositions and objects of verbs. It is this last function that is relevant for this feature set. When attaching to verbs, -kom pluralizes only the P argument4 (4). The clitic -kom is a loan from Cariban languages, where a morpheme komo has been reconstructed with the same range of functions (Rose 2011: 111-116).
| (4) | ɨɳ-a-te | o-zika-kom. |
|---|---|---|
| DEM-a-FOC | 3.I-kill-PL | |
| ‘It’s him who killed them.’ | ||
Ese Ejja has a morpheme -ka that optionally appears only on intransitive verbs when the S argument is 3rd person plural (which we count as monoexponential for the same reasons given for Tapieté). The examples below show that the single argument is necessarily plural (and necessarily 3rd person) when -ka appears on an intransitive verb (cf. 5a vs. 5b); the corresponding form with no -ka (in 5b) could still refer to 3rd person plural, but also to any other person and number; to mark person (and number), there are (non-obligatory) second position clitics (5c and 5d) in the language for 1st and 2nd but not for 3rd person (Vuillermet 2012: 374-375).
| (5) | a. | Kwaji-kwaji-ka-ani. |
|---|---|---|
| run~RDP-3S.PL-PRS | ||
| ‘They run.’ |
| b. | Kwaji-kwaji-ani. | |
|---|---|---|
| run~RDP-PRS | ||
| ‘I/you/she/he/we/you.pl/they run.’ |
| c. | Majoya=se | poki-ani. | |
|---|---|---|---|
| then=1INCL.ABS | go-PRS | ||
| ‘Then we (INCL) go.’ | |||
| d. | A'a=mi | poki-ani? | |
|---|---|---|---|
| INT=2ABS | go-PRS | ||
| ‘Do you go?’ | |||
As illustrated in examples (6)-(8), the monoexponential verbal plural marker of Tapieté can indicate plurality of S, A, or P arguments (González 2005: 34, 102, 191).
| (6) | heta | iyeta-reta |
|---|---|---|
| a lot | speak-PL | |
| ‘A lot (of people) spoke.’ | ||
| (7) | hama | ha’e-mi-re | shɨ-kwa-ma-yɨwe-reta |
|---|---|---|---|
| then | (s)he-DIM1-PL | 1SGIN-know-RES-???-NOM-PL | |
| ‘And they already know me also.’ | |||
| (8) | a-mbo’e-po-yɨ-reta-kwi |
|---|---|
| 1SGAC-teach-FUT-FRQ-PL-DISTFUT | |
| ‘I will teach them again.’ |
{ NA | yes | no }
For this feature, it is crucial to establish event (E) plurality in a context with singular arguments, as described in the What? section.
Nuu-chah-nulth has a suffix =ʔał which may indicate plurality of S (15), A (16) or P (17) arguments (Inman 2021: 118-119).6 Because this marker can be used for both the A and P roles, its interpretation can be syntactically ambiguous, though in context the interpretation is typically clear.
| (9) | hitasaƛ=!aƛ=weʔin=ʔał | |
|---|---|---|
| land.on.beach.PF=NOW=HRSY.3=PL | ||
| ‘They landed on the beach.’ | ||
| (10) | ʔuk̓ʷiic=(m)it=ʔaał=ʔał | č̓apac |
|---|---|---|
| own.one’s.own=PST=HAB=PL | canoe | |
| ‘They had their own canoe.’ | ||
| (11) | ƛuł-ʔaqƛ=!ap=ʔic=ʔał | |
|---|---|---|
| good-FUT=CAUS=STRG.2SG=PL | ||
| ‘You (sg) will make them eat well.’ | ||
Note also that E plurality exists in the language, but it is done via reduplication and affixation on the verb, and not via the above plural morpheme.
In Movima, the monoexponential plural marker indicates S or E plurality with intransitive verbs, and P or E plurality with transitive verbs (Haude 2006: 434, 405):
| (12) | kay<ka:>ni |
|---|---|
| die<MLT> | |
| ‘to die one after the other (e.g. cattle).’ |
| (13) | jom-ka-ye:-na |
|---|---|
| dismiss-MLT-BE:person-DR | |
| ‘I've dismissed them (several people).’ ~ ‘I've dismissed him again and again.’ |
Karok has a suffix -va which can target S, P, or E plurality. When used on an intransitive, it can indicate either a plural S (18) or an iterative event (19), while when used on a transitive it may indicate a plural P (20) or an iterative event (21). Data is from Bright (1957: 92-93), cited in Mithun (1988: 215-216).
| (14) | a. | θivrú·htih |
|---|---|---|
| ‘(one object) to be floating’ | ||
| b. | θivrú·htih-va | |
| ‘(several objects) to be floating’ |
| (15) | a. | pasnáp-iš(rih) |
|---|---|---|
| ‘to glue down (one)’ | ||
| b. | pasnápi’šri·h-va | |
| ‘to glue down (several)’ |
| (16) | a. | ikre·myáhiš(rih) |
|---|---|---|
| ‘to start to blow’ | ||
| b. | ikre·myáhi’šri·h-va | |
| ‘to blow off and on’ |
| (17) | a. | taknah |
|---|---|---|
| ‘to hop’ | ||
| b. | takná·h-va | |
| ‘to play hopscotch’ |
{ NA | ; separated list of [ SA:frequency, SP:frequency, SAP:frequency ] }
Note that the AP combination was, as expected, not attested in our sample and is thus not included in the states above.
SAP: Nanti [nant1250] (Arawakan; Peru)
In Nanti, there is a plural suffix that can indicate the plurality of either A or P, as well as that of S, as illustrated with a transitive verb in (22) and an intransitive verb in (23) (examples from Michael 2008: 260, 317).
| (18) | i=kamoso-hig-ak-i=ri |
|---|---|
| 3mS=visit-PL-PERF-REAL.I=3mO | |
| ‘They visited him. / He visited them. / They visited them.’ |
| (19) | hee, | maganiro=ta | no=aNtabaget-hig-ak-i |
|---|---|---|---|
| yes | all=CNGT | 1S=work-PL-PERF-REAL.I | |
| ‘Yes, we all worked.’ | |||
Apart from the plural marker -kom indicating plurality of the P argument, which was mentioned in (4) under MonPl-03, Teko has two more monoexponential verbal plural markers, both indicating S and A plurality (subject plurality): the verbal clause clitic -(o)ŋ and the morpheme kupa5 (Rose 2011: 105-111). Examples with both transitive and intransitive verbs for both morphemes are given in (11-14).
| (20) | e-potar-oŋ. |
|---|---|
| 1SG.II-love-PL.SUBJ | |
| ‘They love me.’ |
| (21) | i-awu-ŋ |
|---|---|
| 3.II-speech-PL.SUBJ | |
| ‘They speak now.’ |
| (22) | oke-o-ker-o | kupa-ɲ. |
|---|---|---|
| RED-3.I-sleep-CONT | PL.SUBJ-CONT | |
| ‘They are sleeping.’ | ||
| (23) | saʔi | e-bo-boiɾ-ne | oɾone-ɨaɾ | ta-oɾo-ho | |
|---|---|---|---|---|---|
| grandmother | 2SG-CAUS-loosen-contrastive | 1EXCL-canoe | EXH-1EXCL-go | ||
| eʔi-te-we | kupa-o | ikeʔɨ. | |||
|---|---|---|---|---|---|
| 3.say-FOC-also | PL.SUBJ-CONT | then | |||
| ‘ “Grandmother, loosen (free) our canoe, that we can go”, they say (the monsters).’ Rose (2023) | |||||
Note that this is again the edge case mentioned in the introduction: 1st and 2nd person are verbal prefixes expressing both person and number, a slot distinct from the two monoexponential plural morphemes, which appear only with 3rd person (which, like all other person markers, is a prefix).
The Shipibo-Konibo plural marker -kan marks the plurality of S (24) or A (25), depending on the transitivity of the verb in which it occurs Valenzuela (2003: 521, 660).
| (24) | Ja | bari | noka-[a]i-tian | joni-bo | onitsapi-kan-a | iki. |
|---|---|---|---|---|---|---|
| that | sun:ABS | eclipse-S-DS | person-PL:ABS | suffer-PL-PP2 | AUX | |
| ‘Since there was an eclipse of the sun, the people were suffering. | ||||||
| (25) | Ja-tian | ani | nonti-n | westíora | atsa | xeati | chomo | |
|---|---|---|---|---|---|---|---|---|
| that-TEMP | big | canoe-LOC | one | yuca | drink | jar:ABS | ||
| na-yasan-kan-ai… | ||||||||
| INTERIOR-seat-PL-INC | ||||||||
| ‘Then they put a jar of yuca beer in the big canoe.’ | ||||||||
Valenzuela remarks on the peculiarity of a highly ergatively aligned language having a nominatively aligned plural marker.
Pilagá has a general plural suffix -lo which marks S/P plurality Vidal (2001:164f.).
| (26) | Plural S | |||
|---|---|---|---|---|
| a. | t-a-t-e-ge'-lo | so' | siyaca-dipi | |
| setA.3-go-PROGR-EP.V-DIR-PL | CL | woman-COLL | ||
| ‘The people are going away.’ | ||||
| b. | na-yapeg-eg'a-lo | hen | ñiyaq-pi | |
|---|---|---|---|---|
| setB.3-run.in.water-DIR-PL | DEM | fish-COLL | ||
| ‘The fish run with the flow.’ | ||||
| (27) | Plural P | |||
|---|---|---|---|---|
| a. | y-i-ketap-i-sem-lo | so' | yawo-dipi | |
| setA.3-point-DIR-PL | CL | woman-COLL | ||
| ‘He points up at the women.’ | ||||
| b. | ami' | si-yacana-lo | ||
|---|---|---|---|---|
| PRO.2PL | setA.1-call-OBJ.PL | |||
| ‘I will call you (pl.).’ | ||||
{ NA | ; separated list of morphemes with [ SAPE & oblique & subordinate clause & noun & pronoun & adjective & demonstrative & numeral & particle & adposition & adverb ] }
Teko has 3 markers, each separated above by the <;> symbol. One morpheme marks SA plurality, another one also marks SA plurality (borrowed from Cariban), and another morpheme marks P plurality, NP plurality, and plurality on possession or postposition.
Nuu-chah-nulth just has the one monoexponential plural marker used to mark SAP plurality in the clause, and nothing else.
The morpheme -reta which is optionally used to mark 3rd person plural on verbs (examples 1a-b and 6-8) is primarily used with nouns as a suffix with plural or collective meaning. It is also sometimes with pronouns and demonstratives (González 2005).
Derived features offer different views of the same data included in the main features. Often, this is done to group together some states of multi-state features, to capture specific similarities. In other cases, derived features are designed so that they are independent from other features used in the same computational analyses that assume feature independence, e.g. the Bayesian software sBayes, which detects areal signal Ranacher et al 2021.
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-03 contains <S> |
| no | if MonPl-01 is <yes> and MonPl-03 does not contain <S> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-03 contains <A> |
| no | if MonPl-01 is <yes> and MonPl-03 does not contain <A> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-03 contains <P> |
| no | if MonPl-01 is <yes> and MonPl-03 does not contain <P> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-05 contains <SA> |
| no | if MonPl-01 is <yes> and MonPl-05 does not contain <SA> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-05 contains <SP> |
| no | if MonPl-01 is <yes> and MonPl-05 does not contain <SP> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-05 contains <SAP> |
| no | if MonPl-01 is <yes> and MonPl-05 does not contain <SAP> |
{ NA | None | ; separated list of [ noun, pronoun, adjective, demonstrative, other ] }
| NA | if MonPl-01 is <no> |
| None | if MonPl-01 is <yes> and MonPl-06 does not contain <noun> or <pronoun> or <adjective> or <demonstrative> or <numeral> or <particle> or <adposition> or <adverb> |
| noun | if MonPl-01 is <yes> and MonPl-06 contains <noun> |
| pronoun | if MonPl-01 is <yes> and MonPl-06 contains <pronoun> |
| adjective | if MonPl-01 is <yes> and MonPl-06 contains <adjective> |
| demonstrative | if MonPl-01 is <yes> and MonPl-06 contains <demonstrative> |
| other | if MonPl-01 is <yes> and MonPl-06 contains <numeral> or <particle> or <adposition> or <adverb> |
{ NA | yes | no }
| NA | if MonPl-01 is <no> |
| yes | if MonPl-01 is <yes> and MonPl-06 contains <noun> |
| no | if MonPl-01 is <yes> and MonPl-06 does not contain <noun> |
We had two hypotheses which motivated coding for this feature: (1) that these kinds of plural markers are, as a whole, more common inside the Americas than outside; (2) that within the Americas, plural markers with different plurality targets pattern together. This represents the first attempt, to our knowledge, to define and test these questions in a large typological survey.
The first hypothesis was partially confirmed by our data. The overall feature of monoexponential plural marking (MonPl-01) is actually more present in our Australian sample (60%) than in our North American (50%) and South American (45%) samples. However, these three regions had far greater presence of the feature than other macroareas (Africa 24%; Eurasia 26%; Papunesia 24%).
The second hypothesis was also partially confirmed, with S/A/P (but not event) monoexponential plurals clearly clustering along the Peruvian coast and northwestern Australia (MonPl-05c), and S/P monoexponential plurals clustering in Rondônia (MonPl-05b). The use of monoexponential verbal plurals with nouns is also a clear areal feature of the Pacific Northwest Coast (although this is likely a side-effect of the region’s famously porous verb/noun distinction) and the northern Australian coast (MonPl-06b). However, these apparent clusterings exist alongside fairly widespread distributions of each typological feature around the world. It is possible that further investigation and different typological parameters would confirm hypothesis 2 more strongly.
Conceptualization: David Inman, Natalia Chousou-Polydouri, Marine Vuillermet, Balthasar Bickel
Data collection: Kellen Parker van Dam, Anna Graff, Selma Hardegger, David Inman, Natalia Chousou-Polydouri, Nora Muheim, Marine Vuillermet
Supervision of data collection: David Inman, Marine Vuillermet
Bright, William. 1957. The Karok Language. (University of California Publications in Linguistics, 13.) Berkeley and Los Angeles: University of California Press. xi+457pp.
Hollenbach, Elena Erickson de. 2013. Gramática del mixteco de Magdalena Peñasco (Sa'an Ñuu Savi). México, D.F.: Instituto Lingüístico de Verano. 475pp.
González, Hebe A. 2005. A grammar of Tapiete (Tupi-Guarani). Pittsburgh: University of Pittsburgh dissertation. (xv+403pp.)
Haude, Katharina. 2006. A Grammar of Movima. Zetten: Radboud Universiteit Nijmegen dissertation. (xvii+597pp.)
Hutchison, John P. 1981. The Kanuri Language: A Reference Grammar. Madison & Boston: Madison: African studies program, University of Wisconsin. 363pp.
Inman, David. 2021. The Exponence and Development of Plurals in Nuuchahnulth. Papers for the 56th International Conference on Salish and Neighbo(u)ring Languages. pp. 111–133.
Leer, Jeff. 1991. Evidence for a Northern Northwest Coast Language Area:Promiscuous Possessive Constructions in Haida, Eyak, and Aleut.
Michael, Lev David. 2008. Nanti evidential practice: Language, knowledge, and social action in an Amazonian society. University of Texas at Austin dissertation. (xxv+462pp.)
Mithun, Marianne. 1988. Lexical categories and the evolution of number marking. In Michael Hammond (ed.), Theoretical morphology: Approaches in modern linguistics. Nueva York: Academic Press.
Mithun, Marianne. 1999. The Languages of Native North America. (Cambridge Language Surveys.) Cambridge, New York: Cambridge University Press. xxi+773pp.
Ranacher, Peter, Nico Neureiter, Rik van Gijn, Barbara Sonnenhauser, Anastasia Escher, Robert Weibel, Pieter Muysken & Balthasar Bickel. 2021. Contact-tracing in cultural evolution: a Bayesian mixture model to detect geographic areas of language contact. Journal of The Royal Society Interface 18(181). 20201031.
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.
Tarpent, Marie-Lucie. 1983. Morphophonemics of Nisgha plural formation: A step towards Proto-Tsimshian reconstruction. Kansas Working Papers in Linguistics 8(2). University of Kansas. Linguistics Graduate Student Association.
Valenzuela, Pilar M. 2003. Transitivity in Shipibo-Konibo Grammar. Eugene: University of Oregon dissertation. (xxxiv+1029pp.)
Vidal, Alejandra. 2001. Pilagá Grammar (Guaykuruan Family, Argentina). Eugene: University of Oregon dissertation. (xviii+391pp.)
Vuillermet, Marine. 2012. A Grammar of Ese Ejja, a Takanan language of the Bolivian Amazon. Université Lumière Lyon 2 dissertation. (xvi+736pp.)
A more exact term for monoexponential verbal plurals would be monoexponential non-NP plurals or even monoexponential non-exclusively NP plurals. We will use the term “monoexponential (verbal) plurals” for the sake of conciseness. ↩
Note that all targeted morphemes are not strictly monoexponential, as most encode one (or several) particular syntactic role(s). We nevertheless consider them “monoexponential plurals” because the additional information about role is purely syntactic, while in the semantic domain, only number is specified. ↩
By “random” here, we mean that the systems do not exhaustively target semantic classes of verbs, for instance only some members of the posture or the motion semantic classes. ↩
Rose notes that the use of -kom with verbs is rare in her corpus, but it is common in elicitation (when the object is animate). In her corpus there are isolated uses of -kom with verbs where it seems to pluralize the S argument for some persons (1st inclusive and generic). ↩
According to Rose, kupa seems to be an independent word and it appears at the end of the clause. It seems to have been borrowed from the closely related language Wayampi [waya1270] (Tupian; Brazil, French Guiana). It is much rarer than -(o)ŋ. ↩
Glosses absent from the Leipzig glossing rules in the following examples: 0: semantically empty; PSSB: possible mood, i.e. an “if” clause; NOW: now or then, i.e. the next event in a sequence; STRG: strong mood, i.e. strong claim to factuality. ↩