See data and maps.
Mukhin, T. (2023). “Demonstratives”. In: Typological Atlas of the Languages of Daghestan (TALD). Ed. by M. Daniel, K. Filatov, T. Maisak, G. Moroz, T. Mukhin, C. Naccarato and S. Verhees. Moscow: Linguistic Convergence Laboratory, NRU HSE. DOI: 10.5281/zenodo.6807070. http://lingconlab.ru/dagatlas.
@incollection{mukhin2023,
title = {Demonstratives},
author = {Timofey Mukhin},
year = {2023},
editor = {Michael Daniel and Konstantin Filatov and Timur Maisak and George Moroz and Timofey Mukhin and Chiara Naccarato and Samira Verhees},
publisher = {Linguistic Convergence Laboratory, NRU HSE},
address = {Moscow},
booktitle = {Typological Atlas of the Languages of Daghestan (TALD)},
url = {http://lingconlab.ru/dagatlas},
doi = {10.5281/zenodo.6807070},
}This chapter is an overview of the demonstrative systems of the languages of Daghestan and adjacent territories. There is a lot of cross-linguistic variation in the systems of demonstrative pronouns. According to typological studies of demonstratives ((Anderson, Keenan 1985); (Diessel 1999); (Dixon 2003); (Levinson et al. 2018)), demonstrative systems may not only differ in terms of the number of pronouns they have, but also in terms of the functions they perform. In addition to distance, demonstratives can mark additional deictic features, for example elevation or visibility. For East Caucasian languages there is a dedicated detailed typological study of demonstrative systems and their diachrony by Schulze (2003).
Below, I address both the most typical and rare features of demonstrative systems expressed by morphological markers (by a dedicated root or a derivational affix). I focus on demonstrative pronouns, leaving aside all other lexical categories of demonstratives (such as demonstrative determiners, adverbs and identifiers) (Diessel 1999: 57).
In Section 2, I look at the number of distance contrasts in demonstrative systems. In Section 3, I focus on systems that show a more-than-two-way distance contrast and divide them according to whether the deictic center in relation to which demonstratives are determined changes (speaker vs addressee) or not (only speaker). In Section 4, I describe the most exotic and typologically rare systems attested in the languages of the sample in terms of distance meaning. In Sections 5, I discuss additional parameters that demonstrative can distinguish, such as elevation. In Section 6 and Section 7, I give examples of dedicated marking of non-deictic uses of demonstratives. Section 8 is the conclusion.
The current chapter investigates the distribution of demonstrative systems depending on the number of distance contrasts expressed by demonstratives.
In the simplest cases, demonstrative systems consist of only one distance-neutral pronoun (cf. German or Supyire systems (Diessel 1999: 38-39)). Alternatively, they may distinguish between two (as in English this vs that) or more degrees of remoteness. In Udi (Lezgic), for example, demonstratives distinguish three degrees of distance from the speaker: me ‘close to the speaker’, ka ‘not close to the speaker’, t’e ‘outside the region of the speaker’ (Schulze 2008: 253). Systems that distinguish more than three contrasts are attested in many languages of the world (see the map in (Diessel 2013)) and the map in (Skirgård, Haynie, Blasi, et al. 2023), (Skirgård, Haynie, Hammarström, et al. 2023)), including languages of the East Caucasian family.
In the case of polycentric systems (see Orientation of demonstrative systems), despite the fact that in terms of distance from the deictic center addressee-based terms may denote the same distance as the corresponding speaker-based term, I counted each demonstrative as expressing separate distance contrast.
Map 1 shows the number of distance contrasts expressed by demonstratives in the languages of the sample.
The number of contrasts not for all systems corresponds to the number of lexemes considered as a demonstrative. For example, in addition to distance-marking demonstratives, some languages of the sample have a demonstrative that is described as ‘aforementioned’ (see Section 6). For example, Ingush (Nakh) has the demonstratives je ‘this’, dwaara ‘that’, and the demonstrative yz ‘aforementioned’, which does not express a distance contrast but is only used for reference tracking in discourse (Nichols 2011: 179). 1 The same seems to be true for Chechen (Nakh) (Nichols 1994: 34). Another such case is Kumyk (Turkic) that has two dedicated stems which differ by givenness but have the same deictic meaning, see Section 5.3. Such demonstratives were not taken into account in counting the number of contrasts, see the corresponding sections for the discussion.
I also do not consider demonstratives whose meaning is not clear from the description, but is interpreted by the author as an independent root expressing a separate contrast. Such cases include Azerbaijani (Širaliev, Sevortjan 1971), Bagvalal (Magomedova 2004) (here I use data from another source), and Budukh (Talibov 2007). These languages are annotated as having two or possibly more distance contrasts (shown as >=2 on Map 1), because from the description it is not clear whether additional demonstratives involve a distance contrast.
Most of the languages feature either two- or three-way distance contrasts. In the languages of Daghestan, languages with a two-way contrast are almost twice less common than those with a three-way opposition, whereas in the languages of the world, two-way contrasts are the most frequent option (Diessel 2013).
In the sample, there are four languages (all belonging to the Andic branch) in which demonstratives express more than three distance contrasts: Godoberi, Botlikh, the dialect of Andi spoken in the village of Zilo and the dialect of Bagvalal spoken in the village of Kvanada.
Languages belonging to the same branch often diverge in terms of the number of contrasts expressed by demonstratives. Only Lezgic languages seem to be consistent in having three-way systems (except for Budukh, where the interpretation is unclear).
Languages of the Andic branch are the richest in terms of distance contrasts.
The current chapter investigates the distribution of demonstrative systems depending on demonstratives use the same deictic center or different ones.
Anderson and Keenan (1985: 282) and Schulze (2003: 297) classify demonstrative systems with three-way contrast into two types: monocentric (or distance-oriented) and polycentric (person-oriented). 2 The main difference between the two types lies in the interpretation of the ‘medial’ demonstrative. In monocentric systems the ‘medial’ demonstrative is used when pointing to an object that is at a medium distance from the speaker, i.e. neither far nor close to the speaker (e. g. in Udi (Schulze 2008: 253)), while in polycentric systems it is defined relative to the position of the addressee and is used to indicate objects that are close to the addressee (e. g. in Kina Rutul (Lezgic) (Mukhin, Nasledskova 2021)), cf. Table 1.
| monocentric (Udi) | polycentric (Rutul) | |||
|---|---|---|---|---|
| me ‘close to the speaker’ | S-prox | proximal | S-prox | mi ‘close to the speaker’ |
| ka ‘not close to the speaker’ | S-dist | medial | A-prox | ha ‘close to the addressee’ |
| t’e ‘outside the region of the speaker’ | S-far | distal | S&A-dist | ti ‘outside the region of both the speaker and the addressee’ |
Both types may distinguish a more-than-three-way contrast, but such rich systems are usually polycentric (Diessel 1999: 40).
Both mono- and polycentric systems are attested in the languages of the sample. Map 2 shows the distribution of the two types of demonstrative systems.
In Map 3, a detailed classification of systems attested in the languages of the sample is presented with an annotation of the distance meanings of each demonstrative root (excluding elevational demonstratives) as explained in the source (for those with a clear deictic meaning), independently of the cross-linguistic validity of this explanation. 3 In the annotation, I follow the approach of Levinson (2018: 36): S-prox = speaker-based proximal term, S-dist = speaker-based distal term, S-far = speaker far-distal term, A-prox = addressee-based proximal, S&A-prox/dist = speaker and addressee based proximal/distal. I also use the labels A-dist = addressee-based distal and S&A-far = speaker and addressee based far distal, but these are not present in Levinson’s approach.
In the sample, there are four languages which have more than three distance contrasts: Zilo Andi, Kvanada Bagvalal, Godoberi and Botlikh (all of them belong to the Andic branch). All systems are polycentric, which agrees with the observations by (Diessel 1999: 40), (Diessel 2013), (Levinson 2018: 37), according to which systems that distinguish more than three distance contrasts are almost always polycentric. 4
Zilo Andi and Kvanada Bagvalal demonstratives express a four-way contrast, Godoberi and Botlikh have yet more complex systems with five-way contrast. The demonstrative systems for all four languages are presented in Tables 2-5.
| S-prox | A-prox | S&A-dist | S&A-far |
|---|---|---|---|
| ‘near the speaker’ | ‘neat the addresse’ | ‘distal’ | ‘distal relatively further away’ |
| ho- | he- | hi(ni)di | hundo- |
| S-prox | S-dist | S-far | S&A-dist |
|---|---|---|---|
| ‘this one, close to the speaker’ | ‘this one, remote from the speaker’ | ‘that one, remote from the speaker’ | ‘that one, remote from the speaker and the listener’ |
| a- | o- | u- | uhu- |
| S-prox | S-dist | A-prox | A-dist | S&A-dist |
|---|---|---|---|---|
| ‘this, close to the speaker’ | ‘at some distance from the speaker’ | ‘this, close to the listener’ | ‘at some distance from the listener’ | ‘aforementioned or that far away’ |
| ha- | hada- | hu- | hudo- | ho- |
| S-prox | S-dist | A-prox | S&A-dist | S&A-far |
|---|---|---|---|---|
| ‘X is in the direct vicinity of the speaker’ | ‘X is in the vicinity of the speaker, a little to the side’ | ‘X is close(r) to the addressee’ | ‘X is not in the vicinity of the speaker or the addressee’ | ‘X is not in the vicinity of the speaker or the addressee relatively further away’ |
| ha- | ga- | hu- | go- | do- |
Rich systems are typologically rare (in Diessel’s sample of 234 languages (2013) only 9 languages have a four-way contrast and 4 languages have a five-way contrast). A four-contrast system similar to that of Zilo Andi is reported for Quileute (Chimakuan) (Diessel 1999: 41) and Hausa (Chadic, Afro-Asiatic) (Diessel 2013), but no system was found to be similar to that of Kvanada Bagvalal. I did not find five-contrast systems similar to those of Godoberi and Botlikh in Diessel’s sample (five-contrast systems in Koasati, Malagasy, Maricopa and Navajo are different in terms of the contrasts involved), but this can be due to the methodological differences in descriptive grammars. Neither of these rich Daghestanian systems corresponds to the types described in (Levinson 2018: 36).
The systems of Kvanada Bagvalal, Godoberi and Botlikh are unexpected since, as far as can be judged from the descriptions, the deictic meanings of some demonstratives overlap (for example S&A-dist vs S-dist/S-far in Kvanada Bagvalal, A-dist vs S-prox/S&A-dist in Godoberi, S-dist vs S&A-dist/S&A-far in Botlikh) and it is not always obvious which parameters influence the choice of a particular demonstrative.
Typologically unusual is also the system of Kryz (Lezgic), which has a three-way system, as shown in Table 6. Unlike the canonical three-term polycentric system, Kryz lacks S&A-dist and has S-dist instead.
| S-prox | A-prox | S-dist |
|---|---|---|
| ‘sphere of the speaker, close’ | ‘sphere of the interlocutor’ | ‘sphere of the speaker, distant’ |
| u- | am | a- |
The systems discussed in this section show patterns that are not attested or are very rare elsewhere. Because the data for the classification is obtained from grammars rather than results from controlled and comparable experiments, it needs to be additionally checked.
This section discusses some additional parameters expressed by demonstratives in East Caucasian languages, namely: elevation, visibility and givenness. Unlike elevation, consistent expression of visibility and givenness is rare in languages of the sample.
For some languages, it is claimed that elevation and visibility play a role in the choice of demonstratives. However, as it is not always clear how strong is the correlation between the relevant component of meaning and the form of the demonstrative, I only consider languages with what is described as a dedicated morphological expression of elevation and/or visibility.
Besides distance contrast, demonstrative systems can additionally distinguish elevation, indicating location above or below the level of the deictic center. Elevation can be expressed by independent pronouns, as in Archi (Lezgic) (see Table 7), or by attaching an affix to the demonstrative stem, 8 as in Tindi (Andic) (see Table 8). The current chapter investigates the distribution of demonstrative systems depending on presence/absence of a dedicated elevational demonstrative pronoun or affix.
For a world-wide sample see the map in (Skirgård, Haynie, Blasi, et al. 2023), (Skirgård, Haynie, Hammarström, et al. 2023).
| S-prox | A-prox | S&A-dist | above | below |
|---|---|---|---|---|
| ja- | jamu- | to- | ʁodo- | godo- |
| S-prox | S-dist | |
|---|---|---|
| at the same level | a-ja- (< ada-) | o-ja- (< oda-) |
| above | a-ɬa- | o-ɬa- |
| below | a-ɟa- | o-ɟa- |
In languages with elevation affixes, the latter can usually be attached to different distance stems, and thus they distinguish elevation relative to different deictic centers. For most languages, it is not specified whether the affixes can be added to all demonstrative stems (e.g. in Tindi) or only to some (e.g. Zilo Andi, where S-prox and A-prox elevational forms are not attested (Kaye et al. 2020: 12-13)). Therefore, I only looked at the presence/absence of elevation markers.
In languages where elevation is expressed by dedicated pronouns, they usually have non-proximal distance meaning, which confirms the results presented in (Forker 2020: 11). However, it is often not clear from the available descriptions relative to which deictic center elevational demonstratives are defined and whether the center can change (see the discussion in (Forker 2020: 10-11)). Therefore, I do not specify this in the markup and use the labels ‘above’ and ‘below’.
Languages with dedicated elevation pronouns usually have demonstratives that express both meanings (above and below the deictic center). The only exception is Godoberi, which, according to the available description, lacks the above-demonstrative and has only a demonstrative he- meaning ‘below’ (which is rarely used (Gisatullina, Toldova 1996: 42)).
Demonstrative systems like Archi, which has independent pronouns for expressing elevation, can have separate stems for each distance and elevation meaning, but in my sample there is one language that expresses different meanings with the same stem. The demonstrative le in Agul (Lezgic) is used both as A-prox demonstrative (1) and as the above-demonstrative (2). Table 9 presents the demonstrative system of Agul.
| le | fi | e, | fi | e | we | χil.i-ʔ | a-je-f? |
| deml | what | cop | what | cop | your | hand-in | (in)be-part-s |
| le | ʁ˳ad.i-l | al-e | ber | faqšaw |
| deml | roof-super | (super)be-part | shovel | re:bring(imp)deml |
| S-prox | A-prox/above | S&A-dist | below |
|---|---|---|---|
| me- | le- | te- | ge- |
Map 4 shows the presence of special demonstrative affixes or independent demonstratives that express the category of elevation.
As shown on the map, elevation is not a rare feature in East Caucasian languages. Almost half of the languages in the sample distinguish elevation by special demonstratives or affixes.
All languages with dedicated elevation pronouns are polycentric, with the exception of Lezgian (Haspelmath 1993: 184, 190).
In terms of individual branches, Tsezic languages do not distinguish elevation. Lezgic languages express elevation by dedicated stems (if at all). The use of special affixes is a characteristic feature of Andic languages, with the exception of Godoberi, which expresses elevation by dedicated demonstratives, and Botlikh, which does not distinguish elevation.
The visual accessibility of the object to the speech act participants is a common demonstrative parameter, typical of, e.g., the languages of the Americas (Diessel 1999: 41). There is only one language in my sample where, according to the available description, visibility is consistently distinguished for all demonstratives: 10 in Bezhta (Tsezic) there is a dedicated visibility prefix wa- that can be attached to all three demonstratives of the system, cf. Table 10.
For a world-wide sample see the map in (Skirgård, Haynie, Blasi, et al. 2023), (Skirgård, Haynie, Hammarström, et al. 2023).
| S-prox | A-prox | S&A-dist | |
|---|---|---|---|
| invisible | hudi | huli | hugi |
| visible | wa-hdi | wa-hli | wa-hgi |
There is one language in the sample where demonstratives are contrasted by whether the object was previously mentioned or not. According to (Dmitriev 1940: 77-79), Kumyk (Turkic) has two S-prox pronouns, one of which is used when the referent is known to the participants and the other is used to denote a new one, cf. Table 11.
| S-prox | S-prox | S-dist | S-far |
|---|---|---|---|
| ‘close and known object’ | ‘close and unknown object’ | ‘an object that is farther than (referred by) šu, but closer than (referred by) o’ | ‘remote object’ |
| bu | šu | šo | o |
Systems with dedicated demonstrative forms for (un)given referenets are attested in other languages, for example in Hausa (Chadic, Afro-Asiatic) (Anderson, Keenan 1985: 289).
In addition to the primary deictic function, demonstratives can often be used in other functions (Diessel 1999: 6), e.g. anaphorically as the 3rd person pronoun. Below I provide examples of languages that distinguish the use of demonstratives in the anaphoric function (following (Diessel 1999), I call anaphoric those endophoric uses that refer to NP antecedents, which is also the most common endophoric function).
East Caucasian languages mostly do not have dedicated anaphoric pronouns and use demonstratives instead, including in independent syntactic positions, e.g. He had a house a long time ago where other languages like English use dedicated forms. Usually demonstratives used anaphorically do not differ morphologically from pronouns used in the deictic function, but some exceptions are discussed below.
Several languages are reported to have a demonstrative that is mainly (but not exclusively) used anaphorically as the third person pronoun. Such demonstratives are often described as expressing the meaning ‘aforementioned’. Apparently, in such languages, the ‘aforementioned’ demonstrative is most frequent in anaphoric contexts, although they are not limited to this use and can also be used deictically.
Map 5 presents which demonstrative is used more frequently as the 3rd person personal pronoun if this information is specialized in the description. In all East Caucasian languages (apparently with the exception of Chechen and Ingush) more than one demonstrative stem is used anaphorically (usually sources do not specify whether each demonstrative stem can be used anaphorically, including elevational ones if there are any, or only a few).
As was mentioned in Section 2, in Chechen and Ingush 3rd person pronouns seem to have lost their original deictic meanings and are used only anaphorically. According to (Nichols 2011: 34), Chechen originally had a three-term person oriented system consisting of S-prox hara ‘this’, S-dist iza ‘aformentioned’, S-far dʔaːranig ‘that’ that transformed into two-term S-prox hara ‘this’, S-dist dʔaːranig ‘that’ with one with iza ‘aformentioned’ grammaticalized into anaphoric function. The ‘aformentioned’ demonstrative yz in Ingush is described as a neutral demonstrative but it is not clear from the description whether it preserves deictic uses. For these languages distance meanings of the demonstrative were annotated in brackets (e.g. (S-dist)), since probably in the modern language, the demonstrative pronoun used anaphorically is no longer used deictically.
As can be seen from Map 5, there is no language that uses the S-prox demonstrative as the most frequent anaphoric mean, with the possible exception of Kvanada Bagvalal where there are two such demonstratives S-prox and S-dist (A. E. Kibrik et al. 2001: 161, 658, 670) (I do not show this data on the map). In three-term polycentric systems the S&A dist demonstrative prevail, while in monocentric ones S-dist and S-far compete.
This partially disagrees with the observation by Anderson and Keenan that at least in languages with three-terms systems usually the middle term (A-prox or S-dist) is grammaticalized as the 3rd person pronoun (Diessel 1999: 39).
Other cases where anaphoric uses are somehow separated from deictic ones include Hinuq (Tsezic) that has three demonstrative series, the first of which is almost exclusively used anaphorically (Forker 2013: 132-140), cf. Table 12. 11
| S-prox | S-dist | |
|---|---|---|
| 1st series | ha- | hay- |
| 2nd series | hiba(ha)- | hiba(ha)y- |
| 3rd series | iza(ha)- | iza(ha)y- |
It is often not clear from the available descriptions whether the ‘aforementioned’ pronouns are strictly limited to anaphoric uses and cannot be used deictically.
Among the languages of the sample, I found only one language (with the exceptions of Chechen and Ingush discussed above) where demonstratives are reported to consistently differ morphologically depending on what function - deictic or anaphoric - they convey. Hunzib has a dedicated prefix yǝ-, which is attached when a demonstrative is used anaphorically (Berg 1995: 60), cf. Table 13.
| S-prox | A-prox | S&A-dist | |
|---|---|---|---|
| deictic | bǝd | bǝl | ǝg |
| anaphoric | yǝ-bǝd | NA | yǝ-ʔǝg |
In this Section, I discuss the reflexive uses of pronouns that are also used as demonstratives, although usually the reflexive function is not considered as a function of demonstratives.
In Tsez and Khwarshi (both Tsezic), there is a pronoun that can be used both as a demonstrative and reflexive pronoun.
Khwarshi has a rather complex system that consists of S-prox demonstrative idu ‘this’ and S-dist inu ‘that’. In addition, there is a pronoun žu described as S-dist demonstrative whose relation to the S-dist inu in terms of distance is unclear (Khalilova 2009: 143). 12 At the same time, the pronoun žu ‘that’ with the emphatic particle -č (optional in the oblique forms) is used in reflexive third person contexts, cf. (3) where its suppletive oblique form ise is used (Khalilova 2009: 161):
| obu-t’-i | ise.ise | xwaλ-i. |
| father-obl-erg | refl.erg | shave-pst.w |
Another example is Asakh Tsez, which also has the S-prox demonstrative ža with the meaning ‘this’ that can be used as a third person reflexive pronoun and as an NP modifier (Polinsky 2015: 20). Cf. (4) and (5) where the demonstrative ža ’this’ (neła in oblique) is used both as a demonstrative modifier and as an independent reflexive pronoun.
| Dunyal | nełaq | ža | šet’u | b-uti-x. |
| earth.abs.iii | refl.nI.poss.ess | dem | around | iii-turn-prs |
| ža | k’et’u |
| dem | cat.abs.iii |
| neła | k’et’u-r |
| dem.nI.obl | cat-lat |
The use of the same pronoun in both reflexive and demonstrative contexts is rare in the languages of the sample (and perhaps even in other languages of the world). However, historically demonstrative and reflexive pronouns are closely related, since both often develop from the same source: intensifier pronouns (König, Siemund 2000: 58).
The demonstratives žu ‘that’ and ža ‘this’ are described as part of the demonstrative system, 13 but morphologically differ from other demonstratives. It is not clear enough from the descriptions how common the demonstratives žu ‘that’ and ža ‘this’ are in each of the functions, as well as how they relate to other demonstratives of the system in terms of deictic meaning. The status of the pronoun itself is also not entirely clear, since different sources consider different functions to be the main one. In many other languages of the sample there are pronouns that are used as 3rd person reflexives, but not as demonstratives morphologically similar to žu ‘that’ and ža ‘this’ in Tsez and Khwarshi, for example Avar (Avar) and Zilo Andi (Andic) ži ‘self’ ((Alekseev, Ataev 1997: 55); (Kaye et al. 2020: 11).
The demonstrative systems in East-Caucasian languages are rich in terms of the number of distance contrasts they may distinguish. They often also encode a number of additional deictic features. Some of the systems mentioned above are quite unusual from a typological point of view.
In Daghestan and its adjacent territories, the most common type of systems encodes a three-way contrast, whereas elsewhere in the languages of the world systems with a two-way contrast prevail.
In the sampled languages, monocentric and polycentric systems are almost evenly distributed, or at least are equally reported in grammatical descriptions. The distribution does not show genetic or areal patterns: in Tsezic, Lezgic and Andic branches both types are present.
The richest languages in terms of the number of contrasts belong to the Andic branch (Zilo Andi, Kvanada Bagvalal, Godoberi and Botlikh). Despite close genetic relationship, the inventory of demonstratives differs greatly in terms of particular meanings of the lexemes.
Elevation is a common feature of demonstrative systems in the languages of Daghestan and its adjacent territories. Affixal marking of elevation is characteristic of Andic languages, while Lezgic languages express it by dedicated stems, and Tsezic languages are not reported to have it altogether.
Distinctions in terms of visibility, givenness and dedicated anaphoric marking are rare. A visibility affix is only attested in Bezhta, whereas an anaphoric affix is attested in Hunzib. Only Kumyk uses different S-prox demonstrative pronouns depending on whether the referent was mentioned previously or not.
According to data from written sources, there is no language where the S-prox demonstrative pronoun is used as the main 3rd person pronoun.
In two languages of the Tsezic branch Tsez and Khwarshi, there is a pronoun which is used as both a demonstrative and a reflexive pronoun. This may be a rare feature not only in the languages of the sample, but also in other languages, and this requires a separate investigation.
As for languages not from the East Caucasian family, they do not distinguish elevation and have a three-way contrast (with the exception of Tat (IE) and possibly Nogai (Turkic), where it is a two-way contrast). Both mono- and polycentric systems are attested in these languages.
Since the deictic meanings of demonstratives imply a reference to an object in the surrounding world, and the data for this overview are mostly taken from written sources rather than experimental data, some seemingly exotic systems deserve a separate experimental study.
12 — ; 2 — second person; abs — absolutive; add — additive; add1 — ; add2 — ; adj — adjective; attr — attributive; cm — class marker; com — comitative; cop — copula; d — determiner; dem — demonstrative; deml — ; emph — emphatic; erg — ergative; ess — essive; i — class I; iii — class III; imp — imperative; in — inessive; int — interrogative; iv — class IV; j — ; lat — lative; lnk — linker; m — masculine; n — neuter; n2 — ; num — numeral; numcm — ; numn2 — ; obl — oblique; ord — ordinal; part — partitive; pfv — perfective; poss — possessive; prs — present; pst — past; re — refactive; refl — reflexive; s — single argument of canonical intransitive verb; sg — singular; sgemph — ; super — superlocative; w — witnessed
Historically it originated from the medial-distance demonstrative (Nichols 2011: 658).↩︎
My data contain three languages for which a polycentric system with only two demonstratives described as ‘this one, from the 1st person sphere’ and ‘that one from the 2nd person sphere’ is reported: Akhvakh, Chamalal, and Tindi (Andic) ((Magomedova, Abdulaeva 2007: 681-682); (Magomedova 1999: 423); (Magomedova 2012: 126). This interpretation is doubtful, since in such a system the demonstratives would only cover part of the distance scale, and differs from other descriptions ((Creissels 2020: 13) for Akhvakh; (Bokarëv 1949: 66) for Chamalal; (Authier 2020: 10-11) for Tindi)↩︎
The distribution of languages by type does not always correspond to the classification in (Schulze 2003: 302). For example, there are no languages with a monocentric system with three-way contrast (S-prox, S-dist, S-far), only two-way. The reason for this difference may lie in the methodology adopted: most of the systems that I marked as S-prox, S-dist, S-far are systems that are not described in sufficient detail (for example, when the deictic meaning is given using only Russian pronouns or without explicit indication of whom the reference is made to: the speaker or the addressee, cf. Azerbaijani bu ‘indicates a nearby item’, o ‘indicates a distant object’, həmin ‘this/that one’ (Širaliev, Sevortjan 1971: 82)). That is, I used the label S-prox, S-dist, S-far as the default value for three-way systems.↩︎
Some examples of monocentric systems marking more than three-way contrast are attested in (Anderson, Keenan 1985: 286-288).↩︎
The demonstrative ho- is also described as conveying the meaning ‘aforementioned’ but as far as can be judged from the description, it is also used in deictic function.↩︎
Godoberi also has an elevational demonstrative, see Section 5.1.↩︎
The interpretation of the meaning of the demonstrative -ga as S-dist is arguable due to the ‘a little to the side’ part of the translation. I interpreted this as a difference in distance but, perhaps, it means that the demonstrative -ga differs from the demonstrative ha- in visibility. Excluding the demonstrative -ga the Botlikh system coincides with the Zilo Andi one.↩︎
When the elevation is expressed by affixes, location at the level of the deictic center is usually also marked by a special affix.↩︎
Note that Tindi demonstratives a- (S-prox) and o- (S-dist) are also used without elevation affixes -d, -ɬ, -ɟ.↩︎
There are quite a lot of languages for which visibility is mentioned as determining the use of demonstratives (distal terms is often associated with invisible objects), but since it is not marked by a dedicated morpheme in these languages, I do not consider them here.↩︎
According to (Forker 2013: 136, 139), the second series is usually used in exophora/situational anaphora contexts (A. A. Kibrik 2011: 511), where ‘the reference is immediately given’, while the demonstratives of the third series occur only in the deictic function (and in direct speech).↩︎
From these pronouns, other demonstratives of the S-prox and S-dist series are derived (e.g. awedu vs. owenu, hobodu vs. homonu) (Khalilova 2009: 146). The difference in their meaning is not clear from the description. Although the author clarifies that in a previous study by Imnajšvili (Imnajšvili 1963: 117), the meanings of hobodu and homonu were described as ‘close to the 2nd person’ and ‘far from the 2nd person’ , in the modern language this difference is not attested.↩︎
Interestingly, they have different deictic meanings, although the languages are related.↩︎