Navigation – Plan du site

AccueilNuméros69From copula to focus, vice versa,...

From copula to focus, vice versa, or neither?

Information structure and non-verbal predication in Mano
De la copule vers le focus, ou le vice-versa, ou ni l’un ni l’autre ? La structure informationnelle et la prédication non-verbale en mano
Мария Леонидовна Хачатурьян. Показатель фокуса из копулы? Наоборт? Ни то, ни другое? Информационная структура и неглагольная предикация в мано
Maria Khachaturyan
p. 87-120

Résumés

Cet article étudie le système de prédication non-verbale du mano, une langue mandé-sud, basé sur des données de première main incluant des données élicitées et un corpus de discours spontané. La principale question analytique que j'aborde concerne la relation entre la prédication non-verbale et le marquage du focus. Je soutiens qu’en mano, la valeur de focus émerge comme une interprétation pragmatique de constructions non-verbales identificatrices ou présentatives dans des conditions contextuelles particulières. En conséquence, il peut ne pas être nécessaire de distinguer les marqueurs de focalisation en tant que classe syntaxique et sémantique distincte des prédicateurs non-verbaux. Cette discussion vise à éclairer l'origine et la fonction interactionnelle du focus dans les langues mandé.

Haut de page

Texte intégral

1. Introduction

1Among the treasures that Mande languages have offered to linguistic typology are non-verbal predications: both typological peculiarities of individual constructions (Creissels 2022) and the remarkable diversity and originality of systems of non-verbal predication in individual languages (Vydrin 2020).

2This paper contributes to the discussion of Mande non-verbal predication by describing the system of Mano, a South Mande language spoken in Guinea and Liberia (ISO-639 code: mev). The data is drawn primarily from the Maa dialect of the language spoken around Nzérékoré. The main analytical question that I address concerns the relationship between non-verbal predication and focus marking. Let’s consider the following examples:

3Identifying construction in Mano

  • 1 att stands for attention management. The function of attention management is not discussed in this (...)

(1)

Gààzù

dōó

wɛ̄

lɛ́

nɔ́

wɛ̄.

mirror

one

dem

att1

just

dem

4It is the same mirror’ [MOC]

5Focus construction in Mano

(2)

Wìì

lɛ́

zɛ̄

ɛ̄.

animal

att

1sg.pst>3sg

kill

dem

6‘[I did not hit a person,] it is an animal that I killed.’ [MOC]

7The examples (1–2) include the same marker, lɛ́, in different functions: a function of identifying a non-verbal predicator in (1) and serving as a focus marker in (2). In this paper, I argue that focus meaning arises as a pragmatic interpretation of identifying or presentative constructions with non-verbal predicators under particular contextual conditions. As a result, focus markers may not need to be distinguished as a syntactic and semantic class separate from non-verbal predicators. This discussion aims to shed light on the synchronic and diachronic interaction between focus marking and non-verbal predication in Mande.

8The data in this paper are a mix of elicited (marked as [el.]), semi-naturalistic (continuous recordings of an experiment with visual stimuli, marked with [reflexive experiment]; the exact stimuli are given in the Appendix) and naturalistic data, coming from a corpus of narratives, conversations and religious ceremonies [MOC]. In one case, the published translation of the New Testament was also used (UBS 1978).

9I begin by introducing non-verbal predication in Section 2. Section 2.2 describes non-verbal predication with PPMs, Section 2.3 discusses the verb kɛ̄ ‘to be’ and Section 2.4 introduces non-verbal predication with negative copulas. In Section 3, I focus on presentative and identifying predicators. I discuss basic constructions with these markers in Section 3.1 and different syntactic extensions, including those with a clause, which, in some contexts, receive a focal reading, in Section 3.2. I discuss the findings in Section 4, where I place the Mano data in a broader typological and theoretical context.

2. Non-verbal predication in Mano

2.1 General remarks

10Following Hengeveld (2011), I define non-verbal predication as predication where the predicative value is expressed by an element whose default use is not predicative and which is therefore not a verb. Such elements are most commonly referential or adverbial expressions. In addition to the predicative part, non-verbal predication usually includes another element — an argument. Non-verbal predication also typically includes supporting predicative elements. Copulas, in Hengeveld’s terminology, are defined as supporting predicative elements that are semantically empty and whose main function is to enable a non-verbal predicate to fulfill its function. If a supporting predicative element can, in a different context, also combine with a verb and/or is not semantically empty but has an additional verbal or temporal/aspectual/modal meaning, the term “copula” does not apply. Instead, the predicator is ascribed to a broader class of auxiliaries.

  • 2 In Valentin Vydrin’s terms, PPMs are defined as elements which occur only in verbal predications. I (...)

11Mano shows several different strategies for non-verbal predication constructions. Different types of Mano non-verbal predications involve different types of predicative elements. The most prominent type — and the most frequent, also covering most of the semantic classes of non-verbal predication — are constructions involving elements which, in the Mandeist tradition, are called pronominal predicative markers, or PPMs (Vydrin 2020).2 These elements form a central part of a finite construction, expressing tense, aspect, modality and polarity, but also function as a site of subject indexation. Constructions with PPMs of the existential series, which form the bulk of non-verbal predication in Mano, are addressed in Section 2.2. For non-stative aspects and non-present tenses, other PPMs are used in combination with the verb kɛ̄ ‘to be’ (Section 2.3). The Mano negative copula wɔ́ is addressed in Section 2.4.

  • 3 According to Denis Creissels (2022, p.c.), they are semantically similar to either referential iden (...)

12In addition to the aforementioned types, another cross-linguistically common type of non-verbal predication includes predications with an argument and no other predicative part besides the supporting predicative element (ex. 1).3 The most central functions of such predication are referent identification and ostensive presentation. The predicative elements are ascribed to the class of predicative demonstratives (Killian 2022). Such predications, addressed in Section 3, comprise the main subject of this paper.

2.2 Non-verbal predication with existential PPMs

13The Mano series of existential predicative markers can be used in verbal and in non-verbal predications. The word order in Mano verbal predications is Subj-ppm-V, where Subj is the subject, ppm is the pronominal predicative marker and V is the verb. (In transitive verbal clauses, there is another obligatory element, namely, the direct object in the position between the PPM and the verb.) Similarly, in non-verbal predications with PPMs, the word order is Subj-ppm-pred, where pred is the non-verbal predicative element. The subject np is optional, so the subject may be expressed solely by the person-number index on the PPM. In a typical non-verbal predication describing a situation in the present (as in I am a teacher), the PPM series used is the existential one. Table 1 below lists all the PPMs of that series. The second row gives portmanteau forms merged with a 3sg pronoun.

1sg

2sg

3sg

1pl

2pl

3pl

Basic series

ŋ̄

lɛ̄

kō

kā

Portmanteau series (merged with 3sg pronoun)

māā

ɓāā

lāā

kɔ̄āā

kāā

wāā

14Table 1. Existential PPM series

15The PPM construction is used for most non-verbal predications types, including those involving referential identity, class membership, attribution and possession, but also existential predication and location, discussed below. In attribution and class-membership constructions, the predicate is expressed by a postpositional phrase with the postposition ká ‘with.’

16Referential identity

(3)

Gɔ̰́

wɛ̄

lɛ̄

ŋ̄

dɛ̰̄

ká.

man:h

dem

3sg.ex

1sg

husband

with

17‘This man is my husband.’ [el.]

18Class-membership

(4)

Sèé

lɛ̄

kàlàmɔ́ɔ̀

ká

Se

3sg.ex

teacher

with

19‘Se is a teacher.’ [el.]

20Attribution

(5)

Gɔ̰́ā̰nɛ́ɛ́

vɔ̀

tííkpé.

cat

pl

3pl.ex

small

21‘Cats are small.’ [el.]

22Possessive

(6)

Gɔ̰́ā̰nɛ́ɛ́

vɔ̀

ŋ̄

kɛ̀lɛ̀.

cat

pl

3pl.ex

1sg

hand

23‘I have cats’ (etymologically: cats are in my hand). [el.]

24Constructions solely predicating the existence of a referent (existential predications, dubbed as “hyparctic” predication in Haspelmath (2022)), use the existential PPM in combination with the particle ɓē ex.

25Existential predication

(7)

Bīē

vɔ̀

ɓē.

elephant

pl

3pl.ex

ex

26‘Elephants exist.’ [el.]

(8)

Lɛ̄

ɓē

kō

dàāmì

mɔ́ɔ̀ŋwɔ̀mɔ̀.

3sg.ex

ex

1pl

owner

because.of

27‘[People’s body is not for the sin, but] it exists for our Lord.’ (UBS 1978: Mano New Testament, Co1 6:13)

28Location is usually expressed via adverbs, including demonstrative locative adverbs (zèē ‘here,’ dḭ̀ḭ̄ ‘there’), postpositional phrases, or nouns in a special locative form, all following an existential PPM.

29Location

(9)

lòkó

fòtōò

séŋ́

séŋ́

lɛ̄

zèē.

3sg

mother

photo

all

all

3sg.ex

here

30‘Her mother’s photos are all here.’ [MOC]

(10)

kɛ́ì.

2sg.ex

house.loc

31‘You are in the room.’ [el.]

  • 4 As discussed by Creissels (2019), this is a typical feature of the languages of the Sudanic Belt.

32Thetic locational constructions where a referent is introduced at the same time as its location is being specified (existential constructions, in terms of Haspelmath 2022, or presentative locative constructions, in terms of Hengeveld 2011) are expressed the same way as locational construction.4

(11)

Mīā

zèē.

person.pl

3pl.ex

here

33‘There are people here.’ [MOC]

34To conclude the discussion of constructions with existential PPMs, locational non-verbal predications have given rise to durative constructions. Indeed, durative constructions consist of the existential PPM and the main verb in the infinitive form with the suffix -pɛ̀lɛ̀, which derives from the noun pɛ́lɛ́ ‘place’ in the low-tone construct form.

35Durative construction

(12)

Cèe_ Jùú

lɛ̄

ló-pɛ̀lɛ̀

ò!

Prop.N

3sg.ex

go-inf

inj

36‘Ce Jules is leaving, oh!’ (lit. “Ce Jules is at the place of going”) [MOC]

2.3 Non-verbal predication with the verb kɛ̄ ‘to be, to become’

37The previous section described basic non-verbal predications in the stative present, all involving the PPMs of the existential series.

38Class-membership constructions (and other constructions with a non-verbal predicative element) can also be used in other tenses and aspects. In that case, the verb kɛ̄ ‘to be, to become’ is used as the supporting predicator in the respective aspectual and temporal construction, such as the stative past (13), perfect (14) or imperfective present (15).

39Stative Past

(13)

dàā

kɛ̄

dɔ̄mì

ká.

3sg

father

3sg.pst

be

chief

with

40‘His father was a chief.’ [MOC]

41Perfect

(14)

dàā

āà

kɛ̄

dɔ̄mì

ká.

3sg

father

3sg.prf

be

chief

with

42‘His father has become a chief.’ [el.]

43Imperfective present

(15)

dàā

lɛ́ɛ̀

kɛ̀

dɔ̄mì

ká

3sg

father

3sg.ipfv

be:ipfv

chief

with

44‘His father is becoming a chief.’ [el.]

45As I will show in the next section, negative constructions in stative present are expressed with the copula wɔ́. In other tenses and aspects, negative constructions are expressed with the verb kɛ̄ in the respective tense and aspect form. The example below illustrating the stative past is formed with the 3sg negative PPM lɛ̀ɛ́ and the negative particle gbā.

46Negative stative past

(16)

gí

lɛ̀ɛ́

gbā

kɛ̄

mɔ̀

3sg

stomach

3sg.neg

neg

be

3sg

on

47‘She wasn’t pregnant (lit.: her stomach wasn’t on her).’ [MOC]

2.4. Negative non-verbal predication with the copula wɔ́

48All types of non-verbal predication (including location) discussed thus far use the same strategy for negation in the stative present, namely, the negative copula wɔ́. The examples below illustrate negative referential identity, class membership, attribution, possession, existential predication and location, respectively. The word order is invariably Subj–wɔ́pred, where pred is the main predicative element.

49Negative referential identity

(17)

Gɔ̰́

wɛ̄

wɔ́

ŋ̄

dɛ̰̄

ká.

man:h

dem

cop.neg

1sg

husband

with

50‘This man is not my husband.’ [el.]

51Negative class membership

(18)

Sèé

wɔ́

kàlàmɔ́ɔ̀

ká.

PropN

cop.neg

teacher

with

52‘Se is not a teacher.’ [el.]

53Negative attribution

(19)

Gɔ̰́ā̰nɛ́ɛ́

vɔ̀

wɔ́

tííkpé.

cat

pl

cop.neg

small

54‘Cats are not small.’ [el.]

55Negative possession

(20)

Gɔ̰́ā̰nɛ́ɛ́

vɔ̀

wɔ́

ŋ̄

kɛ̀lɛ̀.

cat

pl

cop.neg

1sg

hand

56‘I don’t have cats’ (etymologically: cats aren’t in my hand). [el.]

57Negative existential predication

(21)

Bīē

vɔ̀

wɔ́

ɓē.

elephant

pl

cop.neg

ex

58‘Elephants do not exist.’ [el.]

59Negative location

(22)

Ɓ

wɔ́

kɛ́ì.

2sg.emph

cop.neg

house.loc

60‘You are not in the room.’ [el.]

61The copula wɔ́ can also be used in negative identifying constructions. In that case, a postpositional phrase à ká is used with a 3sg pronoun, making the construction structurally equivalent to referential identity or class-membership constructions. The singular pronoun is used even with plural referents (23b), which is why it should be interpreted as semantically empty.

62Negative identifying construction

(23a)

Ŋ̄

lēkè

wáá

ká,

ŋ̀

nɛ́

lɛ̄.

1sg

younger.sibling

cop.neg>3sg

with

1sg.poss

child

ident

63‘It is not my younger sibling, it is my child.’ [MOC]

(23b)

Ŋ̄

lòkóò

nì

wáá

ká.

1sg

mother

pl

cop.neg>3sg

with

64‘It is not our mother and others with her.’ [MOC]

3. Non-verbal predications with other predicators

65I finally turn to the main subject of this paper – non-verbal predications with predicators. Mano has five predicators: one negative (wɔ́), used in negative non-verbal predications discussed in the previous section, and four affirmative (lɛ̄, lɛ́, wɔ́ and gɛ̰̀), used in identification and presentative constructions. The latter four predicators do not serve to introduce a non-verbal predicative part. There appear only in very limited contexts where the predicators do not combine with any other element than their arguments (see Section 3.2.2 on class-membership constructions with lɛ́ and wɔ́). For this reason, only the negative predicator wɔ́ can be considered a true copula in the sense of Hengeveld (2011). For the other four markers, I prefer the term “predicator.”

3.1. Basic constructions with predicators

66Mano predicators are typically used without a predicative NP or adverbial phrase, with only one argument, although in some cases, a locative adjunct can be used. They are restricted to the present tense. The word order is Subj-Predicator.

67The predicators wɔ́ and gɛ̰̀ are used in deictic presentation constructions. Gɛ̰̀ derives from the verb gɛ̰̀ ‘to see’ and is typically used to refer to visible referents.

68Predicator gɛ̰̀: basic presentative function

(24)

Dìì

dōó

gɛ̰̀,

ɓū

wɛ́lɛ́

ɓɔ̀ɔ́

dōó

gɛ̰̀.

cow

one

pres.vis

rice

bone

bag

one

pres.vis

69‘Here is one cow, here is one bag of rice.’ [MOC]

70Gɛ̰̀ can also be accompanied by the existential particle ɓē. The semantic contribution of ɓē is not clear. Similarly to the existential predication with the PPM and the particle ɓē (ex. 7), the construction “Subj gɛ̰̀ ɓē can imply that the existence of the referent denoted by the argument of the predicator gɛ̰̀ is predicated at the same time as the referent is presented (25), but this is not always the case (26). Note that the constructions with gɛ̰̀ (26a) and with gɛ̰̀ ɓē (26b) are synonymous.

71Predicator gɛ̰̀ + existential particle ɓē

(25)

Lēē

gbē

gɛ̰̀

ɓē.

woman

2sg

son

pres.vis

ex

72‘[When Jesus saw his mother and the disciple there whom he loved, he said to his mother,] Woman, here is your son.’ [MOC, spontaneous translation of John 19:26]

73Predicator gɛ̰̀ + existential particle ɓē

(26a)

Pèé

là

sɔ̄

gɛ̰̀.

PropN

3sg.poss

clothing

pres.vis

(26b)

Pèé

là

sɔ̄

gɛ̰̀

ɓē.

PropN

3sg.poss

clothing

pres.vis

ex

74‘Here is Pe’s clothing’ [el.]

75Gɛ̰̀ can also cataphorically refer to the forthcoming discourse, which can also be treated as a token of metaphorical presentation of discourse.

76Predicator gɛ̰̀: cataphoric reference to discourse

(27)

Pɛ́

kɛ̄

lɛ́

ɓò

gɔ̰̄

zò

wáá

gé

thing:h

3sg.pst

do

then

goat

man

heart

neg.cop>3sg

stomach

dem

yē

gɛ̰̀.

3sg.emph

pres.vis

77(An introduction to a folktale) Here is the reason why the goat does not have a heart in his stomach’ (lit.: the thing that did so that the goat’s heart isn’t in his stomach, here it is). [MOC]

78Gɛ̰̀ can optionally take a locative adjunct.

79Predicator gɛ̰̀: locative adjunct

(28)

Mīā

gɛ̰̀

zèē

gó

kwíí

pàà.

person.pl

pres.vis

here

3pl.pst

leave

European

at

80‘These people here (lit.: here are people here), they came from Europe’ [el.]

81Gɛ̰̀ can combine, without any additional meaning, with a marker coinciding in form with the demonstrative ɓɛ̄, but it does not combine with yā. The existential particle ɓē is preferred to the demonstrative ɓɛ̄ in elicitation, so it is likely that ɓɛ̄ is actually a phonetic variant of ɓē in this context.

(29)

??Pèé

là

sɔ̄

gɛ̰̀

ɓɛ̄.

Pe

3sg.poss

clothing

pres.vis

ex?

82‘Here is Pe’s clothing.’ [el.]

83Turning to the next presentative predicator, wɔ́, it can introduce visible (32) or out-of-sight referents (30), as well as refer to discourse (31).

84Predicator wɔ́

(30)

Nɛ́ŋ̀pà

kpɔ́à

yē

wɔ́

yā.

PropN

trace

3sg.emph

pres

dem

85‘The remains of Neinpa (a name of a village), here they are’ [MOC]

86Predicator wɔ́

(31)

Pɛ́

kɛ̄

lɛ́

ɓò

gɔ̰̄

zò

thing:h

3sg.pst

do

then

goat

man

heart

wáá

gé

yē

wɔ́

ɔ̄.

neg.cop>3sg

stomach

dem

3sg.emph

pres

dem

87(A conclusion to a folktale.) ‘That is the reason why the goat does not have a heart in his stomach’ (lit.: the thing that did so that the goat’s heart isn’t in his stomach, that’s it). [el.]

88Wɔ́ is obligatorily followed by a demonstrative (wɛ̄~ɓɛ̄ or yā~ā). Thus, the presentative predicator wɔ́ (32) can be distinguished from the negative predicator wɔ́ (33), which does not combine with a demonstrative. Example (32) also illustrates that the affirmative wɔ́ can take a locative adjunct.

(32)

Líà

wáá

mɔ̀

yā.

blood

pres>3sg

on

dem

89‘Here is blood on her’ [MOC]

(33)

Líà

wáá

mɔ̀.

blood

cop.neg>3sg

on

90‘There is no blood on her’ [el.]

91The predicators lɛ̄ and lɛ́ are used primarily for identification of non-visible (34) and abstract referents (36), as well as of discourse referents (35), but both can be used for presentation of visible referents as well (37), and thus have the presentative function.

92Lɛ̄ always occurs in a clause-final position and does not combine with demonstratives or locative adjuncts. Lɛ̄ is a homonym of the 3sg existential PPM, but can be distinguished from it because it can combine with arguments of all persons and numbers without changing its form (e.g., its use with 1sg in 34 and 3sg in 35; compare the use of PPMs with subjects of different persons and numbers in ex. 9–11 above).

93Predicator lɛ̄

(34)

Mā

lɛ̄.

1sg.emph

ident

94‘It is me’ [MOC].

95Predicator lɛ̄

(35)

Wálà

léwè

lɛ̄.

God

speech

ident

96‘(This is) the word of the Lord (lit: It is the speech of God, said after the end of readings in church).’ [el.]

97Predicator lɛ̄

(36)

Ŋwɔ́

lɛ́

nū

ká

yē

lɛ̄.

problem

att

3sg.pst

come

3sg

with

dem

3sg.emph

ident

98‘The problem that he brought, that’s it.’ [MOC]

99Lɛ̄ is the most common predicator in answers to the questions of the type ‘What/Who is it?’.

100Predicator lɛ̄

(37)

Pèé

là

sɔ̄

lɛ̄.

Pe

3sg.poss

cloth

ident

101‘[What is it (pointing)?] It is Pe’s piece of clothing.’ [el.]

102Lɛ̄ can occur in focal contexts, as illustrated by example (23) repeated below.

103Predicator lɛ̄: focal context

(38)

Ŋ̄

lēkè

wáá

ká,

ŋ̀

nɛ́

lɛ̄.

1sg

younger.sibling

cop.neg>3sg

with

1sg.poss

child

ident

104‘It is not my younger sibling, it is my child.’ [MOC]

105Lɛ̄ is also used in resultative constructions where the argument of the predicator is a non-finite predication with a gerund. Such usage can also be considered an extension of an identifying construction, where what is identified is a state of affairs rather than a referent. In this function, lɛ̄ can combine with a postpositional phrase expressing the verbal argument.

106Predicator lɛ̄: resultative construction

(39)

Ŋ̄

zò

dɔ̄-ɔ̀

lɛ̄

mɔ̀.

1sg

heart

install-ger

ident

3sg

on

107‘I believe in it (lit.: my heart is laid on it). [MOC]

  • 5 It is not entirely clear whether tɛ́ is a variant of lɛ́ or a separate morpheme with largely overla (...)

108The marker lɛ́ never occurs in a clause-final position and requires another element after it. Most commonly, it is a demonstrative: wɛ̄~ɓɛ̄ or yā~ā. Lɛ́ can also have the variants nɛ́ (in a nasal context), tɛ́5 or a floating high tone. Lɛ́ can be used with visible referents (40), but it is more common with invisible and abstract referents (41).

109Predicator lɛ́

(40)

kɔ́nɔ́

lɛ́

wɛ̄.

2sg

food

att

dem

110‘Here is your food.’ [MOC]

111Predicator lɛ́

(41)

Ŋwɔ́

tɛ́

māē

māà

gɛ̰̀

mɔ̀

ɔ̄

problem

att

1sg.emph

1sg.prf>3sg

see

3sg

on

dem

yē

lɛ́

ā.

3sg.emph

att

dem

112‘That’s the way I see the problem (lit.: the problem that I have seen, this is it).’ [MOC]

113The predicator lɛ́ can also be used with the existential particle ɓē.

114Predicator lɛ́ + ɓē

(42)

Kálá

kɛ̄

sɛ̀lɛ̀

dìè

lɛ́

ɓē.

but

3sg

do

village:cstr

real

att

ex

115‘[Lying, we do it a lot,] but it is the real village of liars’ (lit.: real village of doing it). [el.]

116Lɛ́ also combines with ɓē in questions of the type ‘What is it?’. In the two examples below, nɛ́ is a variant of lɛ́ used in a nasal context (after mɛ̄ ‘what’) and wē is a variant of ɓē:

117Predicator lɛ́ + ɓē

(43a)

Mɛ̄

nɛ́

wē?

what

att

ex

118‘What is it?’ [el.]

(43b)

Dēŋ̀

lɛ́

wē?

who

att

ex

119‘Who is it?’ [el.]

120In most affirmative examples encountered, the construction Subj lɛ́ ɓē occurs within a comparison clause.

121Predicator lɛ́ + ɓē

(44)

Ŋ́ŋ̀

lō

kɛ̄-ɛ̀

lɛ̀

zī

nɔ́

pɛ̄kúlú

kpá̰á̰

lɛ́

ɓē.

1sg.ipfv

go:ipfv

be-ger

like

way

just

iron

bare

att

ex

122‘[If love has finished in my life,] I will become just like bare iron’ (lit.: I will become like, it is iron). [MOC, oral translation from French of 1 Corinthians 13:1]

123The three examples with lɛ́ ɓē above demonstrate that the existence of the referent, the argument of the construction, is not necessarily predicated: in (42), the speaker discusses the very village where he is located, whose existence is common ground for everybody present during the discussion; in (44), the argument is a non-referential, generic noun. In (43), in the question, the existence of the referent which the question seeks to identify must be part of the common ground of the interlocutors in order for the question to be felicitous. Thus, the semantic contribution of ɓē in the construction Subj lɛ́ ɓē is not clear and is not related to predicating existence, similarly to the aforementioned construction Subj gɛ̰̀ ɓē.

124Lɛ́ can also combine with locative adjuncts.

125Predicator lɛ́ + locative adjunct

(45)

Gààzù

láá

kɛ̀lɛ̀

wɛ̄.

mirror

att>3sg

hand

dem

126‘[A: This one is a photographer.] B: That is a mirror in her hand (meaning: it is not a photograph, it is a mirror).’ [reflexive experiment, stimulus 1]

127Finally, lɛ́ can combine with the interjection , with a mirative-like meaning.

128Predicator lɛ́ + interjection

(46)

Ŋ̀ŋ́gɔ̰̀

Marie

gbē

lɛ́

ò.

isn’t.it

Marie

son

att

inj

129‘Isn’t it, it’s a son of Mary (how unusual for such a man to be of such a modest upbringing)!’ [MOC]

130To summarize this section, I have discussed simple constructions with the predicators wɔ́, gɛ̰̀, lɛ̄ and lɛ́. All these predicators have the function of referent presentation and all but gɛ̰̀ can also serve to identify referents. However, wɔ́ more frequently occurs with visible referents and therefore is more compatible with the presentative function, while lɛ́ and lɛ̄ are more compatible with the identifying function (we will see in Section 4, however, that these two functions are not easy to distinguish). Some of the predicators can combine with locative adjuncts, demonstratives, and the existential particle ɓē. Some of these combinations are obligatory. Table 2 summarizes the possible combinations.

Locative adjunct

ɓē (existential particle)

Demonstrative

lɛ̄

no

no

no

wɔ́

yes

no

yes, obligatory

lɛ́

yes

yes, especially in comparison constructions (obligatory presence of ɓē or another final element after lɛ́)

yes (obligatory presence of a demonstrative or another final element after lɛ́)

gɛ̰̀

yes

yes

?no (only ɓɛ̄, likely a variant of ɓē)

131Table 2. Possible combinations of predicators with demonstratives, existential particle ɓē and locative adjuncts

3.2. Extensions of constructions with predicators

3.2.1. Extension with a clause and focus meaning

132All predicators, with the exception of lɛ̄, can occur in a construction where the predicator is followed by a clause. Consider the following example:

133gɛ̰̀ + clause: presentative reading

(47)

lé

lɛ́

lèŋ̄.

3pl

mouth

att

probably

Ō

gɛ̰̀

ò

pɛ̀lɛ̀

yī.

3pl

mouth

pres.vis

3pl.sbjv>3sg

wash:ipfv

there

134‘[The thing that they are doing,] it is probably their mouth, here is their mouth, they are washing it.’ [reflexive experiment, stimulus 2]

135Example (47) illustrates a sequence of two presentative/identifying constructions, one with the predicator lɛ́, another with the predicator gɛ̰̀, extended with a clause. The argument of the predicators is the NP ō lé ‘their mouth.’ The example is taken from a continuous recording of an experiment with visual stimuli where the participants are asked to describe a set of pictures. The function of the constructions is thus identification through visual presentation. Example (47) is a description of a group of children washing their mouths. Note the PPM used in the clause. belongs to the subjunctive PPM series, which never occurs in main clauses. In the main clause, for the same aspectual meaning (imperfective), an PPM of the imperfective series (óò 3pl.ipfv) is used; see ex. (15) illustrating imperfective in a main clause.

136The following example illustrates a construction where gɛ̰̀ is followed by wē (a variant of ɓē) with an intervening clause. While basic presentative constructions can have the form of Subj – gɛ̰̀, constructions Subj – clause – gɛ̰̀ always have a final element, very frequently, ɓē~wē, but it can be also yī ‘there,’ as in (47) above. It is possible that other elements can also occur in this position, but they have not been attested yet.

137gɛ̰̀ + clause + ɓē: presentative and focal readings

(48)

Yē

gɛ̰̀

bḭ́ḭ́

lɛ̄

gààzù

wɛ̄

yí

wē.

3sg.emph

pres.vis

3sg

image

3sg.ex

mirror

dem

in

ex

138‘[A woman is holding a mirror. Her friend that is behind her,] here is her image in the mirror.’ [reflexive experiment, stimulus 3]

139The argument of the presentative gɛ̰̀ in example (48) stands out as one alternative picked among other alternatives. Indeed, the speaker is trying to explain what a visual stimulus intends to depict: a woman looking at another woman behind her through a mirror. Thus, in addition to being presented with gɛ̰̀, the referent is interpretable as contrastive focus: it is the woman behind the woman with the mirror that the latter sees. The focalization is done by means of a cleft-like construction with gɛ̰̀, consisting, as is typical of clefts (Lambrecht 2001), of a non-verbal predication (‘here it is’), which can potentially stand on its own (see Section 3.1), and an ensuing clause, in this case, also non-verbal (‘her image is in the mirror’).

140wɔ́ + clause: presentative reading

(49)

Léé

tɔ́ɔ̄

là

ŋwū

kɛ̄

yē

wɔ́

woman:h

dem.vis

3pl.pst

3sg.poss

head

do

3sg.emph

pres

dìè

gɛ̰̀

gààzù

yí

wɛ̄.

3sg.sbjv

3sg.refl

int

see:ipfv

mirror

in

dem

141‘This woman, she has been braided, here she is, she is looking at herself in the mirror.’ [reflexive experiment, stimulus 4]

142Similarly to (47–48), (49) is another example of a visual presentation of a referent, this time with the predicator wɔ́. The referent is first introduced by a NP with a proximal exophoric demonstrative, léé tɔ́ɔ̄ ‘this woman.’ The entire construction is concluded by the demonstrative wɛ̄ — as discussed earlier, wɔ́ is always accompanied by a demonstrative. Note that the subjunctive PPM is used here as well (, a 3sg form).

143Examples (47–49) illustrate clear cases of ostensive demonstration, where the construction accompanying a presentation of a visible referent is extended by a clause. The context of the activity which the utterances are part of — a description of visual stimuli — and the speakers’ pointing gestures support this interpretation. Consider example (50), by contrast.

144wɔ́ + clause: thetic reading

(50)

Sálámá

wɔ́

kɛ̄

là

ā.

chance

pres

3sg.pst

be

3sg

on

dem

145‘(A group of women is discussing another woman who almost lost her child in an accident.) She was lucky (lit.: this is a chance, it was on her).’

146Here, the referent introduced by wɔ́ is an abstract notion (sálámá ‘chance’). As we saw in Section 3.1, wɔ́ can indeed be used to introduce invisible and abstract referents. Yet, the utterance does not function as an extended presentation of sálámá. Instead, the best interpretation is that of sentence focus where the subject is focalized. The focalization is done by means of a cleft-like construction with wɔ́, consisting, again, of a non-verbal main clause (lit. ‘this is a chance’) and a dependent clause (lit. ‘it was on her’). Supporting the interpretation in terms of sentence focus, the translation that my consultant offered was c’est qu’elle avait la chance ‘It is that she was lucky,’ where c’est que ‘it is that’ serves to introduce an explanation. This is a typical function of sentence-focus utterances, and subject focalization is also a typical way of expressing sentence-focus (Vydrina 2020).

147Many examples with wɔ́ can yield a sentence-focus interpretation. Consider, for instance, the following:

148wɔ́ + clause: presentative and thetic readings

(51)

Car

wɔ́

lō

ā.

truck[fr]

pres

3sg.sbjv

go:ipfv

dem

149‘Here is a truck passing’ [MOC]

150Here, the speaker is drawing attention to a passing truck. It is not a referent alone, but a referent in action that is presented. Thus, it is an all-new, completely out-of-the-blue description of a situation, which is compatible with both a sentence focus and a presentative interpretation.

151Let’s now consider examples with the predicator lɛ́. Example (52) illustrates a case of referent presentation extended by a clause.

152lɛ́ + clause: presentative reading

  • 6 This usage of tɛ́, which is likely a variant of lɛ́, is not covered by the present paper. Here, tɛ́(...)

(52)

Léé

tɛ́6

wɛ̄

tɛ́à

wɛ̄

woman

att

dem

bowl

dem

yē

lɛ́

yà

là

nɛ́

ŋwíí

ā.

3sg.emph

att

3sg.pst>3sg

put

3sg.poss

child

head

dem

153‘This woman, this bowl, it is it (that) she put on her child’s head.’ [reflexive experiment, stimulus 5]

154The context for example 52 is the same as above—it is a description of visual stimuli. Here, several referents are introduced: a woman, her child and a bowl that she is putting on the child’s head. In configuring his response, the participant first introduces the woman by saying léé tɛ́ wɛ̄ (as noted above, tɛ́ is a variant of lɛ́). Furthermore, the speaker introduces the bowl just with a demonstrative, followed by a 3sg pronoun referring to the bowl, followed by lɛ́ and a clause: ‘this bowl, it is it that she put on her child’s head.’ Both times the speaker points to the referent in question, so lɛ́ is used for referent presentation. The following example illustrates a similar pattern.

155lɛ́ + clause: presentative reading and focal reading

(53)

Photo

wɛ̄

yē

lɛ́

zɔ̰̀ɔ̰̀

lɛ̀ɛ̄

āā.

photo

dem

3sg.emph

att

3sg.sbjv>3sg

show:ipfv

3sg

to

dem

156‘[A: This man has drawn a picture of the woman, what is he doing now? – I think that] Here’s the photo he is showing/this photo, it is it (that) he is showing.’ [MOC]

157Here, the speaker first introduces an NP referring to a photo, followed by a 3sg pronoun referring to the bowl, followed by lɛ́ and a clause. In addition to the presentative interpretation, supported by pointing gestures, another interpretation is possible: that of, once again, sentence focus marked through a cleft-like construction involving a non-verbal predication. As Vydrina (2020) demonstrated for Kakabe, in utterances with highly topical subjects—in the case of (52) and (53), subjects are expressed by 3sg pronouns—the focus marking conveying sentence focus can be placed not on the subject (as in 50) but on the direct object.

158The following example, taken from a similar elicitation setting (see example 48 and the description of the context), involves a contrast; therefore, a focal function—this time, contrastive argument focus—is a central part of the function of the utterance.

159lɛ́ + clause: focal reading

(54)

Mí

tɛ́

dɔ̀á

mɛ̀ŋ́

wɛ̄

man:h

att

stand>ger>with

3sg

behind

dem

yē

lɛ́

bḭ́ḭ́

lɛ̄

gààzù

yí

ā.

3sg.emph

att

3sg

shadow

3sg.ex

mirror

in

dem

160‘The person who is standing behind her, it is HER IMAGE in the mirror (lit.: it is her (such that) her image is in the mirror)‘ [reflexive experiment, stimulus 3]

161The stimulus picture offered to the speaker in 54 depicts a woman who is holding a mirror with a reflection of a woman’s head; it is not the woman herself who is reflected, but another woman behind her. Example (54) offers this corrected interpretation. Example (2) above (repeated below) also illustrates contrastive object focus.

162lɛ́ + clause: focal reading

(55)

Wìì

lɛ́

zɛ̄

ɛ̄.

animal

att

1sg.pst>3sg

kill

dem

163‘[I did not hit a person,] it is an animal that I killed.’ [MOС]

164The preceding context (‘I did not hit a person’) clearly introduces a referent to which the focalized constituent (‘animal’) offers an alternative, which warrants a contrastive focal intepretation.

165Finally, the following example shows that lɛ́ and wɔ́ can be used interchangeably in a focal context.

166lɛ́/ wɔ́ + clause: focal reading

(56a)

Ká

tɛ́

dɔ̀á

wɛ̄

mā

lɛ́

house

att

stand>ger>with

dem

1sg.emph

att

mā

dɔ̄

ɔ̄.

Dēŋ̀

lɛ́

dɔ̄?

1sg.pst>3sg

build

dem

who

att

3sg.pst>3sg

build

167A: The house which is standing there, it is me who built it. Who built it?

(56b)

Mā

wɔ́

ŋ̄

tá̰à̰

píé

dɔ̄

yā

kɛ̀ɛ̀?

1sg.pst>3sg

pres

1sg.pst

ground

towards

build

dem

isn’t.it

168B: ‘It is me (lit.: here is me) who built the site up, isn’t it?’ [MOC]

169The example is taken from a recording of several children riding in a car in an unfamiliar city and discussing what they see around them. In (56), they see a building and playfully claim ownership of it: first a boy in (56a), and then a girl in (56b). The boy does so with lɛ́, and the girl with wɔ́. The referents in the example are first person pronouns (‘it is me who built…’), which are discourse-given and do not require identification per se, but can appear as one alternative among a possible set and, for this reason, can be interpreted as focused. The focal reading was indeed suggested by my language consultant, who translated both examples with a cleft construction: c’est moi qui… (‘It is me who…’).

170To summarize this section, we have seen that non-verbal clauses with predicators gɛ̰̀ and especially wɔ́ and lɛ́ can be extended by a finite clause. The clause has signs of syntactic dependency: the imperfective PPM is not used there and is replaced by the subjunctive PPM, which does not occur in main clauses. Crucially, I have shown that some utterances where a presentative/identifying construction with wɔ́ or lɛ́ is extended by a clause can have a focal interpretation: either as argument focus, or, through formal marking of an argument, sentence focus. This observation will be discussed in Section 4.

3.2.2. Extension with a comitative PP and argument-predicate reversal

  • 7 The crucial difference is that in the sense used by Creissels, true argument-predicate reversal con (...)

171The predicators lɛ́ and wɔ́ can occasionally take an argument, expressed by a NP with a comitative postposition ká, the same as the one used in constructions of referential identity and class membership. Such constructions are used to put into correspondence two referential expressions, just like referential identity and class membership constructions formed with the existential PPMs (described in Section 2.2). Constructions with the predicators lɛ́ and wɔ́ differ from constructions with the existential PPMs; in the case of the former, the NP referring to a given referent about which the predication is made is put in the postpositional phrase, while the NP referring to the semantic class of the referent (in case of class membership) or an identical referring expression is put in the position of the argument of the predicator. These examples are thus similar, but not identical, to argument-predicate reversal constructions found in several other Mande languages (Creissels 2022).7 Let’s consider the following example:

172lɛ́: argument-predicate reversal in a class-membership construction

(57)

K

líà

lɛ́

ká.

1pl

blood

att

2sg

with

173‘You are OUR BLOOD.’ [MOC]

174The example is taken from a folk story. Here, deceased parents are coming to their son in a dream, complaining that he did not properly organize their funeral, but nevertheless offering help. The point of the utterance is to communicate that the man is their child, despite his wrongdoings. The semantic predicate of the class-membership construction is kō líà ‘our kin (lit.: our blood),’ which is put in the position of the argument of lɛ́, while the semantic argument of the construction which refers to the addressee is put in the postpositional phrase, ī ká. Crucially, kō líà ‘our blood, our kin’ and ‘you’ are not in the referential identity relation, as the speakers have more kin than their son.

175In the next example, which illustrates a construction that is ambiguous in terms of class membership and referential identity, the argument similarly bears an additional emphasis — this time, through contrast.

176lɛ́: argument-predicate reversal and constrastive focus

(58)

Kō

bɛ̰̀ɛ̰̄

tɔ̀ŋ̀píé

mìà

lɛ́

kō

ká.

1pl.emph

too

PropN

person.pl:cstr

att

1pl

with

177‘[The joking relationship is between us and the Maa people.] As for us, we are the TONPIE PEOPLE.’ [MOC]

178Here, the argument of the construction, tɔ̀ŋ̀píé mìà ‘the Tonpie people,’ a clan name, is in contrast with the name of another clan introduced in the previous clause, màá ‘the Maa people.’ The clan name is the semantic predicate of the construction, while the semantic argument is the first person plural pronoun put in the PP position. It is not clear what the scope of ‘we’ is here, whether it refers to the entire group of people who call themselves Tonpie or just a subgroup — such as residents of the village where the discussion takes place. Therefore, there is ambiguity between referential identity and class membership. Crucially, however, in (58), unlike (57), the argument-predicate reversal is connected to a more specific focal reading – contrastive focus.

179Let’s turn to the next example, this time with wɔ́.

180wɔ́: argument-predicate reversal and contrastive focus

(59)

Kɛ̄

kɔ̰̀ɔ̰̄lèē

pèèlɛ̄

wáá

nɔ́.

isn’t.it

young.mother

two

pres>3sg>with

just

181‘It is two young mothers here, isn’t it?’ [reflexive experiment, stimulus 6]

182The example is taken from the context of an experiment with visual stimuli. The picture that the respondent is describing represents an activity sequence: a woman is putting her child on her back, including the beginning stage of the activity, where she is attaching the cloth, and the end stage, where the cloth is attached. The respondent mistakenly identified the picture as representing two women, not one. Yet the existence of the other alternative (that the picture represents something else than two women) is presupposed and conveyed by the tag question marker, ‘isn’t it.’ The NP ‘two young women,’ the semantic predicate of the class-membership construction, is put in the position of the argument of the predicator, while the PP is filled with a 3sg pronoun. Note here that the presentative function is combined with the function of predicating class membership and a contrastive value.

183However, not all attested examples with lɛ́ have the order of the argument and the predicate reversed, the semantic argument being put in the postpositional phrase and the semantic predicate in the position of the argument of the predicator. Below, the predicate is in its syntactic place in the postpositional phrase.

184lɛ́: referential identity construction without argument-predicate reversal

(60)

Yē

lɛ́

kā

ká

ɓɛ̄.

3sg.emph

att

2pl

with

dem

185‘[They said: who is your mother and your father? He said:] It is you!’ [MOC]

186Example (60) is not a class-membership construction, but rather a referential identity construction, where both NPs have the same referents (both the 3sg pronoun yē and the 2pl pronoun kā refer to the speaker’s mother and father). The motivation for using lɛ́ instead of an existential PPM, which is the default option for referential identity constructions, is not in this case related to the argument-predicate reversal. Example (60) serves as an answer to, and thus structurally mirrors, the question given in (61), in which the NP that refers to the argument of lɛ́ in (60) is also put in the position of the argument of lɛ́, and the same for the NP in the PP. Note also that the question word in (61) is put in the PP. The motivation for the order in the question is explained in the next section. In short, it is related to the additional (almost exasperated) attention which the speakers draw to the referent, ‘your mother and your father.’

187lɛ́: referential identity construction in a question

(61)

lèē

wà

dàā

lɛ́

dēŋ̀

ká?

2sg

mother

3pl.ip

2sg

father

att

who

with

188‘Your mother and your father is who?’ [MOC]

3.2.3. Question

189As pointed out in Section 3.1, lɛ́ (and especially its variant nɛ́) can be used in questions. It can be used in yes–no questions, as in (62), which closely resembles utterances with focal interpretation:

190lɛ́: Y/N question

(62)

Ŋwánà

lɛ́

ɓà

gèē

ā?

truth

att

3sg.ipfv>3sg

say:ipfv

dem

191‘It is truth that you are saying?’ [MOC]

192Lɛ́ can also be used in wh-questions, typically following the left-extraposed question word (exsitu). All questions to subjects are exsitu.

193lɛ́: ex-situ wh-question

(63)

Dēŋ̀

lɛ́

bɛ̀ī

pá

who

att

3sg.ipfv

be.able:ipfv

3sg.conj

touch

3sg.refl

zì

mɔ̀?

poss.int

on

194‘Who can give an example about himself? (lit.: who can touch for himself?)’ [MOC]

195Note that the following example with a question to the subject ends with wē (ɓē).

196lɛ́: ex-situ wh-question

(64)

Tíé

gḭ̄ḭ̄

kpɛ̄ɛ̀

lɛ́

yà-à

lɛ̄

lūú

wɛ̄

mɔ̀

wē?

fire

smell

which

att

3sg

sit-ger

ident

bush

dem

on

ex

197‘What kind of smell of fire is spread (lit. put) in the bush?’ [MOC]

198Example (64) is structurally like the construction with Subj lɛ́ ɓē discussed in (42), except that there is a clause between lɛ́ and ɓē and there is a question word in the subject NP (kpɛ̄ɛ̀ ‘which, what kind of’) and, therefore, the utterance functions as a question.

199Question words, such as dēŋ̀ ‘who,’ can also be put in situ, as in (66). In that case, lɛ́ is not used.

200In-situ wh-question

(65)

Ká

gbṵ̀mɔ̀

nɔ̄

dēŋ̀

lɛ̀ɛ̄?

1pl.conj

help

give

who

to

201‘You help whom?’ [MOC]

202Consider finally a pair of referential-identity constructions framed as questions. Note that the question word, dēŋ̀ ‘who,’ can be put in the PP (66b) and in the subject position (66a). The former case is an example of an in-situ question and the latter of an ex-situ question.

203lɛ́: ex-situ and in-situ wh-questions

(66a)

lèē

wà

dàā

lɛ́

dēŋ̀

ká?

2sg

mother

3pl.ip

2sg

father

att

who

with

204‘Your mother and your father is who?’ [MOC]

(66b)

Dēŋ̀

lɛ́

lèē

wà

dàā

ká?

who

att

2sg

mother

3pl.ip

2sg

father

with

205‘Who is your mother and your father?’ [el.]

206Example (66a) is taken from a narrative. The speakers are parents of a misbehaving man who appear to him in his dream, since he does not honor them enough. It goes without saying for the Mano that disrespect to parents is one of the greatest sins that a person can commit. Thus, (66a) is not a simple question about the identity of one’s parents; in fact, it is a rhetorical question. This is what motivates the extra attention that is drawn to the referent, the usage of lɛ́ and the argument-predicate reversal, whereby the NP ‘your mother and your father’ is put in the position of the subject of lɛ́ and the question word in the PP.

4. Discussion: interaction between non-verbal predication and focus

207While the main subject of this paper concerns non-verbal identifying and presentative constructions (and their syntactic and pragmatic extensions), they have been put in the context of the overall strategies of non-verbal predication in Mano. It turns out that the identifying and presentative constructions formed with an invariable predicator constitute a marginal syntactic type among other non-verbal strategies. Indeed, most types of non-verbal constructions are formed with an PPM (in the stative present) or with an PPM and the verb kɛ̄ ‘to be’ (in other tenses and aspects). Negative non-verbal predications (including identifying ones) are formed with the predicator wɔ́, which, in contrast to identifying and presentative predicators, obligatorily takes not only an argument but also a NP, a PP, an adverb or another element fulfilling the role of a non-verbal predicative part. The small share of non-verbal predications expressed through copulas and other predicators (only negative constructions and class-membership and referential identity constructions with argument-argument reversal) is quite remarkable, compared to other Mande languages. For example, although non-verbal predication through PPMs is also a common strategy in Eastern Dan, it still uses copulas in locative and possessive predications. Manding languages, by contrast, rely exclusively on copulas in non-verbal predication (see example 67 from Mandinka below). It is also worth noting that non-verbal constructions are another area where Mano predication is strongly asymmetrical, in the sense that not all affirmative constructions have a negative counterpart (there is no negative presentative) and also in terms of constructional asymmetry: affirmative constructions are based either on PPMs or on predicators, while negative constructions are based on a copula. For details about asymmetry in the Mano negation system, see Khachaturyan (forthc.), on (a)symmetry in negation from a typological perspective, see Miestamo (2005).

  • 8 Or, rather, the connection between predicative and adnominal demonstratives is not paradigmatic but (...)

208My discussion of identifying and presentative predicators, the main object of this paper, is based on the seminal publication by Killian (2022). Killian’s paper introduces predicative demonstratives as a separate typological class. Predicative demonstratives are often, but not always, part of demonstrative paradigms and have morphological parallels to other demonstratives in a language. Crucially, predicative demonstratives fulfill the syntactic function of a predicative center and occur most centrally in identifying and presentative constructions. Because of the syntactic and semantic function of the predicators in Mano, I consider them instances of predicative demonstratives, despite the fact that there is no morphological connection to adnominal demonstratives.8

209According to Killian’s (2022: 14) definition, presentative demonstratives “introduce and verbally highlight a referent in the environment or discourse in the canonical demonstrative sense, directing the addressee’s attention to the object.” By contrast, “demonstrative identifiers are used for identifying, presenting, or indicating a referent.” Crucially, as Killian acknowledges, identifiers frequently overlap with presentatives in terms of usage; this echoes a similar observation by Diessel (1999: 79). In general, identification and presentation as functions are difficult to differentiate. Many of the examples discussed in Section 3 come from recordings of elicitation sessions with visual stimuli, where the participants were asked to describe a set of pictures; while doing so, they were orienting to, identifying and at the same time presenting (often with a pointing gesture) different objects on the pictures. The difference between identifying and presentative uses in these data was thus not very clear. Key parameters could be the discourse status of the referent (new or given) and its physical properties (visible, invisible, abstract), where new and visible referents are more compatible with the presentative function of the demonstrative. More semantic work to distinguish between the presentative and identifying functions is needed, however. While this is beyond the scope of the present paper, in a preliminary fashion I can define the core functions of the Mano predicators in the following way. Because gɛ̰̀ refers almost exclusively to discourse-new and visible referents, I gloss it as pres.vis, ‘visible presentative.’ Wɔ́ also tends to occur with visible and discourse-new referents, but it can also combine with invisible referents, which is why it is glossed just as pres, ‘presentative.’ Lɛ̄, by contrast, is used for identification but also in derived functions, such as resultative ones. Therefore, the gloss is ident. Lɛ́ has the broadest functionality of all, having both presentative and identifying functions, but also a broader non-predicative function of attention-drawing, which is not discussed in the present paper. Its gloss is thus att.

210Interestingly, Killian observes that presentative demonstratives, unlike identifiers, may have a distributional restriction, namely, that they may not occur in focus constructions (2022: 20). Yet both the presentative wɔ́ and gɛ̰̀ and the identifying lɛ́ have been attested in constructions with focal readings. Moreover, an ostensive presentation of a referent, as in the context of description of visual stimuli, does not exclude a focal reading, which indicates that the restriction that Killian observed in some languages in his sample is not supported by the Mano data.

211Another typological observation concerns the position of presentative predicators: in Mano (and, likely, in Mande more generally), they systematically occur in the clause-final position, while cross-linguistically they favor the clause-initial position, like in French, English or Russian: voici une pomme / вот яблоко here is an apple’ (Killian 2022: 12).

  • 9 Unfortunately, Creissels does not offer a sufficient explanation why he treats the PP Làamínì ti (...)

212Next, I would like to turn to the argument-predicate reversal in class-membership constructions. A recent article by Creissels dedicated to the topic offers an insightful discussion of such constructions in Mande languages. Creissels (2002) suggests a connection between word order and information structure in that the neutral word order is Argument–Predicate, and whenever the predicate comes first, it is motivated by a particular information-structural configuration: the argument-first order corresponds to the neutral topic-comment information packaging, while the predicate-first order corresponds to a comment-afterthought packaging.9

213The discussion of Mandinka data is particularly revealing in that regard. Indeed, in Mandinka, while both orders are possible, the predicate is consistently marked with the focus marking , which indicates its consistent status as the comment (or part of the constituent in focus). Thus, according to Creissels, in an example with argument-predicate reversal (67b), the argument of the copula mú, which is the semantic predicate of the non-verbal predication, has focus marking, while the constituent belonging to the PP has the information status of an afterthought and does not have focus marking.

214Argument-predicate reversal in Mandinka

(67a)

Làamínì

mú

màndìŋkôo

lè

tí.

PropN

cop

Mandinka.d

foc

as

215‘Laamin is Mandinka.’

(67b)

Màndìŋkôo

lè

mú

Làamínì

tí.

Mandinka.d

foc

cop

PropN

as

216‘He is Mandinka, Laamin that is.’ (Creissels 2022)

217In both Mano and Mandinka, constituent order in class-membership constructions can be inverted. In Mano, the neutral argument-predicate order occurs in the construction with an existential PPM (Subject – ppm – predicated semantic class or identical referent), while use of the marker lɛ́ as a predicator is usually associated with the reverse order (predicated semantic class or identical referent – lɛ́ – semantic argument). As shown in Section 3.1, lɛ́ can also be used in constructions with focal interpretations. Similarly, in Mandinka, the marker that is used with the constituent to the left of the predicator in the construction with reverse order can be used to express focus stricto sensu.

218In both Mandinka and Mano, however, the marker which elsewhere can give rise to a focal interpretation does not necessarily yield that in class-membership constructions. Thus, in Mandinka, the focus marker is an obligatory part of the predicate and does not convey contrast and other focus-related meanings. Similarly, while some of the Mano examples of argument-predicate reversal, such as in (58), do indeed involve a contrastive focus interpretation of the argument of lɛ́, the focal interpretation is not compatible with all examples, such as (61), where the argument of the PP is an in-situ question word. While in-situ wh-questions do not have to be focused (Aboh 2007), they cannot be afterthoughts. Moreover, some examples with lɛ́ do not involve argument-predicate reversal, such as in (60), where the argument of the PP is the semantic predicate of a referential identity construction.

219The following pair of examples shows even more clearly that in Mano the inverse order is actually not connected to focus per se, and that focus is marked differently. Consider the first example.

220Argument-predicate reversal: not a focal reading

(68)

Mouchoir

lɛ́

mí

tɔ́ɔ̄

ká

ā.

handkerchief[fr]

att

person:h

dem.vis

with

dem

221‘It is a handkerchief, this thing (lit.: this person).’

222Here, the semantic argument of the class membership construction is referred to by a NP with an exophoric demonstrative tɔ́ɔ̄ and the predicate is mouchoir ‘handkerchief.’ The example was translated by my language consultant as Cet autre est un mouchoir ‘this other one is a handkerchief’ where ‘this other one’ is a translation of tɔ́ɔ̄, used in exophoric reference to objects that require a redirection of the addressee’s attention (see Khachaturyan 2020).

223Now consider a slightly modified version of the example with addition of the existential PPM:

224Class membership construction with focalization

(69)

Mouchoir

lɛ́

lɛ̄

mí

tɔ́ɔ̄

ká

ā.

handkerchief

att

3sg.ex

person:h

dem.vis

with

dem

225‘It is a handkerchief, this one (and the other one is not).’

226This time, as my consultant explained, “tu fais le choix entre deux qui ne sont pas tous des mouchoirs, c’est cet autre qui est le mouchoir” (‘you make a choice between two [objects] that are not all handkerchiefs, and it is this other one which is a handkerchief’). In other words, only (69), and not (68), clearly has a (contrastive) focal interpretation. The function of lɛ́ in (68) is not of focus, but of referent identification.

227The discussion of argument-predicate reversal which is irreducible to information packaging prepares us for the discussion of the status of information packaging and focus marking and the relationship between focus marking and non-verbal predication in Mano. It has been established in the literature that non-verbal predicators in Mande languages are prone to co-occur with focus marking within identification construction, as in the following example from Guinean Mandinka.

228Identifying construction in Guinean Maninka

(70)

lè

yé.

that

foc

cop

229‘That’s it’ (Vydrin 2020: 89)

230In the next example, the copula yé is omitted, so the focus marker is reinterpreted as a predicative element.

231Identifying construction in Guinean Maninka

(71)

Kó

dúman-nɛn`

nè.

matter

good-dim\d

foc

232‘It's a good matter’ (Vydrin 2020: 89)

233Thus, the example above illustrates a common tendency in Mande, namely, that copulas, especially those used in identification constructions, often originate from focus marking.

234In addition, in Mande, and especially in South Mande (see a comprehensive review by Kuznetsova 2023), there is often a homonymy between copulas and focus markers at the synchronic level. The following two examples illustrate this tendency in Guro with the marker lē: (72) is an identifying construction and (73) is a construction with subject focus, which is formed similarly to a cleft construction, with a non-verbal identification construction as a main clause and a finite verbal clause as dependent clause.

235Guro

(72)

Kɔ̰́

.

house

ident

236‘It is a house.’ (Kuznetsova 2023: 232)

(73)

yɩ̀ɩ́

jɛ̄

lè

vɔ̌

ā̰

ɓā̰

ɓɛ̄.

yɩ̀ɩ́

jɛ̄

lē

vɔ̌

ā̰

ɓā̰

ɓɛ̄.

ref.ctr

disease

ident>3sg.sbj

penetrate\pfv

1sg

on

def

237‛It is THAT disease which attacked me (lit.: it is that diseases, it attacked me).’ (Kuznetsova 2023: 245)

238Thus, in Guro, a marker used in focus constructions derives from the identifying copula. In Guinean Maninka, in contrast, a focus marker acquires functions of a non-verbal predicator. The two types of development illustrate the grammaticalization paths cop > foc and foc > cop suggested by Creissels (2022, citing Idiatov p.c.).

239The Mano data shows however that the focal meaning does not characterize particular kinds of syntactic constructions. Indeed, the same types of structures with Subj – lɛ́ (or wɔ́) – clause – dem can have functions that are compatible with a predicative reading of lɛ́ and wɔ́ (as an identifier or a presentative predicator, especially when the referent is discourse new and/or is being pointed to), with a focal reading (when the referent of Subj represents a choice among available alternatives) or both. Both interpretations are particularly common in ostensive demonstration, when what is being demonstrated is an activity, rather than a referent. The systematic co-occurrence of a focal and a non-focal readings make it difficult to single out a distinct focus construction, let alone a focus marker. Therefore, there is no form-meaning correspondence between a focus construction and a focal meaning. Overall, the interpretation — in terms of focus as well —derives from the property of the referent and from the discourse context, rather than from formal marking and especially the choice of construction. The focal interpretation is not proper to a particular focus construction but rather is a result of cooptation (Mauri & Sansò 2011) of a construction with a broader meaning of attention drawing. Thus, instead of choosing between the two grammaticalization paths (cop > foc and foc > cop), I claim that lɛ́ and wɔ́ function, depending on context, as non-verbal predicators or as focus markers.

240By deriving a focal interpretation from another function, not related primarily to information packaging, I follow the bottom-up approach to information structure suggested by Matić and Wedgwood (2013) and Ozerov (2018). Connecting the information-structural uses of the markers lɛ́ and wɔ́ to other uses helps better understand their function in communication, which is irreducible to the rigid categories of information structure concentrated on information update. That the primary function of the markers under discussion is interactional, rather than strictly semantic or information-structural, is also compatible with Killian’s observation that predicative demonstratives fulfill a discourse function that is irreducible to just predicating identity or referent presentation: they “organize and frame chunks of text, such as signaling the opening or closing of texts, or recapitulating referents,” they are also used for indicating a speaker’s belief or attitude towards an idea” (Killian 2022: 7). Finally, should we decide to ascribe to lɛ́ and wɔ́ distinct statuses as focus markers, we would end up needing to explain why a language needs two focus markers, instead of one. Instead, adopting the logic of cooptation and communicative deployment of grammatical markers, we can see how two markers with different core values (presentation for wɔ́ and identification for lɛ́) can end up having a substantial area of functional overlap, in particular in their capacity to express focus.

241The last remark concerns the type of data used for this paper. Most of the examples are taken from a naturalistic corpus which includes, in addition to narrative data, data still quite rarely used in Mande studies: spontaneous interactions. Continuous recordings of picture-based elicitation sessions, which provide semi-spontaneous data on picture descriptions, have also been shown to be of great use. It is not impossible that the differences in analysis of information structure markers which I sketched in this section arise not only from different conceptual foundations (such as a tendency to treat information structure as a distinct category). Instead, it is plausible that by using comparable types of data, scholars of other Mande languages will arrive at similar analytical conclusions as those discussed in this paper.

Acknowledgements

242As usual, I am thankful to Pe Mamy for helping me in data collection and transcription and for endless discussions of the utterances and their contexts. Special thanks to Tilda Carbo for drawing the picture stimuli and to Albion Butters for proofreading. I would also like to thank Denis Creissels and Natalia Kuznetsova for insightful comments on earlier drafts of this paper. The paper was inspired by discussions with Don Killian, who also generously commented on an earlier draft, and with Pavel Ozerov, who drew my attention to the predicative function of lɛ́, which tied everything together, and suggested the invariant meaning–attention drawing (pun unavoidable). Research on this paper was supported by core funding of the Helsinki Collegium for Advanced Studies, and the picture questionnaire was collected and annotated thanks to a grant from the Kone Foundation (201907156).

Abbreviations

2431 – first person

2442 – second person

2453 – third person

246att – attention-managing marker

247conj – conjunctive

248cop – copula

249cstr – construct form

250ctr – contrastive

251d – determiner

252def – definite

253dem – demonstrative

254dim – diminutive

255el. – elicited

256ex – existential

257emph – emphatic

258ex – existential

259foc – focus

260fr – French

261ger – gerund

262h – high tone

263ident – identifier

264inf – infinitive

265inj – interjection

266int – intensifier

267ip – inclusory pronoun

268ipfv – imperfective

269jnt – conjoint

270loc – locative

271MOC – Mano Oral Corpus

272neg – negative

273pfv – perfective

274pl – plural

275poss – possessive

276ppm – pronominal predicative marker

277pres – presentative

278prf – perfect

279PropN – proper noun

280pst – past

281ref – referential

282refl – reflexive

283sbj – subject

284sbjv – subjunctive

285sg – singular

286vis – visual

The stimuli are given in the order they appear in the paper. The author of the stimuli is Matilda Carbo.

Stimulus 1 regarder_se. See example 45.

Stimulus 1 regarder_se. See example 45.

Stimulus 2, brosser_dents_pl. See example 47.

Stimulus 2, brosser_dents_pl. See example 47.

Stimulus 3, regarder. See examples 48 and 54.

Stimulus 3, regarder. See examples 48 and 54.

Stimulus 4, montrer_maitresse. See example 49.

Stimulus 4, montrer_maitresse. See example 49.

Stimulus 5, bol. See example 52.

Stimulus 5, bol. See example 52.

Stimulus 6, mettre_enfant. See example 59.

Haut de page

Bibliographie

References

Aboh, Enoch Oladé. 2007. Focused versus non-focused wh-phrases. In Enoch Oladé Aboh, Katharina Hartmann & Malte Zimmermann (eds.), Focus strategies in African Languages: The interaction of focus and grammar in Niger-Congo and Afro-Asiatic, 287–314.

Creissels, Denis. 2019. Existential predication in the languages of the Sudanic belt. Afrikanistik-Aegyptologie-Online. https://www.afrikanistik-aegyptologie-online.de/archiv/2019/4860/view.

Creissels, Denis. 2022. A typological rarum in Mande languages: Argument-predicate reversal in nominal predication. Mandenkan 68. 3–42. https://doi.org/10.4000/mandenkan.2780.

Diessel, Holger. 1999. Demonstratives: form, function and grammaticalization. Amsterdam: John Benjamins.

Haspelmath, Martin. 2022. Nonverbal clause constructions. https://www.academia.edu/71539642/Nonverbal_clause_constructions.

Hengeveld, Kees. 2011. Non-verbal predication: Theory, typology, diachrony. De Gruyter Mouton. https://doi.org/10.1515/9783110883282.

Kalbertodt, Janina, Beatrice Primus & Petra B. Schumacher. 2015. Punctuation, prosody, and discourse: Afterthought vs. right dislocation. Frontiers in Psychology 6. https://doi.org/10.3389/fpsyg.2015.01803.

Khachaturyan, Maria. 2020. Common ground in demonstrative reference: The case of Mano (Mande). Frontiers in Psychology 11. https://doi.org/10.3389/fpsyg.2020.543549.

Khachaturyan, Maria. forthc. Negation in Mano. In Ljuba N. Veselinova & Matti Miestamo (eds.), Negation in the languages of the world. Language Science Press.

Killian, Don. 2022. Towards a typology of predicative demonstratives. Linguistic Typology 26(1). 1–41. https://doi.org/10.1515/lingty-2021-2078.

Kuznetsova, Natalia. 2023. Homonymy and polysemy of lē in Guro: identificational, quotative, complementiser, focus and relative-possesive functions. In Ivan Kapitonov, Maria Khachaturyan, Sofya Oskolskaya, Nina Sumbatova & Samira Verhees (eds.), Song and trees. Papers in memory of Sasha Vydrina, 227–284. Saint Petersburg: ILS RAN.

Lambrecht, Knud. 2001. A framework for the analysis of cleft constructions. Linguistics 39(3). 463–516. https://doi.org/10.1515/ling.2001.021.

Matić, Dejan & Daniel Wedgwood. 2013. The meanings of focus: The significance of an interpretation-based category in cross-linguistic analysis. Journal of Linguistics 49(1). 127–163. https://doi.org/10.1017/S0022226712000345.

Mauri, Caterina & Andrea Sansò. 2011. How directive constructions emerge: Grammaticalization, constructionalization, cooptation. Journal of Pragmatics 43. 3489–3521. https://doi.org/10.1016/j.pragma.2011.08.001.

Miestamo, Matti. 2005. Standard negation: The negation of declarative verbal main clauses in a typological perspective (Empirical Approaches to Language Typology 31). Berlin: Mouton de Gruyter.

Ozerov, Pavel. 2018. Tracing the sources of Information Structure: Towards the study of interactional management of information. Journal of Pragmatics 138. 77–97. https://doi.org/10.1016/j.pragma.2018.08.017.

UBS. 1978. Wè-dɔ-wè dɛɛ lɛ́ lɛ Wálà wà mia nì o fĩ̀ã̀ā. New Testament in Mano. Translation by June M. Hobley Jackson and Stanley Younguoi. Monrovia: The Bible Society in Liberia.

Vydrin, Valentin. 2020. Non-verbal predication and copulas in three Mande languages. Journal of West African Languages 47(1). 77–105.

Vydrina, Alexandra. 2020. Topicality in sentence focus utterances. Studies in Language 44(3). 501–547. https://doi.org/10.1075/sl.18069.vyd.

Haut de page

Annexe

Appendix. Visual stimuli used in the study.

Haut de page

Notes

1 att stands for attention management. The function of attention management is not discussed in this paper. See Section 4 for a discussion of the glosses.

2 In Valentin Vydrin’s terms, PPMs are defined as elements which occur only in verbal predications. If an element can occur in both verbal and non-verbal predications, as existential auxiliaries in Mano, it is not ascribed to the class of PPMs but to the class of bifunctional auxiliaries (Vydrin 2020). To avoid postulating a part of speech with only one series of elements, I depart from this taxonomy and define existential auxiliaries as PPMs.

3 According to Denis Creissels (2022, p.c.), they are semantically similar to either referential identity or to class-membership constructions, depending on the type of predication involved. The NP in such constructions should be analyzed as the predicate (the predicated class or the identical referent), while the argument (what is being identified) is absent. Although such analysis is very plausible from a semantic point of view, it runs into a difficulty. In the presentative constructions, the NP preceding the predicator should be considered as its argument. Crucially, presentative and identifying constructions in Mano (and in some other languages as well; Killian 2022) share the same predicators and the difference between the two interpretations is not always easy to draw, especially in the case of invisible and abstract referents (see ex. 31). Thus, if we were to adopt Creissels’ syntactic analysis, we would have to deal with a marker included in two dramatically different syntactic structures with still highly compatible and often ambiguous interpretations. Because the fuzzy boundary between presentative and identifying functions, especially in the focus reading, is a central object of this paper (see Section 3.2.1), I prefer to adopt the same analysis for both these constructions by treating the NP preceding the predicator as an argument, while recognizing that for identifying constructions, postulating an implicit argument instead has its advantages. I leave it for future studies to propose a syntactic analysis of identifying constructions which would address both the problem of the implicit argument and the proximity between identifying and presentative constructions.

4 As discussed by Creissels (2019), this is a typical feature of the languages of the Sudanic Belt.

5 It is not entirely clear whether tɛ́ is a variant of lɛ́ or a separate morpheme with largely overlapping functions.

6 This usage of tɛ́, which is likely a variant of lɛ́, is not covered by the present paper. Here, tɛ́ is used in a construction NP tɛ́ dem. Similarly to the construction examined in Section 3.1, lɛ́ (and, more specifically, its variant tɛ́) is used to identify the referent. The broader function of this construction in discourse, however, is not identification as a separate predication, but drawing attention to a referent. Such attention-drawing can be accompanied by a pointing gesture.

7 The crucial difference is that in the sense used by Creissels, true argument-predicate reversal constructions are only class-membership constructions. I use the term “argument-predicate reversal” for all constructions where the semantic predicate is put in the syntactic position of the argument, be it class-membership, referential identity constructions or that are ambiguous in terms of the two reading.

8 Or, rather, the connection between predicative and adnominal demonstratives is not paradigmatic but diachronic in nature: the exophoric demonstrative tɔ́ɔ̄ likely derives from fusion between the predicator tɛ́ (a variant of lɛ́) and the demonstrative wɛ̄.

9 Unfortunately, Creissels does not offer a sufficient explanation why he treats the PP Làamínì tí in (67b) as an afterthought, as a non-fully integrated constituent: whether such an interpretation emerges on a residual basis, from the fact that what appears to be new, focal information is put in the beginning of the utterance and whatever is added thereafter should a fortifiori be treated as afterthought, or because there is independent evidence for such treatment. Such independent evidence could be provided by prosodic effects (Kalbertodt, Primus & Schumacher 2015).

Haut de page

Table des illustrations

URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-1.png
Fichier image/png, 66k
Titre Stimulus 1 regarder_se. See example 45.
URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-2.png
Fichier image/png, 57k
Titre Stimulus 2, brosser_dents_pl. See example 47.
URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-3.png
Fichier image/png, 57k
Titre Stimulus 3, regarder. See examples 48 and 54.
URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-4.png
Fichier image/png, 95k
Titre Stimulus 4, montrer_maitresse. See example 49.
URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-5.png
Fichier image/png, 122k
Titre Stimulus 5, bol. See example 52.
URL http://journals.openedition.org/mandenkan/docannexe/image/3070/img-6.png
Fichier image/png, 87k
Haut de page

Pour citer cet article

Référence papier

Maria Khachaturyan, « From copula to focus, vice versa, or neither? »Mandenkan, 69 | 2023, 87-120.

Référence électronique

Maria Khachaturyan, « From copula to focus, vice versa, or neither? »Mandenkan [En ligne], 69 | 2023, mis en ligne le 02 septembre 2023, consulté le 09 février 2025. URL : http://journals.openedition.org/mandenkan/3070 ; DOI : https://doi.org/10.4000/mandenkan.3070

Haut de page

Auteur

Maria Khachaturyan

University of Helsinki

Articles du même auteur

Haut de page

Droits d’auteur

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC BY-NC-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search