by Nina Dobrushina and Johan van der Auwera and Valentin Goussev
In this chapter, the term optative will refer to an inflected verb form dedicated to the expression of the wish of the speaker. An example comes from Bagvalal (Nakh-Daghestanian; eastern Caucasus):
(1) Bagvalal (Dobrušina 2001: 328-329)
Den
hanč’u-b
b-as-inā-X
b-is-a-nā,
dē
w-ič’a-be-la!
1sg.erg
lying-n
n-tell-impf-conv
n-find-pot-cond
1sg
m-die-imp-opt
‘Would that I died if I’m lying!’
This usage of the label optative is rather uncontroversial (Bybee et al. 1994: 179, Chung and Timberlake 1985: 247, Sadock and Zwicky 1985: 164), but other terms are also found in descriptive grammars. The speaker’s wish can be expressed by different grammatical means across languages, for instance, by the use of modal verbs, such as English may in (2), or by a Russian elliptical conditional in (3):
(2)
May God always help you!
(3) Russian (personal knowledge)
Jesli
by
ona
vernu-l-a-s’!
if
subj
she
return-pst-f-refl
‘If only she came back!’
This chapter restricts attention to languages with a dedicated inflectional optative form as in (1), and disregards non-inflectional means for expressing the speaker's wish such as (2) and (3).
The two values shown on the map are given in the feature-value table:
| Go to map | ||
|---|---|---|
| Value | Representation | |
 |
Inflectional optative present | 48 |
 |
Inflectional optative absent | 271 |
| Total: | 319 | |
Though these values seem straightforward, in practice there are various problems. On the one hand, constructions are sometimes called “optative” which we would call something else. The construction called “optative” in Slave by Rice (1989: 1111), for instance, has a very wide range of uses, and the expression of a wish does not seem to be particularly central. Whatever this is, it is not our optative. Very striking in this respect is also the tradition in Turkology to apply the term “optative” to what we would treat as a first person imperative or hortative (Tenišev et al. 1997: 212, 364, 402). A tendency to refer to certain imperative categories with the label “optative” is found outside Turkic grammars as well. Thus, in Fortescue’s (1984: 24) description of West Greenlandic, the set of what we take to be 1st person singular, 1st person plural exclusive and 3rd person imperatives or hortatives is distinguished from other imperative forms on formal grounds and called “optative”. On the other hand, the semantic category which fits our notion of optative is often listed under another label, such as “desiderative” in Aymara (Hardman de Bautista 1974: 227), Guaraní (Gregores and Suarez 1967: 132), and Maricopa (Gordon 1986: 110), and “benedictive” in Ladakhi (Koshal 1979: 226).
It is important to note that the first value is assigned to languages that have the optative as an inflectional category. This means that it is not sufficient for the language to exhibit a dedicated optative construction. Consider Mao Naga (Tibeto-Burman; India). Its “blessing and cursing” optatives involve combining the particle peno with an imperative verb form.
(4) Mao Naga (Giridhar 1994: 314)
Pfo
peno
thi
mozhu-ṭio!
he
opt
die.imp
soon
‘May he die soon!’
This peno construction is dedicated to the optative function, but it is not inflectional.
A dedicated optative form may occur in a special syntactic construction. For instance, the optative in Kumyk (Turkic; eastern Caucasus) is a combination of a participle in –gaj/-ɣaj and an auxiliary verb edi :
(5) Kumyk (Gadzhiaxmedov 2000: 246)
Jaxšy
jangur
jav-ɣaj
edi.
good
rain
rain-opt
cop
‘Would that it rained!’
The strategy is analytic, but we would nevertheless say that Kumyk has optative morphology: the suffix -gaj/-γaj is fully dedicated to the expression of optatives — it is indeed the optative marker most typical for Turkic in general. Until recently, the Kumyk optative occurred without an auxiliary, but nowadays the pattern without auxiliary sounds obsolete. Even so, the need for the auxiliary does not change the fact that there is morphology that is specifically optative in meaning.
Another restriction is that we only consider forms that are available for all persons. For instance, although Bashkir (Turkic; Russia) has a 2nd person optative to express a wish that something bad should happen (curse) (Juldašev 1981: 288), it is shown on the map as lacking an optative.
It is important also to distinguish the optative from related categories. We will here discuss subjunctives, desideratives and imperatives (hortatives).
The notion of subjunctive, first of all, is itself a problematic notion. One can choose to define it as a mood that is typical for subordinate clauses, or at least some subtypes, or one can associate it with non-actuality and thereby bring it closer to notions like “irrealis” and “potentialis”. Either way, there is a link with the expression of wishes. First, with Russian we have just illustrated in (3) the fact that the optative may use subordination strategies, and the very combination of a past tense and the particle by seen in (3) has been called “subjunctive”. Of course, this Russian subjunctive is not a morphological category, and the Russian facts pose no problem for us. The link between the expression of wishes and the subjunctive category is, however, independent of whether the subjunctive is realized by a syntactic pattern as in Russian or by morphology only, as in Eastern Armenian.
(6) Eastern Armenian (Kozintseva 2001: 247)
Du
oɣĵ
mnas,
aɣĵik-s.
you
well
remain.subj.fut.2sg
daughter-poss
‘I wish you good health, daughter.’
To distinguish the morphological optative – the topic of this chapter – from the wish-expressing morphological subjunctive, such as that of Armenian (6), we stipulate that the expression of the wish must be the main function. This is sometimes clear enough; but when the category called “optative” has other uses, too, the difference between a multifunctional optative and a multifunctional subjunctive may become very small, and grammarians may disagree. Swahili, for instance, is commonly taken to have a multifunctional subjunctive – Wilson (1970: 118-121) lists as many as seven functions – but Loogman (1965: 206-209) speaks of the “Wishing-Form”. Multifunctional optatives used in non-wishing contexts are illustrated by (7) from Archi (Nakh-Daghestanian; eastern Caucasus) (a conditional use) and by (8) from Maricopa (Yuman; Arizona) (what Palmer (1986: 110) calls a “deliberative question” use).
(7) Archi (Kibrik et al. 1977: 227)
Noc’
sas
et̅itʼu-t̅an,
immaq˚’i.
bird
catch
be.able-opt
leave.imp
‘If you can’t catch the bird, just leave it’
(8) Maricopa (Gordon 1986: 110)
Kawish
'-we-lya?
what
1-do-opt
‘What can I do?’
Another notion related to optative is desiderative (Plungian 2000: 317) (see also Chapter 124). Both categories involve the expression of a wish, but whereas the optative is used for the wish of the speaker, the desiderative is used for the wish of a participant in the state of affairs referred to in the utterance. A language with a dedicated inflectional strategy for the desiderative is Nganasan (Songhay, Uralic; northern Russia).
(9) Nganasan (Valentin Goussev, field notes)
D’ojbaru
i-n’antu-g-tu-m/-Ø.
orphan
be-desid-iter-pres-1sg.sbjv/-2sg.sbjv
‘I / you want to be an orphan.’
The third related notion is the imperative(see also Chapter 72). Both the imperative and the optative refer to a wish of the speaker. With the optative, the state of affairs wished for is typically outside the sphere of influence of the speaker; see (10a) from Aghul (Nakh-Daghestanian). With an imperative, however, the speaker launches an appeal to the hearer to fulfill the wish. In Aghul, the optative simply cannot be used in this situation, see (10b).
(10) Aghul (Solmaz Merdanova, p.c.)
a.
Wun
ǯehlem-di-s
ušu-raj!
you
hell-obl-dat
go-opt
‘Go to hell!’
b.
*Wun
mexteb-i-s
ušu-raj!
you
school-obl-dat
go-opt
‘Go to school!’
In other languages, the optative is allowed, but the appeal still cannot be a simple command. In Santali (Munda; India; data obtained with the help of Dmitrij Sitchinava), optatives are allowed for addressing the hearer and the effect is that of a mild suggestion.
(11) Santali (Bodding 1952: 73)
Mit’
taka-m
em-k-iñ-a?
one
rupee
give-opt-1sg-ind
‘Shouldn’t you give me a rupee?’
The problem is that there may be an overlap between optatives and imperatives or hortatives in the third person. Third person imperative (hortative) may be used as third person optative and vice versa. Consider the uses of the third person imperative in Mapudungun (Chile). A true imperative use is shown in (12a) and a semantically optative one is shown in (12b).
(12) Mapudungun (Smeets 1989: 234)
a.
Chewrumé
rüngkü-pe!
wherever
jump-imp.3
‘Let him jump wherever he wants to!’
b.
Küme-le-pe
ñi
fochüm!
good-stat-imp.3
my
son
‘May my son be all right!’
In Lealao Chinantec (Oto-Manguean; Oaxaca, Mexico; Rupp 1989: 95) and in Lezgian (Nakh-Daghestanian; Haspelmath 1993: 151-152), on the other hand, it is the optative third person that does double duty.
There are languages with more than one morphological optative, and in that case they will have different meanings. Thus in Abkhaz (Northwest Caucasian), one of the optatives, glossed as opt1 here (Hewitt's (1979: 198) "subjunctive"), is used to curse and to bless, while the other one, the opt2, is used to express a wish or a dream of the speaker.
(13) Abkhaz (Hewitt 1979: 198; Hewitt, p.c.)
a.
Bzia
wy-ba-aait'!
good
you-see-opt1
‘Greetings!’ (lit. ‘May you see something good!’)
b.
Ma
a-dzə
l-zhʷə-nda.
pcl
def-water
3.f-drink-opt2
‘I wish she‘d drink the water.’
The map documents 319 languages. Worldwide, so it seems, morphological optatives are a fairly infrequent phenomenon (48 languages in our sample). Our sample does not contain any languages with inflectional optatives from Australia or from Africa; we have just two cases in Europe (Albanian and Karaim), one in New Guinea, and only a few in the Americas and Asia outside of the Caucasus and the area of northern India and Nepal.
As far as the Caucasus is concerned, the optative was first of all found in all of the North-Caucasian languages for which we were able to find reliable information. Northwest Caucasian languages, furthermore, seem highly unusual in that they have two optatives; Ingush and Chechen, which are Northeast Caucasian (= Nakh-Daghestanian), have two optatives too, though one of them is unproductive (Johanna Nichols, p.c.). Unfortunately, data from the South Caucasian (= Kartvelian) languages proved less accessible. Georgian, at least, does have a morphological optative. Whether the Caucasian optative is a phenomenon due to areal contact or to a common origin is not clear, in part because the genetic affinity of the Northwest and Northeast languages is still under discussion. Interestingly, both Turkic languages of the Caucasus, Karachay-Balkar and Kumyk, have a full personal paradigm of a morphologically dedicated optative, which is untypical for Turkic.
A second area that seems to be characterized by the morphological optative is northern India and Nepal. The optative occurs in some of the Tibeto-Burman and Munda languages, and interestingly, the only Indo-Iranian language we found to have a morphological optative is Nepali (Korolev 1965: 80), probably as a result of the influence of neighboring Tibeto-Burman languages.