by Johan van der Auwera and Andreas Ammann
In this chapter, we will examine the expression of the equivalents of English may and can as in (1).
(1)
a.
The children can swim across the lake.
b.
You may leave now.
c.
One can get to Staten Island using the ferry.
In (1a) can expresses ability and in (1b) may expresses permission. Both permission and ability have to do with a notion of possibility. These two kinds of possibility belong to a subtype which also includes readings like can in (1c), where neither an individual’s ability nor an authority granting permission is involved. What the three examples have in common is that the speaker merely describes a possibility that exists in a given situation. For this reason we call this subtype situational (also termed “deontic” or “root” in the literature). It is different from the kind of possibility illustrated in (2).
(2)
John may have arrived.
In (2) the possibility concerns the speaker’s evaluation of the truth-value of a proposition about a situation, that is, it expresses a degree of the speaker’s certainty. This kind of possibility is commonly called epistemic (see Chapter 75).
Possibility stands in direct contrast with necessity, and together they constitute the domain of modality. Just as there are two types of possibility, there are two types of necessity, a situational one, as in (3a), and an epistemic one, as in (3b).
(3)
a.
John has to leave now.
b.
John must have arrived.
The map gives some indication about the kind of strategy used to express situational possibility in positive main clauses. Three types of languages are distinguished. The definitions make reference to three marking strategies: affixes on verbs, verbal constructions, and other constructions, ordered in terms of increasing cross-linguistic frequency. The first type on the map comprises languages that use the least frequent strategy, whether or not they also use the more frequent ones. Languages of the second type use the medium frequency (but not the low frequency) strategy, whether or not they also use the most frequent ones. Languages of the third type use only the most frequent strategy.
| Go to map | ||
|---|---|---|
| Value | Representation | |
 |
The language can express situational possibility with affixes on verbs | 63 |
 |
The language cannot express situational possibility with affixes on verbs, but with verbal constructions | 158 |
 |
The language cannot express situational possibility with affixes on verbs or with verbal constructions, but with other kinds of markers | 13 |
| Total: | 234 | |
If one compares the frequency of the types of markers and the frequency of the types of languages, the most remarkable fact is that languages of type three, which only use the most frequent type of marking, are not very frequent.
The first type includes languages that express situational possibility with verbal affixes. An example is Lavukaleve (Solomons East Papuan; Solomon Islands) and the relevant affix is -nen~-nan. In the context for sentence (4), the speaker has just heard that a feast is planned.
(4) Lavukaleve (Terrill 2003: 344)
Valai!
Ngai
a-vo-nen.
how
I
1sg.subj-come-pos
‘What? Can I come?’
Verbal affixes may differ in their degree of dedication to the modal meaning. “Less dedicated” means that the marker has other meanings, or – in part depending on the point of view of the grammarian – that the marker has only one meaning but various uses, with a meaning that is vague and thus wider than just situational possibility. These other meanings or uses are often future or irrealis: if you can or must do something, you normally have not done it yet. An example of a morphological verbal marker that denotes both future and situational possibility is shown from Daga (Dagan; Papua New Guinea).
(5) Daga (Murane 1974: 58)
War-ait.
get-fut.1pl
‘We will get (it).' or ' We can get (it).’
It is not always easy to decide whether a marker is a verbal affix or a free word, a verb or a particle, for instance. For Gadaba (Dravidian; India) the element -occuni is considered to be an invariant modal (subjunctive) suffix by Bhaskararao (1998: 349), yet it is written as a separate word. Since we take it that it is more interesting to draw attention to a low frequency phenomenon (affixal coding) than a higher frequency phenomenon (verbal construction), Gadaba is listed as a language that can express situational possibility with an affix on the verb.
The second type includes languages that do not express situational possibility with verbal affixes, but that employ verbal constructions. This type is well-known from Standard Average European. French pouvoir ‘may/can’ is clearly a verb, appearing as a 1st person singular present in (6a), as a 2nd plural future in (6b), and as a past participle in a ‘have’ perfect in (6c).
(6) French
a.
Je
peux
le
lui
donner.
I
can.ind.pres.1sg
it
him
give
‘I can give it to him.’
b.
Vous
pourrez
me
visiter.
you.pl
can.ind.fut.2pl
me
visit
‘You will be able to visit me.’
c.
J’
ai
pu
le
voir.
I
have.ind.pres.1sg
can.pret.ptcp
him
see
‘I was able to see him.’
In Taba (Austronesian; Halmahera, Indonesia), the modal is a verb, too, but it differs from French in that it can take part in a serial verb construction, i.e. a concatenation of two finite verbs. In (7) the “modal evaluator” -ahan ‘to be able’ is juxtaposed to a lexical verb which bears the same inflection marker (that of the 3rd person singular).
(7) Taba (Bowden 2001: 316)
N=pe
n=ahan.
3sg=do
3sg=be.able
‘He can do it’.
Then there are languages in which the modal verbs have reduced verbal properties – or special ones – so that grammarians may assign them labels like “irregular verb” or “auxiliary verb” and thus distinguish them from regular and lexical verbs. Criteria for such a status, however, differ from language to language and therefore no such distinction will be made here.
Not all verbal constructions that can express situational possibility are dedicated to this function. French pouvoir, for instance, can also serve to convey epistemic possibility (see Chapter 76).
In the third type, the language has neither verbal affixes nor verbal contructions to express situational possibility, but there are other markers. These markers may be particles, adverbials, nouns, or adjectives, as well as some more complex clausal constructions. Chukchi (Chukotko-Kamchatkan; eastern Siberia) makes use of a particle mecənkə for the expression of situational possibility.
Mecənkə
mət-ra-jalɣət-ɣʔa.
enough
1pl-fut-move.camp.th
‘We can move camp.’
An adverbial marker is illustrated with Slovene.
(9) Slovene (Derbyshire 1993: 109)
Lahkó
me
poklíčete,
kàdar
kóli
želíte.
easily
me
call.ind.pfv.pres.2pl
when
ever
wish.ind.ipfv.pres.2pl
‘You may call me whenever you wish.’
An adjectival use is exemplified with the adjective ihoc ‘able’ from Amele (Madang; Papua New Guinea) in (10).
(10) Amele (Roberts 1987: 265)
Ija
nu-ec
nu
ihoc.
I
go-inf
for
able
‘I am able to go.’
In Korean (11), it is a noun that is to be associated with the situational possibility meaning.
Halapeci-nun
wuncenha-si-l
swu(-ka)
iss-usi-ta.
grandpa-top
drive-subj.hon-pres
way(-nmlz)
exist-subj.hon-decl
‘My grandpa can drive.’
Note that the possibility marker in (11) is strictly speaking not the noun on its own, but the combination of the noun and the existential copula. In this sense, one could argue for coding it as belonging to the verbal construction type. We will not do this, however: in (11) it is not so much the copula that has a modal sense, but the noun.
For an example of a complex clausal type we can also turn to Korean. (12) literally says that even if you go, it is good.
Ne-nun
ka-to
coh-ta.
you-top
go-even.if
good-decl
‘You may go.’
Just as with modal affixes and modal verbal constructions, the markers of the third type need not be fully dedicated to the expression of situational possibility. The Chukchi particle mecənkə, for instance, illustrated in (8), can also mean ‘enough’ (Dunn 1999: 76, and p.c.).
Languages that cannot express situational possibility in the verbal domain, whether with separate verbal constructions or with affixes on verbs, seem rare. They are not reported for Europe and most of Eurasia, nor for Africa, South Asia or South America. Elsewhere they are scattered.
Languages that may express situational possibility with verbal affixes are more frequent, particularly in North America and in New Guinea. It is not surprising that this group includes polysynthetic languages, among them West Greenlandic (Eskimo), Kiowa (Kiowa-Tanoan: central United States), Nunggubuyu (Gunwinyguan; Northern Territory, Australia), Yimas (Lower Sepik-Ramu; Papua New Guinea) and Ket (Yeniseian; Russia). Note, however, that the existence of high polysynthesis need not entail expressing situational possibility with verbal affixes. Thus Chukchi or Lakhota (Siouan; Nebraska and Minnesota), both having a high degree of synthesis, do not seem to choose verbal affixes for expressing situational possibility.
The most frequent type of language allows the expression of situational possibility with the use of more or less dedicated verbal constructions. This type is particularly common in Europe, Africa, South Asia and Southeast Asia. It is rare only in New Guinea and in Australia.
Situational possibility and, more generally, situational modality have received considerable attention in the study of European languages in particular, because the markers employed in these languages are multifunctional and they have properties that clearly single them out as a special subclass of verbs, i.e. auxiliary verbs. English is one of these languages, and because English is one of the best-studied languages, the literature on the modals of English is voluminous. The multifunctionality of the English and English-like modals has been of interest for both semanticists (e.g. Palmer 1979, Coates 1983) and historical linguists (e.g. Traugott 1989), and the latter nowadays usually approach the topic from a grammaticalization point of view. Within grammaticalization theory typological work has been done by Bybee et al. (1994) and by van der Auwera and Plungian (1998). The status of auxiliaries has been approached within that framework as well (Heine 1993), but it has also received much attention from formal grammarians, from Ross (1969) onwards.