1. Definition of values

This map shows the grammaticalization of a prohibition addressed to a single addressee. The category will be called the “second singular prohibitive”. Four values will be distinguished.

Values of Map 71A. The Prohibitive

Go to map
	Value	Representation
	The prohibitive uses the verbal construction of the second singular imperative and a sentential negative strategy found in (indicative) declaratives	113
	The prohbitive uses the verbal construction of the second singular imperative and a sentential negative strategy not found in (indicative) declaratives	182
	The prohibitive uses a verbal construction other than the second singular imperative and a sentential negative strategy found in (indicative) declaratives	55
	The prohibitive uses a verbal construction other than the second singular imperative and a sentential negative strategy not found in (indicative) declaratives	146
Total:		496

In the first type of language shown on the map, the second singular prohibitive is a combination of the verbal construction used for the second singular imperative plus a sentential negative strategy found in declaratives, or in indicative declaratives, if the notion of indicative is relevant. This type can be illustrated with Turkish.

(1) Turkish (Ceyhan Temürcü, p.c.)

a.	Okul-a	git!
	school-dat	go.imp.2sg
	‘Go to school!’
b.	Okul-a	git-me!
	school-dat	go.imp.2sg-neg
	‘Don’t go to school!’
c.	Okul-a	gid-iyor-sun.
	school-dat	go-cont-ind.pres.2sg
	‘You are going to school’
d.	Okul-a	git-m-iyor-sun.
	school-dat	go-neg-cont-ind.pres.2sg
	‘You are not going to school’

The prohibitive second singular git-me  in (1b) is composed of the imperative second singular shown in (1a) and the negative me, illustrated in the indicative declarative in (1d). As the translation indicates, English belongs to this type as well. What makes English special is that the negation not only involves the negative word not, but also the auxiliary do, and that is the auxiliary which enters the construction of the second singular imperative.

The second type employs a second singular imperative construction, but the sentential negation strategy is not found in declaratives or, if the notion of indicative is relevant, in indicative declaratives. In Vietnamese, declaratives are negated with chăng  or không, whereas prohibitives use the negations chó  and dung. With respect to the verbal forms used, however, imperatives and prohibitives do not differ, nor do they differ from declaratives.

(2) Vietnamese (Thompson 1965: 221; Victoria Rosén, p.c.)

a.	Uông	ruou!
	drink	alcoholic
	‘Drink alcohol!’ or  ‘I/you/he/etc. is/are drinking alcohol.’
b.	Chó	uông	ruou!
	neg	drink	alcoholic
	‘Do not drink alcohol!’
c.	Không	uông	ruou.
	neg	drink	alcoholic
	‘I/you/he/etc. are not drinking alcohol.’

In the third type, the second singular prohibitive employs a sentential negative that is used in declaratives or, if the notion of indicative is relevant, in indicative declaratives, but the verbal construction is not the one found in second singular imperatives. In Spanish the sentential negation for both negative indicative declaratives and prohibitives is the marker no. However, the prohibitive verb is not an imperative, but a subjunctive.

(3) Spanish

a.	Pedro	canta.
	Pedro	sing.ind.pres.3sg
	‘Pedro sings.’
b.	Pedro	no	canta.
	Pedro	neg	sing.ind.pres.3sg
	‘Pedro does not sing.’
c.	Canta!
	sing.imp.2sg
	‘Sing!’
d.	No	cantes!
	neg	sing.sbjv.pres.2sg
	‘Don’t sing!’

In the fourth type, the verbal construction of the second singular prohibitive is different from that of the second singular imperative, and the negative strategy is different from the sentential negative found in declaratives or, if the notion of indicative is relevant, in indicative declaratives. In Zulu (Bantu; South Africa) the imperative second singular ends in -a. In the prohibitive second singular, however, we do not find an imperative form but an infinitive, combined with a special negative marker musa, different from the indicative declarative sentential negative marker.

(4) Zulu (Poulos and Bosch 1997: 19; Khosi Mnyakeni, p.c.)

a.	Shay-a	inja!
	hit-imp.2sg	dog
	‘Hit the dog!’
b.	Mus-a	uku-shay-a	inga!
	neg.imp.aux-2sg	inf-hit-inf	dog
	‘Do not hit the dog!’
c.	U-ya-shay-a	inja.
	2sg-ind.pres-hit-pres	dog
	‘You hit the dog.’
d.	A-wu-shay-i		inja.
	neg.ind.pres-2sg-hit-neg.ind.pres		dog
	‘You do not hit the dog.’

Several complications occur. First, the prohibitive negation may be partially identical to an (indicative) declarative negation. In Yakoma (Adamawa-Ubangian; Central African Republic) there are four declarative negations and two prohibitive ones. The latter consist of two parts, and the second part is in each case identical to a declarative negation. Negative markers in Yakoma are given in example (5).

(5) Yakoma (Boyeldieu 1995: 131-132)

a.	Declarative	b.	Prohibitive
	mɔ́ … mɔ̄		tá … mɔ̄
	mɔ̄
	lá ̀
	āpɛ̀		tá … āpɛ̀

Partial similarity implies difference, however, and a language like Yakoma will therefore be taken to have a special prohibitive negation.

A second complication is that a language may have more than one second singular imperative. In this chapter, however, we restrict our attention to the most common or neutral imperative. Special future or special polite imperatives, for instance, are not included. This imperative, however, must have a corresponding prohibitive. Thus Kerek (Chukotko-Kamchatkan; eastern Siberia) has both perfective and imperfective imperatives, and the perfective imperative is more frequent than the imperfective imperative (Volodin 2001: 150). However, the prohibitive only has an imperfective paradigm (Volodin 2001: 153). Hence for Kerek, only the imperfective has been taken into consideration.

A third complication is that a language may possess more than one prohibitive strategy. In case the strategies are of the same type, there is no problem. In Yakoma, there are two prohibitive markers, but both are different from the declarative ones, so the language can be said to be of the special prohibitive negation type. Whether there is just one such negative or two is irrelevant. But the strategies may also be of different types. In Apalaí (Carib; Brazil), the prohibitive normally uses the copula, as in negative declaratives, and the declarative negative suffix -pyra. But there is also a pattern with a special prohibitive prefix os-, which combines with the verb in its immediate past form.

(6) Apalaí (Koehn and Koehn 1986: 62)

a.	Yto-pyra	eh-to-ko!
	go-neg	be-2pl-imp
	‘Do not go (you all)!’
b.	Tupito	epery	os-enah-no,	n-ase.
	field	fruit	neg-eat-ind.imm.pret	3-say.ind.pres
	‘ “Don’t eat the fruit of the field”, he said.’

This second strategy is considered to be rare, however. In such cases, we have coded the language according to its most frequent pattern. For all the relevant languages on the map, we have made a frequency-based decision.

Note that we do not rule out the possibility that a language may have a special prohibitive negation which has an additional function in non-negative contexts. Iraqw (Southern Cushitic; Tanzania) distinguishes between the non-prohibitive negative -ká  and the prohibitive m–  (Mous 1992: 151, 168). This m–, however, is also used to make questions, and Mous (1992: 152) thinks that this is no accident, for there are both formal and semantic similarities.

2. Geographical distribution

It turns out that most of the world’s languages do not form second singular prohibitives with a combination of the second singular imperative and the declarative (indicative) negation. This is not to say that this pattern is rare. It is well attested in the Americas and in non-Bantu Niger-Congo; in Eurasia it is typical for Germanic, Slavic and Turkic. However, the strategy is rare to non-existent in South and Southeast Asia and in Africa, other than in the northern non-Bantu Niger-Congo belt. Of the three remaining strategies, the one that is least common is to use the (indicative) declarative negation in combination with a verbal form that is not found in the second singular imperative. In Europe, this is typical for the Romance languages; it is sporadically found across the globe, except for South Asia, Southeast Asia, and New Guinea. The strategy of combining the second singular imperative with a special prohibitive negation is widespread. It seems typical for Celtic and perhaps also Khoisan. Areally, it is most typical for Southeast Asia and the Far East. Finally, it is also common to have both a special prohibitive negation and a verbal form different from the one used for the second singular imperative. This is found everywhere except in western Europe (apart from the Balkans) and in Southeast Asia and the Far East. There seems to be a cluster in the southern and eastern part of India. One can also generalize over languages with a verbal construction not found in indicative declaratives, whether or not they also employ a special prohibitive negation. This is typical for the Bantu languages, as has been remarked by Kamba Muzenga (1981: 256-265). A generalization about languages that have a special prohibitive negation, whether or not the verbal construction differs from the one used in second singular imperatives, holds for India. Given the relatively small number of languages investigated, it is difficult and usually impossible to make statements about smaller areas or language families. Just one illustration: with just two Arawakan languages on the map, Bare and Warekena, which both employ a special prohibitive negative strategy, we cannot improve upon Aikhenvald’s (1999: 96) statement that the Arawak family as a whole tends to have a special prohibitive negative strategy.

3. Theoretical issues

Though negation is a major theme in both typology and formal grammar, prohibitive negation has not figured very prominently. For typology, one should mention Kahrel and van den Berg (1994), Kahrel (1996), Hovdhaugen and Mosel (1999), and Xrakovskij (2001). Both Kahrel and van den Berg (1994) and Xrakovskij (2001) contain contributions on languages all over the world, but the sample is limited. Kahrel (1996) is a study based on 40 languages and it offers some worldwide areal generalizations. Hovdhaugen and Mosel (1999) focuses on one specific language family, viz. Oceanic. Generative grammarians of the Romance languages have tried to account for why prohibitives often do not use the same verb forms as imperatives (Zanuttini 1997). Explanations are offered in terms of type and placement of negation and in terms of the morphological dedication of the imperative.