Doubling of negation

Clause-initial negation, which contributes with the negative semantics in the clause, can be divided into three subcategories according to Lindström (2009:168): (i) responsive negation, (ii) interrogative negation, and (iii) additive negation. Out of these three subtypes, the responsive negation, which is prosodically unstressed and primarily used in dialogues, is possible with clause-final negation yielding thus doubling of negation. Clause-final negation does not contribute with independent negative semantics and it is as such dependent on the presence of a negative element in the clause proper.


Introduction
Doubling of negation, where a declarative is initiated and finished with negation as shown in (1) below, consists of two different structures: clause-initial negation and clause-final negation, and these need not be related.
(1) Inte har jag sett honom inte.(Fenno-Swedish) not has I seen him not 'I have certainly not seen him' (Bergroth 1928: 159) Clause-initial negation, which contributes with the negative semantics in the clause, can be divided into three subcategories according to Lindström (2009:168): (i) responsive negation, (ii) interrogative negation, and (iii) additive negation.Out of these three subtypes, the responsive negation, which is prosodically unstressed and primarily used in dialogues, is possible with clause-final negation yielding thus doubling of negation.Clause-final negation does not contribute with independent negative semantics and it is as such dependent on the presence of a negative element in the clause proper.

Nordic Syntax Database (NSD)
Clause-final doubling of negation has been tested in the NSD with one sentence for both Swedish (2a) and Norwegian (2b).
( The realization of clause-final negation as the negative marker is in contemporary North Germanic restricted to Swedish, and Huldén (1995) suggests that it enforces the negative reading.Important to note is that its presence is not obligatory in the sense that it yields ungrammaticality, if absent.According to

Other data sources
The examples of clause final negation doubling are easily found on the Internet, as a number of examples in ( 12) show.This is not surprising, as the construction is known from both literary and colloquial Swedish, cf.section 1 above. (

Age variation
There is a certain differences in acceptance of clause-final negation doubling between younger and older informants in the NSD.Older informants are more likely to accept the construction compared to the younger informants.This is shown in Map 2 and Map 3.

Historical development
Clause-initial and clause-final negation was previously more widespread across the Scandinavian varieties than the contemporary picture shows.They seem to occur independently of each other.Clauseinitial negation is well known from Old Norse, and an example of this is given in (13).However, Old Norse is not known to have clause-final negation of the modern North-Germanic type.According to Lindström (2009) clause-initial negation was more common in Swedish spoken in Sweden around 1900 than today.It could also be the case of Norwegian dialects, cf.Aasen (1864), and see the . The Norwegian variant (2b) is always rejected by all Norwegian informants with the exception of two: a younger woman in Trondheim, who judges #863 as possible, and an older woman in Oppdal, who judges it as marginally possible.Map 1: Acceptance of clause-final doubling of negation.(#863: Inte har jag smakat på maten inte.'I haven't tasted the food.')(White = high score, grey = medium score, black = low score)2.2Nordic Dialect Corpus (NDC)There are no spontaneous examples of clause-final doubling of negation in the NDC; it may however depend on the small size of the Swedish part of the corpus, as such examples are easily found on

Map 2 :
Acceptance of clause-final doubling of negation among older informants.(#863: Inte har jag smakat på maten inte.'I haven't tasted the food.')(White = high score, grey = medium score, black = low score) Acceptance of clause-final doubling of negation among younger informants.(#863: Inte har jag smakat på maten inte.'I haven't tasted the food.')(White = high score, grey = medium score, black = low score) On the other hand, one finds places like Våxtorp in the Western Swedish province of Halland, where the construction is accepted by younger, but not by older informants, see Map 3 above.
't want to kill your childern.'(Völsunga saga) (Munther 2007ut and behave badly because of that.'(Faarlundetal.1997:874)Clause-finalnegation can be realized with different negative elements such as the negative marker, the negative interjection nej/nei 'no' and local varieties such as the element e in the Swedish dialect of Delsbo(Munther 2007).The negative interjection is possible and common in certain varieties of Norwegian, and it is also used in Swedish, see (8).It is occasionally used in Danish (Østbø to appear).