Double negatives: morphology in contact
Prof. Maria Cristina Cuervo
Natalia Mazzaro
Sentential negation with preverbal n-words displays a crosslinguistic contrast in the presence/absence of a negative element, as illustrated with Spanish (1) and French (2) below. Analyses of this contrast account for this variation in terms of some kind of syntactic parametric variation –in strength of features, movement, etc.– (see Bosque, 1994; Laka, 1994; Haegeman, 1995; Zannuttini, 1997; among others).
1) Nadie (*no) abrió la puerta
Nobody NEG opened the door
'Nobody opened the door'.
2) Personne *( n)’a ouvert la porte
Nobody NEG has opened the door
'Nobody opened the door'.
This contrast is not only found across languages but also within dialects of one language (see Franco & Landa 2006 for Basque Spanish, for instance). In Corrientes – Argentine - Spanish (CS), a variety in contact with Guaraní, the negative clitic no can appear with a preverbal n-word, as shown in (3). Interestingly, the Standard Spanish (SS) variant without no is also accepted (4):
3) Nunca no nos pasó nada (D:164)
Never NEG us happened nothing
'Nothing ever happened to us.'
4) Nunca nos controlaron (D:237)
Never us check
‘They never checked on us’
We present new data which show that this variation exists not only within the same variety of Spanish, but within the same speaker as well. We propose a morphological account of the phenomenon which draws a parallelism between variation in the cooccurrence of a preverbal n-word and the negative clitic –to which we refer as negative doubling- and variation in direct object clitic doubling. Previous approaches seem ill suited to account for variation within the same speaker, since they would suggest speakers have two distinct grammars.
In order to determine the prevalence of negative doubling in the speech community, we conducted and analysed 12 sociolinguistic interviews, each an hour and a half long. The 12 native speakers of CS are distributed along the social categories of sex, age, social class and literacy. The analysis yielded a total of 134 tokens of standard preverbal negation and negative doubling. The results show that negative doubling has an overall rate of occurrence of 17% and that it is present in all layers of society, with women leading in the rate of its use (68%). Concerning the linguistic factors tested, analysis of specificity of the n-word indicates that while the standard variant can have either a specific (44%) or non-specific (56%) reference, negative doubling constructions are almost categorically specific (88%). Specificity, however, was difficult to test with tampoco ‘neither’ and ni ‘nor’.
Negative doubling emerges as the lack of complementarity in distribution of agreeing no and a preverbal n-word, just as clitic doubling is lack of complementarity between an argument DP and an agreeing clitic. We propose, in the spirit of Watanabe 2001, that no is the negative clitic which spells out the Neg head as default, that is, when the specifications of no other lexical items are met. In SS, the item ø rather than no is inserted when SpecNeg is filled. Under this view, the variation found in CS is not a difference in structure or feature strength, which would imply subject-internal variation in core syntax. Here the variation is reduced to variation on the specification of the lexical items ø and no, and is localized in the lexicon (Adger & Smith, 2002). Although of similar nature, this approach crucially differs from Franco & Landa’s (2006) in that it does not depend on their claim –proven wrong by data such as in (1)– that preverbal n-words are contrastively focused and receive main stress. This account can also capture the fact that negative doubling not only occurs with argumental n-words but with negative adverb tampoco ‘neither’ and conjunction ni ‘nor’.
Natalia Mazzaro
Sentential negation with preverbal n-words displays a crosslinguistic contrast in the presence/absence of a negative element, as illustrated with Spanish (1) and French (2) below. Analyses of this contrast account for this variation in terms of some kind of syntactic parametric variation –in strength of features, movement, etc.– (see Bosque, 1994; Laka, 1994; Haegeman, 1995; Zannuttini, 1997; among others).
1) Nadie (*no) abrió la puerta
Nobody NEG opened the door
'Nobody opened the door'.
2) Personne *( n)’a ouvert la porte
Nobody NEG has opened the door
'Nobody opened the door'.
This contrast is not only found across languages but also within dialects of one language (see Franco & Landa 2006 for Basque Spanish, for instance). In Corrientes – Argentine - Spanish (CS), a variety in contact with Guaraní, the negative clitic no can appear with a preverbal n-word, as shown in (3). Interestingly, the Standard Spanish (SS) variant without no is also accepted (4):
3) Nunca no nos pasó nada (D:164)
Never NEG us happened nothing
'Nothing ever happened to us.'
4) Nunca nos controlaron (D:237)
Never us check
‘They never checked on us’
We present new data which show that this variation exists not only within the same variety of Spanish, but within the same speaker as well. We propose a morphological account of the phenomenon which draws a parallelism between variation in the cooccurrence of a preverbal n-word and the negative clitic –to which we refer as negative doubling- and variation in direct object clitic doubling. Previous approaches seem ill suited to account for variation within the same speaker, since they would suggest speakers have two distinct grammars.
In order to determine the prevalence of negative doubling in the speech community, we conducted and analysed 12 sociolinguistic interviews, each an hour and a half long. The 12 native speakers of CS are distributed along the social categories of sex, age, social class and literacy. The analysis yielded a total of 134 tokens of standard preverbal negation and negative doubling. The results show that negative doubling has an overall rate of occurrence of 17% and that it is present in all layers of society, with women leading in the rate of its use (68%). Concerning the linguistic factors tested, analysis of specificity of the n-word indicates that while the standard variant can have either a specific (44%) or non-specific (56%) reference, negative doubling constructions are almost categorically specific (88%). Specificity, however, was difficult to test with tampoco ‘neither’ and ni ‘nor’.
Negative doubling emerges as the lack of complementarity in distribution of agreeing no and a preverbal n-word, just as clitic doubling is lack of complementarity between an argument DP and an agreeing clitic. We propose, in the spirit of Watanabe 2001, that no is the negative clitic which spells out the Neg head as default, that is, when the specifications of no other lexical items are met. In SS, the item ø rather than no is inserted when SpecNeg is filled. Under this view, the variation found in CS is not a difference in structure or feature strength, which would imply subject-internal variation in core syntax. Here the variation is reduced to variation on the specification of the lexical items ø and no, and is localized in the lexicon (Adger & Smith, 2002). Although of similar nature, this approach crucially differs from Franco & Landa’s (2006) in that it does not depend on their claim –proven wrong by data such as in (1)– that preverbal n-words are contrastively focused and receive main stress. This account can also capture the fact that negative doubling not only occurs with argumental n-words but with negative adverb tampoco ‘neither’ and conjunction ni ‘nor’.
