On the morphology of *no* in Spanish

Marianna Pool Westgaard

It would appear that *no*, the pre-verbal element of negation in Spanish, behaves more like a clitic than a word. To show that this assertion is correct, I shall appeal to the traditional notion of what a clitic is and to the pertinent arguments in Arnold Zwicky's "Clitics and particles" (*Language*, 61 (1985), pp. 283-305). The tests he proposes thereare directed towards showing that clitics are morphologically more closely akin to inflectional affixes than to indepedent words. The tests proposed are phonological, morphological and syntactic in nature--having to do with contrastive stress, morphological closure, ordering and distribution of morphemes, morphological complexity, gapping, substitution, and movement of elements in the sentence. I shall use each of these tests to show that Spanish *no* is morphologically (at the very least) a clitic.

First and foremost--and the traditional way of identifying clitics--is the fact that no independent word or phrase may come between the pre-verbal *no* and the part of the verb to which it adheres (the leftmost element in the V, including modals, auxiliaries and pronominal clitics).

1. No *{tu'/probablemente/a veces/por lo menos} le-has-mandado una carta

Some of Zwicky's tests may be applied as follows.

Phonologically, only independent words may show independent stress in non-contrastive contexts. In Spanish, clitics, as well as modals and auxiliaries in these environments, must be unstressed. The same is true of *no*:

2a. *Juan PUEde cantar (Modal)

2b. *Juan HA cantado (Auxiliary)

2c. *Juan LE canta (Pronominal clitic)

2d. Juan CANta (Word)

2e. *Juan NO canta (Negation)

Morphologically, clitics/inflectional affixes are essentially fixed with respect to their distribution and ordering. This is certainly the case with *no*: a quick descriptive rule (the one in (3)a, perhaps) suffices to state the distribution. This is in contrast to independent words, whose distribution is generally subject to a large number of syntactic rules. The ordering of clitics and inflectional affixes within a word is not subject to any stylistic whim, either, but is quite rigid. *No* conforms to this pattern in (3)b-e:

3a. NO--(PRONOMINAL CLITIC(S))--(AUXILIARY(/IES)/MODAL(S))--VERB

3b. no-me-lo-ha-dicho nadie (vs. *me-no-lo-ha-dicho, *me-lo-no-ha- dicho, etc.)

3c. no-se-ve el cielo (vs. *se-no-ve)

3d. Mari'a no-esta'-trabajando (vs. *esta'-no-trabajando)

3e. Seve no-cuenta mentiras

Syntactically, the tests of gapping and movement show that in no case does *no* function independently:

4a. Mari'a lo trajo y Juan (e) tambie'n

4b. Mari'a no lo trajo y Juan (e) tampoco

4c. *Mari'a no lo trajo y Juan no lo (e) tampoco

4d. *Mari'a no lo trajo y Juan (e) lo trajo tampoco

5a. Mari'a le-dio el libro a Pepe

5b. ?Le-dio Mari'a (e) el libro a Pepe?

5c. *?Dio Mari'a le (e) el libro a Pepe?

5d. Mari'a no-vio a Pepe

5e. ?No-vio Mari'a (e) a Pepe?

5f. *?Vio Mari'a no (e) a Pepe?

Now, if Spanish pre-verbal *no* is morphologically a clitic, its relationship to the other negative element in constructions such as *Miranda no recibio' ninguna carta* remains to be determined. We know that Spanish is a language that allows for a wide range of doubling phenomena between a clitic and its referent: among datives this possibility is almost unlimited; in certain dialects and under certain conditions, it is also permissible in accusatives. If our *no* is truly a clitic, it is not unreasonable to postulate that the post-verbal negative doubles that clitic in the same way an NP may double a dative. A common analysis of this type of construction has been one of Spec-Head agreement. I propose that this is the case for double negation in Spanish as well.

Go back to the LSoRB Program

---