Acceso usuarios
Carlos Cid Abasolo 
The relative clause in Basque is a topic that has aroused great interest since the publication of De Rijk's doctoral thesis, one of the first studies to point out the complexity of this subordinate clause which is one of the mainstays of Basque syntax. This interest has been sustained on the one hand by the typological wealth of the subject, and on the other by the difficulty of determining its grammatical limits.
As regards the first point, it is worthy of note that all research to date has chosen to focus on the study of only one type of relative clause: that with a finite verb and the invariable relativiser -n. In this essay, an attempt has been made to examine problems raised by other types of relative clause, such as those with non-finite verbs.
Nevertheless, the second point is the area of study of which all efforts ought to be concentrated. How to determine which cases or grammatical functions can or cannot be relativised, and how to establish restrictions on the movement of relativisation, are essential questions for which no. more than very general hypotheses and principles have been given. It must be known how far one can relativise; or, to put it another way, what the limits of this syntactic process are. The Basque relative of Romance origin, constructed from pronouns and interrogative adverbs, presents no problem at all: any case can be relativised, since the relativised case is marked in the utterance, being attached, specifically, to the aforementioned pronoun or adverb.
However, in all the other relative clauses there is no visible trace of the case or grammatical function; nothing is specified but the function of the relative head with respect. to the main clause. The key, then, lies in the linguistic and extra-linguistic indications that permit identification of the erased/relativised case, the most important of these being morphological in nature. Basque, then, presents the same problems as Japanese, which also has an invariable relativiser in which the relativised grammatical function cannot be marked. The speaker has to fall back on various strategies in order to unearth a hidden linguistic relation; were it not disclosed, comprehension would be prevented. This state of affairs is even more pronounced in the Basque relative clauses with a non-finite verb, where the absence in the verb of any morphological marking of the so-called �grammatical cases� (absolutive, ergative and dative) forces the speaker to have recourse to more diffuse indications, like pragmatics in order to attribute to the relative head its exact grammatical function in the relative clause.
