On the supposed depronominalization of relative

It has been argued that the adjectival clauses that contain an anaphoric pronoun referred to the antecedent of the relative are anomalous, since this relative loses its pronominal nature in order to become a mere conjunctive nexus. it is thus assumed that, in an expression such as «hay gente que le gusta vivir», que no longer substitutes for a noun which represents the content of gente, for this role is now played by the anaphoric le. The author of this essay sets out to demostrate (a) that the so-called relative pronouns are anaphoric determiners devoid of nominal content; (b) that they never substitute for an antecedent, whether nominal or not; (c) that the article or the prepositions that can go before it do not take this position depending on an imaginary nominal condition, which these relatives lack, but depending on the syntactic relation that their antecedent may establish with the verb in the subordinate clause; (d) that the concept of 'anomaly' arises from a mistaken grammatical analysis of the relative clause, as it is evidenced by the fact that these constructions are attested in all the stages of the Spanish language and in all speech registers; (e) that, nevertheless, this phenomenon varies considerably with usage, and this fact allows us to establish several degrees of acceptability, according to the different types of grammatical structure that contribute to the classification of the examples. The author ends by sketching a set of rules concerning both the grammaticality and the acceptability of every specific type into which these constructions can be classified