Cambios en la colocación de los pronombres átonos en la historia del espa~nol

This study analyses the position of the unstressed personal pronoun (UPP) with regard to the verb in Old Spanish. The investigation, which includes quantitative as well as qualitative data, is of a diachronic nature and is based on a corpus of examples dating from the thirteenth to the seventeenth century. In chapter 2 we discuss the most relevant literature relating to the position of the UPP in Old Spanish. In doing so we note that, whereas almost all studies present a detailed description of the possible positions of the UPP in different syntactic contexts, generally, they are not able to relate the position of the UPP to any general syntactic or functional principle of pronoun placement. Moreover, studies that describe the position of the UPP diachronically do not link the different rules for pronoun placement they find in different periods, neither do they seek for an explanation of the changes these rules have undergone. In chapter 3 we analyse the position of the UPP on the basis of our own corpus, after studying its nature in Old Spanish. Based on a quantitative analysis we find that to a certain degree the UPP is syntactically independent and shares certain characteristics with an nominal phrase, since, just as a nominal phrase, it can appear before as well as after the verb and in preverbal position it can be separated from the verb by another constituent (interpolation). However, phonologically we are dealing with a weak, unstressed, enclitic pronoun that always needs another preceding stressed word to lean on. Especially the thirteenth-century texts show this last characteristic with their numerous cases of apocope. Chapter 4 deals with the position of the UPP with regard to the finite verb in the thirteenth and fourteenth century. Although our first analysis, based on three texts which cover the two centuries just mentioned, seems to demonstrate that prepositioning of the UPP increases steadily, on closer investigation this turns out not to be the case. Dividing the data into main en subordinate clauses we find that in both categories the percentage of preposed UPPs stays more or less the same in the thirteenth and fourteenth century: whereas preposition is little frequent in main clauses, the UPP is almost exclusively preposed to the verb in subordinate clauses. We conclude that the general increase of preposition is caused by the fact that in our texts the number of subordinate clauses has grown in the course of time. Further analysis shows that variation in pronoun placement occurs especially in a certain group of main clauses. In these clauses we subjected the position of the UPP to a pragmatic analysis. Our starting-point was the idea that the preverbal position of the UPP is numerically marked and, thus, prepositioning may increase the attention of the hearer. It is then argued that this increased attention is used by the speaker to underscore the relative importance of the referent of the UPP. The relative importance of a referent is reflected by a number of hierarchies that are possibly relevant for the study of the position of the UPP, namely, the person hierarchy, the animacy hierarchy, the gender hierarchy and the agentivity hierarchy. An analysis of our examples based on these hierarchies shows that the importance of the referent of the UPP in Old Spanish conforms to the person hierarchy, i.e. UPPs which refer to a first or second person are almost exclusively preposed, whereas UPPs which refer to a third person most frequently appear behind de verb. Nevertheless, UPPs referring to a third person are also preposed to the verb, which implies that in these cases the speaker also wishes to draw the hearer’s special attention. This special attention turns out to be necessary if the referent of the UPP is less predictable, either because there is a large distance between the UPP and its nominal phrase referent or because there is some other nominal phrase present between the UPP and its real nominal phrase referent which could be interpreted as the referent of the UPP. We conclude, therefore, that pronoun placement in Old Spanish serves a clear communicative goal and is based on pragmatic considerations. The speaker attracts the hearer’s special attention to the interpretation of the UPP, either because its referent is more important or because it is less predictable. In chapter 4 we also treat the absolute prepositioning or postpositioning in the remaining main clauses and in subordinate clauses. In our discussion the clause-initial position P1 (cf. Dik 1989: 348f, 359f) plays a central part. It is argued that in clauses with absolute preposition, P1 is filled with a P1-constituent, which is a constituent that can only appear in clause-initial position. In these cases, P1 cannot be used for pragmatic purposes, i.e. to prepose a constituent with topic or focus function, which is possible in other clauses. The lack of pragmatic possibilities for P1 coincides with the impossibility to prepose the UPP for pragmatic reasons. We argue that the absolute prepositioning of the UPP in clauses with a P1-constituent in P1 is accounted for by the principle of increasing complexity, a constituent ordering principle which states that relatively less complex constituents tend to appear earlier in the linear sequence than other more complex ones with a similar function. In clauses with only postposition of the UPP, P1 is occupied by the verb for pragmatic reasons. Although theoretically the UPP could be preposed to attract the hearer’s special attention, prepositioning is blocked phonologically, since in these cases the UPP, a weak, enclitic element, would be the first constituent of the clause and, therefore, would lack any preceding stressed word to lean on. This explains why the UPP always is postponed when the verb is clause-initial. In chapter 5 we describe the period that starts with the fifteenth century. In this century the number of cases with preposition of the UPP grows explosively. We find that the increase of preposed UPPs occurs in the group of main clauses that in the preceding period presented pronoun placement based on pragmatic considerations. A further analysis shows that the increase can be attributed to an overuse of the pragmatic principle of pronoun placement, especially to the factor of distance between the UPP and its referent; the UPP is increasingly preposed if this distance is relatively small. An increase of the number of cases with preposition involves a situation in which the position before the verb numerically becomes increasingly less marked and therefore pragmatically less effective to draw the hearer’s special attention. In that situation, it is claimed, the real motive for prepositioning can become blurred and this, in turn, can lead to a reanalysis in which prepositioning is related to the presence of one or more words before the verb, which, incidentally, always is the case with prepositioning. This reanalysis can originate a new rule for pronoun placement according to which the UPP is placed before the verb if at least one word precedes. Whether P1 is occupied by a P1-constituent or a pragmatically important constituent is now no longer relevant for the positioning of the UPP. In the fifteenth century we find the UPP appearing invariably before the verb if more than one word precedes; postpositioning is usual if only one word precedes or if the verb stands in clause-initial position. This development, so it is claimed, will be favoured by the fact that prepositioning leads to a phonological pattern that fits the general Spanish stress-pattern, which stresses the last or penultimate syllable of a word. Next, we describe the last step in the development, i.e. the extension of prepositioning to all contexts with a finite verb. This step is a logical one in view of the fact that by the end of the sixteenth century we encounter a situation in which the majority of the finite verbs shows prepositioning of the UPP and only a very small part -if the verb stands in clause-initial position- maintains postpositioning, whereas with non-finite verbs postpositioning is usual. Moreover, in this period variation of pronoun placement is no longer due to pragmatic reasons but only a residue from an earlier linguistic phase. Nevertheless, the last context of postpositioning can only be conquered when the UPP ceases to be phonologically dependent on the preceding word and the verb becomes the decisive factor that motives the positioning of the UPP. These circumstances occur from the fifteenth century onwards, which the disappearance of cases of interpolation, as well as cases of apocope, seem to prove. To conclude the analysis with regard to finite verbes, we argue that cases of the imperative do not follow the general tendency of prepositioning, because with these verb forms postpositioning is communicatively functional. Since these forms are often identical with forms of the indicative or subjunctive mood, the position of the UPP is the only indication which form is meant and therefore, which speech-act the speaker intends to perform. Chapter 6 treats the development of the position of the UPP with non-finite verbs, i.e. the infinitive and the gerundio. Apparently, these forms seem to develop steadily towards prepositioning of the UPP, which abruptly stops by the end of the sixteenth century. However, a further analysis shows that the percentage of preposed UPPs increases only in cases in which the infinitive is introduced by a preposition and that in this construction from the fourteenth century on prepositioning is almost absolute; with the remaining nonfinite verbs prepositioning has always been very rare. In order to explain the cases of preposed UPPs with the infinitive introduced by a preposition, we again turn to the notion of P1. We claim that the preposition is a P1- constituent that fills P1 of the infinitive-clause. As in clauses with a finite verb and a P1- constituent in P1, it is argued, prepositioning of the UPP in infinitive-clauses introduced by a preposition obeys the principle of increasing complexity. The appearance of the first cases of preposition in the fourteenth century is attributed to the fact that, whereas in the preceding century the only frequently occurring preposition was por, by then the infinitive-clause is characterized by a large increase in the use of different prepositions. We claim that the use of different prepositions will strengthen the notion that we are dealing with a construction in which P1 is occupied by a P1-constituent and, subsequently, the principle of increasing complexity can become operative, which favours prepositioning. Our analysis demonstrates that the first cases of prepositioning with an infinitive-clause introduced by a preposition occur especially with the preposition de. This, in turn, is explained on the basis of the syntactic similarity between de and que, the latter being an outstanding example of a P1-constituent. Because of this similarity the principle of increasing complexity could have become operative earlier in cases with de than with other prepositions. The development of the positioning of the UPP with non-finite verbs comes to a standstill by the end of the sixteenth century, when the examples show a return to postposition. In view of the fact that the verb has come to play a central part in pronoun placement -the UPP has become syntactically as well as phonologically dependent on the verb- and in view of the situation of pronoun placement existing by the end of the sixteenth century -mainly prepositioning with finite verbs, mainly postpositioning with non-finite verbs- it is claimed to be hardly surprising that the postponed UPP increasingly drives back its preposed counterpart. This leads to the modern rule for pronoun placement, in which the presence of a P1-constituent in P1 has ceased to be relevant for the placement of the UPP; instead, the verb is the basis for pronoun placement. In chapter 7 we present our most important findings. While describing the development of pronoun placement as a case of grammaticalization, we discuss the question why at a certain moment in time the pragmatic principle is being used more frequently, which leads to the far-reaching consequences mentioned above. It is argued that the increase of the preposed UPP cases with infinitive-clauses introduced by a preposition could have given the first impulse to the increase of prepositioning with finite verbs that show variation in placement. Because of the former increase prepositioning is becoming more frequent overall and thus less marked. In spite of this, it should be borne in mind that this change could only take place because the pragmatic principle of pronoun placement showed some weak spots already. Although communicatively it was a useful device to the speaker, it could be used only in certain contexts, i.e. in part of the main clauses. The fifteenth-century placement rule, in which prepositioning of the UPP is related to the presence of some elements that precede the verb, is based on a principle directly observable in the context and is valid for all finite verbs. Nevertheless, it does not apply to non-finite verbs. The modern placement rule meets the requirement of direct observability as well as the requirement of general applicability, which unquestionably caused its final and definite success. However, the UPP’s change in nature, from a syntactically independent pronoun that phonologically depended on the preceding word to a pronoun that depended syntactically and phonologically on the verb, must have been an important condition for the introduction of the modern placement rule.