<em>Diálogo</em> versus <em>Cartas</em>: A Probe into Juan de Valdés's Linguistic Behaviour

This study is the first investigation of its kind into Juan de Valdés's language usage against his own linguistic prescription and value judgement, and the data examined have proved to be more interesting than was anticipated. Out of the 24 variables studied, there is no clear case in which Valdés's prescription or recommendation falls in line with his own usage. In principle, I have not carried out a comparative examination of usage in the Cartas and the Diálogo. It is a decision made not at the conception of the study, but rather in the process of data collection and analysis, when the possibility of effectively attesting sufficient evidence of variation in his own language use within his letters alone gradually emerged. It is only when variation could not be attested, as in a handful of cases (i.e. variables (4), (11), (13), (21) and (23)), that actual usage in the Diálogo was resorted to. In the process, Valdés's usage in the Diálogo has shown deviation from the prescribed variant for variables (13) and (21). The remaining three produce no variation, but are, all the same, inconclusive. For (4), such forms as sutil, sutileza, etc., not subtil, subtileza, are found in the Diálogo, which share the same phonological structure (and also the variation pattern) with the form dubda, etc. Variable (11) does have 1 count of the form ahora in the Diálogo, but considered against 58 occurrences of agora, the variation has been judged not to be sufficiently safe, on the grounds of the possibility of a printing or editing contribution to the single case of ahora. Then variable (23) is impossible to control, as there is no appropriate communicative context in the Diálogo for the preterite forms of the verb traer to have been used outside their mention for commenting on. Thus, of all the variables examined, twenty-one of them (87.5%) have been ascertained to exhibit evidence of Valdés's deviation from his supposed preferences (19 within the Cartas alone, and a further 2 with the help of the Diálogo). The remaining 3 have been judged to be borderline cases within the confines of the data studied and, therefore, inconclusive. The question then arises as to whether Juan de Valdés, in acting as a linguistic informant to himself (i.e. relying on his native-speaker intuition of upper-class usage) did really make corrections in discussing his language. It transpires that the evidence for that is overwhelming, a fact that should be of utmost importance to knowledge of sixteenth-century Castilian. A crucial component of this consideration is that of how Valdés's comments - his description, recommendation, prescription and proscription - should be construed and utilized by the modern historical sociolinguist working on Renaissance Castilian. They have been enthusiastically acclaimed or tacitly accepted as an authoritative testimony of sixteenth-century Castilian and have been employed as such to confirm the status of linguistic features and their variants at the time (in the light of the standard that was eventually prescribed by the Academy from the eighteenth century). The following statements from some of the best-known scholars in the field may be representative of the perception and reception of the Diálogo: Es un primor de arte y estilo, y será siempre monumento clarísimo de nuestra lengua, y autoridad la más alta entre todas las de los tiempos del Emperador Carlos V (Conde de Viñaza 1893). [Uno de los valores que reúne el Diálogo de la lengua es] como testimonio del estado de la lengua española durante el primer tercio del siglo XVI (Lope Blanch 1984). Juan de Valdés, impulsado por el afán de reglamentar usos, formula muchas normas arbitrarias; pero la mayoría de las que da son exactas, y tiene un sentido muy certero de los usos preferibles en los casos de duda (Lapesa 1981). Thus, we find such sets as cevil/civil, deferir/diferir, vevir/vivir, menguar/ minguar, joventud/juventud, sofrir/sufrir, robí/rubí, mochacho/muchacho. However, it is clear that, in the sixteenth century, educated taste was already beginning to prefer the forms we now regard as standard, since Juan de Valdés, writing in about 1535, recommends vanidad, invernar, aliviar, abundar, cubrir, ruido, etc., over their still frequent competitors vanedad, envernar, aleviar, abondar, cobrir, roido, etc. (Penny 2000). Valdés admits to the widespread use amongst speakers from the lower strata of the speech community of the linguistic variants that he rejected. He claims his opinions were largely about usage in writing, not speech. Viewed against this background, his own frequent use of the same features in writing creates a pattern of speaker behaviour well known in modern sociolinguistics. Silva-Corvalán (2001:3) recently reiterates: 'La introspección y los juicios de aceptabilidad ocupan un lugar muy limitado dentro de este tipo de estudios [es decir, la sociolingüística], pues es sabido que en casos cruciales para el desarrollo de un análisis lingüístico las intuiciones de los hablantes a menudo no coinciden y no son, por tanto, confiables.' The implication of this to what we know about the broader status of the Castilian language at that time is self-evident. The fact that much later the Academy's choices for the standard form of the language coincide with most of his preferred variants (as stated in Lapesa's words quoted above) does nothing to mitigate the potentially misleading nature of Valdés's comments, sociolinguistically speaking. Evidence from this investigation strongly points to an incontrovertible fact: unquestioned belief in Valdés's authority and the concomitant use of his Diálogo as an automatic yardstick does not do justice to the history of the language or to the Diálogo itself. Perhaps a focused study of the Diálogo within sociolinguistic parameters would refine the prevailing perception of the treatise and how it should be employed for a more realistic picture of sixteenth-century Castilian usage