30 Alberto Medina, Jose del Volle, Henrique Monteagudo
Academy. Villa's study argues that control over this institution constituted a necessary tool in the hands of a government that understood language and grammar as disciplinary practices in the constitution of national citizenship (on the relation between grammar, the state and modern subjectivity see also Arnoux in this volume). If every Spaniard was to be educated as an effective citizen contributing to the coherence, strength and prosperity of the nation, the decisions surrounding the codification of language used in the classrooms could not be left in the hands of institutions located outside the orbit of government control. Following the trend slowly built through the eighteenth century and consecrated by the French Revolution, belonging to the nation necessarily implied the use of a common language, tool and symbol of its unquestionable unity. In 1857, the education law (known as Ley Moyano) was a transparent formulation of those ideas: a centralized and homogeneous education with the help of institutions such as the RAE was both a mirror and the condition of possibility for Spain as a modern nation.7
In the final article of this part, Monteagudo analyzes the moment of crisis of that paradigm. In the context of the Second Republic, the modern success of that idea of nation had reached everywhere and, as a result, competing national projects enter a symbolic and political conflict. If the unity of language had been a priority for the modern Spanish nation and its self-constitution, other nations within that political unity called Spain to start thinking along similar lines. In that new reality, different intellectuals directly involved in parliamentary discussions around the 1931 constitution are forced to confront the choice between identification of Spanish language and nation on one hand and, on the other, a more complex dissolution of that privileged relationship in favor of a plurilingualism that may avoid political conflict. It was an attempt to make compatible the stability of the political unity and the ascendant demands of competing self-conscious national identities that made the recognition of other official languages along with Spanish possible, if only for a very short period of time.
But that moment, the political making of other languages within Spain, is perhaps a window to see the stories told in this part just as a threshold to a more complete exercise of storytelling that can only be conceived as polyphonic. The story of a dominant voice is, to a certain extent, a fiction that this volume wants to reveal as such and, in that sense, necessarily intertwined with many other silenced stories to which these studies are only a prelude.
7 Metalinguistic discourses in the context of the development of educational systems - and the connection of both to nation-building and colonialism - are addressed in several chapters throughout the volume: Villa, Arnoux, Barrios, Dubord, Fernandez-Gibert, Castillo Rodriguez.
The prehistory of written Spanish and the thirteenth-century nationalist Zeitgeist
Roger Wright
Introduction
For many centuries, Romance morphology, syntax, vocabulary and semantics developed gradually, and could be represented in writing according to the old Latin writing systems. The idea of the deliberate invention of new Ibero-Romance ways of writing, intentionally distinct from the inherited Latin norm, was catalyzed by the import from France of the phonographic mode of reading Latin aloud in Church with a sound for each written letter. This requirement led first to the existence of two ways of saying the same word, and then, after the invention of Romance writing, to two ways of writing it. The word "Romance" was first applied to the new ways of writing, rather than to speech. Written and spoken "Medieval" Latin (known usually as grammatica) was in theory standardized over Europe. Written Ibero-Romance, used in many contexts by the mid-thirteenth century, might have developed as a single unit, but increasingly during the thirteenth century the separate political units of the Iberian Peninsula wanted to elaborate their own national written form; that is, the conceptual split between Ibero-Romance languages came after, but only shortly after, the conceptual split of "Medieval Latin" from Romance.
The development of written Spanish was not a single event, even though the appearance of complete texts in a deliberately new scripta, around 1200, was relatively sudden. Their arrival had a long prehistory, as Romance features had been present in texts for centuries, and no political overtones. But soon after the appearance of complete Romance texts, the mode was adopted for non-linguistic purposes only indirectly connected with its original point.
Romance languages developed from spoken Latin over many centuries. The choice of language names is notoriously tricky for the philologist studying these centuries, as within the Iberian Peninsula we are faced with the possibility of using "Vulgar Latin," "Late Latin," "Proto-Romance," "Early Romance," "Ibero-Romance" and even "el espanol primitivo"; the speakers themselves created no new language names, and used the phrase lingua latina to refer to their language throughout the whole period before the deliberate invention of new ways of writing. For .it does seem that the word "Romance" (variously
31
32 Roger Wright
The prehistory of Spanish and the 13th-c. Zeitgeist
33
spelled) was first applied to the novel techniques of writing rather than to any register of speech. The modern analyst, on the other hand, naturally wishes to distinguish between the spoken language states of the fourth, eighth and twelfth centuries AD, which is why we feel justified in using a procedure which on the whole historical linguists would do well to be wary of; that is, we tendjo apply to their language some names which its speakers did not Use themselves. In particular, it has seemed reasonable to many to call the spoken language of the sixth to the eleventh centuries "Romance," or "Early Romance," which in this perspective comes later than "Latin" but earlier than the geographically restricted "Ibero-Romance" (or "Old Spanish").
In the first part of the present study we shall glimpse the way in which, within the Peninsula, Early Romance appeared in written form. I am not here referring only to the linguistic mistakes which were unintentionally made in the Latin of inscriptions and manuscripts. There were also occasions when writers in the Iberian Peninsula were deliberately and consciously trying something different, when their non-standard written form was not just a straightforward, spontaneous and natural reflection of their speech, but an intellectual attempt to achieve something new. The presence of both intentional and unintentional phenomena of these types in the Peninsula are collectively summarized here as being the "prehistory of written Spanish," using the language label "Spanish" in the way that Ramon Menendez Pidal (e.g. 1972) and Ines Fernandez Ordonez (e.g. 2010, 2011) have done, to mean what other scholars mean by "Ibero-Romance"; that is, it is not to be identified solely with Castilian.
Menendez Pidal paid a great deal of attention to this prehistory in his Origenes del espahol (see Del Valle's brief discussion in Chapter 1), which was devoted to the ninth, tenth and eleventh centuries. He pointed out, for example, quite rightly, that we can learn from the way names of non-Latin places and people were written, since the scribes often had no canonical inherited form available and were left to their own devices. In general, those written" features in the texts of this age which we can interpret as being inspired by Romance phonetics can indeed help us assess what those phonetics were; where it seems that Menendez Pidal was probably overconfident was in the converse case, where he tended to suggest that the presence of the traditional written form of a word implied that the scribe used old-fashioned phonetics rather than merely that he had learned how to spell the word correctly. Several of the graphical choices made by scribes in the Peninsula during what has come to be known as the epoca de origenes were not destined to become standard in the subsequent writing systems of Old Spanish, such as the spelling of the palatal affricate sound with a double letter gg in such names as Sanchez; that is, the sound now represented by the letters ch. Peterson (2009: 254) has pointed out that in the early eleventh century the same person's name, now written as Ochoiz, of Basque etymology, was spelled in documents in twenty-one
different ways, none of them Ochoiz- But, as Antonio Emiliano has pointed out (e.g. 1991/1996, 2003), it is probably the case that most of the individual graphical (in his word, "scripto-linguistic") solutions eventually adopted in the semi-standardized methods of writing Ibero-Romance in the early thirteenth century had in fact been used on occasion already, within documentation that was ostensibly Latinate.
The decision to represent in a new written format whole texts, rather than individual phonetic features or non-Latin words, when it was consciously taken in some areas of the Peninsula in the second half of the twelfth century, did not thus involve starting from scratch. To many modern analysts it has seemed natural to see Latin and Romance as already being separate whole languages before that time; but the features of both had coexisted in the texts of earlier centuries, and to understand the process that led to their eventual separation as two distinct modes it is worth deconstructing the phenomena into their component parts, taking what Angel Lopez Garcia (2000) would call a "modular" approach. In the next sections we will look in order at syntax, semantics, the lexicon, morphology, phonetics and orthography.
Syntax
Several aspects of Latin syntax changed over the years. But not as much as we might think. Within the Roman Empire itself, written Latin of a practical kind, such as that written by architects, doctors, gardeners, cooks and other professionals, has sometimes been described, inappropriately, as "Vulgar Latin" (see Herman 2000). This is widely regarded now as an unhelpful way of describing these works, if only because what is often called "Vulgar Latin," as opposed to "Classical Latin," was in reality just "Latin," spoken by all, whereas "Classical Latin" was a recherche register written by only a few and spoken by perhaps nobody. The grammar of texts prepared in non-literary registers during the Roman Empire is markedly more like that of later Romance than is the grammar of the highest literary register of that period; it is well known now, for example, that the default word order of the late fourth-century Peregrinatio, composed in a colloquial register by the probably Peninsular nun Egeria, is more or less the same as that of the average thirteenth-century Peninsular composition. Similarly, tenth-century Peninsular documents often attest a VSO (or VS) order, particularly in subordinate clauses, more characteristic of later Romance than of most Latin works written during the Roman Empire. We can also watch, without much surprise, as the documentation of this epoca de origenes uses more prepositions (particularly de) and fewer case inflections than would have been the case during the Empire, although the prepositions themselves are not new. Robert Blake (e.g. 1991), who has shown how the syntax of many documents of the age is effectively that of the Romance of the
34 Roger Wright
time rather than that of the Latin of the distant past, refers to these documents as being "Latinate"; that is, Romance in syntax, although the scribe is still aiming to make the document look like Latin through his use of the traditional spelling. These word-order and other syntactic phenomena are probably not intentional, though; the scribes used those developed features of syntax in writing merely because they used them in speech, rather than out of any conscious desire to revolutionize the written form itself.
Semantics
The study of the semantics of the documentation of that age has not been undertaken to any great extent. Even so, there is a sizeable number of words whose meaning changed between the first and the twelfth century. If we find these words being used in ostensibly Latinate texts with their evolved meaning, that is, with their Ibero-Romance meaning rather than their older Latin one, this supports the hypothesis that the scribe was writing his natural Romance language by disguising it under the old-fashioned orthographical mode, which was the only one available to him. This phenomenon was examined in a paper given to the 2003 conference on the origins of the Romance Languages in Leon (Wright 2004), using as the main example the third-person present subjunctive forms of the Latin verb SEDERE, traditionally written in the singular as SEDEAT, which had been changing its meaning in the Peninsula from "to be seated" to "to be" (whose infinitive would later be written as ser). In the eleventh-century Riojan Glosses, we find this subjunctive form written as sie-gat, and used in those glosses with both the semantics and the grammar that it had in contemporary Ibero-Romance (that is, as a passive auxiliary); in this way, it exemplifies the glosser's attempts to write a feature of his natural language in a novel way. This same word appears in many of the documents of the period which are kept now in the archives of Leon Cathedral; up to the thirteenth century it was always written in the traditionally correct form sedeat, but with the meaning of "be," including in several uses as a passive auxiliary, and without the original meaning of "is seated." Thus both the syntax and the semantics of this word in these ostensibly Latinate documents in Leon are those of Ibero-Romance, as they had been in the Riojan Glosses. When we move ahead a few years to investigate the early thirteenth century, and look at what is probably the first written Romance document produced in either the Leonese or the Castilian chancery, the Treaty of Cabreros of 1206 (both versions are edited in Wright 2000), we find that same word (already pronounced [se-a]) written several times with the letters sea, a three-letter combination which had not been attested in writing in the chanceries before (so far as we can tell). This item is the same entity as the sedeat written in the previous documentation in Leon (and the siegat of the Glosses), from a semantic and syntactic point of
The prehistory of Spanish and the 13th-c. Zeitgeist 35
view, and probably also from a phonetic point of view in speech, but written in a new way. Many other words which had undergone semantic changes had their new meaning attested in old graphical form in the epoca de origenes as well; thus the eventual change to Romance writing in these cases involved a change to a new spelling only, because the semantic and syntactic entity had been present in writing for a long time already.
This conclusion is unsurprising and indeed only natural; when semantic and syntactic changes occur in a language there is no need to change the language's writing system at all if we wish to represent them on paper, because new word orders and new grammatical and semantic uses for existing words can be represented in writing just as well as the traditional orders and uses can. That is why I called my edition and study of the Treaty of Cabreros a "sociophilological study of an orthographic reform" (i.e. of the existing language) rather than "of a new language."
The lexicon
Lexical change involves new words, and thus is far more salient to the language user than semantic change is. There can be consequences for the written form, particularly, as already noted, in names of places and people. Words borrowed into Ibero-Romance from Germanic, and even more so from Arabic, often had no obvious written form in the Roman alphabet. Borrowings from Greek had previously led to the adoption in writing of the Greek kappa, and the Greek upsilon and zeta had inspired counterparts in the letters y and z which were added to the end of the Roman alphabet; but that had already happened long before the Ibero-Romance period, and these letters were no longer seen as novelties. Thus the letter z appears in many documents of the epoca de origenes, particularly in the names of witnesses (representing the sound [dz]), and the k especially in the dating formula, where kalendas tended to be abbreviated with a k. In the Iberian Peninsula there seem to have been no attempts to transfer Arabic letters into the Roman alphabet to represent unfamiliar sounds heard in the etyma of Arabisms; instead the scribes tried to find suitable Roman alphabet equivalents for the unfamiliar sounds, without notable success in a number of cases, as attested by the multiple spellings of some of the more phonetically complicated lexical items involved. This phenomenon helps our reconstruction of the phonetics, as for example when Arabic words with an initial aspirate [h-] turn up in Latinate or Romance texts with an initial letter/- (e.g.fasta for [has-ta], modern hasta, [as-ta]) and thus help us argue that/- in (e.g.) the Poema de Mio Cid often represented [h-].
The presence of originally non-Latinate lexical items in writing is another sign that the scribes were merely aiming to represent their own language on parchment, rather than trying to recreate the Latin of the distant past in
36 Roger Wright
The prehistory of Spanish and the 13th-c. Zeitgeist
37
which such entities were not yet present. And it happens more often than we might expect that there are apparently non-Latinate lexical items in tenth-century documentation which we cannot now easily recognize or interpret. Two examples: the phrase sagia vizione in a legal judgment from late tenth-century Vairao in the diocese of Porto, which might be used to refer to an ecclesiastical sash with ribbons on (Wright 2010c); and the strange word raisce used in several of the documents examined in David Peterson's thoughtful study (2005) of those documents in the San Millan cartulary which were brought there from the Valle del Alto Tiron in the eleventh century. This is a word which seems from the context to mean a meal held in celebration of a successful transaction, but is etymologically unexplained. There are several others, and in general the scribes saw no reason not to use in writing terms which were current in their own community, whether or not they would count technically as Latin to an admirer of Lewis and Short.
As Steven Dworkin pointed out (1995), the eventual change to the new Romance writing system involved losing from the written registers a number of old-fashioned words which probably played little role in speech; but the converse is also true, that most of the Romance vocabulary of that time, whatever its etymological origin, had also been attested earlier in ostensibly Latinate documentation. This fact has consequences for the elaboration of etymological dictionaries; there is little validity in Corominas's normal procedure (Coromi-nas and Pascual 1980-91) of regarding the earliest attestation of an Old Spanish lexical item as necessarily being its earliest attestation in reformed spelling, since most of the words in question turn up in unreformed spelling before the date given in that dictionary. But they are the same item from an etymological point of view, however they happen to be spelled (see Wright 2010b).
Morphology
Morphology is interesting in the present context, and worth separating out into nominal and verbal morphology. Nominal morphology had slowly but permanently simplified over the years, losing all the original case-endings other than the accusatives from Latin nouns and adjectives, preserving [-s] (and -s) in a new function as a plural marker. Most texts prepared in a non-literary register, during these centuries previous to the development of written Romance, do not attest the ablative cases of nouns, because speech had lost the ablative; and they rarely kept the genitives, datives and (in the Iberian Peninsula) nominatives either. Speech eventually preserved only the originally accusative forms of nouns and adjectives, but the use and the understanding of the nominatives and datives survived in the spoken pronouns, and still do (e.g. ILLE > el, ILLIS > les); and genitives continued to be written throughout the period in a number of proper nouns such as the names of churches (e.g. ecclesia
Sanctae Mariae), so it would be an exaggeration to say that the other cases had been dropped from speech. Verbs are slightly different, in that Romance verb morphology, although much evolved, was no simpler than that of Latin. For example, we have already seen the attested spread of the passive auxiliary (< SEDERE), and during these centuries in writing the tense of these passive compounds was often determined by the tense of the auxiliary, as in Romance, rather than that of the participle, as in Latin. Overall, it is reasonable to see the verbal morphology of most non-literary texts of the age as mostly representing that of the speech of the scribe and/or author.
Thus the main difference, between "Latinate" texts of the época de orígenes (that is, written before the advent of the Gregorian and Carolingian reforms) and the subsequent texts in the new Romance writing, lies only in the spellings attempted by the scribe; Romance grammar, vocabulary and semantics were already representable and represented on parchment. The reform of spelling in the Peninsula was preceded and precipitated by a phonetic reform in the nature of the official Latin used in church, and subsequently in the first universities; the official change of rite in Castile and León after 1080 imported the French liturgy and backgrounded the native liturgy which had been elaborated in the seventh century by Isidore of Seville and his colleagues, and brought with it the requirement that in church all texts should be read aloud using the method of giving each written letter of the text a sound, and in theory the same sound each time. Thus most words sounded different in the liturgy from the way they sounded in other circumstances; e.g. episcopus would be pronounced there as [episkopus] rather than [obíspo], zná fecit as [fékit] rather than [fidzo]. These developments were an integral part of the Gregorian reforms of the late eleventh century. French clerics were needed to train the native priests in the new techniques (Wright 1982, 2003; Emiliano 2003 traces the effect of these developments in Portugal). The theoretically biunique connections between each individual written symbol and its related spoken sound, as established at that time in what we now call Medieval Latin, underlay the eventual development of the same correspondences between vernacular sound and letter in the new Romance writing modes. But since contemporary Romance morphology, syntax, vocabulary and semantics could already be, and regularly were, represented up till then in the old writing methods still instilled in scribes during their training, the novelty of such officially inspired texts as the Romance Treaty of Cabreros of 1206, prepared in the Castilian royal chancery with a copy also made in the Leonese chancery, lay almost exclusively in the spellings.
Written Romance, which is what the word "Romance" was initially reserved for at the time, was thus not the sudden eruption of a complete novelty. All aspects of the spoken language other than the phonetic had for a long time already been directly represented on parchment; in exactly the same way, modern English and Spanish grammar, morphology, semantics and vocabulary
38 Roger Wright
are written down by us all on paper every day in the traditional spellings which we learned at school, far from being a phonetic script though these are. The newly developed spellings were a new way of writing the same language, rather than a totally new language: That idea, of a new way of writing the same language, had been gestating for a century already. The famous Riojan Glosses of the 1070s have sometimes been described as the "birth certificate" of the r Spanish language. This idea is not taken literally any more, but the Glosses certainly attest the birth of something, for they show us the start of a new idea within the scribal mentality: the idea that their words could be deliberately written in a non-traditional way. It seems reasonable to suggest that this idea is a by-product of the new atmosphere which accompanied the arrival of the reforms in the 1070s and 1080s.
The reforms themselves, though, took over a century to become generally operative. The twelfth century in the Peninsula presents a kind of patchwork in this respect, as the reforms were pursued more enthusiastically in some groups (such as the new religious orders) than elsewhere (see Fernandez Ordonez 2011; Hernandez 2009; Wright 2003: chapter 17). In some places where the reformed technique of reading Medieval Latin aloud was introduced, an incentive to represent vernacular spoken registers on parchment in a novel way, and thence to create a conceptual distinction between Latin and Romance based on the two modes of writing, was likely to follow, though not necessarily at once. But these new written Spanish modes had a prehistory, which Hispanists are well advised to bear continually in mind; in particular, those who study Ibero-Romance grammatical features of the thirteenth century should be aware that they are almost certainly attested in the preceding centuries also, in texts whose graphical form might make them look ostensibly as if they are not Romance at all.
The reasons for the emerging preference for writing in a new Romance style in the early thirteenth century have been much discussed by modern analysts, with no clear consensus emerging. As was probably the case earlier in Carolingian France, there is probably a connection with ease of reading aloud comprehensibly, which had been made a more complex task by the arrival of the new Medieval Latin reading style (which required, and requires, a sound for every written letter). In my view, and that of several others, the monk (or monks) who wrote the glosses of San Millan and Silos was probably writing for his own subsequent benefit (or possibly also the benefit of a foreign visitor) when later reading the texts aloud; but this is not a universally accepted interpretation. What we call Old Occitan (which it would really be more appropriate to call Young Occitan) may well have been a model in the back, or even in the forefront, of the minds of those who tried the initial experiments. Shortly afterwards, the Languedoc certainly seems to have inspired the twelfth-century developments. Hernandez's 2009 study pointed out that where scribes were
The prehistory of Spanish and the 13th-c. Zeitgeist 39
trained is more important for their writing techniques than the place their documents now happen to be kept, which latter assumption underlay most of Menendez Pidal's organizational rationale in his Documentos Lingiiisticos; rearranging the evidence in this way suggests to Hernandez that the members of the new religious orders of the age, particularly the Cistercians and the Premonstratensians who had come from the Languedoc, where texts in written Occitan Romance had been known for some decades already, were particularly | interested in developing an Ibero-Romance written mode. Geographically, the ' ensuing patchwork of old and new seems incoherent; in terms of where writers were trained, the pattern makes sense. The result in Burgos, for example, was that the Cistercian nuns at Las Huelgas welcomed the new modes from at least 1188 and the cathedral did not do so until the mid-thirteenth century (Hernandez 2009: 267). Aguilar de Campo (where several scribes had Jewish names) adopted the novelties soon after it affiliated to the Premonstratensians in 1169, San Millan did not till well into the following century, and so on. The idea had come from over the Pyrenees, and was continually reinforced by the contacts in both directions between the Peninsular and mother houses, but naturally the details needed to be newly invented in the Peninsula, given the phonetic and other differences between the two varieties of Romance.
In Toledo, local government still mainly used Arabic in written texts in the twelfth century, but towards the end of that century, as knowledge of written Arabic was gradually diminishing, the new written Romance seems to have taken up some of the slack, probably encouraged by the Cistercians who had created the military order of nearby Calatrava (Hernandez 2009: 280). None of this seems to have been the result of any political or national perspective; one prime mover may well have been the Archbishop of Toledo, Martin Lopez de Pisuerga, who was granted oversight of the royal chancery by Alfonso VIII of Castile in July 1206, but the arguments between himself and the official chancellor, Diego de Campos (Diego Garcia), who would have preferred to use traditional written Latin consistently, were probably religious and moral in nature rather than political. Diego de Campos was eventually relieved of his duties in 1217, and immediately (1218) produced his extraordinarily reactionary and unsurprisingly little-read Latin work entitled Planeta, in which he expresses . . the idea that the world is falling to pieces, even implying that the written Latin forms of words have an intrinsic mystical value and that Romance writing, by definition, therefore, is heretical.
As Tore Janson's work has established (Janson 2002), the concept of Romance as a separate language from Latin followed the elaboration of the new written mode, rather than inspiring it. Judging by some of the data recently adduced by Fernando Tejedo Herrero (2008), that idea seems to have taken j a remarkably long time to take complete hold. The metalinguistic distinction between different Romance languages came later than that between Latin and
40 Roger Wright
The prehistory of Spanish and the 13th-c. Zeitgeist
41
Romance. For it is clear, from detailed comparative analysis of the two texts of the Treaty of Cabreros of 1206, that although in that context the distinction between written Latin and written Romance must have been obvious to the chancery scribes, and the choice was probably a matter for conscious discussion and argument, there was at that point no real desire or instinct to distinguish metalinguistically between different geographical kinds of written Romance. Their heads had enough to do getting round the promotion of a diastratic register distinction (high style versus low style) to the status of a linguistic one (Medieval Latin versus normal Romance), without doing the same for the diatopic (low style in area x versus low style of area y). That is, there is no sign in the Treaty evidence that the scribes and notaries of the chancery made any clear distinctions in their minds between one geographical kind of Romance and another. This is hardly surprising from a linguistic point of view, since (as Fernandez Ordonez has established and as we would in any event expect) isoglosses did not then bundle neatly along the political frontier between Leon and Castile; metalinguistically, we have no evidence to support the idea that they made a conceptual distinction yet between Castilian and Leonese as whole dialects. The words castellano and leones existed, but not with the metalinguistic meaning. The two surviving versions were written by notaries from the chanceries of Castile and Leon respectively, and show a large number of minor differences that are most easily attributable to their being two different transcriptions taken from the same dictation; but these differences, much more often than not, did not correspond to isoglosses running between Leon and Burgos. Usually they correspond to two possible ways of spelling the same word within either region (e.g. a versus ha).
Linguistic analysis of the Castilian version of the Treaty and of other Romance texts written in Castile during that decade also shows no consistency in the representation of details; that is, there was no common Castile-wide idea of how to write the Romance of their kingdom. Such consistency as there was tended to operate at the level of the local cultural centre, if at all, not at that of the kingdom as a whole. The idea of writing the Treaty in the new way, and the original preparation of the text, must have come from Castile rather than Leon. The so-called Posturas of the Cortes of Toledo of January 1207 (Hernandez 1988) were prepared in Romance also, as were a few other documents from Toledo at the time (1206-8), including one or two brief fueros (not to mention the Poema de Mio Cid). Rodrigo Ximenez de Rada, however, replaced Martin Lopez de Pisuerga after the latter's death in August 1208, and rehabilitated Chancellor Diego de Campos (as can be seen in the details at the end of the chancery documents of these years). Between them these two forceful personalities and excellent Latinists seem to have discouraged such frivolities as the use of written Romance in the Castilian chancery. Indeed, it is tempting to suggest that this may have been one of the motives for the appointment of
Ximenez de Rada in the first place, at precisely the time when the masters in his alma mater, Paris, were expressing hostility to all the novelties coming into France from Toledo, although the fact that he was said to have been elected with the unanimous support of the cathedral chapter may cast doubt on such a hypothesis. In any event, once Fernando III came to the throne in 1217 the writing of Romance texts seems to have slowly come back into political favour in Castile, and then to have been more decisively rehabilitated at an official level after the Council of Valladolid of 1228 (Wright 2000; 2003: chapter 18). Meanwhile in Leon, after the Treaty of Cabreros, there is no other surviving Romance text from the chancery until after the union of the two kingdoms under Fernando III in 1230. It is striking to note that there is a document from the Leonese chancery, dated September 1207, which refers extensively to the stipulations of the Treaty of Cabreros of the previous year, but which was written once again in the traditional Latinate mode (other than the many toponyms, most of which are spelled there in the same obviously non-Latin way as they were in the Treaty; see Wright 2010a).
Written Romance was being developed in Castile as an occasional proper medium for legal texts throughout the first half of the thirteenth century, a development which was accelerated in the 1240s, when many monastic and other centres saw a marked rise in the proportion of documents transacted in Romance. This change of taste may well have been connected with the spread of fueros. The first fueros were in Latin, such as the highly influential Fuero de Cuenca of c.l 190 (Powers 2000). But some brief ones were being redacted in Romance in New Castile already in the early 1200s, and the general view now among scholars is that this choice was connected with their nature as public documents to be read aloud intelligibly to a wide audience, rather than private texts to be read in a monastic cell, for which the traditional Latin mode was not only more appropriate but actually easier to follow than the new written Romance, for those who had learned to read Latin. New Castile seems to have been the main area for these novelties. The Fuero de Madrid, which was elaborated in several stages, has both Latin and Romance sections; the long Fuero de Alcald de Henares (Torrens Alvarez 2002) shows that if the notaries wished to write lengthy texts in the new mode in the first half of the thirteenth century, they could. Henrique Monteagudo's study (2008) of the Foral (fuero) of Burgo de Caldelas in Galicia shows that the idea was also finding favour there in 1228. The Leonese were less keen to change the written mode in this way, and were to continue to be so; in the second half of that century, Gil de Zamora wrote in Latin, which he probably thought was more respectable than Romance, with the result that almost no modern scholar reads his work.
It has been traditional to see Alfonso X as the crucial figure in the development of written Castilian Romance. As Pedro Sanchez-Prieto Borja (e.g. 2004) and others (including Torrens, Harris-Northall 2007 and myself) have been
42 Roger Wright
arguing, this downplays the role of those who worked out how to represent Romance in writing in the first half of the thirteenth century. But it still seems viable to see Alfonso X as the one who brought in the political, nationalistic and metalinguistic dimension which has characterized inter-Romance rivalry in the Peninsula ever since. Before his time, the arguments, including those between Martin Lopez and Diego Garcia (Diego de Campos) in the Castilian chancery at the start of that century, seem to have been practical, rather than nationalistic. But in the 1240s, when many of the intellectual protagonists of the first half of the century (including Ximenez de Rada) died, the pendulum suddenly swung to the benefit of those who preferred to write in Romance. Berceo, the skilled Latin notary who always said he wrote his verse in romanz (or romance) rather than any geographically more restricted mode of Romance, had started his second career as a writer of McGonagallesque verse in Romance after, and perhaps as a result of, the Council of Valladolid of 1228. At first he must have seemed somewhat eccentric; by the time of his death in the 1250s, his genre, the four-line stanzaic mester de clerecia, was becoming something of a cliche.
It is plausible to attribute the startling metalinguistic success of Romance in Castile, and then its reconceptualized mode as Castilian Romance {romance castellano), to the support given to it by Alfonso X even before he ascended the throne in 1252. He was active both intellectually and in practical, including military, matters before then, and his attitudes are likely to have catalyzed the general shift to Romance documentation in many literate centers during the 1240s. His father, Fernando III, had decreed that the Fuero given to Cordoba in 1241 should be translated into Romance (in vulgarem), although so far as we can tell this did not actually happen until Alfonso's reign. Alfonso's interest in harmonizing, as far as he could, the legal systems in his expanding realms started in the 1240s (if not before), and his collaborators collected together fueros from several areas, probably in both Romance and Latin, which became synthesized in his Fuero Real of 1255-6. The use of written Romance in such an authoritative text, guaranteeing its validity even in legal documentation, can be seen now, and probably also was then, as a definitive indication of the status J of Romance as a separate language from Latin. Such a sociolinguistic change I needs official blessing and prestige, andJEejise of written Romance in the chancery and the law gave it that.
The idea that there was a specifically Castilian Romance followed rtris SQon afterwards. Berceo never seems to have bothered to decide what geographical kind of Romance he was writing (which is why it is anachronistic to refer to him as writing in Castilian). Alfonso's choice of Romance for the written works produced at court was initially conceptualized as the alternative to Latin, as a part of the avant-garde Zeitgeist, but since the linguistic details corresponded / more often than not to features found on the eastern side of any isoglosses
The prehistory of Spanish and the 13th-c. Zeitgeist 43
which ran north to south in the Leonese and Castilian realms, Leonese elements were downplayed, often, probably, intentionally. There had been identifiably ' ' Leonese documentation in the central years of the thirteenth century, but there was no political need to use more than one Romance written mode at court, and written Leonese gradually lost political prestige and practical value (Morala 2004). In the 1270s, Alfonso also helped establish written Galician as a separate respectable entity through his own use of it for poetic compositions of a broadly "lyrical" nature, allying the choice of language with choice of genre, but in Castile romance castellano was always the one with the political prestige. ' °^ Alfonso wanted Castile to have prestige on an international level, as evidenced in his pursuit of the imperial throne, and the language was part of this project. Politics, in the writing of castellano, had thus taken over from the initially specifically linguistic purpose of writing Romance (which was, in particular, to aid reading aloud).
Meanwhile, Portuguese and Galician;vvere acquiring separate metalinguistic identities as a result of Portugal and Gijjicia having become separate political entities during the twelfth century; -and the Catalans, happy to write in a Romance mode based on Provencal features in the twelfth century, developed their own independent written form once the Battle of Muret in 1213 had definitively separated Provence and Catalonia politically. Thus the independent written modes, on which conceptually distinct Ibero-Romance languages were based, were allied closely to independent political units, and each kingdom wanted to be able to claim its own language; by the fourteenth century, they -were able to do so.