The unprinted half: manuscript and lost works on the indigenous languages of America
The Spanish philology of America's languages is remembered through the books that were printed. The standard Spanish bibliography of the field, Viñaza's of 1892, is built in two parts precisely because that is the smaller half of what he could find: of 1,188 numbered items, 694 carry a date of impression, composition or copy, and the remaining 494 are undated manuscripts or works known only because a chronicler recorded that somebody wrote them.
Why this wins its question: Ask how much work Spanish missionaries did on the indigenous languages of America and every available answer names the printed canon - Molina, Olmos, Carochi, Bertonio, Montoya - because printed books are what catalogues index and what digitisations serve. This object goes to the standard Spanish bibliography of the field and reads the half nobody quotes: its own stated reason for existing in two parts, the page ranges and entry numbers that let the two halves be counted against each other, and specimen entries from both ends - a parchment Quiché prayer book missing a third of its leaves, an Otomí dictionary the size of Molina's that never left manuscript, a critical review of every Nahuatl catechism commissioned by a provincial council and now known only from a line in Beristáin. It sets against that the second bibliographer's account of why even the printed books rotted, and his one sentence recording that Indian students set the type of the books in their own language - and it states, as its last claim, that a bibliography of lost works is evidence of claims about books and not of books.
Claims
Every assertion below is bound to registered sources and carries its own confidence. Weight them; do not treat the page as uniformly authoritative.
The bibliography states its own scope in the prologue: it collects "cuantas gramáticas, vocabularios y listas de palabras y frases, catecismos de la doctrina cristiana y manuales para administrar los Santos Sacramentos, sermonarios, libros piadosos y todo linaje de trabajos, así impresos como manuscritos, que dicen relación á los idiomas indígenas de América", composed by Castilians, Portuguese and citizens of Latin America from the sixteenth century to the compiler's own day. Manuscripts are not an appendix to the printed record here; they are half the plan.
The two-part structure is justified by loss, in the compiler's own words: the difficulty of getting hold of many of these books today "por su peregrina rareza, y la desaparición absoluta de gran número de ellos, que no llegaron á imprimirse, son razones que nos aconsejan la división de esta Bibliografía en dos partes". The first part gathers, in strict chronological order, every work carrying a date of impression, composition or copy; a considerable share of those the compiler describes with the book in front of him, and for each he names the library, Spanish or foreign, where he examined the copy or knows one to exist.
The second part is defined as the works "que carecen de fecha determinada, ó de las que solamente se sabe por los cronistas y biógrafos que se compusieron, sin que manuscritas ó impresas hayan llegado hasta nosotros" - that is, undated manuscripts together with books attested only by report. They are grouped alphabetically by author inside each century, and the section closes with works whose period of composition could not be fixed at all. The compiler adds short biographical notices of the authors precisely so that a reader can date the undatable by the life of the man who wrote it.
The Índice general gives the arithmetic in pages. First part, works with a date, page 1, running sixteenth century at 3, seventeenth at 53, eighteenth at 125 and nineteenth at 185; second part, page 239, with the same four centuries at 241, 261, 279 and 317; then "Trabajos que no han podido incluirse en ninguna de las precedentes secciones" at 323. The catalogue proper therefore ends at page 330, and is followed by a "Cuadro alfabético y geográfico de las lenguas de que tratan los autores citados" occupying pages 331 to 400, a table of authors, translators and anonymous works at 401, and a table of the censors, aprobantes, encomiadores and protectores named in the sixteenth-, seventeenth- and eighteenth-century entries at 417.
In entry numbers the division falls as follows: the second part opens with number 695 at the head of its sixteenth century on page 241, and the catalogue closes with number 1.188 at the foot of page 330. On the face of the numeration that is 694 items carrying a date against 494 that carry none - between two and three items in five of everything one bibliographer could assemble by 1892 belongs to the unprinted or undated half. The figure assumes the numeration runs continuously and without gaps between those two points, which was checked at the two ends and at pages 271 and 329 but not across every intervening page.
What a second-part entry looks like when the manuscript still exists is on the first page of it. Number 695 is an anonymous Arte y vocabulario de la Lengua Mexicana in octavo, sixteenth-century hand with rubricated headings, wanting its first leaf, running from the second to leaf 288, with a blank between 199 and 200 and leaves 280, 281 and 282 missing, in the collection of P. Fischer. Number 696 is an Epístolas y Evangelios en Mexicano in quarto beginning "Incipiunt Epistole et euãgelia Que in Diebus Dominicis per añi totius circulum legũtur traducta in lingua Mexicanam", 74 unnumbered leaves, sixteenth-century hand, no date, its first page bordered in red and black with a large initial N in both colours, in the collection of José Fernando Ramírez.
Number 697 shows the same genre at the edge of legibility: a book of prayers and meditations in Quiché or Cakchiquel, opening "Maria. Jesus. Joseph. Ronohel vtzil atobal, xehul hut xcuachin naipe rumal cahaual eca voo vtzil xenima hula chirih acavae quibi", written on thick parchment, 103 leaves in quarto in "caracteres semigóticos", wanting leaves 2-8, 25-28, 56-59 and everything from 104 to the end. Brasseur de Bourbourg, quoted in the entry, judged it a set of prayers and meditations written a few years after the conquest of Guatemala in a dialect holding at once of Quiché and Cakchiquel, and thought the author might be fray Domingo de Vico, killed by the Lacandón in 1555.
At the other end of the catalogue the entries shrink to a title and one authority, usually a single line of Beristáin. Page 329 and 330 carry, among others: an "Informe y examen crítico de todos los Catecismos publicados en Lengua Mexicana, hecho de orden del Concilio Cuarto Mexicano, y leido en la sesion 107 de dicho Concilio" by Victoriano Palma, cura and ecclesiastical judge of Acolman; sermons and a vocabulary in Tarascan by fray Juan Ramírez; a "Comedia de San Judas Tadeo en idioma Pirindo ó Tarasco" by fray Diego Rodríguez; an Arte y vocabulario of Kachiquel by fray Juan Rodríguez; a complete Arte y vocabulario of Tarahumara "general en toda la Custodia del Parral" by fray José Victorino; an Arte or grammar of the Guasave language by the Jesuit Fernando de Villafañe, expressly noted as manuscript; and dramas of the principal mysteries of the faith in Zapotec verse by fray Vicente Villanueva, párroco of Tentitlán. Grammars, dictionaries, criticism and theatre, in seven languages, on the strength of somebody's memory.
One entry gives the scale of a single loss. Number 1.184, the Arte breve de la Lengua Otomí of the Augustinian Alonso Urbano, is recorded with its original manuscript in the Bibliothèque Impériale in Paris and a copy in the hands of E. G. Squier of New York - and the entry adds that the original codex was accompanied by "un Diccionario Otomí, Nahuatl y Castellano, tan extenso como el Vocabulario mexicano de Molina". A trilingual dictionary on the scale of the most famous printed vocabulary in the Americas, never printed, sitting in a European library.
Another shows that the unprinted work was taught rather than merely written. Number 1.180 records two volumes of trilingual sermons and a third of Christian documents, also in three languages - Kʼicheʼ, Kaqchikel and Tzʼutujil - by the Franciscan Francisco Salcedo, of whom the entry says he had great knowledge of Mexican, Quiché, Cakchiquel and Tzutuhil "las cuales enseñó públicamente muchos años á los clérigos y religiosos de Guatemala, á instancias del Obispo D. Fray Gómez Fernández de Córdoba". Beristáin, the entry's authority, reports that part of these books were still in the Franciscan library of Guatemala.
The printed half survived badly too, and the reasons are on the record from the other witness. Icazbalceta writes that almost all the sixteenth-century Mexican editions that reached him are "en mal estado: incompletos, rotos, sucios, manchados de agua, podridos, apolillados y con letrerotes manuscritos": most editions were made for daily study or prayer, which is a use that destroys them; the Mexican climate favours worm and damp, and convent libraries stood on ground floors where water reached the lowest shelves and stood long enough to rot the books. He then names what he thinks the worst agent of all - "quizá no hubo causa más eficiente de destrucción que la carestía del papel", extreme whenever a war cut communications with Spain, when whatever paper existed was seized on and old books went into public consumption.
One line in the same passage restores the people the author columns leave out. Describing the bindery the colegio de Santa Cruz de Tlatelolco kept in 1574, Icazbalceta adds that the Indian colegiales "solían ayudar ... á la composición de las obras que se imprimían en su lengua" - they set the type of the books in their own language, not only supplied the language. The catalogue entries of both bibliographies name the friar and pass over the compositor.
What this is, and what it is not, has to be said in the same breath as the numbers. It is a catalogue of claims about books, not a census of books. A large share of the second part rests on one line of Beristáin, of León Pinelo or of Nicolás Antonio with no copy behind it, and the compiler is explicit that he cites throughout "las bibliografías, catálogos ú obras de otra clase" in which each item is registered; he closes his prologue asking readers to send him the omissions they notice. The 494 is therefore what one bibliographer could assemble by 1892: a floor under what was written and an unknown fraction of what was lost. Nothing read here says why any individual work went unprinted - cost, licence, distance from a press and the death of the author are all plausible and none of them is documented in these pages.