(focused on multiple versions)[1]
(Variants 5. Texts in Multiple Versions – Histories of Editions,
Ed. Luigi Giuliani, Herman Brinkman, Geert Lernout and Marita Mathijsen
Amsterdam/New York, pp. 205-220)
José Camões
Centre for Theatre Research, University of Lisbon
When dealing with the history of Portuguese theatre – or rather, with the history which is limited to an index of published authors and texts written to be performed – the researchers usually have a preference for the 16th century. However, a mere glance at what is being studied and published easily shows that all of them have a common theme, which can be summarised as "Gil Vicente and those other ones".
In fact, few scholars have approached the so-called followers of Gil Vicente, either those who came after him or those who followed his style and are commonly known as "the school of Gil Vicente".
This I found three years ago, when I undertook the edition of the Portuguese 16th century theatre, after having directed the research team who prepared the CD-ROM edition of Gil Vicente. Todas as Obras two years earlier.
I began by re-reading all the histories of Portuguese theatre, all the histories of Portuguese literature, and perusing repertoires and bibliographical dictionaries, as ground material for establishing an index of authors and plays for the project. The outcome of this was rather frustrating. Only 79 texts from that period have survived.
We must bear in mind that 46 plays by Gil Vicente alone have subsisted. I find it very improbable that one single author could have produced 36,8 % of such a scanty catalogue of 125 plays.
A number of authors have written those 79 plays; some were such important literary figures as Sá de Miranda, António Ferreira, and Luís de Camões. It is not surprising that their dramatic works are few, for it is plausible that their writings for the stage were a digression from their poetry; Sá de Miranda wrote two comedies, António Ferreira wrote two comedies and a tragedy, and Luís de Camões three comedy plays, or "autos".
What is indeed surprising is that we find some authors – they are little more than mere names now – who are supposed to have written only one text in their whole lives, and oddly enough that single work is a theatre play. Such is the case of Jorge Pinto, Anrique Lopes, and Jerónimo Ribeiro. Moreover, it is quite surprising that only an average of four plays have survived by authors who wrote exclusively for the stage — or playwrights — as we would call them today, such as Afonso Álvares, Baltesar Dias, and António Ribeiro Chiado.
A good example is given by a volume published in Lisbon in 1587, bearing the title Primeira parte dos Autos e Comédias Portuguesas, feitas por António Prestes e por Luís de Camões (first part of the autos and comedies made by António Prestes and by Luís de Camões). The publisher included merely two plays by Camões (and he probably mentioned him on the title to catch the attention of the buyers). As far as the title indicates, we have reason to believe that the publisher had intended to produce at least a second volume, but he never did, or, at least, it has never been found. With the exception of the plays by Camões, all the texts in this collection had to wait until the 19th century for a second edition.
This means that we must go on searching and rummaging through libraries and archives in Portugal and abroad.
For the moment, my concern will be only the prose plays: the Comédia dos Estrangeiros by Francisco de Sá de Miranda, a poet that lived between 1485-1558, and is a landmark in Portuguese literature, for he his considered the first one to use the Italian manner in the Portuguese XVI century.
There are no autograph plays by 16th century Portuguese writers; manuscript "witnesses" are rare, and these are, for the most part, single copies. When we weigh the Portuguese scarcity against the Spanish profusion the contrast is appalling; and yet, if we think of Shakespeare, then we know that we are not in bad company[2].
Sá de Miranda was left almost unharmed by this doom[3]. Some of his autographs have survived, along with reliable copy-texts, which are sufficient in number for us to understand his writing process and to establish a normalized edition of his poetry, as his first modern historian, Carolina Michaëlis de Vasconcelos, had wished[4].
Little was added to this source of knowledge for almost a century, until José V. de Pina Martins published his studies[5] and Arthur Lee-Francis Askins edited the Portuguese "cancioneiros"[6] that include some of Miranda’s poetry. Even so, the plays by Sá de Miranda did not appeal to researchers or scholars; a few minor references were occasionally made in connection with other authors or with the classical period in general, but there is only one modern edition of his comedies[7]. T. F. Earle wrote the most thorough and innovative study on Os Estrangeiros[8], going far beyond the mere attempt to identify the sources of the play as the occasional surveys did[9].
1559 – Comedia dos Estrangeiros, Ioam de Barreyra, Coimbra
1561 – Comedia, intitvlada os Estrangeiros, António de Maris, Coimbra
1595 – Comedia dos Estrangeiros in As obras do celebrado lvsitano O doutor Frãcisco de Sá de Mirãda, Collegidas por Manuel de Lyra, Lisboa
1622 – Comedia dos Estrangeiros in Comedias Famosas Portuguesas dos Doctores Francisco de Saa de Mirãda, e Antonio Ferreira, Antonio Alvarez, Lisboa
1784 – Os Estrangeiros in As obras do doctor Francisco de Sá de Miranda, Tipografia Rollandiana, Lisboa
1804 – Os Estrangeiros, in As Obras do celebrado lusitano o doutor francisco de Sá de Miranda, Impressão Régia, Lisboa
1937 – Os Estrangeiros in Obra Completa II, ed. Rodrigues Lapa, Sá da Costa, Lisboa (3rd ed., 1977)
1989 – Os Estrangeiros in Poesia e Teatro, ed. Silvério Augusto Benedito (excerpts), Ulisseia, Lisboa
1994 – Facsimile of the 1595 edition, Universidade do Minho, Braga
In 1896, Fernando Palha, a collector of ancient books, published the catalogue of his library; the entry no. 1218 registered a hitherto unknown copy of Os Estrangeiros published in 1559. He made a note about the discrepancies between this edition and the well-known volume of 1561; he copied out the first scene of Act I of the play from both editions, using the left column for the transcript of the earlier edition, and the right column for the 1561 edition[10]. However, as the same T. F. Earle pointed out, «the lead has never been followed up, not until this very day»[11].
I concur with the British professor's opinion, and this paper will try to make up for the flaw that he mentioned. I'll follow up Palha's lead, but having other tools to work with, I present a project designed for the electronic medium, making the transcript of both texts available to the reader, and establishing the first modern edition of the 1559 text. This is not a simple task.
The transcript of the 1561 text, which was the basis for all the subsequent editions of the comedy, doesn't entail any major problems. Browsing the 20th century edition of the comedy, we'll find that its editor, Rodrigues Lapa, registered 28 orthographic differences between the 1561 and 1595 versions; all these differences are typographical errors that can be put down to a careless printer or to different orthographical criteria, for example: auto / acto; deixar / dexar; já / lá. But the transcript of the 1559 text is a much more difficult task.
T. F. Earle analysed the differences between the two versions, and derived different kinds of information from his reading, such as the date of its writing (1523 or 1524) and the author's reasons for specific rewriting options, as for example:
«A comparison between the two versions of the text shows that Sá de Miranda’s struggles were great indeed. He had no particular problems with the structure of his play, but seems to have found it very difficult to give his foreigners a voice. So in both versions the characters are the same and, with one or two minor exceptions, have the same names. There are the same number of acts and scenes, and the complex plot develops in the same way. However, the way in which the characters express themselves is surprisingly different. Very few speeches are the same, and even minor changes show signs of rewriting. It seems likely, however, that the better-known version of 1561 represents Sá de Miranda’s final thoughts about his comedy».
We are tempted to think that there was a version of the comedy for the readers (1559) and another (1561) for the stage, since the first version clearly shows that the author was more concerned with literature than with theatre.
I must emphasize that both editions were printed after the author's death, which probably occurred in 1588. Therefore, we can only infer that the printers had access to, and used, two different autograph texts. Coincidentally, the sequential order in which the editions were published may have been consistent with the sequential order of the writing of the texts.
The history of this comedy's editions can be summarised as follows. There are two versions from the 16th century, one of which gained prominence and was used by all subsequent reprints. These reprints are – or were – quite unproblematic from the point of view of textual criticism, since, as I have mentioned earlier, there are only minor typographical differences of orthography among the surviving "witness" printed texts; there are no textual variants. The 1559 edition has never been reprinted. The one other "witness" that we knew of, the Asensio manuscript[12], varies slightly from the printed edition; therefore, a critical edition would not involve too many difficulties.
The stemma would present a simple correspondence between the author's autograph manuscripts (A and C) and the texts published in 1559 and 1561, and the Asensio Manuscript, perhaps with the intervention of one or two copy-texts (α and γ) of the first autograph, and another (δ) of the second autograph:
Isabel Carlos, one of the researchers who are preparing the CD-ROM edition of Teatro de Autores Portugueses do Século XVI, delved into the Hollis online Catalogue of Harvard University, and found the entry MS Port 13.
Under that entry she found the following title from the Houghton Library holdings: Obras: do excellente poeta Francisco de Saa de Miranda dirigidas ao Principe Dom Joao, filho del Rey Dom Joao terceiro de Portugal: manuscript, [15--].
A Full View of Record was requested, and the results were highly surprising, as the summary provided information that was unknown until then:
Author : Sá de Miranda, Francisco de, 1481?-1558.
Title : Obras : do excellente poeta Francisco de Saa de Miranda dirigidas ao Principe Dom Joao, filho del Rey Dom Joao terceiro de Portugal : manuscript, [15--]
Locations/Orders : Availability
Location : Houghton MS Port 13 Holdings Availability
Description : 603 p., bound; 18 cm.
Summary : Two comedies and numerous poems in an unidentified hand; there are textual variations among those published.
Provenance : Purchased with the Amy Lowell fund, 1962.
Notes : Cf. watermarks to Briquet nos. 11058, 5688, 5703/4.
In the original limp vellum wrapper.
Cite as : MS Port 13. Houghton Library, Harvard University.
Subject : Portuguese poetry – 16th century
Portuguese drama (Comedy) -- 16th century
Form/Genre : Poems -- Portual [sic]-- 16th century.
HOLLIS Number : 009267767
Apparently, this was a manuscript from the 16th century that had never been mentioned in any bibliography or bibliographical repertoire. Until now, and after perusing many bibliographies about the poet, I have in fact found no other reference to this manuscript.
Apart from the two comedies, the manuscript consists of poems already included in other manuscript collections; the Epístola de Dido a Eneas of which there is a copy in the Cancioneiro de Luís Franco Correa, with an appended note attributing the work to Francisco de Andrade; and two sonnets by unidentified poets that I didn't find in any other copy-text.
The first reading of Os Estrangeiros revealed that differences between this text and the printed versions are in such large number that we must reach the conclusion that it is in fact a witness of intermediate version of the play situated between the two printed texts of 1559 and 1561. The number of acts and scenes is the same, which allows for a handy examination. The most significant changes and features can be reflex of the author’s experiments on writing and rewriting towards a scenic version: the Harvard Ms maintains some text of 1559, it ads text that will be preserved in 1561 and has some entirely new text that was not present in the 1559 and that will disappear in 1561. Words counted, each text is shorter than the previous one.
This leads to a new stemma, where B is an autograph manuscript, chronologically placed between A and C, which was the source of the Harvard manuscript, either directly or through the unknown copy text (β).
To summarise, we have two printed texts that are very different (1559 and 1561), a manuscript (the Asensio manuscript) with slight variations regarding the earlier printed text, and another manuscript (the Harvard manuscript) with extensive variations when compared to both of them. This manuscript is particularly interesting to the issue of editing multiple versions of a text in a computer.
The word processors' design is based on the model of the printed page – an apparently perpetual model. And the fact is that word processors' users ask no more from them than to be provided with an application that they can handle as a sophisticated typewriter.
In our previous experience – Gil Vicente. Todas as Obras – there was a particular setback, as we had to insert specific characters that were not provided by the existing Windows fonts. Those characters are the ũ, ẽ, and ĩ, which can be inserted through the symbol box, and the abbreviations for que and para/por (² and ®), which were designed exclusively for the project. Following the example of the Spaniards who were intelligent and strong enough to carry an international campaign that demanded a key for the ñ character, even though that character could be easily inserted using the traditional diacritical mark on the keyboard, I consider that a similar claim ought to be made in order to extend this function to all vowels and consonants – which would result in a more economic process of entering symbols.
The corporate decision should derive from a simple consultation of the users, rather than conforming to a few people’s typing routines as if no other writing requirements existed. I am not aware of any important technical difficulty in providing tildes for five vowels instead of the two (ã and õ) that are available on the Portuguese keyboard; perhaps it’s a question of tradition, or of ignorance. The technique of marking letters is a legacy from the typewriting age, and it’s not a bad option, although the old typewriter allowed the use of any mark over any letter, a possibility that is not allowed by computer programs.
We should understand that computer technology can assist researcher by using the exact same processes that are available to common tasks, and that it is not necessary to devise new programs: a small improvement of the existing programs would suffice.
Other potentials must be explored, and even created, which require new technological devices and a proficient staff, an association of resources that is often difficult to find. To bring together a team of professionals from different areas (computer engineers, software programmers, textual critics, palaeographers, essayists, and so forth) who can establish an effective communication among themselves is hard work indeed, and more time-consuming than what is usually expected[13]. However, based on our work Gil Vicente. Todas as Obras, we began the production of an object where hypermedia would produce the proclaimed liberation of the codex[14].
This new project is more innovative than the previous one, for it involves new tools to insert the text markers, which were designed as the team visualised the final stage of the edition – that is to say, the design was focussed on producing researcher-friendly forms and user-friendly applications.
Since there is no specific software available in Portugal for the processing of ancient texts, the more low-priced way to solve the problem was to use applications that are easy to find and create connections among them. Microsoft Word seems to have monopolised the market, even for those who use Apple computers, and that was the application we chose.
By the request of the editors, computer technicians created new commands in order to run Visual Basic applications from the Microsoft Word toolbar, such as the insertion tags for music, images, dictionary, themes and notes.
The text is marked using Word macros that assign a database index to the selected text. Microsoft Jet Database Engine manages the links among the different databases that are being created and the application. The making of a Glossary takes the following steps: when a word is selected the editor runs a macro that searches a database to check if there is a record of that word in the same database. As a result, a box is displayed with a list of indexes assigned to that word, which means that that word has already been marked and recorded in different contexts, with different meanings or in different languages (Spanish or Portuguese, mainly) or modern spelling. The editor chooses the one that matches the word he wants to mark and clicks in the button insert.
If none of the indexes fit the needs of the researcher, he has to create a new one, choosing the button nova (new), which will at once be recorded in the database that refreshes automatically, so that in the next search it can be already available. At the same time, the editor can review his work and correct misspelling and other mistakes.
The some procedure is applied to all types of text marking demanded by the project: quotations, notes or a thematic field predefined to be searchable, as, for instance, onomastics.
The result for the read will be something like this:
In the case of texts with multiple versions, new macros had to be created to generate new tables in the database, one for each witness, in order to show the differences with a click. For the 1559 text of the Estrangeiros, the transcript of each witness (print and Asensio manuscript) carries the marking of the differences between each other, separately, relating to different tables. In order to consider the differences between them and the Harvard manuscript, this last one had to be marked as well. Thus, each text was marked with two different tags, one for each one of the other two.
In order to achieve the following result, after assigning a different colour to each version, in order easily recognize the compared text, and the buttons that allow the view of the facsimile if desired:
This process of comparison lead us to the conclusion that the Asensio manuscript is not a version of a text, but a simple witness that present some important variants, a better lesson – it corrects some of the mistakes that the 1559 printers made, it is not censured, and therefore, the one that must be used when critically editing the 1559 text. On the other hand, the Harvard manuscript proved to be a real version, showing some differences that make a new text. This electronic edition provided with these tools shown above, allows the editor to economically establish the texts and at the same time allows reader to choose from all the material that is offered and assemble his own edition.
In conclusion, I believe that it is essential to treat the various versions of a text as separate texts, that is, to edit all of them. It is not possible, nor desirable, to make a critical edition with a choice of variants, since these are not actually variants but separate objects which seem to be the outcome of two expressions of the author's will.
[1] Project POCTI / ELT / 33464 / 2000, from the Foundation for Science and Technology (Portugal).
[2] The reason for the scarcity of surviving texts may be related to the consequences of the 1755 earthquake, and in particular the fire that destroyed a large number of houses in Lisbon. Other and less natural causes may also be involved in the mysterious loss of autograph manuscripts.
[3] Five centuries later, Fernando Pessoa made amends for it by offering us enough works and enough identity problems to keep the editors active with textual objects that probably were not even meant to exist.
[4] «...reconhecemos cada vez mais a necessidade de substituir as edições conhecidas e muito defeituosas por edição normal». [We are increasingly aware that the variety of very faulty editions that exist ought to be replaced by a normalized edition] in Poesias de Francisco de Sá de Miranda, Halle, Max Niemeyer, 1885, p. L (Facsimile, Lisboa, Imprensa Nacional-Casa da Moeda, 1989).
[5] Sá de Miranda e a Cultura do Renascimento. I – Bibliografia, Lisboa, 1971
[6] The Cancioneiro de Cristóvão Borges, Barbosa & Xavier Limitada, Braga, 1979, and Cancioneiro de Corte e Magnates, University of California Press, Berkley and Los Angeles, 1988.
[7] Obras Completas, vol. II (edition, notes and foreword by Rodrigues Lapa), Lisboa, Sá da Costa, 1937 (3ª ed., 1977).
[8] The comedy of the Foreigners. Renaissance Sicily through Portuguese eyes. Clarendon Press, Oxford, 1997; «Rhetoric and drama: the two versions of Sá de Miranda's Os Estrangeiros» in Culture and Society in Habsburg Spain, ed. N. Griffin, C. Griffin, E. Southworth, C. Thompson, New York, Tamesis Books, 2001, pp. 35-44; «Sá de Miranda's Roman Comedy» in Cultural links between Portugal and Italy in the Renaissance, ed. K. J. P. Lowe, Oxford University Press, 2000, pp. 153-166; «Traição e Amargura nas comédias de Francisco de Sá de Miranda» in Em Louvor da Linguagem – Homenagem a M. Leonor Buescu, Lisboa, Colibri, 2003, pp. 87-96.
[9] The most informative being Joaquim José Ferreira's Fontes de «Os Estrangeiros» de Miranda: Plauto e Terêncio (a dissertation for his bachelor's degree in classical filology at the Faculdade de Letras da Universidade de Lisboa, 1940).
[10] Catalogue de la bibliothèque de M. Fernando Palha, 4 vols., Libânio da Silva, Lisboa, 1896.
[11] The comedy of the Foreigners. Renaissance Sicily through Portuguese eyes, p. 19.
[12] A handwritten miscellanea from the first half of the 16th century at Biblioteca Nacional de Lisboa (Res 6032 P.).
[13] Gary Taylor has written a witty description of his first experience in collective electronic edition, which illustrates the difficulties I mention: «c:\wp\file.txt05:41 10-07-98» in The Renaissance text. Theory, editing, textuality, ed. Andrew Murphy, Manchester University Press, 2000, pp. 44-54.
[14] Cf. Graham D. Caie, «Hypertext and multiplicity: the medieval example» in The Renaissance text. Theory, editing, textuality, ed. Andrew Murphy, Manchester University Press, 2000, pp. 30-43: 32, and Jerome J. McGann «The Rationale of Hypertext» in Electronic text. Investigations in method and theory, ed. Kathryn Sutherland, pp. 19-46: 21-22.