Computers & Texts No. 11
Table of Contents
March 1996

Review: In Principio

C. W. Dutschke
Butler Library
Columbia University

An enthusiastic 'hurrah' for In Principio! This Brepols CD-ROM, now available in its third edition, includes over 400,000 incipits of classical, medieval and early renaissance Latin texts, with a planned leap up to 750,000-800,000 in the forthcoming editions. Negotiations are underway for yet further expansion (my personal hope is for the incipit files in the Cambridge University Library, representing the CUL itself, the college libraries and the Fitzwilliam - English manuscripts not yet constituting a formal presence). This radically new richness employs as access to texts the age-old device of incipits, or opening words, as the CD's punning title tells us. Incipits were, for the pre-moveable type world, what title pages are to us today - the weight-bearing identifiers of a text. Some remnants of this habit have stayed with us, in the way in which the prayer entitled the Lord's Prayer is frequently referred to as the 'Our Father'.

The genesis of this new tool is not without significance for an understanding of its virtues and vices. It goes back to the late 1930s when the Institut de Recherche et d'Histoire des Textes (IRHT) was founded in Paris with the mandate of creating a centralized repository of information to study the transmission of classical texts. Since that time, teams of French scholars have systematically transcribed incipits from printed catalogues of manuscripts and from the manuscripts themselves - mainly in the Département libraries, but also in those of Paris (excepting the Bibliothèque Nationale), with large numbers devolving from extended periods of work in the Vatican Library.


The third edition of In Principio initiates the incorporation of incipits from the files of the Hill Monastic Microfilm Library of St. John's University, Collegeville, Minnesota (HMML). In the mid-1960s, HMML began microfilming medieval manuscript holdings of libraries outside of France: Austria, Germany, Switzerland, Spain, Malta. Their incipit files derive partially from these microfilms, and mainly from systematic progress through printed catalogues of other collections; they have built up by far the largest such tool in the United States. Of the estimated 400,000 HMML incipits, around 40,000 have now joined their Parisian confrères; the remainder will be added at a planned rate of 100,000 per year. The two incipit collections have, not surprisingly, remarkably little overlap, given the different and complementary geographic areas of coverage: an initial estimate placed 'duplicate' entries at approximately 10% ('duplicate,' here, meaning entries that contain the same information generated from the same source).

But 'duplicate' entries are a desideratum in the sense of the same incipit and the same author but with differing bibliographic citations, whether to manuscript or printed source. In fact, In Principio is intended to produce a number of matches, thus allowing the scholar to locate additional copies of his text, and to begin tracing the patterns of reception of a text (with which other texts did it circulate?) and patterns of transmission of a text (where and when was it copied?). In this sense, In Principio's 'duplicate' entries are a boon to the encompassing reaches of intellectual history as well as to the very basic techniques of research on primary sources.

Inventiveness and flexibility on the part of the scholar will put In Principio to yet other uses. For example, when one of the 'duplicate' entries is to a Vatican manuscript, a scholar need only take the Vatican call number and look it up in the recent Studi e Testi volumes (318-319, 342, 361) by Marco Buonocore or by Massimo Ceresa, to obtain with remarkable ease a bibliography on a given text. In addition, although In Principio presently carries only Latin incipits, it can also be made to help with vernacular texts: Columbia University, Butler Library, Plimpton MS 170 is a hitherto unidentified Italian computus, beginning, 'Conputo si e una scienzia per sapere ciertifichare del tempo . . .'; the most obvious Latin translation gives as key words 'scientia certificandi'; when requested to search the Latin, In Principio duly produced 'Computus est scientia certificandi tempus. . .' and identified the author, Bono da Lucca, confirmed by further research (see screen shots 1& 2).


In Principio has many virtues, and a few vices as well. So many individuals worked over so many years in building up the IRHT incipit files that some ageing and some inconsistencies are inevitable: as the relatively younger HMML files enter the system, there will be a gradual updating of scholarship regarding attributions of authorship. For example, In Principio (second edition), in accord with a considerable number of manuscripts, still assigns the text that begins, 'Primus profectus est in fervore noviciatus . . .' to Bonaventure, De exterioris et interioris hominis conditione. The enterprising scholar, who knows now the recognized/accepted title of his text, may chose to search this title on DBI-Link, the on-line catalogues of medieval manuscripts held by German libraries: he will find many other copies of the text, with correct attribution to David of Augsburg.

Inconsistency, they say, is a hobgoblin, and the trained Latinists who keyed the 400,000 IRHT incipit cards into the computer did a remarkably coherent job. Inevitably some differences in treatment crept in: a search on the call number 'Bodl. 527' produces seven results, but one must insist and try yet again - to be rewarded with yet a different entry under 'Bod. 527.' A search for 'Si quicquam est o paula et eustochium' (the program ignores capitalization and punctuation) cites a manuscript in Admont and one in Reims; but a third entry for this text (the manuscript in Brugge) will not come up because the variant reading was included as if part of the incipit, 'Si quitquam [quidquam] est o paula et eustochium,' nor will one retrieve the manuscript in Douai, since the database's incipit for that text is shorter, 'Si quicquam est paula.' None of this is unfamiliar to us, as we remind ourselves of the old dictum that computers recognize differences but not similarities. The very well written handbook that accompanies In Principio emphasizes that one form of a search is never enough with any database, that we need to give the machine as many options as possible.


The searching interface between the data and the scholar goes a long way towards overcoming the computer's rigidity and our impatience. It automatically adds/deletes the letter 'h': 'Eustochium' and 'Eustocium' as searches bring equal results, with the screen displaying sometimes one sometimes the other spelling. This is particularly useful in the 'nichil/nihil' and 'michi/mihi' alternation. The interface adjusts silently for variant spellings with 'i/j,' 'u/v,' 'ae/e,' 'oe/e' and 'y/i'; it does the same with consonant variations such as 'c/k,' 'ph/f,' u/w' and double consonant variations especially with the letters 'm' and 'p' ('mn/mpn,' 'mt/mpt' and so forth). It also performs key word search, accepts truncations to the left or to the right, and allows for Boolean searches and searches across fields: 'paula' and/or/not 'eustochium,' as well as 'paula et eustochium' and 'Brugge' (location), or 'paula et eustochium' and '12' (date by century), or 'paula et eustochium' not 'Hieronimus' (author - which gives one the text by Paschasius Radbertus). An easily accessible Browse function will tell us, if we take the time to look, that Bodleian manuscripts of the Bodley fonds are cited both as 'Bodl.' and as 'Bod.'; it reassures us that our search for a given author is most quickly conducted under an established form of the author's name, while the program has already built links between names such as 'Jheronimus,' 'Jeronimus,' 'Jeronymus,' 'Hieronimus,' 'Hieronymus' - any of these spellings will pull out the same 3265 entries. The database and its interface can be more flexible than an alphabetically-arranged card file, which adheres strictly to word order: only 'Paula et Eustochium' and never 'Eustochium' and 'Paula.' In short, although instability of both medieval and modern scribes is reflected in some instability of forms in In Principio, the user may readily compensate for it with thoughtful and varied searches.


In recognition of the international community that needs this tool, In Principio offers its pull-down menus, help screen and handbook in both French and English. English-speakers, and in particular those of us in the United States, have particular need of In Principio since our national utilities, RLIN and OCLC, do not allow searching by incipit, even as these two databases are more and more being employed as a cataloguing location for medieval manuscripts. One wonders if it should be proposed that we all contribute our 'own' incipits to In Principio in order that we all might benefit from its ever extending reach.

The simple fact that In Principio exists is a cause for celebration. We owe a debt of gratitude to the members of the IRHT, to the staff at the HMML, and to Brepols, with its long tradition of excellence in academic publishing, for this magnificent tool.


To search DBI-Link telnet to; at the prompt for net command enter 'o dbilink' and at the prompt for user-number enter 'sbbhand'; the search command for all fields is 'f ft=[title]', the command specific to incipits is 'f in=[the incipit itself]' but note that not all entries have incipits; the command specific to authors is 'f le=[the author's name]; type 'show' to see the hits in brief form; type 's f=all' for the long form; type 'stop' to exit the system.

[Table of Contents] [Letter to the Editor]

Computers & Texts 11 (1996), 20. Not to be republished in any form without the author's permission.

HTML Author: Michael Fraser (
Document Created: 25 April 1996
Document Modified: 27 April 1996

The URL of this document is