Part of the inspiration for the database came from a work of W. H. Frere, Antiphonale Sarisburiense, a facsimile edition of the Sarum Antiphoner. Although most of that publication is devoted to the reproduction of the antiphoner, important scholarly contributions are to be found in the introduction and index.
In the introduction, Frere identifies the salient features of the chants of each type, and within type, each mode. In the index, he lists the chants to be found in the facsimile, including not only the page number but also the liturgical occasion on which the chant is sung, its mode, and--in the case of antiphons--differentia, variatio, and group. If Frere has discussed a particular chant in his introduction or provided an identifying letter for its melody there, a page reference for that is also given.
The system of numbering that Frere uses for differentia and variatio comes from the Sarum Tonale, edited by him in The Use of Sarum. For example, in its presentation of the antiphons of mode 7, differentia 1, it identifies twelve variationes, each of which has a distinctive opening figure. Only a small number of examples are provided--two for the first variatio, two for the second, one for the third, two for the fourth, three for the fifth, and so on.
If, by contrast, one goes through Frere's index counting the number of antiphons that are assigned there to each of these variationes, one finds fairly small numbers of antiphons in some of the variationes but at least fourteen in variatio 3 and thirty-five in variatio 5. This work of classification was apparently done by Frere himself.
In the introduction, Frere presents (once again) a very small number of examples for each melody type. If the reader wishes to examine other melodies assigned by him to a particular variatio, he must search through the index for them. He can either read through the index line by line, marking all of the chants to which he will return, or devise some way of sorting the index in such a way as to bring together the listings in the index of antiphons of the same mode, differentia, variatio, and group. Having completed the selection or the sorting, he must then go through the facsimile edition examining the melodies one by one.
In the 1970s, before many scholars had learned to employ computers in their research, sorting the index had to be done manually. One way of doing this was to make a photocopy of it and cut it apart line by line. The tiny strips of paper for mode-1 chants went in one place, those for each of the other modes in other places. When the job was done the strips had to be laid in order very gently on pieces of typing paper and taped into place.
It was clearly work that could be done better on a computer. In his 1980 Ph. D. dissertation at Catholic University, Ronald T. Olexy made an index of the responsories in Toledo, Biblioteca capitular, 44.2, taking the data file he had created on a mainframe computer and sorting it in several different ways.
That paved the way for a pilot study in which Frere's index of antiphons was copied into a computer file with a few minor changes: mode became an Arabic, rather than a Roman, numeral; differentia became a normal number rather than a superscript, and so on.
Even as data entry was still under way, errors and inconsistencies in the work of Frere and his typesetter became evident. Yet when the file was sorted, the results were gratifying, though it was apparent that the computer could not be used effectively in carrying the project further until the standard of accuracy in the data had been elevated. What had been good enough for Frere's purposes was no longer adequate: there would need to be a great deal of checking. Hence, improving Frere's index, or creating an analytical list of the chants in a different source, could not be planned as a one-person job: it would require a team.
Providentially, in 1987 at Catholic University a small amount of funding became available for work of this type, and well-qualified graduate students were on hand to carry it out. One was Joseph Metzinger, who had come to Catholic University after work experience making adaptations of off-the-shelf computer programs for specific business applications. The other was Lila Collamore, who was able to contribute creativity, clear thinking, and flawless data entry to the work in progress.
The new staff made it their first goal to correct, improve the format, and enrich the content of the file of Sarum antiphons. They standardized the spelling of the Latin and added a field that permitted the chants on each page to be sorted into the order in which they appeared in the source.
Using Olexy's work as a model, but introducing necessary modifications, they transformed the data file into a form that was much like that of the current CANTUS database. Fields were added for concordances; a code was introduced for verses of responsories sung to independent melodies rather than modal formulas, and so on. The results were such that an application to the National Endowment for the Humanities for long-term funding seemed a logical next step.
The central purpose identified in the 1987 proposal was the creation of indices similar to Frere's for sources other than the one he had studied. The first two projects selected were an index of Toledo, Bibl. cap., 44.2 that would incorporate work already done by Olexy and a complete index of Karlsruhe, Badische Landesbibliothek, Aug. LX. Each is a complete (or nearly complete) antiphoner representing an important tradition--Aquitanian and German, respectively--and each is relatively early among sources of its type.
One essential aspect of the proposal was the providing of reassurance to those who were to judge it that the Latin texts and rubrics in the manuscripts would be interpreted correctly. Fortunately, Prof. F. A. C. Mantello of the Department of Greek and Latin at Catholic University, a noted specialist in medieval Latin, was willing to serve as an Associate Director of the project.
He began by supervising the transformation of the erratic Latin orthography of the text incipits in Frere's index into standardized forms consistent with the practice of Dom René-Jean Hesbert's Corpus Antiphonalium Officii, which had been adopted as a model for CANTUS; and he continued to provide major assistance to the project throughout its duration. He found and corrected thousands of errors in the copying of incipits in files prepared by the staff, in contributed files.
The project received generous but intermittent funding from the NEH between 1988 and 1997. Support also came from the Dom Mocquereau Foundation and the Catholic University of America.
The project director traveled to a number of conferences to speak about the project, carrying with her handouts that included (at various times) diskettes containing files and a complete printout of the CANTUS file for Bamberg, Staatsbibliothek, lit. 25. An advisory panel of distinguished medievalists assisted with long-range planning and the selection of sources for indexing. The CANTUS staff expanded.
Some files were published in book form; the books were sent to journals for review; and the reviews attracted new users to the Gopher site (and later the Website) that offered the indices for searching and downloading. Scholars at other institutions in the US, Canada, and Europe contributed files.
Joseph Metzinger's menu for the Gopher site was reproduced in the America Online Tour Guide; the Gopher address was published in TIME magazine. By the time work ended at Catholic University in March 1998, there were indices of approximately 40 sources in the database, comprised of approximately 194,500 records.
When the National Endowment for the Humanities indicated that it could no longer continue its support, a proposal was prepared for the Social Sciences and Humanities Research Council in Canada by Professor Terence Bailey of the University of Western Ontario. The success of this application entailed the moving of the project to that university, where it continues to thrive.