Since this proposal is for a project whose principal aim is to establish a digital data resource, it is appropriate to provide a technical appendix as required in the guidelines on Joint C&IT policy issued by the AHRB and AHDS.
There are no significant existing data resources in the field of this application. A number of pieces of lute music have been made available in various data formats (MIDI files; audio files; specialist music-files; pdf or PostScript files, etc) by enthusiasts belonging to various organisations on the Internet [The largest such collection is at the following URL:
http://www.cs.dartmouth.edu/~wbc/lute/lute.html.
In no such case to the applicants’ knowledge has any scholarly rigour been applied to the selection or editorial procedures involved. The aims are purely recreational. One CD-ROM of digital images of lute manuscripts has been produced commercially [TREE Edition, Lübeck: a CD-ROM of graphic images of 10 lute manuscripts from the Goess collection, Schloss Ebenthal, Carinthia].
Two forthcoming projects, however, are likely to be of significant scholarly value:
Dinko Fabris and John Griffiths, eds., a CD-ROM edition of Berlin/Krakow MS 40032, c1580, a large and important collection of Italian and Spanish music for the lute or viola da mano; [Personal communication from the editors.]
Franco Rossi, ed., a CD-ROM edition of an important 16th-century MS discovered by the editor in Castelfranco, which contains unique material by, inter alia, Francesco da Milano. [This is a secretive project, whose existence is suggested by little more than hearsay, but it seems plausible.]
As far as catalogues and bibliographical resources are concerned, the printed music is well covered in H. M. Brown’s Instrumental Music Printed before 1600; the manuscript sources are being covered in C. Meyer et al’s Catalogues des Sources Manuscrits en Tablature, which will supersede Boetticher’s RISM BVII catalogue. But both publications lack any musical content, and identifications of related works are presented as faits accomplis.
Some ‘thematic’ catalogues of incipits exist: Flotzinger (Kremsmünster MSS); Maier (Vienna baroque lute MSS); Rudén (tablature manuscripts in Sweden); Dr Josef Klima’s privately-printed thematic catalogues of various mss; Kirsch and Meierott’s catalogue of tablatures formerly in the Prussian State Library, Berlin, now housed at the Jagiellonian Library, Cracow; the series of thematic catalogues of Bavarian libraries (Nuremberg, Augsburg, Munich, Regensburg etc). But these all cover a limited range of sources, and, useful though they are, they are highly inconsistent in method. Currently, there is no mechanism for including musical data from lute tablatures in the RISM A series of thematic catalogues, and technical problems mean that this is unlikely to occur in the near future.
There are several reasons why the proposed project will achieve goals that are inconceivable without the creation of a data resource of this nature. As well as the points mentioned above in the general project description, additional benefits of a comprehensive data resource of lute music include:
collaborative data-entry, with no restriction on the geographical location of participants;
a ‘cumulative’ attitude to the repertory studies (the same study can be carried out at different stages of data-accumulation as new sources are added, and the changing results used to build up a gradually more focussed picture);
simultaneous performance of several different investigations, either as alternative methods of analysis or to provide several ‘dimensions’ of result-data;
the identification of much anonymous material through concordance studies;
the use of modern IT techniques for the study of individual composers in the Corpus (especially studies of chronology and mutual influence) or for studies and presentation (using graphic techniques) of filiations and stemmatics;
the identification and assessment of neglected areas of the repertory, and the possibility of addressing this by online or printed editions;
the provision of online critical editions, allowing variants to be presented in hypertext format;
comparison with non-lute repertories that exist in machine-readable form (e.g. a number of correspondences between the lute music of Weiss and the keyboard music of Bach have already been noticed in private research and there is no reason to suppose that many more could not emerge from such a study).
For lute-players, the usefulness of the Corpus is obvious. For other musicians it may not be so clear. Fundamentally its utility centres around the nature of the music’s performance-oriented notation, tablature. This gives a ‘recipe’ for performance rather than a diagrammatic representation of the music, and thus it is focussed on certain technical aspects of performance (left-hand positions, right-hand fingerings) rather than on its structure. (Since the lute’s sound did not sustain for long, the notation assumes that each note will ‘die naturally’ and consequently the actual durations of notes are never generally indicated.)
The Corpus will provide a convenient library of repertory for players, allowing the comparison of different versions of pieces, where these exist, and facilitating choices between them. These differences sometimes amount to recompositions of an existing piece, or vestiges of the aural tradition of improvised elaboration; in either case, the availability of such variant versions will be highly useful in understanding the manner of performance insofar as it affected the actual notes.
For performers and scholars, there are many important opportunities the Corpus presents. For example, in certain parts of the repertory, the different versions represent another aspect of performing practice. Variants often correspond to aspects of the conventional style of performance on the lute and sources vary in how they represent such explicit details. (Examples include the use of ‘notes inegales’ and the characteristic spreading/arpeggiation of chords which leads to the so-called ‘style brisé’.)
Sometimes pieces of lute (and other) music are based on ‘parody’ techniques, in which a section of, or a complete, pre-existing work is used, either as a ‘skeleton’ around which a new composition is elaborated, or as a source of extracts which are ‘glossed’, or commented upon. Occasionally such references are buried within a composition and not acknowledged in the source; IT techniques can be used to locate more such references than human observers may notice, thus allowing major insights into creative influence across the repertory.
The Corpus will also offer a new view on the transmission and reception of the music, especially when the musical data is combined with the codicological and historical information that is known about the sources themselves. The complex patterns of scribal connections and ownership, intimately bound up with the teacher/pupil relationship, might be ‘mapped’ in ways hitherto impossible.
For those who do not play the lute yet need or wish to discover the repertory and its extent, the Corpus will add a very significant adjunct to the somewhat arbitrary resource of commercial recordings. Although live performance on historically-appropriate instruments is taken as a sine qua non by the project, the crude approximation offered by MIDI playback will at least show something of the richness of the repertory as yet not recorded. It will become possible for the first time for non-players to appreciate the significance of the instrument at all stages of its existence.
From the above it can be seen that the data resource proposed here, although controlled strictly by the highest scholarly criteria, will represent a significant benefit to a large musical constituency, within and without the academic community. All classes of this potential ‘audience’ will be consulted, by announcements on the very lively ‘LuteNet’ as well as through more conventional procedures. As far as the detailed methods of presentation are concerned, an extensive programme of consultation is planned for the later, more public, phases of the project. A great advantage of a World-Wide-Web, Java-based, approach, is that changes to the user-interface can be made with little interruption to the functioning of the system, and it is foreseen that user-feedback will play a large part in the final design of the UI.
As far as the selection of content is concerned, this will be under the control of the project’s scholarly team, and will be directed towards the research goals of the pilot ‘case studies’, rather than a response to ‘requests’ from outside. As it is impossible to predict the reaction of copyright holders to requests for permission to reproduce sources in this fashion, the extent to which the coverage can be fully comprehensive is equally unpredictable. But it is hoped that a satisfactory method for restricting ‘public’ access to certain materials may satisfy reluctant copyright owners without invalidating the scholarly goals of the resource.
The project description and justification for data development above outline the nature and extent of the data being selected for inclusion.
The project description and justification for data development above indicate ways in which copyright issues affect both the selection of data sources and the degree of public access that may be provided. It should be stressed again that the reaction of copyright holders is entirely unpredictable at this stage, and is likely to vary widely.
The essential data resource that is unique to this project is the lute tablature itself. This will be stored in TabCode, an ASCII-based encoding devised by Mr Tim Crawford some years ago precisely for this purpose. [T. Crawford, ‘TabCode for Lute Repertories’, in Computing in Musicology, 7 (1991), 57-59] This simple coding scheme is easily created manually in two forms: a ‘minimum’ form in which the basic musical material is encoded and a ‘full’ form in which details such as lines, slurs, text-items and other extra features are also incorporated. Mr Crawford has also developed The Tablature Processor, a lute-tablature editing, printing and playback program for the Macintosh computer which can save and read the TabCode file-format amongst several others. During the project a Windows version will be developed and distributed to those collaborating in data entry. The same program (or, rather, modules from the same program) gives the viewing and audio playback facilities for the user-interface to the ECOLM system.
Since TabCode is an ASCII code, it is assumed that standard data-compression methods (such as zip) will be adequate, although in practice they may not even be necessary. Some experimentation will be necessary to determine whether a combination of the ‘minimum’ TabCode (highly economical in disk-space, but not all-inclusive) with medium-resolution graphical images (expensive in disk-space, but including all the information from the source) might not be the most effective mode of storage. The graphic files (monochrome) would mostly be created by scanning from 35mm microfilm, and would be highly susceptible to standard compression techniques.
The analysis of the musical data will be carried out using various techniques specific to the musicological studies being undertaken. The most important tool will be the searching and retrieval techniques being developed at King’s College (Depts of Computer Science, Music and Electronic Engineering), hopefully with funding from the joint JISC/NSF International Digital Libraries Program. The data can easily be translated into the formats required by David Huron’s HUMDRUM suite of Unix programs, which provide a comprehensive range of specialised programs for music-analysis including some specifically designed with tablature notation in mind.
The project will be designed throughout as a World-Wide-Web-based resource, and access for use of the data will be provided via the usual browsers, with specially-written Java applets making use of the sound and visual display facilities increasingly becoming available as standard in that environment. One special feature will be the provision of music fonts as appropriate for display. It is hoped that the Unicode standard (which provides a useful set of standard music-characters) will be implemented fully in time to be used during the project.
Trial and user promotion (see 5, above) will be carried out internally, at King’s College and at collaborating institutions (Universities of Tours and Strasbourg), and publicly via announcements on various Internet mailing lists (med-and-ren-music; LuteNet etc).
See Justification for data development, question 1, above. The applicants are fully informed as to the current state of developments in this area, and maintain regular contacts with organisations such as the Center for Computer-Assisted Research in the Humanities (Stanford University), the leading forum for computer-assisted musicology.
The project has the support of King’s College’s Departments of Music and Computer Science, and Centre for Computing in the Humanities. The Computing Centre at KCL also has much experience in backup and maintenance of data integrity. The present application includes the cost of a single computer to be used in the trial stages as a development platform and as a server. In the future it may become necessary to seek further funding for a dedicated server for the data resource.
Externally, expertise is required the form of the services of an expert programmer to carry out certain parts of the system design. This is likely to be necessary on a half-time basis for the first two of the five years of the project. Where possible any other assistance will be sought from within the expertise available at KCL.
No difficulties are anticipated in making the data resource avalable to the AHDS for the purposes of preservation and subsequent distribution for secondary scholarly research. AHDS guidelines will be followed for documentation of the resource, and consultation with the PADS will be undertaken throughout the project as appropriate.