CONTENTS
Return to Contents
Introduction
When preparations were being made for the forthcoming catalogue of the
Edinburgh University Collection of Historic Musical Instruments,
cataloguing standards were established. These were found particularly
necessary to ensure uniformity of presentation because a number of
consultant authors were involved. More importantly, the new catalogue
has been designed from the start to be produced using computers to
facilitate revision of the data at any time when justified by additions
to the Collection or the acquisition of further information about items
already held. No satisfactory cataloguing standards could be found in
the literature, although a number of published catalogues served as
exemplars.
Having, after consultation, established cataloguing standards for a
medium-sized general collection of instruments, I felt that these could
be generalised and be of use to other museums involved in cataloguing,
and could in fact serve as the basis for an international standard for
cataloguing musical instrument collections. The rubrics which follow,
therefore, are being put before CIMCIM for discussion in the hope that
an agreed International Standard may result.
Return to Contents
Cataloguing Objectives
What business are we in ? The music business ? Education ? Conservation
? I suggest we are primarily in the information business and that our
contribution to the making of music or to education is to a large extent
dependent on our storage and transmission of information. Conservation,
the most exacting of our responsibilities, is not an end in itself, but
is directed at maximising the inherent information content of museum
objects for this and for future generations.
A large element of the curator's work is the recording of that part of
the information about museum objects which is not inherent in the
objects themselves - provenance, usage, value etc. Added to this are
the recorded observations and opinions of experts relating the objects
to the worlds of musical practice and scholarship. The typical products
of such documentation have been the accessions register on cards or
computer and the published catalogue in print and photograph.
I propose that where we deal with a well-defined area of museum activity
such as musical instruments, we can standardise our methods to a large
extent in order to raise standards (where necessary), improve
communication and still allow scope for inspiration, ingenuity and
excellence. By selecting the best from each other's methods we can
better communicate information for the research of today and better
accumulate new information for the benefit of future generations.
The main purpose in compiling the information content of a catalogue of
instruments is to relate the knowledge concerning the craftsmanship of
instrument making to our knowledge of historical practice in music
making. A catalogue is deficient which concentrates only on the
physical description of objects (as is common) or (more rarely) on the
performance possibilities of instruments. What is it that we appreciate
about the great cataloguing masterpieces of the past ? The pioneers,
Mahillon, Bessaraboff and Heyde have displayed the greatest sureness of
purpose, clarity of thought, accuracy of expression, and awareness of
the cultural climate in which instruments originated.
I am proposing today that we should look at the categories of
information that we record, with a view to promoting such clarity of
thought, accuracy of observation and creating a structure for our
records which will facilitate the synthesis of research information in
the areas of both lutherie and performance practice. It is this
synthesis which makes our requirements different from those of other
museum curators.
Our categorisation of information and our structuring of records have to
be sufficiently detailed to promote thoroughness in registration and
cataloguing and sufficiently simple to be readily intelligible to museum
visitors and users of published catalogues. Our methods should be
adaptable to the use of collections with small numbers of instruments
and to museums without the staff time or expertise to produce scholarly
catalogues. At the same time, our methods should be capable of
accommodating the most detailed work of collections taking the lead in
the production of research- based catalogues or of prestige illustrated
coffee-table catalogues.
Any international standard for musical instrument cataloguing must also
serve the needs of both large general collections and specialised
collections. I suppose our largest instrument museums (or groups of
museums sharing a union catalogue) will have to catalogue 10,000 or so
objects. Others of us may be faced with more unusual cataloguing
challenges - 700 concertinas, 800 marching band instruments or 900
incomplete instruments acquired from a maker's workshop.
These problems are not new, or entirely our own. They have been
effectively dealt with by the library profession, who have developed an
international standard (The Anglo-American Cataloguing Rules) which is
widely used (not only in English-speaking countries). These rules
structure the recorded information about published books and other
materials into precisely defined fields, with procedures for recording
the information in cases capable of differing interpretation. The same
book, catalogued by librarians working in different places, will be
given identical, or at least compatible, descriptions.
A standardised approach is, of course, made both more possible and more
necessary by the adoption of computers and modern data communication
methods. The system and methods I propose, however, are not specific to
any one computer system or database software. They can be used with any
word-processor or even without a computer at all, by those of us still
using pieces of paper and cards for primary data recording.
I do recommend, however, the approach we are taking at the University of
Edinburgh. This is to separate the recording of information and the
production of a published catalogue as far as possible. Our computer
files containing both registration and catalogue information are made
using the simplest possible structure, using only basic ASCII coded
characters and containing no formatting instructions. They are ready
either for further editing and data processing in Edinburgh or for
transmission to anyone else who can make use of them using any computer
system capable of interpreting ASCII codes.
In fact, we are aiming at both a printed catalogue, which will be
produced from our basic files by a series of database and desk- top
publishing procedures (in the short-term) and at a searchable database
(in the long-term).
Return to Contents
Standardisation
If the rules for cataloguing are too detailed, it is to be expected that
excessive discussion and delay would occur before agreement was reached,
and they would, when published, be less likely to be followed. If the
rules are too diffuse, than they will be ineffective. We should accept
the fact that we cannot prescribe cataloguing rules that will apply to
every single instrument, and that cataloguers will have to approach
exceptional cases by following the underlying principles of the
cataloguing standards. Standard rules should be hospitable to changes
in emphasis and to the introduction of new categories of information if
and when any aspects of the description of instruments achieve greater
importance.
Return to Contents
Cataloguing Principles
In a catalogue, the needs of the reader should determine the content,
presentation and form.
Typical readers of catalogues are players of instruments and instrument
makers. The former are interested in the characteristics of historical
models and require information which places the instrument in its
musical, cultural and social context, whereas makers require different
information. Since any catalogue has to be written with the average
reader in mind, some knowledge of instruments can be assumed. More
detailed information should be given where possible for instruments
remote from the cultural starting-point of the anticipated readership.
If the content of the collection is ethnocentric, the catalogue will be
ethnocentric also.
The information needs of instrument makers are unpredictable and usually
demand greater detail of description and measurement. In a catalogue
published for general readership we can aim only to provide descriptions
of the instruments that will enable the maker to decide whether to
purchase a workshop drawing, to write for more detailed information, or
to visit the Collection in person. Sufficiently detailed information
should be given in the catalogue to differentiate apparently similar
instruments in the Collection.
The catalogue can be updated as instruments are added to the Collection
and as knowledge is acquired about existing holdings through
organological scholarship.
It is assumed that a published catalogue will consist of an
introduction, the main body of entries, illustrative material and
indexes. The latter should include indexes of instrument names, makers
names, places of origin, acquisition numbers, etc.
If practicable, the language of the catalogue should be English, French
or German. If the suggested standard field names are used, a reference
table can be drawn up in any appropriate languages.
Return to Contents
Cataloguing Procedures
Musical instruments are designed to play music and do not always
conveniently fit into the schemes of classification by which
musicologists and curators arrange their information. Instruments can
be classified according to their place, time and culture of origin,
their morphology, their materials of construction, their function
(either at time of manufacture or at the time of their most recent
regular use), etc. Any classification bringing together instruments
which share one characteristic will separate instruments which share
other characteristics.
Where appropriate, the catalogue should be arranged in the order of the
Hornbostel-Sachs classification tables, as modified by the CIMCIM
Working Group on Classification. A supplementary general rule for
ordering the catalogue is that similar instruments (e.g. clarinets of
whatever keywork) should be listed in order of descending nominal pitch.
Items which can be readily separated and used for their proper musical
purpose should be treated as separate entities for cataloguing purposes.
Brass instrument mouthpieces, percussion instrument beaters and bows for
stringed instruments are in practice frequently used for playing various
brass, percussion and stringed instruments: they should be catalogued
separately even when associated by manufacture or long usage with
particular instruments. Bassoon crooks and bagpipe chanters are
somewhat less readily interchanged between instruments, and (together
with ephemeral items such as reeds and strings) can be described in the
catalogue as playing accessories for the instrument they have been
associated with most recently, if any. In all cases, the cataloguer
should state whether the accessories are known to be those originally
supplied by the maker for a particular instrument, whether they are
probably original, or whether they are only possibly original.
Non-playing accessories would not be recorded with this group.
An instrument should be described in the functional state in which it
left the workshop of its last repairer or, if in original condition, its
maker. Any faults, broken or missing parts that have occurred since
then can be listed in a Faults field; any opinions about a former
state can be given in the Repair history field. For example, if
a clarinet was built as a five-key instrument, had a sixth key added
later and has since lost a couple of keys, it would be catalogued as a
six-key clarinet, the loss of two keys being entered as a fault and the
addition of one as repair history.
The system for description of keywork given here is that developed by
John Dick, and propounded by him at the meeting of the Galpin Society in
Edinburgh in 1986. This does not attempt to give an analysis of the
mechanism from the engineering point of view, or to describe the keywork
fully, but rather to indicate the facilities offered to the player.
Where the system of keywork is common and well described in the
literature (e.g. for 8-key flutes or for 5-key clarinets) the fingering
is not spelled out in detail in each case. In the case of more complex
instruments a detailed description in words would be difficult to write
and read: a diagram can be given if within the capability of the
computing processes chosen for producing the catalogue. Cases of
greater interest or complexity should be illustrated photographically.
Here, as elsewhere in the catalogue, the guiding principle is to
describe distinguishing features in more detail while indicating the
presence of common features as concisely as possible. Where an
instrument requires unusual cataloguing procedures, these should be
explained.
Return to Contents
Measurements
There is a wide variety of practice amongst cataloguers with regard to
the publication of measurements. The overall length is traditional in
museum catalogues, and it is of use in the case of instruments which are
illustrated to give an indication of scale. Many of the measurements
sometimes published are likely to be of interest only to an instrument
maker, who would for any practical purpose require more information than
can be given in a catalogue. Where measurements are of value in
providing an indication of playing characteristics, such as drum
diameters, trombone bores, guitar string lengths and violin body sizes,
these are worth giving.
All measurements should be given in international (S.I.) units.
Consequently, physical dimensions will be in millimetres only.
Cataloguers should state the degree of accuracy associated with their
measurements, whether of length, pitch, or any quantity expressed
numerically. If, say, the diameter of a woodwind finger-hole is stated
to be 3.25mm, it should be clear from the introduction or statement
following the figure whether it is within the range `plus or minus
0.01mm' or `plus or minus 0.25mm'. Where appropriate, the equilibrium
temperature and relative humidity at the time of measuring should be
stated.
If measurements in units in use at the time and place of manufacture of
the instrument are thought to be significant, these may follow the data
in S.I. units.
The Helmholtz and other unsatisfactory pitch nomenclatures, although
still widely used, should be rejected in favour of the American National
Standard nomenclature. The note letter is followed by a number (where
possible written as a subscript) denoting the octave. C4 to B4 is the
octave from "middle" C to B in the centre of the treble clef;
higher octaves are denoted by higher numbers, lower by lower. Thus the
A which is commonly 440 Hz is A4. The cataloguer should make clear
wherever necessary whether the actual or the transposed
("written") pitch is meant. For regional instruments where
international pitch nomenclature is inappropriate, pitches can be given
in terms of the ellis (El). The octave number as above (where possible
written as a subscript) is followed by the pitch in cents above the
nearest C below the pitch being described. Thus the pitch of
"middle" C is 4/0 El; 440 Hz is 4/900 El.
Return to Contents
Cataloguing Data Sheet Format
The record for each instrument (or other catalogue entity) is divided
into fields. Some fields are required for the description of
instruments of all classes; some are required only for the description
of instruments in particular classes. For any one instrument, the
information required to enter the data in a particular field may be
unknown, or that field may be inappropriate for describing the
instrument. In these cases, the field is left blank. When the text is
being processed for publication, blank fields can be omitted.
It is unlikely to be necessary to designate fields which would be used
in describing fewer than 10-12 instruments in a Collection. This is
because there is less need for automatic processing, not because lower
standards are required. Thus for types of instrument less well
represented in a Collection, most measurements and other details will be
entered in the Technical Description field.
If too many fields are used, the database software may not cope. The
criteria for recording data in a separate field should be whether the
data needs to be categorised for recall in a database system,
manipulated as a unit in an editing algorism for publication, or
separated for collection management purposes. Acting as a reminder to
record some piece of information is not a sufficient justification for
designating a separate field.
If the volume of data is too large for inclusion, the most important
information should be given, together with an indication that it is
incomplete and a reference to the location of the remainder of the data
outwith the computer record. This may apply, for example, to the
literature references concerning important instruments, or to the
detailed measurements of complex instruments.
Where the information given is only the opinion or speculation of the
cataloguer, it should be followed by a question mark in brackets (?).
The structure of the record for each instruments is as follows:
- Fields used in identifying the instrument
- Fields
describing the instrument
- Fields giving performance characteristics
- Fields giving the history and use of the instrument
- Curatorial
management data
A printed catalogue can be automatically extracted from the data in the
standard format, and prepared for publication if required by `desk-top'
publishing methods. On-line access to the database is also quite
feasible. The same data, without further keyboarding and proofreading
can be used to generate collection management documentation.
Return to Contents
Abbreviations
Abbreviations of the field names have been devised using two or three
characters. Only the fields with two-character abbreviations would be
used for a published catalogue. The abbreviations have been devised to
be as brief as possible while retaining some mnemonic value in English
at least.
Return to Contents
Data Entry
Data should be entered following the most appropriate field name,
leaving a clear line between each field and terminating the entry for
each field with a full stop. Where a field entry runs to several lines,
the text should be grammatically correct, not in `telegraphese'. Parts
of the field entries (such as the fingering) may, however, be in tabular
form.
At the editing stage, it is suggested that the field names given below
in square brackets should be removed, so the text of the entry should be
complete in sense without the presence of the field name. Where the
field name is merely followed by a colon, the name and the colon should
be retained when editing.
Some notes on the standard of data and format of entry for each field
follow: