Contents - Previous - Next
5.4 Intellectual control
Helen P Harrison
Documents That Move And Speak ICA conference session May 2nd 1990. Consists of introduction and conclusion to the session.
Material in an archive or library is subject to two types of control:
1.- stock control - where the material is located - is it on loan to anyone, if so whom, - what format is it available in, physical description etc.
2.- Intellectual control which includes content and physical description to help the user locate the material required in amongst a lot more material. This paper deals with intellectual control.
Intellectual control used to be known as cataloguing and classification especially in the library world. A generic term was required for the control of books and audiovisual materials - bibliographic control suggested that it was only books which needed finding aids. But I do not think the term is here to stay as intellectual control has overtones of `big brother' and forcible control, and it is bound to offend some people some of the time. The concern during this session is the descriptive entry of the content of the material. Sometimes the same record will be capable of containing both stock and content control items, at other times they may be separated, especially if the material is handled by different departments, for example a preservation/conservation or technical record need not be contained within a descriptive content record although even in these circumstances, and especially in a small collection an integrated record may be necessary. Computers may (or may not) be able to encompass both within the same record, and if you do put it all into one record it could put an unnecessary strain on any system. Also you do have to ensure with the use of selective indexing and print-outs that not all the record is produced at one time.
But what is the purpose of the exercise? Is a description of contents of material which is supposed to be seen and heard necessary? The aim of descriptive cataloguing and the retrieval of information is to produce a written or readable record of the contents of a piece of audiovisual material which will reduce the necessity for overviewing/listening or use to determine its contents. It should enable the user to find his way quickly and efficiently to those items which are of particular relevance to him in his search or research. By using a relatively simple information retrieval system the user can decide on a topic, follow it through the catalogue and its subject terms, to find out how many items of relevance exist and then proceed to look at related material - either summaries of contents, scripts, transcripts or sequence lists refining his terms as he goes until only a few relevant items may remain. Without the aid of even an unsophisticated retrieval system the user can spend many hours viewing for the sake of a few minutes relevant material. This may appear a simplistic argument, but it is the one that counts - saving the users time - who may also have to be paying viewing/listening fees as well. The more effective the retrieval system the better in terms of human and financial resources.
Using the system of information retrieval which the archivist or librarian can build up, however minimal these may be, or ravaged by economic expediency it should be possible to lead the potential user to a few relevant items, or even perhaps a few relevant sequences within the archive stock, and reduce the need for him to spend precious hours viewing material which is useless for his purposes. The user has to bring a certain amount of know-how and old fashioned intelligence to bear on his search, but the skilled indexer can help him to pinpoint certain sections and lead him to some of the bits he needs if he follows the signs.
Some of the questions whch may be posed are:
a) The depth of indexing. Should it be by collection, series, title, sequence, shot or frame.
Some discipline is required - but which is the best? Should we have hard and fast rules for control of av, and if so can they be universally applied. A pragmatic approach is needed and systems have to be adapted to suit the conditions and environment in which the documentalist finds himself. This question is also a matter of economics and resources. How much time and effort can any archive afford on producing descriptions to the nth degree, or how much support can be achieved using related material - scripts, transcripts, synopses etc.
What of the differences of presentation, feature films, documentaries, series, serials, newscasts, newsreels or newsclips. All require a different approach and an archive which holds more than one type of material needs to be able to tailor the indexing to the different materials. For example a feature film may only require technical and production details, and credits plus a brief summary of the plot. But of course one has to catalogue the item in hand especially within an archive - archive copies may have bits missing, or additional to the normal distribution copies. All such details have to be included to describe the item in hand. It is important for the researcher to know which copy each archive has - is it the truncated, or worse, censored version; is it the restored version, what is its ratio, is it cinemascope or normal. Is it the television or cinema version. It is also important to include all of these elements in case the archive should contemplate producing a published catalogue or if it intends to include the material in a union catalogue of holdings. At such a point the mere cataloguer has to be considering all such eventualities, if he is not to waste his present time and that of his successors who have to fill the gaps.
b) What are the differences between describing audiovisual documents and textual documents.
Av materials require more description and therefore more descriptive powers from the indexer. A book has many tools to guide and inform the user as to its content: contents list, visible title page, index and general immediate accessibility to the content. Most av material is blind. One may be able to judge from its title, cover or related material. But immediate information is seldom available with audiovisual materials - they are difficult to browse. Such information has to be created or composed by the indexer and this is where the skill come in.
c) What progress has been made towards descriptive standards for the `new media'
There are several standards for descriptive cataloguing already in existence. There are rules from the British Film Institute, National Film Archive, from the Library of Congress, Motion Picture Division, the Aslib Film Production Librarians Rules, Library Association/National Council of Technology (LANCET) and Anglo American Cataloguing Rules (AACR). These last two are primarily for librarians, although the LANCET rules working party had a number of archivists as members. None of these is wholly satisfactory but elements can be taken from several for use in today's archive. Other standard rules for libraries and archives include the more recent ISBD (International Standard Bibliographic Description) rules for different publications, including a set for nonbook materials (ISBD-NBM). The Museums Association in England has a Documentation committee (MDA) which produces guidelines and standards for description, although these are usually for still visuals and museum objects. The International Associations of FIAF, FIAT and IASA have their own ideas for descriptive cataloguing. All ideas are investigated and cooperative ventures are made, but no complete solution has been found.
d) Will computerisation bring library and archive approaches together or only serve to define the differences.
Archives and libraries are different, there should be no argument about that, but computerisation should not have the effect of emphasising this difference. The two types of institution could be interdependent, cooperating with one another for the general good. Also computers are tools - very useful tools, but they should not alter our lives to the extent of rethinking the principles of descriptive cataloguing. Computers have a great deal to do with improving practice, but the principles will usually remain as they were, especially if they were good principles to start with. For example computers can help in the efficient storage and retrieval of information, but the entry has to be produced in an understandable form for publication in `bibliographies'. Hence the descriptive entries should be uniform across as many archives as possible, so that they can be ordered into a standard format for external use. People tend to forget the input time and effort, but no archivist should. He is the one who may well end up doing it!
e) Can material be accessed by users in their own homes? This was a matter for detailed discussion in the session on intellectual control, but the potential and possibilities are there, there is an awful lot of hardware already in place and satellite broadcasting is making it possible with the use of decoders and other devices, to access specific archives, databases, television stations and so on already. The potential is there, it is only the economic resources which may be missing.
f) Are there areas of intellectual control which cannot be addressed by computers or descriptive standards? Undoubtedly. Intellectual control is carried out by human beings - human beings have limitations. Human beings program computers and the database is, in the end, only as good as the people who set it up and who put the data in.
g) How to decide on the number of levels of control and the percentage of holdings at each level?
This is one of the main arguments, and it goes back to point a). Should the archivist attempt to achieve maximum coverage and detailed description of each item or should he concern himself with collections. Cataloguing a collection as an entity may be opting out of some of the responsibility. Collections are not necessarily homogeneous in content - therefore it becomes difficult to describe an entire collection without the assistance of a lot of related material. We should also remember that we are dealing with audiovisual documents which are even less easy to access than written documents and more vulnerable to damage by casual or unnecessarily frequent consultation.
The percentage of holdings which are catalogued and described at each level is a corollary of the previous point. It is important to establish the differences in detail of cataloguing which can be achieved before deciding which collections or materials can be subjected to each level. This also extends into the principles of selection, another problem for the archivist.
The archivist has to balance the ideal with the realistic. This may often be dictated by economics, but economics should not be allowed to dictate an inadequate control. We should be prepared to look for the minimum entries, but make sure that these are in line with the maximum information which can be provided within these parameters. Keywords if handled properly within a controlled thesaurus of terms can solve some of these problems and provide both sides with some sort of a solution.
AV materials have different requirements as far as descriptive entries are concerned and a multi media catalogue is not always helpful. Although av is searched for, first by subject, in which case a subject enquiry is useful, but also by format. People want to see something in moving image form, or in sound form, or in written form for their own purposes. They seldom, except perhaps for research purposes, need to know absolutely everything which has been produced about a particular subject.
The would-be intellectual controller of archive material walks many tightropes. How to balance the effort of bibliographic description of the material in hand against the possibility of future damage and destruction.
Conclusion to session.
This session was a practical exercise in the principles of intellectual control. Any archive has to consider its main functions and the potential use to be made of its stock before embarking on a policy of intellectual control. Although the session was concerned with film cataloguing or cataloguing of moving images, I remind participants that the Symposium is entitled Documents that Move and Speak, many of the principles indicated will apply to sound but the requirements of the two materials are not necessarily identical. It may be even easier to catalogue what you can see. Sound recordings have their own problems of versions, performances etc. and catalogue entries can be even lengthier than those for moving images.
The session panellists came from archives sufficiently different to highlight some of the differences in emphasis on the detail and content of the cataloguing needed to serve the ultimate users of these archives and fulfil the purposes of the archive collections. We have representatives from the National Archives of Canada, a national archive with responsibility for a range of material including broadcast and newsreels; the Imperial War Museum - a National Archive with a specialist subject interest; the Canadian Broadcasting Corporation, a broadcasting archive with special responsibility towards production; and Cinemateca Brasiliera - a cinemathèque as well as an archive.
The extract which we were all given to analyse provided an example of many different types of television: a newscast, a journal, a series, and a documentary. This gave plenty of scope for variation in the cataloguing treatment.
It was an interesting exercise which helped to highlight some of the problems faced by an archive in cataloguing its stock, keeping control over the collections and providing information to the potential user which will assist his search for relevant material within the often vast collections held by archives. It was also interesting in the mix of people in the audience. This was not a case of cataloguers talking to cataloguers, an esoteric exercise if ever there was one, rather it was an exchange of information between the users, archivists and cataloguers, all of whom could learn from each other to help improve the service which intellectual control can offer. In the long run catalogues may be too important to leave entirely to the cataloguers.
This was therefore not just an exercise in cataloguing practice, but it was an exercise and the results need not be held as sacrosanct. We have always to consider the nature and purpose of the archive involved in deciding upon principles of intellectual control. How much detail, what sort of information, who are the end user(s), and what requirements will they have. In many ways the exercise was a false one as some of the archives concerned would almost certainly not have gone into the requested detail for material which was not destined to be a permanent part of their collections - for example neither the Imperial War Museum nor the Cinemateca Brasiliera would normally collect this material, nor would they collect material in a foreign language, nor others would collect material which was not their own copyright material. The CBC is the closest to the real situation here. But all this is not to denigrate the exercise, simply to indicate that there are different levels of cataloguing which should be applied to different situations, always bearing in mind the function and purpose of the archive concerned.
But of what benefit is this activity of intellectual control - the fact that we all do it to a greater or lesser degree does not automatically justify its existence. Any cataloguer has to know why he is cataloguing material, and for whom before he can judge the most useful method to employ. Intellectual control serves several purposes. There are the archive requirements where control is used to determine the content of the collection, and to prevent too frequent access to a waste of archive time, user's time and unnecessary deterioration of the material.
Material is retrieved for research, for use, for entertainment, and for information. The existing expensively produced material is also required for new productions and in new situations. One of the major activities of archives is preservation of material - but for what purpose? So that material can be made available for future use. Exploitation of the material helps to recover some of the cost of preservation, but in order to exploit the materials they need to be accessed. This is where the activity of intellectual control comes in. Making it easier for the potential user to find the material.
Effective intellectual control assists users to locate relevant material quickly and with reasonable accuracy. There is also the reusers requirement to know the provenance and copyright of an item.
There is another requirement - to describe the item in a printed catalogue or on a database holdings of a particular archive.
It has to be said that the activity of cataloguing or information retrieval is very costly of human and financial resources. It can be relatively even more expensive than preservation. It is a very labour intensive activity, and difficult to short cut. It is also a very time consuming activity. Even to produce minimal entries for a piece of material entering an archive it has to be viewed or listened to in real time. Archives should not enter stock which has not been checked to ensure that it is correct and that the archive knows what is in the can or on the tape. Minimal documentation including title, origin, credits, even genre has to be extracted from the material itself, but that may not always be complete and further information may have to be gleaned from related materials such as a cameraman's dope sheets or shotlists, scripts, production scripts, newspaper and other clippings for new material. As we see from these entries it takes at least twice as long as the extract to produce even minimal entries. It can take several hours if any detailed research has to be undertaken and some of the maximum entries produced today will have taken a few hours to produce when one looks at the detail taken from the tape itself, even without further research. Although not everyone cited the time involved in producing the entry for the 16 minute extract the Cinemateca Brasiliera took about 40 minutes to compile the minimum entry and about 2 hours to compile the maximum entry in Portuguese.
Intellectual control is also one of those activities which does not have an immediate dramatic appeal. It is difficult to persuade paymasters or funding agencies that this activity is essential. Only when it is absent is it noticeable. As one observer noted at the session `up front indexing in detail is very costly - but equally costly at the retrieval end if you do not do it in some detail'.
Serendipity may be a useful attribute, as is the memory of an archivist. But if the archivist goes under a bus (and it has happened) or the serendipity is too limited there is trouble. This is one of the major benefits of adequate cataloguing.
Maximum vs Minimum Entry
One of the exercises the panellists were given in preparing the session was to produce both minimum and maximum entries. They were also asked to confess how much of the holding of their own archive was catalogued according to these parameters. The figures quoted range from 90% on minimum level down to NAC 25% at maximum level, or IWM 15% at maximum. Other archives have less than 5% fully catalogued. The detail of the cataloguing will always relate to the function and usage of the library. If we consider this extract I would expect that the broadcasting company, in this case CBC, would produce a full entry at some speed as opposed to a national archive. In the first place the broadcasting company, or newsreel company would need to produce detailed entries as quickly as possible for re-use purposes. In most broadcasting companies the greatest re-use potential of material is within 6 months of the action and detailed cataloguing is required to cope with this. With an archive, the time-scale is less critical and they can afford to wait for maximum cataloguing of news items, but minimum cataloguing is always essential to prevent the problems involved in backlog.
There are further decisions to be made according to the type of archive holding certain materials. How much of the detail of the extract for instance would be needed in the different archives considered - how would they deal with this type of material and why would they deal with it in the way that they do.
Many of the speakers indicated that they were doing things in a certain way out of expediency. There is nothing wrong with that. Each archive must decide on its own priorities in the light of its function, usage and resources. If it works for you, seriously consider using it until someone comes along with a better idea, but use it; do not sit on your hands and expect the job to be done for you - that will only result in the nightmare backlog we all have to contemplate.
Rules and Standards.
But how to achieve this intellectual control. Are there rules and standards which can be applied? There is one short answer to the question whether there are universally accepted standards for archive cataloguing - No!. But that is not to say that there are no standards in existence which cannot be adapted to an archives' needs. Each archive will look at available rules and decide for itself which of these is useful to adopt and /or adapt for its own purposes. Because the rules and standards are fluid or not in place, it is better to do the job according to these standards which will assist archives NOW rather than wait for the ultimate or definitive standard which will never arrive, and leave you instead with a horrendous backlog. The only way to keep pace with avm is to do something as it is produced or acquired.
All of us on this panel are pragmatists. Take what is suitable when you can, when it is available and adapt it to your own situation. Then improve upon it at a later date. This implies that you need a flexible system which can be adapted. There are no hard and fast rules which can be used and this is probably just as well because each archive has its own constituency, its own function and its own needs. This is not to say that anarchy should prevail because this would make for difficulty in international exchange of materials and information.
Of course there are several rules already available and we have seen examples of archives which use AACR2, FIAT and IFLA ISBD-NBM. Wendy White-Hensen's Archival Moving Image Materials - a cataloguing manual based on AACR and emanating from the Library of Congress has also been mentioned. These rules are either specifically for archives - in the case of FIAF film (moving image cataloguing) or for generalist collections in libraries AACR2 and IFLA. I could throw in a few more such as LANCET, which led up to AACR2 but was specifically for non-book materials, FIAT guidelines on data elements for the description of television materials, MAD - Manual of Archival Description, a general archive cataloguing manual which includes moving images, sound and photographs. And of course there are several computer systems available for dealing with control of documents that move and speak. None represents the universal standard, but audiovisual documentalists are cooperating to produce standards and work is being carried out to draw the standards together.
The state of the art of intellectual control is very fluid - it takes to adaptation - not a reinvention of the wheel but not rigid principles either. No one set of standards or rules will apply to all situations, but there are Minimum and Maximum parameters.
One of the archivists Yvette Hackett in NAC mentioned that NAC is working in a certain way now, prepared to adopt new standards, but not doing their own thing or going off at an unproductive tangent. Meanwhile they are maintaining intellectual control over the material and providing access to it now!
Sometimes a conflict is expressed between conventional cataloguing standards and computer retrieval. Both of course have their own mystique, but they are not in opposing camps.
Computer retrieval is surely a tool - if we have to amend cataloguing standards to suit so be it - as long as the end result is:
- comprehensive and produces the detail required
- easy to handle and understand by the person who consults the catalogue
- successful in retrieval
Dotting `i's and crossing `t's and punctuation go out of the window today - until you come to publication of course, where a uniform (not a pedantic) style are required.
Most databases have a standard format which comes in the software package or can be created and adapted by the cataloguer. There is a need to ensure that all the information you will need to retrieve is on that format. If not, amend the record until it is.
The clever bit comes in setting up a content 'description' or further, constructing a retrieval system - thesaurus, free text searching call it what you will.
Any catalogue entry is based on data elements and we have been shown those data elements in some detail. Minimum data elements have been suggested as follows.
National Archives of Canada
- Accession or item number
- Medium of production (original)
- Statement of responsibility (Production Company).
- Country of production
- Production (film) or release/transmission date (tv).
- Subject description
- Subject terms - indexing
- What is material
- What is it about
- Basic characteristics
- How did it come into archives
- By whom
- To whom does it belong
Both of these entries would also contain basic technical information. This last point led to some discussion on the amount of technical information which should be included in the main item record.
In the first session we noted that there are two necessary records for the material in an archive:
- that for technical or stock control and that
- for intellectual control.
We have been dealing with the latter, but it should always be remembered that the other records exist.
Detailed technical records are usually compiled outside the cataloguing area and kept separately. Technical records would swamp an entry and apart from the immediately available format what use is a detailed technical entry on the state of the material, or the restoration processes to the user. All he needs to know is can he get hold of the material to reuse or view, and of course he will not be accessing the original.
A question was raised about copyright during discussion, something which could lead to a whole nest of vipers. It was raised in the context of locating the owner of the material shown in the extract. How much of the detail can be included in a catalogue entry, and indeed how much should be included. It does not always concern the viewer or user as to where the material has come from, but it does concern the person who wishes to re-use the material, and of course the clearances given to the archive for reuse. We may be talking about copyright, or donor rights. Both of which should be logged somewhere but not necessarily in the main item entry.
In addition many of the broadcasting companies of course do not hold (retain) material which is not their own copyright. It is transmitted but not retained. However with newsfilm, current affairs, documentaries the situation becomes very complex, and decisions have to be taken about how much copyright information to include. It may also be very difficult to tell where all the material has come from. The company which produced the programme would be expected to hold detailed records. Normally archives have sales areas which deal with copyright clearances using records other than the cataloguing record. Acquisition files, programme files, donor files etc. Once again sufficient information may not be available to the cataloguer. As long as information is available somewhere in the archive the catalogue record need not show all the detail.
One other aspect was raised in questions, that of provenance. The user of an archive needs to know exactly what he is retrieving and the cataloguer should go to some effort to ensure that cataloguing details are accurate. The user wants to retrieve, locate and use the genuine image - not a reconstruction or fudging of one situation into another. No matter how hard we try there will be errors. The use of a notes field in the entry can be invaluable to show details of where information in the entry came from. This enables a researcher to check detail for himself and it will save the time of future cataloguers.
One of the most important elements in an item entry is the description of content and the system which the cataloguer sets up for the retrieval of the subject matter of the material. Archives dealing with any film other than fiction (which is usually searched for by title) find that the material is searched for primarily by subject matter. We spoke earlier about the level of cataloguing, whether it should be by shot, sequence, item, series or collection, and most of the panellists seem to agree that researchers are interested at item level, and need a description of content at this level.
I admire the ability of a National Archive to produce a detailed synopsis for a relatively short piece of film, but if you consider a newsfilm agency it would probably put even more detail into a shotlist for a 1 minute film than has gone into even the most detailed entry here. Yet the newsfilm which goes out on syndication could be only a minor part of an extract.
It is a matter of scale and considerable discretion on the part of the cataloguer and archiver to achieve a balance between being overwhelmed by the minutiae or getting the message across.
But there are probably always going to be two or more levels of description operating in any archive. A preliminary record may be done at the of accessioning. A more complete record by specialised cataloguers can be done at a later date. It would be a pity to spend resources on detailed documentation of an item which disappears through deterioration. Conservation and preservation should go hand in hand with documentation.
Language. Description is usually done in the language of the document. Some archives such as those in Canada will use bilingual entries, others will use a single language. Of course most national archives will hold material in their native language, or if foreign language material is held it will be subtitled for exhibition purposes. But note the difficulties which Cinemateca Brasiliera had with translation, and the amount of time it took - 5 hours to produce an English description. Scripts should assist here if available.
Lastly we come to indexing. Most items of this nature will be searched for by subject. Each archive will probably have to devise its own indexing terms as those subject headings which exist in published form are primarily for printed materials. Free text indexing and the use of keywords are popular methods of retrieval at present. The terms have to be in natural language however and strictly controlled. Keywords and thesauri are all very well if the ordinary user can understand it. One or two comments from the floor suggested that the keywords chosen by the documentalist did not always achieve successful hits with the researcher. This is always a problem, and suggests that the documentalist and the user should have a closer understanding. And any thesaurus which is used in compiling the index should be made available to the user before he starts his search.
Contents - Previous - Next