2. The bibliography contains some 1,100 items used for teaching and research, and is expanding fairly rapidly. all anthropologists have access to consult it or to cite items from it; and they can add to it indirectly (precautions against corruption and accidental destruction seem necessary). They can also create their own bibliographies wither by copying items from the main one, or by entering their own items: this has the advantage that consultation is quicker with a small bibliography than with the large one.
The essence is that full bibliographic details are stored in a UNIX file - at Kent it is called makhzan. Each item consists of up to nineteen subsections which contain specific information (e.g. Author's names; date of publication; title; name of journal; volume). These are permanent records, and they can be manipulated to provide different kinds of output. In reading lists for first year students, for example, we include the library class-mark, but for an article in a journal that information is omitted. The format can also be varied, to give different flavours of citation: Man requires ³Ellen, R.F., while American Anthropologist prefers ³ELLEN, Roy F.².
The programs allow you to stipulate the format in which citations are made, in texts which are then processed using nroff or troff programs. You can re-cast an article rejected by Man into the style required by American Anthropologists by altering one word in the command you give to the computer.
3. Bib and refer have different capabilities: refer has an excellent semi-automatic program for entering items, but the more complex format adjustments are made with bib. We have worked out what we think is a sensible combination of these programs.
We have found that neither always lives up to its description in the official documentation and, with the help of our user-friendly computing laboratory, have been able to modify them to make them do what they should.
Nor do they cope well, in the form in which they are issued, with certain kinds of item (edited volumes of essays; books with volumes with different titles) which crop up in social anthropology apparently more than they do in natural sciences. We have modified them to so that they do manage these, without garbling them.
We have also exploited the flexibility of the bib programs so that it is simple to produce the following kinds of document:
i)
Reading lists, with full citations in Man plus library class-marks,
and other comments.
ii)
articles and books with brief citations in the text and a consolidated
list of references at the end in Man style.
iii)
articles and books with brief citations in the text and a consolidated
list of references at the end in American Anthropologist.
iv)
articles and books with citations in footnotes at the bottom of the
page, in Man.
We now know how to make the instructions fairly quickly, and could produce formats for Penguin or Duckworth to order.
4. None of this is any use to you unless you have access to a computer using the UNIX operating system. But if you do, and if you are interested in exploiting bib and refer, you may find our experience useful. They require modification before they will work properly; they require combination before they will do what anthropologists need; and they require some expertise to exploit the in-built flexibility to produce suitable style of citation. It is in fact quite interesting to acquire the expertise; but time is short, you may not want to duplicate effort, and the inadequacies of the original programs are very frustrating indeed.
We have produced a description of the programs and of how to use them which we think is accurate, unambiguous, intelligible to computer novices and is anthropology-oriented (i.e. the examples are anthropological). Some desirable features (e.g. an easy quick editor, an automatic filter to prevent duplicate entries) have yet to be written, but the system is workable as it stands.
If you are interested in any of this, write to John Davis. It might be sensible first to look at the documentation, for although it does have some sections of purely local use (during the present transitional phase at Kent) it explains the general possibilities of the programs.