Content-Type: text/html **This message is being cross-posted to other discussion lists. Please excuse any duplication.** The problems of citing and referring to electronic resources of different kinds have been elucidated in many contexts. As I think every normative prescription as to how it should be done reflects a decision which is dependent of some scientific view, it is most important to take into account decisions already made in Library and Information Science and related disciplines. So in our efforts to elaborate recommendations on this matter, we would be most grateful if anyone could give us some guidance and possibly pointers to some existing recommendations on the web. The problems do not restrict themselves to concerns about how to represent resources in the list of references. There are also quite a few other points of discussion, as we see it. 1. Imagine an electronic journal with a particular file structure and with a highly persistent URL for its home-page. The researcher wants to cite a particular sentence in one article, which consists of five sequential files. At this point we have at least three relevant identifiers. It seems natural to choose the identifier for that particular file which carries the sentence in question. Though, as files structures may change and the article has another vague but persistent identifier depending on its time of publication in the journal, say No. 2 1998, maybe it would be more useful to give the identifier (URL) for the journal’s home-page. Even more, the cited sentence located in file 3 of 5, has a context for which the identifier of the index or title-page would be useful. As not everybody may be capable of navigating back to that point, be it of unfamiliarity with the technology or of the absent navigational aid, that identifier may be the best to give in the bibliographic reference. (This problem - as you may know - have been dealt with in the Dublin Core effort by introduction of the RELATION-element) 2. Another question of importance may be extracted from the fact that the methods for preservation of digital material still are not able to guarantee that objects on the web will be accessible in the near future. Thus a text consisting of references to a large amount of web-objects in a few years may lose some of its value and usability. It would be wise then to give as much data about the object in the list of reference as possible, and to find some way to guarantee the persistency of the object, possibly by saving it in some way and let it accompany the printed text. But this copying may conflict with copyright legislation. In this regard we have a few questions on the actual practice around the world: I. I know that for example the MLA Style of Citation and the ISO 690-2 standard gives recommendations on how to make references, but I can not find any clear guidance on problem no. 1 mentioned above. Especially when it comes to the question on how to cite highly dynamic objects. So what is recommended practice? II. The problem no. 2 above is a difficult one, but in the context of works like master theses and doctoral theses there must be some method in practice on how to secure the possibility to check citations and the like. So what is the practice here? III. The efforts of ISO, MLA, APA and the like to produce standards are important, but to what extent are their efforts taken into account by different scientific disciplines? Please respond to the list or to: Mikael Gunnarsson Swedish School of Library and Information Science 501 90 Boras Sweden tel: +46 33 16 41 38 fax: +46 66 16 40 05 e-mail: [log in to unmask]