[ The following text is in the "utf-8" character set. ] [ Your display is set for the "iso-8859-1" character set. ] [ Some characters may be displayed incorrectly. ]
(2) On Fri, 25 Apr 2008, Jean-Claude Guedon wrote: > The work is a content, not an instantiation of same in a particular form. > What is protected by copyright is a form of expression that has taken > material form, not a particular material instantiation of the mode of > expression. But we are not talking about copyright. We are talking about whether or not page numbers are needed in order to quote passages from a self-archived OA version that lacks the page numbers of the publisher's toll-access PDF. Indeed, we are not talking about copyright, but I used copyright to help demonstrate the reality of the distinction between content and material form. > Likewise, peer review deals with a content, a form of expression, not a > particular material form of instantiation. As a result, peer review is > not tied to a particular material form... [T]he publisher gives it a > particular material form, be it print or bits. All true, but not at all relevant to what we are discussing: [snip] Idem: the reference to peer review was to show that it was addressed to content, not to a particular material form of the content. > Citation, on the other hand, is tied to a particular material because > citation is aimed at helping retrieval. The cited work is the published work. We are simply talking about where in the cited work the *quoted* (or noted) passages occur. In the absence of the pagination, specifying the section heading and paragraph number is just fine. The cited work is the published work. What does "publish" mean? Until now, the only possible publication of the work was the printing of the article within a journal. Nowadays, the work can be published in a variety of ways. As a result, I believe we need to be able to cite any one of these publication forms so long as we have some guarantee that they are equivalent in content. > If we cite down to the page level in many disciplines, it is to help > check that the citation is indeed correct, that the text cited is not > cited out of context, etc. It is fundamental for the work of scientists > and scholars. That's fine. The section heading and paragraph number accomplishes exactly the same ends. As Jean Kempf has rightly noted, this would not be accepted in many disciplines. However, Stevan's point is right here. If the version in the IR is citable, nothing should prevent from pinpointing the location of a citation within a text by using paragraph numbers. Let us only hope the author is not Proust... :-) > Because of its retrieval function, citation has provided an important > handle for publishers to make their material version of a text the > "canonical" version. However, if the IR material version of the same is > declared citable by declaring that it is a certified copy of the peer > reviewed "work", then it can be cited directly. The certification comes > from the good name of the hosting institution - presumably a university > or reputable research centre. The author's refereed final draft can be both cited and quoted directly whether or not it has been "certified" (or "validated": http://valrec.eprints.org/ ) as being verbatim identical with the published version. (Let us suppose, for the sake of argument, that it has indeed been checked and so certified.) The issue was: having been "certified" as verbatim-identical with the publisher's version *except for the pagination* what do we cite and quote, and how? Answer: We continue to cite the published version, as always. And we pinpoint quoted passages by section heading and paragraph number. We could, as Jean-Claude suggests, instead pinpoint quoted passages based on the divergent pagination of the self-archived version in the IR, but it seems to me that that would not be a very useful practice (e.g., for readers of the citing article who have access to the journal version: it confuses them, needlessly). But nothing much hangs on this, one way or the other. It seems to me a very minor, even trivial, contingency. The point of my intervention, of course, is to give make possible citing the IR version and give it the full status of a "published version". > In line with this analysis, many of us have begun providing the on-line > address of the documents we cite when they are in open access so as to > facilitate the retrieval needs of our colleagues. One small, extra step, > is needed to complete the process: just declare the IR version to be an > acceptable source for a citation. Of course all citations (of the published journal article!) should be accompanied by the URL of the OA version (so those who can't afford access to the publisher's version can still access the article). Everyone who provides or uses OA content should do that. Hence my remark that the IR version should have equal status with the publisher's version. From that point on, the IR version is more convenient to use for everybody (except those who do not have access to the internet or a very poor access to it). Those who can't afford the publisher's version can not only access the OA version, but also cite that version as a good, published, version. [snip] No one using OA articles has felt the need for the IR to "declare the IR version to be an acceptable version." (The problem so far has been to get the author to provide an IR version at all!) But this "declaration" is possible (e.g., via http://valrec.eprints.org/ ). [snip] I am afraid I have, and for quite a while now. [snip] Jean-Claude Guédon Université de Montréal