Commentary on
Vol. XXIII, No. 1
Last Updated 6 May 2010
|
Commentary on |
|
Last Updated 6 May 2010 |
The bottom line from my perspective is that when in doubt authorship should be attributed and clear citation should be possible, regardless of the nature of the document (electronic text, database, GIS layer, image, whatever). One ought to be able to cite, say, a database with clear attribution of authorship and a clear and easy way for someone to find the version cited. (A URL and a date do not suffice for this purpose.) That means that in the case of living documents and databases that might be in a nearly continuous state of growth or editorial modification the author(s) will have to freeze them periodically and give them version names, much as books may have multiple editions. What's so hard about that? Keith Kintigh and Frank McManamon are figuring out how to do this with things archived by Digital Antiquity. Seems to me that the worst that can happen is that stuff will get ignored and thus not cited, but that is (or should be) the case with a large fraction of traditional publications already. Anyway, this is a big issue, I think, but one that has some clear solutions that will be a boon to researchers.
I fully agree with Dean Snow's comments. I want to emphasize even more the time element - when reading most pages we have no idea about the validity of their content and its changes. One solution was proposed by Dean Snow, but I am rather pessimistic because it is time-consuming, needs staff responsible for editorial work, and slows down the content evolution. It would be good for larger teams and in case of major changes. Another possible solution is a sub-page with a chronicle of all changes. That is cheaper and easier but less convenient from the users' point of view. We tried this in our handbook of Aegean archaeology and our students find it useful, especially when facing exams.
Every individual record in OpenContext has a citation. (This has been a priority since the project's inception in late 2006.) These citations are all expressed for use with our underlying software and Zotero, an open source tool for managing bibliographies and references. At the project level, the citation information was in the project's coding but hadn't been "turned on" for humans to read. This was an oversight and we have now turned on the citation information for the project overviews.
[Editor's note: the article has been updated to reflect the OpenContext change.]
There has been more interest lately, in our field of virtual heritage, in the issue of transparency. That is, in guiding the user/viewer through the 3D model and VR creation process so that scholars, students, whoever, will understand how the evidence links to the final 3D models and virtual worlds. Further, there is more tendency now to show what areas in the models are conjecture and what are based directly on the evidence. These are not easy tasks, but guidelines, like the London Charter (www.londoncharter.org) and potentially the upcoming Seville Charter (which will expand on our work in London), will offer more concrete procedures.
Our web site tries to attribute each page and sections of description within each page to rightful authors, whether individuals or companies. We believe that these and other citation and attribution formalities are necessary in scholarship - for the reasons outlined in the article and simply because without them the very nature and fabric of "scholarship" fails.
Websites like the new Unvarnished.com and the innumerable other sites where anonymous or aliased users submit comments have their place (and there have been many pshychological studies on how people rate things and how and why people will say certain things online). What has always set scholarship apart has been the name trail, because history matters and a person's (or web site's) professional credentials are part of the critical analysis process.
Trust is important in writing, photographs, 3D models, and virtual worlds - and web sites.
I am pleased that those who have thus far taken the trouble to comment on this issue (including Arizona State University Professor Michael E. Smith on his blog at publishingarchaeology.blogspot.com/) have agreed that scholarly material on the web needs attribution, but I did not expect dispute there. I did expect some dispute as to what should be considered scholarly material, and it seems that the respondents have agreed that databases, CAD models, GIS data, and other non-text resources should be attributed. What I have not yet seen -- and was hoping to have -- is any support for leaving some materials without attribution. Since it seems to me that there are kinds of web pages that do not need attribution, I was hoping that there would be some discussion of where we should draw the line between that which requires attribution and that which does not.
Absent suggestions, let me therefore make my own suggestions for that line on attribution to provide a target for respondents.
1. Anything containing data that might be used by another scholar (a database, a model, a text item, a GIS data set, whatever) should have an author or responsible person and date.
2. Any interpretive material should have a name and date if the interpretation(s) arise from the author's own study and are not merely repetition of the work of others.
3. Materials that merely summarize the work of others for the sake of background -- operating something like a Wikipedia article -- surely need no author. I do not think that should be the end of the story, however; I am inclined to think that some indication of the nature of the piece is required to let the reader know what the absence of an author's name means. A general summary may or may not read as if it were only that to a reader, particularly a non-professional one, and it seems to me that the reader should know that the absence of attribution is not simply accidental or a result of sloppiness.
Table of Contents for all CSA Newsletter issues on the
Web