Physics Roundtable William Armstrong Notes: Emily Poworoznek and Jennifer Edelman The roundtable consisted of speakers and an audience/participant discussion centered around pre-selected topics. The room was full, and participation proved to be quite active. After the speakers were introduced, brief self-introductions by each of the audience members followed. I. Speakers Douglas LaFrenier, American Institute of Physics Bob Michaelson II. Audience/Participants’ Discussion of Selected Topics Some of the specific comments made by audience participants are included below: Discussion Topic 2: Permanence (or lack thereof) and preservation of electronic archives of physics documents (or electronic documents in general) Another issue is the affordability of the archives; should permanence be dependent on continuous payment? Should the author or subscriber be paying? The following remarks and questions addressed these issues: Discussion Topic 3: Landolt-Börnstein Time for the roundtable ran out quickly, and the discussion was brought to a close by the moderator, who thanked the speakers and audience for their participation.
notwwa@lsu.edu
Jerry Cowhig, Institute of Physics Publishing
Cowhig provided updates on improved access to IOPP publications, the schedule for the full journal archive, new journals and services, and editorial initiatives which include adding physicists to staff and making connections with physicists on an international basis. He pointed to trends: for example, article downloads per year increased by a factor of ten from 1998 to 2002, and it's expected that 2002 article downloads will reach 2.7 million. All journal articles are now freely accessible for the first 30 days after publication (used to be free until next issue). Non-subscriber downloads are currently about 10% of the total, but subscriptions have begun to rise since 2000.
The journal archive has been extended to 1968 and by December will go back to 1874, with full-text searching and reference links in and out. Access during 2002 is free; paid options thereafter include an annual fee (2003: $550, discounted for major package subscribers). The archive will also be available for local loading at a one-time cost. Cowhig explained that the annual fee, while low, is important in principle due to the heavy responsibility of maintaining an archive. The current subscription will include a rolling 10 years of access, while the archive covers from v.1 to 10 years prior to the present, and includes all titles (whether currently subscribed or not).
The IOPP now publishes online JHEP (Journal of High Energy Physics) on behalf of SISSA, and Nuclear Fusion for the IAEA, both still maintained at their original hosts, as well as two Chinese journals. JHEP access is free after two years. Cowhig said that JHEP will be added to Packages B and Z or be made available singly for $900 per year. Nuclear Fusion will be included in Package Z.
Electronic journal services now include Vivisimo which clusters search results; a special Web info link for librarians; and is working toward ADA compliance.
LaFrenier discussed the AIP/APS Virtual Journal program. Virtual journals focusing on cutting-edge research were developed to slice through the content of big journals, providing specialized content and current awareness (TOC alerts) without creating new niche journals. He announced the fifth virtual journal produced by the program: Virtual Journal of Ultrafast Science, edited by Philip Bucksbaum. While this will be the first VJ with Optical Society of America journals, it is hoped other VJs will also be permitted to carry OSA journals. LaFrenier explained that the selection process includes computerized article discovery plus editorial review; editors may also add articles. Non-AIP/APS journals covered include Science, Nature, SIAM J. on Computing, and OSA titles.
As to the success of the VJs, LaFrenier went on to say they are not profitable, but authors love the increased distribution, and there is growth in TOC alert registrations (due to word-of-mouth). Lack of marketing limits corporate growth. Credit card purchases are available for those who don't subscribe to the original journals online, and a deposit account for corporations and other subscription models are being considered. It's not clear that the VJs are slowing down the start-up of new niche journals, but more subscribing institutions are linking to the VJs. LaFrenier added that there are no digital object identifiers (DOIs) associated with these journals, and citations of articles found in VJs should be to the original publication, not the VJ. For more info on this program, see the web site or contact Doug LaFrenier or Barbara Hicks (APS).
Michaelson described the PAM Physics Translation Project that he and Dana Roth (Caltech) conceived in response to PAMnet queries for article translations. Please see his session at the PAM-Wide Roundtable for basic information.
Questions included the idea of providing DOI links (which aren’t currently available) and providing indexing for early years of translation journals. The translation journals won't be covered, but perhaps a list of cover-to-cover translation journals could be added eventually. Regarding funding for the project, server use has been offered permanently and time is being volunteered. The project has been undertaken under PAM auspices, using volunteer PAM labor. However, use of the PAM logo on the web interface will have to be authorized by SLA. As mentioned earlier, the translation database will be a freely available web service. For cost recovery, AIP has offered some funding if needed.
Discussion Topic 1: Where are our physicists getting their information?
What sources are they using? What should they be using, or is there even a consensus? Are they, in many instances, relying exclusively on ArXives, pre-print servers? Is this adequate? What long-term effects might such reliance on arXiv.org have on physics libraries, e.g. subscription to source materials as well as provision of indexes for locating such? Are we canceling journals due to electronic access to sources such as lanl.arXiv.org? What of more traditional sources of bibliographic information, e.g. INSPEC? Should we be promoting this more? What are institutions doing that can't afford INSPEC?
Some physics communities appear to use e-journals heavily. High-energy physicists are definitely using arXiv more, while string theorists probably use arXiv almost exclusively. Nuclear physicist use of the SLAC (Stanford Linear Accelerator Center) database preprints is also very high. Solid-state/condensed-matter and acoustics folks still use the library. Until the peer review process changes, journals will maintain their place; arXiv does not provide that process. Those who don't use arXiv often use Web of Science first, rather than INSPEC, the latter being a "hard sell" despite its usefulness. It simply does not seem to fit the way people work now.
When the publishers at the roundtable were asked their views on maintaining print, IOPP responded that it will continue print if the customers want it. AIP maintained it has discounted online-only by 25% for 3-4 years, but only 3% of subscriptions are online-only. They would like, however, to get rid of print. A representative from Springer added that despite the value-added online and two major electronic-only deals, only 8% of their print subscriptions migrated to e-only. In some countries, a physical purchase is needed for the contract. It seems that it will be several years before any large publisher gets rid of print.
We had INSPEC on CD-ROM; then, through a consortium, got INSPEC through the Web. We’re promoting it, but physics faculty don’t want to be bothered.
For many physicists, increasing their use of INSPEC is a hard sell. Other things have a better fit with the dynamics in their discipline. Unless INSPEC adopts a more modern search engine and becomes easier to use, they won’t use it.
The physicists are not relying solely on e-print archives, but what they are doing is invisible to librarians.
Certain fields have a low usage in ScienceDirect journals; they’re using arXiv instead.
We’re paying $20,000 to get Nuclear Physics electronically from Elsevier. People still cite the preprint from SLAC even after the final version has been published, but the faculty don’t want to cancel the subscription!
People go to the Web of Science before INSPEC, because it’s interdisciplinary and links out.
It depends on the discipline: solid state physics and acoustics faculty use the library more than others. We lack data on usage.
If we change to e-only, no one complains. Is this because no one noticed? Is no one using the library?
LANL model and "official" models coexist.
I know one person who would look at the table of contents in the print copy, then download the article from arXiv.
Until someone develops a new model with preprint peer review, journals are the only opportunity for peer review.
Physicists don’t use current awareness searches.
Depending on the institution: undergraduate and graduate students use journals; undergraduates don’t like paper; if it’s not online, undergraduates won’t touch it; older faculty stick with print.
Librarian concerns: Electronic versions should not be an exact copy of the print, or why bother? Access to archives.
(Related to topic 1) – This is an issue of particular importance as it is likely a large percentage of our physicists are relying on electronic documents from pre-print servers. As more and more access becomes strictly electronic, how do we guarantee survival of the electronic data? It’s not just the storage and retrieval formats which are at issue, but the very nature of electronic data itself. Are we in a "golden age" of information which may be as short-lived as a few decades, only to be followed by a dark age, marked by ever growing gaps in information? What is being done or can be suggested to help ensure electronic document (file) integrity and endurance?
There was the general feeling that publishers’ online journal archives are still laden with issues of technology and trust. There is a need of a guarantee to readers and also to authors. Indeed, one member remarked that at an ACM talk 4-5 years ago, there was a suggestion to get rid of online articles that no one has read in the last 10 years. IOP thinks it would never get rid of an article that hasn’t been read. It has a commitment to authors. It has multiple copies of backfiles and a permanent depository, using the "LOCKSS" (Lots of Copies Keep Stuff Safe) policy. APS thinks it is participating as an observer in LOCKSS, but not using it.
Another member asserted that we’re not saving cross-references or functionality in a reliable way, eliciting the comment that preservation is out of librarians’ hands. We have to rely on the publishers. To this, APS responded that it has a deposit at the Library of Congress and a mirror at Cornell; that gives us more trust in the permanence of its archives. Someone else then reminded folks that arXiv is now at Cornell – a library has inherited it, thus providing at least one exception to the assertion that preservation is out of our hands.
Can we continue to afford access for our patrons? With the paper model, we always have access to a journal once we’ve bought it. We would cancel print if we knew we had permanent access/ ownership.
IOP: We can sell you CD-ROMs, but other than that, we can’t promise permanent access.
In the past, we paid for journals, ended the subscriptions, and still had copies on the shelf. Now we pay annually and have the choice to keep paying or lose it. Maybe the author should pay.
Authors paying would never work as a commercial model. It’s not viable; there’s not enough money from the authors to cover the cost of the journal.
How many people are canceling commercial print, not society print? (A quick poll on print cancellations showed more people canceling commercial print subscriptions than society ones – possibly the trust factor?)
We trust societies more.
As the discussion continued, it was asserted that information will have an outlet; scientists will find a way to talk to each other. But this brought up the question as to whether they would be able to continue to talk to each other over the centuries, given the potentially fragile nature of digital information. Someone else pointed out that paper gets lost over centuries as well. It was then questioned whether older information is even used; some has been lost. Others responded quite emphatically that it is indeed used and is quite important.
In recent years, many libraries have had to cancel this series because of high costs and probably low usage. With the advent of the online version, is it time for libraries to re-visit this issue? How well is the new online version working? Will the convenience and ease of access and retrieval of information through the online version make this a more attractive resource? Or is the locating and retrieval of information via the online version efficient and easy enough?
A quick poll of the room showed only six print subscriptions to the Landolt-Börnstein Handbook series. Most have cancelled or didn't ever have it. Dissatisfaction was expressed with the online version, with a general consensus being brought out in the following remarks:
The online version needs to be searchable and linkable by and to other information sources, and searching needs improvement overall. The chapters need to be “unbundled.” Right now, one still needs to understand the structure of the print publication in order to use the online efficiently. The index is hard to use; the most use is from people who have a citation to it, not people who use it as the source of information. Americans, by and large, seem to be ignorant of the set, either of its existence or how to use it, while German professors are more familiar with the series.
