Re: Parallel journals

From: Arthur Smith <apsmith_at_APS.ORG>
Date: Tue, 6 Oct 2009 11:26:37 -0400

There is, it seems to me, an unacknowledged non-dysfunctional purpose
served by journal issues that would be difficult to satisfy with any
search interface: non-duplication. Once a researcher has scanned the
table of contents for an issue looking for articles of interest, that
person feels no need to look at those articles again. If those
previously scanned and ignored articles show up repeatedly in a search
query (worse yet, if there are half a dozen instances of the same
article content in the search, the same or slightly different versions
from different sources) that is not only annoying, but a waste of the
researcher's time spent keeping up with the literature.

News readers for RSS feeds of journals handle this non-duplicative
purpose nicely; a search query-based RSS feed could serve the same
purpose but the technical requirements are somewhat different from just
searching. Though perhaps that's what you were implying by "the power of
online boolean search", the need for tracking what has previously been
looked at (whether read or not) needs to be acknowledged.

Arthur Smith

Stevan Harnad wrote:
> On 5-Oct-09, at 8:49 PM, Klaus Graf wrote:
>
> > 2009/10/6 Stevan Harnad <harnad_at_ecs.soton.ac.uk>:
> >
> > > (5) A journal issue is just a hodge-podge of mostly unrelated
> > > articles; no
> > > need to "reconstruct" that; open access to all the articles plus good
> > > boolean search power is all that's needed.
> >
> > I do not think that you can prescribe readers how to browse journal
> > issues. There are lots of theme-issues where browsing makes sense.
>
> Then just retrieve all and only the entire theme-issue with a suitable
> boolean descriptor...
>
> (You vastly underestimate the power of online boolean search over an
> OA inverted full-text corpus -- and you also overestimate the
> persistence of obsolete, dysfunctional habits, once more powerful
> means become available. But the real reason all this prognostication
> is missing the mark is the persistent absence of most of the target OA
> full-text corpus. The latent power is not at all evident from the
> arbitrary, sparse OA subset we have so far. Until the token drops and
> we realize that the only thing standing between us and that full
> corpus is author keystrokes -- and that all that is needed to inspire
> those author keystrokes is Green OA self-archiving mandates from
> universities and funders -- our imaginations will continue to fail,
> and mislead us.)
>
>
Received on Tue Oct 06 2009 - 21:03:27 BST