Re: SHERPA will take over the Romeo Publisher Policy Table

From: Stevan Harnad <harnad_at_ecs.soton.ac.uk>
Date: Fri, 2 Apr 2004 16:19:58 +0100

There have been some valuable improvements to the Romeo directory of journal
policies on author self-archiving in the new SHERPA/Romeo version
http://www.sherpa.ac.uk/romeo.php
but unfortunately some valuable features of the original Romeo version
http://www.lboro.ac.uk/departments/ls/disresearch/romeo/Romeo%20Publisher%20Policies.htm
have also been lost in the new version, particularly the individual journal
information and summary.

Fortunately, SHERPA have kindly agreed to allow Southampton to download their
database and present and maintain the all-important summary information
on the number and growth-rate of green journals and publishers. This will
be combined with the corresponding information on the number and
growth-rate of gold journals and publishers from the Directory of Open
Access Journals (DOAJ)
http://www.doaj.org/
which DOAJ have also kindly agreed to let us download.

This summary growth chart of OA will be completed by data kindly
provided by OAIster
http://oaister.umdl.umich.edu/o/oaister/
on the number and growth-rate of OA Archives (both green institutional
archives and gold journal archives), as well as the number of records they
contain.

Users -- who will include (1) prospective OA-providing authors, (2)
prospective OA-providing institutions, (3) prospective OA-archive-maintaining
institutional libraries, (4) prospective OA-mandating research-funders, and
(5) journals and publishers contemplating becoming either green or gold --
will be able to see at a glance where things stand in OA provision along
the green and golden roads.

This information will also be supplemented by growth charts for the
contents of institutional archives worldwide
http://archives.eprints.org/eprints.php?page=all
as well as a list of institutions that have adopted OA provision policies
http://www.eprints.org/signup/sign.php
and descriptions of those policies:
http://www.eprints.org/signup/fulllist.php

On Fri, 2 Apr 2004, Bill Hubbard wrote:

> * The primary user-group of the list are those people who are
> archiving material and want to know details of the copyright transfer
> required by a publisher for the purposes of archiving an article.

Correct, But they will want to know it (1) not just on a publisher basis,
but on a *journal* basis: They may know the name of the journal but not
the publisher; they may, as Arun has suggested, want to scan the journal
listings, not the publisher listings; and they may want to know the overall
numbers of green journals and publishers. The overall figures will not only
encourage authors to join the growing number of self-archivers, but it will also
encourage journals and publishers to join the growing number of green journals.

> * The list attempts to clarify what conditions and rights form part
> of a publisher's CTA and so part of a contract between author and
> publisher.

And it lists some very important details (such as whether or not the author may
use the publisher's PDF; whether or not the self-archived version needs to contain
a cross-link to the publisher's own website). These valuable and substantive
details are provided following an asterisk.

> * Changed from a scrolling list to a database solution.

Much better! And also retains the scrolling option with "view all publishers"
option.

What is missing, though, is the at least equally important "view all journals"
option that Arun suggested! This could easily be gleaned from Ulrich's
http://www.ulrichsweb.com/ulrichsweb/analysis/
by pulling out all the journals for each publisher, and then just cloning the
information SHERPA/Romeo already provides for each publisher, listing it also
for each individual journal!

This would be a *huge* increment in functionality for very little individual
effort.

> * Created the ability to search for a publisher.

Excellent. But now needed per journal too.

> * A link to the publisher's home page has been provided from their
> name.

This could be done for the journals too.

> * There is an on-line suggestion form for users to contribute
> corrections, updates and changes.

Excellent!

> * There is a similar on-line form for users to suggest new
> publishers.

Excellent!

> * The award and use of colours has been maintained, as colours have
> proved to be a useful "shorthand" in writing and in conversation about
> the general type of rights assigned or retained with the CTA of a
> publisher.

Here I think there is a way to make the colors much more transparent, heuristic
and useful. In the original Romeo list, the colours evolved adventitiously, but
they still had not been optimised or etched in stone (or perma-ink!):

In the original (scrolled) Romeo, the colours did a double duty because
their two parameters yes/no preprint self-archiving and yes/no postprint
self-archiving did not have their own separate lookups or entries in the scrolled
list. But in the SHERPA version there is no need for the colours to do that
double-duty, as the parameters are listed, explicitly and separately for each
publisher (journal)! It serves no purpose to have each combination of the
parameters now also get its own colour code!

That is not the purpose of the colour code. The colour code has *exactly* the same
purpose as the summary table: It answers the two critical questions about
publisher self-archiving policy:

(1) Has this publisher (journal) given the green light to self-archive the
published postprint? If yes, the code should be BRIGHT GREEN.

(2) Has this publisher (journal) given the green light to self-archive ONLY
the unrefereed preprint? If yes, the code is PALE GREEN.

If the publisher has not given the green light to self-archiving *either*
the preprint or the postprint, the code should be GRAY. (White is fine too,
but not such a good colour for display if the background is white!)

But note that there is no need *whatsoever* for colour coding anything other than
these three categories. Those are the categories we want to know about at a
glance. If we want details, the individual listings indicate whether a publisher
(journal) is Pre+/Post+ or just Pre-/Post+ That distinction is of no
overall interest, redundant with what is already there under Pre and Post,
and should not spawn two more colours to cause confusion and obscure the real
message of the (2.5) colours (GRAY vs PALE/BRIGHT GREEN), namely, whether a
publisher has given the green light to the postprint self-archiving,
or only to preprint self-archiving.

A little reflection will show that that is the what the colour code is
for; the rest is just detail for the individual listings themselves. And
those are also the categories that need to appear in the summary table and
growth chart. None of the other colours are of any use or interest. They
merely distract and confuse.

> * The original RoMEO colour-categories have been amended. The
> original list categorised a publisher as "blue" if pre-print archiving
> rights *or* if post-print archiving rights were permitted.

The Romeo colours were improvised on the fly. In retrospect, it was a
mistake to code Pre+ Post- as "blue": It should have been pale-green!
That way the eyeball sums up all the greens (of either shade), in order
to visualize immediately how many publishers have given the green light
to Pre or Post or both, and a growth chart can graph it with pale-green on
top of bright-green, displaying the relevant partials and totals.

> We have removed this ambiguity by categorising pre-print "friendly"
> publishers as "yellow" and post-print "friendly" publishers as "blue".

"Friendly" is a nice and apt metaphor but "giving the green light"
to self-archiving (Pre- pale-green, Post- bright-green) reflects the
face-validity of the code, and is intuitively obvious.

It has to be stressed that the distinction between the Post+/Pre+ and
Post+/Pre- category is of *no overall interest*! There is no point
coding this distinction or total. No one cares. I want to know how
many publishers give the green light to my self-archiving my article
-- that's the postprint, to a first approximation, and that's the
bright-green code. If that number is N, I don't care, or want to know,
how many of those N do or don't *also* give their green light to my
*also* self-archiving the preprint! If I want to know that detail, I
can look at the individual item in the SHERPA list, where that detail is
clearly given. (I can always self-archive the preprint without a green
light anyway!)

On the other hand, I *do* want to know about publishers who do *not*
yet give their official green light to postprint self-archiving, but *do*
give it to preprint self-archiving: That's the pale-green code!

I strongly urge the SHERPA group to rationalise their colour code, and lose all
the colours except bright-green, pale-green and gray (or white)!

> A publisher allowing both pre- and post-print archiving rights to be
> retained or assigned is is then made "green", as in the original RoMEO
> listing.

Publishers who give their green light to postprint self-archiving OR to both
postprint and preprint self-archiving should be BRIGHT-GREEN.

Those giving the green light to *only* preprint self-archiving should be
PALE-GREEN.

The details should be left to the lookups, not coded into a bigger and more
arbitrary and confusing array of colours.

> * There are four main possible publisher policies and these are all
> materially different in terms of what can or cannot be self-archived.
>
> It is important we distinguish between each of them.

They are *already* fully distinguished in the individual listings by
Pre: +/-
Post +/-

They do not, in addition, all require a colour code! This is about
publishers giving their official GREEN lights to self-archiving.
There are no yellow or blue lights!

> The publisher might allow posting of pre-prints only - "yellow"

PALE-GREEN

> The publisher might allow posting of post-prints only - "blue"

BRIGHT-GREEN

> The publisher might allow posting of both pre-and post-prints - "green"

BRIGHT-GREEN

> The publisher might allow none of the above - "white"

(preferably gray)

> * Clicking a colour in the colour-key shows the list of publishers in
> that category - all green publishers, or all white publishers, etc

This feature would be preserved, without the needless and confusing profusion of
arbitrary colours, if the colour array simply consisted of colours plus these
words:

(GRAY) Pre-/Post-
(PALE-GREEN) Pre+/Post-
(BRIGHT-GREEN) Pre-/Post+
(BRIGHT-GREEN) Pre+/Post+

All the functionality would be there, all the distinctions, all the options -- but
no superfluous colours, and the colours would wear their intuitive meanings on
their sleeves.

> * Ticks, crosses and question marks have been used to give a clear
> summary on the rights granted as part of each publisher's CTA.

Here too, I think that the new SHERPA list has taken on a needlessly negative and
restrictive look: The green ticks are fine (for the green lights).
But why the shrill RED CROSSES and "CANNOTs!

Again, the exact same information can be conveyed without the shrillness and
negativity as follows:

Pre+: (written in pale green) Publisher (Journal) has given the
      green light to preprint self-archiving

Pre-: Blank

Post+: (written in bright green) Publisher (Journal) has given the
       green light to postprint self-archiving

Post-: Blank

Please look at the listing http://www.sherpa.ac.uk/romeo.php?all=yes
and imagine it coded instead as above, and ask yourself which format
is more likely to invite a positive response from either an author
contemplating self-archiving or a publisher contemplating going green!

And bear in mind that the information conveyed by the two formats
is *exactly the same*!

> * Conditions have been separated into different lines for clarity and
> categorised into "standard" conditions where possible.

Excellent!

(But I would suggest losing the nonsensical distinctions between personal
and institutional websites! They are contained in the publisher's policy
statement, also linked. There is no need for the Romeo list to become
a megaphone for amplifying noise! Only list the substantive conditions.)

> * Note that the conditions given reflect those laid out in each
> publisher's CTA. Some of these conditions are ambiguous or unclear.
>
> However, since these form part of a contract between the author and the
> publisher, these are listed as they stand. Authors may wish to
> negotiate exemptions or clarify interpretations before they put their
> name to documents with ambiguous terms. Advocates within an institution
> with a knowledge of their institutional policies might be of assistance
> to authors in interpretation of some of these ambiguous conditions.

Only the substantive conditions should be listed. This list should not
amplify noise. The information is all there in the publisher's policy link.

> * In response to user-feedback, as in the original RoMEO list,
> statistics are available on the different colour-categories of each
> publisher. Note, however, that these statistics relate to only those
> publishers held by the list and the basis for comparison will change:
> for example, when further publishers are added at the suggestion of the
> user community.

This table with its 4 colours manages to fail to provide precisely the
information that is actually needed!
http://www.sherpa.ac.uk/romeo.php?stats=yes

There is no need for the four colours: They are arbitrary, uninformative
and confusing.

Instead of:

    ROMEO colour Archiving policy Publishers %

    green can archive pre-print and post-print 28 33
    blue can archive post-print (ie final draft post-refereeing) 12 14
    yellow can archive pre-print (ie pre-refereeing) 7 8
    white archiving not formally supported 38 45

I suggest:

    ROMEO colour Publisher (journal) green light to self-archive: Number %

    bright-green post-print (ie final draft post-refereeing) or both 40 52
OMIT pre-print and post-print 28 33
OMIT post-print (ie final draft post-refereeing) 12 14
    pale-green pre-print-only (ie pre-refereeing) 7 8
    white (no green light yet) 38 45

Also, since there are publishers that publish 1 journal and publishers that
publish 1700 journals, it is extremely important that this summary table
also contain the numbers and percentages by *journal*, not just by publisher!
Authors are interested in and informed by journal numbers, not publisher numbers.

> I hope that the archiving community finds the list useful.

It is very useful, but could be made vastly more useful with some simple
changes in the coding (and by including the individual journal titles).

Stevan Harnad

Prior Topic Thread:

      "SHERPA will take over the Romeo Publisher Policy Table"
       http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/3100.html
Received on Fri Apr 02 2004 - 16:19:58 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:26 GMT