Prof Harnad,
On Tue, 6 Jul 1999, Stevan Harnad wrote:
> On Tue, 6 Jul 1999, J.W.T.Smith wrote:
>
> > There are problems with this algorith.
> >
> > Applying it precisely would make review articles (for which the author
> > received a small honorarium), some editorials and all commissioned
> > surveys/reports (copies of which may be given away on request) 'trade' -
>
> For 99% of the refereed literature this algorithm applies perfectly
> well. The case of trivial honoraria was covered in the discussion of
> the hybrid journals like Nature (below). The case of editorials is
> trivial. Uncontested give-aways are irrelevant.
>
> The point of raising these inconsequential details is not at all clear.
You claimed it was "precise". I pointed out it was not as "precise" as you
claimed. I think that was fair criticism - if you had said it was capable
of deciding between 90% of cases I would have agreed. In addition, later
in your note you used the 'preciseness' of your algorithm to claim what it
decided was "FACT" (your emphasis). When you make such strong claims you
should expect readers to look critically at the basis of your claim.
> > However the problem could be avoided altogether if a publishing model was
> > adopted that separated the evaluation/quality control role from the
> > publishing/archiving role (making available) from the distribution role
> > (making aware). Strangely enough :-) this is the core of my Distributed
> > Journal model
>
> You have repeated this point a number of times and it would now be
> useful to move on to matters of substance.
Notice the smiley? I was aware I was repeating myself. You habitually
sprinkle your notes with repeated links to your own papers and the same
related URLs. I was just following your example.
> The reply has in each
> instance been that self-archiving is self-archiving; peer review is
> peer review; there's the separation of function and that's all there is
> to it.
But how to organise peer review and self-archiving is another question, So
is how to fund both.
> Your own "model" introduces spurious (1) add-on quality markers (which
> have to be paid for separately by the consumer)
I assume you are talking about SFPs here as no other part of my model
requires payment from the consumer. They are not "quality markers" -
they are subject related filters.
> and (2) multiple
> "journals" for the same article (a profligate use of a scarce resource:
> peer reviewers).
You are confusing two things here probably because you are still using
the term 'journal' which carries too many connotations from the past.
To clarify:
Any SFP can point to any item (article) the producers feel falls into
their subscribers sphere of interest. SFPs are not journals - they *only*
play the alerting role of the current journal.
My independent evaluators (a cumbersome phrase I must admit) are also not
journals - they only play the quality control role of the journal. My
suggestion that authors could submit a 'manuscript' to more than one
independent evaluator for their 'seal of approval' (I now prefer your term
'certification') is just a way of opening up this whole area of quality
control. Also there may be other forms of quality control in addition to
the current operation of peer review. It certainly shouldn't be considered
heresy to think about these things. My model doesn't say the current
journals can't play the independent evaluator role. Neither does it say
they can't set up SFPs but it does say these two activities must be kept
separate - that neither evaluators nor SFPs can 'own' items in the way
current journals 'own' the articles they carry.
The purpose of this apparent complexity is that I am trying to escape from
the limitations of a 17th century paper based publishing model as we move
to a totally new publishing medium (and incidentally the 21st century)
while at the same time retaining the important aspects of that model. My
justification for my approach is summed up in
http://www.ukc.ac.uk/library/papers/jwts/Prolegomena.htm
> It does not clarify the trade/nontrade distinction.
That's because this distinction is immaterial in my model :-) All that
matters is that the items are publically visible on the net.
That's one of the most interesting things about paradigm shifts. An aspect
that is vitally important in one paradigm can disappear completely in the
paradigm that replaces it.
> ... Let's just make sure all authors publicly self-archive all
> their journal articles. The rest will take care of itself.
On this I agree. Once there is a pool of refereed (and non-refereed but
acceptable quality) items out there completly new services will arise to
help academics (and others) make use of this new information resource (and
in some cases make money as well). I cannot predict what these services
will be but I am sure they will not fit neatly into any model carried over
from the paper based publishing world.
Regards,
John Smith.
University of Kent at Canterbury, UK.
Received on Wed Feb 10 1999 - 19:17:43 GMT