Uncitedness and self-citations: assumption of guilt?

[Tabah Albert]

I agree with Eric Ackermann that one needs to look into citer motivation.
However, one also needs to look into the citer's past work and the place
that work takes in his/her specialty area.

First, we must discount new or emergent fields. In the first few months
after a discovery one would expect to see a high self-citation rate because
the migration of researchers into that field and their article production
has not yet taken place.

Second, it is difficult to discuss this topic without dealing with specific
cases and naming names.  Perhaps I can be allowed to name two researchers
in our field with a relatively high self-citation rate but the value of
whose contribution should leave no one in doubt: Eugene Garfield and Anthony
van Raan.

Especially in his Current Comments over a few decades, Garfield's
articles contain a large number of self-citations. But those can be
explained (I think, I can). Either they involve the use of an instrument of
which he is the sole proprietor (his SCI and SSCI databases) or topics and
work that no one has written about. In science it happens often that so long
as an instrument or material is available in very limited quantities, all
work relating to that material is published by a small group of collaborating
scientists. As soon as the material becomes available "en masse", others flood
in, publish, and the rate of self-citations drops significantly - as in the
case of fullerene research in the early 1990s. Much of the useful data
available from the SCI and SSCI databases for years have been available to
ISI insiders or very expensive to obtain. Given the volume of data, their
wide relevance and the volume of Garfield's writing, a high self-citation
rate is almost inevitable.

A cursory look at "Essays of an Information Scientist" makes it clear that
there is hardly anything that Garfield has not written about. Especially
when he makes connections to previous articles in series dealing with
"Most-cited ... papers" or "Journal Citation Studies. ...", again a high
self-citation rate becomes inevitable. Of course, one also needs to admit
that there is a certain amount of marketing effort for his products.

As to Tony van Raan, he heads a large and important research group that
has been very productive and authoritative over the last fifteen years.
Inevitably, a significant amount of his present work builds on previous
work that is both unique and consistent. A large number of his self-citations
or citations to authors within his research group can be explained by the
fact that his/their work has not been replicated or supplanted by anyone.
While I may at times wonder why they have not cited one or two other items,
I personally am comfortable with the high self-citation rates exhibited by
their work.

As it is, we are talking about rather productive authors. In the case of
authors with little work and a high self-citation rate, it becomes more
difficult to avoid thoughts about self-promotion.

I think, what I am trying to say is that, we omit self-citations in order
to evaluate the true impact of one's work on others. However, by doing that,
at times, we end up throwing out the baby with the bathwater. Where there is
"prior art" that is unique or impossible to avoid, a high self-citation rate
may become quite acceptable.


Albert Tabah
University of Montreal
Montreal, Canada


---------------------------------------------------------------------------
Albert N. Tabah

École de bibliothéconomie et des sciences de l'information
Université de Montréal
CP 6128, Succ. Centre-Ville
Montréal, Québec, Canada
H3C 3J7

Tel.: 514-343-7204
Fax:  514-343-5753
albert.tabah at umontreal.ca