Benzon, Empirical Memetics, and ISI

Timothy Perper/Martha Cornog (
Wed, 11 Jun 1997 06:25:33 -0500

Message-Id: <>
Date: Wed, 11 Jun 1997 06:25:33 -0500
From: (Timothy Perper/Martha Cornog)
Subject: Benzon, Empirical Memetics, and ISI

Bill Benzon wrote:

>Tim Perper says:
>>The past-master at this sort of citation counting is Eugene Garfield,
>>founder of the Institute for Scientific Information in Philadelphia, which
>>publishes Current Contents, Science Citation Index, and Social Science
>>Citation Index. Garfield was instrumental in developing "bibliometrics,"
>>which is the study of cross-reference patterns among journals, scholars,
>>and so forth. One of the big battles about bibliometrics was the use of
>>"citations to" an author as a measure of his or her "productivity" for the
>>purposes of tenure and promotion. I knew one case that almost went to a
>>lawsuit on this issue.
BB> Yes indeed.
TP>> Anyone who wants to follow up citation patterns cannot do better than
>>contact ISI -- they have immense amounts of data on it. It's no small
BB> And they have a website (I think it's
>In the early 80s they were working on something called co-citation analysis
>as a way of discovering lines of investigation emerging in the "cracks"
>between disciplines. Imagine that George is an evolutionary biologist and
>Mary is an egyptologist. They work in different departments, belong to
>different professional associations, and publish in different journals.
>They don't know of one another's work. However, you look at the citations
>in their articles and you see that both cite Dawkins and Dennet &
>Cavalli-Sforza etc. so that there is an interesting overlap in their
>bibliographies. Beyond that, you find that Bob and Carol and Ted and Alice
>along with Harpo, Chico, Groucho & Zeppo etc. also cite those folks.
>Perhaps these folks are moving toward a new discipline.
>I have no idea whatever came of this, but it seems to be the sort of
>technique which would be of use in empirical memetics.

TP: Odd that you should mention that project. If it's the one I'm thinking
of -- and I doubt if there were two ISI projects on co-citation analyses
from the late 70s'-early 80s -- it was being done by a sociologist named
Tyler Thompson, who died in the early 1980s of cancer. He was then in his
late 30s. He was a good friend of mine, and brilliant. He did cluster
analyses from the co-citation matrices, and the "big map," as he called it,
covered virtually all the journals in SCI and SSCI. It was an incredible
piece of work -- in page after page of computer print-outs, you could see
the exact citation structure of the sciences. It was also three inches
thick. None of it was ever published, so far as I know.

In those areas of the sciences with which I was familiar, Tyler's maps
corresponded *precisely* to what I knew about the fields. His idea was to
perform such cluster analyses for a number of years sequentially, and then
correlate them with shifts in resources, federal funding, and manpower
(which today would be called something like human resources allocation).
He died before he could do that work.

If Tyler had lived, there is no doubt that he would be considered one the
great sociologosts and information scientists. He had developed algorithms
for dealing with "mostly sparse" matrices (you see, most journals do NOT
cite each other) that I understand were genuine breakthroughs, at least for
the period.

One of Tyler's great interests -- which is also why I think you may mean
his work when you mention the ISI project -- was the emergence of
interdisciplines, as you say, in the cracks between the sciences. Tyler's
maps could trace the emergence of these interdisciplines first by the
appearance of co-citations between journals of hitherto isolated fields
(which of course represents citations between authors in previously
separate fields) and then by the emergence of new journals with
interdisciplinary titles. As the interdiscipline grows, so do the number
of these interdisciplinary journals. Although the idea is intuitive, it
was incredible to see how this process actually worked in Tyler's cluster

He was also interested in shifts in funding, and their policy implications,
when interdisciplines emerge. He suggested, to me at least, that citation
analyses could detect emerging interdsciplines far more effectively than
any other means.

ISI provided him with the data tapes, but not, I think, any funding. So he
was working alone. He had a genuine grasp of the dynamics of information
flow and exchange, and it was from many conversations with him in those
years that I got my first interests in information science. One of the
points that I have been stressing in these posts -- that memes represent or
concretize the sense of information *flow* in a society -- comes directly
from Tyler, and in some ways was his major insight: information does not
reside in static units, but is essentially *dynamic*, because it *must* be
transferred to have any effects on anything.

This view is consistent with biological concepts of information as well as
with the sort of memetics that Mario Vaneechoutte has been discussing.

So it's strange to see you refer to what must be the same project, because,
as I said, I doubt if ISI had two co-citation projects going on
independently of each other. I'm also glad to remember Tyler, and to say
that his death was a serious loss to our understanding of how science and
information co-exist in a complex society.

This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)