More on biogeographic homology

P.Hovenkamp hovenkamp at RHBCML.LEIDENUNIV.NL
Tue May 18 09:57:46 CDT 1999

At 07:06 AM 12-05-99 -0400, John Grehan wrote:
>P. Hovenkamp expressed doubt over my perspective on biogeography and
>outlined his approach as follows:
>>Similarly, when I am correlating taxon- with earth history, I am using the
>>best available data for both. If better data turn up for either, I may have
>>to think again.
>In this procedure, a biogeographic pattern is being correlated with a
>geological hypothesis. Since the geological hypothesis is just that,
>what biogeographers do is "explain" a distribution on the basis of a
>story generated by another discipline (geology). Whether or not the
>story is good or not, the procedure renders the biogeography (the spatial
>pattern) empty of  empirical content - the biogeography itself contains
>no empirical data to generate a historical hypothesis.

Biogeography is an wide field - in my paper in Cladistics (1997) I tried to
distinguish between the two main approaches. What we are arguing about here
is a matter of "taxon history" (I think initially spawned by a question
from John Grehan about the explanation of a single taxon occurring in three
widely dispersed localities). Taxon history attempts to explain the current
distributional pattern of a taxon on basis of the best information about
its history combined with the best available information about earth
history. It does not attempt to "generate a historical hypothesis".

On the other hand, if we want to generate a historical, geological
hypothesis from biogeographic data, we are practising what I termed the
"Earth history approach". A different approach, with different goals and
different methods.

>Geologists on the other hand analyze geological patterns to
>generate historical hypotheses. They do not rely on biogeographic
>stories about the past to generate their stories.
>Panbiogeography is the only method I am aware of (I am open
>to correction) where biogeographic patterns are analyzed to
>generate historical hypotheses from the distribution data.

This clearly is "Earth history", and vicariance biogeography was originally
developed to do exactly that (see Rosen, 1988).

I cannot let the occasion pass to insert a reference to Van Steenis as
well, who, on basis of biological distribution patterns argued that land
connections between various parts of the world had to have existed at some
time in the past, in the face of the then accepted geological theories.

>>I think that is all fairly normal. Science is fallible, and depends on
>>empirical data.
>I agree. In panbiogeography the empirical data are distribution patterns. So
>panbiogeography has scientific content.
>>Somehow it seems to me that what John Grehan wants is an explanation that
>>remains correct even if the data it tries to explain is changed.
>This was not evident to me.
>Hovenkamp states that he does not understand the problem of biogeographic
>content. I don't know how many others feel that way, but it represents I
>think, an important distinction between panbiogeography and other
>biogeographic practices that I am familiar with. In panbiogeography the
>distribution pattern - the spatial structure - is seen as informative. As a
>parsimony criterion it is necessary to chose a particular spatial structure
>out of all those possible. The criterion chosen is minimum distance
>as the initial hypothesis.

Maybe my lack of understanding was due to the fact that I only recognised
in this mailing exchange that the confusion between the two biogeographic
approaches (earth- vs. taxon history) was again at the root of some of the

P. Hovenkamp
Rijksherbarium, Leiden
The Netherlands
hovenkamp at

More information about the Taxacom mailing list