Skip this thread if Biogeography bores you - especially those recurrent discussions about Croizat

Tue Feb 5 15:38:50 CST 2002

At 11:28 AM 2/4/02 +0100, P.Hovenkamp wrote:
>Hi all those not deterred by mentioning the name of Croizat,
>Readers without access to the writings of Croizat may from the following
>easily gain the impression that Croizat was seeing things (even if not
>writing) clearly and was recognizing the importance of  hybridization:
>>  More
>>>importantly, it is becoming increasingly obvious that reticulate evolution
>>>is important in many, if not most, sexual groups.
>>Actually I would suggest that it has been as 'obvious' all along - just not
>>popular with
>>traditional evolutionary models (I would call them 'Darwinian' models -
>>pejorative or
>>not). Croizat referred to reticulate evolution as character recombination.
>>John Grehan
>But this last statement is not exactly true. At least, if Croizat referred
>to reticulate evolution as character recombination, he did not do so in
>writing. I checked the quite extensive index on "Things and concepts" in
>Croizat's "Space, time form" (1964) - unfortunately, his later works are
>not so fully indexed.
>There are c. 27 entries on character recombination, and c. 20 on
>hybridization and interbreeding (as placeholder for reticulation, which is
>not a word Croizat indexed on). Only on 2 pages the two concepts are
>treated together, and looking up those pages it seems clear that the two
>are seen as completely different processes:
>p. 273, footnote: "Adding to this the effect of (...some processes...)
>hybridizations and recombinations (...more distinct processes...) form
>making may reach in time a hardly credible density"
>p. 520: "(...) hybridization took place here and there during spells of
>mobilism followed in immobilism (C.means during subsequent spells of
>immobilism, note PH) by re-combination of characters of permanent nature,
>whatever the technical process in detail."
>To me, it is clear that hybridization and character recombination are to
>Croizat two completely different processes, which need not coincide in time.
>This is of course rather surprising, as hybridization offers such a good
>expanation of observed character recombination. Yet Croizat seems not to
>have thought of that. But then, what he called character recombination is
>not what we casually call character recombination.
>His concept of character recombination refers to parallel evolution of
>similar combinations of character states. To Croizat, organisms were a bit
>like banks of dip-switches. Ancestral forms had all the switches in the
>same position, and in descendant forms different combinations of switches
>would be "up" or "down". Recombination of characters refers to the
>possibility that the same combination of "up" and "down" could occur in
>"unrelated" forms. Throughout evolution, new "banks" of switches could not
>be added: "in evolution old characters are being combined and recombined
>without cease rather than new characters constantly called into being" (p.
>309). Thus, it is not necessary to invoke hybridization or reticulation -
>it was enough to recognize that "form making" could simply flip any number
>of switches at any time.
>I will not go into John Grehans first statement - that "traditional
>evolutionary models" have neglected reticulate evolution. It just shows
>that he's not a botanist.
>Peter Hovenkamp

But yet botanical cladograms are constructed the same way as others, aren't
they?  If so, and for as long as this has occurred and continues to occur,
that is a warning flag for the methodology as a whole, from my
perspective.  And how about the degree to which cladistic (and other?)
botanical studies use chloroplast and ribosomal rather than nuclear genes
from which to generalize?

Steve Manning

>P. Hovenkamp
>Nationaal Herbarium Nederland - Leiden
>PO Box 9514
>2300 RA  Leiden
>The Netherlands
>hovenkamp at

More information about the Taxacom mailing list