Hennig on spatial characters

pierre deleporte pierre.deleporte at UNIV-RENNES1.FR
Thu Jul 4 17:24:29 CDT 2002

Hi John,

thanks for quotation. I'll have to check in Hennig for the example I
referred to.
Anyway, the quotation below is suficiently explicit that in Hennig's mind
spatial data matter some way.
Now, it seems to me that there is an implication (and also in your own
example), that the taxa in question be already identified as close
relatives, thus there is already a "larger level" phylogenetic information

As for "reciprocal clarification" (or "reciprocal illumination"), I can't
remember where I read, in the eighties I think, the following statement in
the writings of an author criticizing cladistic methods :
"reciprocal illumination is a nice term for circular reasoning" (if anybody
has a hint for the source...)

I don't think so. Combining evidence involves a requirement of independence
of sources of evidence, and true tautology is rarely encountered. It is
conceivable that some biological information is used at some phylogenetic
level, and some spatial information tentatively used, in this context, at a
lower level. It seems that your examples illustrate this, as well as Robert
Mesibov recent ones about hybridization.
No circularity is involved whent  the hierarchy (or combining rules) of
arguments in the reasoning is clearly stated and logically consistent. This
is also the case in the procedure sometimes called "reciprocal
illumination", consisting in applying a congruence criterion in phylogeny
inference from a data marix.
Everything boils down to setting coherent combining rules.


A 10:54 04/07/2002 -0400, John Grehan wrote:
>Pierre wrote a while back
>But I don't remember that Hennig suggested to put characters of spatial
>location inside the data matrix of *taxa + inheritable biological traits*
>in order to infer the phylogeny. I remember Hennig reasoning in terms of
>"oriented dispersal" when a morphocline for some characters fitted to a
>series of spatially oriented locations for these taxa. This can be extended
>from morphocline for a character to phylogenetic relationships between taxa
>(which the morphocline was supposed to fit).
>I admit to not being a scholar of Hennig's works so I would appreciate any
>insight to my current understanding of Hennig's use of geographic
>characters. What I was referring to is his 'vicariance criterion'. In the
>Craw et al book we quote Hennig as follows:
>"The geographic distribution of organisms provides another way of checking
>the reliability of systematic results. Following the principle of
>reciprocal clarification it is possible to reverse order to use geographic
>distribution for determining the phylogenetic relationships themselves. In
>general, two taxonomic groups that standing spatial vicariance relationship
>to each other are more closely related than either is to any other
>taxonomic group".
>Here at least is one explicit model for generating spatial characteristics.
>Agreed, this is not combining spatial and biological data at the outset,
>but using one to check on the other. I applied Hennig's procedure (without
>being aware of Hennig's formulation at the time) with some moths where two
>vicariant genera were geographically adjacent to each other. I proposed
>that they were more closely related to each other than to many other
>members of their family. At the time all they obviously shared was a common
>feeding mode as there were no morphological character studies. Subsequently
>morphological character analysis corroborates the prediction although a
>thorough character analysis admittedly has yet to be produced.
>John Grehan

More information about the Taxacom mailing list