Re: Lumper or Splitter?

Gerrit Hanenburg (
Sun, 08 Dec 1996 16:23:49 GMT

"Rohinton Collins" <> wrote:

>In an attempt to inject some *scientific* discussion into this group lets
>have some opinions on the following:
>1. Is the 'hit-and-miss' process of classification of fossil taxa more
>likely to lead to lumping or splitting?

Doesn't that depend on who does the classifying? Some people are more
prone to being impressed by differences than others.
In my opinion it's better to be a splitter than a lumper because
different taxa can always be recombined after their phylogeny has been
resolved and they emerge as sister taxa.

>2. Is it safe to assume that variation within extinct taxa would be similar
>to the variation seen within extant taxa, bearing in mind that millions of
>years may separate fossils believed to be in the same species?

Variation in certain characters in a long-lived extinct species may be
greater than in extant ones when these characters have changed
significantly through time,but it need not necessarily be so.
I've calculated the CV for Homo erectus (sensu lato) cranial capacity;
CV=14.9% (n=26,values taken from table 27 in Rightmire (1990)).
I do not have this value for a sample of modern humans (anybody?) but
the CV doesn't seem to be exceptionally high despite the enormous time
range (~1.8 - 0.25 myr).

>3. Following on from 1 and 2, do you think that the specimens classified as
>H. habilis are representative of a single species?

I doubt it. It would be a highly sexually dimorphic species with an
unlikely pattern of dimorphism (e.g.with females having larger
browridges and stronger prognathism than males;compare KNM-ER 1813
with ER 1470). The same problem exists with relation to
postcrania,with OH 62 being even more primitive than AL 288-1 in many
respects (Hartwig-Scherer and Martin 1991) while a specimen like
KNM-ER 1481 is much more derived.
A complicating factor is that no sufficiently complete associations
between crania and postcrania are available. It is possible that ER
1481 is Homo erectus.

>I think that it is dangerous making any hard and fast rules about
>classification of extinct taxa. With extant species we can test for
>inter-fertility. This is not possible with extinct species. The fact that a
>species may span several million years makes this irrelevant anyway. For
>example early eastern H. erectus specimens may differ sufficiently from
>later H. erectus for some to identify them as different species. Has
>speciation occurred? How long before speciation may be said to occur in a
>species experiencing gradual evolution (as opposed to punctuated
>equilibrium when a species with a markedly different morphotype arises in a
>very short period of time)? To a certain extent it is simply a question of

We can only define palaeontological species on the basis of observed
morphology. Palaontological species are thus minimal morphological
clusters of individuals deemed useful to establish. We can only hope
that they're also species in an evolutionary sense.

Hartwig-Scherer,S.and Martin,R.D.(1991), Was "Lucy" more human than
her "child"? Observations on early hominid postcranial remains.
J.of Hum.Evol.21:439-449.

Rightmire,G.P.(1990), The evolution of Homo erectus. Comparative
anatomical studies of an extinct human species. Cambridge University