taxacom taxacom at IBSS.DVO.RU
Mon Jan 28 11:18:47 CST 2002

To: e.parmasto at

Dear Erast Khansovich,

It is nice to hear from you, even if you ask difficult questions about the
definition of 'natural'. I'm afraid that neither Gilmour nor Linne did not
inspire me with my notion of 'natural', but this notion cannot do without
'Linnaean Hierarchy'. (Recently, I wrote several articles in defence of the
Linnaean Hierarchy in - maybe, vain - attempt to resist to the genuine
'crusade' against Linnaeus in modern 'phylogenetic systematics'. Benton's
interesting paper is also for Linnaeus but my views are diverged in some

1. Gilmour. - I don't share his following thesis (Gilmour, 1940 : 466):
"NATURAL groups class together individuals which have a large number of
attributes in common, whereas in artificial groups the individuals concerned
possess a much smaller number of common attributs". Thus: "A natural group
being based on a large number of attributes..."

Of course, not! Or, more exactly, taxonomists should use as many characters
as possible but one must weigh characters (instead of mere addition
together). Quantity is not decisive at all. Vavilov's law of homologous
variation demonstrates the weakness of Gilmour's 'natural' groups especially
well. For example, for Secale and Triticum, Vavilov (1922) counted about 30
similarities in their inner polymorphism, but only 1-2 characters compose a
difference between these genera (Probatova, 1985). Thus the 'overall
similarity' is higher then the difference (of higher level and,
consequently, of higher value), and the unification of two groups would be

2. Linnaeus. - Much more complicated. He admitted natural species and
genera, even natural orders and classes: "most of the classes and orders are
natural"; "the artificial classes will take the place of natural ones [only]
until the latter have been discovered" (cf. Stafleu, 1971: 66-67). In other
words, he admitted the natural taxa. The problem (for most modern
taxonomists) is the fact that Linnaeus associated the search of natural
groups with the search of their 'essences' (while a lot of nonsensical words
about essences were written in taxonomic literature).

"The knowledge of the 'essences' of the 'natural' taxa, however, is not
easily gathered" (l. c. : 73). Exactly! Linnaeus did not suggest the clear
method to find natural taxa. He only knew that 1) "the crux of the matter is
that the taxonomist notices the DIFFERENCES, and that these differences are
the key to the essence" (l. c. : 52) (Exactly! after weigting!) and 2) that
'essences', or distinguishing characters, are distributed among different
levels in taxonomic hierarchy. His main weakness was A PRIORI fixation of
some characters (especially those of fructification) at a certain level. A
priori, a taxonomist can only find tentative 'species' or 'genera'; some of
them may appear to be quite natural, others - not. Linnaeus was a genius, he
guessed a lot of 'good' species and genera persisting today.

What I suggested is the addition of A POSTERIORI testing of tentative
groups. By the way, Linnaeus himself emphasized that a posteriori knowledge
increases the naturalness of genera (cf. his Philosophia botanica, 159).

In my work on several orders of fungi I tried to find taxa of equal level
among numerous chaotically segregated species, genera and families and
started from the most frequent differences between them. This process may be
called an aposteriori character weighting, while taxa under consideration
undergo a kind of test for the equality of level. Usually, many species and
genera stand such a test, while some others should be united or removed to
other groups because of the wrong former placement. The majority of familiar
and traditional taxa remain almost unchanged after such an analysis but they
may be placed into the more natural system. Despite there are no absolute
species characters, one can use so many differences as to retain the largest
number of familiar species and, then, keep the species level for their
differences. The same procedure may be accomplished with tentative genera
and families. As the result, the traditional Linnaean hierarchy remains as
good as it was before but it may be quite compatible with evolutionary
thinking, with the principle of common descent and even with the
phylogenetic systematics.

Best wishes,


----- Original Message -----
From: "erast" <e.parmasto at>
To: <taxacom at IBSS.DVO.RU>
Sent: Saturday, January 26, 2002 4:21 AM
Subject: Re: paraphyly

> Larissa Vasilyeva wrote:
> ........
>     For instance, a genus contains a hundred of species and all of them
are descendants of a single
>  ancestor. One can arbitrarily divide the genus into several groups of
species, and all of those
> groups may be meaningless (paraphyly is mistake). However, suppose, one
finds such
> characters that divide a genus into quite natural subgenera. Then,
subgenera are paraphyletic
> extensionally since they do not include all of the original ancestor's
descendants, but they
> are monophyletic intensionally...
> ...........
>    The question is (in this case): what means "NATURAL subgenera".
> Without clear definition of "natural" this is not understandable. -
> Natural sensu Gilmour? Or Linne?
>    Erast Parmasto
> *************
> Erast Parmasto
> Institute of Zoology and Botany
> 181 Riia St., 51014 Tartu, Estonia
> Tel -372 7 383 027   Fax -372 7 383 013

More information about the Taxacom mailing list