Darwin-L Message Log 12: 1–25 — August 1994

Academic Discussion on the History and Theory of the Historical Sciences

Darwin-L was an international discussion group on the history and theory of the historical sciences, active from 1993–1997. Darwin-L was established to promote the reintegration of a range of fields all of which are concerned with reconstructing the past from evidence in the present, and to encourage communication among scholars, scientists, and researchers in these fields. The group had more than 600 members from 35 countries, and produced a consistently high level of discussion over its several years of operation. Darwin-L was not restricted to evolutionary biology nor to the work of Charles Darwin, but instead addressed the entire range of historical sciences from an explicitly comparative perspective, including evolutionary biology, historical linguistics, textual transmission and stemmatics, historical geology, systematics and phylogeny, archeology, paleontology, cosmology, historical geography, historical anthropology, and related “palaetiological” fields.

This log contains public messages posted to the Darwin-L discussion group during August 1994. It has been lightly edited for format: message numbers have been added for ease of reference, message headers have been trimmed, some irregular lines have been reformatted, and error messages and personal messages accidentally posted to the group as a whole have been deleted. No genuine editorial changes have been made to the content of any of the posts. This log is provided for personal reference and research purposes only, and none of the material contained herein should be published or quoted without the permission of the original poster.

The master copy of this log is maintained in the Darwin-L Archives (rjohara.net/darwin) by Dr. Robert J. O’Hara. The Darwin-L Archives also contain additional information about the Darwin-L discussion group, the complete Today in the Historical Sciences calendar for every month of the year, a collection of recommended readings on the historical sciences, and an account of William Whewell’s concept of “palaetiology.”

--------------------------------------------
DARWIN-L MESSAGE LOG 12: 1-25 -- AUGUST 1994
--------------------------------------------
_______________________________________________________________________________

<12:1>From DARWIN@iris.uncg.edu  Mon Aug  1 00:07:15 1994

Date: Mon, 01 Aug 1994 01:06:57 -0500 (EST)
From: DARWIN@iris.uncg.edu
Subject: List owner's monthly greeting
To: darwin-l@ukanaix.cc.ukans.edu
Organization: University of NC at Greensboro

Greetings to all Darwin-L subscribers.  On the first of every month I send
out a short note on the status of our group, along with a reminder of basic
commands.

Darwin-L is an international discussion group for professionals in the
historical sciences.  It is not devoted to any particular discipline, such
as evolutionary biology, but rather endeavors to promote interdisciplinary
comparisons among all the historical sciences, from historical linguistics
and geology to archeology, systematics, cosmology, and textual criticism.
The group is now eleven months old, and we have just 600 members from 30
countries.  I am grateful to all of our members for their interest and their
many contributions, and for helping to make Darwin-L one of the most
professional and successfully interdisciplinary discussion groups around.

The Darwin-L gopher contains logs of all our past discussions, as well as a
collection of files and network links of interest to historical scientists.
The Darwin-L gopher is located at rjohara.uncg.edu; on most mainframe systems
you can simply type "gopher rjohara.uncg.edu" to get there.

Darwin-L is occasionally a "high-volume" discussion group.  Subscribers
who feel burdened from time to time by their Darwin-L mail may wish to take
advantage of the digest option described below.

Because different mail systems work differently, not all subscribers can
see the e-mail address of the original sender of each message in the message
header (some people only see "Darwin-L" as the source).  PLEASE include your
name and e-mail address at the end of every message you post so that everyone
can identify you and reply privately if appropriate.  Remember also that in
most cases when you type "reply" in response to a message from Darwin-L your
reply is sent to the list as a whole, rather than to the original sender.

The following are the most frequently used listserv commands that Darwin-L
members may wish to know.  All of these commands should be sent as regular
e-mail messages to the listserv address (listserv@ukanaix.cc.ukans.edu),
not to the address of the group as a whole (Darwin-L@ukanaix.cc.ukans.edu).
In each case leave the subject line of the message blank and include no
extraneous text, as the command will be read and processed by the listserv
program rather than by a person.  To join the group send the message:

     SUBSCRIBE DARWIN-L <Your Name>

     For example: SUBSCRIBE DARWIN-L John Smith

To cancel your subscription send the message:

     UNSUBSCRIBE DARWIN-L

If you feel burdened by the volume of mail you receive from Darwin-L you
may instruct the listserv program to deliver mail to you in digest format
(one message per day consisting of the whole day's posts bundled together).
To receive your mail in digest format send the message:

     SET DARWIN-L MAIL DIGEST

To change your subscription from digest format back to one-at-a-time
delivery send the message:

     SET DARWIN-L MAIL ACK

To temporarily suspend mail delivery (when you go on vacation, for example)
send the message:

     SET DARWIN-L MAIL POSTPONE

To resume regular delivery send either the DIGEST or ACK messages above.

For a comprehensive introduction to Darwin-L with notes on our scope and
on network etiquette, and a summary of all available commands, send the
message:

     INFO DARWIN-L

To post a public message to the group as a whole simply send it as regular
e-mail to the group's address (Darwin-L@ukanaix.cc.ukans.edu).

I thank you all for your continuing interest in Darwin-L and in the
interdisciplinary study of the historical sciences.

Bob O'Hara, Darwin-L list owner

Robert J. O'Hara (darwin@iris.uncg.edu)
Center for Critical Inquiry and Department of Biology
100 Foust Building, University of North Carolina at Greensboro
Greensboro, North Carolina 27412 U.S.A.

_______________________________________________________________________________

<12:2>From schoenem@QAL.Berkeley.EDU  Tue Aug  9 01:12:02 1994

Date: Mon, 8 Aug 1994 23:11:09 -0700 (PDT)
From: Tom Schoenemann <schoenem@qal.Berkeley.EDU>
Subject: Re: Deep roots and measuring diffusion
To: darwin-l@ukanaix.cc.ukans.edu

It is really too bad that Greenberg is not on this list himself, in order
that he might respond directly to comments such as those recently given by
Thomason.  Anyone interested in getting the general tenor of what he might
say should read the first chapter of Greenberg's book, "Language in the
Americas"  (from which all the quotations below are taken).  He makes a
number of crucial points that appear to be quite convincing, at least to
this humble non-linguist.  For example:

The idea that sound-correspondences are necessary to prove a genetic
relationship is simply naive.  He gives numerous quotes from early work
done on the reconstruction of proto-sound systems indicating that none of
them believed the validity of the genetic groups they worked with DEPENDED
on such reconstructions.  For example, he quotes Delbruch (in what
Greenberg states is "frequently looked on as the basic manifesto of the
Neo-Grammarians", p 30, LIA)

"My starting point is that specific result of comparative linguistics that
is not in doubt and cannot be in doubt.  It was proved by Bopp and others
that the so-called Indo-European languages are related.  The proof was
produced by juxtaposing words and forms of similar meaning.  When one
considers that in these language the formation of the inflectional forms
of the verb, noun, and pronoun agrees in essentials and likewise that an
extraordinary number of inflected and uninflected words agree in their
lexical parts, the assumption of chance agreement must appear as absurd."

Greenberg also points out that Albanian is universally recognized as an
Indo-European language, yet this was accomplished without showing how
Albanian forms can be derived using 'regular sound laws' from
reconstructed Proto-Indo-European.

Greenberg notes that: "In the last five decades or so, we may note,
Hittite, Luwian, Lycian, and Palaic, and other Anatolian languages
have been universally recognized as Indo-European, as has Tokharian with
two dialects, Tokharian A and B.  In no case did anyone publish the sorts
of articles with tables of correspondences and asterisked forms so common
in the pages of the INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS, which
are believed to reflect the methodology of Indo-European comparative
linguistics.  In the case of Tokharian, one must believe that all three
(or four) consonant manners had merged into one, and then each phoneme,
now the sole representative of the position of articulation, split into
two under phonetic circumstances that no one has been able to state.

"Why, then, have all these language been accepted as Indo-European?
The reason is the existence of a considerable number of word-stems
resembling those that are widespread in Indo-European, and a number of
highly characteristic grammatical formatives involving sound and
meaning.  Thus the existence of even a few froms such as Hitite eszi 'he
is' and asanzi 'they are' (cf. Latin et/sunt, Sanskrit asti/santi) is
quite sufficient to exclude accident." (p. 19)

Later, Greenberg points out the inefficiency of focusing efforts at lower
level groupings first, before looking at broader groupings.  Greenberg
writes: "There has grown up, as a corollary of present doctrine, the
notion that one must first reconstruct the proto-language of each lower
grouping, thus proving its validity, before proceeding to the
reconstruction of the higher level groupings.  Such a stepwise procedure
appears to be very virtuous, but in fact is an illusion.  The
reconstruction will itself be a poorer approximation to the truth if it is
confined to a restricted group.  In the model case of Indo-European, it
was the broader group that was first reconstructed.  In fact, many
phenomena of narrower groups can only be understood historically by
outside evidence from within the broader stock.  An example is the
Germanic consonant alternations that arose as a result of the change
expressed in Verner's law.  These alternations require a knowledge of the
Indo-European pitch accent that did not exist in Germanic in the historic
period.  After all, in historical matters, the earlier explains the later.
If the earlier is not directly attested, one looks backward by looking
sideways, which is precisely the comparative method." (p. 36)

Greenberg also points out that there are no such things as perfect
sound-correspondences.  He uses French-English correspondences to
demonstrate the tremendous degree of inconsistency that is typical even of
languages who's affinity no one doubts:

"The existence of the same correspondence in several different etymologies
certainly adds to the probabilities of each being correct.  Moreover, such
correspondences are our chief methodological tool in reconstruction.
However, what many linguists fail to appreciate is that anything
approaching a complete and highly convincing reconstruction on the basis
of recurrent correspondences is in general possible only with languages so
closely related that it is unnecessary anyway.  Even here we have cases
like Athabaskan in which the reconstructions not only differ, but are
confined to initial and non-final consonants.  Where the separation is
greater, as we have seen, the reconstructions are so underdetermined by
the data that deviations from a particular theory of reconstruction can be
accommodated by a whole series of strategies [of which he outlines at
least 10 that have actually been used by linguists].  Some etymologies
will always remain uncertain.  In others the lesser claim of cognation can
be maintained, even thought the reconstructive explanatory theory remains
uncertain or in dispute.  A far more convincing refutation is to show an
incorrect morphological analysis, not deviation from a predicted
phonological outcome." (p.33-34)

Another comment might be made.  Thomason's point about not knowing in
the case of two isolated languages whether their similarities are due to
borrowing or to common ancestry serves to highlight a general
misunderstanding of Greenberg's method and parsimony in general.  In a
case such as this, where there really are no related languages around, it
is true that the similarities COULD be due to borrowing, but borrowing is
not as simple an explanation as common ancestry.  Borrowing does not
involve simply hearing a foreign word once, it involves a significant
degree of cultural contact, mixing, exchange, etc.  This is not as simple
an explanation as common ancestry, and the simplest explanation must
always be favored over the more complicated.  The reason this is so is
quite simple.  Given this hypothetical situation that Thomason describes,
we are also not able to exclude the possibility that both languages
borrowed the same words from a third, now extinct language.  It does no
good to argue that we should reject this as a favored explanation for the
resemblances because it is "less likely," since exactly the same logic
holds for borrowing between different cultural groups.  We must pare-off
unnecessary complications until other evidence suggests otherwise.

Greenberg's method involves multiple comparisons, which means that cases
of borrowing become much more apparent.  In fact, borrowing can only be
detected in the context of the broader grouping that the two languages
come from.  Greenberg points to the situation with Turkish and Arabic,
which share numerous apparent cognates.  He writes, "One obvious
consideration, of course, is that these loanwords are not basic
vocabulary items.  But we do not even need such a hypothesis.  The most
powerful proof is, once more, distribution across languages.  Turkish
and Arabic are not mutually intelligible and are obviously distinct
languages.  Hence, if Turkish were really a Semitic language, it would
show some independence within that family.  But Turkish never has a
Semitic morpheme unless it occurs in close to the same form in Arabic.
In the absence of direct historical evidence, which is of course present
in this case, this is the most powerful evidence for borrowing." (pp. 22-23)

In other words, it it the patterns across groups of languages that allow
us to ferret out borrowing.

In summary, it appears that: 1) regular sound correspondences have never
been used to prove genetic relationships, 2) regular sound correspondences
always have many exceptions, 3) sound laws are best reconstructed AFTER
the genetic grouping is determined, and 4) instances of borrowing are best
detected by examining the context given by the broader genetic grouping.

I would be interested to know how linguists who disagree with Greenberg
would address these points.

P. Tom Schoenemann
Department of Anthropology
University of California, Berkeley
(schoenem@qal.berkeley.edu)

_______________________________________________________________________________

<12:3>From sally@isp.pitt.edu  Tue Aug  9 06:29:10 1994

To: darwin-l@ukanaix.cc.ukans.edu
Subject: Re: Deep roots and measuring diffusion
Date: Tue, 09 Aug 94 07:29:05 -0400
From: Sally Thomason <sally@isp.pitt.edu>

  Tom Schoenemann is quite right in saying that, to nonlinguists,
Greenberg's methodological arguments and data look very convincing:
that has to be the reason for the surprising (to most historical
linguists who have examined his data) acceptance of his results by
nonlinguists.  I've taken up too much space on darwin-l lately,
so won't give a detailed response to Schoenemann's points, but (fairly)
briefly, here's how the response would go:  (1) The crucial issue is
not absolutely regular sound correspondences.  The crucial point is
some evidence of SYSTEMATIC correspondences -- recurring sound
correspondences -- in forms (including inflectional affixes,
as in the Delbruck quote) of similar meaning.  That's exactly
what Delbruck is talking about, in the passage quoted.  This
also makes Schoenemann's point (2) moot:  no one, but no one,
is insisting on absolute regularity.  (3) Of course no
reconstruction can be done until after one suspects one has a
genetic grouping.  But (we believe) there are sure to be related
languages on which no reconstruction can be done, because you can't
reconstruct anything if you don't have systematic correspondences.
That's why we insist on the *possibility* of reconstruction; you
have to be able to show that the correspondences are there, and
showing that some reconstruction is possible is the surest way
to do that.  (4) Of course you need patterns across languages
to detect borrowing, *at levels of connection where patterns
are still detectable at all*.  That's just it: in Greenberg's
data, there are no patterns -- either of the sort that would
permit reconstruction or of the sort that would permit us to
separate a few remaining remote borrowings from a few remaining
inherited words.  Parsimony doesn't really enter in here,
unless of course one starts with the methodological
assumption that inheritance is always more likely than
borrowing as an explanation for shared words.  (It's not just
inheritance vs. borrowing, though: for most historical linguists
who have examined Greenberg's data, the main concern is
historical connection -- whether inheritance or borrowing --
vs. chance.  Chance is what, in most specialists' opinion,
Greenberg has not excluded.)

  The real issue is time depth.  The Indo-European examples
aren't really relevant, because there we're talking about a much
shallower time depth than with Greenberg's "Amerind"; and, even
more significantly, what made it possible to see immediately
that those newly discovered "small" languages were Indo-European
was that there was already a very solid matrix of information
about Indo-European structure and lexicon to fit them into.
With "Amerind", there is no such matrix.

  For the rest, see the papers that I and other linguists have
mentioned on darwin-l.

  Sally Thomason
  sally@isp.pitt.edu

_______________________________________________________________________________

<12:4>From peter@usenix.org  Tue Aug  9 07:15:23 1994

Date: Tue, 9 Aug 94 05:15:05 PDT
From: peter@usenix.org (Peter H. Salus)
To: darwin-l@ukanaix.cc.ukans.edu
Subject: Re: Deep roots and measuring diffusion

Sally,
Well, as a historian, I want to point out the internal
contradiction of (a) citing Delbrueck and the stating (b):

>>also makes Schoenemann's point (2) moot: no one, but no one,
>>is insisting on absolute regularity.  (3) Of course no

as it is precisely Brugmann, Delbrueck, etc., through
Hermann Paul and Streitberg who insist on the "Ausnahmslosigkeit
der Lautgestetze."

Peter

_______________________________________________________________________________

<12:5>From GOLLAV@axe.humboldt.edu  Tue Aug  9 13:57:57 1994

Date: Tue, 09 Aug 1994 11:59:23 -0700 (PDT)
From: GOLLAV@axe.humboldt.edu
Subject: Re: Deep roots and measuring diffusion
To: darwin-l@ukanaix.cc.ukans.edu

Sound "laws" -- regular phonemic correspondences in cognate forms in
a group of genetically related languages -- are part of the fabric of
historical explanation.  They are not a precondition for proposing a
genetic grouping, but neither are they (as Greenberg sometimes says)
mere pedantic frills that one can postpone working out until one has
time for the exercise.  Historical research is rarely if ever deductive;
it proceeds by building up a thick fabric of interlinking probabilities.
In historical linguistics, I think it is fair to say that no language
family can be considered ESTABLISHED -- that is, accepted as a valid,
productive historical hypothesis -- until etymologies are thick on the
ground.

Etymologies are historical hypotheses about individual words and
grammatical features.  It is one thing to say that there are striking
resemblances among English, French and German, and that these support
a genetic explanation.  It is quite another to show that the resemblances
among _fire_, _Feuer_, and _feu_ result from genuine cognacy between
the English and German forms, but from chance in the case of French
(which is from Latin _focus_).  In general, you can't propose meaningful
etymologies without hypothesizing sound correspondences and reconstructing
proto-languages at various time depths.  (The explanation of the "fire"
words requires the reconstruction of Proto-Germanic to account for English
and German, and Proto-Romance (if Latin weren't already attested) and
Proto-Indo-European to account for the French form.)  The historical
understanding of the genetic relationship between English and German, and
of the other historical relationships, both genetic and otherwise, between
French and either of the other two, are codified in etymological dictionaries
and comparative grammars.  It is NOT in any significant way based on
either abstract "sound laws" or on gross counts of the numbers of words
that look alike.

On the other hand, genetic groupings can be PROPOSED before such detailed
historical work has been done.  In the Americas this has generally been the
case, from Albert Gallatin's time onward.  The relatively few scholars who
have concerned themselves with American Indian linguistic diversity on the
grand scale have mostly been content to make sweeping classificatory
statements, more in the nature of forecasts or prognostications than as
serious historical proposals.  (Edward Sapir's 1929 scheme of 6 "superstocks"
for North and Central America is one of the best known.)  But such proposals
have no HISTORICAL standing until subjected to real historical (i.e.,
etymological, reconstructive, sound-correspondence-postulating) investigation.
As A. L. Kroeber once wrote (with Sapir evidently in mind) a classification
of this sort "is in no sense whatever a definable or controllable method of
science or scholarship."  It is simply a prognostication, a suggestion that
future research should turn in certain directions.

Greenberg's hemisphere-wide web of interlocking genetic proposals in
LANGUAGE IN THE AMERICAS is in the same league as similar speculations of
Sapir's (or Brinton's, Radin's, Loukotka's, and Swadesh's).  From Amerind
on down, Greenberg's proposals are interesting and may well goad research,
although not significantly more than previous speculations.  In fact, many
of the specific groupings Greenberg has proposed overlap considerably with
groupings perviously proposed, including the separation of Na-Dene from
the rest of American languages -- an old idea of Sapir's.  What is different
about Greenberg's scheme is the wrongheaded vehemence with which he and a
few partisans have argued that his "method" (a very crude survey of super-
ficially resemblant vocabulary) miraculously yields true historical results,
with no further research required.  This astonishingly unscientific claim
has essentially isolated Greenberg from most of his peers and subjected his
work to scathing reviews.

As far as most responsible historical linguists are concerned, the matter
is settled.  Through a stream of popular and self-serving articles and books,
Greenberg has chosen to appeal his case to non-linguists.  Claiming as he
does to have to have spun the gold of history out of the flax of simple
vocabulary comparison, his work is bound to attract attention, especially
from those people who think the minutiae of real linguistic scholarship are
hopelessly remote and useless.

--Victor Golla
  Humboldt State University
  gollav@axe.humboldt.edu

_______________________________________________________________________________

<12:6>From delancey@darkwing.uoregon.edu  Tue Aug  9 17:33:00 1994

Date: Tue, 9 Aug 1994 15:32:51 -0700 (PDT)
From: Scott C DeLancey <delancey@darkwing.uoregon.edu>
Subject: Re: Deep roots and measuring diffusion
To: darwin-l@ukanaix.cc.ukans.edu

Tom Schoenemann quotes from Greenberg's defense of his methods:

> For example, he quotes Delbruch (in what
> Greenberg states is "frequently looked on as the basic manifesto of the
> Neo-Grammarians", p 30, LIA)
>
> "My starting point is that specific result of comparative linguistics that
> is not in doubt and cannot be in doubt.  It was proved by Bopp and others
> that the so-called Indo-European languages are related.  The proof was
> produced by juxtaposing words and forms of similar meaning.  When one
> considers that in these language the formation of the inflectional forms
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> of the verb, noun, and pronoun agrees in essentials and likewise that an
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> extraordinary number of inflected and uninflected words agree in their
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> lexical parts, the assumption of chance agreement must appear as absurd."
  ^^^^^^^^^^^^^
No one would disagree with this.  In fact, exactly this point in made
by Ives Goddard, who is perhaps Greenberg's severest critic, in an
article on proving relationship where he looks at Algonquian-Wiyot-
Yurok evidence.  But keep in mind the sort of evidence that Delbruck
is talking about here.  Not just lists of words with similar phonetic
forms and similar or potentially relatable meanings, which is what most
of Greenberg's evidence is.  He's referring to resemblances
in fine and specific detail among the morphological systems
of the languages in question, specifically to the inflectional
systems of the verb, noun, and pronoun, which are strikingly similar
across the older and more conservative IE languages.

If Greenberg, or anyone else, could produce evidence like this for
Amerind, or any subgroup of it, the lists of resemblant forms which
make up the bulk of his book would be unnecessary.  But without this kind
of evidence, the argument from resemblant forms remains weak.

Greenberg's critics, and historical linguists in general, are by no
means as fixated on the criterion of regular correspondence as he,
for rhetorical purposes, likes to make out.  It's not that regular
sound correspondences are the necessary and sufficient proof of
relationship--Greenberg, and Tom, are quite correct that they are
neither.  It's that the evidence from vaguely resemblant forms,
selected from the extremely large body of data available, is too
weak--we simply have no way of determining whether it is statistically
significant.  A few robust regular correspondences would strengthen this
evidence considerably by reducing the possibility that the resemblances
are simply random.

_______________________________________________________________________________

<12:7>From delancey@darkwing.uoregon.edu  Tue Aug  9 17:39:16 1994

Date: Tue, 9 Aug 1994 15:39:09 -0700 (PDT)
From: Scott C DeLancey <delancey@darkwing.uoregon.edu>
Subject: Re: Deep roots and measuring diffusion
To: darwin-l@ukanaix.cc.ukans.edu

On Tue, 9 Aug 1994, Peter H. Salus wrote:

> Well, as a historian, I want to point out the internal
> contradiction of (a) citing Delbrueck and the stating (b):
>
> >>also makes Schoenemann's point (2) moot:  no one, but no one,
> >>is insisting on absolute regularity.  (3) Of course no
>
> as it is precisely Brugmann, Delbrueck, etc., through
> Hermann Paul and Streitberg who insist on the "Ausnahmslosigkeit
> der Lautgestetze."

I think two different issues are being confused here.  Neogrammarian
orthodoxy insists on the exceptionlessness of sound change as a
theoretical principle--i.e. the claim is made that in language
change, it is sounds, not words, that change, and thus that as
a change is actually occurring, it is exceptionless.  No one, not
Brugmann, Paul, or anyone else, has ever insisted that when you
look at comparative data from related languages, you will necessarily
find the regular correspondences to be without exceptions.  No one
could, because this is clearly not true.  A major component of the
development of historical linguistics was the discovery and cataloguing
of various explanations for residue, i.e. for apparently cognate
forms which seem to be exceptions to regular correspondence.

Scott DeLancey			delancey@darkwing.uoregon.edu
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA

_______________________________________________________________________________

<12:8>From hale1@husc.harvard.edu  Tue Aug  9 22:06:19 1994

Date: Tue, 9 Aug 94 23:06:12 -0400
From: hale1@husc.harvard.edu
To: darwin-l@ukanaix.cc.ukans.edu
Subject: Deep Roots & Science

Sorry to jump in in the middle of a discussion I have
not been privy to the start of, but, at the risk of
repeating arguments made formerly (for which I apologize
in advance) and since Prof. Schoenemann states that
he is interested in hearing from historical linguists),
I would like to address a few of the issues raised
by the most recent posts of Profs. Schoenemann & Salus.

Prof. Salus notes, quite correctly, that the Neogrammarians
believed that there existed a particular type of change,
the Lautgesetz or 'Sound Law', which had as one of its
attributes 'exceptionlessness' in original application.
[I believe they were right, as do many 'modern' linguists.]
It is however quite mistaken, as even cursory familiarty
with the works of the Neogrammarians would show, to
conclude that they ever believed that the existence of
regular sound change implied that all sound correspondences
would be 'absolutely regular'.  This follows of course
clearly from the fact that in defining 'regular sound
change' they were establishing a type of historical event
which they believed, correctly in my view, had distinct
properties (e.g., obeying certain constraints) from
other types of change.  Since the other types of change
(including 'sporadic' sound change, borrowing, analogy)
by hypothesis also exist (that's what makes 'regular
sound change' so special and interesting) and since all
three of these other types of diachronic events will affect
the 'regularity' of correspondences used in comparative
linguistics, Sally was correct in noting that even
Delbrueck did not expect, nor does Neogrammarian doctrine
entail, absolute regularity of correspondences.

More troubling, since the point discussed above is a more
intellectual-history issue than anything else, is the
assertion by Prof. Schoenemann that 'common ancestry' is
to be favored over 'borrowing' in cases where either is
a possible explanation.  I should note first that I agree
with Sally that Greenberg's comparisons do not pass the
first test which linguists apply to comparative data
(before consideration of 'borrowing' vs. 'direct inheritance'
even comes up): to wit, is there anything here that needs
to be explained, i.e., is the pattern of data attested
different from randomness.  This is of course not a particular
constraint of the comparative method: this is a basic
constraint on all scientific activity.  Before a scientist
offers an 'explanation' for a pattern attested in his/her
data, s/he must establish that there is in fact a pattern
(as distinct from random distribution), for in the latter
case (i.e., in the case of randomness) there is nothing for
science to do.

At any rate, this test successfully passed (not the case, in
my view, with Greenberg's data) the scientist must then
abandon 'chance' and move to the next most restrictive hypothesis.
This is obviously 'borrowing', since we have constraints on
the types of patterns we expect borrowing to display (e.g, it
should affect free morphemes before bound ones, it should
pattern within the lexicon based on the type of contact -- we
don't expect basic numerals, kin terms and body parts to
be borrowed in the absence of more general lexical adoptions,
we can sometimes check the claim of borrowing against known
history of the peoples in question, etc.).  It is only when
borrowing and chance have been excluded that we invoke the
'last resort' hypothesis: genetic affiliation.  This follows
for the simple reason that whereas the likelihood that a
given pattern is due to borrowing can sometimes be rendered
too low to warrant borrowing as an explanation, genetic
affiliation is not similarly restricted.  If the similarities
are more than chance would allow and we were to next invoke
genetic affiliation, saving 'borrowing' for only cases where
genetic affiliation could be disproven, our work would be
much less productive.  Again, I believe this follows from
general principles of scientific methodology.

It should be clear that 'related' is being used in two
completely different ways in the discussion I've seen: first
to mean 'demonstrably related using empirical methods' and
the other to mean 'having a common origin'.  All historical
linguists recognize that there may be many language related
in the second sense whose relatedness is of no scientific
value because it cannot be empirically demonstrated (and
therefore cannot be built upon for further linguistic -- or
other -- scientific work).  'Relatedness' in the absence
of scientific demonstration of the relationship is a
matter of faith, not science.  I don't have a problem with
people adopting whatever nonscientific beliefs they may
favor, but the scientific discourse has to distinguish
between personal prejudice/belief and demonstrated
scientific relationships.

Mark Hale
Dept of Linguistics
Harvard University

_______________________________________________________________________________

<12:9>From mghiselin@casmail.calacademy.org  Wed Aug 10 18:19:26 1994

Date: Wed, 10 Aug 94 16:44:30 PST
From: mghiselin@casmail.calacademy.org (Ghiselin, Michael)
To: darwin-l@ukanaix.cc.ukans.edu
Subject: Darwin CD ROM

          I am about to begin serious work on the next edition of the
          Darwin CD ROM which Pete Goldie and I produced a couple of
          years ago, and was wondering if any of you had any advice to
          offer.

          The first edition contains the Journal of Researches, Origin
          of Species, and Descent of Man, all in the final editions,
          plus the full text of my 1969 book a chronology,
          bibliography and the like.  For the next edition we want to
          produce variorum versions of Origin and Descent and add The
          Expression of the Emotions and other works depending on what
          becomes available.  A great deal of progress has already
          been made extending the bibliography and improving the
          chronology.

          The text of the Monograph on the Sub-Class Cirripedia is now
          being keyed in, and we hope to do some really creative
          things with this one, such as provide synonymies and various
          things that might serve as a readers' guide.  We chose the
          monograph(s) because as a natural history museum we at the
          California Academy of Sciences are in a particularly good
          position to do this.  Furthermore it will help to give
          systematics and phylogenetics in particular the kind of
          appreciation that seems needed.

          The book on coral reefs is high on my list of priorities
          too.

          I am applying for grant support for this project and just
          expressions of interest might help.

_______________________________________________________________________________

<12:10>From sturkel@cosy.nyit.edu  Thu Aug 11 10:54:46 1994

Date: Thu, 11 Aug 1994 11:50:35 -0400
From: sturkel@cosy.nyit.edu
To: darwin-l@ukanaix.cc.ukans.edu
Subject: Re: Darwin CD ROM

Variora of the Origin and the Descent would be great!

Another very helpful aid would be illustrations of the species of
plants and animals mentioned.  Most of the students have not seen these
organisms and many do not have access to good atlases.

sturkel@cosy.nyit.edu

_______________________________________________________________________________

<12:11>From john.wilkins1@udev.monash.edu.au  Sun Aug 14 18:51:08 1994

Date: Mon, 15 Aug 1994 09:48:28 +1000
From: John Wilkins <john.wilkins1@udev.monash.edu.au>
Subject: Re: Darwin CD ROM (sturkel@cosy.nyit.edu)
To: darwin-l@ukanaix.cc.ukans.edu

>Variora of the Origin and the Descent would be great!
>
>Another very helpful aid would be illustrations of the species of
>plants and animals mentioned.  Most of the students have not seen these
>organisms and many do not have access to good atlases.

And a concordance? That should be relatively simple to set up with the
appropriate software.

_______________________________________________________________________________

<12:13>From witkowsk@cshl.org  Mon Aug 15 08:42:32 1994

Date: Mon, 15 Aug 1994 09:43:40 -0500
To: darwin-l@ukanaix.cc.ukans.edu
From: witkowsk@cshl.org (J. A. Witkowski - Banbury Center, CSHL)
Subject: Re: Darwin CD ROM

>          I am about to begin serious work on the next edition of the
>          Darwin CD ROM which Pete Goldie and I produced a couple of
>          years ago, and was wondering if any of you had any advice to
>          offer.

Is there any chance of including materials from the CUP Correspondence of
Charles Darwin? Some sort of cross referencing between what Darwin used in
his books and his sources would be interesting. In any case this is a
really worthwhile project. One thing that comes over from the
Correspondence is that Darwin was a very great scientist, not just an
amateur naturalist. He was continually exploring and experimenting,
proposing hypotheses and testing them.  He deserves the sort of treatment
that you are giving him.

Jan A. Witkowski, Ph.D.
Banbury Center
Cold Spring Harbor Laboratory
PO Box 534
Cold Spring Harbor, NY 11724-0534
(516) 549-0507
(516) 549-0672 [fax]
witkowsk@cshl.org

_______________________________________________________________________________

<12:14>From Neve@ecol.ucl.ac.be  Tue Aug 16 02:26:44 1994

Date: Tue, 16 Aug 94 09:29:49 +0200
To: Darwin-L@ukanaix.cc.ukans.edu
From: Neve@ecol.ucl.ac.be
Subject: Musical on Darwin

Dear Networkers,

At the moment, I am reading the interesting biography of the Huxley family,
by Ronald W. Clark ("The Huxleys", Heinemann, London, 1968).

On 24 November 1959, Julian and Juliette Huxley were in Chicago for the
celebration of the Centennial of the publication of Darwin's "Origin of
Species". Clark mentions that the Centennial, apart from symposia, lectures
and debates, also included the presentation of "Time Will Tell", a musical
play based on the life and times of Darwin.

The great debate of Saturday, 30 June 1860, between bishop Wilberforce and
Thomas Henry Huxley, forms a high spot in which THH sang that

  I don't see that the Bishop has reason to sneer,
  And I have no wish to abuse him;
  But taking his line,
  If I had to incline,
  Would I choose him ?

Unfortunately Clark does not give any more details on the musical in his
book, and Julian Huxley does not mention it in his autobiography (Memories
II, George Allen & Unwin, London, 1973).

Does anybody on Darwin-L have more details on "Time Will Tell" (author,
composer, performers, ...) ? Has a recording of it ever been published ?

Any information welcome.

Gabriel

===========================================================
Gabriel NEVE                                  o   o
Unite d'Ecologie et de Biogeographie           \ /
Universite Catholique de Louvain           ***  Y  ***
Croix du Sud 5                            *   * I *   *
B-348 Louvain-la-Neuve                    *    *I*    *
Belgium                                   *    *I*    *
                                          *   * I *   *
EMAIL: NEVE@ECOL.UCL.AC.BE                 ***  I  ***
Fax  : +32/10/473490
Tel  : +32/10/473495
"The death of the butterfly is the one drawback to an
entomological career"
 - Margaret E. Fountaine (1892)
===========================================================

_______________________________________________________________________________

<12:15>From d_baum@huh.harvard.edu  Tue Aug 16 15:00:57 1994

Date: Tue, 16 Aug 94 16:00:50 EDT
To: darwin-l@ukanaix.cc.ukans.edu
From: d_baum@huh.harvard.edu ("David Baum")
Subject: The one-way mirror

I have observed an interesting sociological/psychological phenomenon in
systematics and was wondering whether a similar one exists in other historical
sciences.  I call this phenomenon the "one-way mirror."  First I need to give
some background.

Among phylogeneticists/cladists (i.e., those systematists who consider
themselves intellectual descendants of Hennig) I perceive the existence of two
interpretations of what systematics is all about:

1) Systematics is concerned with detecting and summarizing the hierarchic
relationships among taxa.  In so doing all unnecessary assumptions are
avoided.  In particular, assumptions about the causes of the hierarchic
patterns observed. (i.e., evolution) are not permitted - the argument being
that by so doing the patterns can validly be used to "test" the theory of
evolution. Because evolution is not assumed, the entities studied (i.e.,
taxa/species) may not be defined historically.  Instead taxa/species are
defined based on the distribution of characters.  In a logical sense,
therefore, all that is studied in systematics is the distribution of
characters.  Thus, taxa are viewed as properties of characters rather than
the reverse (Nelson, 1989).  Given this view of taxa the first sentence of
this paragraph may be rewritten: Systematics is concerned with detecting and
summarizing hierarchic patterns in the distribution of characters.  Hence I
call this approach the character-based view.  It approximates the view called
by some "pattern-cladism."

2) Systematics is concerned with reconstructing the HISTORICAL relationships
among organism and taxa.  Taxa/species are defined on the historical
relationships among the constituent organisms.  Characters are viewed as
merely EVIDENCE used in discovering taxa and species rather than as DEFINING
ATTRIBUTES.  Under this view it is assumed that evolution occurred and, hence,
our understanding of the evolutionary processes is validly used to refine our
methods for detecting historical relationships and thus in delimiting taxa.
Among these methods "claidistic analysis" using parsimony is certainly
important but it is by no means the only approach.  Approaches using distance
methods, maximum likelihood etc. are permissible.  Similarly some evidence
can be gained from information other than characters, e.g., biogeography,
palaeontology, breeding relations etc.  Nonetheless. because taxa/species are
defined based on history, and history is not directly knowable, there will be
some taxa/species whose existence could not ever by determined and some groups
that we incorrectly conclude to be taxa/species.  In view of its emphasis on
history, I call this second approach the history-based view of systematics.

At the end of this message I have attached a table that may help to highlight
the differences between these two approaches.  Here, I do not want to get into
the argument as to which of these views is correct.  My point here is merely
that they are distinct and each is logically consistent by which I mean that
neither makes any illogical statements.  For further discussion of the two
approaches see de Queiroz and Donoghue (1990: Cladistics).

Are intermediate positions acceptable?  For example, would it be valid to
claim: "Groups of organisms with a unique common history and groups with
diagnostic characters have the same boundaries (are "coextensive"), therefore
BOTH characters and history may be used to define taxa?"

It would ONLY be logically valid if the following statement were rejected
(and/or key terms in it redefined): "The genealogy of organisms is
independent of the characters they manifest, hence organisms with similar
characters need not be related historically."  However, rejecting this
statement would also imply rejection of the history-based view described
above.  Thus the position would no longer be intermediate.

Thus I have made two arguments so far: 1) there are two distinct views of
systematics, 2) intermediate positions are untenable.  Now to introduce the
one-way mirror phenomenon.  Proponents of the character-based view simply do
not, and I believe cannot, "see" the existence of the history-based view
whereas proponents of the history-based view (such as myself) can "see" both
approaches.  I feel like I am in a room looking into another room through a
one-way mirror and the people in that room cannot see the room I am in.  (the
analogy is imperfect as they believe they CAN see me in the room with them,
whereas I see them in a separate room).  For example, Nelson and Patterson,
1993 (Biol. Phil. 8:441-444) say in response to Donoghue's 1990 (Biol. Phil.
5:459-472) discussion of the two views of systematics:

"We do not see Donoghue's idealogical boundary, between characters and descent.
We accept as true that, in his words, taxa are defined by characters, and that
both characters and taxa owe their existence to descent.  Donoghue's
idealogical competitors are not in competition." (p. 441-442)

In other words they claim that there is no distinction between using characters
or descent (=history) to define taxa because characters' existence is due to
history.  By analogy it would be like discussing whether Bob O'Hara exists as
an individual due to a) the interconnectedness and common history of the cells
of his body or, b) the fact that he has certain emergent properties (e.g.,
running DARWIN-L) but then concluding that there is no difference between these
alternatives because he could only have emergent properties if his cells were
interconnected.  Defining Bob based on interconnectedness or emergence is NOT
the same!!  (Whichever is better).  Similarly defining taxa based on characters
vs. history is NOT the same.  Why, then, can Nelson and Patterson not see this?

The complete inability to see the distinction between character-based and
history-based views is intriguing.  How can a distinction that, even in the
abstract is totally distinct, not be apparent to some (very intelligent)
scientist?  I am mystified.  Maybe somebody out there can explain it.  Also, I
wonder whether anyone has come accross analogous blind-spots in other fields.

David Baum
D_Baum@HUH.Harvard.edu

P.S Here is the table I promised (I don't stand by every entry - many need
further discussion).  I hope the formatting survives.

                           Character-based         History-based

Higher Taxa defined:        Synapomorphy           Monophyly

Species defined       Basal, diagnosable group     Basal, monophyletic (or
                                                   exclusive group) - I could
                                                   write more here!!!

Characters are:       Defining attributes of taxa    Attributes observed by
                                                     humans - evidence
                                                     of history

Homology                =Synapomorphy           Relative propinquity of
                                                descent of character-states(?)

Other evidence relevant to
identifying taxa?             None             Yes: biogeography, ecology etc.

Methods for phylogenetic
analysis  :               Parsimony only       Parsimony/distance/likelihood...

Statistics appropriate?    No! If the          Yes, for evaluating the support
                           characters have     for the historical relationships
                           been scored right   hypothesized
                           there is no need

Speciation involves:     Fixation of a new      Isolation followed by the
                             Character          extinction of gene-lineages

                                                shared with organisms outside
                                                the species.
All organisms are
in a species                   Yes              No (some are in no basal
                                                group) - needs more
                                                discussion another time.

Overarching philosophy:    Operationalism       (Naive?) Realism

__________________________________
David Baum
Harvard University Herbaria
22 Divinity Ave
Cambridge, MA 02138
Tel:(617)496-6744/496-8766
Fax:(617)495-8944
D_Baum@HUH.Harvard.edu
__________________________________

_______________________________________________________________________________

<12:16>From TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU  Wed Aug 17 07:56:23 1994

Date: Wed, 17 Aug 1994 05:31:39 -0700 (PDT)
From: "Elihu M. Gerson" <TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU>
Subject: Re: The one-way mirror
To: darwin-l@ukanaix.cc.ukans.edu

The one-way mirror which David Baum describes is not so simple. It is
generally a poor research strategy for sociologists, historians, etc
to question the motives or intellectual capacities of opponents in a technical
debate, because such a strategy cuts off research and blocks the possibility
of discovering a superior interpretation.

In the case of "pattern cladists" such as Nelson and Patterson, it is
a good idea to start with their position as quoted by Baum, and see what
else they have to say on the issue. Nelson has pointed out many times, e.g.,
that the *only* data biologists have for constructing phylogenies is
character data (he and Patterson have an argument against the privileged
use of paleontological data too). Constructing good classifications on
the basis of this character data is thus (in the pattern cladist view)
the necessary first step in any further work. Thus, introducing hypotheses
about descent before the classification is complete can lead to circular
reasoning, and hence should be avoided.

This is a very stringent position, but it isn't illogical or
counterfactual. Pattern cladists might claim that there is a one-
way mirror between them and descent-oriented systematists, who don't
"see" that they use character and only character data in constructing
phylogenies, and thus are reasoning in a circle when they begin with
ideas of descent. I think students of this kind of debate (historians,
sociologists, etc) have the problem of understanding their organization
(e.g., how many one-way mirrors? How arranged?) and the reasons for that
organization (e.g., what role for disciplinary boundaries? academic
customs? differences between academic and museum settings?).

Elihu M. Gerson
Tremont Research Institute
458 29 Street
San Francisco, CA 94131
415-285-7837  tremont@ucsfvm.ucsf.edu

_______________________________________________________________________________

<12:17>From d_baum@huh.harvard.edu  Wed Aug 17 10:25:12 1994

Date: Wed, 17 Aug 94 11:25:05 EDT
To: darwin-l@ukanaix.cc.ukans.edu
From: d_baum@huh.harvard.edu ("David Baum")
Subject: Re: The one-way mirror

Elihu Gerson misunderstands what I mean by the "one-way mirror" and I
wanted to clarify the issue before it gets overly confused.  He seems to
think that the one-way mirror involves pattern cladists not being able to
see that the history based view is correct.  This is not my point - they do
not even see that a history-based view exists!

My point is that whereas I can see the logical consistency of
character-based approaches (and admire it for that if, in the end,
disagreeing), character-based cladists seem unable to see any alternative
to their views.  Their logic excludes the descent/character dichotomy
out-of-hand.  This is frustrating - how can one debate the virtues of two
positions if your opponents deny that the positions are different!

I perhaps made the mistake of viewing the one-way mirror as a
psychological/sociological phenomenon.  Perhaps it rests in the very logic
underpinning the character-based view.  Either way it makes it very hard to
continue a constructive debate one the issues.
__________________________________
David Baum
Harvard University Herbaria
22 Divinity Ave
Cambridge, MA 02138
Tel:(617)496-6744/496-8766
Fax:(617)495-8944
D_Baum@HUH.Harvard.edu
__________________________________

_______________________________________________________________________________

<12:18>From TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU  Wed Aug 17 13:12:58 1994

Date: Wed, 17 Aug 1994 09:59:10 -0700 (PDT)
From: "Elihu M. Gerson" <TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU>
Subject: Re: The one-way mirror
To: darwin-l@ukanaix.cc.ukans.edu

I didn't misunderstand Baum's position; I pointed out two flaws in it,
both of which remain. They are:

1) Ad hominem argument is a poor strategy for understanding debates. This
is in addition to the point that ad hominem argument is fallacious; it is
a point about the methods of science studies, not the methods of the
scientists studied.

2) I think the claim that Nelson, Patterson, and other "pattern cladists"
don't see that a history-based view exists is false. These authors have
each written many many papers concerned with evolutionary theory. They
have many objections to different aspects of evolutionary theorizing, but
they've always accepted the descent-with-modification hypothesis. Their
objection to systematic methods based on the hypothesis is that *starting*
with the hypothesis can lead to circular reasoning.

There is a symmetry with Baum's position here: pattern cladists could
easily claim that history-oriented biologists "can't see" the problems
which arise from circular reasoning. It is certainly the case that such
biologists only rarely (very very rarely) explain why this criticism is
invalid. If it is. The same issue is at the heart of considerable debate
on another front-- the analysis and role of homology. For this, see
B.K. Hall, ed. _Homology_ San Diego, Academic Press, 1994. This volume
commemorates the 150th anniversay of Owen's distinction between homology
and analogy, and contains some interesting historical material.

Perhaps the difficulty to which Baum refers arises from constructing
the distinction as a logical alternative between character and descent views.
If one starts with the assumption of descent, then there are many
interesting and important questions about characters which follow.
Conversely, if one starts with characters, one can perform many different
kinds of phylogenetic analysis. There is no reason why one cannot move
back and forth between the two starting assumptions in different studies
(or have the different studies going on in parallel). There is no
*necessary* contradiction between these two approaches. On the contrary--
in the logic of the comparative method, they should come up with mutually
supportive results. If the results conflict, then that is very strong
evidence against the validity of the theory (the descent hypothesis in
this case).

So the opposition between character and descent views is a misconstruction,
and the pattern cladists don't fail to see that the descent-oriented
approach exists. And the idea that disagreement implies mental defect
isn't acceptable.

Elihu M. Gerson
Tremont Research Institute
458 29 Street
San Francisco, CA 94131
415-285-7837  tremont@ucsfvm.ucsf.edu

_______________________________________________________________________________

<12:19>From Kim.Sterelny@vuw.ac.nz  Wed Aug 17 18:06:36 1994

Date: Thu, 18 Aug 1994 11:06:25 +1200
To: darwin-l@ukanaix.cc.ukans.edu
From: Kim.Sterelny@vuw.ac.nz
Subject: Re: The one-way mirror

On Wednesday , 17 Aug 1994  Elihu M. Gerson wrote:

>I didn't misunderstand Baum's position; I pointed out two flaws in it,
>both of which remain. They are:
>
>1) Ad hominem argument is a poor strategy for understanding debates. This
>is in addition to the point that ad hoinem argument is fallacious; it is
>a point about the methods of science studies, not the methods of the
>scientists studied.
>
>2) I think the claim that Nelson, Patterson, and other "pattern cladists"
>don't see that a history-based view exists is false. These authors have
>each written many many papers concerned with evolutionary theory. They
>have many objections to different aspects of evolutionary theorizing, but
>they've always accepted the descent-with-modification hypothesis. Their
>objection to systematic methods based on the hypothesis is that *starting*
>with the hypothesis can lead to circular reasoning.
>
>There is a symmetry with Baum's position here: pattern cladists could
>easily claim that history-oriented biologists "can't see" the problems
>which arise from circular reasoning. It is certainly the case that such
>biologists only rarely (very very rarely) explain why this criticism is
>invalid. If it is. The same issue is at the heart of considerable debate
>on another front-- the analysis and role of homology. For this, see
>B.K. Hall, ed. _Homology_ San Diego, Academic Press, 1994. This volume
>commemorates the 150th anniversay of Owen's distinction between homology
>and analogy, and contains some interesting historical material.
>
>Perhaps the difficulty to which Baum refers arises from constructing
>the distinction as a logical alternative between character and descent views.
>If one starts with the assumption of descent, then there are many
>interesting and important questions about characters which follow.
>Conversely, if one starts with characters, one can perform many different
>kinds of phylogenetic analysis. There is no reason why one cannot move
>back and forth between the two starting assumptions in different studies
>(or have the different studies going on in parallel). There is no
>*necessary* contradiction between these two approaches. On the contrary--
>in the logic of the comparative method, they should come up with mutually
>supportive results. If the results conflict, then that is very strong
>evidence against the validity of the theory (the descent hypothesis in
>this case).
>
>So the opposition between character and descent views is a misconstruction,
>and the pattern cladists don't fail to see that the descent-oriented
>approach exists. And the idea that disagreement implies mental defect
>isn't acceptable.

two points

i. One important issue of over pattern cladism is not whether they claim
that systematic views informed by evolution are circular, but whether they
are right to do so. If they are not right, then Baum is entitled to regard
their views as logically deficient. Mark Ridley, in his Evolution and
Classification, gives an extensive anaylsis of this claim and I think shows
its wrong. The circularity idea would be justified only if systematic
hypotheses informed by some evolutionary hypothesis insulated, or tended to
insulate that hypothesis from correction; it would be circular if there
were no way evolutionarily-informed systematics could feedback in a
correctional loop on the hypotheses that inform it. Pattern cladists, to my
knowledge anyway, have made no case that such a loop is impossible; they
typically assume that is there is feed forward from evolution to
systematics, that establishes right there that there must be circularity in
any use of systematics in the evaluation of an evolutionary hypothesis. And
inspection of cladistic texts (Brooks and McLennan, or Harvey and Pagel) I
think supports Ridley's views here.

2. I think it rather unfortunate that Gerson keeps taking Baum to task for
ad hominem language; if any of the two have a hectoring face, it's Gerson.

kim sterelny
philosophy
victoria university of wellington

_______________________________________________________________________________

<12:20>From TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU  Wed Aug 17 20:54:57 1994

Date: Wed, 17 Aug 1994 18:25:08 -0700 (PDT)
From: "Elihu M. Gerson" <TREMONT%UCSFVM.BITNET@KUHUB.CC.UKANS.EDU>
Subject: Re: The one-way mirror
To: darwin-l@ukanaix.cc.ukans.edu

Sterelny's comments about pattern cladists' arguments are interesting,
but they don't address either my point or Baum's (as I understand it).
The issue is why (or if) pattern cladists "can't see" the historically-
oriented argument about evolution. The validity of the circularity
argument isn't at stake here; I mentioned it only in order to show that
pattern cladists could (if they wished) make a reciprocal "can't see"
argument against Baum's position.

As I noted in both previous postings, my point about ad hominem argument
is about the methods of science studies-- it's a bad strategy for those
*studying* debates (not necessarily for those participating in them) because
that kind of argument discourages further exploration of alternative
explanations of a debate and its outcomes.

Elihu M. Gerson
Tremont Research Institute
458 29 Street
San Francisco, CA 94131
415-285-7837  tremont@ucsfvm.ucsf.edu

_______________________________________________________________________________

<12:21>From sally@isp.pitt.edu  Thu Aug 18 07:51:57 1994

To: darwin-l@ukanaix.cc.ukans.edu
Subject: Re: The one-way mirror
Date: Thu, 18 Aug 94 08:51:46 -0400
From: Sally Thomason <sally@isp.pitt.edu>

Elihu Gerson writes, quoting Nelson, that "the *only* data biologists
have for constructing phylogenies is character data".   Does
"character data" include only the characters themselves, or does
it also include hypotheses (based on indirect evidence of what has
happened in partly analogous instances of descent with modification)
about directionality, such as "X is likely to change to Y [in the
presence of character Z] but not vice versa"?

Also -- sorry to ask such an ignorant question -- could someone
provide a reference or two to places where Nelson and/or Patterson
discuss these issues?  Maybe I'm not the only non-biologist who
doesn't know where to start.

   Sally Thomason
   sally@isp.pitt.edu

_______________________________________________________________________________

<12:22>From ahouse@hydra.rose.brandeis.edu  Thu Aug 18 08:29:37 1994

Date: Thu, 18 Aug 1994 09:31:50 -0400
To: darwin-l@ukanaix.cc.ukans.edu
From: ahouse@hydra.rose.brandeis.edu (Jeremy Creighton Ahouse)
Subject: Re: The one-way mirror (intro refs)

>  Also -- sorry to ask such an ignorant question -- could someone
>provide a reference or two to places where Nelson and/or Patterson
>discuss these issues?  Maybe I'm not the only non-biologist who
>doesn't know where to start.
>
>   Sally Thomason
>   sally@isp.pitt.edu

        Though this isn't a reference to Nelson or Patterson and Kim
Sterelny mentioned this book in his post I want to encourage those who are
interested but unfamiliar with this discussion to grab a copy of Evolution
and Classification: The Reformation of Cladism by Mark Ridley (Longman
Group Limited 1986 [QH83.R49 1986]).  This book is a quick read.  Ridley
manages to criticize the pheneticists and "evolutionary" taxonomists, make
the case for cladistics, and then argue against the "extension"
(regression?) of cladistics to natural order systematics (yet another
synonym for pattern cladistics).

        This discussion should interest those who are interested in
applying Kuhn's notion of science change to a particular area in biology.
There was (it seems to me) a real sense of the incommensurable when
cladistics first faced off with the "evolutionary" taxonomists.  The
pronouncements by the evol taxonomists  (oft represented by E. Mayr) felt
mysterious and not particularly coherent to  cladists.  But with work (see
Ridley) it is possible to extract the goals of taxonomists who would blend
together both phenetic and phylogenetic approaches.  So maybe we don't
really have strong incommensurability but rather a good old fashioned
dispute.  (Though one gets the sense that evol tax supporters don't quite
"get" the excitement engendered by phylogenetic systematics.)

        David Baum notes a single direction inability to admit (even if to
disagree with it) the position of cladists by transformed cladists.  Is
this an incommensurable disconnect or just obstinance or a rhetorical move
or something else altogether?

        - Jeremy

p.s. must mention David Hull's _Science as Process_ esp chapters 4-7.  This
160 pp. will give you a good read for a Sunday afternoon.  Hull provides
the history of these arguments and some of the personalities.

_________________________________________________________________
Jeremy Creighton Ahouse (ahouse@hydra.rose.brandeis.edu)
Biology Dept.
Brandeis University
Waltham, MA 02254-9110
(617) 736-4954
(617) 736-2405 FAX
__________________________________________________________________

.. animals are divided into (a) those that belong to the emperor, (b)
embalmed ones, (c) those that are trained, (d) suckling pigs, (e) mermaids,
(f) fabulous ones, (g) stray dogs, (h) those that are included in this
classification, (i) those that tremble as if they were mad, (j) innumerable
ones, (k) those drawn with a very fine camel's hair brush, (l) others, (m)
those that have just broken a flower vase, (n) those that resemble flies
from a distance.
    from the Celestial Emporium of Benevelent Knowledge (Jorge Luis Borges)

_______________________________________________________________________________

<12:23>From MNHPB019@SIVM.SI.EDU  Thu Aug 18 08:49:01 1994

Date: Thu, 18 Aug 1994 09:40:59 -0400 (EDT)
From: "Nancy E. Todd" <MNHPB019@sivm.si.edu>
Subject: Re: The one-way mirror
To: darwin-l@ukanaix.cc.ukans.edu

      ETE Program, Dept. of Paleobiology
      Smithsonian Institution

Character data, as Sally Thomson asks, can include hypotheses of descent if one
is analyzing the data through tranformation series, in which you determine and
set which character state leads to which, i.e. a - b - c - d. You end up doing
this for each character. You can also run the data unordered.

My own personal argument about multistate characters is that if you do
transformation series, then you are overlaying an idea of descent on the
dataset because you are hypothesizing the track of character change. I would
hope that the counterargument for this would be that transformation series are
to be used as a "tool" to examine patterns in the dataset after the initial
trials. I have been discussing this for a couple of years now with one of my
advisors, and we can't really agree on it, although we each understand the
other argument.

I have been interested in this argument, and am responding for the first time,
having lurked about in the background for almost as long as Darwin-L has been
around. The dichotomy that Baum raised does seem to be a problem on the
surface, but I would argue that it is not a one-way mirror, particularly if one
is working with paleontological specimens. I am working with fossil elephants,
many of which are defined by morphological characters that are parts of
evolutionary trends. In this case, there is no getting around the fact that the
characters are historically based, although I hope to do so. In fact, if I run
the characters ordered in the direction in which the characters trend, the
resulting tree does not match the trends. With this type of circular analysis,
one would think that the tree would plot the hypothesized direction of the
trends. Thus, I feel that it is impossible to separate history from character
based views in this case. I hope I am wrong.

Nelson and Patterson tend to be very theoretical about some of their work, and

_______________________________________________________________________________

<12:24>From MNHPB019@SIVM.SI.EDU  Thu Aug 18 08:49:11 1994

Date: Thu, 18 Aug 1994 09:47:23 -0400 (EDT)
From: "Nancy E. Todd" <MNHPB019@sivm.si.edu>
Subject: Re: The one-way mirror
To: darwin-l@ukanaix.cc.ukans.edu

      ETE Program, Dept. of Paleobiology
      Smithsonian Institution

I am very sorry, I slipped and hit send instead of add a line.

Just to finish my statement. We need to be able to comprehend the theoretical
and also be able to use the practical applications of character-based and
historical views.

I would be happy to discuss this further with anyone off the list.
Comments and suggestions are welcome.

Nancy Todd
mnhpb019@sivm.si.edu

_______________________________________________________________________________

<12:25>From mwinsor@epas.utoronto.ca  Thu Aug 18 10:25:52 1994

From: Mary P Winsor <mwinsor@epas.utoronto.ca>
Subject: about "ad hominem"
To: darwin-l@ukanaix.cc.ukans.edu (bulletin board)
Date: Thu, 18 Aug 1994 11:25:16 -0400 (EDT)

From Polly Winsor mwinsor@epas.utoronto.ca.  Warning, this is pretty
long. I was interested in Gerson's comment that saying Patterson
couldn't see his opponents's views as clearly as the opponents saw his
was an ad hominem judgement, and thus one Gerson would avoid making.

Eli Gerson makes a distinction of great interest to me,
for I am an historian of biology, and I try to teach
graduate students how to be historians of science.
He says there exist two distinct worlds: one consisting
of scientists, and the other of those study about the
scientific enterprise (historians, sociologists,
philosophers).  For the purposes of this current
discussion, I'll call the two worlds S (science) and
C (commentators on science).

Gerson says that within S, it is sometimes successful
strategy for a debater to use ad hominem argument
("Professor X has a blind spot in his mental apparatus,
so you need not give credence to what he says").  Notice
that Gerson is speaking as a member of C world, because
within S world the rules of debate say "use only facts
and reasoning, not the Medieval scholastic tricks of
appeal to authority, rhetoric and so on - when you
catch someone using ad hominem, denounce them for using
unscientific tactics."  But as a solid member of C, Gerson
knows that scientists break their rules all the time.  Much
literature in C claims that not only do members of S break
the rules [this is demonstrably so], but it works, that is,
debates are sometimes decided that way rather than by the
weight of evidence.  [This claim is harder to demonstrate,
since there are always multiple factors at work, so
evaluating the actual contribution of one is hard.]

Notice that there are deep issues of value at stake:
members of S believe that their ideals are Good and to
break the rules is Wrong and to win an argument thus is
Cheating, whereas members of C would claim that, as neutral
observers, they must abstain from such judgements, limiting
themselves to declaring a strategy "successful."

Gerson explains that within C, members of C have learned
that it is poor strategy to employ individual human
differences (genius, stupidity, courage) to explain events
in S; I teach my students the First Rule: that their prime
working assumption must be the reasonableness of the person
whose writings they are analyzing.  (This is like Lyell's
assumption that natural forces just like those at work around
us should be what we use in explanations of the past.  If we
read someone who is insisting that worms are spontaneously
generated, we must seek to understand why any reasonable
person in his position would think such a thing, and this
leads us to discover that the array of facts before him,
plus the absence of facts we take for granted like how small
organisms reproduce, do make the belief unsurprising after all).

Things are not so simple, however.  Both within C and within S,
the reality is not as neat as I described it, and things really
get complicated when we include the relationship between S and C.

1. The ideal of S is modelled on very simple mathematical systems
like Euclidean geometry and Archimedean statics, which Copernicus,
Kepler, Galileo, Descartes, and Newton used with dazzling success.
Within mathematics, you don't debate a proof, though you might
demonstrate a flaw in the logic, and expose the error of what was
thought to be a proof.  Against such a simplistic ideal, no wonder
the actual practice of science can always be shown to be much
messier.  Members of S do not, and never have, limited themselves
strictly to fact and reason (nor could they, as is obvious if you probe
into what the words "fact" and "reason" can realistically mean).  Of
course authority and rhetoric are at work within S. [but how important
is another issue]
>
2. Yet also within C, things are much messier than Gerson implies.
First of all, this First Rule (which warns against "ad hominem"
judgements like "he didn't comprehend") is merely wise advice, urging
one to search for all relevant factors, rather than being information
that humans are all identical.  Their individual differences, which
actually do include character and intelligence, do affect their actions.
In the world of C are people with very different approaches, and some
would say history is a social science and some would say it is one of
the humanities.

3. But the complications have only just begun.  Members of S are quite
free to dip into C for information to help them get on within S.  That
might mean reading or talking to members of C, but it might also mean
taking time out of the lab to think about what they are doing, the past
of their field, the structure of current debates and so on.

So here we have the interesting case of a member of S asking members of
C for assistance in a current debate.  ("Tell me the philosophical or
sociological labels to help me describe the stubbornness of pattern
cladists.")  Then the member of C says "if you ignore the First Rule, I
will not recognize you as a colleague within C with whom to discuss S
objectively, though of course I would not be surprised if within
S you fight with the gloves off."  Gerson may be nervous that any
comments from C on the substance of the debate will be put to use.

I know about nervousness of that kind, for I uncovered, in the
archives of the Systematics Association in the late 1930s, the record of
a debate centering on this same "circular-argument-if-you-assume-
evolution" idea.  Even though there was no clear winner (World War II
intervened) I am learning that it is impossible to tell the story
without my own symptathies (for phylogeny) showing through.

Polly Winsor    mwinsor@epas.utoronto.ca

_______________________________________________________________________________
Darwin-L Message Log 12: 1-25 -- August 1994                                End