5.0278 Dendrology and Cladistics (4/269)

Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Thu, 15 Aug 1991 15:57:33 EDT

Humanist Discussion Group, Vol. 5, No. 0278. Thursday, 15 Aug 1991.

(1) Date: Sat, 3 Aug 1991 15:40:46 EDT (165 lines)
From: Robert O'Hara <MNHVZ028@SIVM>
Subject: Textual Criticism Challenge

(2) Date: Sun, 4 Aug 1991 12:16:17 EDT (59 lines)
From: Robert O'Hara <MNHVZ028@SIVM>
Subject: Linguistics, stemmatics, cladistics

(3) Date: Mon, 12 Aug 91 18:50:50 CST (27 lines)
From: (James Marchand) <marchand@ux1.cso.uiuc.edu>
Subject: dendrology

(4) Date: Wed, 14 Aug 91 15:39:42 EDT (18 lines)
From: dthel@conncoll.bitnet
Subject: dendrology-addendum

(1) --------------------------------------------------------------------
Date: Sat, 3 Aug 1991 15:40:46 EDT
From: Robert O'Hara <MNHVZ028@SIVM>
Subject: Textual Criticism Challenge

I found Peter Robinson's "Textual Criticism Challenge 1991" (to
reconstruct a stemma for a collection of Old Norse MSS) very exciting
for a number of reasons, and I hardly know where to begin. I had been
considering bringing up this general topic on HUMANIST for a while, but
didn't feel I had the time at this particular moment. I still don't
have the time, but I'm going to do it now anyway. :-)

My training is in evolutionary biology, and I am much interested in the
relationships between the humanities and the sciences; in particular,
between history and the "historical sciences" of cosmology, geology, and
evolution. One of the most active areas of theoretical research in
evolutionary biology in the last 25 years has been the study of
phylogeny, the evolutionary history of life. If we accept that
evolution has occurred, then all of the millions of species of organisms
we see around us today are related to one another in a great tree of
life, an evolutionary genealogy that we can, in principle, reconstruct.
The reconstruction of this great tree of life, the reconstruction of
evolutionary history, is the concern of the field known as
_systematics_, or more specifically _cladistics_, and these are my
particular specialties. (Systematics is to cladistics about as
philology is to stemmatics.)

The reconstruction of evolutionary trees is an inferential discipline.
We do not directly observe sequences of past events; we use the pattern
of similarities and differences that can be observed among organisms in
the present to infer the sequence of steps by which these current morpho-
logies arose, and the sequence of splitting events (speciations) that
led to the presently observed diversity. Now as you may be able to see
already, there are many parallels, some nearly exact, between the
reconstruction of phylogeny (evolutionary trees), and the reconstruction
of both linguistic trees and stemmata of manuscripts. Peter Robinson
suggests a variety of conclusions anyone taking up his challenge might
be able to arrive at, and many of these are the very same conclusions
a systematist trying to estimate a phylogeny would try to reach. Peter's
"genetic groups" are what we would call _clades_ (Gr. clados, "branch").
Rather than speaking of a word or phrase that exhibits a variety of
readings, we speak of a _character_ that exhibits several _states_; characters
are the raw data, the observed differences, from which we infer past
events. The "readings characteristic of particular groups" are what
we would call the "apomorphies" or "derived character states" of a clade.
"Contamination" in stemmatics is what systematists would call "horizontal
transmission" or hybridization. It is very much a nuisance in systematics
just as it is in stemmatics. "Readings independently conceived by
different scribes" are what we call convergences or "homoplasies" (Gr.
"molded the same"). And "readings that have spread by virtue of the
common descent of all these manuscripts from a single parent manuscript"
are what we would call "plesiomorphies" or "primitive character states."
Primitive character states/readings do not define clades/genetic groups:
if you have two sets of organisms/MSS, one with state/reading "A" and the
other with state/reading "B", you have to know which of these is the
primitive state and which is the derived state in order to tell which
set of organisms/MSS is the clade/genetic group, derived from somewhere
within the other group. In systematics the determination of the direction
of change between two or more character states is called "polarity
determination", and is ordinarily done according to a method called
"outgroup comparison". I don't think stemmatics has an exact parallel to
outgroup comparison; instead, "reading polarity" is judged according to
various principles like "lectio difficilior", etc. Systematists have also
developed a variety of computer programs for analysing character data
and reconstructing evolutionary trees, but I won't say too much about them
because that's the approach I'm going to use when I take a crack at
Peter's challenge, and I don't want to give all my secrets away. :-)

I have followed as an amateur the recent popular literature in linguistics
on the "Nostratic" and "proto-World" questions, and there are many additional
parallels to be found there between systematics and historical linguistics.
The technique advocated by Greenberg called "mass comparison" is very like
the systematic method, now largely discredited, called "phenetics". Further,
the linguistic school called "glottochronology", popular I believe in the
1960s, made assumptions similar to those made by systematists who use
biochemical data and argue for the existence of a "molecular clock" that
can be used to date past evolutionary events.

The literature on phylogeny estimation in systematics has increased
enormously in the last few years, and it has from time to time gotten
very rancorous. Many of the detailed debates on these issues have
taken place in the journals _Systematic Zoology_ and _Cladistics_.
Unfortunately there are relatively few general or introductory treatments
available, since most of these methods are still fairly new. One of
the best analyses of the theoretical foundations of phylogeny estimation

Sober, E. 1988. Reconstructing the Past: Parsimony, Evolution, and
Inference. Cambridge: MIT Press.

Sober is a philosopher, and he does a very good job of explaining how
many of the theoretical problems of phylogeny estimation are special
cases of general problems that have been discussed in the philosophical
literature for some time. I'm quite sure that anyone interested in the
theoretical aspects of either historical linguistics or stemmatics would
find much of interest in his book. A couple shorter papers that discuss
the general nature of contemporary systematics are:

de Queiroz, K. 1988. Systematics and the Darwinian revolution.
Philosophy of Science, 55:238-259.

O'Hara, R. J. 1988. Homage to Clio, or, toward an historical philosophy
for evolutionary biology. Systematic Zoology, 37:142-155.

These do not examine particular methods of reconstructing evolutionary
trees, but rather describe the nature of the systematic enterprise, and
its recent history.

The parallels between systematics and both historical linguistics and
stemmatics have not gone unnoticed, but I think that up to now it's been
like two groups of people speaking different languages -- there are
a few people interested in learning the other language, but there really
aren't any bilingual people yet who can help the others along. One
fascinating pioneering volume has appeared, but much more work of this
sort is needed:

Hoenigswald, H. M., & L. F. Wiener, eds. 1987. Biological Metaphor and
Cladistic Classification: an Interdisciplinary Perspective. Philadelphia:
University of Pennnsylvania Press.

Another interest of mine in this context has been the history of the diagrams
that have been drawn to represent evolutionary history, and natural
diversity more generally. We now draw evolutionary trees of course, but
in the pre-evolutionary period (early 19th C. and before) natural diversity
was not necessarily visualized as tree-like, and circles, stars, maps,
and other metaphors were used to represent what was called "the Natural
System" (whence the name "systematics"). If anyone knows of historical
studies in linguistics or stemmatics that have particularly examined the
development of linguistic or stemmatic tree diagrams I would be very
interested to hear of them. A reference to my own work on this topic is:

O'Hara, R. J. 1991. Representations of the Natural System in the
Nineteenth Century. Biology and Philosophy, 6:255-274. [I would be
happy to provide anyone who is interested with reprints of either
of my papers; just send me a snail-mail address.]

Indeed, the rise of historical linguistics in the late 18th and early 19th
Centuries was happening at precisely the same time as the rise of historical
thinking in geology, zoology, and botany. The best two references on this
"historical revolution" from the perspective of the sciences are:

Greene, J. C. 1959. The Death of Adam: Evolution and its Impact on
Western Thought. Ames: Iowa State University Press.

Toulmin, S., & J. Goodfield. 1965. The Discovery of Time. New York:
Harper & Row. ["In the whole history of thought no transformation in
in men's attitude to Nature -- in their 'common sense' -- has been more
profound than the change in perspective brought about by the discovery
of the past." p.17]

I will be moving to the University of Wisconsin in a few weeks, so this
month is rather hectic for me, but I hope that any of you who find these
issues interesting will give some thought to them, and if I can't respond
completely at the moment I will be sure to bring up the subject again in
a month or two. At Wisconsin I will be working with Elliott Sober, the
author of the first book mentioned above, on an NSF postdoc to examine
the theoretical foundations of the idea of "a character" in systematics,
so if anyone happens to know of any similar discussions in the textual
analysis literature on what the concept of "a reading" means I would be
glad to hear of them.

My thanks to Peter Robinson for putting forth such an interesting
challenge. I hope that many more people will take it up.

Bob O'Hara, MNHVZ028@SIVM.bitnet
National Museum of Natural History, Smithsonian Institution
(2) --------------------------------------------------------------66----
Date: Sun, 4 Aug 1991 12:16:17 EDT
From: Robert O'Hara <MNHVZ028@SIVM>
Subject: Linguistics, stemmatics, cladistics

After posting my response to Peter Robinson's "Textual Criticism Challenge"
I went back through the Hoenigswald & Wiener volume I mentioned, and it
really is quite excellent. Since the title might not particularly catch
the eye of a linguist or philologist, I reproduce here the table of contents.
I highly recommend the volume to anyone interested in the theoretical aspects
of historical reconstruction in language and manuscript studies.

Hoenigswald, H.M., & L. F. Weiner, eds. 1987. Biological Metaphor and
Cladistic Classification: an Interdisciplinary Perspective. Philadelphia:
University of Pennsylvania Press.

Part One: Historical Perspectives

1. Biological analogy in the study of languages before the advent of
comparative grammar. (W. Keith Percival)

2. The life and growth of language: metaphors in biology and linguistics.
(Rulon S. Wells)

3. "Organic" and "organicism" in Franz Bopp. (Anna Morpurgo Davies)

4. On Schleicher and trees. (Konrad Koerner)

5. A legal point. (Boyd H. Davies)

6. Haeckel's variations on Darwin. (Jane M. Oppenheimer)

Part Two: Methodology

7. Cladistic and paleobotanical approaches to plant phylogeny. (Peter R.
Crane & Christopher R. Hill)

8. Pattern and process: phylogenetic reconstruction in botany. (Peter F.

9. Characters and cladograms: examples from zoological systematics.
(Michael J. Novacek)

10. Reconstructing genetic and linguistic trees: phenetic and cladistic
approaches. (Maryellen Ruvolo)

11. Of phonetics and genetics: a comparison of classification in linguistic
and organic systems. (Linda F. Weiner)

12. The upside-down cladogram: problems in manuscript affiliation. (H. Don

13. Representing language relationships. (William S.-Y. Wang)

14. Language family trees: topological and metrical. (Henry M. Hoenigswald)

15. Computational complexity and cladistics. (David Sankoff)

Bob O'Hara, MNHVZ028@SIVM.bitnet
National Museum of Natural History, Smithsonian Institution
(3) --------------------------------------------------------------37----
Date: Mon, 12 Aug 91 18:50:50 CST
From: (James Marchand) <marchand@ux1.cso.uiuc.edu>
Subject: dendrology

When I made the statement about dendrological thinking being hard-wired
into humans, I did not even think about it being controversial. I am
opposed to dendrological thinking just because it is so pervasive. I
really know of no culture in which the members do not draw "trees" to
illustrate relationships. The ready acceptance of Darwinianism in the
19th century, though the seedbed had been prepared by Herder and Goethe
and historicism in general, shows that this sort of thinking was
widespread at least. As to the "family" metaphor, I have no idea when
it arose as applied to languages; it goes back at least to the 17th
century. The tree metaphor as applied to languages, particularly in the
form given it by Schleicher, brought about a counter- theory, the wave
theory (Schmidt), but we still draw (leafless) trees to illustrate
kinship, manuscript stemmata, language relationships, etc., without
thinking that this requires an either/or but never both (exclusive or)
kind of thinking, which I think led to the Darwinian rejection of
speciation by convergence. As humanists, we very seldom have to deal
with things which are discrete, easily lumped. As the case of Yiddish
and German shows, it is often even hard to tell when is a language. At
any rate, dendrology is an ever present danger for the humanist, be it
hard-wired into us or not. As Gottfried put it: si bernt uns mit dem
stocke schate, und nicht mit dem gruenen meienblate "They offer us shade
with the (bare) branches, and not with the green leaf of May."
Jim Marchand
(4) --------------------------------------------------------------22----
Date: Wed, 14 Aug 91 15:39:42 EDT
From: dthel@conncoll.bitnet
Subject: dendrology-addendum

The network transmission of my comment on "dendrology" somehow lost an
important line. The omission concerned Martin Bernal's asertion that the
tree metaphor applied to languages provided a powerful paradigm that excluded
the possibility of allowing the influence first, of non-IE languages on the
IE family; this preserved the purity of the "Sprache", and sewrved as the
foundation of what he calls the Aryan model. And second, the tree metaphor
for languages was extended to culture as well, excluding here too the influence
of non-IE cultural factors (he has in mind specifically Semitic and African
factors).Thus Bernal asserts in his book Black Athena that racism and
anti-Semitism were sustained by and promoted the adoption of the tree metaphor
as well as the family metaphor. I wonder how many agree with his ideological
thesis, which at the very least asserts that rather than being hard-wired as
Marchand suggested the metaphor is the child of extra-rational forces. Dirk
Held, Classics, Connecticut College.