Professional Documents
Culture Documents
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Oxford University Press, Taylor & Francis, Ltd., Society of Systematic Biologists are
collaborating with JSTOR to digitize, preserve and extend access to Systematic Zoology
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
Syst. Zool., 31(1), 1982, pp. 25-34
JOHN BEATTY
Abstract
Beatty, J. (Department of the History of Science, Harvard University, Cambridge, Mas-
sachusetts 02138) 1982. Classes and cladists. Syst. Zool., 31:25-34.-Disillusions concern-
ing evolutionary theorizing have forced a split among cladists. The disillusioned group of
"pattern" cladists seeks an evolutionarily neutral brand of cladistics. But pattern cladistics is
not, after all, evolutionarily neutral. Rather, it is at odds with evolutionary theorizing. [Classes;
cladistics; evolution; phylogenetics; Popper.]
Hull (1976, 1978) has argued that if erties that distinguish Ursus from other
species are interpreted as classes, as is genera of its family, and the properties
traditional, then a little problem arises- that distinguish its family from other fam-
namely, species cannot evolve. But ac- ilies of its order, and so on. To say that
cording to our best accounts, they do. species so designated are "classes" is just
Thus there is an inconsistency in our un- to say that they are collections of objects
derstanding of species. In Section 1, I re- that share the defining properties of the
formulate the difficulty in order to reem- species name.
phasize the problem. My ultimate Two kinds of classes, hence two kinds
purpose, however, is to discuss the dif- of class interpretations of species, are
ferences between two ever more distin- commonly distinguished (e.g., Beckner,
guishable groups of cladists in terms of 1959, Chapter 5; Caplan, 1980). The
the difficulty that this problem poses for name of an "Aristotelian class" is defined
each. In Section 2, I discuss the back- in terms of properties that are collective-
ground to the widening split between ly necessary and sufficient for member-
phylogenetic and pattern cladists. Final- ship in the group. Members of a "cluster
ly, in Section 3, I argue that pattern clad- class," on the other hand, need only sat-
ism has conceptual drawbacks that stem isfy most of the defining properties of the
from the problem outlined in Section 1. class name. So any defining property of
the name of an Aristotelian class is pres-
1. Evolutionary Theory vs. a
ent in 100% of the members of that class,
School of Systematics
while any defining property of a cluster
1.1- The sort of interpretation of species class is present in a high proportion of
that Hull finds most objectionable from members of the class.
the viewpoint of evolutionary biology is One can imagine still other sorts of
one according to which a species name class interpretations of species. For ex-
is defined in terms of the set of properties ample, the defining properties of a
operationally employed to recognize and species name need not be the usual sorts
distinguish members of that.species from of properties operationally used to rec-
other species. Thus, we recognize and ognize and distinguish members of dif-
distinguish polar bears (Ursus mariti- ferent species. They might instead be re-
mus) from brown bears (Ursus arctos) lational properties like interbreeding and
by, among other traits, their white coats. ancestor-descendant relations specified
Consequently, we might include the dif- with regard to model members of the
ferentiating property "white coat" in the group. Alternative class interpretations
definition of "Ursus maritimus." We like the latter can be worked out. But it
would also include the other properties is the interpretation in terms of the usual
that distinguish maritimus from other recognition properties that is at issue in
species of Ursus, as well as those prop- Hull's work and in the recent division
25
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
26 SYSTEMATIC ZOOLOGY VOL. 31
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
1982 CLASSES AND CLADISTS 27
evolution of species. I would argue, more class. The lineage interpretation thus
specifically, that unless we interpret skirts the difficulties of the class interpre-
species entirely genealogically, we are tation discussed earlier.
forced into the position that species can-
2. A School of Philosophy vs.
not evolve with regard to their defining
properties, and hence we are forced to Evolutionary Theory
accept the problematic consequences of Unfortunately, even after Hull's origi-
that position, as discussed above. Species nal arguments, remnants of the class in-
must be lineages if we are to make sense terpretation are still around. What is most
of the evolution of species in the manner surprising/distressing is that a class inter-
we presume to. The reasons are as fol- pretation should be espoused in recent
lows. cladist literature. Given their original
Organisms make up a lineage by virtue emphasis on representation of genealogy
of their interbreeding and ancestor-de- (Hennig, 1966), one would expect clad-
scendant relations, not by virtue of sat- ists to be most amenable to a strictly lin-
isfying the sort of membership-require- eage view of species. For instance, Wiley
ment properties traditionally thought to (1980; 1981, Chapter 2) settles the lin-
define species names. A lineage can eage-class issue decidedly in favor of the
evolve with regard to any of the latter former. But the lineage interpretation-
sort of properties and still remain the indeed, the whole emphasis on genealo-
same lineage. So a name given to a lin- gy-is too theoretical to suit the empiri-
eage is not properly defined in terms of cist standards of an ever more clearly dis-
any particular properties of that sort. To tinguishable subgroup of cladists. Let's
be sure, the organisms that constitute a call the former subgroup, of which Wiley
lineage may, at any one time, have prop- is a member, the "phylogenetic" cladists,
erties in common that are useful for rec- in recognition of their adherence to Hen-
ognizing that lineage, and for distin- nig's original goal of phylogenetic sys-
guishing it from its contemporaries. But tematics. And let's call the latter group
such distinguishing descriptions should the "pattern" cladists. I will get to the
not be confused with definitions. In this reason for that reference shortly.
respect, at least, lineages and their names In order to understand the rise of pat-
are similar to individual organisms and tern cladism, we have to consider, if only
their names. "The author of the Origin" briefly, the recent appeal of Popper's phi-
is a description that distinguishes Charles losophy of science (e.g., Popper, 1959,
Darwin from everyone else. Anyone who 1963) among evolutionists and system-
asserted that Darwin did not write the atists. The lesson usually drawn from
Origin would be viewed suspiciously, Popper is that science is readily distin-
but they would not be accused of contra- guishable from pseudoscience and non-
dicting themselves, as they would if science by virtue of the degree of falsi-
"Charles Darwin" were defined in terms fiability of scientific hypotheses, and by
of the characteristic in question. As in the virtue of the earnest intent of scientists
case of individual organisms, so too in the to submit their hypotheses to crucial
case of lineages: distinguishing descrip- tests, to reject those that fail the tests, and
tions associated with their names are not to resubmit those that pass. The logic of
definitional constraints on the respects in testing is supposedly straightforward.
which they can change. Hence, we can From the hypothesis ostensibly under
meaningfully talk about the increase in test, and from additional "auxiliary" hy-
frequency of any of the properties of the potheses, a prediction is deduced. If the
organisms that constitute a species-qua- prediction fails to obtain, it follows that
lineage, whereas we cannot meaningful- at least one of the hypotheses from which
ly talk about the increase in frequency of it was deduced is false. Of course, an os-
the defining properties of a species-qua- tensibly refuted hypothesis can be
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
28 SYSTEMATIC ZOOLOGY VOL. 31
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
CLASSES AND CLADISTS 29
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
30 SYSTEMATIC ZOOLOGY VOL. 31
xX y
3.2 What is a group? Originally, clad-
ists assumed that one branching lineage
had been produced in the course of evo-
w A B lutionary history. With respect to such a
tree, a natural group was conceived to be
FIGS. 3A and 3B. See text for explanation. any one stem broken at a branching
point, together with all the branches de-
scended from it, as in Figure 3A. Natural
ods, I will leave it to the cladistic com- groups were said to be "monophyletic"
munity as a whole to decide. In one not in this sense. Assuming also descent with
uninteresting sense, of course, cladistics modification, the original cladists figured
is just what cladists do. And cladists are that they could recognize and distinguish
the ones that belong to the Hennig So- monophyletic groups on the basis of evo-
ciety and direct its course, etc. lutionary novelties that arose in the stem
So besides the "negative" program of lineage and were passed on unchanged
getting rid of all those evolutionary as- to the descendant branches. In other
sumptions that the pattern cladists want words, different groups should be char-
to avoid, there is also this "positive" pro- acterized by different sets of such shared
gram of getting rid of whatever assump- derived characters, or "synapomorphies."
tions they feel they can just as well do This is still the phylogenetic cladists'
without. It is hard to tell whether the pos- conception of natural groups. It is impor-
itive or the negative aspects are the great- tant to recognize that on this conception
est incentives for the development of the of groups, what "makes" a group a real
movement. For instance, in his recent group is its genealogical history, not the
Hennig Society lecture, Patterson (1981) synapomorphic characters that are used
stressed what he seemed to consider a to recognize it.
positive aspect-namely that pattern cla- Note that groups so conceived and rec-
distics is more "theory neutral" than the ognized can be ordered strictly hierar-
original brand. But it is hard to tell chically in terms of their distinguishing
whether theory neutrality in this case is sets of synapomorphies. If group W in
the primary goal (and why), or whether Figure 3A is recognized and distin-
it is a byproduct of the apparent relief guished in terms of characters a and b, X
that accompanies severing the connec- in terms of c and d, Y in terms of e and
tions with the ever more burdensome f, and Z in terms of g and h, then the
discipline of evolutionary biology, "where organisms that make up the groups fall
nothing seems accessible to investigation into the set-subset relations represented
or test, but where fantasy has free play" by the Venn diagram in Figure 3B. There
(Patterson, 1981). is no partial overlapping here, only inclu-
But the neutrality of pattern cladism sion and exclusion.
with respect to evolutionary theory is, I So here is the pattern cladists pattern
believe, a myth. I will not argue that it again. But the pattern cladists don't be-
reflects or reinforces any particular evo- lieve we need evolutionary assumptions
lutionary theory-i.e., that it is positively about genealogy, monophyly, or anything
theory laden. I will argue instead that it else to generate it or to interpret it. For
is theory antagonistic with respect to evo- the pattern cladists, groups are just col-
lutionary theory. It is at odds with current lections of organisms, distinguished by
evolutionary theorizing. And it under- the sorts of characters that allow the col-
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
1982 CLASSES AND CLADISTS 31
As far as I can see, that's all there is to cladis- What justification is there? How about
tics. ... And I want to emphasize that in pre- the discrepancies with evolutionary the-
senting it, I had no need to use any of the words ory noted in Section 1? That is, evolu-
that have caused all the arguments over the last
15 years-words like monophyly, paraphyly, spe-
tionary theorists do not recognize any
ciation, dichotomy, ancestry, adaptation-don't traits with respect to which species can-
need them. Now, of course, all those words have not evolve, but pattern cladists insist that
to do with evolution, and I didn't need that word there are such traits-namely, the defin-
either. Nor did I need any branching diagrams.
ing traits. This discrepancy may seem in-
And, of course, half the quarrels over cladistics
are due to the mistaken belief that branching dia- consequential to pattern cladists, who are
grams are about evolution. As I understand it, not particularly keen on evolutionary the-
cladistics is theoretically neutral so far as evo- ory anyway. But it's one thing to strive
lution is concerned. It has nothing to say about
for a brand of systematics that is neutral
evolution. You don't need to know about evolu-
tion, or believe in it, to do cladistic analysis. All
with respect to evolutionary theory, and
cladistics demands is that groups have charac- another thing altogether to pursue a
ters, and that the groups are nonoverlapping. brand of systematics at odds with current
(Patterson, 1981) evolutionary theorizing.
The usual rejoinder to this sort of crit-
3.3 So much-though not much-for icism is that classifications should be
the pattern cladists' groups. Species are based on the world, not on theories. A
similarly construed: they are just the system of classification based on evolu-
smallest groups. As Nelson and Platnick tionary theory would tell us more about
construe them, "species are just the that theory than it would tell us about the
smallest detected samples of self-perpet- world. This sort of reasoning has, I think,
uating organisms that have unique sets of unreasonable appeal. In the first place,
characters" (1981:12). the rejoinder is irrelevant in this case.
Now this does not seem very different The question at issue here is not so much
from the traditional class interpretation of whether systematics should be theory
species-the interpretation that led to the neutral, but whether systematics should
difficulties discussed in Section 1. That be theory antagonistic. In the second
"having unique sets of characters" in- place, the rejoinder seems to suggest that
deed means "having defining characters" we can build classifications on the basis
is acknowledged by Nelson and Platnick: of the world in the same manner that we
build museums on the surface of the
To state that a cladogram is a synapomorphy earth. We cannot build classifications "on
scheme invites the rejoinder that a cladogram
the world," but only on what we know
must, therefore, be a phyletic concept. Not so,
for by "synapomorphy" we mean "defining char- about the world. And what we purport to
acter" of an inclusive taxon. True, all defining know about the world is contained in our
characters, in the phyletic context, may be as- best theories. As it happens, our best the-
sumed to be evolutionary novelties. But making ories about patterns of nature are evolu-
that assumption does not render it automatically
true; nor does it change the characters, the ob-
tionary theories. What nonevolutionary
servations on which the characters are based, or rivals better explain geographic distri-
the structure of the branching diagram that ex- bution, the fossil record, developmental
presses the general sense of the characters: i.e., similarities and differences, as well as
that there exist certain inclusive taxa ... that
other adaptive and nonadaptive similari-
have defining characters. (1981:324)
ties and differences? There are difficul-
It would appear, then, that systematics in gen-
ties with our evolutionary theories, as
eral consists of the search for defining characters
of groups. Admittedly, the search seems to have
with all theories. (Among our best evo-
been abandoned, on occasion, by persons who lutionary assumptions, however, some
would search instead for overall phenetic simi- are less problematic than others. On any
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
32 SYSTEMATIC ZOOLOGY VOL. 31
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
1982 CLASSES AND CLADISTS 33
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms
34 SYSTEMATIC ZOOLOGY VOL. 31
logenetic analysis and paleontology (J. Cracraft tions of cladistic analysis. Second Annual Meet-
and N. Eldredge, eds.). Columbia Univ. Press, ing of the Willi Hennig Society, Ann Arbor.
New York, 233 pp. PLATNICK, N. I. 1977. Cladograms, phylogenetic
GHISELIN, M. T. 1966. On psychologism in the log- trees, and hypothesis testing. Syst. Zool., 26:438-
ic of taxonomic principles. Syst. Zool., 15:207- 442.
215. PLATNICK, N. I. 1979. Philosophy and the transfor-
GHISELIN, M. T. 1969. The triumph of the Darwin- mation of cladistics. Syst. Zool., 28:537-546.
ian method. Univ. California Press, Berkeley, 287 PLATNICK, N. I., AND E. GAFFNEY. 1977. System-
PP. atics: A Popperian perspective. Syst. Zool., 26:
GHISELIN, M. T. 1974. A radical solution to the 360-365.
species problem. Syst. Zool., 23:536-544. PLATNICK, N. I., AND E. GAFFNEY. 1978a. System-
GOULD, S. J., AND N. ELDREDGE. 1977. Punctuated atics and the Popperian paradigm. Syst. Zool.,
Equilibria: the tempo and mode of evolution 27:381-388.
reconsidered. Paleobiology, 3:115-151. PLATNICK, N. I., AND E. GAFFNEY. 1978b. Evolu-
GOULD, S. J., AND R. C. LEWONTIN. 1979. The tionary biology: A Popperian perspective. Syst.
spandrels of San Marco and the Panglossian par- Zool., 27:137-141.
adigm: A critique of the adaptationist programme. POPPER, K. 1957. The poverty of historicism. Bea-
Proc. Royal Soc. London, B205:581-598. con, Boston, 166 pp.
HENNIG, W. 1966. Phylogenetic systematics. Univ. POPPER, K. 1959. Logic of scientific discovery. Ba-
Illinois Press, Urbana, 263 pp. sic Books, New York, 480 pp.
HULL, D. L. 1976. Are species really individuals? POPPER, K. 1963. Conjectures and refutations. Ba-
Syst. Zool., 25:174-191. sic Books, New York, 417 pp.
HULL, D. L. 1978. A matter of individuality. Phil. RUSE, M. 1977. Karl Popper's philosophy of biol-
Sci., 45:335-360. ogy. Phil. Sci., 44:638-661.
HULL, D. L. 1980. Individuality and selection. WILEY, E. 0. 1975. Karl R. Popper, systematics,
Ann. Rev. Ecol. Syst., 11:311-332. and classification: A reply to Walter Bock and oth-
LAKATOS, I. 1970. Falsification and the methodol- er evolutionary taxonomists. Syst. Zool., 24:233-
ogy of scientific research programmes. Pp. 91-196 242.
in Criticism and the growth of knowledge (La- WILEY, E. 0. 1979. Cladograms and phylogenetic
katos and A. Musgrave, eds.). Cambridge Univ. trees. Syst. Zool., 28:88-92.
Press, Cambridge, 257 pp. WILEY, E. 0. 1980. Is the evolutionary species fic-
LAUDAN, L. 1977. Progress and its problems. Univ. tion? A consideration of classes, individuals, and
California Press, Berkeley, 257 pp. historical entities. Syst. Zool., 29:76-80.
LEWONTIN, R. C. 1978. Adaptation. Sci. Amer., WILEY, E. 0. 1981. Phylogenetics: The theory and
239:212-230. practice of phylogenetic systematics. Wiley, New
NELSON, G., AND N. PLATNICK. 1981. Systematics York, 439 pp.
and biogeography. Columbia Univ. Press, New
Yoik, 567 pp. Manuscript received November 1981
PATTERSON, C. 1981. The goals, uses, and assump- Revised December 1981
This content downloaded from 128.110.184.42 on Mon, 27 Jun 2016 02:31:01 UTC
All use subject to http://about.jstor.org/terms