DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

download DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

of 13

Transcript of DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    1/13

    352

    Syst. Biol. 48(2):352364, 1999

    Charles Darwins Views of Classication in Theory and Practice

    KEVIN PADIAN

    Department of Integrative Biology and Museum of Paleontology, 1101 VLSB, University of California,Berkeley, California 94720, USA; E-mail: [email protected] u

    Abstract. It has long been argued that Charles Darwin was the founder of the school of evolu-tionary taxonomy of the Modern Synthesis and, accordingly, that he recognized genealogy andsimilarity as dual, synergistic criteria for classication. This view is based on three questionableinterpretations: rst, of isolated passages in the 13th chapter of the Origin of Species; second, of onephrase in a letter that Darwin wrote about the work of an author he had partly misunderstood;and third, of his taxonomic practice in the barnacle monographs, which only implicitly embody hisphilosophy of classication, if at all. These works, seen in fuller context and with the perspectiveof extensive correspondence, are consistent with the view that Darwin advocated only genealogyas the basis of classication, and that similarity was merely a tool for discovering evolutionaryrelationships. Darwin was neither a Mayrian taxonomist nor a cladist, and he did not approachsystematic issues in the same terms that we do in the late 20th century. [Cladistics; classication;Darwin; systematics.]

    What criteria did Charles Darwin regardas the best basis for classication of organic

    beings? And did he follow his own philos-ophy in practice? Darwins views on clas-sication have been extensively researchedand debated, especially in the decades fol-lowing the rise of the competing evolution-

    ary, phenetic, and cladistic schools ofsystematics (e.g., Ghiselin and Jaffe, 1973;Nelson, 1974; Mayr, 1982, 1995; Ghiselin,1985). Some scholars have advanced theview that Darwinian classication (in the-ory, in practice, or both) involved dual cri-teria of genealogy and degree of similarity,at least to some extent (e.g., Ghiselin and

    Jaffe, 1974; Mayr, 1982, 1994a, 1994b, 1995).Others (e.g., Nelson, 1974; Ghiselin, 1985;

    Desmond and Moore, 1991; Padian, 1994)have stressed Darwins insistenceon geneal-ogy alone. Most scholars agree that Darwincannot be regarded simply as an evolution-ary taxonomist or a cladist, or as a patronsaint to any current school. However, is-sues and terms have shifted so much overthe years that it has become both difcultand inappropriate to shoehorn the ideasand language of Victorian science into cur-

    rent debates. Nevertheless, Darwins ideasabout classication are interesting and im-portant for their own sakefor what theytell us about the changing climate of thoughtduring the era when he wrote, and for therelationship that they bear to his other ideas

    about evolution (e.g., Desmond and Moore,1991).

    New publications of Darwins letters andrecent advances in scholarship have pro-vided some fresh insights into Darwinsthoughtin thecontextof his times (Desmondand Moore, 1991; Browne, 1995; and notably

    the series, Correspondence of Charles Darwin,edited by Burkhardt and Smith [1985, 1988,1990, 1991, 1993], denoted here as CCD withappropriate volume and page numbers).Current historiographical methods aim tosettexts and authors in their appropriate cul-tural placeand time, acknowledging that themeanings of words as well as the groundsof basic issues and even philosophical stan-dards have changed. To understand what

    Darwin actually said and thought, it is im-portant rst to read his passages in context,not in isolation (Ghiselin and Jaffe, 1973).Darwin often developed his ideas over aseries of pages, and an isolated quotationcan be misleading. Second, it is important toknow when Darwin is stating his own views,and when he is summarizing, however sym-pathetically, the views of others with whomhe disagreed (Ghiselin, 1985:460). Darwin

    was extremely deferential to the views ofothers, particularly when he was about toadvance an alternative, and he tried scrupu-lously to be fair about representing estab-lished views. He also went out of his way toconsider the views of those who he thought

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    2/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 353

    would be most likely to object to his own,regardless of whether he thought their crit-icisms relevant or justied (his self-styledtruckling passages). Frequently, Darwinsletters to his friends contain franker andmore elaborate statements of his views, andthese have been effectively analyzed andcompared in understanding the develop-ment of his ideas (e.g., Ghiselin and Jaffe,1973; Ospovat, 1981; Desmond and Moore,1991; Browne, 1995). Third, it is importantto read what Darwin himself read. Oftenin his published writings as well as his let-ters, his language reected arguments andterms of others that were current at thetime; for example, the Quinarian classica-tion scheme of W. S. MacLeay, which in-trigued many biologists in the 1820s and1830s (Ospovat, 1981:110114) but is largelyforgotten today. Finally, one must take intoaccount that the meanings of some crucialwords have changed since the 1840s. Dar-win, for example, had to ask G. R. Water-house in 1843 what he meant in his letters

    by relationship in classifying organisms(CCD 2:378). Waterhouse replied by rela-tionship I mean merely resemblanceI useit in a most vague way and he differ-entiated a relationship of afnity (similarityshared by members of a natural group)from one ofanalogy (convergent similarity).For Waterhouse, relationship was purely aquestion of degree of resemblance (CCD2:381382), a meaning that 20th century sys-tematists would not accept because it doesnot strictly presume a genetic basis. This isanother hazard of taking Victorian vocabu-lary out of context.

    In this essay I make three points aboutDarwins views of classication. First, Dar-win advocated and insisted on genealogy asthe criterion for classication. He did notconnect or correlate it with any other cri-terion, though he recognized factors suchas degree of similarity and correlations inembryological stages as important in reveal-ing genealogy (and this is a crucial distinc-tion). He maintained this position in lettersfrom 1843 until after the initial publicationof the Origin of Species (1859) and throughevery later edition, as well as in The De-scent of Man (1871). Second, the argument

    that Darwin advocated dual criteria of ge-nealogy and degree of similarity is not sup-ported by his writings on how classicationshould be done, but is rather an interpreta-tion that hinges on a single statement, takenout of context, that Darwin wrote about an-other authors workone that he himselfdoes not seem to have fully understood. Infact, on several occasions Darwin denied ex-plicitly that similarity should be the basisfor classication. Third, in his work on liv-ing and fossil barnacles, Darwin apparentlydid notfollow his own stated theoretical pre-cepts, in the sense that he did not make ge-nealogy a focus of discussion as the basisof classication, and in fact barely referredto it. Nor did he develop in these works anyphilosophy of classication. However, I sug-gest that his practical taxonomic work didindeed support his view that classicationshould be based on genealogy. The difcultyof his taxonomic subjects simply preventedhim from achieving his ideal and has pre-vented us from understanding why.

    DARWINS EMPHASIS ON GENEALOGY

    Mayr (1982:209) regarded Darwin asthe founder of the whole eld of evolu-tionary taxonomy, and quoted Simpson(1961): Evolutionary taxonomy stems ex-plicitly and almost exclusively from Dar-win. Mayrs (1982) Chapter 5, reiterated inpart more recently (1994a, 1994b), summa-rizes his view of taxonomy and of Darwinsrole in the development of taxonomy. Afterciting Simpson with approval, Mayr quotedsome famous passages from the 13th chap-ter of Darwins Origin of Species (1859:411413) to the effect that classication is not ar-

    bitrary, like the groupings of stars into con-stellations, but attempts to follow a NaturalSystem. Darwin said that what naturaliststhen called the Natural System took sev-eral forms (Ghiselin, 1969): It was variouslya scheme for separating and grouping un-like and like living objects (why did Dar-win not use the term organic beings, ashe did throughout the Origin?) according toresemblances, key characters, embryology,or other factors; or it was an attempt to un-derstand the plan of the Creator (as Natural

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    3/13

    354 SYSTEMATIC BIOLOGY VOL. 48

    Theology did), for which no natural ruleshad been adduced; or it was a combinationof these factors (which could be said to have

    been the goal of Linnaeus, among others in-cluding Owen, Sedgwick, and even Lyell).

    For Darwin, the only real difference be-tween these traditional systems and whathe regarded as the proper basis of classi-cation was that descent with modicationreplaced divine plan or other philosophicalconsiderations. But he still needed a prac-tical way to discover genealogical relation-ships. He recognized that many of the tradi-tional criteria, such as degree of similarity,embryological resemblances, and so on, re-ected genealogical proximity. In these re-spects, he appreciated that traditional tax-onomy had probably been replicating moreor less faithfully the pathways of diverg-ing parentage. In one sense evolutionarytaxonomy follows Darwins view, in that itaccepts genealogy as the underpinning ofmost similarity, and only inherited similar-ity can be a basis for taxonomy. The foundersof evolutionary taxonomy explicitly tracedthis rationale to Darwins doorstep. It couldeven be argued that Darwin condoned pa-raphyletic taxa (Ghiselin and Jaffe, 1973), asticking point between the evolutionaryschool and the cladistic school, inasmuchas some of his barnacle taxonomy admittedthem. But in the full compass of his writings,Darwin regarded such groups as necessaryevils only because the complete genealogywas not known (thanks to differential ex-tinction). Without these gaps, classicationwould be purely genealogical.

    Mayr (1982) was correct that Darwin wascontemptuous of the Natural System. Dar-win (1859) traced this system, in its variousforms, to Linnaeus and other authors of the18th and 19th centuries, noting that, at leastin its mature form in mid-Victorian England,it incorporated both similarity and DivinePlan. However, the various versions of theNatural System were sufciently confus-ing that he had to apologize profusely toRichard Owen for including him with thespecial creationists in the rst edition of theOrigin.

    Let us rst see what Darwin thoughtabout the Natural System and why. In 1843

    G. R. Waterhouse wrote to him to ask hisadvice on the ranks of classication in theNatural System. Darwin characterized theNatural System as follows in his response(CCD 2:375376; emphasis in original):

    Most authors say it is an endeavour to discover the

    laws according to which the Creator has willed toproduce organized beingsBut what empty high-sounding sentences these areit does not meanorder in time of creation, nor propinquity to anyone type, as man.in fact it means just nothing.According to my opinion, (which I give everyoneleave to hoot at . . . ) classication consists in group-ing beings according to their actual relationship, i.e.,their consanguinity, or descent from common stocks.

    Darwin paraphrased this passage, some-what toned down, 16 years later, on page

    413 of the Origin of Species, which suggeststhat his views did not change in the ensuingyears, and that he felt that this was a suf-ciently clear statement of them. But some fol-lowing passages in his letter to Waterhouse(CCD 2:376) are equally revealing:

    To me, of course, the difculty of ascertaining truerelationship [i.e.,] a natural classication[,] remains

    just the same, though I know what I am looking for.. . . we ought to look at classication as a simple log-

    ical process, i.e.[,] a means of conveying much infor-mation through single wordsnow the laws of clas-sication tryed by this ruleit is clear that neithernumber of speciesnor grade of organization oughtto come in, as an element. . . . the only cheque to thesplitting of groups,appears tobe convenience; . . . I re-peat, that until you can dene your object in classify-ing,youhavenorighttointroducenumberofspecies,or (arbitrary) grades of organization, but are boundto follow simple amount of differences of organiza-tion [as, he notes, botanists do]. . . . This is clearly therule, on the view of classication, being a genealogi-

    cal process, exhibiting literal or actual relationship.

    In this passage, Darwin links genealogyand amount of differences, but he is not say-ing that they are dual criteria for classica-tion. He says in both passages that classica-tion should be strictly genealogical, and thatnumbers of species or arbitrary grades of or-ganization (for example, whether a barnacleis stalked or not) should notcharacterizema-

    jor groups. The reason he advocates follow-ing simple amount of differences of orga-nization in classication is that progressivedifferences suggest progressive divergence,reecting the gradual process of evolution

    by which genealogy can be traced, given a

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    4/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 355

    sufcient series of forms. In another letterto Waterhouse a few days later (CCD 2:377379, in which he also protested against num-

    ber or grades of organization being used inclassication), he emphatically stated thatall rules for a natural classication are futile un-til you can clearly explain, what you are aimingat (emphasis in original). Darwins under-pinning idea of the consanguinity of all crea-tures was clearly what he was aiming at, butthat idea was not generally accepted in 1843,or he would not have risked shocking Wa-terhouse by his views (he ended his earlierletter, Have you had patience to read thusfar.). This letter, again, says nothing aboutdegree of similarity; it speaks only of re-lationship by descent.

    These views of Darwins were neitherhastily composed nor temporarily held.Fourteen years later, Darwin carried on asimilar correspondence with Huxley. Twoletters, from 26 September and 3 October1857, reiterate his views on the Natural Sys-tem and on genealogy in classication (CCD6:456):

    In regard to Classication, & all the endless disputes

    about the Natural System[] which no two authorsdene in same way, I believe it ought, in accordancetomy heteredox notions, tobe simply genealogical.But as we have no written pedigrees, you will, per-haps, say this will not help much; but I think it ul-timately will, whenever heteredoxy becomes ortho-doxy, for it will clear away an immense amount ofrubbish about the value of characters &will makethe difference between analogy & homology, clear.The timewill come I believe, though I shall not live tosee it, when we shall have very fairly true genealog-ical trees of each great kingdom of nature.

    Darwin regarded the lack of intermediateforms in both the living world and the fossilrecord as the main impediment to achiev-ing a fully genealogical classication (e.g.,CCD 2:376, 415). Huxley wrote back to Dar-win (CCD 6:461) that Cuviers denition ofthe object of Classication seems to me toembody all that is really wanted in Scienceit is to throw the facts of structure into the fewest

    possible general propositions. (emphasis inoriginal) Darwin replied,

    I knew, of course, of the Cuvierian view of Classica-tion, but I think that most naturalists look for some-thing further, & search for the natural system,forthe plan on which the Creator has worked &c &c.

    It is this further element which I believe to be simplygenealogical.

    And he went on to give Huxley an ex-tended example of his view (emphasis inoriginal) that it would be possible to makeclassications purely genealogical if missing

    intermediates were known (CCD 6:463):Grant all races of man descended from one race; grantthat all structure of each race of man were perfectlyknowngrant that a perfect table of descent of eachrace was perfectly known.grant all this, & then doyou not think that most would prefer as the best clas-sication, a genealogical one, even if it did occasion-ally put one race not quite so near to another, as itwould have stood, if allocated by structure alone.Generally, we may safely presume, that the resem-

    blance of races & their pedigrees would go together.

    He ended his letter to Huxley, who hadpushed development as a prime cog in thenascent machine of a Scientic & logicalZo-ology & Botany (CCD 6:461), with the fol-lowing postscript: It might be asked why isdevelopment so all-potent in classication,as I fully admit it is; I believe it is, becauseit depends on, & best betrays, genealogi-cal descent; but this is too large a point to

    enter on. Huxley completely missed thepoint,writing back, Your pedigree businessis a part of Physiology. . . no more to dowith pure Zoologythan human pedigreehas with the Census.Zoological classica-tion is a Census of the animal world (CCD6:462).

    Darwins own words, cited above, indi-cate his frustration with the Natural System(which he put in quotation marks) as it was

    practiced in his day, and his strong advo-cacy of genealogy as the central and only cri-terion for classication. They also illustratehis frustration with the gaps that exist be-tween living forms and in the fossil record,without which the establishment of a clas-sication by genealogy would be much eas-ier. The Natural System, to him, could not

    be truly natural unless genealogy were itsbasis, and this is why he viewed the Natu-

    ral Systems of his time with frustration andcynicism.The most crucial passage on classication

    in the Origin of Species (Chapter 13, p. 420)has been widely cited, but often interpretedout of context. As this passage begins, Dar-

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    5/13

    356 SYSTEMATIC BIOLOGY VOL. 48

    win has just nished discussing the ideasof various authors (not his own) on crite-ria for developing classications (adaptive,inadaptive, reproductive, rudimentary, andembryological characters), and how mutu-ally conicting these characters can be. Hethen advances his own view that all of thesefeatures can be explained by descent withmodication; and hence, if one were to ac-cept this view, the characters which natu-ralists consider as showing true afnity be-tween any two or more species, are thosewhich have been inherited from a commonparent, and, in so far, all true classicationis genealogical. Note that he is not sayingthat this is how taxonomy was practiced orunderstood in his day (it was not); rather, heis saying that if the patterns of variation can

    be explained by descent with modication,then a true classication would be strictlygenealogical.

    Moreover, for Darwin the resemblancesthat naturalists have been using all along arereally genealogical. He immediately con-tinues in the same vein (emphasis added):that community of descent is the hidden

    bond which naturalists have been uncon-sciously seeking, and not some unknownplan of creation, or the enunciation of gen-eral propositions, and the mere putting to-

    gether and separating objects more or less alike.Darwins statement here reveals that simi-larity is not in itself a criterion for classica-tion, but a means to understand genealogy.In that nal phrase, in fact, he could be readas throwing cold water on the idea that de-gree of difference should have an intrisicallyimportant role. Rather, he used similaritiesin (what we would call) both labile and con-servative features, as well as in all stages ofembryology, to try to understand genealog-ical relationships.

    Ghiselin has long argued that genealogywas the only criterion of Darwins philoso-phy of classication, though some of his ear-lier work (Ghiselin and Jaffe, 1973) regardedDarwins view as closer to Mayrs than toHennigs. In his 1969 book, The Triumph ofthe Darwinian Method, Ghiselin stated simply(1969:89) : Ithas already been demonstratedthat Darwin considered propinquity of de-scent the basis of natural classication. His

    motives have been explained as the conse-quence of a particular kind of nominalism:a taxon is a genealogical entity, not a class ofmorphologically similar organisms. Ghis-elin (1985) claried this apparent paradox.He noted that Darwins view of the Natu-ral System in use at his time only roughlycorresponds to the genealogical nexus. Sincethe amount of difference is expressed, it hasto be done by means of categorical rank.However, [Darwin] did not endorse thatstep, but only enunciated what could better

    be read as a conditional statement.Ghiselin, like Nelson (1974), detailed Dar-

    wins treatment of the classication of hu-mans in The Descent of Man as an example ofhow he subjugated ideas about degreeof dif-ference to those of genealogy, where it could

    be known (Darwin, 1871:181) :

    [N]aturalists have long felt a profound convictionthat there is a natural system. This system, it isnow generally admitted, must be, as far as pos-sible, genealogical in arrangementthat is the co-descendants of the same form must be kept togetherin one group, apart from the co-descendants of anyother form; but if the parent forms are related, so willtheir descendants, and the two groups together will

    form a larger group.

    Darwin argued in the following pages thathumans have separated themselves fromother animals largely by their degree of dif-ference. If man had not been his own classi-er, he would never have thought of found-ing a separate order for his own reception[p. 188]. . . under a genealogical point of viewit appears that this rank is too high, andthat man ought to form merely a Family, orpossibly even only a Sub-family [p. 187].Darwin recognized that man is unquestion-ably an offshoot of the Catarrhines ( OldWorld monkeys), and that, under a ge-nealogical point of view he must be classedwith [them]. (p. 189) Classication shouldnot, toDarwin, be based on amount of diver-gence, no matter how important the differ-ences, but rather should be rooted strictly bythe recognition of closest divergence from acommon ancestral stock. When this point isknown, the taxonomic rank will follow, ashis examples above suggest.

    Peter Stevens, in reviewing an earlier draftof this paper, astutely pointed out the im-

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    6/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 357

    portance of distinguishing among Darwinsusesofthewordsgenealogy, arrangement,andclassication (as for example in Chapter 13of the Origin, discussed above). I had over-looked arrangement as a weak synonym ofother terms, and agree that this should be re-garded as a separate intellectual activity forDarwin. Some kinds of arrangements (i.e.,genealogies with all intermediates present)could not be represented as classications,and this was a problem that worried Darwin(see below). A set of related species can bearranged according to given criteria of char-acters and an idea of the rationale for thisactivity (what you are aiming at in Dar-wins words); a classication can be drawnfrom the results. These three terms are as dif-ferent as are systematics, taxonomy, and clas-sication (pace Simpson, 1961; Mayr, 1982).Systematics can be seen as the philosophyof organizing nature; taxonomy as the useof sets of organic data guided by system-atic principles, which sort of sets will dif-fer among taxa (teeth for fossil mammals,genitalia for owering plants); and classica-tion as the tabular or hierarchical end resultof this activity. Darwin regarded genealogyas the overriding principle by which organ-isms should be ordered; he saw the arrange-ment of organisms as basically genealogi-cal but varying in placement according tothe evolutionary differentiation of featuresand the work of extinction in eliminating in-termediate forms (see below); and classica-tion as an expression of genealogy, ironicallymade possiblein part because extinction hadcreated visible divisions in the natural ge-nealogical ow.

    DARWINS DUAL CRITERIA? GENEALOGY VERSUSSIMILARITY

    Up to this point, I have concentrated onclarifying Darwins disregard for the Nat-ural System of his times, and how, in theOrigin of Species and The Descent of Man,

    Darwins statements on classication advo-cate genealogy as its unique criterion (Nel-son, 1974; Ghiselin, 1985). In contrast, Mayr(1982, 1994a, 1994b) has long championedthe view that Darwin advocated the two cri-teria of genealogy and degree of similarity in

    classication. Much of the basis of this inter-pretation by Mayr and others (e.g., Ghiselinand Jaffe, 1973) has traditionally come froma passage in Chapter 13 (p. 413) of the Originof Species:

    I believe that something more is included [in our clas-

    sication], than mere resemblance; and that propin-quity of descentthe only known cause of the simi-larity of organic beingsis the bond, hidden as it is

    by various degrees of modication which is partiallyrevealed to us by our classications.

    It would be possible to infer from this iso-lated passage that Darwin was saying thatresemblance should be involved in classi-cation, but that another element, propin-quity of descent, is what is really important.

    But Darwin is making a deeper point. Inhis other writings Darwin clearly says thatthe reason that we are able to rely as muchon resemblances in our classication as wedo is that common descent is what actu-ally causes these resemblances, and that iswhy traditional classications (unintention-ally) reect genealogy so well. In a passagesome pages following the one quoted above(p. 420), Darwin attempts to explain his

    meaning more fully (emphasis in original):I believe that the arrangement of the groups withineach class, in due subordination and relation to theother groups, must be strictly genealogical; but thatthe amount of difference in the several branches orgroups, though allied in the same degree in blood totheir common progenitor, may differ greatly, beingdue to the different degrees of modication whichthey have undergone; and this is expressed by theforms being ranked under different genera, families,sections, or orders.

    Mayr (1982, 1994a, 1994b) regards this astwo criteria, but this interpretation is un-dermined by the plain difference betweenthe must be of strict genealogy and themay of the amount of difference. Darwinchose his verbs carefully. As in the previouspassage, Darwin is clearly saying here thatthe arrangement, and hence the classica-tion, must be genealogical, even though dif-

    ferences among descendant taxa may varywidely. (He goes on to explain that dif-ferential extinction has forged these vari-able distinctions among taxa, whose rankssimply reect the current absence of nowextinct, transitional forms descended from

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    7/13

    358 SYSTEMATIC BIOLOGY VOL. 48

    common ancestors.) The third verb in thispassage expresses the fact that in currentpractice (in his day), this degree of modi-cation is expressed by degrees of ranking,which may reect the intermediate gaps, buthe is not advocating this ranking practicefor classication (i.e., the practical outcomeof the arrangement of groups with respectto other groups). His following discussionsexplain his illustration of what might becalled a genealogical arrangement, the onlyillustration in the book; and his analogy tothe classication of languages that follows(p. 422423) is strictly genealogical with-out any criterion of degree of similarity ordifference entailed. Ghiselin (1985:460) hascited this passage, emphasizing Darwins(1859:422423) acknowledgment that differ-ences among languages would have to beexpressed by groups subordinate to groups;

    but the proper or even only possible ar-rangement would still be genealogical; andthis would be strictly natural, as it wouldconnect together all languages, ancient andmodern, by the closest afnities. [I]f wehad a real pedigree, Darwin continued, agenealogical classication would be univer-sally preferred. This point is clearly borneout in the passages on classication of hu-mans cited above. So in these passages Dar-win does not seem to be advocating dualcriteria of genealogy and similarity.

    Mayr (1982:210) has forcefully stated hisview that, In the Origin, and in his corre-spondence, Darwin stresses again and againthat all true classication is genealogical(p. 420), but genealogy by itself does notgive classication (L.L.D., II:247). Fromthis, Mayr (1994a) developed his conceptthat Darwin prescribed dual criteria for clas-sication, which differentiate it from theHennigian system (emphasis is added):

    There are now two systems of ordering organisms inuseDarwinian classication, by which organismsare grouped according to both similarity and geneticrelationship, and Hennigian ordering, by which or-

    ganisms are grouped according to the branch of thephylogenetic tree on which they occur.

    Mayr (1994b) reiterated this claim with re-spect to the Originthough note a subtlechange in wording:

    Darwin showed that all classications must be ge-nealogical in order to reect his theory of commondescent. Furthermore, Darwin strongly emphasizedthat a second criterion was needed to achievea soundclassication, that of representing degree of similar-ity [1859:420]. Again and again Darwin stated ge-nealogy alone does not give a classication. Dar-wins advice tobase classications on twocriteria, ge-

    nealogical relationship and degree of similarity, wasmore or less followed by all good taxonomists in thenext 90 years. There was no need to designate thismethod as Darwinian because it was the only prop-erly presented method of classifying.

    In the passage quoted previously, Mayrhas Darwin saying again and again thatall true classication is genealogical, whichI have emphasized with Nelson and Ghis-

    elin is a correct statement. But in the pas-sage just quoted, Mayr (1982:210; 1994a) alsohas Darwin saying again and again thatgenealogy alone does not give a classica-tion, which I think is demonstrably false,and which removes the underpinnings fromMayrs dual criteria argument. Darwinnever said anything like this in the Origin ofSpecies. The Concordance to the Origin (Bar-rett et al., 1981) shows no use of the word

    degree in the book in association with thewords of similarity or of difference inthe context of classication, and no similarconstruction associated with any other wordin this phrase. Darwin also never said thisagain and again. He used the words ge-nealogies and genealogical sixteen timesin the Origin, and when he did so in thecontext of classication, it is clearly the onlycriterion under discussion. Mayr (1982:210)

    was originally correct in saying, In theOrigin, and in his correspondence, Darwinstresses again and again that all true classi-cation is genealogical (p. 420), in contrastto his 1994a statement. But Mayr (1982:210),I think, was not correct in linking this fact tothe conditional statement, but genealogyalone does not give classication. The lat-ter phrase is a comment Darwin made in aDecember 1859 letter to Joseph Hooker, his

    friend and close correspondent, about theFrench botanist C. V. Naudin. But Mayr hastaken it out of context, and its origins are

    both complex and revealing.In the months following the publication

    of the rst edition of the Origin of Species,

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    8/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 359

    the ush of criticism from people who mis-understood and wanted to pillory Darwinfor his ideas was bizarrely complemented

    by others who wanted to take credit for them(see, e.g., Hull, 1983; Desmond and Moore,1991). Naudins colleague Joseph Decaisne,an editor of the Revue Horticole, had repre-sented to Hooker that Naudin had alreadypublished many of Darwins ideas in that

    journal in 1852. Darwin replied to Hooker,who had sent him Naudins article. Darwinwas mostly concerned with his own prior-ity for the idea of Natural Selection, but hedirected a few comments at Naudins state-ments on classication in the last long sen-tence of this passage (CCD 7:443444):

    I am surprised that Decaisne shd say it was same asmine. Naudin gives articial selection as well as ascoreof English writers; & when he says species wereformed in same manner I thought the paper wouldcertainly prove exactly the same as mine. But I can-not nd one word like the Struggle for existence &Natural Selection. . . . He assumes (like old geologistsassumed the forces of nature were formerly greater)that species were at rst more plastic. His simile oftree & classication is like mine (& others), but hecannot, I think, have reected much on subject, oth-erwise he would see that genealogy by itself does

    not give classication. I declare I cannot see muchcloser approach to Wallace & me in Naudin than inLamarckwe all agree in modication and descent.

    What did Naudin really say? The longpassage from Naudin that concerns us here(1852:105106) begins with the recognitionthat species are simply varieties that have

    been separated for a much longer period oftime. Taken a step farther, all ofnature sharesa consanguinity that reects a commonal-

    ity of origin, and in turn is reected by acommonality of organization that is classi-cation. Exactly as Darwin did in the Ori-

    gin, Naudin went on to criticize the con-temporary system in which resemblanceswere thought to be fortuitous, merely adap-tive, or independently created. To the con-trary, these resemblances are the conse-quence of a real parentage and inheritancefrom a common ancestor. Hence, Naudin

    said (1852:105106; my translation; empha-sis in original),

    In our system, by contrast, these resemblances areat the same time the consequence and the proof ofa parentage, no longer metaphorical, but real, that

    they inherit from a common ancestor, from whichthey havedescended over more or less remote epochsand through a more or less numerous series of inter-mediates; such that we would explain the true rela-tionships among species by saying that the sum oftheir reciprocal analogies [similarities] is the expres-sion of their degree of common parentage, as thesumof their differences is that of their separation from the

    common stock from which they originated.Seen from this point of view, the vegetable king-

    dom would no longer present itself as a linear serieswhose endswould run along lines ofincreasing or de-creasing complexity of organization, whether one ex-amined it beginning at one end or the other; it wouldno longer be a disorganized tangle of intercrossedlineages, nor even a geographic map whose regions,different in shape and extent, would be connected bya greater or lesser number of points; it would be atree whose roots, mysteriously hidden in the depthsof cosmic time, would have given rise to a limited

    number of stems that successively divided and sub-divided. These rst stems would represent the pri-mordial types of the kingdom; their last brancheswould be the living species.

    It follows from this that a perfect and rigorous clas-sication of organic beings of the same kingdom, or-der, and family would be nothing but the genealogi-caltree of these species itself, showing therelative ageof each one, its degree of specieite [he means by thissomething incorporating both degree of divergenceand also the integrity of the phenotype of a species],and the ancestral lines from which it descended. By

    this would be represented, in a somewhat palpableand material way, the different degrees of commonancestry among species, as well as those of groupsof diverse degrees, going all the way back to the pri-mordial types. Such a classication, summarized in awritten table, would be perceived with equal ease bythe mind as well as the eyes, and would present the

    best application of the principle generally admittedby naturalists: that nature is miserly with its causes andprodigal with its effects.

    It is clear from the rst paragraph thatNaudin fully accepts descent with modi-cation and goes on to say that classicationshould reect this. But he never said thatgenealogy gave classication in practice; likeDarwin, he said that it should do so, in con-trast to the system in use in their time, andperfectly in agreement with Darwins views.In fact, Darwin could almost have writtenNaudins last paragraph himself.

    How did the confusion arise? One in-

    terpretation is that Darwin appears not tohave recognized that Naudin was writingin the conditional tense, manifested in everyverb in this passage (resulterait instead ofre-sulte; seraient instead ofsont). In other words,everything that Naudin was saying repre-

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    9/13

    360 SYSTEMATIC BIOLOGY VOL. 48

    sented an ideal, not reality. Could Darwinhave misinterpreted Naudin? The phrase inquestion is only one point of several thatDarwin treated in his hasty letter to Hooker(Darwins haste is evidenced in his run-onsentences, abrupt transitions of topic, and

    omissions of articles that were typical of hisquick notes to his friend). By all accounts,Darwins French was less than uent, andhe may have read Naudins passage quickly.Darwin was far more interested in Naudinstreatment (or omission of) articial selec-tion, natural selection, the struggle for ex-istence, and decreasing variability throughtime, than in the treelike approach to clas-sication. In 1860, when he determined to

    add an historical sketch of Natural Se-lection and related ideas to the second edi-tion of the Origin of Species, Darwin wroteto Hooker (CCD 8:60) asking for a sentencefrom Naudins paper dealing with the prin-ciple of nality. In his sketch he treatedNaudins views on selection and the forma-tion of species with his usual polite circum-spection, but he could not resist a jab aboutNaudins underlying explanation of nal-ity. He did not mention Naudins viewson classication and never seems to havethought of them again.

    Regardless of Darwins reading of him,Naudins meaning is clear. But more impor-tant for our purposes is what Darwin wastalking about in his letter to Hooker. Dar-win did not mean should when he said ge-nealogy alone does not give classication.

    In other words, he meant that in practice itdoes not, and he could not see how Naudin(as he thought) could say that it did. But inall other statements on the subject, as shownabove, Darwin rmly believed that ideally,genealogy should give classication.

    This statement of Darwins concerningNaudins views is essential to the argumentthat Darwin felt that classication shouldnot be based on genealogy alone (Mayr,

    1982, 1994b, 1995). Because it is plain thatDarwin misinterpreted Naudins views, thephrase in question provides no support forthat argument. The contradiction betweenthis statement of Darwins and his otherviews on classication did not escape Ospo-

    vat (1981:172), who reconciled it by showinghow Darwin used the principle of genealog-ical classication as a basis for the principleof divergence.

    DARWINS BARNACLE MONOGRAPHS: CONFLICTBETWEEN THEORY AND PRACTICE

    Darwins barnacle monographs (1851a,b1854a,b) are the only taxonomic works thatDarwin ever did, and he spent some 8 yearsat them (amid other studies). What can theytell us about Darwins systematic philoso-phy? It is fruitless to try to read Darwinspractice in these monographs as evolution-ary, cladistic, or akin to any 20th cen-

    tury school. Darwin was a man of his times,though he transcended them, and his taxo-nomic work must be seen in the context ofits times. It is almost universally acknowl-edged that Darwins larger oeuvre (not the

    barnacle monographs) legitimized descentas a basis for classication, but as Alvar El-legard (1958) noted, it was descent, not nat-ural selection, that was accepted (Ospovat,1981). It is equally acknowledged that tax-

    onomy was not revolutionized in practice inthe decades following the increasing accep-tance of evolution; indeed, Darwins mono-graphs had no effect at the time, or since, oncontributing to this eventual revolution.

    Among recent works, Ghiselin and Jaffe(1973) revived the consideration of whatDarwin was trying to accomplish system-atically in his barnacle monographs, andthey were the rst to recognize the primacy

    of genealogy in Darwins view of classi-cation (see also Ghiselin, 1985). Burkhardtand Smith (1988, Appendix II to CCD 4:388409) examined historiographically Darwinsreasons for undertaking the study as wellas the philosophical and technical bases formuch of the work. They stressed Darwinsuse of homology (recently standardized byOwen, in place of the vaguer afnity), inparticular, using all ontogenetic stages (not

    merely the adult) to discover the evolution-ary relationships of the barnacles. And New-man (1993), in a masterful study that should

    be read in complement to Burkhardt andSmiths, reviewed the quality and characterof Darwins observations on barnacles, espe-

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    10/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 361

    cially in the light of more recent discoveries,interpretations, and classications.

    The barnacle papers are in most re-spects characteristic of taxonomic mono-graphs from Darwins day to this, in thatevolution and genealogy are topics subju-gated to enumerating, describing, distin-guishing, and evaluating characters andtaxa, and such judgments have their ownreasonable subjectivity and logic. In fact, inthe barnacle monographs, there is hardlyany attempt to document evolution; thereis no statement of systematic philosophy;and virtually the only attempt to put anykind of evolutionary perspective to the clas-sication is Darwins inference that sessile

    barnacles evolved from stalked ones, be-cause the latter are geologically older. Ghis-elin (1969) masterfully showed how muchDarwin transcended most systematic workof his day by introducing considerations ofcorrespondence in larval form, sexual di-morphism, and adaptive modications totaxonomy, in an effort to reect evolution-ary change; but these considerations werelargely inferential and implicit, not explicit.That is, we can read them now if we look forthem, but few in Darwins time would havegrasped these subtleties. Ghiselin (1969:105)acknowledged that Darwin nowhere dis-cussed the relationship between evolution-ary theory and systematics and nowherelaid out a philosophy of classication.

    One can, of course, search the mono-graphs for evidence that Darwin followedone or another practice of recent schools ofthought. He eliminated some polyphyletictaxa, and he reduced some paraphyly: Forexample, he could have used the amount ofdifference to separate the Cirripedia fromthe Crustacea, but he did not (Ghiselin,1985; see also Ghiselin and Jaffe, 1973). Onthe other hand, he chose not to group the11 genera of Lepadidae into subfamilies(1851b:64), even though he recognized theirlong history and considerable (overlapping)divergence. But he was not wrestling withour problems. His rst problem was to de-termine the point at which cirripedes could

    be separated from other crustaceans, and soreveal their taxonomic rank. This was dif-cult, as he showed (1854b:1718), because

    authorities of the time conferred degree ofrank based on factors such as species diver-sity, degree of difference, or perfection oforgans as being high or low in rela-tion to a standard type (letters from Darwinto Waterhouse in 1843 [CCD 2:378] expresshis frustration with this). Darwin wrestledwith all these, hoping to offend no one, andnally sided with Dana and against Milne-Edwards (to whom the work was dedicated)in recognizing Cirripedia as a subclass ofCrustacea, rather than a subgroup of ento-mostracan crustaceans. Genealogy, at thispoint, never explicitly entered the question.Darwin does not say thatthe differences thatseparate barnacles from other crustaceansreect their deep genealogical divergence;he simply contrasts their similarities and dif-ferences. Yet even a decade earlier he wasvehemently stressing to Waterhouse thatclassication should be based on genealogyalone, and he was still doing it in The Descentof Man in 1871. This is the real paradox of the

    barnacle monographs.Among the characters Darwin consid-

    ered, he tried to look for unusual or uniqueresemblances, and some of these are clearlywhat cladists would call derived. But itcannot be said that Darwin methodicallysearched for what we would call synapo-morphies. He gave considerable attentionto weighting embryological and other con-servative characters traditionally consid-ered important in taxonomy (Ghiselin and

    Jaffe [1973] and Newman [1993] providedwell-detailed examples). For example, hechose to evaluate the unusual cirripedeCryptophialus on the basis of its conserva-tive female morph, not on the complemen-tal male form. But at the same time, he wastempted to group Cryptophialus, to which hegave its own order, with Alcippe on the ba-sis of what we would call uniquely sharedfeatures (1854b:564566).

    On the other hand, Darwin seemed touse both general afnity and degree ofsimilarity to classify some barnacles. Hiserection of the lepadid genus Oxynaspis(1851b:133), for example, reected its sim-ilarities in various features to those ofLepas,Paecilasma, and Scalpellum without beingable to place it in any of these. This would

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    11/13

    362 SYSTEMATIC BIOLOGY VOL. 48

    seem to support Mayrs view that Darwinused both genealogy and similarity as tax-onomic criteria. But Darwin was not beingprecise about genealogy here (Ghiselin and

    Jaffe, 1973:Fig. 1); in fact, genealogy wasshrouded by the intricate and parallel adap-tive modications of these taxa. Rather, Dar-win understood that at some level all theseforms, and doubtless others, had evolvedfrom a singlecommon ancestor. In the courseof their divergence they retained and elab-orated some characteristics of that lepa-did ancestor, while differentiating in otherways; and this is what might have beenexpected, Darwin concluded, for it is themost ancient family, and extinction has doneits work, separating genera which, in accor-dance to analogy, we may suppose were oncemore nearly connected by intermediate forms(Darwin, 1854b:32; emphasis added).

    In other words, Darwin was trying tosolve the same problem over and over. Vi-sualize an evolving lineage growing up-ward, and separating as it grows into vari-ous branches; now transect the trees upperreaches with a horizontal plane. The planerepresents our current slice of time. Fromthis slice, the problem is to trace the diver-gence of the branches, using only presentforms, given that extinction has removedthe intermediates. This is how genealogywould have to be discovered. Darwin didnot explicitly search for synapomorphies;he searched for any similarities that wouldreveal genealogy, as taxonomists still doon the basis of what seem to be valuablecharacters.

    As Ghiselin and Jaffe (1973) noted, Dar-win was formulating a system for histimes, and the times were not yet ripe for astrictly genealogical arrangementat leastone that would have had to be argued asstrongly as cladists have argued their phi-losophy since the 1970s. It is unlikely thatDarwin wished to take up that particularcudgel at the time (or could have), given hislarger plan. I think, with Ghiselin (1985), thatit is a mistake to read too much systematicphilosophy into these monographs. For onething, Darwin never classied any barnacleon the basis of genealogy, which is indis-putably regarded as his central criterion

    even though, by the 1850s, it clearly under-lay his ideas (as his 1843 letters to Water-house show). To argue on the basis of the

    barnacle monographs that Darwin regardedsimilarity as an equally important criterion,merely because he uses similarities in histaxonomy, makes no sense, because the en-tire concept of genealogy is suppressed.

    Why didnt Darwin bring genealogy morestrongly into his barnacle monographs? Thesimplest explanation, though circumstan-tial, is consistent with Darwins long secrecyon many of his views. To establish geneal-ogy as the basis of classication for barna-cles, he would have had to establish the ev-idence for the common origin and descentof all living things at the same time. Withhis great book in a premature stage of ges-tation, he obviously did not want to let thecat out of the bag too early (Desmond andMoore, 1991). Accordingly, he produced aconventional taxonomy that addressed theissues of his day. Recall that he undertookthe wholeexercise to begin with because of astray remark in a letter from Hooker (regard-ing Gerards views on species: CCD 3:250256), who seemed to suggest that Darwinsown theories might not be taken seriouslyunless he had himself labored in the taxo-nomic vineyards. Darwin used similaritiesnot to construct classications, but to revealcommon ancestry on which classicationsshould be based. His greatest impedimentsto this were the gaps among living forms,and between living and fossil forms. Heused neither primitive nor derived characterstates on principle; he did not mind what wewould call paraphyletic taxa if he couldnot resolve their membership more de-nitely; and he did not advertise his view thatclassication should be based on genealogy.

    CONCLUSION

    In his book on the fertilization of orchids(1862:330331), Darwin wrote,

    To make a perfect gradation, all the extinct formswhich have ever existed, along many lines of descentconverging to the common progenitor of the order,would have to be called into life. It is due to their ab-sence, and to the consequent wide gaps in the series,that we are enabled to divide the existing species intodenable groups . . . . If there had been no extinction,

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    12/13

    1999 PADIANDARWINS VIEWS OF CLASSIFICATION 363

    there would still havebeen great lines, or branches, ofspecial development,. . . butancient and intermedi-ate forms, very different probably from their presentdescendants, would have rendered it utterly impos-sible to separate by distinct characters . . . one great

    body from [an]other.

    Because living things have suffered differ-

    ential extinction, we have only incompleteseries of related forms; because we cannotseriate their entire genealogies, we must relyon their shared similarities to reconstructthat genealogy. This seems to me the clos-est approach that we can make to Darwinsview of classication.

    The preponderance of available evidenceindicates that Darwins only criterion forclassication was genealogy, a point he sum-

    marized in the last chapter (p. 486) of theOrigin of Species, when he was describingthe effects that will accrue (again, in a con-ditional future) to natural history if andwhen common descent is recognized as thecentral underlying principle of biologicalphenomena: Our classications will cometo be, as far as they can be so made, genealo-gies; and will then truly give what may becalled the plan of creation. The rules for clas-

    sifying will no doubt become simpler whenwe have a denite object in view. This lastsentence, as the reader will see, recurs tothe phrase in the letter he wrote long beforeto Waterhouse, quoted above. Hence, I con-clude, it is difcult to accept Mayrs case fordual criteria of genealogy and similarityin Darwins views on classication. There iscertainly enough ambiguity to allow roomfor interpretation in many of his passages,

    taken by themselves, but in larger contextthe argument does not hold up very well.The interpretation that Darwin used simi-larities to get at genealogy seems to explainmore of the available evidence. Moreover,Darwin in taxonomic practice was not par-ticular about what we would call primi-tive and derived similarities, nor aboutparaphyletic groups when strict genealogycould not be discovered. It is not relevant to

    force Victorian sensibilities into 20th centurymodes. Just as Darwin was neither a cladistnor a Linnean, it is not historically faithfulto graft the taxonomic philosophies of theModern Synthesis or those of phylogeneticsystematics onto his ideas and practices.

    ACKNOWLEDGEMENTS

    I thank Stan Rachootin, Adrian Desmond, Kevin deQuieroz, David Cannatella, David Lindberg, WendyOlson, M. P. Winsor, Keith Benson, Mickey Rowe, and

    Jeremy Ahouse for comments and helpful conversa-tions, while holding them blameless for any errors onmy part. I am particularly grateful to Michael Ghiselinfor his patience and generosity with ideas, and to PeterStevens for his trenchant insight on the difference be-tween classication and arrangement. Special thanks toPeter-John Leone and Robin Smith for providing copiesof Darwins edited correspondence, and to Jim Valen-tine for the long loan of rare Darwiniana. This is Uni-versity of California Museum of Paleontology Contri-

    bution No. 1685.

    REFERENCESBARRETT, P. H., D. J. WEINSHANK, AND T. T.

    GOTTLEBER. 1981. A concordance to Darwins Ori-gin of Species, 1st ed. Cornell Univ. Press, Ithaca, NewYork.

    BROWNE, J. 1995. Charles Darwin: Voyaging. Prince-ton University Press, Princeton, New Jersey.

    BURKHARDT , F., AND S. SMITH, eds. 1985. The corre-spondence of Charles Darwin, Volume 2: 18371843.Cambridge University Press, Cambridge, England.

    BURKHARDT , F., AND S. SMITH, eds. 1988. The corre-spondence of Charles Darwin, Volume 4: 18471850.

    Cambridge University Press, Cambridge, England.BURKHARDT , F., AND S. SMITH, eds. 1990. The corre-

    spondence of Charles Darwin, Volume 6: 18561857.Cambridge University Press, Cambridge, England.

    BURKHARDT , F., AND S. SMITH, eds. 1991. The cor-respondence of Charles Darwin, Volume 7: 18581859 (supplement 18211857). Cambridge UniversityPress, Cambridge, England.

    BURKHARDT , F., AND S. SMITH, eds. 1993. The corre-spondence of Charles Darwin, Volume 8: 1860. Cam-

    bridge University Press, Cambridge, England.DARWIN, C. 1851a. A monograph on the fossil Lepa-

    didae, or pedunculated cirripedes of Great Britain.Palaeontographical Society, London.DARWIN, C. 1851b. A monograph on the sub-class Cir-

    ripedia, with guresof allthe species: The Lepadidae;or, pedunculated cirripedes. Ray Society, London.

    DARWIN, C. 1854a. A monograph on the fossil Bal-anidae and Verrucidae of Great Britain. Palaeonto-graphical Society, London.

    DARWIN, C. 1854b. A monograph on the sub-class Cir-ripedia, with gures of all thespecies: The Balanidae,(or sessile Cirripedes); the Verrucidae, etc., etc., etc.Ray Society, London.

    DARWIN, C. 1859. On the origin of species by meansof natural selection, or the preservation of favouredraces in the struggle for life. John Murray, London.

    DARWIN, C. 1862. On the various contrivances bywhich British and foreign orchids are fertilised byinsects, and on the good effects of intercrossing. JohnMurray, London.

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/
  • 7/29/2019 DARWIN EVOLUTIONARY VIEW OF CLASSIFICATION.

    13/13

    364 SYSTEMATIC BIOLOGY VOL. 48

    DARWIN, C. 1871. The descent of man, and selectionin relation to sex, Volume 1. John Murray, London.

    DESMOND, A.J., AND J. MOORE. 1991. Darwin. TimeWarner, New York.

    ELLEGaRD, A. 1958. Darwin and the general reader.

    Goteborgs Univ. Arsskrift, Goteborg, Sweden.GHISELIN, M. T. 1969. The triumph of the Darwinian

    method. Univ. California Press, Berkeley.

    GHISELIN, M. T. 1985. Mayr versus Darwin on para-phyletic taxa. Syst. Zool., 34:460462.

    GHISELIN, M.T., AND L. JAFFE. 1973. Phylogenetic clas-sication in Darwins Monograph on the Sub-class Cir-ripedia. Syst. Zool. 22:132140.

    HULL , D. L. 1983. Darwin and his critics: The recep-tion of Darwins theory of evolution by the scienticcommunity. Univ. Chicago Press, Chicago.

    MAYR, E. 1982. The growth of biological thought.Belknap Press, Cambridge, Massachusetts.

    MAYR, E. 1994a. Cotylorhynchus: Not a mammal. Sci-ence 264:1519.

    MAYR, E. 1994b. Ordering systems. Science 266:715.

    MAYR, E. 1995. Systems of ordering data. Biol. Phil.10:419434.

    NAUDIN, M. 1852. Considerations philosophiques surlespece et la variete. Rev. Hort. (Ser. 4) 1:102109.

    NELSON, G. 1974. DarwinHennig classication: A re-ply to Mayr. Syst. Zool. 23:452458.

    NEWMAN, W. A. 1993. Darwin and cirripedology.Pages 349434 in Crustacean issues (History of Car-

    cinology, Volume 8). A. A. Balkema, Rotterdam, TheNetherlands.

    OSPOVAT, D. 1981. The development of Darwins the-ory: Natural history, Natural Theology, and naturalselection, 18381859. Cambridge Univ. Press, Cam-

    bridge, England.PADIAN, K. 1994. Ordering organisms. Science

    265:1017.SIMPSON, G. G. 1961. Principles of animal taxonomy.

    Columbia Univ. Press, New York.

    Received 21 August 1998; accepted 11 November 1998

    Associate Editor: R. Olmstead

    atUniversityofGhana,Legonon

    February6,2013

    http://sysbio.oxfordjournals.org/

    Downloadedfrom

    http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/http://sysbio.oxfordjournals.org/