© Det här verket är upphovrättskyddat enligt Lagen (1960:729) om upphovsrätt till litterära och konstnärliga verk. Det har digitaliserats med stöd av Kap. 1, 16 § första stycket p 1, för forsk- ningsändamål, och får inte spridas vidare till allmänheten utan upphovsrättsinehavarens medgivande. Alla tryckta texter är OCR-tolkade till maskinläsbar text. Det betyder att du kan söka och kopie- ra texten från dokumentet. Vissa äldre dokument med dåligt tryck kan vara svåra att OCR-tolka korrekt vilket medför att den OCR-tolkade texten kan innehålla fel och därför bör man visuellt jämföra med verkets bilder för att avgöra vad som är riktigt. This work is protected by Swedish Copyright Law (Lagen (1960:729) om upphovsrätt till litterära och konstnärliga verk). It has been digitized with support of Kap. 1, 16 § första stycket p 1, for scientific purpose, and may no be dissiminated to the public without consent of the copyright holder. All printed texts have been OCR-processed and converted to machine readable text. This means that you can search and copy text from the document. Some early printed books are hard to OCR-process correctly and the text may contain errors, so one should always visually compare it with the images to determine what is correct. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 C M 0 1 2 3 4 5 6 7 8 9 10 11 12 IN C H Göteborg University Faculty of Natural Sciences Tree-thinking and Nemertean Systematics Mikael Härlin Dissertation Department of Zoology Section of Zoomorphology Medicinaregatan 18 S-413 90 Göteborg Sweden Avhandling för filosofie doktorsexamen i zoologi (examinator: Professor Theo van Veen), som enligt den biologiska sektionsstyrelsens beslut kommer att offentligt försvaras fredagen den 26 april 1996 kl. 10.00 i föreläsningssalen, zoologiska institutionen, Medicinaregatan 18, Göteborg. Disputationen sker på engelska Fakultetsopponent: Dr. Jon L. Norenburg, Department of Invertebrate Zoology, Smithsonian Institution, Washington D.C., USA. . & -• i "Jag sa : Kom s å ger vi fan i debatten om betydelsen i konsten. Här ha r vi penslar och färg, det räcker till och vi målar liksom som vi vill. För livet är lek och spel, och vi leker våra spel, och går en stund på jor den, ja vi går e n stund på jorden". Ulf Lundeli - Och går en stund på jorden (1976) "Du kan foga de j i flocken Du kan kuva ner din kraft Men aldrig se dej i dom ögon du en gång haft" Ulf Lundell- Laglös (1982) •mmmsii "Th e act of arranging things into groups seems virtually indispensable to every sort of intellectual endeavor. In all branches of learning we encounter kinds and classes. Chemists have their molecules and elements, botanists their phanerogams and cryptogams, moralists their sins and peccadillos, grammarians their nouns and verbs, dramatists their comedies and tragedies. It is hard to imagine how we could get along without classifying. We cannot generalize when we concern ourselves merely with single obje cts. Ev en where we are mainly interested in one individual thing, we compare it and its attributes to others. Whatever the topic of discourse - be it gods, organisms, or statues - we inevitably talk and think about groups. He nce the problems of classification should interest everyone who seeks to view the various branches of knowledge in a larger perspective" Michael T. Ghiselin (1980) Tree-thinking and Nemertean Systematics Mikael Härlin Department of Zoology, Section ofZoomorphology, Göteborg University, Medicinaregatan 18, S-413 90 Göteborg, Sweden. E-mail: m.svensson@zool.gu.se Abstract: Classification and systematization are fundamentals in the organismal world because they form the basis for our ability to generalize and distinguish various things. Systematics is thus of major imp ortance in biology. This thesis is a result of phylogenetic and biogeographic studies of a group of marine nemerteans (Nemertea, Hoplonem ertea, Eurepta ntia). The main theme is trees and tree- thinking, covering aspects of phylogenetic systematics like homology vs. homoplasy (I), cladistic analysis (II, III), historical biogeography (III, IV), and philosophy of taxonomy and systematics (V, VI). Using a set of more than 100 cladistic analyses (I) we show that nemerteans are no more homoplastic than any other groups of animals or plants. Furthermore, we argue that homoplasy and homology are a strictly historical concepts without relevance outside a phylogenetic hypothesis. Hence , a phylogenetic analysis is a prereq uisite for all homology/ho moplasy statements. The cladistic analysis of the eureptantic nemerteans (II) is the first phylogenetic hypothesis of the group and shows that the previous classification, in many parts, is non- monophyletic. Hence there are no logical arguements for keeping the old classification. Therefore, the hypothesis is used as the basis for a systematization of these nemerteans (VI), where all names strictly refers to monophyletic assemblages. In the same paper (VI) I also provide a historical and theoretical account of tree-thinking in nemertean systematics. Paper V explores the philosophy of names and is thus a theoretical basis for the systematization in paper VI. The evolution of the eurepantic nemerteans is studied from a biogeographical point of view in paper III. Eastern Indian Ocean is suggested as the most likely present area being a part of the ancestral area (sea) of the Eurep tantia. Besid es studying the evolution of the Eurep tantia, I present a hypothesis of the historical relationships between the modem oceans. This latter hypothesis, is based on a cladistic biogeographic analysis of a combined data set of eureptantic nemerteans and acanthuroid fishes. I also provide a discussion on concepts and components in cladistic biogeography (IV). Key words:- Cladistics, Phylogeny, Classification, Systematics, Systematization, Taxonomy, Hom ology, Homo plasy, Taxa, Clade, Evo lution, Refe rence, Meaning, Individual, Philosophy, His tory, Names, Definition, Diagnosis, Biogeography , Evol ution, Nemertea, Hoplonemertea, Polystilifera, Eureptantia Göteborg University 1996 ISBN 91-628-1952-6 This thesis is based on the following papers, referred to by their Ro man numerals in the text. HÄ RLI N and SVEN SSON refers to the same person. I. SUNDBE RG, P., AND M. SVEN SSO N. 1994. Homoplasy, chara cter function, and nemertean systematics. Journal of Zoology, London 234: 253-2 63. II. HÄ RL IN, M., AND P. SUNDBE RG . 1995 . Cladistic analysis of the eureptantic nemerteans (Nemertea, Hopl onemertea). Invertebr ate Taxon omy 9: 1211-1229. III. H Ä R LIN, M. 1996. Biogeographic patterns and evolution of the eureptantic nemerteans.Biolo gical Jour nal of the Linnea n Society (In press). IV. H Ä R LIN, M. (submitted). Concepts and components in cladistic biogeography. V. HÄRL IN, M., AND P. SUNDBE RG , (submitted). Taxonomy and philosophy of names. VI. H Ä RL IN, M. 1996. Tree-thinking and nemertean systematics, with a systématisation of the Eurep tantia. H ydrobiologia (accepted). Co ntents Preface Introduction (Page 1) Trees and Clades (3) Naming Clades (9) Phylogeny and Bio geography of the E ureptantia (14) Summary (18) Acknowledgements (18) Refere nces (20) Papers Homop lasy, Char acter Function and Ne mertean Systematics (I) Cladi stic Analysis of the Eureptantia (Nem ertea : Hoplone mertea) (I I) Biogeog raphic Pa tterns and the Evolution of Eureptantic Nem erteans (II I) Conc epts and Co mponents in Cl adistic Biogeo graphy ( IV) Taxo nomy and Philosop hy ofN ames(V ) Tree- think ing and Neme rtean Systematics, with a Systematization of the Eureptantia (V I) Front cover by Carina Härli n Preface In Octo ber 179 1 George Cuvier wrote a letter to Christian Heinrich Pfaff expressing his views on Jussieu's wo rk. "His merits is not really in very detailed descriptions, or in a large number of species being described, which is often only a sign of a lack of criticism, but it is in the philosophical manner of seeing things, and finding the delicate threads by which plants may be held together, making the whole a painting" (from St evens, 1994). My impression is that something similar goes for the present thesis, even though not claiming that it is comparable to the work of Jussieu. Nevertheless, the merits of this thesis is not in the descriptions of nemertean morphology, others have done, and continue to to do that much better than I do, but rather in the study of relationships among natural groups. One o ften hears that no hypothesis is stronger than the observations (data) used to erect the hypothesis with the corollary that we must study our organisms more carefully. This is probably true, but it is based on the notion that observation and theory are independent. Howev er, I believe it is equally true that no observations (data) are stronger than the hypothesis within which the observations are made. Thus, there are no such things as theory independent observations. My hope is that the reader will keep this in mind when browsing this thesis. Mikael Härl in Göteborg February, 1996 / V - Ï9W;:M30SSPMÏM§ "The great tree of life which fills with its dead and broken branches the crust of the earth, and covers its surface with its ever branching and beautiful ramifications" (Darwin, 1859 ) Intr oduction Natural order, in various guises, was the aim for scientists long before Darwin. The Greeks searched for the great chain of being ( scala naturae), where continuity of characters and taxa from "sim ple" to more complex were of paramount importance. This idea have persisted consciously, or unconsciously, as a central tenet in western philosophy of natural order ever since (Lovej oy, 1936; Stevens, 1994). In France during the 18th and 19th centuries Lamarck, Jussieu, and to some extent de Candolle and Cuvier strived to represent the natural order as a continuous series of taxa from simple to complex in their classifications, that is, the scala naturae was the world-view for them and their contemporaries. It was, however, becoming common to view natural order as an analogy with geographic maps, de Candolle and Cuvier, for instance, argued more in favour of discontinuities but without making a clear brake with the tradition of continuity (Stevens, 1994). Stevens (1994) even argues that the notion of continuity lived on, at least unconsciously, well into the 20th century. Exa mples of this can be seen within nemertean systematics (paper VI; Sundberg et ai, 1996). Jussieu's work is generally acknowledged to be the most influential of early attempts to elucidate the structure of natural relationships. His a pproach was widely adopted by botanists and it also influenced zoologists like Cuvier. By the end of the 18th century it had become commonplace to claim that analytical classifications (systems), in which groups where successively subdivided, were necessarily artificial while the formation of successively larger groups (upward classification, method or synthesis), was believed to lead to natural arrangements of organisms (Stevens, 1994). Ou tside France, scientists at this time did not discuss the natural system explicitly even though their systems more or less were reflections of the scala naturae, or at least of the continuity of taxa and characters. It was not until 185 9, when Darwin published his now famous The Or igin of Species, the world-view changed and the tree of life became the metaphor of natural groups, and evolutionary history have occupied the human mind ever since. Geological information and the geographic distribution of organisms, among other things, led Darwin to introduce an evolutionary world-view and among ".. . historians of science, the Darwinian revolution' has always ranked alongside the 'C opernican revolution' as one of those episodes in which a new scientific theory symbolizes a wholesale change in cultural values" (Bo wler, 1989). Biog eographic data was of maj or importance for Darwin - 2 - when reaching his evolutionary theory. Ghiselin (1969) even argues that "In The Origin of Species, the strongest positive argument for evolution is the geographical one Darwin (1859: 300-303) himself wrote "In considering the distribution of organic beings over the face of the globe, the first great fact which strikes us is, that neither the similarity nor the dissimilarity of the inhabitants of various regions can be wholly accounted for by climatal and other physical conditions" and continues "A second great fact which strikes us in our general view is, that barriers of any kind, or obstacles to free migration, are related in a close and important manner to the differences between the productions of various regions" and "A thi rd great fact... is the affinity of the products of the same continent or of the same sea, though the species themselves are distinct at different points and stations". He concludes "We see in these facts some deep organic bond, throughout space and time, over the same areas of land and water, independently of physical conditions. The naturalist must be dull who is not led to inquire what this bond is. The bond is simply inheritance, the cause which alone, as far as we positively know, produces organisms quite lik e each other, or, as we see in the case of varieties, nearly alike". For Darwin and contemporaries like Wallace, biogeographic distributions was used as an argument for evolution while today, phylogenies are also used to study the historical distributions of biological clades (IE, IV) . Darwin used a single illustration in The Origin of Species - a phylogenetic tree. The metaphorical tree was important for Darwin when describing and discussing the evolutionary history of species, and it was furthermore the beginning of what now is known as phylogenetic systematics. Even though systematists grasped the theory of evolution and incorporated it into their work, they did so only as an after-the-fact explanation of their taxonomic groups. Thus, the same taxa that was recognized before Darwin were still valid after Darwin's book. Evolution was only used to explain their existence. It took almost 100 years from the publication of The Origin of Species until someone provided a method for detecting natural relationships. Zimmerman (1931) and Hennig (195 0, 1966) developed a method, or rather, a philosophy for how to detect genealogy (phylogeny). They were the first who clearly distinguished monophyly from paraphyly and argued that only monophyletic assemblages are natural groups. If Darwin started a revolution in human thinking in general and biological thinking in particular, Henn ig provided the basis for a revolution within biological systematics. Today, phylogenetic systematics qualif ies as a school book example of a scientific revolution sensu Kuhn (1970). Anomalies are accumulating and basic concepts like species, taxon, homology are questi oned and rapidly changing meaning. Present tree-thinking is quite different from Darwin's and his contemporary practising taxonomists (paper VI). Species concepts within the Darwinian paradigm may serve as the best example of a scientific revolution; from a biological or neo-darwinistic species concept to the cladistic or phylogenetic species concepts discussed today (e.g. Ehre shefsky, 1992; Baum and Donoghue, 1995). One major advantage to be gained from these discussions is the importance to distinguish phylogeny from taxonomy (VI) and thus relegating the ranking - 3 - problem to a secondary and arbitrary endeavour. The result will be that we get over rather than solve the problems with species concepts and taxa in general. He nce, "... intellectual progress usually occurs through sheer abandonment of quest ions together with both the alternatives they assume - an abandonment that results from their decreasing vitality and a change of urgent interest. We do not solve them: we get over them. Old que stions are solved by disappearing, evaporating, while new quest ions corresponding to the changed attitude of endeavour and preference take their place" (Dewey, 1910). The main theme in this thesis is trees and tree-thinking, covering aspects like homology and homoplasy (I), cladistic analysis (II, III), historical (cladistic) biogeography (III, IV), philosophy of taxonomy and systematics (V, VI). In this summary I will try to develop some of the ideas in these papers and provide a discussion aiming to put the papers into a broader context. I also give a short review of the phylogenetic and biogeographic results obtained in papers II and in. Trees and Clades Begin ning with Darwin, the tree-icon have developed into one of the most important and influential illustrations in biology. Darwin argued that descent with modification is the cause for such a tree-like pattern. Fairly little have been devoted to this aspect of evolution compared with how much effort that have been put into the process of natural selection. Haeckel (1866,1868) developed and discussed the tree-icon of evolutionaiy history in more detail and was also the one who popularized Darwin's ideas (Opp enheimer, 1987). He transformed Darwin's diagrammatic tree into more tree like pictures like oak trees (Op penheimer, 1987), thus, in a sense he illustrated Darwin's words (the citation in beginning of this summary may serve as an example of Darwin's writing). Oppe nheimer (1987) , however, argues that Haeckel ra ther than trying to illustrate Darwin's words tried to improve on nature. As a skilful artist he was not always so accurate with facts and he was often accused of scientific falsification (Opp enheimer, 1987). The more artistic (or derivations of it) way of illustrating phylogenetic trees have dominated many text books in the 20th century (e.g. Romer and Parsons, 1978; Willmer, 1990) which often makes them difficult to interpret. Cladistics returned to the Darwinian way of illustrating phylogenetic relationships. Howev er, the interpretation of trees have varied (O'Ha ra, 1992; paper VI) through time. In 1895, for instance Biirger made c omments like "Hubre chtia desiderata ist die einzige mir bekannte Nemertine, whelche die Klu ft zwischen Proto- und Heter onemertine überbrückt" and "Mit d ieser an Formen ausserordentlich reichen Familie [Lineidae ] schliess t die Ordnung de r Heteronemertinen ab . Sie steht and der Spitze des einen Haupta stes, whelcher von dem kurzen Stamme der Protonemertinen abgeht..." . In a similar vein Brinkmann (1917: flg. 28) argues that "Die Entwicklungsreihe Probalaenanem ertes - Bal aenanemertes zeight auch, was dieses Organ [Blindda rm] betrifft, eine schöne Reih e - 4 - fortscreitender Red uktionen..." . Similar statements are common at this time and they indicate a core of continuity (scala naturae) rather than discontinuities despite their relationship being depicted as a tree (VI), and O'Ha ra (1992) argues that sequ encing of contemporary taxa in a link from primitive to more advanced is a common narrative device in natural history. Nemertean systematics still suffers from a inconsistent interpretation of evolutionary history and a seriously flawed tree-thinking (VI; Sundberg et al., 1996; Sundberg, 1993). One re ason for this is the continuous emphasis on anagenetic change as a useful tool in both phytogeny reconstruction and taxonomic practice. The introduction of cladistic philosophy (Hennig, 1966) changed the way systematists view and interpret phylogenetic trees. Contemporary taxa are no longer sequ enced from primitive to advanced. Instead they are viewed as sister taxa sharing a common ancestry. This is a corollary of only cladogenetic events being recognized as important in the sense of conveying evolutionary relationships. Anagenetic change can never be q uantified, and in such a continuum only arbitrary delineations can be made (Lidén, 1990; Frost and Kluge , 1994). Cladogenetic events, on the other hand, offer non-arbitrary delineations of historical entities. Given the phylogenetic hierarchy, it is important to distinguish between phylogeny and taxonomy (VI), or, between grouping (phytogeny reconstruction) and ranking (taxonomy) in the words of Donoghue (1985 ). O'Ha ra (1988, 1992, 1993) in his historical accounts of systematics called it chronicle and history and identified it as a general distinction made in all historical sciences. A chronicle is a series of statements arranged in chronological order but not accompanied by any explanation or interpretation, while a historical statement contains the explanation of events and interpretations of their significance. Using the terminology of O 'Hara (1988), phytogeny is the chronicle and the taxonomy (classification, or rather systematization) is a historical statement representing the phylogeny. A major advantage gained from separating phylogeny from taxonomy is that both can be investigated in some detail, without too much confusion (V). Phytogeny is fractal (self-similar) (Green, 1991) and as such possible to generalize. Given the historical and fractal nature of phytogeny, we must realize that we live in the midst of the evolutionary process giving rise to the ever branching phylogenetic hierarchy and our statements (taxonomy) about segments (clades) of this hierarchy are historically constrained (O'H ara, 1993; Danto, 1985; Atran, 1990). O'H ara (1993) used an analogy between cartography and phylogenetic trees to illustrate the nature of generalizations and their importance for phylogenetic systematics. Just as maps are representations of the earth and subjec ted to what is called cartographic generaliztion, so are diagrams of the natural system like phylogenetic trees representations of the evolutionary chronicle. Depending on ones terminal units (genes, organisms, clades) the hypotheses of monophyly may differ. For instance, two genes may have different evolutionary histories and at the level of organism or clade they may lead to different hypotheses of monophyly. Thus, it is important to keep in mind that we search for monophyletic assemblages from an a priori chosen starting point, i.e. monophyly is a relative concept (Lidén, 1990; O'Har a, 1993; Frost and Kluge , - 5 - 1994). Lidén (1990) illustrated this by pointing out that the zygote is not a monophyletic unit itself with regard to its cells, but together with all cells descending from it, it forms a monophyletic assemblage. How to relate characters and organisms to groups and groups to organisms and characters is an old ques t in biological systematics (Stevens, 1984, 1994). A unifying concept in this discussion have been homology. The distinction between affinity (recognized by homology) and parallelism (recognized by analogy) was first made by MacLeay (1821), but later discussed and clarified by Owen (1843) and Strickland (1846). The concept of homology itself is, however, much older and Russell (1916) traces it back to Aristotle. Inference of animal relationships and constructing classifications were, in the early 19th century, based on comparative anatomy as exemplified by the studies of Owen, Cuvier and Geoffroy (see Panchen, 1994 for further discussion). The data were obtained by dissection and close observation. Owen (1843) distinguished analogue (a part or organ in one animal which has the same function as another part or organ in a different animal) from homologue (the same organ in different animals under every variety of form and function). At this time homology was more or less equa l to similarity. Hence , if a organ was considered to be similar enough in two species - it was considered homologous. More specifically, according to Owen, two structures are homologous because they are derivations of the same structure in the archetype. Owen's archetype represented a generalized and primitive condition from which, for instance, the skeletons of real vertebrates were derived - the human skeleton, representing the nearest approach to perfection, was furthest removed from the archetype (Panchen, 1994). Rupke (1993) argues that this kind of distinction is of considerable theoretical importance not only to Owen 's views but to all non evolutionary explanations of homology. As mentioned by Panchen (1992), Owen' s concept is Platonic, i.e. similar to Plato's theory of forms or ideas. In the beginning these theories were generalizations from classes of object s. Later, however, Plato claimed that the only real things were the ideas. This is exemplified by his famous metaphor of the cave, where prisoners in the cave are constrained so that they can see only the wall opposite the entrance; the ir view of the world outside (the world of ideas) is simply that of shadows cast by objec ts in the real world. The shadows are the objects of sensory perception: the real world consists of ideas. Hence, at t his time homology was a non evolutionary concept. For Darwin (1859 ) the meaning of homology is different. In the glossary he explains homology as "Tha t relation between parts which results from their development from corresponding embryonic parts, either in different animals as in the case of the arm of man, the fore-leg of a quadrupe d, and the wing of a bird; or in the same individual, as in the case of the fore and hind legs in q uadrupeds, and the segments or rings and their appendages of which the body of a worm, a centipede, &., is composed. The latter is called serial homology. The parts which stand in such a relation to each other are said to be homologous, and one such part or organ is called a homologue of the other" ( emphasis in original). Analogy, on the other hand is explained as "The resem blance of structures which - 6 - depends upon similarity of function, as in the wings of insects and birds. Such structures are said to be analogous, and to be analogues of each other". Exp laining his views further, Darwin (1859 :351) argues "... if I do not greatly deceive myself, ... the Natural System is founded on descent with modification;- that the characters which naturalists consider as showing true affinity between two or more species, are those which have been inherited from a common parent, all true classifications being genealogical;- that community of descent is the hidden bond which naturalists have been unconsciously seeking, and not some unknown plan of creation, or the enunciation of general propositions, and the mere putting together and separating objec ts more or less alike". T hus the Darwinian view of homology is one of characters related through common descent. At some points, Darwin is very close to the phylogenetic view of homology (or rather phylogenetic systematics is very close to Darwin's views) as determined by congruence among all possible characters. He w rites "T he importance, for classification, of trifling characters; m ainly depends on their being correlated with many other characters of more or less importance" (Darwin, 185 9). Despite, the Darwinian heritage of a phylogenetic view of homology, the concept is still controversial and often discussed (e.g. paper I, IV; Ghiselin, 1984; Rieppe l, 1992; Panchen, 1994; Lauder, 1994; Nelson, 1994). In fact, an edited book was recently devoted to the topic (H all, 1994). Even am ong advocates of a phylogenetic or cladistic homology concept, there sometimes is a (maybe unconsciously) echo of pre-Darwinian reasoning. For instance, Lauder (1994) in his discussion of a phylogenetic homology concept agrees with Patterson (1982) who argues that homologous similarities are those that define natural or monophyletic groups of organisms. How ever, natural groups do not posses defining properties (Ghiselin, 1966, 1969, 1981, 1984, 1987, 1995; Hull, 1989; II, IV, V). Natural groups are like individuals rather than classes and as such they cannot be defined, they are whether we find them or nor. Concreteness and non-instantiability are the fundamental criteria of individuality (Ghiselin, 1995 ). The individuality thesis developed by Ghiselin (1966, 1969, 1974 , 1987) and later by others (e.g. Hull, 1989) have had major impact on systematic thinking and is also the foundation in the present thesis. It is not only species that are individuals; a ll taxa are. Even though these ideas are getting more and more general acceptance it is not uncommon that clades (individuals) are defined rather than discovered. I believe it is fundamental to distinguish between two ways of viewing phylogenetic analysis in order to get a better understanding of the groups we are studying and avoid unnecessary class connotations. On the one hand a phylogenetic analysis can be seen as a tree building or reconstruction procedure and on the other as a way of choosing among possible trees. A tree building approach implies that the characters give the tree, i.e. that the clade exists because of the characters supporting it, while in the latter approach congruence among characters are seen as a means for choosing tree(s) from the pool of all possible trees. For a given number of terminal taxa there are a given number of trees, and these trees exists irrespective of the characters. For instance, with three taxa A, B, and C there are only three - 7 - possible resolved rooted trees. Eithe r is A sister to B leaving C out, or A is sister to C leaving B out, or B and C are sisters leaving A out (FIG. 1). C A B B A C A B C v vy FIGUR E 1. The three possible resolved rooted trees, given the terminal taxa A, B, and C. With more terminal taxa, the number of possible trees soon get astronomically high. Nevertheless, the task of a systematist i s to chose among possible trees (II, III). The tree- building approach fails because it does not distinguish between epistemology and ontology. The tree-choosing approach does, and is therefore more in line with the individuality thesis and therefore preferable within an evolutionary paradigm. A corollary of the individuality thesis is that the ordering in phylogenetic systematics should not be viewed as classification but instead as systematization (Griffiths, 1974; de Queir oz, 1988). Classification is the ordering of things into classes defined by their attributes, while systematization is the ordering into systems or wholes where parts are related through a process (Griffiths, 1974; see paper VI for a example and discussion). Within the phylogenetic framework discussed here, the most logical homology concept is one strictly coupled to tree topology. After an initial observation of structures in two or more organisms, our senses may tell us that the structures look very much the same, which in turn leads us to infer a horizontal similarity relation. These hypotheses, in turn serves as raw data in a congruence analysis (parsimony) of all possible characters which reveal the phylogenetic importance of any given character. The congruent distribution of characters in a phylogenetic tree is the relation called homology (I). Thus, it is not the presence of a specific character in an organism that make it a homology but rather its congruent relationships with all other possible characters in a tree. While the similarity relation is between two or more organisms in a horizontal fashion, the homology relation is a hierarchical concept based on congruence. This is a crucial advance from the pre-Darwinian homology concepts which only rested on similarity. Ghiselin (1966) puts the problem with similarity like this, "Similarity is a relation; things are not twice as "similar to" any more than they are twice as "around". When I say that x is hotter than y, I do not mean that it contains a greater number of hotter things, or a greater quantity of heat. When I say that rats are similar to mice, I do not mean that they are compose d of an equivalent number of comparable entities. If someone says that two organisms differ in 75% of a sample of characters, while two others differ in only 65% of these, he cannot meaningful assert that - 8 - one of these pairs possesses a greater amount of difference, in the same sense that one animal may be said to have more mass than another. One would obtain j ust as meaningful a figure by adding two oranges, a glass of water, and a telephone number" ... "For this reason, there can be no quan tifiable evolutionary rate in either the degree of morphological diversification or the level of anagenesis, and qua ntitative similarity is a metaphysical delusion". Trees and tree-thinking have been of major im portance in the advancement of biological systematics. With the tree as a central tenet in all systematic endeavours, concepts like homology and taxon have changed into a more precise and evolutionary meaning. Ont ological understanding as well as operational discovering procedures of phylogenetic phenomena have gained on the tree-icon. Not only phylogenetic systematics have gained from the development of tree-thinking, also historical biogeography, the field that meant so much to Darwin when formulating his evolutionary theory (above) have gained tremendous. Biog eographers work under the premises that biological diversity and distributions offer information on the history of geographic areas and waters, but also that geography is a potential explanatory source of biological diversity. With the development of cladistic or phylogenetic systematics both of these approaches have become more explicit. From a phylogenetic tree of the study group(s) a reduced area cladogram is estimated using one (or more) of several methods (e.g. component analysis (Nelson and Platnick, 1981; Page, 1988, 1993), Broo ks parsimony analysis (Wiley, 1988; Klu ge, 1988; Brook s, 1990), three-area statements (Nelson and Ladiges, 1991)). The main concern of these methods is the study of correlated patterns from independent taxa inhabiting the same areas (e.g. Nelson and Platnick, 1981; Hump hries and Parenti, 1986), with the aim to get a general pattern of an areas history. Information from as many independent clades as possible is thus considered important. Br emer (1992) brought back the focus to the study of the origin and evolution within a single clade, and thus closed in on Darwin's original view. He d eveloped a method to estimate how likely it is that a certain area inhabited by extant clade parts was part of the ancestral area (see below and paper HI for a discussion and example). Analysing the geographic distribution within a particular clade from a historical perspective contributes to the knowledge of the evolution of the group and have the potential to enhance the general understanding of the evolution of areas. One problem suffered in cladistic biogeography is that it have relied too much on a "pa ttern" cladistic approach, not incorporating the individuality thesis (IV). This have caused a confusion of epistemology and ontology and misunderstandings concerning widespread species and areas of endemism (IV). In conclusion, trees and clades have thus been of paramount importance in more than one historical aspects of biology. - 9 - Naming Clades During the time humans have classified (or systematized) they have also given names to the things classified. Naming is a way of handling our knowledge in order to be able to communicate, while classifications (systematizations) are necessary for generalizations. Single obje cts alone makes generalizations impossible (Ghiselin, 1980). A world with a name for every unique observ ation and thing would be a very isolated world. Wittgenstein (195 3) developed this to some extent when he explored the possibilities of a private language, something he in the end declared impossible. How names refer and the difference between meaning and reference are important not only in analytical philosophy. Issues like this may also obscure the advancement of systematics (V). There is an incongruence among systematists and biologists in general in how they use a taxon name when talking about a particular specimen. It is not so common to see that one refers to a specimen of Homo s apiens as a part of Hom o sapiens, rather it is much more common to see it referred to as a Homo sapiens-, implying the existence of several Homo sapiens. Rowe (1987) stumbles on a similar problem when arguing that "... one may easily delimit individuals such as Michael Ghiselin from other Homo sapiens..." (emphasis added). This I believe exemplifies both Ghiselin (1966, 1974) and de Queir oz (1988) point of an failure to distinguish class from individual but also an inability to use the language in a consistent way. The definition of taxon names are often taken for the meaning of the name rather than just being a method of fixing the reference, de Queir oz (1994) explicitly states that "... the meanings of taxon names are stated in terms of phylogenetic relationships rather than on organismal traits...". Putnam (1975 ) described the traditional theory of meaning with the following lines: "On the traditional view, the meaning of say, lemon, is given by specifying a conjunction of properties. For each of these properties, the statement 'lemons hav e the property F is an analytic truth; and if Pi, P2, Pn are all of the properties in the conj unction, then anything with all of the properties P j, ...., is a lemon is likewise an analytic truth". The conj unction of properties associated with the term is the intension and the intension determines the exte nsion of the term. Hence, the intension is the definition (meaning) of the name while the extension is everything that fits the definition. Often the intension of a term is taken for the essence of the kind or thing named (Schwartz, 1977 ), something which also is common in biological taxonomy even though biologists tend to reje ct the term essence. Donnellan (1966, 197 2) and Kripk e (1971 , 1980), among others, broke with this traditional view of names. They argued that names lack intension and conseq uently cannot be defined in a traditional sense. Kripke took his starting point in the notion that identity must exist by necessity, something that rule out contingent identity. While Frege explained contingent identity like "a = b" (given that "a = a" and "b = b") (see paper V), K ripke argued that it did not exist (yet another example of getting over, rather than solving a problem). - 1 0 - Traditionally, a priori and necessity have always walked hand in hand, as have a posteriori and contingency. Krip ke (1980) argued that this need not be the case, and that both necessary a posteriori truths and contingent a priori truths are possible. As an example of the relation between contingency and aprioricity, Kripke (1980) discusses the reference meter in Paris. The argument goes something like this. If we define 'one meter' to be the stick S at time to, is it then a necessary truth that stick S is one meter at time l(p. If we believe that everything we know a priori is necessary, we may argue that this is the definition of one meter. That is, by definition stick S is one meter long at to and conclude that it is a necessary truth. Kri pke disagrees, since the use of definition in this case does not give the meaning of what is called the 'meter ', but it fixes the reference. There is a difference between the two statements 'one m eter' and 'the le ngth of stick S at to'- Th e first phrase rigidly designates a certain length in all possible worlds, which in the actual world happens to be the length of stick S at to• The statement 'the length of stick S at to', on the other hand, does not designate anything rigidly. In some counterfactual situations the stick might have been longer and in others shorter, but the name 'one meter' s till refers to the same stick. He nce, it is not a necessary truth that S is one meter long at to- The reason is that one designator ('one meter') is rigid and the other designator ('the length of S at to') is not. Epis temologically we have a priori knowledge of 'on e meter' refe rring to 'the st ick S at to', but this does not necessarily mean that stick 5 at to' is one meter long. In this sense we can talk about contingent a priori truths. Hence, Kripk e argues that names lack intension and thus cannot be defined in a traditional sense. The referential part of a name is more important to Kripke and Donnellan. They argue that proper names refer independently of identifying (definite) descriptions. Donnellan (1966) showed that reference can take place not only in the absence of identifying descriptions but even when the identifying description associated with the name do not correctly apply to the individual to whom the name refers. Kripke (1980) called such a name a rigid designator since it refers to the same individual in all possible worlds in which it is present. Possible worlds are given by the descriptions we associate with them, i.e., a possible world is stipulated not discovered. In phylogenetic systematics, possible worlds belongs to the pool of all possible trees as discussed below and in paper V. A name (if rigid) will refer to the same individual whether or not he (or she or it) satisfies some list of commonly associated descriptions. When we use a name, Donnellan writes, we use it as a referential description to refer to some definite individual, independently of descriptions. If names are used to refer to whoever fits the identifying descriptions associated with them we get certain paradoxical results. For instance, we might have to conclude that an individual did not exist because he did not fulfil the definite description of his name. It is even impossible to discover things at all about carriers of proper names - because they are their names. In general, biological taxonomy have not kept up pace with the developments in the philosophy of names. With few exceptions, biological taxonomy is still to a large extent -11 - based on theories of meaning (V). Howeve r, early in the development of phylogenetic systematics, Ghiselin (1966, 1969, 1974) claimed that species are individuals rather than classes and argued that their names are proper. He also argued that such names only could be defined ostensively. Thus, the basis for a change in the philosophy of taxonomic names towards a reference system like Kripke 's (above and V) have been present just a s long as the philosophy for detecting phylogenetic relationships (Henn ig, 1966). Despite that, taxon names are commonly defined by the properties of a type specimen. The type specimen, in turn, is used to infer the extension of the taxon. This invokes one immediate problem concerning meaning and reference; they d o not have the same extension. The reference is to a particular individual (type specimen) while the meaning of the name covers this specific individual but also other specimen judged similar enough to the type. Furthermore, in this case reference and meaning belongs to two different metaphysical inquiries ; the re ference is to an individual while the meaning is a class of similar things (V). In the early 1990's, de Queir oz and Gauthier (1990, 1992, 1994) picked up where Ghiselin left off and approached the problem with a method for defining taxon names phylogenetically. They proposed three classes of phylogenetic definitions; node -, stem-, and apomorphy based definitions (FIG. 2). A B C A B C A B C vy v FIGURE 2. The three classes of phylogenetic definitions of taxon names suggested by de Queir oz and Gauthier (1990). A node-based definition (1) can be formulated either as the clade stemming from the most recent common ancestor of B and C (e.g. de Queir oz and Gauthier, 1990, 1992, 1994) or as the least inclusive clade comprising B and C (e.g. Schänder and Thollesson, 1995; VI). The stem-based definition (2) can be phrased either in the spirit of de Queiroz and Gauthier like "taxon B and C and all taxa sharing a more recent common ancestor with them than with A" o r in the words of Schänder and Thollesson (1995) "the most inclusive clade comprising B and C" . The apomorphy-based definition (3) refers to the clade stemming from the first ancestor possessing the particular apomorphy (indicated by a black cross-bar). The thicker black lines indicates the inclusiveness resulting from the various types of definitions. The apomorphy based definition have later been criticized for being ambiguous since it, given a change of phylogenetic hypothesis, can turn out homoplastic (Br yant, 1994; Schänder and Thollesson, 1995). Also the stem-based definition have been considered ambiguous since it may refer to a non-existing clade, given a change of phylogenetic hypothesis (Schänder and Thollesson, 1995). The node-based definition, however, seems - 1 2 - unambiguous within the system presented by de Queiroz and Gauthier. These effects are due to phylogenetic taxonomy being an application of the traditional theory of names (V). To a large extent, de Quei roz and Gauthier' s suggestions have passed rather silently in the literature both practically (but see VI) and theoretically. Beside s the contribution of some theoretical comments by de Queiroz (1992,1994,1995 ) himself, few other have entered the discussion (but see Bryant , 1994; Rowe a nd Gauthier, 1992; Sundberg and Pleij el, 1994; Schänder and Thollesson, 1995 ; Ghiselin, 1995; paper V). The comments that have been made are mainly positive, but none of them (except Ghiselin, 1995 ; and de Que iroz's own papers in 1992, 1994 and 1995 ) have considered the philosophical background (V). All of the contributions by the original authors are based on the assumption that common ancestry can be used as defining properties of names and de Que iroz (1994) specifically argues that "de fining formulas of taxon names can be stated in terms of logically necessary properties, in other words, that the names of taxa have intension" (emphasis in original). de Queir oz and Gauthier's system is an application of the traditional view of names as theories of meaning, although in a phylogenetic suite. Consider the following example. If the meaning of "CD" is given by the definite description "the least inclusive clade comprising C and D" (FI G. 3) the identity judge ment "CD is the least inclusive clade comprising C and D" b ecomes an analytic truth - a tautology; true through linguistic laws. Such statements can be false since it is possible that the specific clade never comprised C and D. In a possible world where another clade comprised C and D, the statement " the least inclusive clade comprising C and D" re fers to that clade, while the name "CD" still refers to CD in the actual world. In this sense proper names are rigid designators since they refer to the same individual in all possible worlds (FÏG . 3). This is not the case for definite descriptions since they refer to whatever fits the description. The name and the definite description can therefore not have the same meaning. "C D" "C D" FIGURE 3. CD refers to the least inclusive clade comprising C and D (1). Howeve r, C and D are not logically necessary properties of CD since the clade (C, D) does not exist all possible worlds, e.g. tree 2. This is furthermore an example of what can happen if definite descriptions like the least inclusive clade comprising C and D are considered the meaning of the name - the name will refer to whatever clade that fits the description (2). - 1 3 - Possible worlds are stipulated and not discovered, which means should our phylogenetic hypothesis change in a way affecting our named clade so does the resulting possible worlds. Howe ver, it is important that we consider the nature of a phylogenetic individual and what is meant by the same individual if we are to evaluate the two approaches to names and naming. K ripke's philosophy of names stem from a world view where we can observe the individual we name. We observe when a child is borne, and we then decide to give that particular child a name. Phylogenetic individuals, on the other hand, can never be observed since they are historical individuals. Such individuals are nested within larger individuals forming a phylogenetic hierarchy and there are no objective criteria for what clade to name. It is equa lly logical to name all as none. Furthermore, phylogeny is a ongoing process incorporating both births and deaths of clades. The crucial point, however, is that we never know when we find the true phylogeny and in that sense de Queiroz an d Gauthier's s ystem will work as a operational system since it always will refer to a phylogenetic individual, but maybe not the particular one intended in the actual world. Accepting de Queir oz and Gauthier's suggestion we must give up the idea that the clade named (taxon) in the actual world is a real thing, we must accept that it is a human construction made to facilitate communication of phylogeny. One major advantage with this approach is that we gain a sort of operationality but we lose the reality of the clades being named. Krip ke's appr oach, on the other hand, takes for granted that the individuals we have are real and that it is them we refer to by the name. This approach lose in operationality since it is possible that our phylogenetic hypotheses will change with improved methods of inference and with new interpretation of data. A possible way out of this dilemma is to argue that phylogenetic statements like "the least inclusive clade comprising C and D" always refer to the same individual in all possible worlds, even though the inclusiveness of the name will change. Thus even if our knowledge (epistemology) of the clade may change (i.e. inclusiveness), the ontology of the clade will not change. One draw back, of course, is that we then let our language determine what historical individuals to name rather than the individuals being the subjec t of naming. Hen ce, we give primacy to names rather than clades. Kri pke's philosophy, on the other hand give primacy to the clade being named and also allows for discoveries of new parts of that clade. de Queir oz and Gauthier (e.g., 1990) tend to confuse epistemology with ontology, and this may be the reason why they argue that taxon names are defined by logically necessary properties. For instance, they claim that phylogenetic definitions of taxon names distinguish between a definition and a diagnosis and argue that this is a difference between an entity itself (ontology) and evidence for its existence (epistemology) (de Queiro z and Gauthier, 1990). I do not agree, both definition and diagnosis are epistemological matters. The ontology of phylogenetic systematics are individuals (clades) and they cannot be defined (Ghiselin 1966, 1969). We can only make more or less accurate reference to particular clades. Furthermore, when arguing in favour of "pro perties" used in phylogenetic definitions of taxon names being defining and logically necessary properties, de Queir oz - 1 4 - (1995) uses an example aiming to define the name "Mammalia". He writes "B eing a part of a particular clade that eventually and contingently gave rise to horses and echidnas in the actual world is what is logically necessary to be a mammal, not being a part of a clade that immediately and necessarily gave rise to horses and echidnas in all possible worlds. Indeed, the existence of horses and echidnas is not even necessary to define the name of the clade/ance stor in ques tion, for house mice and platypuses, or monotremes and therians, would do just as well". This is a confusion of what is a priori knowable with what is metaphysical necessary. Contingency and necessity are metaphysical categories; the contingent obtaining in some worlds and the necessary in all. Clades are not necessary since they do not exist in all possible worlds (FIG. 3), unless we accept that whatever the phylogenetic definition of the name refers to is the same individual. Matters of epistemic primacy (what is a priori or not) are not quest ions about worlds but of knowledge. Metaphysical necessity is one thing, epistemic aprioricity another (Kripke , 1980; Nelson, 1992). The statement "C D refers to the least inclusive clade comprising C and D" (FIG. 3) is a contingent but a priori truth. Things could have been different, for instance we could have chosen to give the name "CD" to the least in clusive clade comprising A and B instead. Furthermore, the tree topology could have been different (FIG. 3, right). But given the present situation, the name "CD" is a matter of epistemic aprioricity, while the fact that it refers to a particular ontological individual is contingent. All clades are contingent and then- names are based on a priori knowledge of the evolutionary model. Defining and logically necessary properties are not part of a phylogenetic ontology. H ence, it is not a matter of necessity for a horse to be a mammal but given the present usage of the name, it is a matter of a priori knowledge. Phy togeny and Bio geography of the Eu rep tantia Eurep tantia (Nemertea, Ho plonemertea, Polystilifera) is a group of marine nemerteans distributed world-wide. The Indonesian Archipelago and the Mediterranean Sea are the two main areas of present distribution. Until recently (II, VI), the 46 described species were grouped into 9 families and 24 genera; many of them monotypic. This traditional classification (II, table 1) was mainly constructed on the views of Stiasny-Wijnhoff (1936, and references therein). Recen tly, we (II) analyzed the cladistic relationship among a majorit y of the eureptantic species, based on most of the characters that traditionally are used in eureptantic classification. As a comparison, we used the traditional classification as a constraint in an identical cladistic analysis. All genera, families, and higher categories were constrained to be monophyletic, but with their internal relationships unresolved. The original analysis resulted in six, highly congruent, most parsimonious trees with a consistency index (CI) of 0.30, a retention index (RI) of 0.62, and a tree-length of 175 steps. Even though the CI is slightly below the empirically expected (Sanderson and - 1 5 - Donoghue, 1989), it is not significantly lower than in any other group of animals or plants (paper I). The tree distribution is significantly skewed to the left (gl = -0 .355), indicating phylogenetic information in the data set (Hillis and Huelsenbeck, 1992). Inae quif urcata Aeq uifurcata Eurepta ntia FIGURE 4. Phylogenetic hypothesis of the Eur eptantia (II). The names from the systematization in paper VI are used in the tree. For instance, Eu reptantia refers to the least inclusive clade comprising Uniporus and Aequifurcata. The old names are mentio ned within parenthesis. Besid es, for systematic purposes it is not necessarily a data set with a high CI (i.e. little homoplasy) one need but rather a good possibility for tree choice. "... when actually analysing data cladists probably worry less when they find high levels of homoplasy and - 1 6 - one or a few trees, than when they find hundreds or thousands of equa lly parsimonious trees. That is the right thing to do" (Goloboff, 1991). Compared with the traditional classification (as resulting from the constrained analysis), the most parsimonious solution is 16 steps shorter. Furthermore, our data set deviate from a random set of characters since on average (based on 5 randomized data sets using the shuffle option in MacClade) such trees are much longer (230 steps), have lower CI (0.24), and are less skewed to the left (gl = -0 .082). In summary, the hypothesis in paper (II) and FIG. 4 offers the best estimate of eureptantic phylogeny presented up to date. Many of the previous recognized taxa are rendered non-monophyletic status as shown in FIG. 4. The results obtained in paper (II) ju stifies a reorganization of the eureptantic nemerteans in the form of a systém atisation (sens u Griffiths, 1974; de Queiro z, 1988). A new system of names is suggested in paper (VI, table 1) and shown in FIG. 4, which strictly make reference to phylogeny rather than to organismal traits or Linnean categories. All references are made to nodes in the phylogeny and as many of the traditional names as possible have been kept. H owever, I am eager to point out that even though some of the names refer to the same species, they do that on totally different grounds. For instance, Paradrepanop horus is traditionally defined by Stiasny-Wijnhoff (1936) as "Atr ium untief oder fehlend. Dorsale Gehirnganglia mit gegabeltem Faserkern. Blinddarm bi s neben dem Magen. Enteron taschen alle gleich gestaltet. Periphere Blu tgefässanastomosen. Proximale Nephridiopori. Gonadentaschen vielreihig bis paarweise angeordnet, vorzugsweise V-förmig gestaltet". Now, the name Para drepanophorus refers to " the least inclusive clade comprising P. corallinicola and P. ob iensis". While the former only make reference (is defined by) a set of characters, the latter make explicit reference to a clade. Thus, the difference is tree-thinking. By making explicit reference to a clade we avoid all unnecessary class connotations and furthermore, we make sure that the reference is to a historically relevant entity. The phylogenetic hypothesis of the eureptantic nemerteans is a first step towards an increased understanding of their evolution. Another step is to investigate their biogeographic relations (III). Given the phylogeny (II) and the known distribution of the Eurep tantia, I recognized six main areas of endemism (see paper IV on areas of endemism); Eastern Indian Ocean (EIO), Western Indian O cean (WIO), West ern Pacific (WP), Eastern Atlantic Oc ean (EA T), Western Atlantic Ocean (WAT), and the Mediterranean (ME ). Based on Breme r's (1992) ancestral area analysis, I identified EIO as the present area being most likely to be part of the ancestral distribution of the Eureptanti a (III). The reduced area cladogram based on the eureptantic phylogeny (paper III, fig. 3 and FIG. 5) suggests that the evolution of the Eurepta ntia have been dominated by vicariance events rather than dispersals. This is deduced from he high consistency index (0.87 ) which suggests a good fit of the characters on the tree topology. Furthermore, the phylogenetic tree (II and FIG. 4) is significantly asymmetric using two of the statistics (N= 11.26 and = 5.24 ) in Kirkp atrick and Slatkin (1993) with most statistical power for trees with - 1 7 - more than 20 species. One exp lanation for asymmetric trees could be that species living in geologically active areas experience frequ ent vicariant events resulting in a biased cladogenesis. A possible cause for the asymmetry in the eureptantic evolution could be that many clades evolved in south-east Asia, which is, and have been, a geologically active area. The many autapomorphies on the EIO branch (III, fig. 3 and FIG. 5) indicate the high cladogenesis in that area during the last 20-30 million years. sa w O w O » •M £ fe? £ ra S s 8 S Ë T 3 < S3 w 9 2 1 26 40 41 42 10 0 o '3 eu c s S3 £ 1 32 6 2-4 24-25 7 33 12 10-11 29 8 34 17 13-14 31 27 65 19 15 35-3 6 28 66 22 18 37 30 53 23 40 6 1 5 61 T3 C a 44 45 46 5 1 52 54 76 8 O T3 C e B 5 IS 23 37 49* 50« •Z 3 < 16 20 5 1 52 53 3 O 38 78 39 43 47 48 84 49 50 55-6 0 62-64 67 FIGURE 5. Red uced area cladogram based on the p hytogeny of the eurepta ntic nemerteans (in). - 1 8 - Summary Hum ans tend to classify or systematize most things, from the less important like nemerteans (e.g. paper VI) to more essential things in life like Scotch single-malt whiskies (e.g. Lapointe and Legendre, 1994). This is not surprising since systematics and taxonomy are more than biological fields of inqui ry, they are part of "a fun damental operation in the acqu isition of knowledge" (Ghiselin, 1981). "Taxonom y is [even ] older than thinking" (Ghiselin, 1981). Howeve r, it is important to keep in mind that there is not just a vast number of things to classify, but also many ways to order things into groups depending on the purpose with the classification. There is not necessarily a bad and good way or a right and wrong way to do systematics and taxonomy, rather certain approaches suite certain purposes better than others. Therefore, I believe that being consistent in ones approaches to taxonomy and systematics, and thus avoid confusing approaches, is of major importance. For instance, we should realize the distinction between classification and systematization (Griffiths, 1974; de Queiroz, 1988) and stop criticizing a classification for not doing the job of a systematization, and vice versa. Both approaches ca n be kept as long as we maintain an awareness of their totally different assumptions and possibilities. Ackno wledge ments Without certain people this thesis would probably never have turned out the way it did. First of all I would like to thank Mikael Svensson for writing the major part, but also Mikael Här lin for adding the finishing touch. Without you ( two) the thesis would never have been written. Per Sundberg provided lots of help and support throughout these years. He s ponsored my (and Mats') trip to all the pubs in the vicinity of Bangor where we were supposed to look for nemerteans. Unfortunately I did not find any (some would probably say I still have not seen any nemerteans), nevertheless it was a very pleasant way to start my PhD studies. Howev er, what I appreci ated the most with Per as a supervisor was his willingness to let me do whatever kind of research I pleased (even though I guess he wanted m e to work with Prostoma tella). Thanks a lot. Lots of people have contributed, one way or the other, to make this a pleasant time. During the first half of this thesis I spent a lot of time at " The Dubliner" drinking beer and playing dart and the second half I spent at "Gyllene Prag" eating cheese and drinking beer. Both of these places ar e gratefully acknowledged for providing both intellectual and relaxing atmospheres. Thanks all of you who shared beers with me at these and other places. To mention just a few, thanks Juha, Claes, Johan, Sven, and Mats. A special thanks to Ulf Lundell and Bruce Springsteen for writing lyrics and music that have stimluated me many times. -19 - Working at the "Gaz a strip" a t the Department of Zoology have been interesting. Down there we had the opportunity to develop a systematic part of the department. From being a warehouse of zoological waste it is now a well working systematic unit and from time to time interesting conversations are going on down there. Thanks Mats, Mikael, Chris, Urban, Lena, Jocke, Marianne, Susanne and all others showing up at Gaza every now and then, and good luck with the molecules. I wonder how many hours (days, weeks, years) Mats and I spent talking during these years, despite (or rather thanks to) that I finished a thesis which I am sure gained tremendously much from these discussions. Mats, thanks for all these strange, abstract and never ending discussions on systematics and life in general. The discussion group at the departments of zoology and systematic botany have been a nice injection , giving rise to several interesting conversations. I must admit that working with nemerteans is lots of fun. Thanks all of you who attended the two latest nemertean conferences in Bangor and Pacific Grove for making them so pleasant A bunch of people have commented and provided useful suggestions of improvement on various parts of this thesis and I thank you all. Howe ver, Mats Env all, Christoffer Schänder, Per Sundberg, and Mikael Thollesson read more than most and deserves a special thanks. I would also like to thank our former professor, Anders Enemar, for his efforts to get as many PhD grants as possible to the section of zoomorphology, which was something I benefited greatly from. Today it is a privilege to be financially supported throughout ones PhD studies, even though it should be self-evident. I am also grateful to our new professor, Theo van Veen, for providing a relaxed atmosphere at our department. My parents, Kurt and Marie-Louise, have always supported me in my hobbies and choice of studies. Their support and understanding have meant more to me than they probably realize. Thanks. My grandfather, Lasalle, was a man who I admired a lot. His attitude and spirit have stimulated me many times. Ooop s! I almo st forgot to thank my wife Carina. I do not dare to think what would have happened if I had. For some reason she often thinks that I work too much and I guess she is the only one with that opinion. And Carina, remember that a beer will always be at least a couple. Nevertheless, she is a tremendous support and invaluable friend. Without you my life would be boring. Thanks for being there, I love you. This study was financially supported by Rådman och Fru Ernst Collianders Stiftelse, Lennanders Stiftelse, various funds from Göteborg University, from the Royal Swedish Academy of Science, Anna Ahrenbergs fond för vetenskapliga ändamål, Wilhelm och Martina Lundbergs Vetenskapsfond, Kungliga och Hvitfeldtska Stipendieinrättningen, and GMF. They are all gratefully acknowledge for their generous support and despite all taxes and fees to the University, these grants covered most of my research. I hope that someone benefited from the taxes and fees. Onc e again - thanks everyone. - 2 0 - Re ferences ATRA N, S. 1990. Cognitiv e foundations of natural history: towards an anthropology of science. Cambridge University Press, England: Cambridge. BAUM , O. A., AND M. J. DONOGH UE. 1995 . Choosing among alternative "p hylogenetic" species concepts. Systematic Bota ny 20: 560 -57 3. BOW LER, P. J. 1989. Evolution. The history of an idea. California University Press. BREME R, K. 1992. Ancestral areas: a cladistic reinterpretation of the centre of origin concept. Systematic Biology 41: 436-445. BR INK MANN, A. 1917. Die pelagische Nemertinen. Berge ns Museums Skrifter 3:1-194. BROO KS, D . R. 1990. Parsimony analysis in historical biogeography and coevolution: methodological and theoretical update. Systematic Zoology 39: 14-30. BRYAN T, H. N. 1994. Comments on the phylogenetic definition of taxon names and conventions regarding the naming of crown clades. Systematic Biology 43: 124-130. BÜRGER, 0.1895 . Die Nemertinen des Golfes von Neapel und der angrenzenden Meeres- Abschnitte. Fauna und Flora des Golfes von Neapel. 22:1-7 43. DANTO , A. 1985. Narratio n and knowledg e. Columbia University Press: New York. DONNE LLAN, K. 1966. Reference and definite d escriptions. Philosophica l Review 75: 281- 304. DO NNEL LAN, K. 1972. Proper names and identifying descriptions. Pages 356-379 in Semantics and language (Harman and Davidson, eds.). Reidel Publishing Comp any, Dordrecht-H olland. DE QUIE ROZ, K. 1988. Systematics and the Darwinian revolution. Philoso phy of Science 55 : 238-259. DE QUE IROZ , K. 1992. Phylogenetic definitions and taxonomic philosophy. Biology and Philo sophy 7: 295-313. DE QUEIR OZ, K . 1994. Replacement of an essentia listic perspective on taxonomic definitions as exemplified by the definition of "Mammal ia". Syste matic Biology 43: 497-5 10. DE QU EIR OZ, K. 1995. The definitions of species and clade names: a reply tho Ghiselin. Biology and Philosophy 10: 223-228. DE QUE IROZ , K„ AND J. GAUTH IER. 1990. Phylogeny as a central principle in taxonomy: phylogenetic definitions of taxon names .Systematic Zoology 39: 307-322. DE QUE IROZ . K., AND J. GAUTH IER. 1992. Phylogenetic taxonomy. Annual Review of Ecology and Systematics 23: 449-480. DE QUEIRO Z, K., AND J. GAUTHIER . 1994. Toward a phylogenetic system of biological nomenclature. 7rendl s in Ecology and Evolution 9: 27 -31. DEW EY, J. 1910.7% influence of Darwinism on philosophy, and other essays in contemporary thought. Henry Hol t and Co., New York. - 2 1 - DONOG HUE, M. J. 1985 . A critiqu e of the biological species concept and recommendations for a phylogenetic alternative.ZÄe Bryologist 88: 172-1 81. EHRES HEFSK Y, M. 1992. The units of e volution. Essays on the nature of species. MIT Press, Massachusetts. FROST, D. R. , AND A. G. KLUG E. 1994. A consideration of epistemology in systematic biology, with special reference to species. Cladi stics 10: 259-29 4. GHISELIN, M. T . 1966. On psychologism in the logic of taxonomic controversies-Systematic Zoology 26: 207-2 15. GHISELIN, M. T. 1969. The triumph of the Darwinian method. California University Press, Be rkeley. GHIS ELIN , M. T. 1980. Natural kinds and literary accomplishments. The Michig an Qua rterly Revie w 19: 73-8 8. GHISEL IN, M. T. 1981. Categories, life, and thinking. The Behavioural and Brain Scien ces 4: 269-313. GHIS ELIN , M. T. 1984. "Defi nition", "cha racter" , and other equ ivocal terms. Systematic Zoology 33: 104-109. GHISEL IN, M. T. 1987. Species concepts, individuality, and objectiv ity. Biology and Ph ilosophy2: 127 -143. GHIS ELIN , M. T. 1995 . Oste nsive definitions of the names of species and clades. Biology and Philo sophy 10: 219-222. GREEN, D. M. 1991. Chaos, fractals, and nonlinear dynamics in evolution and phylogeny. Trend s in Ecology and Evolution 6: 333-337 . GRIFFITHS, G. C. D. 1974 . On the foundations of biological systematics. Acta Bioth eoretica 23: 85 -131. GOLO BOFF , P. A. 1991. Homopla sy and the choice among cladograms. Cladistic s 7: 215- 232. HALL, B. K. 1994. Homology. The hierarchical basis of comparative biology. A cademic Press Inc., San Diego. H AEC KEL , E. 1866. Allgemein Entwicklungsgeschichte der Organi smen. Georg Reimer , Berlin . HAEC KEL, E. 1868. Natürliche Schöfungsgeschichte . Georg Reimer, Berlin . HEN NIG, W. 1950 . Grundzüge einer Theorie der phylogenetischen Systematik. Deutsch er Zentralverlag, Berli n. HENNIG, W. 1966. Phylogenetic systematics. Univ. Illinois Press, Urbana. HlLLIS, D. M„ AND J. P. HUE LSEN BECK . 1992. Signal, noice, and reliability in molecular phylogenetic analysis. Journal of H eredity 83: 189-195 . HULL, D. L. 1989. The meta physics of evolution. SUNY Press, New Y ork. HUMPH RIES, C. J., AND L. R. PARENT! . 1986. Cladistic biogeography. Clarendon Press, Oxford - 2 2 - O'HARA, R. J. 1988. Homage to Clio, or, toward a historical philosophy for evolutionary biology. Systematic Zoology 37: 142-155. O'HA RA, R. J. 1992. Telling the tree: narrative representation and the study of evolutionary history. Biology and Philosophy 7: 135-160. O'HAR A, R. J. 1993. Systematic generalization, historical fate, and the species problem. Systematic Biology 42: 231-246. KIR KP ATRIC K, M„ AN D M. SLATKIN . 1993. Searching for evolutionary patterns in the shape of a phylogenetic tree. Evolution 47 : 1171-1181. KLU GE, A. G. 1988. Parsimony in vicariance biogeography: a quantitative method and a greater Antillean example. Systematic Zoology 37 : 315-328. KR IPK E, S. 1971. Identity and necessity. Pages 135-164 in Identity and individuation (M. K. Munitz , ed.). New York University. KRI PKE , S. A. 1980. Namin g and necessity. Blackwell, Oxford. KUH N, T. S. 1970. The structure of scientific revolutions, 2nd edition. Chicago University Press. LAPO EN TE , F-J., AND P. LEG EN DR E. 1994. A classification of pure malt scotch whiskies. Applied Statistics 43: 237-257. LAUDE R , G. V. 1994. Homology, form, and funct ion. Pages 152 -196 in Homology: the hierarchical basis of comparative biology (B. K. Hall, ed.). Academic Press, Inc. LIDÉN, M. 1990. Replicators, hierarchy, and the species problem. Cladistics 6: 183-186. LOV EJOY , A. O. 1936. The great cha in of being. Harv ard University Press, 382 pp. MACLEAY , W. S. 1821. Horae Entomologicae: or essays on the annulose animals, Vo l. I, Part II. C ontaining an attempt to ascertain the rank and situation which the celebrated insect, Scarabaeus sacer, holds among organised beings. S. Bagster, London. NE LSO N, G. 1994. Homology and systematics. Pages 102-149 in Homology: the hierarchical basis of comparative biology (B. K. Hall, ed.). Academic Press, Inc., San Diego. NE LSON , G., AND P. Y. LADIGE S. 1991. Three-area statements: standard assumptions for biogeographic analysis. Systematic Zoology 40: 470-485. NE LSON , G., AND N. I. PLATNICK . 1981. Systematics and biogeography. Cladistics and vicariance. Columbia University Press, New York. NE LSO N, R. J. 1992. Nami ng and reference. T he link of wo rd to object. Routledge. OPP EN HE IME R, J. M. 1987. Haeckel's variations on Darwin. Pages 123-135 in Biolo gical metaphor and cladistic classification (H. M. Hoenigswal d and L. F. Wiener, eds.). University of Pennsylvania Press. OW EN, R. 1843. Lectures on the compa rative anatomy and physiology of the invertebrate animals, delivered at the Royal College of Surgeo ns, in 1843. Longman, Brown, Gr een, and Longmans, London. PANCHE N, A. L. 1992. Classification, evo lution, and the nature of biology. Cambridge University Press, Cambridge and New York. - 2 3 - PANCH EN, A. L. 1994. Richard Owen and the concept of homology. Pages 21-62 in Hom ology. The hierarchical basis of comparative biology (B. K. Hall, e d.). Academic Press, Inc., San Diego. PATTE RSON , C. 1982. Morphological characters and homology. Pages 21-74 in Problems of phylogenetic reconstruction (K. A . Joysey and A. E . Friday, eds.). Academic Press, London. PUTNAM, H. 1975 . Mind, language, reality. Philosophical papers, volume 2. Cambridge University Press. RIEPPEL, 0. 1992. Homology and logical fallacy. Journal of e volutionary Biology 5: 701 -715 . ROMER, A. S., AND T. S. PARSON S. 1978. The vertebra te body. Shorter version. W. B . Saunders Company. RO WE , T. 1987 . Definition and diagnosis in the phylogenetic system. Systematic Zoology 36: 208-211. ROWE , T„ AND J. GAUTHIER . 1992. Ancestry, palaeontology, and the definition of the name Mammalia. Systematic Biolog y 41: 372 -378 . RUPKE, N. A. 1993. Richard Owen' s vertebrate archetype. Isis 84: 231-251 . RUSSE LL, E. S. 1916. Form and function : a contribution to the history of animal morphology. John Murray, London. SANDERSO N, M. J., AND M. J. DON OGH UE. 1989. Patterns of variation in the levels of homoplasy. Evolution 43: 1781 -179 5. SCHA NDER, C., AND M. THO LLES SON . 1995. Phylogenetic taxonomy - some comments. Zoologica Scripta 24: 263-268. SCHW ARTZ , S. P. 197 7. Introduction. Pages 13-41 in Naming, necessity and natural kinds (S. P. Schwartz, ed.). STEV ENS, P. F. 1984. Metaphors and typology in the development of botanical systematics 1690-1960, or the art of putting new wine in old bottles. T axon 33: 169- 211. STE VEN S, P. F. 1994. The developmen t of biological systematics. Antoine-La urent de Juss ieu, nature, and the natural system. Coulmbia University Press. SHASN Y - Wl JNHOFF, G . 1936. Die Polystilifera der Siboga-Exped ition. Siboga Expeditie 22: 1-214. STRICK LAND, H. E. 1846. On the structural relations of organized beings. Philos. Mag. J. Sei. 28: 354 -364. SUNDBERG , P. 1993. Phylogeny, natural groups and nemertean classification. Hyd robiologia 266: 103-113. SUNDBERG , P., AND F. PLEIJ EL. 1994. Phylogenetic classification and the definition of taxon names. Zoologica Scripta 23: 19-25 . - 2 4 - SUNDBERG, P. , J. MCCLINT OCK TUR BEV ILLE , AND M. HÄRL IN. 1996. Molecular evidence does not sup port Je nsen's theory on nemerteans as vertebrate ancestors. Hydrobiologia WILEY, E. O . 198 8. Parsimon y analysis and vicariance biogeography . Systematic Zoology 37 : 27 1- 29 0. WIL LME R, P. 199 0. Invertebrate relationships. Patterns in animal evolution. Cambridge University Press. WIT TGE NSTE IN, L. 195 3. Philosophical investigations. Macmillan, C hicago. ZIMM ERMA N, W. 193 1. Arb eitswei se der botanischen Phy logenetik und anderer Gruppierungswissenschaften. In Handbuch der biologischen Arbeitsmetoden, Abt. 3,2, Teil 9 (ed. E. A bderhalden). Urban & Schwarzenberg, Berlin . På grund av upphovsrättsliga skäl kan vissa ingående delarbeten ej publiceras här. För en fullständig lista av ingående delarbeten, se avhandlingens början. Due to copyright law limitations, certain papers may not be published here. For a complete list of papers, see the beginning of the dissertation. . • • SÂÇâ SSW œsfe^ SÂïS «®; IisifisiPsamtiit sims't ifs ~ • Tryckt & B unden Vasastadens Bok binderi AB 1996