Primate Taxonomy

John R Grehan. 21st Century Anthropology: A Reference Handbook. Editor: H James Birx. Volume 1. Thousand Oaks, CA: Sage Reference, 2010.

Primate taxonomy contributes to the science of anthropology by providing an evolutionary framework for the biological and cultural origins of human behavior. But taxonomy involves methods and principles that are not always readily accessible to the anthropologist who may depend upon the results of taxonomic research to interpret the evolution of human behavior. This juxtaposition of disciplines is particularly challenging for the anthropologist when taxonomic research either fails to provide a resolved classification for humans or other primate groups, or it requires the reevaluation and even falsification of long established units of classification. These taxonomic challenges are relevant to anthropological interpretation because the units of taxonomic classification are arranged to reflect hypotheses about evolutionary relationships. With these practical and conceptual challenges in mind, this review will examine the principal developments and issues of contention, as well as the main outlines of primate taxonomy at the beginning of the 21st century.


The modern system of classification (the rules for naming and arranging taxa) is derived from the binomial system of Linnaeus who provided a genus and species epithet unique to each species and placed them within a hierarchal arrangement of successfully inclusive groups. These units were proposed at a time when the similarities and differences among groups were seen to be the result of common design rather than shared descent. Following the publication of On the Origin of Species (Darwin, 1859), the widespread acceptance of evolution represented a major theoretical development where taxonomic units could now be seen to represent organisms sharing descent from a common ancestor. In practice, the process of classification was not radically altered as many approaches to recognizing groups or organisms were intuitive. Taxonomists often produced classifications based on characteristics that were thought to be particularly important for the evolution of a group or represented an amalgam of various characteristics to give some level of apparent coherence or stability in classification. Classifications also often emphasized overall similarity so that groups such as the reptiles and the great apes were each grouped together because they appeared to be more similar to each other than to other groups of species. As research into the theory and method of classification developed during the 20th century, there was an increasing awareness that similarity alone could not be assumed to indicate a close evolutionary relationship. The evolutionary challenge for taxonomists was to identify those similarities that provided evidence of common descent rather than those that may have evolved in different lineages or those that are misleading because they represent primitive retentions that have been lost in some members of the group, so the remaining species are grouped together because they share ancestral characters.

Taxonomy and Systematics

A direct correspondence between evolutionary relationship and taxonomic classification was made possible through the development of cladistic methodology. The first detailed outlines of this method were made by Daniele Rosa (1918), but it did not become popular until after the English translation of Hennig’s Phylogenetic Systematics (1966). Following a decade of sometimes contentious debate in the 1970s and early 1980s, cladistics became the standard methodology for reconstructing evolutionary relationships and identifying natural, or monophyletic, groups that include all descendants of a hypothesized common ancestor. Monophyletic groups are recognized by the shared presence of one or more unique features that are assumed to have been present in the common ancestor and therefore retained in all the descendants. As speciation occurs, each individual descendant may evolve further unique features that are in turn inherited by their descendants. This process results in a hierarchy of groups that are each characterized by uniquely shared features, orsynapomorphies. Synapomorphies for a particular group represent ancestral features for member taxa. For example, the presence of at least one flat nail on the foot represents a primate synapomorphy, but within the primates, the presence of a flat nail on any individual primate species provides no information on their evolutionary relationships with other primate species. Similarly, the presence of body hair exemplifies modern mammals as a natural group but is uninformative about the evolutionary relationships of particular mammal groups or species with each other.

Perhaps the most fundamental consequence of cladistic reasoning for primate taxonomy was the requirement that all natural groups be monophyletic. The criterion for monophyly required that all members had to be not only descended from a unique common ancestor but also that all descendants were included in the group. In some noncladistic classifications, there was an emphasis on separating out groups based on their sharing primitive features. This resulted in artificial, or paraphyletic, groups that failed to represent the actual pattern of evolutionary descent by leaving out taxa that had lost the primitive features due to the evolution of more derived features. Only by grouping organisms together according to their shared derived features is it possible to eliminate paraphyly. A classic example is the family Pongidae, which was originally designated to encompass all the great apes. But this group is paraphyletic because at least one great ape species is more closely related to humans than other great apes (see below under monophyly of hominoids).

Cladistic classification is always comparative. It is not sufficient, for example, to note that within the primates, monkeys share with hominoids (apes and humans) the presence of a fused jaw. This comparison only indicates that they are similar to each other in this respect. What is necessary for understanding this similarity as evidence for an evolutionary relationship is that these two groups are more similar in sharing the presence of a fused jaw than a third group, such as the prosimians (that do not have a fused jaw). To have some confidence that the unique similarity of the monkey and hominid jaws is inherited from a unique common ancestor not shared with prosimians (these three taxa together representing the in-group), it is necessary to demonstrate that the feature is absent from a fourth, external out-group. Theoretically, all other life-forms comprise the out-group, but in practice, the out-group is more circumscribed. For example, comparison of Primates with an out-group comprising other mammal orders shows the presence of a fused jawbone to be relatively rare outside the Primates, being found in a few groups such as wombats, mystacinid bats, and pigs. This occasional occurrence in the out-group suggests that within Primates the fused jaw bone represents a separate development supporting the hypothesis that the fused jaw shared by monkeys and hominoids is the result of their being more closely related to each other than prosimians and that together (as anthropoids) they represent a natural phylogenetic group.

If the distribution of synapomorphies was consistent, then there would be only one hierarchal set of characters and only one classification. In practice, the distribution of unique features may indicate more than one possible set of evolutionary relationships and arrangement of taxa. Species A may share one or more unique characters with Species B but also share one or more unique characters with Species C. In this situation, systematists will choose the larger set of characters as being more likely to represent the correct evolutionary relationship because it minimizes the number of times it is necessary to explain the independent origin of false similarities originating in lineages that have already separated. There are also problems created by the loss of features—the Madagascan primate Daubentonia lacks a toothcomb, a feature that otherwise defines all prosimians, for example. If other features are consistent with Daubentonia being a prosimian, then this anomaly is viewed as the loss of a toothcomb rather than its never having been present (which would require that Daubentonia not be a prosimian as currently defined).

When a small number of taxa and characters are being analyzed, it may be a relatively simple procedure to sort out the best supported set of relationships. With larger numbers of taxa and characters supporting different relationships, the sorting procedure becomes much more complicated and requires computer algorithms to sort through the different possibilities. Even then, it may not be possible to find a consistent answer, especially where the numbers of characters supporting different evolutionary relationships are identical or nearly so. One confounding problem in primate taxonomy is that some classifications that are characterized as being cladistic include analysis of features where some or all of the conditions of the in-group are also present in the out-group. This practice is not cladistic and may result in erroneous evolutionary relationships.

It is the nature of systematics that after years of comparative study, it may still be impossible to resolve the evolutionary relationships and classification of one or more taxa, and it is never possible to predict whether further study may not result in the discovery of other derived features that support an alternative to the accepted theory of relationship. It may not be possible for any one technique or only one aspect of an organism’s biology to consistently and accurately reflect evolutionary relationships. It is probably a truism that organisms will always show features that could support more than one theory of relationship. In this respect, some taxonomic categories may always be ambiguous.

Molecular Taxonomy

The application of molecular techniques to primate classification is widely characterized as a powerful tool for resolving the classification of primates, particularly where morphology has previously given uncertain or ambiguous results. Molecular evidence is also widely seen to represent a better alternative to morphology because molecular techniques allow rapid comparison of many taxa without being confronted with the necessity of engaging in extensive and prolonged comparative analysis of individual characters to identify homologies as is the case for morphology. Instead, there is the apparent simplicity of analyzing the presence or absence of just four base pairs that may replace each other in different species. In the absence of each character being individually justified, the molecular technique is widely seen to be reliable because thousands of bases can be compared with the intuitive assumption that the sampling of large numbers increases the likelihood of finding a correct relationship.

The evolutionary problem for primate taxonomy represented by morphological and molecular techniques is that they do not always produce the same result. A widely accepted principle of molecular systematics is that strongly supported molecular similarities are correct even when they conflict with morphologically based classifications—although this determination is not always consistently applied—sometimes, molecular results are rejected when they conflict with strongly accepted morphologically based classifications. Molecular similarity was originally validated as evidence for evolutionary relationships by its general correspondence with well-established morphologically based classifications. Morphological evidence was later treated as suspect and usually rejected when not supported by molecular similarity. Currently unresolved incongruities between molecular and morphological evidence include the monophyly of Madagascan lemurs, the relationships of tarsier and other primates, and the monophyly of great apes.

Although widely seen as a panacea for systematics and classification, molecular systematics lacks a comprehensive theory of homology (a principal of similarity that identifies a unique common ancestry). Evidence for molecular homology is limited to a set of four nucleotides and their relative position (sequence order). As in morphological data, shared similarity in DNA sequences can be due to primitive retention, reversal, or nonhomology (e.g., convergent). DNA sequences are further inherently ambiguous because substitutions leave no evidence of former replacement that would indicate whether matching base pairs represent primitive retention, convergence, or unique derivation. To match homologous base pairs shared between different primate species it is necessary to artificially make the genes the same length by creating changes to the relative positions by creating gaps between bases as inferring associated base substitutions. This alignment procedure requires mathematical models that generate a compromise fit to minimize gaps and substitutions. The resulting homology of each base is the product of the overall similarity created by the algorithms. In this procedure, there is no empirical evidence for the homology of individual bases and no way to know what bases represent primitive retentions or uniquely shared novelties, the latter being essential for cladistic analysis. Other potential problems include inadequate out-group sampling (often just a few species) in characters in the out-group being overlooked, the continued use of noncladistic techniques, such as distance measures that only measure overall similarity, and other methods such as maximum likelihood, which are based on theoretical models of what the correct phylogeny should be.

Taxonomic Categories

Taxonomic classification represents a list of named groups arranged in a hierarchical series of subordinating units based on the species, genus, family, order, and class categories along with various other intermediate or higher ranks. As monophyletic groups, these taxonomic labels may represent a natural classification for evolutionary relationships, but the number of groups, their scope, and their rank are entirely arbitrary. There is no necessary corresponding taxonomic significance for the same rank (e.g., family) compared between different lineages. For example, the family Hominidae may include only 1 (humans) or 6 species (humans and great apes) depending on how broadly the category may be applied. Similarily, the Hylobatidae (lesser apes) encompass a single genus and about 11 species. In contrast, the family Cercopithecidae (Old World monkeys) comprises at least 22 genera and 80 species. In some cases, restricted higher-level taxonomic categories are used to emphasize how particular primates are seen to be more different or more evolutionary significant than others. This has occurred with the use of Hominidae as a restricted family rank. Focusing on taxonomic ranks rather than evolutionary relationships can impede rather than advance the understanding of primate taxonomy.

The smallest evolutionary units of taxonomy represent a major area of research in primate taxonomy. Over the last century, there has been a continual expansion of the number of species with a concurrent debate over whether various subspecies represent geographic variants or distinct species in their own right. Naturally, the search for an objective and universal definition of a species has been a constant concern, not only for primate biologists but also for taxonomy in general. One criterion has been the absence of interbreeding, but this becomes problematic for taxa that are geographically isolated and do not inter-breed simply as a geographic constraint. Some systematists favor the identification of the smallest cluster of monophyletic organisms that can be separated by a unique combination of character states. This approach has been extended to the use of DNA sequences such that any unique DNA sequence shared by some individual organisms may constitute a distinct species. In practice, primate species are identified with respect to particular places and times, and their resulting taxonomic status may or may not be controversial.

Defining Primates

Primate Characters

The modern concept of Primates has its origins with Karl Linnaeus’s 1735 work Systema Naturae, where he proposed the group Anthropomorpha to include monkeys (Simia), sloths (Bradypus), and humans (Homo) based on their sharing the presence of a single pair of pectoral mammary glands. This classification was historically important for including humans within the taxonomy of other organisms for the first time. The name of the order was then changed to Primates in 1758 when Linnaeus also excluded sloths while also adding lemurs, bats (later removed by Johann Friedrick Blumenbach), and apes (the latter under Simia). A more extensive set of similarities was also identified to include the presence of four upper and four lower incisor teeth, parallel-sided lower anterior teeth, a pair of projecting caninelike teeth in the upper and lower jaws, limbs terminating in hands or structures that functioned like hands, arms that were separated by clavicles, locomotion mostly on all four limbs, tree climbing, and fruit consumption. This new definition accommodated lemurs while excluding sloths, and it applied to bats if they were broadly viewed as having grasping feet.

Subsequent classifications frequently separated humans by placing them in their own group such as Bimanes, or Bimana, while the remaining four-handed primates were classified as Quadrumanes, or Quadrumana. Many early taxonomists regarded the possession of two hands in humans as sufficient justification to separate them from the rest of the animal kingdom. In 1811, Karl Illiger separated Homo from all other animals by their erect posture, and he also applied the term Prosimii for lemurs and lorises while Tarsius was placed in the Macrotarsi to emphasize their development of highly elongated tarsal bones. Separation of humans from other animals largely persisted until 1883 when William Henry Flower included the genus Homo in a classification that persisted into the 20th century. He also divided the primates into the suborders Lemuroidea (lemurs, lorises, tarsiers) and Anthropoidea (monkeys, apes, humans). The Lemuroidea was further subdivided into the families Lemuridae (lemurs, lorises, bush babies), Tarsidae (tarsiers), and Chiromyidae (aye-aye). These primates came to be known as the “lower” or most primitive primates with lemurs being the most primitive of all. Tarsiers were initially grouped with galagos before being separated into their own group (Macrotarsi), leaving galgagos, lemurs, and lorises in the Prosimii. As a separate family, tarsiers were then again grouped with lemurs, lorises, bush babies (Lemuridae), and the aye-aye (Chiromyidae) in the suborder Lemuroidea until this name was replaced by the Prosimii. Tarsiers were then separated from the prosimians by Pocock in 1918, who grouped them with anthropoids in a new suborder called the Haplorrhini where the tarsiers were labeled the Tarsoidea and the anthropoids in the Pithecoidea (now Anthropoidea). The remaining prosimians were renamed the suborder Strepsirrhini. The haplorrhine classification was largely overlooked until the 1970s and 1980s when various features were widely seen to indicate a closer relationship between tarsiers and anthropoids than prosimians. Further support for the Haplorrhinii was also found in various molecular comparisons of DNA although other studies continued to support the inclusion of tarsiers within the Prosimii.

Flower grouped the New and Old World monkeys, and hominoids (apes and humans) together under the Anthropoidea. The New World monkeys were further divided into the families Hapalidae (marmosets), and Cebidae (larger prehensile-tailed monkeys). The Old World monkeys were placed in the family Cercopithecidae comprising with two subfamilies, Cercopithecinae and Colobinae. Apes were initially assigned to the family Simiidae while humans were represented by the family Hominidae. The emphasis on distinguishing primitive taxa reflected a widely held notion that it was possible to line up lower primates in a sequence of increasing complexity leading to “higher” primates from prosimians to anthropoids.

The definition of primates also continued to evolve. In an effort to further clarify the primates definition, St. George Mirvart proposed a broad set of features in 1873: presence of nailed digits, clavicles, orbits encircled by bone, three types of teeth during the life of an individual, a brain with a posterior lobe and calcarine fissure, innermost digits opposable for at least one pair of limbs, a thumb with or without a flat nail, a well-developed cecum, and pendulous penis, scrotal testes, and two pectoral mammary glands. The possession of a thumblike digit to one part of limbs accommodated the inclusion of humans. These characters are all taxonomically problematic because they are not specific to primates and therefore not cladistically informative for understanding the monophyly of primates. The three teeth types also occur in some marsupials and placentals, such as pigs and bears. Most mammals also have a well-developed cecum and scrotal testes also occur in carnivores such as cats. Two pectoral mammary glands are present in sloths, bats, and flying lemurs, and a posterior cerebral lobe with calcarine fissure also occurs in tree shrews. The clavicle is found in other mammals and even reptiles. Possession of an opposable thumb or big toe is not restricted to primates, but it is rare in other mammals. Of these features, only the development of at least a flattened nail on the big toe is distinctive for extant primates.

Primates are often characterized as having thumbs and big toes that are divergent, flattened nails at least on the big toe and a postorbital bar behind the eye. These features are individually present (although rare) in some other extant mammals, but only primates have all. But all living primates (tarsiers, lemurs, anthropoids) share at least six unique characteristics of the cheek (molar) teeth: (1) low rounded cusps (bunodont) with the sides of the upper molars being filled out rather than concave, (2) anterior outer cusp (paraconid) of the first lower molar positioned more toward the cheek (buccal) than the anterior of the tooth, (3) upper molars with a shallow angle (entoflexus) between the protocone and metacone, (4) trigonid (anterior half of the lower molar) not much taller than the (posterior) talonid, (5) talonid at least as long and broad as the trigonid, and (6) an oblique crest between the hypoconid and metaconid (cristid obliqua). Since these are all hard tissue features, and teeth are often the only fossil remains of extinct primates, they can be used to determine whether various fossils fall within Primates as characterized by its living members.

Recognizing Fossil Primates

The primate fossil record often presents a taxonomic challenge because in the absence of soft tissue features, all taxonomic assignments must rely on the evidence of hard tissue features. The scope of evidence is further reduced for many fossil taxa where skeletal material is incomplete and in some cases limited to a few bones or teeth. Some fossil taxa may remain chronically indeterminate until sufficiently complete samples are found. Integration of fossils within the taxonomy of living primates requires recognition of the same diagnostic features, but this is often lacking or is poorly documented for fossils, particularly those that are represented by only a few dental or other fragments. A recent example is the Anthrasimias gujaratensis, represented by three molars and one deciduous premolar. It was characterized as the earliest Asian fossil anthropoid recorded to date even though there were no diagnostic characters identified as being anthropoid or even primate. Efforts to classify fossils in relation to living taxa using parsimony or other computational methods may also be confounded by a prevalence of missing character states that contribute to poorly resolved phylogenies, even though the evolutionary relationships and taxonomic groupings may be unproblematic for the living taxa. Many fossil taxa remain, at this time, subject to future evaluation and the likelihood that the monophyly of some groups will be rejected and new taxonomic categories will be recognized. With this contingency in mind, the taxonomic arrangement of all fossil primate taxa must be regarded as subject to future revision.

A major question in primate classification has been the identification of its most primitive representatives in the fossil record that may not exhibit all features present in extant primates. This has been the situation for Plesiadapiformes, a group of fossils from the Paleocene, Eocene, and Oligocene of North America and Europe mostly represented by fragmentary jaws with teeth or isolated teeth. They conform to primates in their dental characteristics, but some skulls (such as for Plesiadapis) lack a postorbital bar. The typical primate’s grasping hands and feet and flattened nails are also missing in Plesiadapis, but Carpolestes had a grasping foot with flattened nail as well as a grasping hand. This range of variation suggests that the defining features for all modern primates may have emerged at different times with mammals first evolving into primates in their dental characteristics and later in their skull, hands, and feet. This pattern of similarity may be interpreted taxonomically in two different ways. Either the definition of primates is limited to dental features in which case Plesiadapiformes are included within the primates, or the definition of primates is limited to the inclusion of all features found in modern primates in which case Plesiadapiformes may be considered an extinct primate relative but not a member of Primates as defined by the living taxa. Either way, the Plesiadapiformes represent the closest fossil primate relative, but this relationship may be represented by two different taxonomic arrangements according to whether the order Primates is defined to include or exclude the Plesiadapiformes (after Schwartz, 1986):

The taxonomic status of other proposed primate fossils are more problematic. Eosimias and Bahinia have been identified as anthropoids even though they apparently lack dental features characteristic of living primates. Other fossils such as Wailkia and Siamopithecus are dentally primate, but their relationships within Primates are problematic because of inconsistencies between apparent similarities that would relate them either to anthropoids or prosimians.

Primate Classification

The living primates are represented by about 355 proposed or recognized species. The taxonomic composition of primates will continually be modified according to new interpretations of evidence about relationships, as well as new discoveries that alter the taxonomic position of known species or add new species and even genera. An increasing number of primatologists working in primate classification have rapidly increased the number of recognized species, although some may be poorly supported and possibly erroneous. There may also be differences over whether individual primate groups are recognized as species or subspecies, so the total number of primates will likely vary over time as well as between different classifications.

Primate classifications of living primates are represented by two alternative subordinal arrangements. The oldest classification divides Primates into the suborders Prosimii (lemurs, lorises, tarsiers) and Anthropoidea (monkeys, apes, humans). Prosimians were originally viewed as an assemblage of primitive primates in contrast to the more “advanced” simians (monkeys, apes, and humans). In this context, prosimians were a group by default—what was left over after exclusion from the anthropoids. Over the latter half of the 20th century, increasing interest has focused on an alternative division of the Primates into the suborders Strepsirrhini (prosimians without tarsiers) and Haplorrhini (tarsiers along with monkeys, apes, humans). Strepsirrhini refers to the presence of a nostril with a lateral slit or crease (strepsis means “twisting,” referring to the upward twist at the back end of the nostril slit), a primitive condition found in most mammals. Haplorrhini refers to a simple or unadorned (haplos) condition, although tarsiers and various New World monkeys are morphologically strepsirhine in having laterally creased nostrils. The contrasting theories of relationship generate two possible classifications depending on whether tarsiers are more closely related to lemurs and lorisiforms (Prosimii) or to the anthropoids (Haplorrhini). Table 61.1 illustrates alternatives with respect to the living Tarsiiformes represented by the single genus Tarsier with about seven species and Lorisiformes represented by the families Lorisidae (lorises of Africa, India, and Southeast Asia) and Galagonidae (galagos and bush babies of Africa). A third infraorder, Chiromyiformes, has also been proposed for the Strepsirrhini to represent the monotypic aye-aye, Daubentonia madagascarensis, which is otherwise usually placed within the Lemuriformes. The three main alternative classifications are illustrated below with example sources.

Tarsier Relationships

Tarsiers are unique among primates in having scales around the nipples and under the tail, a tail that is longer than the body, a head that can turn 180 degrees in either direction, and a bulging, immobile eye larger than the brain. They also have sinus hairs outside the nasal cavity, hairs on upper and lower lips, the longest tarsal bones, and three fully developed cusps on the anterior part of each lower molar and pointed cusps on all teeth anterior to molars, and tarsiers are the only specialist primate carnivore. A principal question in prosimian classification concerns the relative position of Tarsius, a primate genus that has been controversial from the beginning of modern primate classification when it was at first not even recognized as a primate (Linnaeus later classified it as the primate Simia syrichta in 1758).

The distinctiveness of tarsiers has often represented a taxonomic distraction with the differences emphasized by its frequent allocation to a group of high taxonomic rank and uncertainty about its evolutionary relationships with other primates. Tarsiers have been closely identified with both prosimians and anthropoids. Their inclusion within prosimians was emphasized in earlier classifications, but current morphological evidence is limited to relatively few—but potentially significant—uniquely shared features. Tarsiers along with other prosimians have a clawlike nail (called a grooming claw in Lemuriformes and Lorisiformes) on the second toe of the foot (tarsiers are unique in also having a claw on the third toe). Tarsiers are unlike other prosimians in lacking a set of slender, elongate incisors (four or two) and canines that are either horizontal or upwardly tilted. Collectively, these teeth are characterized as a “toothcomb” used in grooming and sometimes feeding. Tarsiers have only two small, vertical anterior teeth, and in this respect, they do not have a tooth-comb. One possibility is that tarsiers have lost the full set of anterior teeth found in other prosimians. This possibility is illustrated by the sifaka lemur where there are only four toothcomb teeth compared with six in other lemurs and lorsis. If tarsiers lost a further two teeth, then there would no longer be a recognizable toothcomb structure, and it has been suggested that the two anterior teeth of Tarsius show a similar transverse rounded shape with lateral borders to the outer anterior teeth of other prosimians.

In contrast to the few proposed uniquely shared similarities between tarsiers and prosimians, a relatively large number have been proposed for tarsiers and anthropoids. If all or most are correct, then they would provide less ambiguous evidence than molecular studies that have been divided between those that support a prosimian relationship and those that support an anthropoid relationship. Some of these features are problematic. Tarsier eyes lack a tapetum (a reflective layer of the retina) and in this respect are more like anthropoids than prosimians where a tapetum is said to be present, but several species of Eulemur and possibly Varecia variegata also lack a visible tapetum (i.e., lack light reflection), although it is not known whether this indicates the absence of a tapetal layer. The completely fused upper lip of tarsiers is often characterized as an anthropoid trait, but it does not apply to various New World monkeys that have a groove (philtrum) down the midline of the upper lip indicating that fusion is not complete. Tarsiers and anthropoids have haemochorial placentation, but the similarity may be superficial, resulting from different developmental paths. The apparently short, anthropoid-like face of tarsiers (that is even shorter than some anthropoids) may be an artifact of enlarged eyeballs extending over the snout, and the partial postorbital closure in tarsiers and New World anthropoids involves different bones. In these and various other characteristics the correct homology and morphology is in dispute while some others such as the inability to manufacture vitamin C in tarsiers and anthropoids may be valid. In view of the morphological and molecular disagreements, the taxonomic position of tarsiers currently remains beyond consensus.

Fossil Prosimians

The fossil record is replete with many prosimian fossils that range from relatively complete skulls that are sometimes associated with postcranial remains and the many more largely fragmentary jaw fragments and isolated teeth. The absence of soft tissues that may otherwise provide critical evidence and the disparity of fossil representation render the accurate placement of many fossil taxa highly problematic, and many may remain irresolvable unless more complete fossils are found in the future. Some fossil prosimians are considered to lie outside the modern families and have been classified as members of families that are no longer extant, including the Protoadapinae, Omomyidae, Microchoeridae, and Adapidae. Membership within these families is sometimes problematic and influenced by assumptions about the relative size, location, and geological age of relevant fossils.

Monophyly of Lemuriformes

The lemurs of Madagascar comprise the families Cheirogaleidae (dwarf/mouse lemurs), Daubentoniidae (the aye-aye), Megaladapidae (Lepilemuridae; sportive lemurs), Lemuridae (lemurs), and Indriidae (indri). These families have traditionally been regarded as a monophyletic group called the Lemuriformes while the remaining prosimians were represented by the Lorisiformes (galagos and lorises). This monophyly has been challenged in some studies that support placement of the Cheirogaleidae within Lorisiformes and isolating the Daubentoniidae as a separate prosimian infraorder. These different systematic relationships may be represented by three principal alternative taxonomic arrangements.

Taxonomy of the Anthropoidea

In anthropoids, olfaction is further diminished while vision becomes functionally dominant and the orbits are more consistently directed forward. Major defining features include postorbital closure of the postorbital bar; fusion of the lower jaws at birth or, as with the frontal bones, early in development; a single-chambered uterus (also in anteaters); hemochorial placentation characterized by concentration of blood vessels into at least one disklike structure; and development of the amniotic sac membranes by folding in and around the growing fetus in contrast to invagination as found in tarsiers. Molar cusps are relatively low in height, the number of premolars is reduced from four to three, the lower molars no long have an anterior cusp, and a large cusp is consistently present on the posterior upper molars. The following taxonomic arrangement within the Anthropoidea (treated here as a suborder) is stable to the superfamily level:

Suborder Anthropoidea (after Groves, 2001)
Infraorder Platyrrhini (New World monkeys)
Family Cebidae
Family Aotidae
Family Atelidae
Family Pithecidae
Infraorder Catarrhini (Old World monkeys, apes, humans)
Superfamily Cercopithecoidea
Family Cercopithecidae (Old World monkeys)
Superfamily Hominoidea (apes and humans)

The New World monkeys are characterized by a platyrrhine nose where the nostrils are oriented laterally and separated by a fleshy septum. Since this feature is also found in many other mammals, it does not constitute a character supporting platyrrhine monophyly. Monophyly of Platyrrhini is supported by several morphological features, including separation of the frontal bone from the sphenoid and the presence of an external auditory meatus. Monophyly of this group is also supported in molecular reconstructions. Old World monkeys and hominoids have a catarrhine nose with a narrow nasal septum separating nostrils that are closely adjacent and more forward directed. They also have only two premolars in the upper and lower jaws (compared with three in New World monkeys) and the possession of an ectotympanic tube forming the opening of the ear. Some Old World Oligocene fossils also have three premolars, and the Old World fossil monkey Aegyptopithecus lacks an ectotympanic tube like the New World monkeys. These fossil taxa may represent extinct taxa that do not lie within the Catarrhini but are nevertheless more closely related to the Catarrhini than Platyrrhini.

Fossil Anthropoidea

Most fossils characterized as anthropoids exhibit sufficient features to be recognized as primates with confidence. But the earliest claimed Asian anthropoid, Anthrasimias gujaratensis of India at 54.5 million years ago (mya), lacks any described anthropoid or even primate features, so uncertainties over the phylogenetic position and taxonomic identification of such fossils remain a persistent problem. Other fossils such as (Branisella) can be recognized as anthropoid but cannot be placed within the two extant infraorders, while other fossils can be placed within one or the other but only as extinct families or superfamilies. Examples include Parapithecidae within the Platyrrhini and Pliopithecoidea (east Asia, Europe) within the Catarrhini. A variety of extinct families has been proposed for the Hominoidea including Dendropithecidae, Dryopithecidae, Proconsulidae, Ramapithecidae, and Victoriapithecidae (Africa) with various levels of general acceptance.

Monophyly of Hominoids

Hominoids comprise humans, great apes, and lesser apes. Their monophyly appears to be well supported and the category is taxonomically stable for living taxa. Hominoids are most well known for the absence of an external tail in contrast to all other primate groups. Hominoids also exhibit major alternations of body shape including a thorax that is broader than deep, a dorsal position of the scapula, an elongate clavicle, a broad separation between the infra-orbital foramen and a suture between the zygomatic and maxillary bones, a variable articulation of the ulna and triquetral, and a postnatal ossification in the distal humeral and proximal radial epiphyses. The lesser apes represent a well-established and stable taxonomic category represented by a single family Hylobatidae (gibbons and siamangs) comprising about 15 species in Southeast Asia. The number of recognized genera varies from one (Hylobates) to four (Hylobates, Hoolock, Nomascus, Symphangulus) due to different perspectives over their overall divergence or phylogenetic age.

The taxonomic arrangement of the remaining hominoids is far less settled with respect to the relationship between humans and great apes. Great apes are represented by three genera, Pongo (one or two species of orangutan in Southeast Asia), Gorilla (two or three species of gorilla in Africa), and Pan (two species in Africa). From the early to mid-20th century, African apes were increasingly seen to be more closely related to humans than orangutans. Darwin had also expressed this view, although this was a speculation based on the presumption that Africa rather than Asia would provide the necessary selection pressures for the evolution of humans. Chimpanzees and gorillas are indeed more similar to humans than the long-armed orangutan, which also exhibits distinctive and unique features, such as cheek pads in males and large vocal sacs and having a much more specialized arboreal existence. In these respects, orangutans are widely seen to be much more distantly related to humans than African apes.

Morphologically, the African apes would seem to represent a natural group since chimpanzees and gorillas share a suite of uniquely shared features, including their specialized knuckle-walking anatomy. But in the 1960s, another line of phylogenetic evidence has led to the now almost universal view that chimpanzees are more closely related to humans than gorillas. This evidence was principally in the form of shared similarities in biological molecules, beginning with protein and amino acid similarities supporting an African ape relationship and later DNA sequence similarities that are greatest between chimpanzees and humans. The chimpanzee relationship was at first rejected by primate morphologists but has since become so firmly accepted that it has become a fact of evolution that is beyond question. The only anomaly with this molecular theory is the lack of morphological features uniquely shared between humans and chimpanzees. A few such features have been proposed, but they have proven to be erroneous or lack verification. This lack of uniquely shared morphological similarity would not necessarily represent a problem as it could be attributed to a paucity of such features in the last common ancestor of humans and chimpanzees or a loss of those features subsequent to their respective divergence. This possibility is, however, confounded by the fact that humans do share a large and broad set of unique or specialized features with the orangutan.

The orangutan relationship was first identified by primate systematist Jeffrey Schwartz (2005) who found at least 40 features that were either uniquely shared or nearly so between humans and orangutans including several dental features, such as cusp configuration and thick molar enamel (African apes are like most other primates and mammals in having thin enamel), a single incisive foramen, a foramen lacerum, an anterior-posteriorly short scapula with a vertical vertebral boarder and a relatively small supraspinous fossa, the widest spaced mammary glands, highest estriol production, absence of tumescence of female genitalia during ovulation, greatest degree of cerebral and Sylvian sulcus asymmetry, beard and mustache in the male, forwardly directed cranial hair, receded hairline at birth, and absence of ischial callosities. For any other group, this preponderance of morphological evidence would be widely seen as critical if not conclusive. But this has been precluded by the majority of primate and evolutionary biologists because it conflicts with molecular evidence that has been already declared infallible. The morphological pattern may, however, suggest that in this case at least the pattern of molecular similarity is misleading, perhaps as the result of pervasive and unrecognized primitive retentions in the distribution of base pair sequences.

Molecular grouping of humans and chimpanzees is also problematic because it is incongruent with the hominid fossil record (hominid here referring to any fossil relatives more closely related to humans than the nearest living great ape). The earliest accepted hominids (Australopithecus) fail to exhibit uniquely chimpanzee features and instead show the same skeletal features as humans and orangutans. In addition, they show typically orangutan features, such as large, vertically inclined cheekbones with anterior facing roots, and a posteriorly thickened posterior palate. Analysis of these and other morphological features suggests that australopiths along with humans and orangutans comprise a monophyletic group. This monophyletic group would include the hominids Orrorin and Kenyanthropus along with the extinct great ape genera Hispanopithecus, Ouranopithecus, Ankarapithecus, Sivapithecus, Gigantopithecus, Lufengpithecus, and Koratpithecus, while excluding the purported hominids Ardipithecus and Sahelanthropus. The taxonomic implications are listed in Table 61.3, although fossil taxa cannot be included in the molecular taxonomy since fossils cannot (with the exception of subfossils) provide evidence of molecular similarity.

Humans and great apes together comprise the large bodied hominoids. Molecular reconstructions have popularized extending the family Hominidae to include this group, including their fossil relatives. In addition to those fossils listed below, the large bodied hominoids include the fossil genera Afropithecus, Ardipithecus, Dryopithecus, Nacholapithecus, Oreopithecus, Otavipithecus, Periopithecus, Kenyapithecus, Sahalenthropus, and Samburupithecus.

Future Directions: Primate Taxonomy in the 21st Century

It is impossible to accurately predict the course of primate taxonomy and systematics further into the 21st century, but at the beginning of this century, several outstanding problems represent actual or potential challenges for the future.

Diagnostic Characters

Even though cladistic principles are now generally accepted as the current standard of systematic analysis and taxonomic classification, the specification of the unique features that place a fossil within a given taxon are often unclear. Fossil descriptions and analyses are frequently muddled without a focus on primitive states that obscures the presence of derived conditions that would link the fossil with other living or fossil taxa. This is particularly evident where new taxa may be distinguished (particularly in the fossil record) without reference to any uniquely derived features to define the taxon or without reference to synapomorphies that would place it within a higher taxon. A classic example is Homo floresiensis, which was first described as a member of the genus Homo related to Homo erectus without citation of supporting evidence. It was later stated that the generic name was selected as a rhetorical device to prevent the fossil falling into obscurity, and the relationship with H. erectus appears to have been determined by default since this was the only other known hominid in the same region at the same time. Subsequent morphological studies have reinforced the status of this fossil as a distinct species (as some argued it was the result of microcephally in a primitive human) while at the same time highlighting features that relate H. floresiensis more closely with australopiths or earlyHomo. Publications on new hominid or presumed hominid fossils are particularly prone to this kind of problem when there is a lack of subsequent detailed studies to follow the initial rapid and superficial announcements (although the study of H. floresiensis is proving to be a notable exception).

Holotype Access

The scientific foundation of taxonomy biological systematics is the ability to provide empirical verification. This is the essence of testing in science. Without verification, scientific propositions lose their scientific status and instead become metaphysical abstractions supported entirely by faith. In the study of comparative biology, this pitfall can be avoided only through access to the original material on which theories about identity and relationships are made. In taxonomy and systematics, this is made possible through access to the type material, the specimens to which taxonomic names are attached and from which character similarities are generated. It is the holotype in particular that provides the critical verification as it is the reference specimen for each species identity and consequently for relating all additional specimens. For each species, there is only one holotype, and it is essential that there is open access as recognized in the International Rules for Zoological Nomenclature. This requirement by the systematics community is, however, all too often thwarted through the deliberate withholding of access to individual researchers. This practice represents a deviation from the principles of science, and for one primate group in particular—the hominids—it has become a pervasive and persistent problem that continues into the 21st century. Another problem in primate systematics is the use of pseudoholotypes in the form of photographs that cannot be subject to further testing or investigation. This practice reduces primate taxonomy and systematics to the level of stamp collecting as it removes access to the original material and therefore fails the scientific requirement of empirical corroboration.

Molecular Incongruence

This is probably the elephant in the room when it comes to primate taxonomy. There has been an increased focus on molecular similarity as the basis not only for classification but also for delineating the smallest taxonomic units, such as species. Some approaches attempt “total analysis” that treats morphological and molecular similarity as the same kind of data, but this overlooks the inherent problems of homology and similarity that have yet to be adequately addressed in molecular analysis. Emphasis on the same result from different genes or from large data sets may seem to further corroborate the molecular result, but if the molecular comparisons are not actually between homologous derived character states, then the large numbers of similarities may reflect a similar prevalence of primitive retentions in molecular similarity as is often found to be the case in demonstrations of overall morphological similarity. The alternative possibility is that robust morphological relationships may represent a viable falsifier of molecular similarity, particularly when the morphological evidence among living taxa is consistent with hypothesized phylogenetic relationships between fossil and living taxa. One possible resolution of incongruence between morphological and molecular similarity may come when the connection between uniquely shared morphological traits and their molecular position within DNA is understood. The current inclination for primatologists to focus on molecular similarity to reconstruct primate relationships and delineate taxonomic categories at the expense of morphological incongruence may represent a critical future problem for primate taxonomy and systematics.

Taxonomic Labels

From Linnaeus to the present, primate taxonomy is represented by a hierarchical system of names. At the close of the 20th century, some systematists have argued for a system of names without rank called the phylocode. Each node on acladogram could be named, but they would have no corresponding rank. There would also be no species and no binomials. The purpose of the phylocode was to introduce stability into taxonomy—given the constant flux in phylogenetic relationships that currently lead to new taxonomic arrangements. This goal may be problematic since the taxonomic meaning of any one node is still contingent upon its relative position on a phylogenetic tree, and if current Linnean names are incorporated, there is no necessary way to determine how their meaning may have changed. The incorporation of additional taxa within a given clade would also change the meaning of that node, requiring that the definition of the name would have to be changed, or the name would have to be redefined in reference to an entirely different node.

The actual and potential problems for the phylocode system appear, at this time, to present no necessary improvement over the problems confronting Linnean taxonomy against a background of uncertain phylogenetic reconstructions. In systematics, all relationships are relative when the primary question is whether any two taxa are more closely related to each other than either is to a third. In this phylogenetic context, taxonomic categories and classifications are irrelevant and unnecessary other than as convenient labels to designate particular groups of organisms. Whether the 21st century is witness to some other entirely novel solution remains to be seen by those fortunate to be around so long.

Conclusion: The Future Nature of Evidence

It is anticipated that the major contentious questions over evolutionary relationships among the primates will continue to dominate taxonomy and systematics into the 21st century, particularly where different sources of evidence continue to give contradictory or ambivalent results. In addition to those questions of classification within the primates, there will no doubt be further exploration of the interrelationships of modern primates to other mammalian groups.

The future status of morphological systematics represents a major future question. Morphology is problematic when different researchers use different characters, and this identifies a future need for comprehensive and illustrated comparative documentation of features used to define clades (and in turn the taxonomic groups linked to those nodes). Detailed comparative morphology is probably better known for some obscure insect groups than many primates, especially not (and perhaps most surprising) for the great apes that have been of so much interest for theories of human origin. In some major respects, primate taxonomy and systematics during the last half of the 20th century exemplifies the severe decline in the science of comparative morphology as the inverse mirror image of the rapid expansion of molecular biology. This decline is often evident in the lack of knowledge on comparative primate anatomy and biology, often including some of the most basic elements of biology, such as reproduction. The foundering of comparative morphology may be inconsequential if the promotion of molecular similarity is justified as the final authority on evolutionary relationships. But this status has yet to be seriously evaluated, and it may yet be possible that the 21st century will take a new look at the role and importance of morphology in the reconstruction of primate relationships and classification.