Elsevier

Journal of Human Evolution

Volume 44, Issue 2, February 2003, Pages 225-254
Journal of Human Evolution

Morphological affinities of the Australopithecus afarensis hand on the basis of manual proportions and relative thumb length

https://doi.org/10.1016/S0047-2484(02)00207-5Get rights and content

Abstract

The hands of apes and humans differ considerably with regard to proportions between several bones. Of critical significance is the long thumb relative to other fingers, which is the basis for human-like pad-to-pad precision grip capability, and has been considered by some as evidence of tool-making. The nature and timing of the evolutionary transition from ape-like to human-like manual proportions, however, have remained unclear as a result of the lack of appropriate fossil material. In this article, the manual proportions of Australopithecus afarensis from locality AL 333/333w (Hadar, Ethiopia) are investigated by means of bivariate and multivariate morphometric analyses, in order to test the hypothesis that human-like proportions, including an enhanced thumb/hand relationship, originally evolved as an adaptation to stone tool-making. Although some evidence for human-like manual proportions had been previously proposed for this taxon, conclusive evidence was lacking. Our results indicate that A. afarensis possessed overall manual proportions, including an increased thumb/hand relationship that, contrary to previous reports, is fully human and would have permitted pad-to-pad human-like precision grip capability. We show that these human-like proportions in A. afarensis mainly result from hand shortening, as in modern humans, and that these conclusions are robust enough as to be non-dependent on whether the bones belong to a single individual or not. Since A. afarensis predates the appearance of stone tools in the archeological record, the above-mentioned conclusions permit a confident refutation of the null hypothesis that human-like manual proportions are an adaptation to stone tool-making, and thus alternative explanations must be therefore sought. One hypothesis would consider manipulative behaviors (including tool-use and/or non-lithic tool-making) in early hominines exceeding those reported among extant non-human primates. Alternatively, on the basis of the many adaptations to committed bipedalism in A. afarensis, we propose the hypothesis that once arboreal behaviors became adaptively insignificant and forelimb-dominated locomotor selection pressures were relaxed with the adoption of terrestrial bipedalism, human-like manual proportions could have merely evolvedas a result of the complex manipulation selection pressures already present in extant non-human primates.Both hypotheses are not mutually exclusive, and even other factors such as pleiotropy cannot be currently discarded.

Introduction

Manual proportions, and especially relative thumb length, are among the features most readily distinguishing the hands of humans from those of other Hominidae. Humans display short handsin comparison with those of great apes, with a particularly long thumb in relative terms (Schultz, 1930, Ashley-Montagu, 1931, Midlo, 1934, Jouffroy et al., 1991, Watkins et al., 1993, Christel, 1993, Susman, 1994). These proportions of the human hand are essential for human manipulative skills related with precision grip capability(Napier, 1960, Marzke, 1971, Marzke, 1997), being essential for bringing together the tips of all the five digits at the axis of the middle finger, and permitting a pad-to-pad (=pulp-to-pulp) contact between the thumb and the remaining digits (Napier, 1960, Napier, 1962a, Napier, 1962b). On the contrary, these human-like precision grip postures cannot be assumed by the great apes, due to the relative shortness of their thumbs (Napier, 1960), especially compared with theotherwise elongated fingers. (Although tip-to-tip and tip-to-side precision grasping occur in great apes, pad-to-pad precision grip is exclusive ofhumans, where it is the most frequently used precision grip posture: Christel, 1993.) The long hands of great apes must be rather interpreted as an arboreal adaptation for functioning as a grasping hook during suspension and/or climbing on large-diameter branches (Susman, 1979, Inouye, 1992). Given the fact that the pattern of long hands with a relatively short thumb is shared by orangutans and African apes, which are successive sister taxa for humans, it is more parsimonious to assume—until fossil evidence demonstrates the contrary—that human manual proportions are apomorphic rather than primitive for the hominine clade. The nature and timing of the evolutionary transition from ape-like to human-like manual proportions, however, are still unclear. Deciphering this has been hampered by the lack of well-founded evidence of manual proportions in Plio-Pleistocene hominines, in its turn resulting from the scarcity of associated fossil hand bones. As a result, a secure assessment of the selection pressures originally involved in the emergence of modern human manual proportions has been thus far precluded.

The involvement of tool-related manipulative selection pressures in the evolution of human manual proportions deserves careful consideration and a precise definition of the terms involved. ‘Tool-behavior’ can be employed to refer simultaneously to both tool-use and tool-making on any type of raw material (Beck, 1980; see also Marzke, 1997, p. 91, footnote 1). ‘Tool-use’, in turn, canbe defined as “the external deployment of an unattached environmental object to alter more efficiently the form, position, or condition ofanother object”, whereas ‘tool-making’ can be defined as “any modification of an object by the user or conspecific so that the object serves more effectively as a tool” (Beck, 1980, pp. 10 and 11). Tool-use has been documented in a wide spectrum of animal species (Alcock, 1972, Beck, 1980),including primates such as baboons (Beck,1974), capuchin monkeys (mostly in captivity) (Visalberghi, 1987), and apes both in captivity and in the wild (see review in McGrew, 1993). It seems, however, that important cognitive differences exist with tool-using between monkeys and apes; as a result, imitational learning by observation is prevented in monkeys (Futuyma, 1986, Whiten, 1989), which would explain why novel tool-related behaviors in monkeys are not culturally disseminated (Beck, 1974, Visalberghi, 1993). Amongst apes, only chimpanzees show frequent and diverse tool-use in wild environments uninfluenced by humans (McGrew, 1993). This happens in several contexts, including aggression, feeding and hygiene (Goodall, 1970, Goodall, 1986, Gould, 1975, Nishida, 1973, McGrew, 1974, McGrew, 1992, McGrew, 1993 and references therein; Teleki, 1974, Sugiyama and Koman, 1979, McGrew et al., 1979, Boesch and Boesch, 1981, Boesch and Boesch, 1984, Boesch and Boesch, 1990, Nishida and Hiraiwa, 1982, Kortlandt, 1986), even associated with cultural variation (Sabater Pi, 1984, McGrew, 1992, Whiten et al., 1999). Chimpanzee tool-behavior not only involves tool-use of natural objects, but also in many instances modification of them(e.g. Teleki, 1974, Boesch and Boesch, 1990), i.e. tool-making. In the wild, however, chimpanzee tool-making behavior is limited to organic raw materials, whereas tool making with lithic raw materials (hereafter termed stone tool-making) has only been documented for humans. Moreover, unlike humans, chimpanzees manufacture all their tools with their teeth and hands, and do not use secondary tools (i.e. tools to make tools), at least in the wild (McGrew, 1993). From the viewpoint of comparative ethology, the last common ancestor of African apes and humans mighthave been either an accomplished tool-user like chimpanzees, a non-tool-user like gorillas, or something intermediate (McGrew, 1989, McGrew, 1993). As pointed out by Susman (1995, p. 589), “we might suspect (although there is no evidence of either tools or anatomy) that the earliest hominids used tools in an apelike fashion”, and it has even been suggested that they engaged in tool-behaviors more sophisticated than those practiced by extant apes (Marzke, 1997). As stressed by Susman (1995), however, “without tools and without any anatomical traces, we can only speculate about tool-behavior at this early point in the human career”. The recent finding that chimpanzee tool-using behavior can leave evidence in the archeological record (Mercader et al., 2002) is very promising for the future discovery of such tool-using activities, but since only lithic industries have been thus far documented, we are restricted to stone tool-making as the hypothesis to be tested (see subsequently).

The morphological features that permit human-like precision grasping have been equated with tool-behavior and “the enhanced ability tomanipulate tools” by Susman, 1998, Susman, 1994, respectively). However, as noted by Marzke (1997, p. 91, footnote 1), in some instances the term tool-behavior has been employed to refer to tool-using and/or non-human tool-making, whereas in other instances (Susman, 1991) it has been employed to refer exclusively to human tool-making. Susman attributes the equation between human-like precision grasping features and tool-making to Napier (e.g. Susman, 1988, Susman, 1994, Susman, 1998), considering them as a ‘toolmaking criterion’ (Susman, 1994, p. 1572). However, if human-like precision grasping features are diagnostic of tool-making, as Susman argues, then it follows that these features, including human-like manual proportions, must have originally evolved as an adaptation for tool-making. Although thisis conceivable, current function must not benecessarily equated with evolutionary origin (Gould and Vrba, 1982), so alternative hypotheses are also possible. In this article, we test the hypothesis that the selection of human-like manual proportions, including an enhanced thumb/hand relationship approximating a human-like precision grasping, was originally linked with manipulative selection pressures driven by stone tool-making selection pressures. To do so, we rely on the manual remains of Australopithecus afarensis, computing its manual proportions and comparing them with those of extant apes and humans by means of univariate and multivariate morphometric techniques. Unlike later australopith species (both Australopithecus spp. and Paranthropus spp.), A. afarensis precedes by about 1 Myr or so the appearance of stone tools and cut-marks in the archeological record at ca. 2.5 Ma of age (Semaw et al., 1997, Delson et al., 2000), thus providing a meaningful test for the mentioned hypothesis. In order to be able to accept hypotheses of adaptation in the fossil record, “it is minimally necessary … that historical concordance be true for the characters under consideration” (Lauder, 1991, Hartwig and Doneski, 1998). Therefore, our null hypothesis would be rejected if it could be demonstrated that A. afarensis already possessed human-like manual proportions, thus considerably predating the appearance of stone tool-making in the record. As we have already mentioned, other tool-related behaviors (tool-using and non-lithic tool-making) cannot be incorporated into the null hypothesis because it is not currently possible to discern between real and artifactual absence of traces of these behaviors in the record. Of course, many other qualitative and quantitative morphological characters, including robusticity and morphology of articular surfaces, are important in relation to human manipulative skills, but their investigation is outside the scope of this paper and will be discussed in detail elsewhere (Moyà-Solàet al., in preparation; Alba et al., in preparation).

The issue raised in this paper has remained unresolved until now because of the controversy surrounding the morphological affinities of the hands of early hominines. The most complete specimen was until recently O.H. 7 from Olduvai Gorge (Napier, 1962b, Leakey and Hay, 1979, Susman and Creel, 1979), which lacks the first metacarpal and proximal phalanx, and in general is tooincomplete to provide a meaningful comparison of manual or thumb proportions. Accordingly, previous assertions on this subject have been restricted to A. afarensis from Hadar. In spite of the abundant manual remains of this species from locality AL 333/333w (Bush, 1980, Bush et al., 1982), two different views can be found in the literature with regard to the morphological affinities of the A. afarensis hand. According to Stern and Susman (1983), the hand of A. afarensis would be essentially ape-like, i.e. “a suspensory adapted hand, surprisingly similar to hands found in the small end of the pygmy chimpanzee–common chimpanzee range” (Stern and Susman, 1983, p. 284). For Susman, this would be true not only in a locomotor context, but also in a manipulative one, asserting that “A. afarensis has a chimpanzee-like thumb” (Susman, 1995, p. 589) on the basis of thumb robusticity (metacarpal I head breadth/length). On the contrary, according to Marzke, 1983, Marzke, 1986, Marzke, 1997, the hand of A. afarensis would already display some human-like characters (i.e. features more similar to modern humans than to living apes). Proportions between several hand bones have not played a very prominent role in this discussion, however, probably due to the problems associated with their quantification, since more than one single bone is required. For example, Stern and Susman's (1983) account of the A. afarensis hand as essentially ape-like made no explicit mention of manual proportions. Later on, Susman (1994) relied on thumb robusticity as a ‘toolmaking criterion’ (p. 1571) additional to the “morphological criteria for toolmaking proposed by Napier in the early 1960s, including, among others, a fully opposable thumb with increased relative thumb length…” (p. 1572). And he did so in spite of recognizing that “The length of the thumb relative to the fingers is the most appropriate comparison” (Susman, 1995, p. 589), because the latter criterion is “limited to individual hominid specimens that are represented by associated hand bones (necessary to calculate thumb and finger proportions)” (Susman, 1994, p. 1572). Marzke, 1983, Marzke, 1997), who nevertheless investigated this issue, concluded that the thumb of A. afarensis “was longer in proportion to the fingers than in pongids” (1983, p. 197). This confirmed the earlier report of Bush (1980, p. 210), according to which relative thumb length in this taxon would be “relatively short compared to H. sapiens but relatively long compared to modern pongids”. The latter two authors therefore suggested derived thumb/hand proportions for A. afarensis, but still intermediate between extant apes and modern humans. This is shown by the fact that Marzke (1997, p. 105) considers that in A. afarensis, “The thumb probably was not long enough to facilitate pad-to-pad precision pinch and handling of small tools by the thumb and distal finger pads”. In any case, these somewhat human-like proportions in the hand of A. afarensis would at least partially result from finger shortening, as shown by Latimer (1991). The latter concluded, on the basis of third proximal phalanx relative length, that the finger bones in Hadar hominids “are shorter than any other hominoid (except Homo) indicating that they had undergone directional selection for length reduction” (Latimer, 1991, p. 169). Trinkaus (1992, p. 347), finally, also concluded that the hands of A.afarensis and other australopiths “lack either thumb foreshortening or elongation of the other fingers”, with “relative finger lengths … similar to those of modern humans”, although he did not provide any indication on what analyses hiscommentaries were based.

With regard to manual proportions, all the above-mentioned analyses must remain tentative because the manual elements from locality AL 333/333w include a mixed array of isolated bones from several individuals. Consequently, it would be possible to argue that the reported, somewhat human-like proportions are merely artifactual, i.e. that they result from computing proportions from bones belonging to different individuals. After a close examination of good-quality casts of hand bones from locality AL 333/333w, we have come to the conclusion that many of these bones probably belong to the same individual (see discussion in Section 2), but since the bones were not foundin anatomical connection, doubts remain. Susman (1998), in particular, has recently criticized the use of relative thumb length in A. afarensis on the basis that the lack of a firm association between the recovered hand bones makes this measurement unreliable. A rigorous quantitative assessment of manual proportions in this taxon, taking thiscaveat into account, is therefore required. In fact, Watkins et al. (1996) already concluded that the bones of digital rays I–IV (but not V) of the A. afarensis composite hand set could be modeled as belonging to a single individual, but unfortunately no further details have been provided since their initial preliminary report. In this paper, we show by means of randomization analyses that firm conclusions can be reached on the manual proportions of this taxon even if the bones do not belong to the same individual. This is very promising, since until more evidence is available for more ancient species such as Ardipithecus (White et al., 1994) and Orrorin (Senut et al., 2001), or from the nearly complete hand (and skeleton) of A. africanus from Sterkfontein (Clarke, 1999), A. afarensis remains central for understanding the origin of human-like manual proportions. We will demonstrate that this confirms the previous assessment of Watkins et al., 1993, Watkins et al., 1996, according to which, in comparison to body mass, the thumb of A. afarensis would be nearly as elongated as in modern humans (in comparison to African apes), whereas the length of the remaining manual rays would be reduced to the same degree. Moreover, we further refine and extend these conclusions, by showing that: (1) besides relative thumb length, A. afarensis has overall human-like manual proportions (an aspect previously not directly addressed by otherauthors), which contradicts the previous interpretation of it having an essentially ape-like, suspensory-adapted hand (e.g. Stern and Susman, 1983); (2) the increased thumb/hand relationship of A. afarensis is not only derived towards the human condition (thus confirming Bush, 1980 and Marzke, 1983), but also that, as in living humans, this mainly results from hand shortening (as first shown by Latimer, 1991) and only to a lesser extent (if at all) from thumb lengthening (thus confirming Watkins et al., 1993, 1996); (3) the relative length of the A. afarensis thumb is not intermediate between modern humans and apes (as previously asserted by Bush, 1980 and Marzke, 1983), but almost fully human, being thusprobably capable of making pad-to-pad precision grasping (contra Marzke, 1997); and (4) these conclusions are valid even if the bones do not belong to a single individual. This allows us to confidently refute the null hypothesis that human-like manual proportions originally evolved as an adaptation to stone tool-making selection pressures. Several possible alternative hypotheses, taking into account both manipulative and locomotor selection pressures, are therefore discussed.

Section snippets

The fossil sample

A reconstruction of the A. afarensis hand when compared with that of living great apes andhumans, based on the assumption that many of the manual remains from locality AL 333/333w belong to the same individual, has been provided inFig. 1;comptd;;center;stack;;;;;6;;;;;width> . This reconstruction includes the following bones: AL 333w-39, AL 333w-48, AL 333w-16,AL 333-56 and AL 333w-89 (metacarpals I–V);AL 333-69, AL 333-93, AL 333-63, AL 333x-19 and AL 333-62 (proximal phalanges I–V) andAL

Overall manual proportions

The results of multivariate canonical (discriminant) analyses have been reported in Fig. 2, Fig. 3;comptd;;center;stack;;;;;6;;;;;width> ;comptd;;center;stack;;;;;6;;;;;width> and Table 2, Table 3, Table 4, Table 5(see also the allometric regression equations derived to compute the logarithmicresiduals in Table 6). All the different types of variables employed performed approximately equally well for discriminating among livingspecies (Table 5), with more than 90% of correctly classified cases

Discussion

Our results confirm the previous assertions that the thumb/hand proportions in A. afarensis are higher than in chimpanzees (Bush, 1980, Marzke, 1983, Marzke, 1997), and also indicate that this proportion is not intermediate between apes and humans, but much more closer (if not equal) to the modern human condition, irrespective of whether ratios or residuals are employed to quantify it. This indicates, contra Marzke (1997), that the thumb ofA. afarensis was probably long enough to permit

Summary and conclusions

In this article, we test the hypothesis that human-like manual proportions, including theenhanced thumb/hand relationship that permits pad-to-pad precision grasping in modern humans, originally evolved as an adaptation to stone tool-making. This hypothesis is tested by investigating the proportions of the A. afarensis composite hand from locality AL 333/333w (Hadar, Ethiopia) by means of bivariate and multivariate morphometric analyses. We show that human-like manual proportions can be already

Acknowledgements

We are indebted to the following curators or investigators for access to skeletal material from living apes and humans (W. van Neer, B. Engesser, A. Malgosa), to T. White and O. Lovejoy forthe loan of casts of the A. afarensis hand, and toE. Sarmiento for kindly making available to us his morphometric database on hominoid hands and for reading a previous draft of this paper. This work has been supported by the Generalitat de Catalunya (Grup de Recerca de Qualitat 1999SGR00018, and also

References (129)

  • W.C. McGrew

    Tool use by wild chimpanzees in feeding upon driver ants

    J. Hum. Evol.

    (1974)
  • H.M. McHenry

    The first bipeds: a comparison of theA. afarensis and A. africanus postcranium and implications for the evolution of bipedalism

    J. Hum. Evol.

    (1986)
  • T. Nishida

    The ant-gathering behavior by use of tools among wild chimpanzees of the Mahale Mountains

    J. Hum. Evol.

    (1973)
  • T. Nishida et al.

    Natural history of a tool-using behavior by wild chimpanzees in feeding upon wood-boring ants

    J. Hum. Evol.

    (1982)
  • G.C. Packard et al.

    The use of percentages and size-specific indices to normalize physiological data for variation in body size: wasted time, wasted effort?

    Comp. Biochem. Physiol. A

    (1999)
  • B. Senut et al.

    First hominid from the Miocene (Lukeino Formation, Kenya)

    C.R. Acad. Sci. Paris

    (2001)
  • G.H. Albretch et al.

    Ratios as a size adjustment in morphometrics

    Am. J. Phys. Anthropol.

    (1993)
  • G.H. Albretch et al.

    Ratio adjustments in morphometrics: a reply to Dr. Corruccini

    Am. J. Phys. Anthropol.

    (1995)
  • J. Alcock

    The evolution of the use of tools by feeding animals

    Evolution

    (1972)
  • M.F. Ashley-Montagu

    On the primate thumb

    Am. J. Phys. Anthropol.

    (1931)
  • B.B. Beck

    Animal Tool Behavior

    (1980)
  • C. Boesch et al.

    Possible causes of sex differences in the use of natural hammers by wild chimpanzees

    J. Hum. Evol.

    (1984)
  • C. Boesch et al.

    Tool use and tool making in wild chimpanzees

    Folia Primatol.

    (1990)
  • C. Boesch et al.

    Different hand postures for pounding nuts with natural hammers by wild chimpanzees

  • M.E. Bush

    The thumb of Australopithecus afarensis

    Am. J. Phys. Anthropol.

    (1980)
  • M.E. Bush et al.

    Hominid carpal, metacarpal, and phalangeal bones recovered from the Hadar Formation: 1974–1977 Collections

    Am. J. Phys. Anthropol.

    (1982)
  • J. Charteris et al.

    Functional reconstruction of gait from the Pliocene hominidfootprints at Laetoli, northern Tanzania

    Nature

    (1981)
  • J. Charteris et al.

    Pliocene hominid gait: new interpretations based on available footprint data from Laetoli

    Am. J. Phys. Anthropol.

    (1982)
  • M. Christel

    Grasping techniques and hand preferences in Hominoidea

  • R.J. Clarke

    Early hominid footprints from Tanzania

    S. Afr. J. Sci.

    (1979)
  • R.J. Clarke

    Discovery of complete arm and hand ofthe 3.3 million-year-old Australopithecus skeleton from Sterkfontein

    S. Afr. J. Sci.

    (1999)
  • R.J. Clarke et al.

    Sterkfontein Member 2 foot bones of the oldest South African hominid

    Science

    (1995)
  • M.I. Coates et al.

    Fins, limbs, and tails: outgrowths and axial patterning in vertebrate evolution

    Bioessays

    (1998)
  • M.J. Cohn et al.

    Molecular control of vertebrate limb development, evolution and congenital malformations

    Cell Tissue Res.

    (1999)
  • R.S. Corruccini

    Shape in morphometrics: comparative analyses

    Am. J. Phys. Anthropol.

    (1987)
  • C. Darwin

    The Descent of Man and Selection in Relation to Sex

    (1871)
  • M.H. Day et al.

    Laetolil Pliocene hominid footprints and bipedalism

    Nature

    (1980)
  • E. Delson et al.

    Body mass in Cercopithecidae (Primates, Mammalia): estimation and scaling in extinct and extant taxa

    Anthropol. Pap. Am. Mus. Nat. Hist.

    (2000)
  • D.M. Fragaszy et al.

    Social processes affecting the appearance of innovative behaviors in capuchin monkeys

    Folia Primatol.

    (1990)
  • D.J. Futuyma

    Evolutionary Biology

    (1986)
  • D.L. Gebo

    Climbing, brachiation, and terrestrial quadrupedalism: historical precursors of hominid bipedalism

    Am. J. Phys. Anthropol.

    (1996)
  • D.N. Gomberg et al.

    Observations on the transverse tarsal joint of A. afarensis, and some comments on the interpretation of behavior from morphology

    Am. J. Phys. Anthropol.

    (1984)
  • J.v.L. Goodall

    Feeding behavior of wild chimpanzees

    Symp. Zool. Soc. Lond.

    (1963)
  • J.v.L. Goodall

    Tool-using in primates and other vertebrates

  • J.v.L. Goodall

    The Chimpanzees of Gombe: Patterns of Behavior

    (1986)
  • S.J. Gould

    Allometry in primates, with emphasis on scaling and the evolution of the brain

    Contrib. Primatol.

    (1975)
  • S.J. Gould et al.

    Exaptation—a missing term in the science of form

    Paleobiology

    (1982)
  • H. Greene

    Diet and arboreality in the emerald monitor, Varanus prasinus, with comments on the study of adaptation

    Fieldiana Zool.

    (1986)
  • W.C. Hartwig et al.

    Evolution of the hominid hand and tool making behavior

    Am. J. Phys. Anthropol.

    (1998)
  • J. de Heinzelin et al.

    Environment and behavior of 2.5-million-year-old Boure hominids

    Science

    (1999)
  • Cited by (114)

    • Tool use and the hand

      2023, Behaviour in our Bones: How Human Behaviour Influences Skeletal Morphology
    View all citing articles on Scopus
    View full text