Tracy L. Kivell1
1Department of Human Evolution, Max Planck Institute of Evolutionary Anthropology, Leipzig, Germany
Department of Human Evolution, Max Planck Institute of Evolutionary Anthropology, Deutscher Platz 6, Leipzig 04103, Germany
Received: 12 Nov. 2010
Accepted: 08 Feb. 2011
Published: 11 May 2011
How to cite this article:
Kivell TL. A comparative analysis of the hominin triquetrum (SKX 3498) from Swartkrans,
South Africa. S Afr J Sci. 2011;107(5/6), Art. #515, 10 pages.
© 2011. The Authors. Licensee: OpenJournals Publishing. This work is licensed under the Creative Commons Attribution License.
ISSN: 0038-2353 (print)
ISSN: 1996-7489 (online)
A comparative analysis of the hominin triquetrum (SKX 3498) from Swartkrans, South Africa
In This Research Article...
• Materials and methods
• Comparative morphological description of SKX 3498
• Results of discriminant function analyses
• Functional implications
• Taxonomic implications
The SKX 3498 triquetrum from Member 2 at Swartkrans Cave, South Africa is the only hominin triquetrum uncovered
(and published) thus far from the early Pleistocene hominin fossil record. Although SKX 3498 was found over
two decades ago, its morphology has not been formally described or analysed, apart from the initial description.
Furthermore, the taxonomic attribution of this fossil remains ambiguous as both Paranthropus and early
Homo have been identified at Swartkrans. This analysis provides the first quantitative analysis of the
SKX 3498 triquetrum, in comparison to those of extant hominids (humans and other great apes) and other fossil
hominins. Although the initial description of the SKX 3498 triquetrum summarised the morphology as generally
human-like, this analysis reveals that quantitatively it is often similar to the triquetra of all hominine
taxa and not necessarily humans in particular. Shared hominid-like morphology between SKX 3498 and Neanderthals
suggests that both may retain the symplesiomorphic hominin form, but that functional differences compared to
modern humans may be subtle. Without knowledge of triquetrum morphology typical of earlier Pliocene hominins,
the taxonomic affiliation of SKX 3498 remains unclear.
Between 1979 and 1986 a complete, undistorted hominin right triquetrum (SKX 3498) was discovered
from Member 2 at the Pleistocene site of Swartkrans Cave, South Africa (1.8 MYA – 1.0 MYA1).
Susman2,3 provided an initial description of this fossil, describing it as ‘
essentially humanlike’ in its overall shape and facet morphology and within the size range
of ‘small (5’0”) modern humans’. Only the elliptical shape of the pisiform
facet was described as being unique and unlike that of modern humans or chimpanzees.2,3
Since this initial description, however, the SKX 3498 triquetrum has been rarely mentioned in the
literature, most likely because it is the only early Pleistocene hominin triquetrum known (and
published) and little comparative data exist.4 Furthermore, because Swartkrans, and
Member 2 specifically, are associated with both Paranthropus robustus and early Homo
(Homo cf. erectus5), the taxonomic affiliation of this carpal bone remains
uncertain. This analysis provides the first quantitative analysis of this fossil in comparison to
the triquetra of modern humans and other great apes (Figure 1) and to published data on hominin
fossils (Ardipithecus ramidus, Homo neanderthalensis and archaic H. sapiens), with
the aim of clarifying its functional and taxonomic interpretation.
FIGURE 1: SKX 3498 triquetrum morphology in comparison to that of extant hominids. SKX 3498 is shown in all views compared to lateral, distolateral and palmar views
of Homo (male modern H. sapiens), Pan (female P. paniscus) and Pongo (male P. pygmeaus).
The qualitative description and measurement of the original SKX 3498 triquetrum were conducted at the
Ditsong (formerly Transvaal) National Museum of Natural History in Pretoria, South Africa. SKX 3498 was
compared to a large sample (n = 254) of extant hominids (humans and other great apes,
Table 1). The modern human sample comprised White and Black individuals and was derived from the Grant
collection (University of Toronto) and the Terry collection (Smithsonian Institute). The extant great
ape comparative sample included Pan paniscus, P. troglodytes, Gorilla gorilla,
G. berengei, Pongo pygmaeus and P. abelii and the individual specimens were housed
at the following institutions: The Powell-Cotton Museum, Musee Royal de l’ Afrique Centrale,
Max-Planck-Institut für evolutionäre Anthropologie, Museum für Naturkunde Berlin, The National Museum
of Natural History, Harvard Museum of Comparative Zoology, The Cleveland Museum of Natural History,
The Royal Ontario Museum and the University of Toronto.
TABLE 1a: Details of the sample of extant taxa used in this analysis.
TABLE 1b: Details of the sample of fossil taxa used in this analysis.
Comparisons to the few other fossil hominin triquetra were made using published data. Lovejoy et
al.6 provided two linear measurements for the Ar. ramidus triquetrum ARA-VP-
6/500–029: the ‘maximum dimension’ and ‘lunate surface breadth’,
assumed to be the equivalent of the maximum breadth of the triquetrum body and lunate facet,
respectively, in this analysis (Table 2 and Figure 2). Trinkaus7 offered several
linear measurements for three adult H. neanderthalensis triquetra: complete left triquetra
from Shanidar 6 and Shanidar 4, the latter of which has some pathology, and a complete right triquetrum
from Shanidar 5 (Table 2a). Finally, Sládek et al.8 provided data on two archaic
H. sapiens specimens, a left triquetrum from the Dolní Věstonice 3 individual and
a more complete right triquetrum from Dolní Věstonice 14.
TABLE 2a: Metric data used in this analysis (given in mm) for extant samples.
TABLE 2b: Metric data used in this analysis (given in mm) for fossil samples.
Nine linear measurements were taken to quantify the size of the triquetrum and its articular facets
(Figure 2). A geometric mean was used as the size variable and was calculated from the raw measurements
for each specimen.9 Each linear measurement was divided by the geometric mean of all
measurements to create a dimensionless shape ratio.10,11 For some fossil hominin
specimens, only a few linear measurements were available with which to calculate a geometric
mean (e.g. two variables for Ar. ramidus). Because a geometric mean is a volume that
requires at least three variables,12 a Spearman rank correlation (rs)
test was used to determine if a geometric mean derived from fewer variables was significantly
correlated with the geometric mean derived from the complete set of variables. Spearman rank
correlation was used, as opposed to Pearson’s correlation, because it is a more conservative
measure when the relationship between two variables is not necessarily linear.13
A significant correlation was determined if rs > 0.80 and p ≤
FIGURE 2: Linear measurements used to quantify the morphology of the SKX 3498 triquetrum.
Differences in shape ratios across extant taxa were assessed using a one-way analysis of
variance followed by a Tukey–Kramer post-hoc test for multiple comparisons.13
All statistical analyses were run with sexes pooled and results were considered statistically
significant at the p ≤ 0.05 level. Differences in triquetrum shape ratios amongst
extant groups and the fossil sample were evaluated graphically with box–and–whisker plots.
Finally, the morphology of the SKX 3498 triquetrum was further quantitatively compared to the
extant sample, Neanderthals and archaic H. sapiens Dolní Věstonice 14 using a
discriminant function analysis (DFA). DFA is a classification technique that generates a
linear combination of variables that maximises the probability of correctly assigning observations
into their predetermined groups.13 DFA of the extant comparative sample was used to
determine the utility of the triquetrum shape ratios to resolve taxonomic and/or functional groups.
Subsequently, DFA was also used to assign ‘unknown’ observations – that is, fossil
specimens – into ‘a priori-defined’ taxonomic groups. Because differences in
group sample sizes can bias the discriminant analysis and classification,14 data were
randomly culled to the lowest sample size (i.e. Pongo, n = 31) to test for
any adverse effects.
Spearman rank correlation (rs) revealed that the four additional geometric means
calculated from the subsets of variables available for fossil specimens (Table 2b) were significantly
correlated with the geometric mean derived from all nine variables: (1) the Neanderthal specimens
(rs = 0.99), (2) Dolní Věstonice 14 (rs = 0.99),
(3) Dolní Věstonice 3 (rs = 0.93) and (4) Ar. ramidus
(rs = 0.93). Therefore comparisons across shape ratios are considered robust.
Figure 3 to Figure 6 provide box–and–whisker plots for the relative differences across
extant and fossil taxa for each shape ratio. In all cases, the plot including comparisons to the
most fossil taxa is shown (i.e. using a shape ratio that is derived from fewer than nine variables)
and, unless otherwise stated, the relationships amongst the taxa did not differ substantially from
those representing a shape ratio derived from all nine variables. In instances where relationships
did change, multiple box–and–whisker plots are shown (e.g. for the maximum breadth of
the triquetrum body).
FIGURE 3: Box–and–whisker plots of the relative (i.e. size adjusted) height (a) and length (b) of the triquetrum body in extant and fossil taxa.
FIGURE 4: Box–and–whisker plots of the relative breadth of the triquetrum body (BTB) in extant and fossil taxa. The relative size of the BTB varied depending on the
variables used to calculate the geometric mean: the BTB shape ratio derived from all nine variables (a), and from the subset of variables available for Neanderthal
specimens (b), Dolní Vestonice 3 (c) and Ar. ramidus (d).
FIGURE 5: Box–and–whisker plots of the relative size of the triquetrum’s lunate and
hamate facets in extant and fossil taxa. The relative length of the lunate facet was
derived from the subset of variables available for the Dolní Vestonice 14 (a) and Ar.
ramidus (b) specimens. The relative height of the lunate facet (c) and the relative
breadth (d) and height (e) of the hamate facet were derived from the subset of
variables available for the Dolní Vestonice 14 specimen.
FIGURE 6: Box–and–whisker plots of the relative size of the triquetrum’s pisiform facet in extant and fossil taxa. The relative breadth of the pisiform facet was derived
from the subset of variables available for the Dolní Vestonice 3 specimen (a) and the relative length was derived from variables available for the Dolní Vestonice 14
Comparative morphological description of SKX 3498
Previous qualitative descriptions of the SKX 3498 triquetrum described the overall shape of the
body and the facet morphology as generally human-like.2,3 The results of this analysis
generally support this assessment, but morphological similarities are often shared with all
hominine taxa (African apes and humans), not just humans, and there are certain aspects of the
morphology that are unlike that of humans (Figures 3–6). The SKX 3498 triquetrum body is
most similar to hominines in its relative dorsopalmar height, as are Neanderthal and Dolní Vě
stonice 14 specimens, and in its proximodistal length (Figure 3). However, the mediolateral
breadth of the triquetrum body (the longest dimension) is, relative to carpal size, broader
than the mean breadth of all extant hominines (Figure 4). The mediolateral breadth of SKX 3498
falls outside or within the upper range of variation in breadth in hominines and, in some
comparisons, is more similar to Pongo (Figure 4c). The broad breadth of SKX 3498 is similar
to that of Neanderthals in some quantitative comparisons (Figure 4b and 4c). Ar. ramidus
also displays a relatively broad triquetrum body (Figure 4d) whilst Dolní Věstonice 3 is most
similar to the mean of modern humans (Figure 4c).
The lunate facet of SKX 3498 is similar to extant hominines2,3 and other fossil hominins
in being slightly concave (unlike Pongo, which is strongly convex) and almost square in shape
(Figure 1). The relative length of the SKX 3498 lunate facet is similar to that of extant hominines
and longer than those of all other fossil hominin taxa (Figure 5a and 5b). However, relative to carpal
size, the height of the SKX 3498 lunate facet is greater than the mean of all extant hominids (humans
and other great apes) and Dolní Věstonice 14 and is only within the upper range of variation of
Pan (Figure 5c). In this way, SKX 3498 is similar to Neanderthals. The SKX 3498 hamate facet
is also expansive, with a slight concavo-convex surface that ‘wraps around’ the dorsomedial
edge of the triquetrum body (Figure 2). Relative to carpal size, the breadth of the SKX 3498 hamate
facet is greater than the mean of all hominine taxa (but less than that of Pongo) and is most
similar to that of Neanderthals and, less so, Dolní Věstonice 14 (Figure 5d). The relative
height of the hamate facet is also comparatively greater than the mean height of other extant hominid
taxa, Neanderthals and Dolní Věstonice 14 (Figure 5e).
The pisiform facet is positioned at the distomedial end of the palmar surface of the triquetrum and
its orientation is mostly palmar, but also slightly medially facing (Figure 1). Results of this analysis
support Susman’s2,3 description of the SKX 3498 pisiform facet as being relatively
small compared to other hominids. The longest axis (which is almost mediolateral) of the pisiform facet
is most similar to African apes in being relatively long and falls outside the range of variation seen
in modern humans and other fossil hominins (Figure 6a). However, proximodistally the pisiform facet is
uniquely short, falling outside the range of all extant taxa, Neanderthals and Dolní Věstonice 14
(Figure 6b). This combination in SKX 3498 produces an elliptically shaped facet that is unlike the more
circular facet of extant hominids.2,3 The palmar surface is also marked by a deep groove at
the distolateral end for the attachment of a well-developed lunatotriquetrum ligament, which appears
similar to that of Ar. ramidus6 and some modern human specimens, but is unlike the
morphology of non-human great apes (Figure 1).
Results of discriminant function analyses
Two DFA were conducted, each based on the subset of shape ratios available for the Neanderthal and Dolní
Věstonice 14 specimens (Table 2b). The DFA results on the culled data set yielded the same shape
ratio loadings on each discriminant function, a similar distribution of taxa in the scatter plot and the
same taxonomic classification of the fossils as the DFA using the complete sample. Therefore, because a
larger sample provides a better representation of the natural variation within a given taxon, only the
results based on DFA of the complete sample are discussed here.
In the DFA based on the Neanderthal shape ratios, the first discriminant function distinguished
Pongo from all other hominines because of its relatively broad triquetrum body and hamate
facet and relatively short triquetrum body and long pisiform facet (Table 3a, Figure 7a). The second
discriminant function distinguished most Pan and Gorilla specimens from Homo
specimens by their relatively tall lunate facet and short pisiform facet. Extant taxa were correctly
classified into their respective taxonomic groups at only 64.3%. SKX 3498 and all Neanderthal specimens
were strongly classified as either Pongo or Pan (Table 3b), which reflects the placement
of all fossil hominins amongst the extreme range of the Pongo on the first discriminant function
and within range of Pan on the second discriminant function. The DFA based on the Dolní
Věstonice 14 shape ratios produced similar results, with similar variable loadings on both
discriminant functions (Table 3a) and an almost identical distribution of taxa in the scatter plot
(Figure 7b). Although the correct classification of extant taxa into their respective taxonomic
groups was even lower (60.5%), the classification of all fossil taxa into the ‘a priori-
defined’ groups was the same (Table 3b). However, it is notable that the Dolní Věstonice
14 specimen was the only fossil hominin to be classified as Homo.
TABLE 3a: Results from two discriminant function analyses (DFA) based on shape
ratios available for Neanderthal and Dolní Vestonice 14 specimens: Correlations
of prediction variables with discriminant functions.
TABLE 3b: Results from two discriminant function analyses (DFA) based on
shape ratios available for Neanderthal and Dolní Vestonice 14 specimens:
Predicted classification of fossil specimens.
FIGURE 7: Scatter plots of the first two discriminant functions based on the shape ratios available for (a) the Neanderthal specimens and (b) the Dolní Vestonice 14
This analysis provided a quantitative comparison of the SKX 3498 hominin triquetrum from Swartkans Cave with
the triquetra of extant hominids and other fossil hominins. Susman’s2,3 original description
of this fossil summarised the morphology as generally human-like. The quantitative analyses presented here
reveal that the morphology of SKX 3498 often is similar to that of extant hominines, but not specifically
humans. Relative to carpal size, the height and length of the SKX 3498 triquetrum body, the shape of the
lunate facet, and some aspects of the hamate and pisiform facets are similar to both modern humans and African
apes. The relative breadth of the SKX 3498 triquetrum body and its hamate facet are intermediate between
that of hominines and Pongo, whilst the length of the pisiform facet is uniquely short, as described
by Susman2,3, compared to all other extant and fossil taxa. Neanderthals share much of this
morphology with SKX 3498, but have a more Pongo-like hamate facet and length of the lunate facet
relative to carpal size. The two morphometric variables available for Ar. ramidus are more similar
to Pongo whilst Dolní Věstonice 3 and 14 are more similar to modern humans.
Both SKX 3498 and all Neanderthal specimens were strongly classified in the DFA as non-human hominids, despite
having morphology that has been generally described qualitatively as human-like.2,3,7 In addition,
the three Neanderthal triquetra are classified as different taxa (Pan vs. Pongo), despite being
from the same taxon and the same site (Shanidar Cave).7 Although this result is consistent with the
many similarities in relative size of the triquetrum body and facet morphology of these fossil specimens to
those of hominines or Pongo, the results are likely confounded by three methodological factors. Firstly,
many of the shape ratios that distinguished Pongo from extant hominine taxa – namely the broader
triquetrum body and hamate facet and shorter pisiform facet – are also shape ratios in which both SKX
3498 and the Neanderthal specimens are quantitatively (i.e. relative to carpal size) more similar to Pongo.
Secondly, in DFA the unique or intermediate morphology of SKX 3498 and Neanderthals must be classified
into ‘a priori-defined’ group. Thirdly, the triquetrum measurements used in this study were poor
in distinguishing amongst extant taxonomic or functional groups and, in particular, the more subtle variation
amongst hominine triquetrum morphology. Thus, although qualitative comparisons describe the morphology of SKX
34982,3(Figure 1) and Neanderthals7 as generally human-like, this similarity was not
captured by the linear measurements used in this analysis. It is possible that with the inclusion of angles
of orientation, curvature or surface area of facets, for example, the variation amongst extant hominines and
the more human-like aspects of fossil hominin triquetrum morphology may be revealed.
Despite these methodological limitations, it is notable that the Dolní Věstonice 14 specimen was
correctly classified as H. sapiens in the DFA, suggesting that: (1) the measurements used in this
analysis were robust enough to distinguish human from non-human triquetrum morphology and (2) the non-human
classification of SKX 3498 and the Neanderthal specimens reflects, at least to some degree, true morphological
differences from those of archaic or modern humans that have not been previously recognised. Thus, the
evolutionary implications of these results need to be further explored.
The SKX 3498 triquetrum, as well as those of Neanderthals and Ar. ramidus, is relatively
more mediolaterally broad than modern humans and African apes. A relatively broad triquetrum body
suggests that the functional role of the pisiform could have been more accentuated compared to that
of humans. The pisiform serves as an attachment site for the tendon of the flexor carpi ulnaris
muscle and the shape, size and position of the pisiform can alter the moment arm of this
muscle.15,16 Relative to humans, the broader body of SKX 3498 would move the already
distomedially placed pisiform facet more distomedially, which, depending on the morphology of the
pisiform, could increase the flexion and adduction moment arm of the flexor carpi ulnaris muscle.
However, such a functional hypothesis depends strongly on the size and shape of the pisiform.
SKX 3498 has a unique elliptically shaped pisiform facet2,3 – morphology that is
not found in any of the extant or fossil taxa in this sample. Unfortunately, little is known about
the relationship between the pisiform facet of the triquetrum and the shape of the pisiform itself
and there is a paucity of both bones in the hominin fossil record. Thus the functional implications
that can be derived from the shape of the triquetrum’s pisiform articulation are limited. The
unique pisiform facet in SKX 3498 suggests that the shape of the pisiform may have been different
from the more circular, small, pea-shaped pisiform of archaic8 and modern humans and
Neanderthals.7 However, it cannot be presumed that the pisiform associated with the SKX
3498 triquetrum was elongated and rod-shaped like that of Au. afarensis17,18 and
great apes. The corresponding triquetrum facet of the Au. afarensis pisiform appears more
circular than the equivalent pisiform facet on the SKX 3498 triquetrum,17 suggesting that
the associated pisiform of SKX 3498 was different from that of Au. afarensis. Variation in
pisotriquetrum facet shapes may instead reflect the degree of movement at this joint, rather than
the shape of the pisiform itself. The synovial pisotriquetrum joint permits a large degree of movement
in the pisiform, being pulled by the attached ligaments and tendons proximally during flexion and
adduction and distally during extension and abduction.15 Thus the functional implications,
if any, of the unique SKX 3498 pisiform facet morphology remain unclear.
The relatively expansive lunate and hamate facets suggest that the SKX 3498 morphology may have allowed
for slightly more mobility at the lunatotriquetrum and hamatotriquetrum joints than is typically found
in extant hominines. Although the triquetrum is firmly bound within the carpus via extrinsic and intrinsic
ligaments to the distal ulna and adjacent carpals, it is still capable of a large degree of movement within
the wrist.19,20 The relatively broad hamate facet suggests that the triquetrum may
have been able to move into a more extreme mediodistal position, rotating along the convex surface
of the hamate, allowing for greater ulnar deviation than is typically found in hominines.20
For example, Pongo, with the broadest hamate facet in the sample, has a much larger range of
ulnar deviation (98 degrees21) than the narrower facets of Pan (70 degrees21)
or Homo (38 degrees22). However, without knowledge of the mutual facets on the
lunate and hamate or soft tissue morphology, inferring mobility in these joints remains speculative.
Given that the breadth of the hamate facet of the SKX 3498 triquetrum and that of Neanderthals
fall within the upper range of the variation in breadth in hominines, the degree of ulnar deviation
in fossil hominins was likely more similar to that of modern humans or African apes than to the
extreme mobility of Pongo.
It is similarly challenging to interpret the functional significance of the SKX 3498 morphology
in terms of tool-use or tool-making capabilities as the triquetrum is rarely discussed in the
analyses of hominin hand use. This absence in the literature is likely as a result of the scarcity
of triquetra in the early hominin fossil record, though new discoveries of relatively complete
australopithecine hands at Sterkfontein (2.2 MYA23,24,25) and Malapa (1.8 MYA26,27)
in South Africa promise to offer a more comprehensive perspective on overall hominin hand function.
Much of the research on hominin wrist and hand function has, understandably, focused on the interface
between the distal carpal row and the metacarpals.6,18,28,29,30,31,32,33,34,35 Analyses
of the nearly complete wrists of Ar. ramidus6 or Neanderthal and early H.
sapiens28,33,34 do not specifically discuss the functional morphology of the
triquetrum. Thus, direct comparisons to other hominin triquetra are challenging. However,
given (1) the general quantitative similarities between SKX 3498 and Neanderthal triquetra
in this analysis and (2) that Neanderthal hand morphology overall has been interpreted as
generally like that of modern humans and indicative of modern human manipulative
capabilities,7,28,33,34 it is reasonable to
assume that the morphological variation in triquetrum morphology
between SKX 3498/Neanderthals and humans implies relatively subtle functional differences
within the wrist as a whole. That said, recent experimental analyses have shown that the
wrist plays a particularly important role in the biomechanics of stone tool
production36 and that the intrinsic muscles of the fifth digit, not just the
muscles of the thumb and index finger, are important for tool use37
(but see Williams and Richmond38). Thus, further functional studies of
the medial wrist bones, particularly the triquetrum and pisiform, may offer new insight
into the hand use of early hominins.
The limited data available for Ar. ramidus (4.4 MYA),6 as well
as much of the morphology of SKX 3498 (1.8 MYA – 1.0 MYA1) and
the Shanidar Neanderthal specimens (approximately 0.11 MYA – 0.04 MYA7,39),
are similar to non-human hominids and are distinct from the Dolní Věstonice specimens
(0.03 MYA8) and modern humans. This division suggests that the triquetrum morphology
documented in Ar. ramidus, SKX 3498 and Neanderthals may be symplesiomorphic for
fossil hominins and that the more derived form of archaic and modern humans may be a
relatively recent development in human evolutionary history.
Unfortunately this analysis provides little resolution regarding the taxonomic attribution of the
SKX 3498 triquetrum. Both P. robustus and early Homo (Homo cf.
erectus5) have been identified at Swartkrans and in Member 2
specifically1,5 and these results demonstrate that SKX 3498 could be attributed to
either of these taxa (or to an as of yet unidentified hominin taxon). The similarities to
Neanderthal morphology might suggest that SKX 3498 is more likely to be early Homo
rather than P. robustus. However, it is unclear if both SKX 3498 and Neanderthals
simply retain sympleisomorphic hominin morphology common to many hominin taxa or if earlier
hominins (e.g. australopithecines and paranthropines) had a different, more great-ape-like
morphology from which SKX 3498 and Neanderthals have derived. Given the modern human-like
morphology of the Dolní Věstonice specimens relative to the other fossil hominins in
this analysis, it may be more likely that similarities between SKX 3498 and Neanderthals are
primitive hominin retentions and that modern human-like triquetrum morphology has evolved
relatively recently (i.e. within H. sapiens only). In such a scenario, the SKX 3498
could be either P. robustus or early Homo. Unfortunately the available information
on the Ar. ramidus triquetrum is not sufficient to establish the morphotype of a potential
last common ancestor of Homo-Pan.6,40 Until descriptions of
australopithecine triquetra from Sterkfontein23,24 and Malapa26
are available, the evolution of hominin triquetrum morphology and the taxonomic affiliation
of SKX 3498 remain ambiguous.
The initial description of the SKX 3498 triquetrum summarised its morphology as human-like, with a unique
pisiform facet.2,3 This analysis generally supports
this description, but reveals that the SKX
3498 morphology is most similar to all extant hominine taxa and not specifically humans. In this way, SKX
3498 shares much of its morphology with Neanderthals and both are often distinct from archaic H.
sapiens Dolní Věstonice specimens and modern humans. Despite the methodological limitations of
this study, the quantitative distinction made between archaic and modern H. sapiens, on the one hand,
and SKX 3498 and Neanderthals (and the limited data available for Ar. ramidus) on the other hand,
suggests true morphological differences between the latter specimens and H. sapiens that have not
been previously recognised.2,3,7
The triquetrum morphology of SKX 3498 and Neanderthals may be
considered symplesiomorphic for fossil hominins whilst the morphology found in H. sapiens likely
evolved relatively recently in hominin evolution. Functional interpretations of the SKX 3498 morphology
are tentative, but may suggest slightly more mobility at the triquetrolunate and triquetrohamate joints
and an enhanced function of the pisiform and flexor carpi ulnaris muscle than is found in modern humans.
However, the morphological similarities between SKX 3498 and Neanderthals, for which in the latter numerous
wrist bones are known and overall wrist function is considered generally human-
relatively subtle functional differences between fossil hominin and modern human triquetra. Without a clearer
understanding of triquetrum morphology in earlier hominins, the taxonomic affiliation of the SKX 3498
triquetrum remains unknown.
I am grateful for the support and hospitality of Stephany Potze and Dr. Martin Kruger of the Ditsong
(Transvaal) National Museum of Natural History and for making SKX 3498 available for study. I am also
grateful to the following curators from whom the comparative sample was obtained: W. Wendelen and M.
Louette (Musée royal de l’Afrique centrale), F. Mayer and S. Jancke (Museum für Naturkunde Berlin),
U. Schwartz (Max-Planck-Institut für evolutionäre Anthropologie), M. Harman (Powell-Cotton Museum), L.
Gordon (Smithsonian Institution), Y. Haile-Selassie and L. Jellema (Cleveland Museum of Natural History),
and J. Eger and S. Woodward (Royal Ontario Museum). Finally, I am thankful to two anonymous reviewers for
constructive comments, Matthew Skinner for helpful discussions that improved this manuscript and
Jean-Jacques Hublin for research support. This research was funded by the Natural Sciences Engineering
Research Council of Canada, General Motors Women in Science and Mathematics and the Max Planck Society.
1. Brain CK. Structure and stratigraphy of the Swartkrans cave in light of the new excavations. In: Brain CK,
editor. Swartkrans: A cave’s chronicle of early man. Pretoria: Transvaal Museum, 1993; p. 23–34.
2. Susman RL. New postcranial remains from Swartkrans and their bearing on the functional morphology and behaviour
of Paranthropus robustus. In: Grine FE, editor. Evolutionary history of the ‘robust’ australopithecines.
New York: Aldine de Gruyter, 1988; p. 149–172.
3. Susman RL. New hominid fossils from the Swartkrans formation (1979–1986 excavations): Postcranial specimens.
Am J Phys Anthropol. 1989;79:451–474.
4. Tocheri MW, Orr CM, Jacofsky MC, Marzke MW. The evolutionary history of the hominin hand since the last
common ancestor of Pan and Homo. J Anat. 2008;212:544–562.
5. Susman RL, de Ruiter D, Brain CK. Recently identified postcranial remains of Paranthropus
and early Homo from Swartkrans Cave, South Africa. J Hum Evol. 2001;41:607–629.
6. Lovejoy CO, Simpson SW, White TD, Asfaw B, Suwa G. Careful climbing in the Miocene: The forelimbs of
Ardipithecus ramidus and humans are primitive. Science. 2009;326:70e1–70e8.
7. Trinkaus E. The Shanidar Neandertals. New York: Academic Press; 1983.
8. Sládek V, Trinkaus E, Hillson SW, Holliday TW. The people of the Pavlovian: Skeletal catalogue and
osteometrics of the Gravettian fossil hominids from Dolní Vĕstonice and Pavlov. Dolní Vĕstonice
Studies 5. Brno: Akademie vĕd Česke republicky; 2000.
9. Jungers WL, Falsetti AB, Wall CE. Shape, relative size and size-adjustments in morphometrics. Yrbk
Phys Anthropol. 1995;38:137–161.
10. Mossiman JE. Size allometry: Size and shape variables with characterizations of the lognormal and
generalized gamma distributions. J Am Stat Assoc. 1970;56:930–945.
11. Darroch JN, Mosimann JE. Canonical and principal components of shape. Biometrika. 1985;72:241–252.
12. Coleman MN. What does geometric mean, mean geometrically? Assessing the utility of geometric mean and
other size variables in studies of skull allometry. Am J Phys Anthropol. 2007;135:404–415.
13. Quinn GP, Keough MJ. Experimental design and data analysis for biologists. Cambridge: Cambridge University Press; 2002.
14. Sanchez PM. The unequal group size problem in discriminant analysis. J Acad Marketing Sci. 1974;102:629–633.
15. Taleisnik J. The wrist. New York: Churchill Livingstone; 1985.
16. Hamrick MW. Functional osteology of the primate carpus with special reference to Strepsirhini. Am J Phys Anthropol.
17. Bush ME, Lovejoy CO, Johanson DC, Coppens Y. Hominid carpal, metacarpal and phalangeal bones recovered
from the Hadar Formation: 1974–1977 collections. Am J Phys Anthropol. 1982;57:651–677.
18. Marzke MW. Joint functions and grips of the Australopithecus afarensis hand, with special
reference to the region of the capitate. J Hum Evol. 1983;12:197–211.
19. Feipel V, Rooze M, Louryan S, Lemort M. Functional anatomy of the carpus in flexion and extension and in
radial and ulnar deviations: An in vivo two- and three-dimensional CT study. In: Schuind F,
An K-N, Cooney WP III, Garcia-Elias M, editors. Advances in the biomechanics of the hand and wrist. New York:
Plenum Press, 1994; p. 255–270.
20. Moritomo H, Goto A, Yoshinobu S, Sugamoto K, Murase T, Yoshikawa H. The triquetrum-hamate joint: An
anatomic and in vivo three-dimensional kinematic study. J Hand Surg. 2003;28A(5):797–805.
21. Tuttle RH. Terrestrial trends in the hands of the Anthropoidea. Proceedings of the 2nd International
Congress of Primatology; 1968; Atlanta, GA, USA. Basel: S. Karger; 1969. p. 192–200.
22. Ryu J, Palmer AK, Cooney WP III. Wrist joint motion. In: An K-N, Berger RA, Cooney WP III, editors. Biomechanics
of the wrist joint. New York: Spinger-Verlag, 1991; p. 37–60.
23. Clarke RJ. Discovery of complete arm and hand of the 3.3 million-year-old Australopithecus skeleton
from Sterkfontein. S Afr J Sci. 1999;95:477–480.
24. Clarke RJ. Latest information on Sterkfontein’s Australopithecus skeleton and a new look at
Australopithecus. S Afr J Sci. 2008;104:443–449.
25. Pickering R, Kramers JD. Re-appraisal of the stratigraphy and determination of new U-Pb dates for the S
terkfontein hominin site, South Africa. J Hum Evol. 2010;59:70–86.
26. Berger LR, DeRuiter DJ, Churchill SE, et al. Australopithecus sediba: A new species
of Homo-like australopith from South Africa. Science. 2010;328:195–204.
27. Dirks PHGM, Kibii JM, Kuhn BF, et al. Geological setting and age of Australopithecus sediba
from Southern Africa. Science. 2010;328:205–208.
28. Niewoehner WA, Weaver AH, Trinkaus E. Neandertal capitate-metacarpal articular morphology. Am J
Phys Anthropol. 1997;103:219–233.
29. Ricklan DE. Functional anatomy of the hand of Australopithecus africanus. J Hum Evol. 1987;16:643–664.
30. Marzke MW. Precision grips, hand morphology and tools. Am J Phys Anthropol. 1997;102:91–110.
31. Marzke MW, Wullsetin KL, Viegas SF. Evolution of the power (squeeze grip) and its morphological correlates
in hominids. Am J Phys Anthropol. 1992;89:283–298.
32. Marzke MW, Marzke RF. Evolution of the human hand: Approaches to acquiring, analysing and interpreting
the anatomical evidence. J Anat. 2000;197:121–140.
33. Niewoehner WA. The functional anatomy of late Pleistocene and recent human carpometacarpal and
metacarpophalangeal articulations. PhD thesis, Albuquerque, University of New Mexico, 2000.
34. Niewoehner WA. Behavioral inferences from the Skhul/Qafzeh early modern human hand remains. Proc Natl
Acad Sci. 2001;98(6):2979–2984.
35. Tocheri MW, Marzke MW, Liu D, et al. Functional capabilities of modern an fossil hominid hands:
Three-dimensional analysis of trapezia. Am J Phys Anthropol. 2003;122:101–112.
36. Williams EM, Gordon AD, Richmond BG. Upper limb kinematics and the role of the wrist during stone tool
production. Am J Phys Anthropol. 2010;143:134–145.
37. Marzke MW, Toth N, Schick K, et al. EMG study of hand muscle recruitment during hard hammer percussion
manufacture of Oldowan tools. Am J Phys Anthropol. 1998;105:315–332.
38. Willams EM, Richmond BG. Hand pressure during Oldowan stone tool production. Am J Phys Anthropol. 2010;141:243.
39. Solecki RS. Shanidar IV, a Neanderthal flower burial in northern Iraq. Science. 1975;190:880–881.
40. Sarmiento EE. Comment on the palaeobiology and classification of Ardipithecus ramidus.