No Association between 2D:4D Ratio and Hunting
Success among Hadza Hunters
Duncan N. E. Stibbard-Hawkes
#Springer Science+Business Media, LLC, part of Springer Nature 2019
The ratio of index- and ring-finger lengths (2D:4D ratio) is thought to be
related to prenatal androgen exposure, and in many, though not all, populations,
men have a lower average digit ratio than do women. In many studies an
inverse relationship has been observed, among both men and women, between
2D:4D ratio and measures of athletic ability. It has been further suggested that,
in hunter-gatherer populations, 2D:4D ratio might also be negatively correlated
with hunting ability, itself assumed to be contingent on athleticism. This
hypothesis has been tested using endurance running performance among runners
from a Western, educated, and industrialized population as a proximate measure
of hunting ability. However, it has not previously been tested among actual
hunter-gatherers using more ecologically valid measures of hunting ability and
success. The current study addresses this question among Tanzanian Hadza
hunter-gatherers. I employ a novel method of assessing hunting reputation that,
unlike previous methods, allows granular distinctions to be made between
hunters at all levels of perceived ability. I find no statistically significant
relationship between digit ratio and either hunting reputation or two important
hunting skills. I confirm that Hadza men have higher mean 2D:4D ratios than
men in many Western populations. I discuss the notion that 2D:4D ratio may be
the consequence of an allometric scaling relationship between relative and
absolute finger lengths. Although it is difficult to draw clear conclusions from
these results, the current study provides no support for the theorized relation-
ship between 2D:4D ratio and hunting skill.
Keywords 2D:4D .Digit ratio .Hadza .Hunting reputation .Foragers .Hunter-gatherers
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12110-019-
09359-z) contains supplementary material, which is available to authorized users.
*Duncan N. E. Stibbard-Hawkes
Department of Anthropology, Durham University, Durham DH1 3LE, UK
Digit [2D:4D] ratio is the length of the second digit of the hand (index finger) over the
length of the fourth digit (ring finger). In many (Manning 2002; Manning and Fink
2008; Manning et al. 1998,2000; Rammsayer and Troche 2007; Trivers et al. 2006),
though not all (Apicella et al. 2016;Brosnan2008; Manning 2002) human populations,
digit ratios show considerable sexual dimorphism; men generally have lower digit
ratios than women. The same pattern of sexual dimorphism in digit ratio has also been
observed in several nonhuman species (Brown et al. 2002a;BurleyandFoster2004;
Cain et al. 2013; Direnzo and Stynoski 2012; Tobler et al. 2011). Several authors (Galis
et al. 2010; Manning and Fink 2008) have argued that sexual dimorphism in digit ratio
is a product of sex differences in fetal androgen exposure. Studies of early sex
differentiation in the digit ratios of human fetuses (Galis et al. 2010), as well as negative
associations between digit ratio and congenital adrenal hyperplasia (Brown et al.
2002b;Oświecimska et al. 2012), appear to confirm this supposition.
The negative associations observed between androgen exposure and digit ratio have
led some authors to infer that digit ratios may act as a good measure of both
developmental and adult testosterone levels. For this reason, authors have looked for,
and in many cases found, negative associations between digit ratio and other measures,
including athleticism, fertility, and fecundity, also thought to be associated with high
testosterone: Among ∼153,000 participants in a BBC online survey which employed
self-measurement (Manning and Fink 2008), lower male digit ratios were associated
with higher lifetime fertility (number of children corrected for age), sensation-seeking
behaviors and also “dominance”(i.e., responses to 10 questions taken from the
International Personality Item Pool (www.ipip.com) concerning the desire to outdo,
surpass, or control peers). Manning (2002) found a negative correlation between digit
ratio and sperm count. Hönekopp et al. (2006) identified a negative relationship
between digit ratio and number of sex partners in heterosexual German
undergraduates. Finally, in a North American and British population, Wlodarski et al.
(2015) observed that low digit ratios occurred slightly more often in more promiscuous
men. However, Manning and Fink (2008) found no such trend. Further, Falcon (2016)
was unable to replicate Wlodarski et al.’s findings and criticized their method of
analysis (though see Dunbar and Wlodarski 2016 for reply).
Digit ratio has also been negatively associated, in many studies, with mea-
sures of athleticism and sporting success in both sexes (Lombardo et al. 2018;
Manning and Taylor 2001; Manning et al. 2007). Male professional footballers
have lower mean digit ratios than non-professional controls (Manning and
Tay lor 2001). In a study of 607 British women, digit ratio was found to be
significantly negatively associated with lifetime sporting achievement (Paul
et al. 2006). Two independent studies found significant negative associations
between digit ratio and endurance running performance (Longman et al. 2015;
Manning et al. 2007); Manning and Hill (2009)foundasimilarnegative
relationship between digit ratio and sprinting ability. Digit ratio has been
associated with rowing performance (Longman et al. 2011), skiing speeds
(Manning 2002), and rugby ability (Bennett et al. 2010), as well as fencing
ability among female, but not male fencers (Voracek et al. 2010). It has been
suggested, for this reason, that digit ratio is in some way connected to
intrasexual competition in humans (Manning and Taylor 2001), as well as
nonhuman primates (Howlett et al. 2015;Nelsonetal.2010).
Related to these findings, the persistence hunting hypothesis has previously been
invoked to highlight a potential link between digit ratio and hunting success among
forager populations (Longman et al. 2015). Proponents of the persistence hunting
hypothesis posit that humans are well-adapted endurance runners. Lieberman et al.
(2007a) have suggested that endurance running is an important evolved trait that has
allowed human hunters to capture their prey by persistence hunting; “following an
animal, keeping it above its trot-gallop transition for several hours, driving the animal
into hyperthermia so that it can be killed safely at close range”(2007a:290). Drawing
on persistence hunting theory, Longman et al. (2015:9) cited endurance running
performance in a half-marathon as a proxy for hunting success. They found that
endurance running performance and digit ratio were inversely correlated and concluded
that “hunting ability might therefore act as a reliable signal of male fitness”(2015:8).
Although complex, Longman et al.’s(2015) argument is based on the following
clearly-set-out assumptions: (a) Hunting, as others have suggested (e.g., Hawkes et al.
2018, though see Stibbard-Hawkes 2019), is a form of male intrasexual competition
and acts as an honest signal (Hawkes and Bird 2002), perhaps of athletic ability. (b)
Athletic ability is positively related to testosterone. (c) Adult testosterone is positively
related to reproductive fitness, including measures of fertility. (d) Adult testosterone is,
furthermore, positively related to developmental androgen exposure. (e) The 2D:4D
ratio is also related to developmental androgen exposure.
Although the argument is clear, there are some complications. First, Longman et al.’s
argument rests on a causally complex chain of assumptions. Therefore, although each
of these assumptions appears evidenced and reasonable, the model lacks parsimony.
Second, the ecological validity of the persistence hunting hypothesis has been the
subject of some debate (Lieberman et al. 2007b; Pickering and Bunn 2007). In
ethnographic and living hunter-gatherer populations, persistence hunting is rare. Even
among the Kua San, one of only two modern forager groups known to employ
endurance hunting, the practice is uncommon, employed far less frequently than
ambush hunting (Pickering and Bunn 2007). Hunters in most ethnographically record-
ed and living hunter-gatherer groups more often either make use of netting and ambush
tactics (Wilkie and Curran 1991) or, in the majority of cases, hunt using long-range,
mechanically projected weapons and weapon poisons (Lee 1979; Marlowe 2010).
Lieberman et al. (2007b) have, however, argued that too great an emphasis is placed
on the ethnographic record, especially given that mechanically projected weaponry and
weapon poisons appear to be relatively recent inventions (Brown et al. 2012; Wadley
et al. 2015). This debate has not been resolved to the satisfaction of all parties
(Lieberman et al. 2007b). It is apparent, however, that marathon performance,
Longman et al.’s chosen proxy for hunting success, is at the very least not an
ecologically valid reflection of hunting practices in most extant forager groups and
perhaps not in the majority of past populations either (Pickering and Bunn 2007). There
is a call to test this hypothesis using measures that are more reflective of real foraging
patterns in living populations.
Third, Longman et al.’s sample of Western, educated participants in a Cambridge-
shire fun-run are likely to be unrepresentative of foragers in several respects, including
diet, activity budget, and pathogen exposure. For most of our evolutionary history,
humans have lived in small-scale societies without schools, hospitals, or supermarkets
and with “greater exposure to hunger, pain, chronic diseases, and lethal dangers”
(Henrich et al. 2010:80). For this reason, Western populations “may often be the worst
population[s] from which to make generalizations”(Henrich et al. 2010:79), and results
typical of studies from Western populations often do not conform to patterns observed
elsewhere (Henrich et al. 2010). Indeed, the pattern of sexual dimorphism in digit ratio
frequently observed in Western populations (e.g., Manning et al. 2000) does not hold
true for the Hadza (Apicella et al. 2016). Despite these concerns, Longman et al. (2015)
set out a reasoned precedent for expecting an inverse relationship between digit ratio
and hunting success.
The Hadza of northern Tanzania are one of very few populations who continue to
subsist through hunting and gathering. Endurance running does not feature in the
Hadza hunting repertoire. However, general fitness and athletic ability, elsewhere
associated with digit ratio, are certainly important; Hunting reputation has been asso-
ciated with several indices of athletic ability (Apicella 2014; Stibbard-Hawkes et al.
2018) and many foraging activities, including climbing for honey, carrying food and
other resources, as well as drawing/aiming bows, require a significant degree of
strength and physical fitness (Apicella 2014; Blurton Jones et al. 2002).
Digit ratio has been previously measured among the Hadza. A recent sum-
mary of published data on Hadza digit ratios (Apicella et al. 2016) showed an
absence of sexual dimorphism in left-hand measurements (0.973–0.997 for men
vs. 0.965–1.01 for women). In right-hand measurements, contrary to the pattern
commonly observed elsewhere, Hadza women had a lower mean digit ratio than
Hadza men (0.982–0.994 for men vs. 0.965–0.980 for women) (Apicella et al.
2016). Butovskaya et al. (2010,2015) did find that Hadza men had lower digit
ratios than Hadza women. Butovskaya et al. (2010) report apparent sexual
dimorphism in both hands (0.95 in both male left- and right-hand mean digit
ratios, respectively), although Butovskaya et al. (2015) report smaller sex
differences (men’s digit ratios had a mean of 0.98 in left-hand measures and
0.97 in right-hand measures, compared with left and right means of 0.99 and
0.98, respectively, for women). Furthermore, Hadza male digit ratios were, in
most cases (Apicella et al. 2016; Butovskaya et al. 2015; though not
Butovskaya et al. 2010), higher than or equivalent to mean digit ratios observed
among women in US samples (e.g., Puts et al. 2004). Although there is no
conclusive explanation for this pattern, Apicella et al. (2016) have proposed
that the Hadza may have different hormonal profiles than people in Western
populations, perhaps because of greater exposure to energetic stress.
Given these results, it is clearly important to test for relationships between 2D:4D
and hunting success in an actual forager population who live in a subsistence environ-
ment more representative of pre-agricultural populations than do participants drawn
from Western, industrialized societies. Indeed, Longman et al. concluded that “further
work is now required to test [for a relationship between 2D:4D and hunting success] in
hunting societies”(2015:9). Despite this, no other studies have, to date, searched for an
association between digit ratio and hunting success among either the Hadza or, to the
author’s knowledge, any other hunter-gatherer group. Stibbard-Hawkes et al. (2018)
recently introduced a novel measure of hunting reputation which acted as a viable
proxy of hunting skills. I use this metric, as well as two further indices of hunting
ability (upper body strength and aim), along with visual acuity, to address this question.
Materials and Methods
The Hadza are an ethnolinguistic group living in the Lake Eyasi region of northern
Tanzania. Around 250 Hadza hunt and gather for the great majority of their diet: >95%
by some estimates (Wood and Marlowe 2013), although it appears probable, from
personal observation, that this proportion has decreased in recent years, following an
increased reliance on grain and other cultigens. Like many other forager groups (Bliege
Bird and Bird 2008; Marlowe 2007), the Hadza have a strong sexual division of labor:
hunting and honey-collecting among the Hadza are predominantly male activities,
while women gather the majority of tubers, berries and baobab fruit. This division of
labor appears at an early age (Crittenden et al. 2013; Froehle et al. 2018; Lew-Levy
et al. 2019). The great majority of Hadza men’sforagingtrips(∼89%) are solitary
(Berbesque et al. 2016). Men hunt using bows and poisoned arrows and generally
ambush and shoot, then follow prey until they succumb to poison-induced cardiac
arrest (Marlowe 2010). The Hadza, like many foragers, never practice endurance
hunting. They have no cause to, partly because of to the efficacy of weapon poisons
and because the Eyasi region is shrubland (Marlowe 2010) with low visibility and
much cover. The Hadza exhibit a high degree of food-sharing and in some datasets
(Hawkes et al. 1991), though not others (Wood and Marlowe 2013; cf. Hawkes et al.
2014), a hunter and his family receive no more of the food he brings back than does any
other person living in the same camp. As with other hunter-gatherer populations, foods
collected by men are both more temporally variable (i.e., unreliably attained) and more
widely shared (Hawkes et al. 1991) than foods collected by women. Hadza men are
more willing than Hadza women to take on risk in experimental settings also (Apicella
et al. 2017). Hadza camps, usually composed of 20–30 people (range = 6–139), are
ephemeral, and most individuals move camp on average 6.5 times per year (Marlowe
2010). The Hadza show no assortative mating preferences for body height, weight, or
grip strength (Sear and Marlowe 2009), although both men and women cite the
importance of character, physical appearance, and foraging ability in a potential spouse
(Marlowe 2004b). Better hunters are also preferred as campmates (Smith and Apicella
2019; Wood 2006). As among many other forager populations (Dyble et al. 2015),
Hadza marital residence is multilocal (Marlowe 2004a); in-settlement relatedness is low
relative to most unilocal populations, and a mean of 27% of camp members share at
least one great-grandparent, only 16% higher than chance (Blurton Jones 2016:101).
Both Hadza women and men are afforded almost complete autonomy over who they
choose to marry or whether to divorce (Marlowe 2010).
In this study, I collected hunting reputation measures for 71 hunters, all male, aged
between 17 and 75 (mean = 39). Of these, 70 provided digit ratio data, 68 provided bow
pull strength data, and 64 participated in an archery contest. Most Hadza are nonliterate.
Accordingly, I explained at the beginning of each camp visit what would be involved in
each measurement. I explained, in lay terms, the purpose of each measurement. Hunters
were drawn from 17 camps during three separate field trips (17 August–17 September
2013; 7 December 2013–6 January 2014; 19 October–27 November 2014). On the last
of these trips, with the help of two research assistants, I conducted hunting reputation
interviews. The 67 interviewees whose answers I included in the final sample were of
both sexes (m = 36, f = 31). Participants were remunerated with gifts: shoes, blankets,
soap, petroleum jelly, plates, hammers, cold chisels, and other useful items. Some
interviewees in the Mangola region, where local laws required it, were remunerated
with money, equivalent to ∼GBP (or ) Springer: choose one or the other 6 per camp
visit and shared between all camp members (including nonparticipants). I assured
people that they were free not to participate in any measures and to drop out of the
study at any time. I also made it clear that those who decided not to participate in some
measurements would still receive gifts when I left camp.
Hunting Reputation Measure
Reputation data interviews were conducted in the seclusion of a field vehicle, where
participants’answers could not be overheard. I showed each participant a high-
resolution face-on photograph of each one of the 71 hunters in my sample. I asked
them to provide the hunter’s first name, his father’s name, and the length of time that
they last lived in the same camp with the interviewee. A hunter’s renown was defined
as the number of interviewees who knew both that hunter’snameandhisfather’sname.
In order to ensure that interviewees were familiar with the hunters they were ranking,
photographs were removed if the interviewee did not answer the first two questions
correctly, or had not lived with the photographed hunter within the previous two years.
Next, I set out the remaining photographs in a random order and asked interviewees to
remove the photograph of the best hunter. The photograph was removed, and the
process repeated, until each interviewee had provided a ranked list of every hunter
they knew in the sample, ordered from best to worst.
These ranked lists were then collated using the methods set out by Stibbard-Hawkes
et al. (2018). I took the proportional rank of each hunter in each list (i.e., the fraction of
the way up each list that each hunter appeared—for example, a hunter halfway up a list
of 20 would score 0.5). I then took the mean of these scores for each of the 71 hunters,
for each list in which that hunter appeared, giving me an aggregated index of “hunting
reputation”for each of the 71 hunters in my sample. The measure is an aggregation of
ordinal rank data, although it behaves and is treated as a continuous variable. A
mathematical formalization of the procedure is provided in ESM §1.
This reputation measure showed a high degree of internal consistency, and it
significantly predicted skill on three hunting tasks as set out further by Stibbard-
Hawkes et al. (2018) and in ESM §2. For this reason, aggregated reputation data
appear to act as a serviceable proximate measure of true hunting skill (i.e., true skill at
finding and killing wild animals). Furthermore, and unlike previously used reputation-
based measures (e.g., Apicella 2014; Blurton Jones and Marlowe 2002; Marlowe
2000), the current method allowed fine-grained distinctions to be made between
hunters at all levels of ability across camps.
I measured the length of the index finger (second digit) and ring finger (fourth digit) on
first the right hand and then the left hand from the center of the basal crease to the distal
tip of the finger using battery-operated digital callipers. Digit ratios may be more
accurately collected using a scanner (Kemper and Schwerdtfeger 2009), but callipers
have been elsewhere shown to be sufficiently reliable (Voracek et al. 2007), have been
successfully implemented in previous Hadza studies (Apicella et al. 2016), and were
deemed the most practicable method under field conditions. It should be noted,
however, that direct measures often yield higher ratios than indirect measures
(Ribeiro et al. 2016). Due to time constraints, digit lengths were measured only once
for each participant by a single observer. This constitutes a limitation.
A participant’s digit ratio was calculated by dividing the length of the second finger
by the length of the fourth finger for both hands. Where participants had sustained
injury to either hand, measures from that hand were excluded from the final analysis.
Hunting Ability Measures
Bow Pull Strength Bow pull strength was measured using an Easton Digital bow pull
scale. Hunters hooked the scale to the string of their own bow and drew the bow at peak
strength for fifteen seconds with their dominant arm. To allow direct comparison with
previous studies (e.g., Blurton Jones and Marlowe 2002), weight at peak pull was
recorded and reported in pounds (lbs).
Bow Aim I measured aim in an archery contest. Following the methods of Blurton
Jones and Marlowe (2002), I constructed a 61 by 61 cm cardboard target with an
opaque outer circle of 33 cm in diameter, a transparent inner circle of 20 cm in
diameter, and an opaque “bullseye”of 4 cm in diameter. The target was attached to a
tree or other solid structure at shoulder height. Care was taken to ensure that terrain was
reasonably flat, and that view of the target was unobstructed by grass or foliage.
Although it was impossible, in field conditions, to completely standardize wind speed,
archery contests were conducted only at a Beaufort wind force of two or less, using the
Beaufort wind scale. Participants fired at the target using their own bows and arrows
from distances of 10, 20, and 30 m. They took three shots from each distance, nine
shots in total. Hits were scored at 100 for a bullseye, 50 on the inner ring and 25 on the
outer ring. It would have been ideal to conduct these tests without observers, although
given the excitement each archery contest generated, this proved impracticable.
I measured visual acuity for each eye using a 3 m Landolt C Optotype. This does not
require literacy or knowledge of the Latin alphabet. Since tests were conducted outside
in sunlight, I could not completely standardize lighting conditions between different
camps. However, I always ensured that the chart was clearly illuminated, free from
glare, and that the sun was behind the participant. Results were recorded as LogMAR
scores, and visual acuity in the best eye was used in the final analysis.
Research Clearance, Data Security and Availability
Research was approved by Cambridge Biological Research Ethics Committee and
conducted with permission from the Tanzanian Commission for Science and Technol-
ogy (COSTECH Permits: 2013–271-ER-2000-80 and 2014–317-ER2000–80).
Clearance to conduct research was also attained at the local level and from all
participants. All data were stored on a password-protected hard-drive encrypted using
Apple Firevault and anonymized using ID numbers known only to a small group of
Hadza researchers. Certain participants are known by their hunting reputations. For this
reason, to protect participant’s anonymity, I have not made hunting reputation, skill, or
age data freely available online. However, finger lengths and associated digit ratios,
which hold no value as individual identifiers, are available online at https://osf.io/ejtx8/.
Mean Hadza digit ratio in the current sample was 1.00 for right-hand measures (n=69,
SD = 0.04, range = 0.93–1.09) and also 1.00 for left-hand measures (n=69,SD=0.05,
range = 0.88–1.12). Neither departed significantly from the expectation of a normal
distribution in a Shapiro-Wilk test of normality (right hand, W=0.98, p= 0.22; left
hand, W=0.98,p= 0.45; Fig. 1). Descriptive statistics are provided for each variable in
the study in Table 1. As expected, in both hands, second- and fourth-finger lengths
showed a significant, strong positive linear relationship. Right second-finger length
increased 0.85 cm for each 1 cm increase in right fourth-finger length (p<0.00; R2=
0.70) while left second-finger length increased by 0.78 cm for each 1 cm increase in left
fourth-finger length (p<0.00; R2= 0.61). The regression lines for both hands had non-
zero intercepts (Fig. 2), and intercepts were significantly higher than zero for both right
(b=9.65, p= 0.03) and left hands (b=14.45, p= 0.01), implying that 2D:4D ratio
decreases as digit length increases in large samples. However, there was no significant
relationship between average digit length and digit ratio in the current dataset for either
the right-hand (F1,67 =0.05, R
2=0.00, p= 0.82) or left-hand (F1,67 =0.00, R
p= 0.98) measure.
0.90 0.95 1.00 1.05 1.10
Right Hand Digit Ratio (2D:4D)
0.90 0.95 1.00 1.05 1.10
Left Hand Digit Ratio (2D:4D)
Fig. 1 Histograms of right (left) and left (right) hand 2D:4D ratio for 67 Hadza men
Hunting reputation showed no significant relationship with digit ratio for either
hand. Although the slope of the regression line was in the expected direction for both
the right and the left hand, the value of R2was near zero in both cases (R = 0.01, 95%
CI = −0.11, 0.02; L = 0.03, 95% CI = −0.15, 0.01). Consistent with findings from many
other populations (e.g., Koster et al. 2019; von Rueden et al. 2008), hunting reputation
showed a quadratic relationship with age. It peaked at roughly 45 and gradually
declined thereafter. For this reason, and following the recommendations of Blurton
Jones (2016), I included both age and age2as controls. The effect of 2D:4D ratio did
Table 1 Descriptive statistics for each variable
Va r i ab l e NMean SD Median Min Max Range
Right Ring Finger (D4) Length 69 64.93 4.75 64.95 53.85 80.32 26.47
Right Index Finger (D2) Length 69 65.04 4.85 64.83 55.28 79.71 24.43
Right 2D:4D 69 1 0.04 1 0.93 1.09 0.16
Left Ring Finger (D4) Length 69 65.51 4.82 65.21 55.66 79.36 23.7
Left Index Finger (D2) Length 69 65.31 4.80 64.9 51.14 80.7 29.56
Left 2D:4D 69 1 0.05 1 0.88 1.12 0.25
Age 71 39.18 14.81 38 17 75 58
Aim 65 143.46 74.91 150 0 350 350
Bow Pull Strength (KG) 67 61.61 16.24 60.8 18.88 92.8 73.92
Best Eye Visual Acuity (LogMar) 70 −0.03 0.23 −0.1 −0.3 0.8 62
Hunting Score 71 0.47 0.17 0.45 0.18 0.88 0.7
Renown (n= 89) 71 53.63 16.69 53 22 84 62
55 60 65 70 75 80
Right Hand 4th Digit Length (mm)
Right Hand 2nd Digit Length (mm)
55 60 65 70 75 80
Left Hand 4th Digit Length (mm)
Left Hand 2nd Digit Length (mm)
Fig. 2 Scatterplot of ring finger length by index finger length for all hunters in both right (left) and left (right)
hand measures, overlaid with a simple linear regression line and 95% confidence band
not approach significance whether or not age and age2were included as controls
(Table 2), nor when including hunter’s“renown”as a control (see ESM §3).
I also tested for a relationship between digit ratio and three other variables: draw
strength, best-eye visual acuity, and aim with a bow and arrow. In a simple linear
regression, there was no significant relationship between either right- (F1,63 =0.00,
R2=0.00, p= 0.60) or left-hand (F1,63 =0.00, R
2=0.00,p= 0.99) digit ratio and draw
strength. I calculated 95% confidence intervals for R2,whichwere−0.09 and 0.03 for
the right-hand measure and −0.06 and 0.06 for the left-hand measure. Similarly, I
found no significant relationship between right- (F1,61 =0.55, R
2=0.01, p=0.46) or
left-hand (F1,61 =2.60, R
2=0.04, p= 0.11) digit ratio and aim with a bow and arrow.
The R295% confidence intervals were −0.03 and 0.11 for the right-hand measure and
−0.18 and 0.00 for the left-hand measure. Furthermore, I found no significant rela-
tionship between either right- (F1,66 =1.7,R
2=0.01,p= 0.20) or left-hand (F1,66 =0.04,
R2=0.001, p= 0.84) digit ratio and visual acuity. The R295% confidence intervals
Table 2 Regression Models of Mean Hunting Score and 2D:4D Ratio with and without Age and Age2,first
for right hand, then for left hand measures. Regression coefficients are reported in their natural units and as
Model B SE B βR2Adjusted R2p
1.1 0.011 −0.004 0.385
Intercept 0.895 0.489 0.072
Right Hand 2D:4D −0.426 0.488 −0.106 0.385
1.2 0.064 0.036 0.111
Intercept 0.829 0.480 0.089
Right Hand 2D:4D −0.463 0.478 −0.115 0.335
Age 0.003 0.001 0.230 0.057
1.3 0.103 0.062 0.068
Intercept 0.452 0.526 0.393
Right Hand 2D:4D −0.321 0.480 −0.080 0.506
Age 0.016 0.008 1.344 0.051
Age2−0.0002 0.0001 −1.133 0.010
2.1 0.028 0.013 0.171
Intercept 1.046 0.419 0.015
Left Hand 2D:4D −0.580 0.419 −0.167 0.171
2.2 0.073 0.045 0.082
Intercept 0.949 0.416 0.026
Left Hand 2D:4D −0.577 0.413 −0.166 0.166
Age 0.002 0.001 0.213 0.213
2.3 0.120 0.079 0.040
Intercept 0.637 0.442 0.154
Left Hand 2D:4D −0.517 0.407 −0.148 0.208
Age 0.016 0.008 1.411 0.036
Age2−0.0002 0.0001 −1.218 0.067
were −0.01 and 0.15 for the right-hand measure and −0.07 and 0.05 for the left-hand
measure. Scatterplots are provided in Fig. 3.
Analogous analyses, accompanied by a widely applicable information criterion
(WAIC) model selection, were also conducted using a Bayesian framework. These
analyses yielded similar results to those reported here and did not provide any more
certainty about the relationships between the variables under investigation. Since
frequentist analyses are still both widely used and more widely understood, I have
opted to report frequentist results in the body of the article. Results of the Bayesian
reanalysis are reported in ESM §4.
Hadza Digit Ratio Compared with Previous Hadza Studies and Other Populations
Mean male Hadza digit ratios in the current study, 1.00 for both right- and left-hand
measures, are higher but almost within the range (0.97–0.99) of mean male digit ratios
in most other Hadza datasets (Apicella et al. 2016;Butovskayaetal.2012,2015)and
notably higher than those reported by Butovskaya et al. (2010) (R and L means = 0.95).
This is not because every hunter had an even digit ratio, but because participants had
digit ratios both above and below 1.0 (e.g., right-hand range = 0.93–1.09). These results
are dissimilar to those often found in Western populations (e.g., means: men = 0.95,
women = 0.97 among undergraduates from the University of Pittsburgh, reported by
Puts et al. 2004) and higher than the upper ranges of male digit ratio reported in a recent
review by Ribeiro et al. (2016) (i.e., 0.98 for right-hand measures in both US and Saudi
populations). Since this study was concerned with hunting, a typically male activity
among the Hadza, I did not collect any data on female digit ratio for comparison. For
this reason, although current data are consistent with Apicella et al.’s(2016)finding
that Hadza men do not have significantly lower digit ratios than women, I lack the data
to replicate these findings in the current study.
It is unclear why Hadza men’s digit ratios are higher than those reported for men in
many other study populations (e.g., Puts et al. 2004). This is relevant to the current
question because, drawing on the method of Longman et al. (2015), digit ratios are
treated in this study as a proxy measure of developmental androgen exposure. It is
possible that, as Apicella et al. have suggested, Hadza men have different hormonal
profiles than those in other populations, and the suppression of testosterone might be
advantageous under conditions of energetic stress, “reducing muscle composition and
the metabolic requirements of its maintenance”(2016:6). A study by Muller et al.
(2009), to my knowledge, contains the only published measure of testosterone levels
among the Hadza. Their results indicate that Hadza adult male salivary testosterone
may indeed be lower than is found in Western populations (pmol l−1151 among the
Hadza vs. ≥pmol l−1250 in US study populations). However, their Hadza sample size,
at 27 individuals, was small.
This question may be fertile grounds for further enquiry. Butovskaya et al. (2015)
compared digit ratios between Hadza foragers and Datoga pastoralists and reported
higher mean digit ratios among the Hadza than among the Datoga (Hadza R = 0.97,
L = 0.98 vs. Dotaga R = 0.96, L = 0.96). Although the Datoga live in the same
0.95 1.00 1.05
Right Hand Digit Ratio
Hunting Reputation Score
0.90 0.95 1.00 1.05 1.10
Left Hand Digit Ratio
Hunting Reputation Score
0.95 1.00 1.05
Right Hand Digit Ratio
Bow Pull Strength (KG)
0.90 0.95 1.00 1.05 1.10
Left Hand Digit Ratio
Bow Pull Strength (KG)
0.95 1.00 1.05
Right Hand Digit Ratio
0.90 0.95 1.00 1.05 1.10
Left Hand Digit Ratio
0.95 1.00 1.05
Right Hand Digit Ratio
Best Eye Visual Acuity (LogMar)
0.90 0.95 1.00 1.05 1.10
Left Hand Digit Ratio
Best Eye Visual Acuity (LogMar)
geographic region as the Hadza, they have much higher rates of polgygny, less paternal
care, and very different subsistence practices (Butovskaya et al. 2015;Mulleretal.
2009). However, Muller et al. (2009) have previously shown that the difference
between both morning and evening salivary testosterone levels from Hadza and Datoga
men was nonsignificant, and both had greatly lower salivary testosterone levels than is
generally observed in adult men from Western (US) populations analyzed in the same
Digit ratios have been reported from at least one other population who have
traditionally practiced hunting and gathering. Pettigrew et al. (2017)reportedthatmean
digit ratios directly measured among the San showed strong sexual dimorphism (Male
R = 0.96, L = 0.95 vs. Female R = 1.02, L = 1.03). San men in this sample had digit
ratios comparable to or lower than those often reported in Western men (Pettigrew et al.
2017). In a separate study, Worthman and Konner (1987) showed that blood plasma
testosterone measures among San hunters were not significantly different to Western
men, a different pattern to that observed by Muller et al. (2009) among the Hadza.
However, since Pettigrew et al. (2017)andWorthmanandKonner(1987)reportedtheir
results 30 years apart, working with different individuals, the grounds for comparison
are limited. Overall it is difficult to draw conclusions from these data, and more cross-
cultural comparative research examining digit ratios in a range of subsistence environ-
ments is warranted.
Digit Ratio and Digit Length
Kratochvíl and Flegr (2009)andLollietal.(2017) have proposed that sexual
dimorphism in digit ratio may not be a direct consequence of differences in
developmental androgen levels between men and women. They have instead
argued that sexual dimorphism in digit ratio may be the consequence of a
nonlinear scaling relationship between digit ratio and absolute digit length. In a
sample of 297 Czech biology students, Kratochvíl and Flegr (2009)notedthat
the relationship between 2D and 4D length had a positive intercept, implying a
decrease in the ratio between finger lengths as absolute finger lengths increase.
This, they argued, might account for patterns of sexual dimorphism in digit
ratio. Lolli et al. (2017) took this idea further and, in a study of 154 men and
262 women, also concluded that an allometric, nonlinear size-scaling relation-
ship was the most plausible model for normalizing 2D to 4D lengths. More-
over, after normalizing for scaling relationships, they found that women’s digit
ratios were actually lower than men’s.
The relationship between the length of the second finger and the length of
the fourth finger in the current dataset shows a very similar pattern to that
reported by Kratochvíl and Flegr (2009). As in Kratochvíl and Flegr’sstudy,
the intercept of the regression line of left- and right-hand second and fourth
digit lengths was positive, implying that digit ratio decreases as finger length
increases (Fig. 2). Hadza men in the current study have lower absolute mean
digit lengths (2D = 65 mm, 4D = 65 mm) than men in Western populations (e.g.,
RFig. 3 Scatterplots of Hunting Reputation Score, Bow Pull Strength, Aim, and Visual Acuity by right- and
left-hand digit ratio, overlaid with simple linear regression lines and 95% confidence bands
4D = 77.46 mm and 2D = 73.04 mm),
suggesting that the relatively high digit
ratio observed in the study population may be a function of allometric scaling.
This evidence may cast doubt on claims that digit ratio is a direct consequence of
developmental androgen exposure. This relationship may, instead, be confounded in
many populations by sex difference in absolute finger length. However, the validity of
this interpretation is in turn unclear for at least five reasons.
First, there is no significant relationship between mean digit length and digit ratio
within the range of finger lengths measured in the current sample and, as with those
data presented by Kratochvíl and Flegr (2009), the assumed allometric relationship
between finger lengths and finger ratio is an implication of the slope of the regression
line. It is not a verifiable pattern in the current dataset. This pattern may be the result of
normal levels of error, rather than genuine evidence of allometry (Forstmeier 2011).
Second, the current data are from adult men only. I collected no data from women,
nor from those under the age of 16, and there are no data on either Hadza fetal androgen
levels or fetal digit ratios. It is, therefore, impossible to say whether the same finger-
length scaling relationship exists at all ages, and in other populations (Galis et al. 2010;
Hönekopp and Watson 2010), this appears not to be the case. A study of the digit ratios
of Hadza children, and perhaps including the Hadza women’s digit ratios gathered by
Apicella et al. (2016), would shed further light on this question.
Third, although they cautioned against using ratios where scaling relationships are
evident, in a reanalysis of Kratochvíl and Flegr’s data, Forstmeier concluded that
“human digits are probably very close to being isometric”and are “fairly independent
of absolute size”(2011:1860, although see Forstmeier 2018).
Fourth, and perhaps most problematic, is the fact that much recent evidence from
Western populations shows an absence of allometry effects on digit ratio. Galis et al.
(2010), in a study of fetuses, showed that sex differentiation in 2D:4D appears as early
as 14 weeks. Manning and Fink (2018) found that, although digit length increased and
sexual dimorphism in digit length changed across development, digit ratio was not
wholly age dependent. Furthermore Lombardo et al. (2018) found that throwing
performance among college-aged women was negatively related to digit ratio indepen-
dent of size-scaling effects.
Fifth and finally, Hadza digit ratios were higher than those reported among San
foragers (Pettigrew et al. 2017), even though San men had mean absolute second digit
lengths comparable to those of Hadza men in the current sample (e.g., Hadza R 2D =
65.04 mm, 4D = 64.93 mm, San R 2D = 65.15 mm, 4D = 67.7 mm).
Therefore, although the digit ratios of Hadza men reported here and elsewhere
(Apicella et al. 2016) are higher than in many other populations, this fact may not be
explained by size-scaling relationships.
Digit Ratio, Hunting Reputation, and Hunting Skill
In the current study there was no significant relationship between digit ratio for either
hand and hunting reputation. Furthermore, there was no significant relationship be-
tween digit ratio for either hand and either aim with a bow or the bow pull measure of
Statistics from a study of 849 Californian men published by Lippa (2003:183), one of the few studies to
report absolute finger lengths.
strength. Eyesight also showed no relationship to digit ratio in either hand. However,
although eyesight has been proposed as important to hunting success (e.g., Blurton
Jones 2016), Stibbard-Hawkes et al. (2018) found that visual acuity was unrelated to
both archery skill and hunting reputation within the normal range of human vision. The
analogous reanalysis using Bayesian methods (ESM §4) yielded comparable results—
relationships between digit ratio and the other variables under investigation were weak.
Only one of the models was substantially preferred to the mean of the outcome variable
(the null model) in a WAIC model selection: bow aim was negatively related to left-
hand digit ratio. However, right-hand digit ratio showed a positive relationship to bow
aim, muddying the results.
Such results are difficult to interpret, and current findings leave open several
possibilities. It is possible that, among the Hadza, hunting might not, as Longman
et al. have suggested, act as either “a marker of testosterone exposure”(2015:3) or “a
signal of reproductive potential”(2015:6). It is further possible that digit ratio is
unrelated, among Hadza men, to developmental androgen exposure or that develop-
mental androgen exposure is unrelated to hunting success. It is finally possible that the
current sample size was not large enough to detect genuine but weak associations
between digit ratio and hunting ability, an interpretation consistent with the small effect
sizes often observed in digit ratio research. These possibilities I discuss further.
As set out in the introduction, Longman et al. (2015) argue that hunting ability
should be negatively associated with digit ratio. Their argument is based on the
following assumptions: (a) Hunting acts as a signal of athletic ability. (b)Athletic
ability is positively related to testosterone. (c) Adult testosterone is positively related to
fertility or other measures of “reproductive fitness.”(d) Adult testosterone is, further-
more, positively related to developmental androgen exposure. (e) 2D:4D ratio is a
marker of developmental androgen exposure.
Longman et al. (2015) present a logical argument, supported by data showing a
significant negative relationship between endurance running performance and digit
ratio. However, the model is complex and relies on a long chain of relationships
between variables. As a consequence, if any one of these associations is not valid or
does not apply in a particular case, the entire model is called into question. If, for
example, Hadza hunting reputation does not reliably reflect athletic ability, a negative
relationship between hunting reputation and digit ratio should not be expected. If digit
ratio is not negatively related to adult testosterone levels, as suggested by Hönekopp
et al. (2007), a negative relationship between digit ratio and reproductive outcomes
might not be expected. Furthermore, if the negative relationship between developmen-
tal androgen exposure and digit ratio is not universal (e.g., Apicella et al. 2016), a
relationship between athletic ability and digit ratio also should not be expected.
Similarly, the relationships between testosterone levels and fertility or other heritable
measures of “reproductive fitness”are not straightforward (e.g., Scott et al. 2012),
further complicating Longman et al.’smodel.
It is difficult to tell which assumption, if any, is incorrect or inapplicable to the
Hadza case. Furthermore, even if all the assumptions of the model are valid, if one or
more of the associations is, in general, weak, then the chain of associations may be
prone to weak or nonsignificant results overall. This may further account for the
absence of a significant association between digit ratio and hunting reputation in the
There are two further difficulties in interpreting current results. Hunting reputation
here is an indirect proxy measure of true hunting skill. The measure is an aggregation of
ordinal rank data, although it behaves and is treated as a continuous variable. Because
of the significant associations found by Stibbard-Hawkes et al. (2018) between hunting
reputation and strength, aim, and ecological knowledge, it is likely that hunting
reputation, to a certain extent, reflects athletic ability and is externally valid. However,
much of the variance in hunting reputation remains unaccounted for, and although
participants showed high levels of agreement concerning the hunting ability of those
they were rating, some error has probably been introduced in aggregating ranked lists
with unequal numbers of cases. For this reason, any associations that exist between true
hunting ability and digit ratio may not be captured in the current measure of hunting
reputation. I here addressed this issue by relating 2D:4D ratio to two other component
hunting skills (a measure of strength and a measure of aim with a bow and arrow) as
well as eyesight. None of these variables show a significant relationship with 2D:4D
ratio in either direction. In the Bayesian analysis (ESM §4), left-hand digit ratio did
negatively predict aim (the expected direction), although conversely right-hand digit
ratio showed a positive relationship to aim.
The lack of significant associations in the current study could also be a consequence
of sample size. Those associations that do exist between digit ratio and “spatial ability,”
strength, sporting achievement, voice pitch, age-specific reproductive success, and so
on, in other populations usually have small or modest effect sizes. The sample sizes in
the current study (n=63–67) are low—a fraction of those used elsewhere (e.g.,
Manning and Fink 2008). It is possible that, if there were small or even moderate
associations between digit ratio and hunting success, they would not be apparent in a
sample of this size. This is, perhaps, grounds for conducting a further study with a
larger sample size. However, given that fewer than 150 Hadza men are still hunting for
the majority of their diet, the current sample includes a reasonable proportion of the
total population. Even a study that included the entire population of hunters would
probably yield similar results. This is especially likely given that, in the current study,
associations between 2D:4D ratio and other variables were close to zero and did not
approach significance, as well as the fact that 95% confidence intervals suggested even
the highest probable value of R2was still low (0.15)—a finding mirrored in the
Bayesian reanalysis (ESM §4).
In summary, results from the current study fail to support the prediction that digit ratio
should be related to hunting ability (Longman et al. 2015). Longman et al.’smodelis
based on a large number of assumptions, and if the results do reflect a genuine absence
of relationship between digit ratio and hunting skill, it is unclear which of Longman
et al.’s assumptions may not apply in the Hadza case. Due to the uncertainty inherent in
the reported results, it is further unclear whether there is a genuine absence of
relationship between the variables under study, or whether this is a consequence of
the small sample sizes typical of forager research, the small effect sizes commonplace
in digit ratio research, or, alternatively, noise and error in the variables used. Given the
small size of the adult male Hadza population, and some unavoidable imprecision in
hunting success measures (e.g., Hill and Hurtado 2009; Stibbard-Hawkes et al. 2018,
further research is sadly unlikely to produce more conclusive findings.
This study confirms that Hadza male digit ratios are higher than those generally
observed in Western populations (see Apicella et al. 2016;Butovskayaetal.2015;but
see Butovskaya et al. 2010) and among the San (Pettigrew et al. 2017). Results
corroborate previous evidence that Hadza digit ratios show little (Butovskaya et al.
2015), no, or reversed (Apicella et al. 2016) sexual dimorphism. Hadza men also have
shorter absolute finger lengths than men from many Western populations. However, the
possibility that high Hadza male digit ratios are the result of a size-scaling relationship
between relative second- and fourth-digit lengths (Kratochvíl and Flegr 2009; Lolli
et al. 2017) is called into doubt by the fact that San men have comparable second-digit
lengths but lower digit ratios than Hadza men (Pettigrew et al. 2017).
Whether or not there is an influence of allometry on 2D:4D ratio, current results
provide reason for caution in generalizing findings from Western, educated, industrial-
ized populations to other groups with differing diets, levels of pathogen exposure,
activity budgets, and modes of subsistence. This is especially important given that
findings from Western populations do not uniformly match findings from non-Western
populations in many other types of investigation (see Henrich et al. 2010). The current
study finds great uncertainty and no clear support for the idea that 2D:4D ratio and
hunting abilities are associated in a hunter-gatherer population. Results suggest that, if
such a relationship does exist, it is unlikely to be a strong one.
Acknowledgments Thanks to all those who have provided support throughout this project. Special thanks
to Frank Marlowe, Robert Attenborough, Jeremy Kendall, Coren Apicella, Sally Street, Enrico Crema, Trevor
Hawkes, and Charles Endeko. Thanks also to all those who participated in the project. Further thanks to
Bernhard Fink, Martin Voracek, and one anonymous reviewer for the help they have provided in revising and
improving this manuscript.
This project was funded with generous support from the Leakey Foundation, Robinson College Cambridge,
the Smuts Memorial Fund, the Ruggles-Gates Fund of the Royal Anthropological Institute, the Cambridge
Department of Archaeology and Anthropology, the Cambridge Centre for African Studies, the Anthony
Wilkin Fund, and the Ridgeway-Venn Fund.
Apicella, C. L. (2014). Upper-body strength predicts hunting reputation and reproductive success in Hadza
hunter-gatherers. Evolution and Human Behavior, 35(6), 508–518.
Apicella, C. L., Tobolsky, V. A., Marlowe, F. W., & Miller, K. W. (2016). Hadza hunter-gatherer men do not
have more masculine digit ratios (2D:4D). American Journal of Physical Anthropology, 159(2), 223–232.
Apicella, C. L., Crittenden, A. N., & Tobolsky, V. A. (2017). Hunter-gatherer males are more risk-seeking than
females, even in late childhood. Evolution and Human Behavior, 38(5), 592–603.
Bennett, M., Manning, J. T., Cook, C. J., & Kilduff, L. P. (2010). Digit ratio (2D:4D) and performance in elite
rugby players. Journal of Sports Sciences, 28(13), 1415–1421.
Berbesque, J. C., Wood, B. M., Crittenden, A. N., Mabulla, A., & Marlowe, F. W. (2016). Eat first, share later:
Hadza hunter-gatherer men consume more while foraging than in central places. Evolution and Human
Behavior, 37(4), 1–6.
Bliege Bird, R., & Bird, D. W. (2008). Why women hunt: Risk and contemporary foraging in a Western Desert
Aboriginal community. Current Anthropology, 49(4), 655–693.
Blurton Jones, N. G. (2016). Demography and evolutionary ecology of Hadza hunter-gatherers. Cambridge:
Cambridge University Press.
Blurton Jones, N. G., & Marlowe, F. W. (2002). Selection for delayed maturity. Human Nature, 13(2), 199–
Blurton Jones, N. G., Hawkes, K., & O’Connell, J. F. (2002). Antiquity of postreproductive life: Are there
modern impacts on hunter-gatherer postreproductive life spans? American Journal of Human Biology,
Brosnan, M. J. (2008). Digit ratio as an indicator of numeracy relative to literacy in 7-year-old British
schoolchildren. British Journal of Psychology, 99(1), 75–85.
Brown, W. M., Finn, C. J., & Breedlove, S. M. (2002a). Sexual dimorphism in digit-length ratios of laboratory
mice. Anatomical Record, 267(3), 231–234.
Brown, W. M., Hines, M., Fane, B. A., & Breedlove, S. M. (2002b). Masculinized finger length patterns in
human males and females with congenital adrenal hyperplasia. Hormones and Behavior, 42(4), 380–386.
Brown, K. S., Marean, C. W., Jacobs, Z., Schoville, B. J., Oestmo, S., Fisher, E. C., Bernatchez, J., Karkanas,
P., & Matthews, T. (2012). An early and enduring advanced technology originating 71,000 years ago in
South Africa. Nature, 491(7425), 590–593.
Burley, N. T., & Foster, V. S. (2004). Digit ratio varies with sex, egg order and strength of mate preference in
zebra finches. Proceedings of the Royal Society B: Biological Sciences, 271(1536), 239–244.
Butovskaya, M. L., Burkova, V., & Mabulla, A. Z. (2010). Sex differences in 2D:4D ratio, aggression and
conflict resolution in African children and adolescents: A cross-cultural study. JournalofAggression,
Conflict and Peace Research, 2(1), 17–31.
Butovskaya, M. L., Vasilyev, V. A., Lazebny, O. E., Burkova, V. N., Kulikov, A. M., Mabulla, A., Shibalev, D.
V., & Ryskov, A. P. (2012). Aggression, digit ratio, and variation in the androgen receptor, serotonin
transporter, and dopamine D4 receptor genes in African foragers: The Hadza. Behavior Genetics, 42(4),
Butovskaya, M., Burkova, V., Karelin, D., & Fink, B. (2015). Digit ratio (2D:4D), aggression, and dominance
in the Hadza and the Datoga of Tanzania. American Journal of Human Biology, 27(5), 620–627.
Cain, K. E., Bergeon Burns, C. M., & Ketterson, E. D. (2013). Testosterone production, sexually dimorphic
morphology, and digit ratio in the dark-eyed junco. Behavioral Ecology, 24(2), 462–469.
Crittenden, A. N., Conklin-Brittain, N. L., Zes, D. A., Schoeninger, M. J., & Marlowe, F. W. (2013). Juvenile
foraging among the Hadza: Implications for human life history. Evolution and Human Behavior, 34(4),
Direnzo, G. V., & Stynoski, J. L. (2012). Patterns of second-to-fourth digit length ratios (2D:4D) in two
species of frogs and two species of lizards at La Selva, Costa Rica. Anatomical Record, 295(4), 597–603.
Dunbar, R. I. M., & Wlodarski, R. (2016). Reply to Falcon. Biology Letters, 12,3–4.
Dyble, M., Salali, G. D., Chaudhary, N., Page, A., Smith, D., Thompson, J., Vinicius, L., Mace, R., &
Migliano, A. B. (2015). Sex equality can explain the unique social structure of hunter-gatherer bands.
Science, 348(6236), 769–798.
Falcon, R. G. (2016). Stay, stray or something in-between? A comment on Wlodarski et al. Biology Letters,
Forstmeier, W. (2011). Women have relatively larger brains than men: A comment on the misuse of general
linear models in the study of sexual dimorphism. The Anatomical Record, 294, 1856–1863.
Forstmeier, W. (2018). Avoiding misinterpretation of regression lines in allometry: Is sexual dimorphism in
digit ratio spurious? bioRxiv, 298786. https://doi.org/10.1101/298786.
Froehle, A. W., Wells, G. K., Pollom, T. R., Mabulla, A. Z., Lew-Levy, S., & Crittenden, A. N. (2018).
Physical activity and time budgets of Hadza forager children: Implications for self-provisioning and the
ontogeny of the sexual division of labor. American Journal of Human Biology, 31(1), 1–13.
Galis, F., Ten Broek, C. M. A., Van Dongen, S., & Wijnaendts, L. C. D. (2010). Sexual dimorphism in the
prenatal digit ratio (2D:4D). Archives of Sexual Behavior, 39(1), 57–62.
Hawkes, K., & Bird, R. B. (2002). Showing off, handicap signaling, and the evolution of men’swork.
Evolutionary Anthropology, 11,58–67.
Hawkes, K., O’Connell, J. F., & Blurton Jones, N. G. (1991). Hunting income patterns among the Hadza: Big
game, common goods, foraging goals and the evolution of the human diet. Philosophical Transactions of
the Royal Society, B: Biological Sciences, 334(1270), 243–251.
Hawkes, K., O’Connell, J. F., & Blurton Jones, N. G. (2014). More lessons from the Hadza about men’swork.
Human Nature, 25(4), 596–619.
Hawkes, K., O’Connell, J., & Blurton Jones, N. G. (2018). Hunter-gatherer studies and human evolution: A
very selective review. American Journal of Physical Anthropology, 165(4), 777–800.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world. Behavioral and B rain
Sciences, 33(2–3), 61–83.
Hill, K., & Hurtado, A. M. (2009). Cooperative breeding in South American hunter-gatherers. Proceedings of
the Royal Society B: Biological Sciences, 276(1674), 3863–3870.
Hönekopp, J., & Watson, S. (2010). Meta-analysis of digit ratio 2D:4D shows greater sex difference in the
right hand. American Journal of Human Biology, 22(5), 619–630.
Hönekopp, J., Voracek, M., & Manning, J. T. (2006). 2nd to 4th digit ratio (2D:4D) and number of sex
partners: Evidence for effects of prenatal testosterone in men. Psychoneuroendocrinology, 31(1), 30–37.
Hönekopp, J., Bartholdt, L., Beier, L., & Liebert, A. (2007). Second to fourth digit length ratio (2D:4D) and
adult sex hormone levels: New data and a meta-analytic review. Psychoneuroendocrinology, 32(4), 313–
Howlett, C., Setchell, J. M., Hill, R. A., & Barton, R. A. (2015). The 2D:4D digit ratio and social behaviour in
wild female chacma baboons in relation to dominance, aggression, interest in infants, affiliation and
heritability. Behavioural Ecology and Sociobiology, 69,61–74.
Kemper, C. J., & Schwerdtfeger, A. (2009). Comparing indirect methods of digit ratio (2D:4D) measurement.
American Journal of Human Biology, 191,188–191.
Koster, J., McElreath, R., Hill, K., Yu, D., Glenn, S. J., Van Vliet, N., Gurven, M., Kaplan, H., Trumble, B.,
Bliege Bird, R., et al. (2019). The life history of human foraging: Cross-cultural and individual variation.
Kratochvíl, L., & Flegr, J. (2009). Differences in the 2nd to 4th digit length ratio in humans reflect shifts along
the common allometric line. Biology Letters, 5(5), 643–646.
Lee, R. B. (1979). The !Kung San: Men, women, and work in a foraging society. Cambridge: Cambridge
Lew-Levy, S., Boyette, A. H., Crittenden, A. N., Hewlett, B. S., & Lamb, M. (2019). Gender-typed and
gender-segregated play among Tanzanian Had za and Congolese BaYaka hunter-gatherer children and
adolescents. Child Development.htt ps://doi.org/10.1111/c dev.13306.
Lieberman, D., Tooby, J., & Cosmides, L. (2007a). The architecture of human kin detection. Nature,
Lieberman, D. E., Bramble, D. M., Raichlen, D. A., & Shea, J. J. (2007b). The evolution of endurance running
and the tyranny of ethnography: A reply to Pickering and Bunn (2007). Journal of Human Evolution,
Lippa, R. A. (2003). Are 2D:4D finger-length ratios related to sexual orientation? Yes for men, no for women.
Journal of Personality and Social Psychology, 85(1), 179–188.
Lolli, L., Batterham, A. M., Kratochvíl, L., Flegr, J., Weston, K. L., & Atkinson, G. (2017). A comprehensive
allometric analysis of 2nd digit length to 4th digit length in humans. Proceedings of the Royal Society B:
Biological Sciences, 248,1–7.
Lombardo, M. P., Otieno, S., & Heiss, A. (2018). College-aged women in the United States that playoverhand
throwing sports have masculine digit ratios. PLoS One, 13(9), e0203685. https://doi.org/10.1371/journal.
Longman, D., Stock, J. T., & Wells, J. C. K. (2011). Digit ratio (2D:4D) and rowing ergometer performance in
males and fe males. American Journal of Physical Anthropology, 144(3), 337–341.
Longman, D., Wells, J. C. K., & Stock, J. T. (2015). Can persistence hunting signal male quality? A test
considering digit ratio in endurance athletes. PLoS One, 10(4), e0121560. https://doi.org/10.1371/journal.
Manning, J. T. (2002). The ratio of 2nd to 4th digit length and performance in skiing. Journal of Sports
Medicine and Physical Fitness, 42(4), 446–450.
Manning, J. T., & Fink, B. (2008). Digit ratio (2D:4D), dominance, reproductive success, asymmetry, and
sociosexuality in the BBC internet study. American Journal of Human Biology, 20(4), 451–461.
Manning, J. T., & Fink, B. (2018). Sexual dimorphism in the ontogeny of second (2D) and fourth (4D) digit
lengths, and digit ratio (2D:4D). American Journal of Human Biology, 30,1–7.
Manning, J. T., & Hill, M. R. (2009). Digit ratio (2D:4D) and sprinting speed in boys. American Journal of
Human Biology, 21(2), 210–213.
Manning, J. T., & Taylor, R. P. (2001). Second to fourth digit ratio and male ability in sport: Implications for
sexual selection in humans. Evolution and Human Behavior, 22(1), 61–69.
Manning, J. T., Scutt, D., Wilson, J., & Lewis-Jones, D. I. (1998). The ratio of 2nd to 4th digit length: A
predictor of sperm numbers and concentrations of testosterone, luteinizing hormone and oestrogen.
Human Reproduction, 13(11), 3000–3004.
Manning, J. T., Barley, L., Walton, J., Lewis-Jones, D. I., & Trivers, R. L. (2000). The 2nd:4th digit ratio,
sexual dimorphism, population differences, and reproductive success: Evidence for sexually antagonistic
genes. Evolution and Human Behavior, 21,163–183.
Manning, J. T., Morris, L., & Caswell, N. (2007). Endurance running and digit ratio (2D:4D): Implications for
fetal testosterone effects on running speed and vascular health. American Journal of Human Biology, 19,
Marlowe, F. W. (2000). The patriarch hypothesis. Human Nature, 11(1), 27–42.
Marlowe, F. W. (2004a). Marital residence among foragers. Current Anthropology, 45(2), 277–284.
Marlowe, F. W. (2004b). Mate preferences among Hadza hunter-gatherers. Human Nature, 15(4), 365–376.
Marlowe, F. W. (2007). Hunting and gathering: The human sexual division of foraging labor. Cross-Cultural
Research, 41(2), 170–195.
Marlowe, F. W. (2010). The Hadza: Hunter-gatherers of Tanzania. Los Angeles: University of California
Muller, M. N., Marlowe, F. W., Bugumba, R., & Ellison, P. T. (2009). Testosterone and paternal care in East
African foragers and pastoralists. Proceedings of the Royal Society B: Biological Sciences, 276(1655),
Nelson, E., Hoffman, C. L., Gerald, M. S., & Shultz, S. (2010). Digit ratio (2D:4D) and dominance rank in
female rhesus macaques. Behavioural Ecology and Sociobiology, 64(6), 1001–1009.
Oświecimska, J. M., Ksiazek, A., Sygulla, K., Pyś-Spychała, M., Roczniak, G. R., Roczniak, W., Stojewska,
M., & Ziora, K. (2012). Androgens concentrations and second to fourth-digit ratio (2D:4D) in girls with
congenital adrenal hyperplasia (21-hydroxylase deficiency). Neuroendocrinology Letters, 33(8), 787–
Paul, S. N., Kato, B. S., Hunkin, J. L., Vivekanandan, S., & Spector, T. D. (2006). The big finger: The second
to fourth digit ratio is a predictor of sporting ability in women. British Journal of Sports Medicine, 40(12),
Pettigrew, J. D., Bhagwandin, A., Spocter, M. A., Davimes, J., & Manger, P. R. (2017). Hands of living San
resemble those in Palaeolithic stencils, not modern Europeans. Transactions of the Royal Society of South
Africa, 73(1), 1–7.
Pickering, T. R., & Bunn, H. T. (2007). The endurance running hypothesis and hunting and scavenging in
savanna-woodlands. Journal of Human Evolution, 53(4), 434–438.
Puts, D. A., Gaulin, S. J. C., Sporter, R. J., & McBurney, D. H. (2004). Sex hormones and finger length: - what
does 2D:4D indicate? Evolution and Human Behavior, 25(3), 182–199.
Rammsayer, T. H., & Troche, S. J. (2007). Sexual dimorphism in second-to-fourth digit ratio and its relation to
gender-role orientation in males and females. Personality and Individual Differences, 42(6), 911–920.
Ribeiro, E., Neave, N., Morais, R. N., & Manning, J. T. (2016). Direct versus indirect measurement of digit
ratio (2D:4D): A critical review of the literature and new data. Evolutionary Psychology, 14(1), 1–8.
Scott, I. M., Clark, A. P., Boothroyd, L. G., & Penton-Voak, I. S. (2012). Do men’s faces really signal heritable
immunocompetence? Behavioural Ecology, 24(3), 579–589.
Sear, R., & Marlowe, F. W. (2009). How universal are human mate choices? Size does not matter when Hadza
foragers are choosing a mate. Biology Letters, 5(5), 606–609.
Smith, K. M., & Apicella, C. L. (2019). Partner choice in human evolution: The role of character, hunting
ability, and reciprocity in Hadza campmate selection. PsyArXiv.https://doi.org/10.31234/osf.io/35tch.
Stibbard-Hawkes, D. N. E. (2019). Costly signaling and the handicap principle in hunter-gatherer research: A
critical review. Evolutionary Anthropology, 28,144–157.
Stibbard-Hawkes, D. N. E., Attenborough, R. D., & Marlowe, F. W. (2018). A noisy signal: To what extent are
Hadza hunting reputations predictive of actual hunting skills? Evolution and Human Behavior, 39(6),
Tobler, M., Healey, M., & Olsson, M. (2011). Digit ratio, color polymorphism and egg testosterone in the
Australian painted dragon. PLoS One, 6(1), e16225. https://doi.org/10.1371/journal.pone.0016225.
Trivers, R., Manning, J., & Jacobson, A. (2006). A longitudinal study of digit ratio (2D:4D) and other finger
ratios in Jamaican children. Hormones and Behavior, 49(2), 150–156.
von Rueden, C., Gurven, M., & Kaplan, H. (2008). The multiple dimensions of male social status in an
Amazonian society. Evolution and Human Behavior, 29(6), 402–415.
Voracek, M., Manning, J. T., & Dressler, S. G. (2007). Repeatability and interobserver error of digit ratio (2D:
4D) measurements made by experts. American Journal of Human Biology, 19(1), 142–146.
Voracek, M., Reimer, B., & Dressler, S. G. (2010). Digit ratio (2D:4D) predicts sporting success among female
fencers independent from physical, experience, and personality factors. Scandinavian Journal of
Medicine and Science in Sports, 20(6), 853–860.
Wadley,L.,Trower,G.,Backwell,L.,&D’Errico, F. (2015). Traditional glue, adhesive and poison used for
composite weapons by Ju/’hoan San in Nyae Nyae, Namibia. Implications for the evolution of hunting
equipment in prehistory. PLoS One, 10(10), e0140269. https://doi.org/10.1371/journal.pone.0140269.
Wilkie, D. S., & Curran, B. (1991). Why do Mbuti hunters use nets? Ungulate hunting efficiency of archers
and net-hunters in the Ituri rain forest. American Anthropologist, 93(3), 680–689.
Wlodarski, R., Manning, J., & Dunbar, R. I. M. (2015). Stay or stray? Evidence for alternative mating strategy
phenotypes in both men and women. Biology Letters, 11(2), 20140977.
Wood, B. M. (2006). Prestige or provisioning? A test of foraging goals among the Hadza. Current
Anthropology, 47(2), 383–387.
Wood, B. M., & Marlowe, F. W. (2013). Household and kin provisioning by Hadza men. Human Nature,
Wood, B. M., & Marlowe, F. W. (2014). Toward a reality-based understanding of Hadza men’swork.Human
Nature, 25(4), 620–630.
Worthman, C. M., & Konner, M. J. (1987). Testosterone levels change with subsistence hunting effort in
!Kung San men. Psychoneuroendocrinology, 12(6), 449–458.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
Duncan Stibbard-Hawkes received his BA and PhD in biological anthropology from the University of
Cambridge. He conducted his PhD research with the Hadza in northern Tanzania, where he examined hunting
reputation and the costly signaling hypothesis. He currently holds an honourary fellowship in evolutionary
anthropology at Durham University.