Rearing condition and willingness to approach a stranger explain differences
in point following performance in wolves and dogs
Christina Hansen Wheat,
Wouter van der Bijl &
Clive D. L. Wynne

Abstract
The relative importance of adaptation and individual ontogenetic experience in dogs’ high levels
of behavioral compatibility with humans has been a topic of intense scientific attention over the
past two decades. Salomons et al. Current Biology, 31, 3137–3144, (2021) recently presented a
particularly rich data set of observations on both wolf and dog puppies that has the potential to
contribute substantially to this debate. In their study subjecting wolf and dog puppies to batteries
of tests, including the ability to follow human pointing gestures, Salomons et al. (2021) reported
that dogs, but not wolves, have a specialized innate capacity for cooperation with humans.
However, upon reanalyzing this data set, we reach a different conclusion—namely, that when
controlling adequately for various environmental factors, wolves and dogs perform similarly in
their cooperation with humans.

Authors’ final draft. For published version go to https://rdcu.be/cXuRd

Hansen Wheat, C., van der Bijl, W., & Wynne, C. D. L. (2022). Rearing condition and
willingness to approach a stranger explain differences in point following performance in
wolves and dogs. Learning & Behavior. https://doi.org/10.3758/s13420-022-00544-2

Rearing Condition…

Hansen-Wheat et al.

1

Rearing condition and willingness to approach a stranger explain differences
in point following performance in wolves and dogs
The claim that dogs may have unique abilities to recognize the implications of human
intentional gestures is not new. In 2002, Hare et al. proposed the domestication hypothesis,
which stated that “dog’s social-communicative skills with humans were acquired during the
process of domestication” (p. 1636). This hypothesis was largely based on findings suggesting
that dogs outperform wolves in following human pointing gestures and that dogs’ performance in
following pointing did not vary with age. In 2008, Riedel et al. added to the domestication
hypothesis by reporting that dog puppies as young as six weeks would follow human pointing
gestures, thereby supporting the narrative that, as a result of domestication, dogs possess a
unique, innate capability to interpret human social-communicative cues.
However, several findings call into question the robustness of the claim that dogs have
unique abilities in interpreting human gestures as a consequence of domestication. First,
methodological concerns have been raised in relation to both studies cited above. Udell et al.
(2008) identified that the wolves in Hare et al. (2002) were tested under different conditions
from the dogs. Further, Udell et al. (2008; see also reanalysis of original results in Udell &
Wynne, 2010) demonstrated that, when tested under identical conditions, wolves may even
outperform dogs in following human pointing gestures. Second, when testing puppies ranging
from nine to 21 weeks of age, Dorey et al. (2010) found a significant developmental trajectory in
the ability to interpret human pointing gestures. Third, various studies have demonstrated that,
if socialized to humans at an early age, a wide range of nondomesticated and domesticated
species—including wolves (Canis lupus; Udell et al., 2008), megachiropteran bats (Pteropus;
Hall et al., 2011), African elephants (Loxodonta africana; Smet & Byrne 2013), pigs (Sus scrofa;
Nawroth et al., 2014), and goats (Capra hircus; Nawroth et al., 2020)—can follow human socialcommunicative cues as given in pointing tests. It has even recently been revealed that handraised wolf puppies as young as eight weeks, without any prior training, will spontaneously
engage in cooperation with a stranger based on human intentional gestures (Hansen Wheat &
Temrin, 2020), thereby indicating that such capabilities were present in the ancestral populations
to domestic dogs. Together, these studies make a strong case that phylogeny alone is not enough
to account for any animal’s ability to follow human social cues. Rather, only when phylogeny is
considered in conjunction with ontogeny can an individual’s success or failure in following
human pointing gestures be properly interpreted (Udell et al., 2010).
Salomons et al. (2021) have now reinvigorated this debate by testing 44 dog and 37 wolf
puppies on a variety of tasks relating to their comprehension of communicative gestures. They
concluded that their data offered “support [for] the predictions of the domestication hypothesis
[that d]og but not wolf puppies are attracted to humans and show early emerging skills for
reading human gestures, even though the wolf puppies received more intense human
socialization” (p. 3141). They further concluded that “dog puppies are specialized for cooperative
communication with humans” (p. 3142).
Salomons et al. (2021) exposed their dog and wolf puppies once each to a battery of tests.
The most relevant tests for this commentary are the ability to follow a human pointing gesture
and the willingness to approach an unfamiliar human (classified by the authors as temperament).
Point-following addresses a central issue in the ongoing debate over the relative importance of

Rearing Condition…

Hansen-Wheat et al.

2

phylogeny and ontogeny in dogs’ adaptation to human-dominated environments. Willingness to
approach an unfamiliar human enables a test of the proposition that any advantage dogs might
show in point following could be due to a hypersocial phenotype (vonHoldt et al., 2017)—in
other words, it could be due to a social rather than a cognitive advantage over wolves.
Clearly, in order to rigorously test for the impact of domestication on any behavioral
differences between dogs and wolves, it is essential to adequately control for environmental
effects. Because the availability of hand-reared wolves is a limiting factor, studies of this kind are
inherently characterized by small sample sizes, and eliminating environmental bias therefore
becomes even more important. In Salomons et al. (2021), wolf and dog puppies received varying
degrees of socialization and environmental conditions prior to testing, and, importantly,
conditions for the two subspecies were not matched. Specifically, wolf puppies were hand-raised,
but received varying levels of socialization, with 24% of the puppies being in contact with their
human caregivers to a far lesser extent: “Wolf puppies remained with littermates but received 12
h (24%) or 24 h (76%) human care from 10 to 11 days after birth” (p. 3138). In contrast, “all dog
puppies remained with their mothers until weaning, around 6 weeks of age, and with their
littermates until ~8 weeks of age. During this time, they mainly socialized with humans during
short routine caretaking tasks. Around 8 weeks of age, puppies were then sent to live with
human families” (p. 3138). That is, dogs, but not wolves, were moved to human homes.
Importantly, 18 out of 31 of the dog puppies were tested in the pointing test at >9 weeks of age
after they had been sent to live with human families. This environmental change strongly altered
their exposure to people and was not matched in the wolf puppies. Furthermore, in the pointing
test dogs were tested up to 17 weeks of age, whereas none of the wolves were older than 13
weeks when tested. Because age and degree of human exposure is inherently confounded in the
dog population in the data set provided in Salomons et al. (2021), but the human exposure is not
controlled for in the original analyses, we consequently reanalyzed Salomons et al.’s data
separating the dogs tested after they had moved to human homes from the wolves and younger
dogs tested while they lived with their mothers. We find that dog puppies only significantly
outperformed wolf puppies after they have been moved to human households (Fig. 1a).
Furthermore, the apparently superior performance of dogs compared to wolves may be a
consequence of their heightened tendency to approach strange humans.

Rearing Condition…

Hansen-Wheat et al.

3

Figure 1

a Proportion of trials in which an animal followed a human pointing gesture, comparing
wolves with young dogs (7–8 weeks, living with their mother and littermates) and older dogs
(10–17 weeks, living with human families). Only the difference: wolves—older dogs (10–17
weeks) is significant. b Proportion of trials in which an animal followed a human pointing
gesture is related to the willingness to approach an unfamiliar human. The line represents the
model fit from a generalized linear mixed model, with the grey area showing the 95% confidence
interval (bslope = 1.14, SE = 0.42, p = 0.006).
Since testing of the dog puppies’ willingness to approach people and follow their pointing
gestures was conducted for some dogs before, and for some after, they were moved to human
homes, the possible effect of this environmental change on point following needs to be
controlled for. Thus, to test whether rearing conditions or species identity better accounts for the
differences in human-gesture-following performance between wolf and dog puppies, we re-ran
the comparison of performance on the gesture-following task, dividing the dogs into two groups:
those still living with their litter and thus had less human contact when tested (i.e., those dogs ≤9
weeks old), and those placed with human families (>9 weeks old). Since only five wolves were
tested before the age of 10 weeks, we grouped all wolves together for our analyses. This
reanalysis (binomial generalized linear mixed model [GLMM]) showed that the wolves differed
from older (bolder dogs - wolves = 1.05, SE = 0.334, p = .0046), but not younger dogs (byoung dogs - wolves =
0.42, SE = 0.338, p = .42), thus contradicting Salomons et al.’s (2021) claim that dogs
outperform extensively socialized wolves in a cooperative communication task with humans
“despite having far less human exposure than the wolf puppies” (p. 3141; Fig. 1a).
Salomons et al. (2021) also reported that dog puppies were around 30-fold more willing
to approach a strange human than wolf puppies of comparable ages. Since gesture-following in
an experimental context necessarily also involves approaching a stranger, this large difference in
willingness to approach could account for the difference in gesture following between the two
species. However, Salomons et al. (2021) stated that after the effect of species was entered in the
statistical model to account for gesture following, approach performance was not a significant
covariate (we note that species is also not significant in this model). We redid this analysis, using

Rearing Condition…

Hansen-Wheat et al.

4

a corrected dataset provided by the authors. We found that the single best model (binomial
GLMM, AIC: 404.25) to explain the gesture following data has only one fixed effect, the
willingness to approach a stranger (Fig. 1b). This is indistinguishable from a model that only has
species as a fixed effect (AIC: 404.75), and a model that has both willingness to approach a
stranger and species as fixed effects (AIC: 405.55). Thus, previously established differences in
willingness to approach a stranger (termed temperament in Salomons et al., 2021) are sufficient
to explain differences in test performance, and there is no need to attribute superior
communicative skills to dogs.
It is well-established that hand-raised wolves, even when extensively socialized, do not
generalize this socialization to strangers and remain fearful of unfamiliar humans (Hansen
Wheat et al., 2022; Klinghammer & Goodman, 1987; Zimen, 1987). Indeed, Salomons et al.
(2021) report that “we attempted to test a total of 49 wolf puppies and were able to collect data
from 37. We were unable to collect any data with the other 12 wolf puppies because they were
too nervous around unfamiliar humans (i.e., experimenters), even though they had been heavily
exposed to humans” (p. e3). Even when successfully tested, hand-raised and socialized wolves
that have been exposed to strangers under controlled conditions during their upbringing still
express significantly elevated stress and fear responses towards strangers in test situations
(Hansen Wheat et al., 2022). On the other hand, prior studies have demonstrated that dogs have
a greatly enhanced propensity to approach and interact with people (Udell, 2015; vonHoldt et
al., 2017) and express little, or no stress or fear behaviors around strangers in test situations
(Hansen Wheat et al., 2022). The use of unfamiliar humans as experimenters in the pointing
tests in Salomons et al. (2021) thus highlights a potential bias favoring the dogs’ performance.
The emotional engagement towards people and lack of fear around strangers is likely a key factor
in dogs’ success in human-dominated environments (Wynne, 2021), and is sufficient to explain
the differences between species reported in Salomons et al. (2021) without the need to evoke
specialized cognitive adaptations.
In sum, Salomons et al. (2021) make an appreciable contribution to the literature by
offering a substantial data set which can be used to test various behaviors in wolves and dogs.
While we agree that wolves and dogs differ behaviorally in important ways, we do not believe
that these data demonstrate that dogs have an evolved capacity to follow human gestures at an
earlier age than wolves. Rather, when rearing contexts and willingness to approach strange
humans are considered, dog and wolf puppies perform very similarly. We further disagree that
rearing condition and willingness to approach a stranger can be ruled out as significant
determinants of cognitive performance in these subspecies. We therefore emphasize the
importance of designing studies in which wolves and dogs are reared and tested under identical
conditions, in order to adequately disentangle domestication effects on behavior.

References
Dorey, N. R., Udell, M. A. R., & Wynne, C. D. L. (2010). When do domestic dogs, Canis
familiaris, start to understand human pointing? The role of ontogeny in the development of
interspecies communication. Animal Behaviour, 79(1), 37–41.
https://doi.org/10.1016/j.anbehav.2009.09.032

Rearing Condition…

Hansen-Wheat et al.

5

Hall, N. J., Udell, M. A. R., Dorey, N. R., Walsh, A. L., & Wynne, C. D. L. (2011).
Megachiropteran bats (Pteropus) utilize human referential stimuli to locate hidden
food. Journal of Comparative Psychology, 125(3), 341–346. https://doi.org/10.1037/a0023680
Hansen Wheat, C., Larsson, L., Berner, P., & Temrin, H. (2022). Human-directed
attachment behavior in wolves suggests standing ancestral variation for human–dog
attachment bonds. Ecology and Evolution, 12, e9299. https://doi.org/10.1002/ece3.9299
Hansen Wheat, C., & Temrin, H. (2020). Intrinsic ball retrieving in wolf puppies suggests
standing ancestral variation for human-directed play behavior. iScience, 23(2), 100811.
https://doi.org/10.1016/j.isci.2019.100811
Hare, B., Brown, M., Williamson, C., & Tomasello, M. (2002). The domestication of social
cognition in dogs. Science, 298, 1634–1636.
Klinghammer, E., & Goodmann, P. A. (1987). Socialization and management of wolves in
captivity. In H. Frank (Ed.), Man and wolf: Advances, issues and problems in captive wolf
research (pp. 31–60). W. Junk Publishers.
Nawroth, C., Ebersbach, M., & von Borell, E. (2014). Juvenile domestic pigs (Sus scrofa
domestica) use human-given cues in an object choice task. Animal Cognition, 17, 701–
713. https://doi.org/10.1007/s10071-013-0702-3
Nawroth, C., Martin, Z. M., & McElligott, A. G. (2020). Goats follow human pointing
gestures in an object choice task. Frontiers in Psychology, 11, Article 915.
https://doi.org/10.3389/fpsyg.2020.00915
Riedel, J., Schumann, K., Kaminski, J., Call, J., & Tomasello, M. (2008). The early ontogeny
of human–dog communication. Animal Behaviour, 75(3), 1003–1014.
https://doi.org/10.1016/j.anbehav.2007.08.010
Salomons, H., Smith, K. C. M., Callahan-Beckel, M., Callahan, M., Levy, K., Kennedy, B.
S., Bray, E. E., Gnanadesikan, G. E., Horschler, D. J., Gruen, M., Tan, J., White, P.,
vonHoldt, B. M., MacLean, E. L., & Hare, B. (2021). Cooperative communication with
humans evolved to emerge early in domestic dogs. Current Biology, 31, 3137–3144.
https://doi.org/10.1016/j.cub.2021.06.051
Smet, A. F., & Byrne, R. W. (2013). African elephants can use human pointing cues to find
hidden food. Current Biology, 23(20), 2033–2037. https://doi.org/10.1016/j.cub.2013.08.037
Udell, M. A. R. (2015). When dogs look back: Inhibition of independent problem-solving
behaviour in domestic dogs (Canis lupus familiaris) compared with wolves (Canis
lupus). Biology Letters, 11, Article 20150489. https://doi.org/10.1098/rsbl.2015.0489
Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2008). Wolves outperform dogs in
following human social cues. Animal Behaviour, 76(6), 1767-1773.
https://doi.org/10.1016/j.anbehav.2008.07.028
Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2010). What did domestication do to
dogs? A new account of dogs’ sensitivity to human actions. Biological Reviews, 85(2), 327–
345. https://doi.org/10.1111/j.1469-185X.2009.00104.x
Udell, M. A. R., & Wynne, C. D. L. (2010). Ontogeny and phylogeny: Both are essential to
human-sensitive behaviour in the genus Canis. Animal Behaviour, 79(2), e9–e14.
https://doi.org/10.1016/j.anbehav.2009.11.033
vonHoldt, B. M., Shuldiner, E., Koch, I. J., Kartzinel, R. Y., Hogan, A., Brubaker, L.,
Wanser, S., Stahler, D., Wynne, C. D. L., Ostrander, E. A., Sinsheimer, J. S., & Udell, M.
A. R. (2017). Structural variants in genes associated with human Williams–Beuren syndrome
Rearing Condition…

Hansen-Wheat et al.

6

underlie stereotypical hypersociability in domestic dogs. Science Advances, 3(7), Article
e1700398. https://doi.org/10.1126/sciadv.1700398
Wynne, C. D. L. (2021). Dogs’ (Canis lupus familiaris) behavioral adaptations to a humandominated niche: A review and novel hypothesis. In M. Naguib & L. Barrett
(Eds.), Advances in the study of behavior (Vol. 53, pp. 97–162). Academic Press.
https://doi.org/10.1016/bs.asb.2021.03.004
Zimen, E. (1987). Ontogeny of approach and flight behavior towards humans in wolves,
poodles and wolf–poodle hybrids. In H. Frank (Ed.), Man and wolf: Advances, issues, and
problems in captive wolf research (pp. 275–292). Springer Science & Business Media.

Acknowledgments
We thank Salomons et al. (2021) for providing us with a corrected data set for this
reanalysis and for being very forthcoming in answering our questions.

Rearing Condition…

Hansen-Wheat et al.

7