63 - 76: Non-Human Primates: A Comparative Developmental Perspective on Yawning

Yawning as a form of ritualized display in nonhuman primates was commented upon by Darwin [1], who described it as an expression of passion and anger in baboons and other Old World monkeys. Many observations of captive primates have supported this view. For example, Zuckerman [2] described ‘bachelor’ Hamadryas baboons about to engage in conflict as ‘yawning’ (he used inverted commas) in association with teeth grinding, grimacing and staring at the intended target of the aggression. Plate XVIII in Zuckerman [2] clearly shows yawning by an adult male baboon described as‘threatening’. Early field studies of Old World monkeys further added to the evidence for yawning as a form of communication. Remarking how yawning clearly displays the large canine teeth of adult males, Hall and DeVore [3] proposed a ‘secondary intimidating effect’ (p. 100) for this behavior in baboons. Those authors published two photographs showing the unworn canines of a young adult male baboon yawning‘under tension’, and the broken and worn canine teeth of an old adult male yawning, respectively. Two more photographs in the same chapter describe adult males as‘harassing’ other adult males, and in each case the harassing male is clearly yawning. Interestingly, in the same volume, Simonds [4] draws attention to the generally better condition of the canine teeth of dominant males in a wild group of bonnet macaques compared to more subordinate males, and he describes how serious wounding may occur during fights. Simonds [4] does not include yawning among the gestures used to determine dominance ranks within the group, but a photograph illustrating the well-developed canines of a fully adult male clearly shows the male yawning (p. 181). Laboratory analysis of the scanpaths of rhesus monkeys visually exploring photographs of conspecifics‘ faces provides good evidence for the ‘canine display‘ aspect of yawning: when presented with photos of adult males yawning, around 50% of viewing monkeys‘ first saccades to the mouth were directed precisely to the canines, whereas photos of open-mouth threat elicited less specific saccades to the mouth area and to the ears [5].

Several lines of evidence link the frequency of yawning in adult male Old World monkeys to dominance rank. In sexually dimorphic polygynous species, such as macaques and baboons, there is a marked sex difference in the frequency of yawning: adult males easily out-yawn other age and sex classes [6]. In Celebes macaques, the rate of yawning by adult males varied with changes in their dominance status [7]. In wild olive baboons, the frequency of ‘nondirected’ yawning in the proximity of immigrant males varied with the condition of the displaying males‘ canines [8], and possibly reflected heightened tension. One adult male with broken canines refrained from yawning [8]. Bertrand [9] reported that only stumptail macaque males of high social rank responded by yawning when she looked at them; other males, over whom Bertand claimed to be dominant, showed no such response. She also noted teeth-grinding accompanying the dominant males’ yawns, and characterized the latter as threat yawns, rather than tension yawns.

When individually housed adult male stumptail macaques were introduced into the same cage, dominant individuals showed more yawning than subordinates [10]. Those staged encounters gave rise to physical attacks; injurious aggression can be avoided while yawning rate is increased in experimental encounters between unfamiliar adult males separated by a transparent partition (rhesus macaques [11]). Adult males may also yawn prodigiously when confronted with their reflection in a mirror (e.g. various species of macaques [11, 12] and patas monkeys [11, 13]). Semifree-ranging adult male Barbary macaques yawned more just after triadic encounters involving an infant (sometimes referred to as ‘agonistic buffering’) than during the encounters. The only one of 7 studied males not to show this effect was the most dominant male in the group [14]. The authors suggested that yawning was actively inhibited during the male-male body contact phase of an encounter, and that physical separation between the males switched off the inhibitory effect and thereby permitted a ‘tension component’ to be expressed in the form of yawning.

Information of the type reviewed in the previous section, specifically the fact that adult male Old World monkeys yawn readily in a range of situations characterized by tension or arousal, and that at least some yawns appears to be communicatory (note the distinction between ‘directed’ and ‘undirected’ yawns [6]), stimulated our first experimental studies of yawning in nonhuman primates. The question was whether adult male macaques could bring yawning under voluntary control. Although there was positive evidence for operant conditioning of a few naturally occurring behaviors in monkeys, such as vocalizations [15] and self-scratching [16], there were no previous attempts to train monkeys to produce yawns in order to obtain food rewards. Our first study used 2 behaviorally normal adult male pigtailed macaques housed in individual cages [17]. During 3 initial phases (each consisting of three 1-hour sessions), each monkey was passively observed by the experimenter, then given 20 pieces of fruit at random intervals during the session, then given fruit every time the facial expression ‘protruded lips’ was shown, respectively. Together these phases are referred to as condition A. There then followed5 sessions in which all yawns were rewarded with a piece of fruit, handed to the monkey by the experimenter. This is condition B. In the first session of condition B, monkey ‘Bernt’ was trained to yawn through a procedure of successive approximations: at first any mouth-opening act was immediately followed by a reward. By the end of the first session, simply opening the mouth went unrewarded; instead only full yawns were reinforced, including full lip retraction, wide opening of the mouth, and a climatic tilting of the head upwards and backwards to expose the teeth. The second male, ‘Charlie’, was rewarded from the start of condition B only for producing full yawns.

Figure 1 shows the huge increase in yawning by both males in condition B (continuous reinforcement) compared to condition A (baseline). It is unclear whether the slight variation in procedure used with the 2 males (successive approximation vs. full yawn required from the outset) played a role in the higher rate of yawning produced by Bernt. Condition C consisted of 2 extinction sessions: no rewards were given; it is clear that both males‘ rate of yawning quickly diminished to baseline levels when no rewards were forthcoming. There then followed another session of continuous reinforcement, which saw yawn rates rise rapidly again (condition D). Finally, a closing extinction phase in which no rewards were given led to another decline in yawning to baseline levels (condition E).

This study provided us with the first evidence that adult male macaques, like humans, could make themselves yawn. As described by Barbizet [18] for humans trying to make themselves yawn, Bernt and Charlie would open their mouths and appear to inhale deeply; both were also seen to abandon some attempts, further suggesting that they were deliberately trying to trigger yawns. These attempts were made while the monkey looked away from the experimenter. Upon successful completion of a yawn, the monkey would then orient towards the experimenter and sometimes extend a hand to receive the reward.

In a follow-up study, we used a similar procedure with 2 adult males of another species - tonkean macaques - with similar results [19]. Two new findings in this second study were: (1) a short-lived increase in yawning rates at the start of extinction sessions, reminiscent of the ‘partial reinforcement effect’ described in the learning literature, and (2) the especially high rates of yawning obtained when a fixed-ratio-3reward contingency was used, i.e. a reward was given only after the monkey produced3 yawns. This second study therefore confirmed that adult male macaques can voluntarily increase their rate of yawning when yawning is associated with obtaining rewards, and inhibit their rate of yawning when no extrinsic rewards are obtained.

One of the tonkean macaque males trained to yawn produced a mixture of ‘slow’and ‘fast’ yawns during the experimental phases. Yawns of the latter type were also frequently produced when this male came into visual contact with his former group, and also when he was confronted with his mirror image [11]. Are fast and slow yawns functionally different? Several authors have made the distinction between yawns whose main function might be communication and those that are simply due to fatigue or boredom. For example, Zuckerman [2] noted that baboons yawned not just as a form of harassment but probably also‘… in the human sense, from fatigue or boredom‘ (p. 262). He did not suggest differences in form between display and fatigue yawns, but various authors have invoked differences in the extent of teeth exposure, ear-flattening and eyebrow-raising, among other morphological aspects [e.g. 3, 9, 20]. Like humans, monkeys may show a peak in yawning shortly after wakening up [7, 21], and it would clearly be wrong to attribute a communicatory function to all yawns by adult males. Indeed, in his careful analysis of hundreds of yawns in groups of Old World monkeys, Deputte [21] concluded that around 90% of all yawns were ‘rest’yawns, whereas only 10% could be described as ‘emotion’ or ‘tension’ yawns.

It would appear that even among Old World monkeys, most yawns are a consequence of states such as fatigue or drowsiness. Nevertheless, if even a subset of yawns can be used as a social signal, it would be useful for an individual to be able to control their production. Our two studies on operant conditioning of yawning in adult male macaques provided evidence of this ability, and raised a number of potentially interesting questions that mostly remain unanswered, such as:

• Is the capacity for voluntary production and inhibition of yawning limited to Old World monkeys, or does it also exist in other nonhuman primates? As mentioned previously, the studies conducted on operant conditioning of yawning specifically used adult male macaques because of their known propensity for yawning in a range of contexts. It would therefore be useful to extend this line of work to other species, including those with different species-typical social structures and degrees of sexual dimorphism in body size. For example, we would expect it to be quite easy to increase the rate of yawning by adult males of other polygynous Old World monkeys in which adult males are bigger than females, with greater canine development. One the other hand, gibbons - the so-called‘ lesser apes’ are generally considered to be monogamous rather than polygynous, and there is much less of a difference in body size and in canine size between males and females than in the Cercopithecidae. Gibbons obviously yawn, but despite the lack of any detailed information on contexts of occurrence and frequency of yawning throughout the day, it seems safe to suggest that yawning in gibbons contains no element of threat. It is not even clear whether there is a sex difference in frequency of yawning (there does not appear to be any sex difference in human adults‘ rate of yawning either [22]). Can adult male or adult female gibbons exert voluntary control over their yawning? Nobody has attempted to find out.

• Can other age and sex classes control yawning? This question applies not just to Old World monkeys, but also to other species. For example, yawning rates in male macaques rise considerably with the onset of puberty, indicating a role for circulating androgens in preparedness to yawn [7, 23]. Castration of adult male rhesus almost totally eliminated yawning in a study of the effects of hormones on sexual behaviour; administration of testosterone reversed this effect [24]. Interestingly, administration of exogenous androgens increases the rate of yawning in adult female macaques to levels typical of adult males, whereas anti-androgen treatment blocks this effect [25, 26]. Such data indicate that, although under normal circumstances adult females yawn much less frequently than males, the requisite neural structures are in place that would allow flexible use of yawning by females. Of course, the development of the ability to control yawn production should also be amenable to study. Yawning clearly starts off as a reflex act; the question is at what point do individuals become able to (at least partly) overcome the reflexive aspect of the act and either voluntarily induce yawns, or inhibit them.

• One potentially interesting case to study might involve a species in which adult females are dominant over adult males. In the lemur branch of prosimian primates, for example, females are often, though not always, dominant over males [27]. Yawning occurs during agonistic interactions in ring-tailed lemurs; yawns that occurred during aggressive encounters including stink-fights were described as tension yawns [28]. They occurred mostly during pauses in the conflict, and were associated with displacement activities such as chewing movements and self-grooming. In other words, there is no evidence for yawns as displays in female lemurs, who in terms of body size are similar to adult males. When an unfamiliar adult female ringtail was introduced into a small captive group containing 2males and an adult female, both females engaged in agonistic interactions, and both yawned repeatedly, particularly in the first few minutes of the introduction;the 2 males did not show aggression-related yawns [29]. It is not known whether yawning can be brought under operant control in lemurs of either sex.

• Can other species control the production of yawns? Concerning non-primates, only 1 study appears to have asked whether yawning rates can be manipulated through reinforcement schedules. It was reported that domestic dogs rewarded for yawning tended to produce ‘pseudo-yawns’, consisting of simply opening the mouth [30]. Dogs would be a particularly interesting species to study in this respect given the recent burgeoning literature claiming to demonstrate impressive social and cognitive achievements in these animals. Indeed, there now exists a claim and a counter-claim concerning whether dogs show contagious yawning after seeing a human yawn. Discussion of these studies will be deferred until later. Before then, we turn to the question of the contagiousness of yawning in primates.

Although yawning may be ubiquitous across vertebrates, it has long been held that the phenomenon of contagious yawning is uniquely human [31, 32]. Although group-living species with highly synchronized behavioral profiles might show bouts of yawning at around the same time, such as upon awaking in the morning, or resting and digesting after feeding (e.g. lions [33]), the literature is remarkably lacking in descriptions of contagious yawning between nonhumans that are not explainable in terms of simple activity synchrony. Humans may see somebody yawning at any time of day, in any one of a wide variety of contexts, and be induced to yawn in turn. However, there is a huge range of both between- and within-individual differences in our susceptibility to contagious yawning. It is not an obligatory response; we do not yawn every time we see somebody else yawning, and in fact some people appear quite resilient to the effect (e.g. autistic children, when compared to typically developing children [34]). In experiments designed to assess susceptibility to contagious yawning, researchers usually ask people to watch controlled stimuli in the form of video images showing repeated yawning. Under these conditions, typically around 40-55%of participants (normal, young adults) report that they yawned [35, 36]. Investigation of factors underlying individual differences in this effect are still at a very early stage, but it would appear that people who score relatively highly on questionnaire measures of empathy and mental state attribution skills are more likely to yawn in response to seeing somebody else yawn [36].

The ease of constructing video sequences to show multiple examples of conspecifics yawning, and the knowledge that chimpanzees are capable of showing empathy and self-recognition, stimulated us to investigate contagious yawning in chimpanzees [37-38]. Two ‘yawn’ videotapes were prepared, each showing 10 short clips (6-8 s each, separated by blanks) of a chimpanzee yawning. One of these videotapes featured familiar chimpanzees from the same research centre where the study was conducted (Primate Research Institute, Kyoto University, Japan), while the second showed unfamiliar wild chimpanzees. The clips variously showed young infants, adult males and adult females yawning, either sitting or lying down. Two control videotapes were matched for general content, but instead of yawns they showed facial expressions such as grinning or threatening, or other mouth movements.

Six adult female chimpanzees, all members of a group with considerable experience of non-invasive behavioral experiments but with no previous experience of such video stimuli, were the participants. If they felt like doing so, they responded to an invitation by the experimenter to leave their group in the outside enclosure and move into a familiar experimental booth. They were thus tested individually, except for3 of them who each brought along their 3-year-old juvenile. Each adult female was tested for 4 sessions, with 1 yawn and 1 control video being presented in each session, for a total of 2 exposures to each video. Following a 5-min baseline/habituation period, a 35-cm video monitor placed at floor level near one glass wall of the booth was switched on, and one of the videotapes was presented (duration: approximately3 min). When the videotape ended, the monitor was switched off and a 3-min post-video observation period started. There then followed a 5-min distraction period, after which the second videotape was presented and followed by another post-video observation period. All occurrences of yawning by the chimpanzees were recorded in real time and verified by subsequent video analysis, as the chimpanzees‘ general behavior during video presentation and the post-video periods was videotaped.

Two striking results emerged from this experiment. First, 2 of the females showed highly significant increases in yawning in response to yawn video stimuli compared to control stimuli; the 4 others showed no clear effects, always yawning relatively infrequently. So, in this small sample, 33% adult females showed contagious yawning. Although this is lower than the 40-50% of contagious yawners typically reported in studies with humans, in some ways the chimpanzee data can be seen as more impressive than the data from equivalent human studies. Note that human participants are usually instructed to self-monitor and report their yawns, and merely by thinking about their yawning, they are more likely to yawn. In contrast, apart from a general expectation about participating in some kind of task involving manipulating objects or a computer screen, the chimpanzees had no idea about why they were in the experimental booth; there was nothing to suggest to them that yawning was of some special interest. We therefore feel that contagious yawning in 33% of naïve naturally behaving chimpanzees compares favorably with data in humans.

The second striking outcome was that none of the 3 juveniles showed any yawning whatsoever during the study, in spite of watching yawns on the video monitor and also seeing their mothers yawn (one juvenile‘s mother was one of the 2 highly contagiously-yawning females described above). This finding is of interest in view of the finding that human children below the age of 5 years also appear to be immune to the yawn-inducing effects of yawns by other individuals [39]. Children ranging in age from 2 to 11years were shown a video of an adult talking about a visit to the zoo and reciting nursery rhymes, and frequently breaking off to yawn. Direct observations of the children during exposure to the videotape and during post-video observations revealed an absence of yawning in preschoolers, and a subsequent gradual increase throughout childhood in the tendency to ‘catch’ another‘s yawn (fig. 2), reaching adult-like levels by the age of 11 or 12 years. In another condition, preschoolers were read a story in which the central character yawned repeatedly; older children read the book for themselves. The overall similarity of the results to those obtained in the video condition suggest that very young children are not vulnerable to ‘psychological’ inductors of yawning, such as seeing someone else yawn, or reading or thinking about yawning. Furthermore, the overall similarity between chimpanzees and humans in the effectiveness of video yawns for inducing yawning in adults, but not young juveniles, implies some commonality of mechanisms underlying the phenomenon of contagious yawning in humans and our nearest evolutionary neighbors. The cognitive correlates of contagious yawning in developing human and great ape children remain to be clarified.

Another aspect of yawning in chimpanzees that indicates greater commonality with its human counterpart than with yawning in Old World monkeys is that it appears to serve no agonistic signaling function. Although chimpanzees may yawn due to tiredness, boredom, or mild tension or uncertainty, and show more than one type of yawn [40, 41] (fig. 3), there is no suggestion of anything like the ‘canine contests’ that occur in adult male macaques and baboons; this is in spite of the fact that chimpanzees show some degree of sexual dimorphism in body size and canine development. However, then it might be asked why, despite hundreds of thousands of hours of observations in the wild, in zoos and in laboratories, has nobody ever described naturally occurring contagious yawning in chimpanzees? One possibility is that it does occur in the wild, but at times of the day when humans are unlikely to see it occurring, such as immediately after waking early in the morning, while chimpanzees are in their arboreal nests, or at dusk, when chimpanzees are starting to settle in their nests for the night. An alternative possibility is that the phenomenon does not exist under natural circumstances in either wild or captive chimpanzees, or if it does, it is extremely rare. It may well be that video sequences showing sequences of yawns constitute a kind of supernormal stimulus that is able to trigger a contagious effect in some, though not all, chimpanzees.

Recent support for a ‘supernormal stimulus’-induced contagious yawning effect in chimpanzees comes from a study involving presentation of computer-animated stimuli representing chimpanzees either yawning or displaying control mouth movements [42]. Two-dozen captive chimpanzees were tested, and on average they yawned over 2.5 times more frequently to the yawn animations than to control animations. Furthermore, it was established that the chimpanzees attended to the yawn and control animations for similar amounts of time; thus, ruling out differential saliency as a factor.

An obvious question is: if video stimuli can induce contagious yawning in chimpanzees even if they do not normally show the phenomenon, would the same happen for monkeys? As already described, ‘canine contests’ can involve two or more individuals yawning at around the same time, but there is no evidence that monkeys show true contagious yawning, rather than simply an escalation of a tension-or display-related behavior in adult male Old World monkeys. To conduct the only study so far to address this question [43], we prepared a video showing 10 naturally occurring yawns by members of a captive group of stumptail macaques; all yawns were considered to occur in relaxed, not tense situations. A control tape showed 10 facial actions such as those that occur during vocalizing, teeth-chattering or chewing. The 2 tapes were matched for general structure and total duration (1 min30 s). The video stimuli were presented via a video monitor placed near a window so that any monkeys in the colony room could clearly see what was on the screen. The experimenter had an unobstructed view of all monkeys is the colony room, and recorded all occurrences of yawning during video presentation and for 3 min afterwards.

Over 14 yawn and 14 control sessions, every monkey except 1 adult female yawned at least once during at least 1 test session. Although there were no significant effects at the individual level, at group level, the average number of yawns in control sessions was 2.4 per individual, compared to 4.3 in yawn sessions, a significant difference. There was no overall relationship between dominance rank, sex or familiarity with the stimulus monkeys on yawning frequency.

So, does this finding of a group-level increase in yawning to the yawn video compared to the control video constitute evidence of contagious yawning in stumptail macaques? The evidence is not compelling, for a couple of reasons. First, as well as yawning more in the yawn condition, the monkeys also scratched themselves more; together, yawning and scratching can be a sign of anxiety or uncertainly. Second, over half of the yawns in the yawn video (6/10) were performed by an adult male. As already described, yawns by adult males draw attention to the canines, and can be interpreted as a kind of threat. Therefore, increased yawning accompanied by scratching in yawn video sessions might have been due to an increased feeling of uneasiness arising from seeing an adult male yawn repeatedly. In fact, 1 of the 2 adult males in the group was responsible for approximately half of all the yawns seen in the yawn condition. Our conclusion at present is that even with a ‘supernormal stimulus’ in the form of video yawn stimuli, there is not yet any strong evidence for contagious yawning in any species other than humans and chimpanzees.

Recently, however, this view has been challenged by a report of contagious yawning in gelada baboons [44]. Using data on over 3,000 yawns collected from a zoo-housed group of 21 individuals organized into 2 one-male units, it was shown that monkeys likely to have seen another individual yawning were more likely to yawn within a5-min period than if the latter individual had not yawned. The effect appeared most strongly for adult females. Furthermore, 3 types of yawn were distinguished, called‘covered teeth’, ‘uncovered’ teeth and ‘uncovered gums’, and the second individual to yawn was significantly more likely to produce the same type of yawn as the model, especially when the pair consisted of females. Finally, within 1 one-male unit (gelada social structure is based on subgroups containing 1 fully adult male and several adult females plus associated offspring), there was a positive correlation between frequency of grooming and temporal proximity of yawns between individuals. The authors interpreted the findings as follows: ‘䀦the matching response recorded between females probably reflects the need and ability of females to stay in tune with each others’ behavioral activities and to form coalitions and alliances based on a precise reading of others behavior’ (p. 4).

Clearly this report raises several questions that future research needs to address before gelada baboons can be admitted to the ‘club of contagious yawners‘. For example, what is the functional significance of the 3 different types of yawn (if any), and why should there be precise matching in contagion? What proportion of contagious yawns occurred between members belonging to different one-male units, as presumably these individuals are less empathic towards each other than those belonging to the same unit? Is it possible that the observed yawning in close temporal proximity indicates general behavioral synchrony of the type seen, for example, in lions, rather than true contagious yawning? Why has contagious yawning never been reported in other primate species in which bonds among females (including members of the same matriline) are crucially important?

It is not just in primates that yawning gives rise to controversy. Reflecting the growing wave of new research investigating the social and cognitive abilities of domestic dogs, one recent report claimed that dogs yawned when they observed a human yawning, but not when the human made control (non-yawn) mouth movements [45]. The size of the effect was impressive, with 21 of 29 dogs tested showing a positive response; indeed this percentage (72%) is considerably higher than the corresponding value in studies with humans (45-60%), although the latter studies usually use video models rather than a live model. This apparent cross-species contagion effect was attributed to dogs‘ advanced capacity for empathy, with the suggestion that it might help ‘coordinate dog-human interaction and communication’ (p. 446); the paper received considerable media coverage. There are, however, several issues that call for prudence before accepting this study as evidence of cross species contagious yawning by dogs. For example, in the ‘yawn’ condition, after attracting the dog‘s attention the experimenter started to simulate yawning accompanied by vocalizations (the nature of the vocalizations is not described); these yawn-vocalization pairings were presented 10-19 times over a 5-min period. In the control condition, non-yawn mouth movements were made, but without vocalizations, and it is unclear whether the frequency of repetitions was matched with that in the experimental condition. It therefore seems possible that the particularly strange behavior of the experimenter in the yawn condition, consisting of repeated yawning movements and vocalizations, caused a state of unease or anxiety in the dogs, giving rise to yawning.

Indeed, another study using controlled video sequences showing humans and dogs yawning or making control mouth movements obtained no strong evidence that dogs ‘catch’ yawns either of humans or conspecifics [46]. Only 1 of 15 dogs tested showed significantly more yawns during yawn videos (5 yawns in total) than during the open-mouth (0 yawns), and even that, the authors point out, could have arisen by chance. It should be obvious that the question of contagious yawning in dogs requires further study before concluding that they might show an even greater effect than humans.

Hopefully, this review of some of the comparative literature on yawning will stimulate further interest in the phenomenon not only in humans, but in nonhumans too. It is clear that we have a lot more to learn about the psychological influences on yawning, how this varies during development, and what the implications are for individuals‘ social and self-awareness. To conclude, I will offer some thoughts relevant to the intriguing question: ‘Why does contagious yawning exist?’ Consider the following: as other chapters in this book show, yawning gives rise to a suite of physiological changes (e.g. in cortical temperature, oxygen uptake and CO₂ dispersal, inner ear pressure equalization). Under normal conditions, these changes are beneficial to the organism. However, in many human societies it is considered impolite to yawn in the presence of others [47]. I suggest that inhibition of yawning due to social pressures and other factors such as a busy working life actually gives rise to a kind of yawn deprivation in many people. In other words, we ‘forget’ to yawn as much as we should, or at least when we normally should. It may therefore be that by virtue of a low-level empathic processes, seeing (or hearing) somebody else yawn, or reading or thinking about yawning, might function to remind our brain that it might be a good idea to yawn, which in turn will lead to some of the beneficial physiological effects outlined above. From this perspective, the lack of contagious yawning in young children becomes easier to explain, as youngsters do not have the same degree of inhibition when it comes to yawning in public. Young children yawn wherever and whenever they feel like it, which is usually when they are tired. Likewise, chimpanzees do not have the same kinds of social constraints on yawning as human adults do; they appear to be liberated yawners [e.g. 41]. Chimpanzees are capable of showing empathy with other individuals, but because under normal circumstances they have little need to catch up on yawning, they show little susceptibility to the contagious yawning effect that we see in human adults. It is only when some form of extra yawning stimulation is given to chimpanzees, for example in the form of repeated video or animated images, that they may be moved to show the effect.