In contrast, anthropological and sociological approaches to humour have focused predictably on the social processes in humour and the social functions humour serves. Some theorists have seen joking as “play” on socially perceived form (including not only linguistic forms but all social actions), while others have gone beyond logical-semantic analyses in the text of jokes (Johnson, 1978) arguing for the constitutive role of context and social response in the joke itself. There is also a long tradition of exploring the role of the fool/trickster who plays a central part in myths and parables in many cultures (Charles, 1945; Imafaku, 1988) [2] either through wit or through a socially permitted (even required,) absurdity, foolishness and profanity. Adult clowns attract us for many reasons–they are creators of freedom, resilient despite degradation, eternal amusers, moralists-in-reverse, they remind us not to become over serious and free us from our normal restraints (Welsford, 1935; Pollio and Edgerly, 1976; Levine, 1961; Charles, 1945; Klapp, 1950; Lorenz, 1966)–but they can only succeed because they have their fingers on the pulse of society and on the reactions of others. In general these approaches place the social and the emotional at the heart of humour.

When humour creation has been looked at in psychological research (e.g. McGhee, 1974, 1980a), research has been largely restricted to verbal reports and verbal humour, ruling out non-verbal participants such as infants. Freud, for instance, saw the first stage in the development of humour as being “play”, consisting of absurd and nonsensical combinations of words or ideas, and starting from around 2 years of age (Freud, 1916, 1991), but ignoring nonsensical actions which may occur earlier. Modern cognitive-developmental accounts largely agree that the perception of humour does not arise until 18 months with the onset of the ability to perceive mis-expected (rather than merely unexpected or novel) “stimuli” as make-believe (Schultz, 1976; McGhee, 1980 b). This has supported the general neglect of humorous interactions in infancy before this period, despite a contrary argument that humour is based on the simpler ability to perceive incongruity as playful rather than as make-believe and is therefore potentially possible even by 4 months of age (Pien and Rothbart, 1980). We know from research on responses to humorous events that infants from at least 7 months of age laugh at “stimuli” which in adults and older children would be called humorous (Sroufe and Wunsch, 1972). [3] Although we know that by the end of the first year infant laughter is increasingly obtained by their own actions rather than just by a passive witnessing of others’ acts (Sroufe and Wunsch, 1972), we know little about when, or even whether, infants deliberately try to make others laugh and what they do to achieve it.

The present paper explores infant creation of humorous engagements before infants are verbally competent, and examines the social and emotional implications of these engagements. The focus is on infant clowning–i.e. on acts which infants repeat deliberately in order to re-elicit laughter from others. Clowns are arch manipulators of our emotional reactions and their ability to sense humorous reactions in others and their desire to elicit and engage with it may be present early in development. The study of humour creation in its developmental origins might help us to better understand the role of clowning in adult interaction and, vice versa, the fact that clowns are such an important feature of even modern society might help us to understand the significance of clowning as humour in infancy.

Data from two longitudinal studies covering other aspects of play as well (Reddy, 1991, 1998), are summarised here, one with 12 infants visited fortnightly in their homes from 7 to 12 months and the other with 20 infants visited at 8 and 11 months. Since critical events are most likely to happen off camera and at times of heightened interaction and emotionality in the absence of strangers, parents were trained to observe and report at interviews on each visit on particular incidents involving laughter and humour. Naturalistic play interactions between parents and infants during the home-visits were video-taped.

Identifying clowning. There are two ways in which one can approach the identification of clowning in everyday interactions: One, through looking for instances where an act which obtains laughter from others is intentionally repeated in order to re-elicit the laughter, i.e. to approach the phenomenon neutral to the content of the act. And two, through looking for similarities with the acts which clowns in many societies are known to do, i.e. focusing on the content of the act, and enabling, for instance, the identification of acts which may only be performed once. In the present studies parents were regularly asked to observe and report on infant clowning using the first criterion, but were asked to give details of the acts which would later allow comparison with the acts of adult clowns. For clowning to be accepted as present at a given age, it was required that parents report at least one incident in detail of the intentional repetition of an act by the infant in order to re-elicit laughter from others.

The prevalence and nature of clowning acts. Parents had little difficulty in identifying instances when their infants did things intentionally to elicit laughter from them. Of the parents who gave clear responses to the question, 87% of 8 month olds, 87% of 11 month-olds (and 100% of 14 month-olds)–were reported to have shown at least one instance, with details, of clear clowning with no significant age differences in the presence of clowning. Some of the parents at 8 months reported only simple acts such as splashing and banging, which increased in frequency and intensity when others laughed. Other parents, even at 8 months, reported more complex and idiosyncratic acts (see Table 1). There were increases with age in the obviousness of the attempts to re-elicit laughter. At 8 months most of the attempts were short-term ones, i.e. repeating an action immediately a few times if it was followed by laughter from others, with only a few actions being repeated across episodes or over several days. However, the majority of infants were reported to have shown at least one incident of long-term clowning. By 11 months long-term clowning was taken for granted.

Adult clowns are known to engage in a variety of foolish, absurd, fake, incompetent and mocking acts, the prevalence of each of these varying across individuals, situations and cultures. Table 1 shows some examples of adult clowning in various forms, categorises infant acts done to re-elicit the laughter of others, and gives examples of infant clowning acts from the two studies.

The most frequently reported acts were actions of the head and face, mainly head shaking, nodding or wobbling, and screwed up faces. The infant actions shown in Table 1 were all ones which we as adults would recognise as being funny, and fall without difficulty into categories of actions derived from adult clowning. This in itself is unsurprising, for it is after all adults or older children who provide the laughter which maintains and allows embellishment to the infants’ acts. It was always, however, the infants’ sensitivity to and interest in the laughter and emotional reactions of others that then allowed these acts to be repeated as clowning.

It is clear that some of the acts of infant clowning derived their meaning only from adult reactions–e.g. patting the mother on the head or holding up “smelly feet”. But most of them were variations of actions which were already in the infant’s repertoire as meaningful acts. The variations on, e.g. normal (and newly learned) walking or normal holding of the head and neck or normal expressions of the face or normal calling, appeared to already interest the infant as variants before their social value as laughter elicitors became apparent. In many cases where parents were aware of the history of a particular clowning action, it was clear that the infant had begun to engage in the act for “solitary” reasons–e.g. that the act in itself was loud or extreme or provided an interesting sensation in the body–but the reactions of those who witnessed the act extended its meaning. In many other cases–such as the teasing–it was the discrepancy between the act and socially acceptable forms that was central and it was the other’s reactions which were sought in the first instance.

Clowning and the appreciation of humour. In repeating acts for re-eliciting others’ laughter and in teasing by deliberately violating shared expectations, understandings and conventions, infants are not just creating “stimulus discrepancy” or merely creating unexpected or novel “stimuli” (McGhee, 1972). They are, first of all, creating these in interpersonally appropriate humorous situations. And second, in many of the acts of clowning and teasing infants are, in a ludic context, creating social mis-expectations. These are undoubtedly very simple mis-expectations, and are subsequently supported by laughter–but then so, one might argue, is some adult clowning. These data provide support for Pien and Rothbart’s (1980) view that the capacity for humour may be present in the first year of infancy.

Further evidence that the infant is not blind to the emotional significance of the laughter that she is eliciting comes from another source. Infants are sensitive to the laughter of others and are interested in the laughter as a social and communicative phenomenon. In the present studies 59% of the parents were certain that their infants at 8 months occasionally sought to join in with the laughter of those around them even when the reasons for the laughter did not involve the infant (23% were uncertain and 17% were certain that this had not happened). Parents also reported that infants had different kinds of laughter, including a fake laugh which was used for joining in others’ laughter (clearly observed by the parents in 50% of the 8 month-olds), and in two infants (one at 8 months and another at 11 months, ) for actually seeking to prompt laughter in others. Two parents reported that their infants also had a polite laugh–which was a scaled-down version of the full laugh, used when others were trying to make them laugh, when the infant knew that a laugh was called for but the attempt had ceased to be really amusing for the infant.

Given that infant acts of clowning are largely determined by the responses of others, one could ask to what extent the infant appreciates the humour in the act. It would appear that the humour is in the responsive context, not solely in the text of the act. One could argue that these acts may serve as a kind of mechanical platform upon which “real” (i.e. contextually independent) humour appreciation may be built. However, it is not the case that even in adulthood clowning is ever independent of either the situation or the response. Neither the existence of a humorous incongruity nor its successful use can be independent of social reactions (Bergson, 1913; Hayworth, 1928; Greig, 1923), but the humour creator needs both to understand the significance of responses and to have a desire to maintain them. A contextual (or social and non-individualist) approach to humour allows us to treat similarities between infant and adult clowning as evidence of potential continuities. A purely textual approach (focusing on individual cognitive advances) would require a priori rejection of these phenomena as being out of bounds.

Mocking the serious. Some categories of adult clowning are clearly missing in the repertoire of the infants at this age. One key instance is lampooning or ridiculing the serious. Infant clowning is at a simpler, less satirical level. The violation of taboos is happening because the infant is recognising and violating the taboo for reasons of testing boundaries and eliciting laughter in the process rather than for creating satire. Is there, in infant clowning, the mocking of the serious that we see in adult clowning? Only to a limited extent can we see this in instances where the infant imitates the serious acts of others. For example, when Sh, 11m, imitates her great grandmother’s snoring face, she is detecting, and offering to others her detection, of something odd about the expression. She is not, as an adult clown might do, linking in with centuries of cultural knowledge and jokes about the infirmities of old age. It is only the embarrassed and suppressed laughter of the others around her that contains that knowledge and makes her act so funny. However, this act, and Rob’s similar imitation of his grandmother’s face, is already the beginning of play with and playful comment upon the serious acts of others.

Teasing, violation of norms and exploration of shared meanings. However, another kind of humour which is beginning at this age is teasing (Hubley and Trevarthen, 1979; Reddy, 1991; Reddy and Williams, in preparation). Three kinds of teasing are common before the end of the first year: teasing with offering and withdrawing of objects, teasing with non-compliance, and teasing with disruption. In these acts infants are showing the beginnings of deliberate violations of norms and socially agreed meanings–of gestural acts (as in offering objects with an extended arm, of others’ intentional actions (as in “she wants to clear the floor”, “she wants the tower to stand”, “she wants to play the recorder”), or of norms or others’ intentions for the infant’s actions (as in “mustn’t touch”, “mustn’t go there”, “mustn’t spit”, etc.). These violations are not the considered satirical comments that adult clowns engage in. However, in these the infants are performing some of the same functions by taking serious intentions and agreed meanings and turning them round, playing with their limits and challenging their incorporation into shared culture.

Clowning and Emotion. Bergson (1913) argued that humour was an essentially unemotional process, that above all it required something like a momentary anaesthesia of the heart. In a similar vein others have argued that “a good clown is like a good anthropologist. He or she observes the social world as a participant, but is clumsily integrated with the surroundings and thus ironically detached” (Manning, 1972, p. 12). In arguing that humour is social but unemotional, Bergson assumes a separation of the social from the emotional which is not easy to sustain at least in infancy. He claimed that “in a society composed of pure intelligences there would probably be no more tears, though perhaps there would still be laughter” (Bergson, 1913, p. 4). On the contrary, it would seem from evidence of the socioemotional basis of clowning and humorous laughter in infancy, that far from being its foe, “being in tune and unison with life” and others is the basis of the comic. Good clowning is based upon a central interest in others’ laughter [4] and a desire to elicit it, and is sustained only in full engagement with others. Both its form and its content is dependent on a sensitivity to and engagement with the emotional reactions of others. Even the momentary dis-engagement that is needed to step back and violate the ongoing sequence of interaction is still sustained by the emotionality of others’ reactions. Our own fondness for intellectualising the processes of engagement may lead us to see success in clowning as a detached intellectual process rather than an involved, emotional one. Even adult clowning consists partly of the engaged sensitivity to and reliance upon, the emotional reactions of their audience, and the nature of adult clowning acts is heavily shaped by the things that matter to those around them (otherwise it would not work).

Conclusions. Clowning as an activity which is sensitive to and plays upon others’ laughter appears to be present in the second half of the first year of human infancy. For the infants, as it is for adult clowns too, the play with others’ reactions is the crucial emotional key to such engagements. Perception of the value to others of taboos and norms gives the clown power in different ways at different ages and times, developing from these simple acts to more complex satirical clowning in adulthood. The developmental origins of humour creation appear to be intrinsically social and emotional; the affective responses of others form a constitutive part of humorous incongruity and thus shape both the form and the content of humour.