2.2.1. Definitions

Humor can be broadly defined as “anything that people say or do that is perceived as funny and tends to make others laugh” (Martin 2007, p. 5). This definition of humor therefore includes both the process by which an individual produces humor and the affective response it elicits in others. It is opposed to a stricter definition, which considers humor as a mental state distinct from other mental states, such as lying or making an error, and whose production is intentional (Hoicka 2014). The first broad definition allows identifying the emergence of humor perception and production as early as the first months of life with the appearance of the first laughter and smiles, while the latter would place the understanding and production of humor at a later age in infancy. Despite this discrepancy in the infant literature, authors agree that humor can be broken down into four constituent elements: the need to be placed in a social context, the intervention of perceptual-cognitive processes (including the perception of the incongruity of a situation), the provocation of an emotional response and the expression of laughter (Martin 2007). In infants, the hide-and-seek game of peekaboo is a perfect illustration of these four elements: it involves the action of a social partner, who hides his or her face before making it reappear, thus creating an incongruity (the disappearance and then the sudden reappearance of the social partner) and a surprise for the infant. The combination of these elements thus provokes a positive emotion, which translates into smiling and/or laughing. For Bergen (2015), humor occurs in a playful context because it is a playful way to interact with others. Humor is therefore a deviation from what is expected. Both children and adults have expectations about the physical and social world and any deviation from these expectations in a non-serious playful social context creates humor. The development of humor therefore depends on the infant’s knowledge and expectations about the world.

2.2.2. The different types of humor in infants

While the elements that constitute the definition of humor are the same for all types of humor, namely, incongruity, playful social context and laughter, the nature of these elements can change. This leads to different types of humor for which individuals will have different sensitivities at different ages in their lives. Some types of humor, such as irony, sarcasm and other forms that are expressed primarily through language, will develop relatively late in school-aged children and will not be the focus of this chapter. In infants, humor is mainly observed in body and facial expressions during the first year of life. Infants tend to laugh when parents make funny faces and sounds, hide their faces before they reappear, imitate the infant by crawling, blow bubbles on the infant’s body, etc. (Hoicka 2014). Infants can also produce humor themselves from the age of 7–8 months, when they voluntarily repeat actions that provoke laughter in others (Reddy 2001). From the age of 1 year, infants begin to enjoy humor involving objects. In a study involving six one year-olds observed in a nursery, Loizou (2005) showed that at this age infants use objects in ways that divert them from their original purpose, demonstrating both knowledge of cultural conventions and their ability to manipulate these conventions (e.g. wearing an apron as a skirt). They can also copy incongruous actions directed at objects (such as putting a shoe on the hand) (Hoicka and Gattis 2008). An incongruous action in a playful context can make 18-month-old infants laugh, such as throwing an object on the floor (Esseily et al. 2015). Furthermore, parents report that a one year-old may voluntarily assume unusual body postures or participate in tickling games (Hoicka and Akhtar 2012). From the age of two years, infants become able to understand and produce a wider variety of humor. In particular, they may begin to engage in verbal humor by intentionally producing sentences that do not make sense (e.g. “Dinosaurs eat the wall”) or by making animal sounds that do not correspond to the intended animal (“moo” for a pig, for example), or by using rude concepts (burps, the word “poop”) (Hoicka and Akhtar 2012). At this age, infants are also able to understand jokes involving word substitutions (e.g. “apple” instead of “banana”), however, parents rarely pick up on this type of production before the age of 3 years (Hoicka and Akhtar 2011, 2012).

2.2.3. The functions of humor

Despite the fact that humor is linked to a playful situation, its presence so early in development and across cultures suggests that it serves an important number of functions that have probably contributed to the survival of the species. Martin (2007) groups the psychological functions of humor into three broad categories: (1) socio-cognitive benefits related to the involvement of positive emotions and feelings of joy; (2) the use of humor as a vehicle for social communication; and (3) stress reduction and improved coping skills (Box 2.1).

Martin suggests that some of these functions, particularly those related to emotions, are common to certain animal species with which we share the ability to laugh (primates, dogs and rats), while others developed following the emergence of language and are therefore specific to humans. This is the case for functions related to communication or stress management. In addition to this social function, another function holds in promoting cognitive development. These two functions are not opposed or exclusive but would both be facilitated by the use of humor, and this from the first months of life (Mireault and Reddy 2016).

2.3.1. Observational studies

An observational study consists of observing the spontaneous appearance of a given behavior in the infant’s natural environment without manipulating or intervening in the observation. The observation of humor in infants has often been done through the observation of laughter. Sroufe and Wunsch (1972) followed 150 infants from 4 to 12 months of age, describing the frequency of laughter and the type of stimuli that provoked it. Between 4 and 6 months, the authors showed that infants laughed in response to auditory and tactile stimuli (e.g. saying “boom boom boom”, blowing on the infant’s body, tickling). Between 7 and 9 months, infants laughed mainly in response to visual stimuli (when an adult put a bottle in their mouth, walk like a penguin, crawl like a baby). This pioneering study confirms, on the one hand, that humor is fundamentally social and is created in situations of interaction and, on the other hand, that humorous situations involve the dimension of incongruity. However, and despite the fact that this study did not specifically use the term humor to describe the situations that provoked laughter in the infant, the authors only studied infants’ perception of humor, with the idea that humor would be unidirectional from the parent to the infant, and that the infant would be inactive in this construction. More recently, Reddy (2001) showed that, as early as 8 months of age, infants are able to intentionally produce humor by repeating what she calls “clowning around” to make those around them laugh (e.g. showing hidden body parts). Mireault et al. (2012) showed that these humorous productions are often preceded by “simple clowning” such as grimacing, which would accidentally provoke laughter in others, as early as 3 months of age. These early unintentional accidents allow the infant to understand that some of his or her actions can make social partners laugh and thus maintain a pleasant and positive relationship, which would provoke a repetition of these actions. These observational studies clearly show that humor develops from the first months of life, and this in an interactional way.

2.3.2. Laboratory studies

Laboratory studies are most often experimental and rely on the manipulation of variables to observe a given target behavior. In contrast to observational studies, laboratory studies generally show a later emergence of humor in infants, around 18 months of age (Sroufe and Wunsch 1972). In addition, it can be complicated to induce laughter in an experimental setup. For example, in Esseily et al.’s (2015) study on the effect of laughter on observational learning in 18-month-old infants, less than half of the infants (16/37) laughed in the condition described as humorous. However, the value of these experimental laboratory studies is that they can manipulate specific environmental factors to test the influence of the physical and social context in the emergence of humor and its potential effects. They can also help to better specify humor in relation to other behaviors or mental states, such as lying or making errors. In particular, these studies have shown that, as early as 18 months of age, infants differentiate mistakes from jokes. In a first study, Hoicka and Gattis (2008) presented infants, aged 19–36 months, with either unambiguous humorous actions accompanied by laughter, unambiguous mistakes accompanied by “whoops”, or ambiguous actions that could be interpreted as either a mistake or a joke, and that were accompanied by either laughter or “woops”. The authors then asked the infants to perform these same actions. All infants spontaneously imitated the humorous action initially accompanied by laughter, while they corrected the action considered to be wrong (accompanied by “whoops”). Furthermore, only the oldest children reproduced the ambiguous action when it was accompanied by laughter and corrected it when it was accompanied by “whoops”. In a second study, Hoicka and Akhtar (2011) showed similar results but this time at the language level, in infants aged 30–36 months. These infants mimicked more the production of distorted words by an adult who spoke the infant’s native language but who imitated an accent by accompanying it with laughter, than by an adult who spoke a foreign language and whose infants then corrected the distortedly pronounced words. These studies show that, from the earliest years of life, infants are able to differentiate humor from other mental states such as error, and that the introduction of social cues such as laughter is very important in understanding humor.

Observational and laboratory studies show that humor emerges from the first months of life. The question is: how do infants learn to perceive and produce humor? At what age do they begin to innovate in humor and what is the role of social learning in the development of humor? The contagiousness of laughter, even if it is less well known than the contagiousness of crying in infants, is a phenomenon observed in young children from the age of 2–3 years. Some authors evoke a facilitation of laughter by the presence of peers and more particularly peers who laugh at the same time (Chapman and Wright 1976; Addyman et al. 2018). Other authors even speak of imitative learning, because children laugh at a stimulus even when they are alone but have previously observed a peer laughing at the same stimulus (Brown et al. 1980). Therefore, it would seem that, social learning plays an important role in the development of humor. A survey of parents and an observation study showed that infants imitate their parents’ humorous productions during the first year of life, and that from the age of two years onward they create new humorous situations themselves, although they also continue to imitate them. However, by the age of 3 years, infants usually stop imitating their parents and their humorous repertoire is completely new (Hoicka and Akhtar 2012). Thus, it is clear that social learning, including imitation, plays an important role in the emergence and construction of humor. We will see in section 2.4 how humor can exert an influence on social learning.

2.3.3. Humor and language

Humor in language, or what Nancy Bell and Anne Pomerantz call “language play” (2015), is the manipulation of language in a non-serious way for personal or shared amusement. Playful modifications can occur at all levels of language: phonological, morphosyntactic, semantic and pragmatic. Language play can cross language boundaries by, for example, crossing words from different languages, especially for bilingual children. Language play often provokes laughter and can thus be considered humor. Humor in language is a specific field of study, with corpus analyses mostly carried out by linguists. This research is mainly concerned with second language learning and the effect of humor on it. Overall, these studies show that students learn a second language better when the material to be learned, as well as the teacher’s instructions, contain humor (see Bell and Pomerantz (2015) for a review of the literature). In infants, only one study has looked specifically at humor in language (Del Ré 2003; Del Ré and Morgenstern 2010). In this study, two Brazilian children and one French child were videotaped monthly from the age of 11 months until the age of 6 years, in routine communicative situations with their parents. The researchers identified verbal and non-verbal cues all related to humorous speech: laughter, smiling and language play. The authors observed that it was possible to identify the emergence of productions of shared amusement when the infant was just beginning to speak, as early as 11 months of age. At this age, it is not yet a question of linguistic production but of shared dialogical pleasure where a cry can be amusing and taken up to create humor. Later, humor in language becomes more elaborate. In Del Ré’s (2003) article, the researchers give the example of a 3-year-old child who sings with made-up words. The researcher asks him what language he sings in and the child replies “that one” while sticking out his tongue (language in French being the same word as tongue). The authors also mention the example of another 4-year-old who laughingly tells his mother that he bought his nose in a nose shop and did not choose a black nose so that his skin would not become colored. These language games, like other forms of humor, clearly emerge in situations of interaction and communication and develop within these interactions. It would be interesting to see to what extent these language games contribute to the emergence and development of language in infants.

2.4.1. Empirical studies

When we talk about humor and learning, we tend to think of the teacher’s use of humor and its effects on formal learning, or humor in language. Studies in this area have shown controversial results on the effects of humor on learning, depending on the type of humor used by the teacher, the content of the humorous material and the cognitive function tested (Savage et al. 2017). We will not expand on this literature here as we are interested in the early development of humor and its effect on learning in infants. However, and given the very limited number of studies in infants, we will refer to the classroom humor literature to try to interpret the results in terms of cognitive and social processes and try to understand the underlying mechanisms associated with them.

In infants, learning is ususally not formal and is most often resulting from their daily explorations and social interactions. Social learning is a form of incidental learning that is very present during the first years of life, contributing greatly to the growth of infants’ motor and language repertoires (Esseily et al. 2010). As we have seen in the previous sections, humor is present very early in life. The question that arises is whether infants and children imitate humorous acts more than non-humorous ones. There are two studies in the literature that have looked at the effect of humor on learning in infants, one investigating the use of humor in books and the effect on language development (Hoicka et al. 2008) and the other investigating the effect of humor in social learning of the use of tools (Esseily et al. 2015).

In the first study, Hoicka and Gattis (2008) investigated the link between humor and learning through abstraction and conceptualization in a dyadic reading context. In the first part of the study, the authors analyzed the content of 20 books for infants aged 1–2 years. Humorous situations were the ones that seemed to be the most represented (a little more than half of the books), in comparison to other situations such as mistakes, lies, false beliefs or metaphors. In a second part of the study, the authors looked at whether the humor contained in the books allowed for a greater abstraction of the parents’ language level, with the idea that humor implies an understanding of both the content and the reader’s attitude toward that content. Thus, parents who read funny stories to their infants had more elaborate language, referring more to abstract concepts than when they read other stories. These abstractions can take many forms, ranging from simply naming the objects in the book to relating several names of objects in the book, to using imagination or abstract concepts to refer to an object in the book. The authors observed 15 mother-infant dyads and 5 father-infant dyads in a reading session of a book, One Gorgeous Baby. The infants were between 19 and 26 months of age. The authors showed that when parents read humorous pages of this book, they used a higher percentage of high-level abstraction in their language occurrences, compared to reading neutral pages. By using these high levels of abstraction, parents referred to both social conventions and the violation of these conventions (e.g. when reading a page where the infant throws his diapers on the floor, a parent said, “Do you throw your diapers on the floor?”). These results suggest that the content of a book can change the way parents talk to their children. Thus, reading humorous books can influence infants’ language and cognitive development in a fun and enjoyable setting. Research shows that higher levels of abstraction in parents lead to higher levels of abstraction in children (Kleeck et al. 1997). Higher levels of abstraction also support the development of children’s comprehension and language production. This study shows that humor can support cognitive and language development in infants.

In a more recent study, Esseily et al. (2015) tested the influence of humor during the social learning of a tool use task. In this study, the researchers presented 18-month-old infants with a humorous demonstration, which involved picking up a rake and using it to bring a small, attractive toy that was out of reach. Once the object was caught by the demonstrator, the humorous action was to immediately throw it on the floor with a smile. The action was based on violating social expectations and norms by throwing a retrieved object to the ground instead of manipulating or playing with the object. This demonstration was repeated eight times, after which the infants themselves had the rake within reach and the small toy out of reach. The infants’ actions were coded and compared to those of another group of infants who saw a non-humorous demonstration, which, as in the humorous condition, consisted of catching the toy with the rake but not throwing it on the floor. Among the infants in the humorous demonstration group, it is noteworthy that less than half of the infants perceived the humor and therefore laughed at the demonstration. The main result of this study is that all the infants who laughed in the humorous condition succeeded in imitating the model and catching the small toy using the rake, compared to only 30% of the infants in the non-humorous condition, and 20% of the infants who did not laugh in the humorous condition. Thus, humor has a significant effect on social learning abilities. Several hypotheses can explain these results, and we describe them in section 2.4.2.

2.4.2. Cognitive processes

Because of its nature, the perception and use of humor involve a complex cognitive process. This process was described by Suls in 1972 in two steps: first, the incongruity of the situation must be perceived and then the cognitive problem related to this incongruity must be solved in order to perceive its humorous aspect. In order for this process to result in humor, the incongruity must not be too complex or take too long to resolve: the longer the incongruity takes to perceive, the less funny it is. Thus, the very activity of perceiving or producing humor requires advanced cognitive resources such as attention, memory, confrontation of the situation with reality and previous knowledge. This allows not only to resolve the incongruity but also to mobilize the cognitive functions necessary for learning. These functions can be mobilized during any situation of violation of expectations that causes surprise (Stahl and Feigenson 2015, 2017). However, humor facilitates learning through two other important factors: (1) a facilitator of social interactions, which is particularly useful during social learning tasks (such as the one described in the previous section) and (2) positive emotions which, as we will see later, promote learning. We will first describe the cognitive processes related to humor, and then the other two factors.

2.4.3. Physiological processes

The positive effects of humor can also be observed at the physiological level, notably with its known antagonistic effect on stress, but also with possible effects on the cardiovascular, immune and endocrine systems (Berk et al. 1989; Lefcourt et al. 1990; Fry 1992; Fredrickson and Levenson 1998; Bennett et al. 2003). The positive emotional experience elicited by humor is directly related to the activation of the brain’s reward system, which becomes more active the more amusement the individual experiences (Mobbs et al. 2003; Amir et al. 2015). Specifically, studies have shown an increase in hemodynamic (blood pressure and flow-related) signaling in the dopaminergic reward system, a system known to play a crucial role in motivational and addictive behaviors, including drugs (see Schultz (2002) for a detailed review of the dopamine-reward link).

Humor will also modulate brain activity in several other cortical regions. As discussed earlier, humor involves the ability to perceive the absurdity or at least the unexpectedness of a situation, which provokes a violation of our expectations. In adults, Amir et al. (2015) showed that the anterior cingulate cortex is activated when subjects observe humorous drawings. This brain area also seems to play a role in detecting errors or contradictions (Botvinick et al. 2004; Brown and Braver 2007; Shenhav et al. 2013). These same authors also observed the activation of certain association areas, whose role is to converge different information. Amir et al. (2015) interpret these results by the need for a humorous situation to appeal to the ability to make “remote associations”, that is associations between several ideas or concepts that are not initially obvious and that help establish the absurd aspect of a situation. The authors give the example of one of the images they used in their experiment, which depicts a pig looking at the titles and covers of books in a library – the absurdity of the situation lies in the fact that pigs do not read. The novelty of the image, the violation of expectations and/or the rejection of the link between concepts when the image is considered absurd will provoke the activation of the association areas, and the maintenance of this activation when the subject must then link this image to a context (e.g. a text associated with this image, as in the case of this experiment).

Other studies have shown activation of the temporo-occipital, which processes visual information, in situations involving incongruity detection (Iwase et al. 2002), semantic processing of jokes (Goel and Dolan 2001) and identification of visual cues to emotion (Geday et al. 2003). However, this activation also includes the left fusiform gyrus, a region that induces laughter as well as a sense of joy, a positive emotion when experimentally stimulated (Arroyo et al. 1993). These studies seem to converge toward the possibility that this cortical region is involved in processing the incongruous or surprising aspect of a joke or funny situation, and could therefore play a crucial role in the cerebral processing of humor.

2.4.4. Positive emotions

The cognitive and physiological benefits of humor can be at least partially linked to the positive emotions associated with it. These play a role in learning, through various mechanisms such as stress reduction, improvement of social affiliation (Somogyi and Esseily 2014) and prosocial behaviors (Isen 2003), memorization and reasoning processes, cognitive flexibility, action planning (Isen 2003) or the activation of the dopaminergic reward system already mentioned in the previous section (Ashby et al. 1999). To date, we do not know whether this system is already activated in this context in infants, but numerous studies have shown the effects of positive emotions on infants’ performance in various social tasks. In this section, we will not provide a comprehensive literature review on the effects of positive emotions on learning, but we have selected some studies in infants or children that could explain the results observed in the study by Esseily et al. (2015).

2.4.5. Positive emotions, humor and curiosity: toward a learning model?

Most of the processes mentioned in this section on the effects of humor and positive emotions in learning can be found in a recent model proposed by Murayama et al. (2019). This model involves a new element, curiosity, as the main driving force of autonomous individual learning. Studies proposing that curiosity plays a crucial role in learning are not recent (Berlyne 1978), but they have long sparked debate about how to define curiosity. The model proposed by Murayama et al. (2019) avoids this pitfall by focusing not on curiosity itself, but on the different elements that are involved in the curiosity-driven learning mechanism, which would be the starting point for autonomous learning (Box 2.3).

This model is particularly interesting for humor because it involves several of its components: surprise, linked to the discovery of something new that we will be able to integrate and memorize, the emotional valence of this new information (which can be positive, negative or neutral – a positive valence being more of a source of motivation to learn), and a reward system. The positive effect of humor observed on certain learning processes fits perfectly into this model since surprise is at the center of the humorous process. The latter will therefore add value to the information acquired in a humorous context compared to a neutral context – but only if the humor is used appropriately and provokes a positive emotion. Indeed, it is known that the use of certain types of humor, or its inappropriate use, can have negative effects on learning (see Banas et al. (2011) for a detailed review).