Reading Time: 14 minutes

Pretending to pretend: 

The trickster’s mind in animals

Carlos H. Guzmán

University of Tartu, Department of Semiotics

[PDF]



Abstract. The ‘trickster’s mind’ is a notion that I have proposed to present a set of cognitive skills that allow human agents to deceive other human and non-human agents by generating false beliefs that make them behave in ways that contribute to fulfilling the deceiver’s agendas. I propose that these cognitive skills have been represented in cultural narratives under the figure of the trickster which, although under different guises, maintains some common elements such as marginality, ambiguity, and playful ingenuity. In this article I analyse how the seeds of this trickster’s mind could be observed in non-human animal behaviours. I will also review some ongoing debates about issues such as the ability of non-human animals to understand false beliefs in others, which are the basis for more complex abilities such as creative deception.

Keywords: trickster, trickster mind, deception, animal behaviour, theory of mind, false beliefs, cognition 

Teesklemist teeseldes: triksteri mõtlemine loomadel

Käesolevad artiklis pakun välja mõiste ‘triksteri mõtlemine’, millega kirjeldada kognitiivsete oskuste kogumit, mis võimaldavad inimestel petta inimesi ja mitte-inimestest loomi, tekitades valesid uskumusi, mis panevad neid käituma viisil, mis aitab kaasa petja plaani täitmisele. Pakun välja, et neid kognitiivseid oskusi on kultuurinarratiivides esindatud triksteri kuju all, mis, ehkki erinevatel kujudel, säilitab mõned ühised elemendid, nagu marginaalsus, mitmeti mõistetavus ja mänguline leidlikkus. Selles artiklis analüüsin, kuidas selle triksteri mõistuse algeid võib täheldada mitte-inimestest loomade käitumises. Pakun ülevaate ka mõnedest käimasolevatest aruteludest teemadel, mis käsitlevad mitte-inimestest loomade võimet mõista teiste valeuskumusi, mis on aluseks keerukamatele võimetele, näiteks loomingulisele petmisele. 

Märksõnad: trikster, triksteri mõtlemine, petmine, loomade käitumine, vaimuteooria, valed uskumused, kognitsioon


The search for signs of awareness and social-cognitive skills in human and non-human animals puts the capacity to develop a functional theory of mind as a key element that must be present for an organism to develop such complex capabilities; although there is evidence to support the idea that some non-human animals like chimpanzees and elephants have a basic understanding of other agents’ actions in terms of underlying goals and intentions (Mitchell, Thompson 1986). When it comes to the point of detecting the ability to understand false beliefs in others there is no experimental data to confirm that non-human animals possess this ability (Call, Tomasello 2008). 

In my thesis The Trickster’s Mind (Guzmán 2017), I propose the idea that the trickster figure, present in most of the cultural traditions around the world, is a way that people have represented and nurtured in new generations a set of cognitive skills that are related to the ability to understand false beliefs; and to intentionally generate false beliefs in others to make them act in ways that advance the goals of the deceiver. This ability leads to the notion of creative deception, that although not observed in other animals is the product of a set of cognitive tools that have been acquired and perfected through evolution.  

The use of traps, disguises, distractions, and other deceptive tactics, responds to a biological necessity to provide organisms new ways to survive and prosper in an ongoing hostile environment. This kind of trickster behaviour has introduced a new element in the logic of survival that deeply modifies the relations established, for example, between predators and prey (Guzmán 2017), “trickster feeds himself where predator and prey meet, but rather than entering the game on their terms he plays with its rules” (Hyde 1998: 24).     

The main purpose of this article is to propose the existence of a ‘trickster’s mind’ in animals, related to a capability to ‘pretend to pretend’ that in a rudimentary way can generate false beliefs in other agents. For this, I take as a starting point the texts of Donald Griffin on the topic of animal awareness, the perspective proposed by Daniel Dennett in the book Kind of Minds, and the critique of Dennett’s ideas by Cary Wolfe from his text about Posthumanism. This approach is an opportunity to refine the ideas that I have presented in my thesis related to the role of deceptive strategies in the emergence of a trickster’s mind.

Can animals lie creatively?

Deceptive animal characters are a common element present in many traditional narratives: Anansi the spider from African folktales, the characters of Raven and Coyote in Native American narratives, Kitsune the fox of Japanese Shinto lore and its counterpart Reynard the Fox from European medieval fables, are just some examples of trickster animal characters. A probable reason for this is that humans have observed, copied, and improved the deceiving behaviours of many animals and have translated these observations and learnings into the narrative figure of the trickster, who is a reflection of the cunning skills that animals, including humans, possess. 

In some way humans have recognised a variety of skills for deception in non-human animals, however it is important to have a clear definition of what is understood as deception and creative lying before we can attribute these traits to non-human animals. For this, I start with the definition proposed by Lilly-Marlene Russow on what constitutes deceiving behaviour: “An agent’s behavior is deceptive if and only if the agent intends that, because of its behavior, another organism will come to have (and perhaps act on) a false belief.” (Russow 1986: 48). 

Following this definition, the actual position is that non-human animals cannot deceive because that behaviour requires the understanding and intended use of false beliefs for which, as stated before, there is no compelling experimental evidence. This understanding of a false belief is described by Call and Tomasello as:

the special case in which an observer predicts or explains the behavior of an actor based on a judgment of what that actor believes to be the case, not what really is the case as the observer knows it. (Call, Tomasello 2008: 189) 

To achieve this level of understanding, the agent must first be able to understand other agents’ goals and intentions, so they can react not only to what others are doing, but also to anticipate what they will do. The big problem, as stated by Call and Tomasello is to discern when an observer is reading another actor’s behaviour and when it is reading the other’s goals.    

In this sense the example that Griffin provides about monkeys and apes trying to hide food from more dominant members of the group is difficult to interpret without postulating at least short-term intentions and plans (Griffin 1981: 96). Griffin defies the idea that animals use signs, but they do not know that they are signs, as in the case of dancing bees that devoid of intention cannot use their dance to lie to other bees about the location of a food source. This leads to the notion of prevarication, understood as the intentional use of communication signals to convey information known to be inaccurate, and this notion was added by Thorpe (1974) to the original list of design features formulated by Hockett and Altman (1968) that qualitatively distinguished human language from animal communication.

A simpler explanation for some deceptive behaviour in animals that could be understood as prevarication, is based on the idea that some animals act based on knowledge gathered through the observation of the behaviours of other animals. On this account, we have the example of the dog that lures her master out of her favourite chair by standing by the door as if wanting to go out to steal the spot as soon as the person stands up, presented by Dennett (1978: 274–276). Dennett argues that the dog by repeated observation has learned that standing next to the door makes the human stand up and does not require any intention to generate a false belief in the human.

In this example we can predict that once the human figures out the ruse, it would no longer be effective, and the dog will be limited to repeat this movement till she understands that it is no longer working. If the dog is not able to change its repertoire to lure out the person from the chair, that would mean, as Hyde explains, that the dog cannot lie creatively (Hyde 1998: 46). But as Russow points out, to accept Dennett’s interpretation as adequate, the intention of the dog, that could be expressed as “I want to run to the door to make the human get up so I can sit on the empty chair” must be related to a belief that the dog has come to have based only on the experience acquired by the repetition of random acts. Acts such as going to the door until she finds out that this triggers the response in the human to get up and open the door without ascribing to this behaviour any communicative intention, and that idea, especially for many dog owners, would be very strange.

Following this line of thought, Dennett says that “many animals hide but don’t think they are hiding. Many animals flock but don’t think they are flocking” (Dennett 2008: 119), and I would dare to add in relation to the dog example the idea that many animals lie but don’t think they are lying. This ‘intelligent but unthinking’ behaviour proposed by Dennett denies the possibility of animals manifesting a ‘third-order intentional system’, the capability of an agent to want other agents to believe that it wants something, even if that is not true, in other words pretending to pretend.  

For Wolfe, Dennett’s scheme is problematic on two accounts: first, the idea of ‘represented knowledge’ that Dennett considers necessary to ascribe conscious thinking to a creature, depends on the assumption that language can provide a user-illusion of intrinsic intentionality, that Dennett has disavowed; and second, this “conceptual and phenomenological restabilization of the subject by means of language […] forms an ontological specificity that is no different in principle from the Cartesianism Dennett rejects” (Wolfe 2010: 38).

The first problem is related to Dennett’s argument that the informational unification required for a human type of consciousness is not part of the organism’s ‘hard-wiring’, in other words, it is not imprinted in our genetic code, but it is mostly a product of our immersion in human culture. Following this argument, Dennett says that “early education produces in us a sort of benign ‘user illusion’” (Dennet 1995: 702). New-born humans and non-human animals do not have this illusion because “there’s no user in there to be fooled” (ibid). 

Going back to the dog example, the lack of a ‘user illusion’ is what would impede the dog to formulate a though on the line of “I’m going to run to the door to make my owner believe that I want to go out and then sit on my favourite spot when he stands up”. The dog, under Dennett’s perspective, doesn’t have what it takes to develop second-order intentional systems: beliefs and desires about beliefs and desires, its own or those of others: (“I can express to my owner my desire to go out by standing on the door”) much less, to have a third-order intentional system: (“I can make my owner believe that I want to go out, when in fact what I desire is to sit on my favourite chair”). Without this insight, the dog should not be able to come up with new ruses to make his owner stand up from the chair, about this Dennett argues that “such virtuosity in a dog would be highly implausible” (Dennett 1978: 276). 

Mitchell and Thompson’s (1986) observational analysis of different play interactions between humans and dogs allows us to move from the mere anecdote as related by Dennett’s to actual examples where it is possible to contrast his posture with what is actually happening in these dynamics between species.

Among the interactions described in Mitchell and Thompson’s study, we can focus on a couple of interactions that involve some deceptive behaviours between a human and a dog. In the first one, a woman pretends that she is going to throw a ball for the dog to catch, as she had done several times before, but without actually throwing it; in the second a dog drops a ball close to the reach of the human but takes it before the human can grab it. 

In the first scenario, the woman pretended that she was about to throw the ball far, and the dog predicted that this action would result in the ball’s landing somewhere behind him, so the dog turned and waited for the ball to land (Mitchell, Thompson 1986: 197), but the ball never leaves the woman’s hand, so the dog was fooled. If we apply Dennett’s interpretation to this, we can say that through previous experience the dog has learned that certain movements from the human indicate that the ball will fly and land at a certain distance and he gets ready to catch it. When the expectation is not fulfilled, the dog has no way to understand that he has been fooled because without a second-degree intentional system he cannot have beliefs about the intentions of the woman, including the intention to deceive. Yet, when the woman tries to do the trick again, the dog, instead of turning around, keeps his gaze on the ball. Mitchell and Thompson argue that this behaviour is consistent with an ability to avoid deception by “focusing on aspects of the deceiver’s behavior the deceiver cannot fake” (Mitchell, Thompson 1986: 200).

 More interesting is the case of the dog as the deceiving actor in the game in the last case described: 

Dog moves closer to Woman and shakes his head, letting the ball drop in front of W; D maintains his gaze on the ball. W backs up, and then moves forward to get the ball but, as she does, D moves toward the ball and grabs it in his mouth just as W reaches for it. D runs, with the ball, away from W. (Mitchell, Thompson 1986: 202)

In this example, trying to explain the dog’s behaviour by saying that he is just a good behaviourist does not seem to justify the dog’s belief that the person will try to grab the ball. The dog is not just dropping the ball he is making an ostentatious display that the ball can be taken by the woman, and here we can find a situation similar to the story quoted by Lacan: “Why do you tell me that you are going to X in order to have me believe you are going to Y whereas you are indeed going to X?” (Lacan 1977: 173). Or in this case, why does the dog pretend that he is not going to take the ball in order to make the woman believe that she can take it if at the last moment he will run away with the ball? The dog seems to be pretending to pretend, something that under Dennett’s behaviourist explanation is not possible. To this I would add the element that the dog is not doing this in response to some basic survival instinct but just for fun, something that would lead to a well-deserved analysis on its own.  

For Mitchell and Thompson, this kind of gameplay is evidence that both “dogs and people recognise the directionality of each other’s projects and actions; if they did not, they would not be able to predict each other’s actions” (Mitchell, Thompson 1986: 202). How much of this behaviour can be claimed to be based on an actual intention to manipulate each other’s mental states, particularly in the case of the dog, is a question that the authors leave open until further evidence is found.       

What I propose is that there should be a fundamental qualitative difference between the kind of ‘ritual’ behaviour that can be observed in experiments like Skinner’s (1992) ‘Superstition’ experiment with pigeons, where an operant conditioning is produced in hungry animals when food is delivered at regular intervals, and the kind of behaviour described in Mitchell and Thompson’s examples of games between humans and dogs.

Dennett’s position is that even if the dog’s actions look very much like true second-order interacting, meaning that they have “beliefs and desires about beliefs and desires” (Dennett 2008: 121), and if we assume that to the dog his master is just a machine activated by different actions, something very similar as the way the pigeon sees the automatic food dispenser,  then “we will have just as good a predictive ascription, more modest but still, of course, intentional” (Dennett 1978: 274). This view contradicts Mitchel’s affirmation that the deceptive actions in the described games between humans and dogs would not be possible unless dogs and people view each other as agents with intentions and not just as machines that react in a fixed predictable way. This also leads to new level related to the issue of social cooperation and social manipulation.   

Social cooperation and manipulation in human and non-human animals

The emergence of social behaviours in animal groups is one of the most intriguing evolutionary adaptations, this could include schools of fish that move in patterns with the purpose to disorient predators, insect colonies with clearly established functions for different members, and hierarchical packs of mammals with different strategies to avoid predators or, conversely, to catch prey.

In the case of large fish schools or a flock of birds, the level of communication and understanding between members of the group appears to be simple, but from simple rules of behaviour impressive group patterns could emerge. In respect to insect colonies, as pointed out by Griffin, there is, as in the case of honeybees “a great deal of communication among the members of a hive” (Griffin 1981: 78) who are extremely interdependent and have developed a complex communication system that allows the colony to perform cooperative activities such as looking for food, selecting a new cavity to migrate to, and regulate the sex and growth of developing larvae. The communication systems used by social insects meet with many of the design features formulated by Hockett and Altman (1968) and Thorpe (1974) including semanticity, arbitrariness, discreteness, displacement, and productivity.

In social mammals, the complexity of communication systems reaches a whole new level, as emerging cooperative interrelations require a continued balance between individual and collective goals. 

For Michael Tomasello (2016: 9) a complex social life is based on achieving a balance between cooperation and competition. In animal societies this balance can be represented by two axes (figure 1), “a horizontal axis of cooperation based in individuals’ propensities (high or low) for affiliating with (or even collaborating with or helping) others of their kind, and a vertical axis of competition based in individuals’ power and dominance (high or low) in contesting resources.” (ibid, 9)

Figure 1. The two dimensions of social life for complex organisms (Tomasello 2016: 9)

This kind of balance can be observed in social groups of chimpanzees and bonobos. Although they are built for competition, they can develop cooperative activities such as foraging for food or defending against attacks from other groups. Yet they still compete for food and even fight for access to females (ibid, 21). In these circumstances, individual chimpanzees and bonobos perceive others as intentional, decision-making agents with whom they must compete (ibid, 22). This leads to different types of deceptive behaviours that could involve communicative gestures used as displays of strength, gratitude, submission, etc., even incorporating objects as part of their communicative efforts as in the case observed by Goodall of the use of kerosene cans by a chimpanzee to generate additional noise with the intention to enhance his dominance display (Griffin 1981: 76).

Cooperative behaviour in animals opens the door to another kind of interaction: social manipulation. The research on this topic, with experimental studies on rhesus monkeys and baboons by Delgado (1963) and Beck (1973) is still inconclusive, but there are observed cases where a chimpanzee can use social skills and some deceiving skills such as exaggerated displays of pain to gain the support of other members of the group to punish an opponent (de Waal, van Hooff 1981). More recent research has revealed that some primates can develop deceptive tactics when using other species members to look for concealed food, as in the experiment with mangabeys’ monkeys by Coussi-Korbel (1994) and the experiment to examine chimpanzee understanding of others’ status of knowledge and ignorance in free interactions without human control (Hirata, Matsuzawa 2001). 

The trickster’s mind between animal instinct and human rationality

After this brief overview of the conflicting ideas about animal intentionality and the capacity or lack of understanding of false beliefs in other agents, I return to my hypothesis that some animals behave as agents with the ability to conceive and enact deceiving schemas in a way that cannot be explained as mere instinct-driven or as a result of a basic capacity for detecting other agent’s behaviour patterns. Instead, I propose that this ability for intentional deceiving leads to consideration of the actual possibility that some animals are more than just thoughtless agents that lie without knowing that they are lying. That they can perceive other agents as individuals with particular beliefs and desires that can be manipulated to obtain some kind of advantage.     

This points to a different kind of behaviour, in some cases improvised, where the mechanisms of opportunity are not activated as an instinctive program but are selected among a diverse and increasing repertoire and used at will to respond to specific situations and, more importantly, to the actions and responses of other animals. This ability to improvise tricks and manipulate other organisms with a previously established intention is what I have linked to the notion of creative intelligence, a fundamental component for the emergence of a trickster’s mind in animals, including humans (Guzmán 2017).

The representation of the trickster as an animal in different narratives can be interpreted as Robert Pelton suggests, as a way to represent the human creative intelligence and the transforming power of imagination (Pelton 1993: 130). Something that humans have inherited or learned by watching other animals, also the animal-human duality of the trickster figure can be a representation of the process leading to the emergence of human culture from the unconscious natural world of animals.

References 

Call, Josep; Tomasello, Michael 2008. Does the chimpanzee have a theory of mind? 30 years later. Trends in Cognitive Sciences 12(5): 187–192.

Coussi-Korbel, Sabine 1994. Learning to outwit a competitor in mangabeys (Cercocebus torquatus torquatus). Journal of Comparative Psychology 108(2): 164–171.

Delgado, José M. R. 1963. Cerebral heterostimulation in a monkey colony. Science 141(3576): 161–163.

Dennett, Daniel C. 1978. Brainstorms. Montgomery, VT.: Bradford Books.

Dennett, Daniel C. 1995. Animal consciousness: What matters and why. Social Research 62(3): 691–710. 

Dennett, Daniel C. 2008. Kinds of Minds: Toward an Understanding of Consciousness. New York: Basic Books.

Griffin, Donald Redfield 1981.  The question of animal awareness: Evolutionary continuity of mental experience. New York: Rockefeller University Press.

Guzmán, Carlos H. 2017. La Mente del Trickster. Master’s thesis from Universidad de Bogotá Jorge Tadeo Lozano, Bogotá D.C., Colombia. Available from: https://expeditiorepositorio.utadeo.edu.co/handle/20.500.12010/2672; accessed on 27 October 2020.

Hirata, Satoshi; Matsuzawa, Tetsuro 2001.  Tactics to obtain a hidden food item in chimpanzee pairs (Pan troglodytes). Animal Cognition 4(3–4): 285–295.

Hockett, Charles F.; Altmann, Stuart A. 1968. A note on design features. In: Sebeok, Thomas A. (eds.), Animal Communication: Techniques of Study and Results of Research. Bloomington: Indiana University Press, 61–72.

Hyde, Lewis 1998. Trickster Makes this World: Mischief, Myth, and Art. New York: Farrar, Straus and Giroux.

Lacan, Jacques 1973. Écrits: A Selection. London: Tavistock Publications.

Mitchell, Robert W.; Thompson, Nicholas S. (eds.) 1986. Deception: Perspectives on Human and Nonhuman Deceit. Albany: State University of New York Press. 

Mitchell, Robert W.; Thompson, Nicholas S. 1986. Deception in play between dogs and peopleIn: Mitchell, Robert W.; Thompson, Nicholas S. (eds.), Deception, Perspectives on Human and Nonhuman Deceit. Albany: State University of New York Press, 193–204.

Pelton, Robert D. 1993. The trickster in West Africa: A study of mythic irony and sacred delight. In: Hynes, William J. (ed.), Mythical Trickster Figures: Contours, Contexts, and Criticisms. Tuscaloosa: University of Alabama Press, 122–140.

Russow, Lilly-Marlene 1986. Deception: A philosophical perspective. In: Mitchel, Robert W.; Thompson, Nicholas S. (eds.), Deception, Perspectives on Human and Nonhuman Deceit. Albany: State University of New York Press, 3–40.

Skinner, Burrhus F. 1992. “Superstition” in the pigeon. Journal of Experimental Psychology: General 121(3): 273–274.

Tomasello, Michael 2016. A Natural History of Human Morality. Cambridge, MA: Harvard University Press.