|Acting under uncertainty: Ritual actions and predictions of success in pragmatic endeavors
Jesper Sørensen, PhD
Institute of Philosophy, Education and the Study of Religions
University of Southern Denmark
In this paper I will discuss a number of possible theoretical models aimed to explain the relation between the performance of stipulated ritual actions and the execution of actions with uncertain result. The British-Polish anthropologist Bronislaw Malinowski noted the cultural ubiquity of such performances already in the early 20th century. However, so far no satisfying theory has been produced that explains why ritual performance are performed in relation to particular types of endeavors (and not others) as well as why they should have an effect on subsequent behavior. Based on results from evolutionary psychology and cognitive models of human action representation, I will propose a number of hypotheses that might help explain this phenomenon.
Keywords. Ritual actions, Magic, Uncertainty, Action representation, Cognitive psychology, Evolutionary psychology.
Almost 100 years ago, the British-Polish anthropologist, Bronislaw Malinowski conducted the field study that would earn him a position as one of the founding fathers of modern social anthropology. Delving into almost all aspects of the lives of the Trobrianders – the inhabitants of a small group of coral atolls northeast of Papua New Guinea – Malinowski paid particular heed to magical and religious behavior. As such, he was in line with and perhaps a product of the early phases of anthropology where magic, religion and questions concerning the alleged rationality or non-rationality of so-called primitive people around the globe were looming large. In contrast to other prolific writers in this subject area, such as Tylor, Frazer, Marett, Durkheim, Lévy-Bruhl and Mauss, Malinowski did not disengage the ritual practices from the immediate context. In line with his psycho-functionalist approach, he argued persistently that ritual practices must be understood in their local pragmatic context rather than as disentangled propositional statements. Malinowski claims that all behavior aims, more or less directly, to enhance he survival of the organism. Besides the direct function of range of physical actions (eating, reproducing, excreting etc.) as well as the derived function of social institutions (family, master-apprentice relations etc.) that ensures the transmission of vital instrumental and cooperative knowledge, the question arises why people engage in religious behavior, i.e. why humans tell mythical narratives and perform religious and magical rituals.
Whereas Malinowski found the function of myth and of religious ritual in their explanatory and social integrative effects respectively, his pragmatist approach led him to focus on the close connection between the execution of instrumental actions and magical rituals. In short, Malinowski gave three functional accounts of the existence of magic. First, some actions have an uncertain result, either by nature or due to limited technology, and in a sense magical rituals take over where technology ends. Second, besides this pragmatic explanation, Malinowski presented a cognitive account as well. Accordingly, human are able to mentally represent possible future outcomes and are therefore able to represent potential failure. In order to alleviate the resultant disruptive feelings of anxiety, magical rituals install a sense of security in the participants by engaging supernatural means of protection and assistance. Third, as predominantly social actions, magical rituals help coordinate communal pragmatic actions by specifying particular temporal constraints (e.g. when to commence the planting of the yams gardens) as well as allocating authority. Thus, as the case of myth and religious ritual, magic has an integrative function as it ensures the performance of biologically necessary and historically tested ways of acting. Magical rituals protect vital and culturally specified forms of behavior by directly addressing specific sociological and, in particular, psychological threats to individual motivation.
By embedding magical actions in this pragmatic context, Malinowski not only disengaged its performance from notions of primitivism, pre-logical mentality and ‘childishness’. He also pointed out that no all actions are related to magical rituals. When engaging in poison fishing in the safe lagoon, no Trobriander performs magical ritual. In contrast, when embarking on deep-sea shark fishing, magical rituals are performed in order to ensure the safety of the sailors and an abundant catch.
To summarize, Malinowski’s had two important observations: First, people perceive some actions as more risky and uncertain than others and risky actions are likely to entice adjacent ritual behavior. Second, the performance of ritual actions has an effect on the likelihood of peoples’ representation of success and thus the likelihood of performing such actions (for a more thorough discussion of Malinowski’s theory of magic, see Sørensen 2007).
Even though Malinowski’s arguments concerning the relation between pragmatic and ritual actions are often favorably discussed in reviews, few if any study have set out to investigate more precisely the relation between judgments of uncertainty and risk and the performance of ritual behavior. One reason might be that scholars have simple noted that there is such a relation and found it evident why this is the cases, i.e. that when people experience uncertainty they seek the assistance and reassurance of superhuman agents. This interpretation, however, presupposes not only belief in such agents, but also that (a) the performance of the ritual is somehow motivated by this belief, and (b) that representations of such agents is present while engaging in ritual behavior. As we shall see below, both of these assumptions are problematic and more simple hypotheses might be posed to account for this relation. Another, and complimentary reason is the rejection of functionalism and behaviorism following Malinowski’s death in 1942 and the subsequent focus on ritual ‘meaning’ in contrast to its immediate pragmatic effect. When rituals are conceived of as primarily transmitting semantic content, the job of the scholar becomes one of hermeneutic encoding ritual elements in relation to s cultural symbolic system. This encoding of meaning might, subsequently be used to explain why a ritual is performed in a particular setting, e.g. to imbue ‘meaning’ to a situation, but in this approach the more direct relation between performing ritual actions and engaging in risky behavior was lost.
The remained of this chapter therefore aims to address two problems originally framed by Malinowski but, of course, reframed in light of recent research: First, what types of actions seems to entice people to perform ritual performance? Second, if ritual behavior really affects people’s representations of success in uncertain and risky behavior, why is this the case? This discussion will lead to a more speculative part posing questions about the possible feedback relations when stipulated ritualized behavior is repeatedly related to particular types of pragmatic behavior.
Judgment Theory and Evolutionary Psychology
But what count as acting under uncertainty? What is a risky action? Economists, psychologists, sociologists have all investigated human actions understood as chancy or uncertain. Most directly relevant for the current investigation is the psychology of “judgment and decision making” that, for two centuries have investigated how humans make particular types of decisions and, in particular, how they ought to make decisions. I will not venture into a long review of this research (see Goldstein & Hogarth 1997).
According to Preferential Choice Theory, humans act in order to enhance utility understood as a relation between probability and subjective assessment of chance of success.
According to Judgment Theory more focus should be directed at the type of information stemming from the environment and the psychological mechanisms underlying the preferential processing of some and not other parts of this information.
Critics of implicit Olympian agents in these approaches have instead argued that humans use a number of heuristic devises when making judgments under uncertainty rather than probabilistic judgment (Tversky & Kahneman)
Critique of “game metaphor” in judgment theory that translates all decisions made under uncertainty into monetary games that can be controlled in the experimental setting.
Instead renewed focus on the potential role of domain-specific cognitive processing in making judgments under uncertainty (Goldstein & Weber 1997).
The role of domain specific cognitive processing brings recent hypotheses made by Evolutionary Psychology into focus.
Human cognitive architecture is to a large extent a result of an evolutionary process.
Therefore, cognitive systems are adaptations to a number of species-specific adaptive problems.
The overall purpose of these systems is to extract relevant information from the environment and execute adequate behavioral responses.
Humans have a large number of such prefixed learning systems
Domain specific processing are adaptations to an ancestral environment and need therefore not lead to adaptive responses in current day environment.
Some types of information (say of predators, potential mates or sources of nourishment) are highly relevant to our cognitive systems, whereas other stimuli are not.
Together these lines of research indicate:
That some types of information and therefore situations of judgment should be highly relevant to our evolved cognitive system while others are not.
That different types of information of information will be processed by different cognitive systems – therefore we cannot equate all situations of uncertainty.
That judgment of likely success and assessments of own emotional reaction to both success and failure will depend on actual domain of processing.
In relation to ritual action it lead to the hypotheses:
That some types of information relevant to particular cognitive domains are more likely to result in the performance of ritual actions.
That the performance of ritual will effect actions performed within distinct domains differentially.
Further, we need to take into consideration the temporal unfolding of the uncertain action. Whereas the game metaphor might adequately map the temporal unfolding of the bifurcation structure in pure win-loose situations, many uncertain situations are not clear-cut but will expose both graded and hierarchical structures.
Bifurcation: either a or b (e.g. lottery coupon)
Graded structure: result somewhere between a and b (e.g. agricultural yield)
Hierarchical structure: result depends on the success of a number of sub-actions (e.g. agriculture)
This points to the fact that most actions form part of larger event-frames, i.e. cognitive representations of how the execution of a number of bodily actions can lead to the attainment of some pre-specified goal.
Human Action Representation
The cognitive systems organizing human action representation is modeled in figure 1:
The ultimate intention specifies a goal of an action, say buying bread. The action sequence itself consists of a large number of sub-actions, whose execution will lead to the goal, e.g. taking money and putting it in my pocket, getting in my car etc.) All of these are determined by a proximate intentions and are temporally and causally related to each other (e.g. taking the car key before getting into the car).
Ritual action introduces a number of changes into this model:
Figure 2 is a model of how rituals modify action representation system: Sub-actions with proximate intentions are stipulated into a sequence with minimal causal specification (but a strong temporal one). Ultimate intention underlying the concrete performance of a particular ritual has next to no effect on the structure of the action itself (intentional under-specified).
Question 1: Will performance of a ritualized action sequence alone produce an enhanced representation of success?
By activating the action representation system in a ritual, all sub-actions are connected and lead to their pre-specified result. This might have a contagious effect on representation on the action performed under uncertainty, i.e. with an uncertain relation between the proximate actions and the attainment of the ultimate goal. Rituals thus condense long and unpredictable actions sequences (hierarchical) or those imbued with large uncertainty due to factors outside the agent into an action sequence with certainty of success. Notice that he ritual is itself intentionally underdetermined: Both actions under uncertainty and ritual action expose lack of agentive control, ritual by being stipulated, uncertainty by lack of control. But in the case of ritual the outcome is certain, in contrast to the pragmatic endeavor.
Question 2: Will involvement of particular objects and actions enhance this effect?
Ethnographic observation points to the involvement of particular types of actions and objects in ritual actions. These are often related to the ritual’s target-domain by due to similarity of contagion (the often noted characteristics of ‘magical rituals’).
Question 3: Are such involvement necessary to produce effect?
In order to stabilize, ritual actions must produce other effects that enhance the likelihood of their performance in concert with particular types of pragmatic actions.
Feedback loop 1: Conditioning
Skinner argued that superstition would arise due to the spatio-temporal contingency of a particular action pattern and its subsequent (but causally unrelated) success.
This might explain some of the stabilizing effects found in the historical and ethnographic record, but it is better understood as a feedback loop, as it does not explain the emergence of ritualized behavior in relation to pragmatic pursuits to begin with. Further, in order to work, success has to follow immediately after the ritual action sequence, and this is in most cases very far from natural condition (e.g. ’farming magic’, where result is not known for many weeks).
Feedback loop 2: Superhuman agents
I have argued elsewhere (Sørensen 2007) that the performance of ritual action in itself will make representations of superhuman agents more cognitively relevant. This is so, because representations of gods, spirits and ancestors explain why (a) we perform exactly those and not other actions (solve the causal under-specification) and (b) how the actions relate to the result (solve the intentional under-specification). This lead to the third question:
Question 3: Will performance of ritual action in relation to semi-random or random events enhance representations of superhuman agents intervening in these?
Feedback loop 3: Demons, witches and other malignant forces
If the performance of ritual actions has the (side-)effect of imbuing the whole pragmatic action sequence with superhuman protection this is likely to result in representations of potential failure as a result of equally superhuman but malignant agents. Failure is then explicable as the result of actions performed by other agents and this is likely to influence judgment under uncertainty.
Question 4: Will representations of intervening superhuman agents enhance representations of malignant agents responsible for pragmatic failure?
If this question finds a positive answer this will have the rather ironic effect of reintroducing a new level of uncertainty – in relation to beneficent and malignant superhuman agents – and this transform the most important variables of pragmatic success into questions of morality and ritual behavior.
A central question following this line of inquiry concerns whether ritual behavior has adaptive value. Two claims will be made:
Claim 1: Enhanced representations of likely success stimulate pursuits of pragmatic goals with inherent uncertainty and risk.
Claim 2: “Insurance paradigm”: Ritual actions result in the spread of responsibility of the pragmatic actions performed to both other humans (ritual participants) as well as to superhuman agents.