Anthropology of Religion
December 1, 2003
Animals make an appearance in many of our accounts of African Witchcraft beliefs – vultures snakes, hyenas, vampire bats, etc. Why is it that animals and witches seem to be so thoroughly intertwined in culture after culture? Use Boyer’s evolutionary approach (and other approaches, too, if you wish) to try and explain this phenomenon.
The most obvious link between animals and witchcraft is that of predation. Boyer explains that our evolutionary past has made us keenly aware of predators who may chase us and cause death (Boyer 217). This creates a particularly eerie human reaction to provocations to may resemble or be reminiscent of predators or predatory behavior (moving silently in the dark, a gazing eye, strange sounds, etc). Interestingly enough, popular descriptions of the practice of witchcraft do resemble predatory behavior. Witches may move silently at night in their brooms, or they may cast an “evil eye” (although this practice is believed to be exercised by anyone) on the victims. So human reaction to these “predatory” activities is not only determined in that they are grotesque activities, but also in their conceptual relationship to fear of being predated.
The effect is intensified if the external appearance of a witch resembles that of an animal – particularly, animals known to predate other animals (as opposed to herbivore ones, as cows) – for the identification with predatory behavior may be much stronger. Assuming that the property observed in the first paragraph – that of acting as if they were predators – is a given, the question arises: why should they also look like animals? Boyer explains that the mind has a tendency to fill the gaps, and once a strong identifier belonging to a certain category is recognized by the human cognitive system, other aspects belonging to the same category are often imagined to complete the picture in a coherent manner. (Boyer 145)
Now we could start by denying the first assumption, and presume that witches are not thought to act like predators. Why should they look like predators, if there is no “picture” to complete? Boyer has also an answer for this question: human cognitive systems have a propensity to be wary of objects that belong to a certain category but portray features from other categories (Boyer 224). For instance, a stool with cognitive features (an identifier of humans) is seen as awry; a human which cannot move (an identifier for vegetables), or animals that speak (human identifier) are seen as such as well (Boyer 67). A violation of one feature while keeping other identifiers of its own category causes cognitive confusion in the human, and it makes the human to be keenly aware and concerned about the object sticking out. Therefore, if a witch, fundamentally a person, displays identifiers that belong to another category (animals), such as the leopards, crocodiles or chimpanzees (Shaw 61), this will cause cognitive confusion and make the person pay heightened attention to the object, and eventually fear.
An explanation based on the same evolutionary assumptions of Boyer’s, but with a different line of argument could be made to account for the fact that two, and not one, identifiers are violated in the pre-established cognitive category (being human) among witches. Boyer also explains that objects that violate more than one category are simply not as fascinating to the human cognitive system as those that violate only one. A pencil (tool category) that walks (animal behavior) is to be wary of. Indeed, it is quite possible that a walking pen may appear in horror movies. However, a walking pen that talks (human behavior) may not be as scary as a pen that solely walks. In fact, humanized walking-talking-pens often appear in children friendly cartoons. Taking another turn, a walking pen that is profusely bleeding is indeed a chilling vision. It may be argued that a bleeding-walking pen has violated two categories and still has managed to stay frightening. However, bleeding and walking both are identifiers of animals, and the pen, belonging to the tool category, has merely displayed two signs of transgressing the cognitive breach towards only one category (animals).
Why does breaching one category cause fear, while two may stay for children? Boyer has not explained this, but my guess is that one cognitive breach is enough for the mind to expand its imaginative explanatory devices to “figure out” what’s wrong with the particular perception. One category breach may be already too much for the human mind. Two or more category violations, in this line of thought, may be considered too bizarre to actually happen in reality, being relegated to the surreal. Because a walking-talking pen is unlikely to happen, it can be deemed safe to conceive and allowed to show as children’s cartoons. Another extension of this thought is that single category violations are quite likely to happen – on a windy night, any tree could look like the shadow of a wolf or any big wild animal. A pen may fall from the table because of the wind. This weekend while writing another paper, the plastic curtains (the ones that need to be pulled down) went up by itself and I had a big scare. Of course I understood that this had happened because I had not fixed the curtain when closing it and that it went up because of spring tension forces, I could not help but recall that most horror movies start with scenes like these (and they do not start with moving curtains that also think) and was startled for quite a while. (Actually, I don’t remember what I did. Maybe I went to the lounge to join others.)
Now back to the question of witches who may act like animals and also look like animals. Following the argument put forward in the last two paragraphs, it could be said that witches are thought to carry one of the indicated breaches, but popular rhetoric has added another category breach to make it surreal and therefore exempt from danger. Thus, a double violation may be thought as a emotional defense mechanism, a way of coping with the fears aroused by an object that violated cognitive categories – by adding yet another violation to it and disappear it from the face from the earth.
But which violation comes first, as to trigger the other violation? Are witches thought to naturally look like animals, and then thought to act like animals to cope with it, or are they thought to initially act like animals? I am inclined to answer in the latter, because to think that someone acts like an animal (when minding its witchcraft business) is much easier to conceive than thinking that the witch actually looks like an animal. Of course, in the particular cases of witchcraft accusation in Malawi, witches are said to transform to animals only when performing witchcraft (van Dijk 105). But regardless, that a person transforms into some other physical entity can be disproved in the evidence of material proof (Shaw 63), while a claim based on the mental faculties of the witch is much harder to dodge. Therefore, my assumption is that witches are thought to act like animals, and to cope with the idea, the populace thinks they look like animals, too.
There is an obvious fallacy to this argument, however. As explained before, a talking-walking pen may not be thought as dangerous, but a bleeding-walking pen is thought as such. The difference is that talking and walking do not belong to the same category, while bleeding and walking do. It should be pointed out that looking like an animal and acting like an animal appear to belong to the same category. Therefore, not two categories have been violated, but only one – that of being a human object with features of the animal category. Indeed, to have another feature from the same deviating category should actually increase the uneasiness towards the object, not decrease it.
In sum, witches are thought to be dangerous on the basis of their direct relationship with predatory animal behavior, and cognitive categorical violation. There are two explanations for the fact that witches need to both act like animals and look like animals that are mutually excluding, but both nonetheless account for an evolutionary explanation of such violations.