Staring back at the evil eye

A few months into my fieldwork in a Romanian village, I was told by friends that I wonder way too much. When visiting people in their homes, I alway noticed something interesting, be it old house architecture, inventive implements, cute animals or anything catching my attention. My mistake, I was told, was expressing my curiosity out loud, wondering how this was made or where that came from. Worse, I was praising my hosts’ properties, mistakenly thinking they wouldn’t mind, or quite feel proud. Instead, I was told people were uncomfortable with such expressions of wonder, curiosity, and praise because they bring misfortune by means of deochi the “evil eye”.

In folklore studies, the evil eye is “a widespread but by no means universal folk belief complex according to which the gaze or praise of one individual at or for another may cause illness or even death to the second individual or to an object belonging to that individual” (Dundes 1981 p vii) [1]. I haven’t thought about this until writing a chapter on the epistemic protection of households in my village. Fear of the magical evil eye blends in nicely with the wider local representations of secrecy and mistrust.

And then I found out my mother hanged in her apartment an Israeli hamsa, my daughter plays with a Turkish blue-eye glass amulet, and my partner has a Moroccan bracelet with blue-eye beads, all gathered from our trips.


In fact, if you google images for “evil eye”, most hits are on the symbol of the blue eye. This however is not an evil eye; it is, on the contrary, a symbol meant to protect the wearer against the malicious agency. Well documented in the study of folklore (Dundes, 1981), variations on the eye motif are widespread in the Mediterranean area. What makes this cultural token so successful?

Evil eye, for all its magical properties, is still an eye. It may dry up cow’s milk, break down things or sicken people, but it does not do so randomly. It follows from the evil person’s visual attention to something. Mere knowledge is not enough, sight is crucial. As I know from ethnography, things kept hidden and out of sight are not in danger of deochi. But of course certain things remain clearly visible and hence vulnerable. What is a good way to prevent this unwanted attention?

Perhaps staring back. Eye symbols may be perceived as a cue of real eyes. They may fall under the actual domain of perceiving eyes as part of faces with further inferential results. You can sense the difference between these two picture, one with three agents looking in the same direction, versus three eyes at a slight angle apparently interested in different things. The wide-open eyeball of the amulet, its conspicuous iris and sclera in particular, sends the viewer a cue of being watched, and perhaps even more – surprise? anger? Hanging such a symbol above your door or at your neck provides cue of an unflinching gaze meeting any and all forms of visual attention, evil eyes included.


Olivier Morin (2013) [2] analysed the cultural attraction of direct gaze representations in portraits, and here the gaze has additional moral power.

There is experimental evidence that presence of eye drawings makes people behave prosocially. When watched, people are more careful with their reputation and adjust their behaviour closer to social norms (Ernest-Jones, Nettle, and Bateson, 2011 [3]; Manesi, Van Lange, and Pollet, 2016 [4]). They pay more often their coffee at unattended machines, and the effect is stronger when symbolic eyes seem to pay attention by their direct gaze. So could it be that decorative amulets really keep away evil-doers by making them feel watched? Not impossible, but hard to test. Or perhaps users of amulets see their monitoring gaze as an intuitive reason to attribute them some efficacy? If the evil eye comes from nefarious attention, one can signal that the attention is recorded. When eyes (people) look at each other, a certain form of shared knowledge appears. I know you are looking at me, and I know you know I am looking at you, etc. If you are observing me, I am observing you. If you have evil thoughts, better look away since someone saw you looking.

Here is another puzzle: natural blue eyes are believed to be particularly prone to evil-eye misdeeds, yet also blue eye amulets are the best defense. Is there a causal link from one representation to another? Seeing stylised blue eyes reminds that the other observer knows about and takes protection against evil eyes in culturally-defined ways. But it is also a visible public token of the tradition of “evil blue eyes”  Does this signal not only “I see you” but also “I particularly see you, the one with the evil eye”? (Why blue of all colors? I have no idea, honestly. Do you?).

There is another way evil-eye averting eyes are culturally successful. Amongst many cultural traditions associated with the evil eyes, this seems one if not the most easy to understand and pick up by foreigners such as tourists and even included in New Age symbolism in societies without displaying the traditional beliefs in the agency of the evil eye. Many of you perhaps encountered them more than other traces of the evil eye tradition. In this case, the artefact encourages two representations. One is the overt idea of magical protection. But it also an intuitive reminder of the evil eye itself. Folklore has other recurrent means to avert misfortune-by-attention, such as spitting, making the object uglier or hanging a flashy red band, but they do not carry a reminder of the original problem. Perhaps that is why they tend to remain local traditions such as those I observed in a Romanian village. The tourist buying the blue eye trinket gets two things at the price of one: a cute amulet as well as an intuitive representation of what it is against. Vicarious perhaps, but easy to explain with a nice story why you put that item on the wall of your condo.

The eye staring back is not the only cultural tool against the evil eye. The villagers I know do not use eye symbols against deochi but hang elaborate red tassels to horses and cows.


Before knowing about deochi I saw them as purely aesthetic. But why this attention-grabbing artefact? Again I don’t have a good explanation, but I know better now than to praise them out loud or look too intensely in their direction, even without an eye amulet watching me.


[1] Dundes, A. (Ed.). (1992). The evil eye: A casebook (Vol. 2). Univ of Wisconsin Press.

[2] Morin, O. (2013). How portraits turned their eyes upon us: Visual preferences and demographic change in cultural evolution. Evolution and Human Behavior34(3), 222-229.

[3] Ernest-Jones, M., Nettle, D., & Bateson, M. (2011). Effects of eye images on everyday cooperative behavior: a field experiment. Evolution and Human Behavior32(3), 172-178.

[4] Manesi, Z., Van Lange, P. A., & Pollet, T. V. (2016). Eyes wide open: Only eyes that pay attention promote prosocial behavior. Evolutionary Psychology14(2), 1474704916640780.

3 Comments

  • comment-avatar
    Dan Sperber 15 February 2018 (18:49)

    Interpretation and Cultural Attraction
    Here, Radu, is a methodological/theoretical question about your enjoyable post. You write:

    …perhaps users of amulets see their monitoring gaze as an intuitive reason to attribute them some efficacy? If the evil eye comes from nefarious attention, one can signal that the attention is recorded. When eyes (people) look at each other, a certain form of shared knowledge appears. I know you are looking at me, and I know you know I am looking at you, etc. If you are observing me, I am observing you. If you have evil thoughts, better look away since someone saw you looking.

    These are speculative interpretations of what might go on in the mind of people who fear the evil eye and take protective measure. Let’s assume, not implausibly, that these interpretations correspond to something many of these amulet-users have in mind. How do such interpretations help us explain the use of eye-shaped amulets as a protection against the evil eye?

    I assume we would agree that interpretations that seem to rationalize individual agent’s cultural beliefs or practices do not by themselves explain these. Why not? Because people encounter such cultural beliefs and practices already when growing up. They don’t need reasons to adopt these beliefs or practices other than the fact that they are already adopted by their elders, who are competent and trustworthy. In general, people don’t need a reason to accept a common cultural practice. On the contrary, they would need a reason to challenge or reject it (which might be a socially costly move).

    Does this mean that what explain such bits of culture is the inertia of conformity? Mere conformity to what others think or do involves no preference regarding content or form. A conformity based explanation would predict that in human culture “anything can happen and it probably will” (to quote one my favourite films), as many anthropologists have actually suggested in more scholarly terms.

    In CAT (Cultural Attraction Theory), we assume, however, that beyond conformity (and other “transmission biases”, to use Rob Boyd an Peter Richerson’s phrase), there are powerful factors of attraction that cause only some representations, practices, and artefacts of specific content or form to reach a degree of dissemination and stability in a population sufficient to make them typically “cultural”. Among these factors of attraction, there are psychological mechanisms that make some items more catchy (or likely to be transformed, in the process of transmission, towards nearby more catchy forms or “attractors”). Now, lending itself to an intuitive interpretation, as you argue do eye-shaped amulets against the evil eye, may be a factor of attraction. This is neither implausible nor obvious, but would need to be argued. If this could be done for most of cultural items, it would suggest a nice articulation, not to say conciliation, of interpretive and epidemiological approaches, but let’s not be carried away.

    In fact merging interpretation and causal explanation (or, more specifically, intuitive interpretability and attractiveness) may itself be an attractor for students of culture. This, however, is a reason to be cautious and to demand precise and testable hypotheses about the cognitive mechanisms that might be involved.

  • comment-avatar
    Pascal Boyer 16 February 2018 (16:37)

    Staring back – In the context of zero-sum sociality
    I suggest we ignore Dan’s wise counsel and blithely speculate, mixing interpretation and explanation in an epistemically unsound manner. (But with an eye, so to speak, in the direction of possibly testable hypotheses).
    Radu Umbres’s ethnography suggests a place where sociality is construed by most people (here I may be simplifying a lot) as the interaction between closed family units, with intense solidarity within the small domnestic unit and equally intense suspicion, jealousy, mistrust between those units. A situation described very vividly, a long time ago, in Banfield’s classic The Moral Basis of a Backward Society and, mutatis mutandis, in Jeanne Favret-Saada’s Deadly Words: Witchcraft in the Bocage.
    In such places, cooperation between family units is restricted to occasional, strict tit-for-tat. You fix my car and I help you move. Or, I work in your field and you work in mine. And even that is rather rare. Open-ended reciprocity – I fix your car but do not ask for anything in return, except some unspecified benevolence – is perceived as threatening.
    In this kind of social environment (again, broad brushstrokes here), the very fact that people outside the group know something about your family, even some trivial detail, is seen as potentially dangerous.
    People explicitly describe the evil eye in terms of that suspicion – others will see that you have something they do not have, and will harm you until that perceived imbalance is eliminated. The economics of that kind of negative competition is described in (Gershman, 2015).
    So “staring back” by wearing, for instance, a necklace with an eye motif, seems to suggest the same tit-for-tat sociality. Others could harm me by just looking at me – I pre-emptively harm others by looking at them. In Jeanne Favret-Saada’s book, “strong” people can look you straight in the eyes. You must stare back. Looking sideways would mean they have invaded your ‘space’ and you let them do it.
    By contrast, I once worked in a place (in central Africa) where people also worry about others’ jealousy, they also “hide” some precious possessions (e.g., by calling their own children “ugly” or “difformed”, to fool spirits that might be jealous of beautiful children) but they do not have evil eye beliefs or anti-evil-eye protective eyes. (And that is in a place where looking people in the eye is considered aggressive, so the potential is there).
    So, moving from unconstrained interpretation to almost-testable explanation: Can we predict that staring-back practices will be more easily acquired, transmitted etc., in places with tit-for-tat sociality?
    The hypothesis might need quite a bit of fine-tuning to be really testable – illustrating that as Dan recommended we should be careful about mixing our interpretive apples and explanatory oranges.

    Gershman, B. (2015). The economic origins of the evil eye belief. Journal of Economic Behavior & Organization, 110, 119-144. doi: https://doi.org/10.1016/j.jebo.2014.12.002

  • comment-avatar
    Radu Umbres 19 February 2018 (00:45)

    Speculations and testing the model
    Thank you Dan and Pascal for these engaging comments, and I would like to get back to some of your suggestions.

    @Dan, I totally on board with your caveats and call for more rigurous testing. But I think there is something empirically interesting about the staring-back evil eye. Let me start from this wonderful quip that in culture “anything can happen and it probably will”. The cultural repertoire against evil eye harm is indeed vast. Some cultural forms are unique to one location, or maybe just the last old woman the researcher found to know about these magical practices. What links token of the motif is the folkorist/anthropologist’s interpretation (which also classifies all evil-eye beliefs as a class). But there are what folkorists call motifs, reccurent themes, concepts, symbols, which appear across locations, in different societies sometimes thousands of years or kilometers apart.

    This raises an empirically interesting question. Why is staring back evil eye a motif, but wearing something the other way around, as I know from Romania, is not? One has more cultural success than the other one, something which can be quantitatively proved. Is there a story of cultural transmission? Of separate converging inventions? A bit of both? I wish I had a good theoretical model, but instead here is another convergene: between traditional and modern, globalised cultures. Why did the staring eye emerge the winner rather than rubbing ash, another widespread folklore motif? Why is the red color string a close second? Mere conformity, in my opinion, cannot account for this string of apparent coincidences.

    @Pascal I have to say I share the epistemically-unsound intuition that tit-for-tat sociality has something important to do with evil eye beliefs and protection devices I would see a connection with limited good approach cultural representations, and beggar-thy-neighbour economic strategies. But I did not think until now that the staring back eye motif can also be narrowed down to one-to-one retaliation. In my fieldwork, holding a strong look and eye contact is also a sign of power and security as in the Bocage, but neither place employs staring back amulets. So is the blue eye amulet just an esthetic frill of a more general stairing-back practice?

    To fine tune a bit the hypothesis, what would be a way to protect against the evil eye in non tit-for-tat sociality? Perhaps an appeal to a higher authority, such as saying a prayer or an amulet with sacred text quote? There is some evidence in this direction, but I stop before making too much of an apple-orange jam of the issue.