The Double Role of the Human Gaze

Dictionary definitions of “creepy” ––– sources of creepy feeling ––– the role of eyes in eliciting creepiness ––– the silent threatening gaze ––– gender differences in ogling ––– the analogous creepy house ––– humans’ evolved eye detection and eye signaling ––– humans’ utilization of cover and their intuitive need of privacy ––– humans’ inadaptedness to virtual ramblings ––– lack of rationale for privacy ––– the dangers of not protecting one’s privacy ––– the absolute grip of power of an all-seeing government ––– the uncertainty of online privacy.

Urban dictionary’s current top entries for “creepy” are:
“adj. An overused slang term for ‘sexually inappropriate’ or ‘perverted’ or for ‘attempting to derive sexual gratification through dishonorable means.’”
“Somewhat scary because of strangeness.”
“By far the most common adjective predicated to quiet people who don’t smile.”
“A man who is honest about wanting to get in a girl’s panties.”

For “creep” we have:
“a dude that tries way too hard with chicks, usually younger chicks. Also usually waits till the chicks are fucked up to take advantage of them”
“A creep is a male that a female is not interested in. A woman may label a man a creep if the man does not communicate to her that he has wealth, power, or influence (such as by
the clothes he wears, the car he drives, the job title he carries, or other external signs of social proof), and the only thing he does show to her, his physical attractiveness, is not enough to interest her.”

For “creepy”, the Oxford dictionary simply gives, “Causing an unpleasant feeling of fear or unease” (and gives as an example sentence one with a creepy house). I think this definition captures the experience of creepiness but fails to uncover the source of it which the urban dictionary entries had a go at (beside the retaliatory and cynical ones, which I have omitted with one exception). The source of the feeling is essential to the definition as it differentiates it from the many other things that “cause fear and unease” without being “creepy”.

The kind of “creepy” I’m interested in in here is the one assigned to “normal healthy looking people”. The adjective modifying places (“creepy mansion”) is of an analogous usage (more about it below), while the kind of creepy you get by google-imaging the word (not highly recommended) seemed to me at first distinct or analogous only by a very wide stretch, but I have come to think that the two concepts are very close to each other. Either way, in this latter case, the creepiness that is evoked by the resultant pictures is a combination of disgust and of the eeriness of the uncanny valley, which in my opinion are one and the same. That is, this eeriness is a kind of “stripped-down” version of disgust, like a sinewave being a thin version of a piano playing a note or what vanilla extract is to vanilla. To me the “pathogen avoidance” theory mentioned in the uncanny valley’s wiki page seems clearly to be the reason for the effect. I think it’s no coincidence that to “creepy” google image provides a mix of bisque and ventriloquist dolls, zombiish creatures and disfigured humans ––– all of which being a point between “regular humans” and other things (inanimate objects, monsters, &c).

As for the kind of creepy I’m concerned with here (behavioural creepiness?), I think the crux of it, of what makes a person creepy, is a combination of interaction and lack of communication of intention. What makes creepy that guy who you see at every other party you go to, where he seems to look at you and even gradually change rooms after you, yet doesn’t approach you, is that it is clear to you that you are in his attention, yet you have no idea what sort of design he has in mind as he doesn’t reveal anything, which also leaves room to imagine the worst.
My personal analysis of creepiness was augmented relatively recently by an observation about eye-contact. Eyes have a dual role. First, they are sensory organs, by which people see. Second, they are instruments of communication. They can communicate content, and do so both actively, as when someone makes eye contact with another and then “gestures” with the eyes in some direction surreptitiously, instead of with the hands, to divert attention thither, and passively, as when someone else’s gaze calls attention to the direction they’re looking at (Imagine the domino effect when in a theater some spectators start looking backwards above the rest of the audience and others follow). But the communication of the eyes is more commonly formal rather than of a communication of content; it’s meta-communication, the communication of the intention of communication, the establishing of communication. If we want to say something to someone engaged in some activity, we’ll call their name and await if not an auditory response (“yes?”) then at least a glance, to know that that person’s attention is to what we are about o say. Or, when someone who is being talked to is gazing elsewhere, it communicates distraction and inattentiveness, even though the person might be listening very carefully. If it goes on for very long, the talker might feel uncomfortable going on. Or consider a talkee’s gesture of closing their eyes, as if to say: “stop talking, I don’t want to hear this anymore”.
Those two roles of the eyes have different objects upon which they act: the sensory eye inspects/looks at/ sees the world in which and upon which the seer acts, while the communicative eye interacts with other people to establish a channel of communication.

Strangers might look at each other on the street because people monitor the environment surrounding them. If in this case an eye-contact is made, they will soon avert their gaze to avoid interaction. If the first looker lingers her look, the other will interpret it as a seeking of interaction, and depending on his will to interact, he will either linger/return his gaze or avert it. If he lingers, he would expect the first onlooker to say something. Of course, on the street, with two people passing each other, a mutual prolonged gaze might amount to nothing other than a possible fleeting mutual interest while the momentum of the movement carries the two persons away from each other. But the case is quite different when the two persons are static or within an enclosed space that renders movement to be only a temporary distancing.
Bob looks prolongedly at Alice. If she notices that, two things can happen: either she “answers” him and an eye contact is established, or she “turns him off” by clearly avoiding this contact. In both cases, two things might happen: either Bob approaches Alice, or he doesn’t. If his gaze had been answered and he approaches Alice (or vice versa), then we have a meeting of strangers being successfully mediated by the play of the eyes. If he has been turned down and he doesn’t approach Alice, we have, too, a scenario in which the eyes adequately fulfil  their roles in this kind of negotiation. The two other cases are hiccups in the smooth functioning of this system, but are overall not so disastrous. If an eye contact is made but no approach ensues, then depending on the situation one might assume that either one and/or the other parties were shy. In the case that Bob has been turned off “ocularly” but still approaches, Alice might feel unhappy about it and feel the situation is “awkward”, but all in all the tension is resolved. Another type of tension might arise, in the case that the interaction between them is unpleasant to Alice and she now seeks ways to “get rid of him”, but this is altogether a different matter. When one places oneself in the public sphere, one must expect to be approached by others which one would not approach oneself, or even others that one did not notice were there at all: not much harm was done. Of course, it’s good for people have tact and a sense of when their presence and attention is unwanted, but again, that’s a different matter than what I want to discuss.
Creepiness arises not by a hiccup of this system, but by its complete breakdown in the form of transformation from a linear process to a limbo of a circular one. Everything is in order in the case that Bob makes a prolonged eye contact, is unanswered, and he doesn’t approach ––– if everything ends here. Order has collapsed if at this point Bob goes back to square one, as it were, and beseeches eye contact yet again –––that is, keeps on staring. That’s when the stare becomes creepy, and potentially threatening. I think two things are going on here with the same end result. The relentlessness of Bob’s staring turns it from a kind of meta-interaction into an “interaction by itself” of a different nature: The starer’s eyes transform from an instrument of communication to an instrument of observation, i.e, they switch “ocular roles”, and the staree transforms –– for those eyes and their new role –– from a person, from a peer, to an object, to something that the starer has designs on (otherwise he wouldn’t focus thus his attention) which require no cooperation (otherwise he would have started a conversation, invited or not). That is, he has designs on the staree’s body1, which is of course unsettling. But even before this transformation, the starer’s, Bob’s, ignoring (or ignorance) of Alice’s meta-communication ––– which is also a communication by itself ––– bespeaks of ill-prospects of a successful peer-to-peer relationship where clear mutual communication is necessary. These together have a dark synergetic effect: an adversary or predator that can’t understand or wouldn’t listen is always more terrifying than an enemy we believe we might parley with ––– we know that if there’s a fight, it will be to the finish.2
A variation of this breakdown of the system can also occur if rather than just keeping on glancing, Bob averts his eyes every time Alice “catches” him ––– so that both the aversion and the initial looking are done by Bob –––and repeatedly returns gazing after she averted hers. In this case Alice becomes aware of being an object of interest for somebody else, but is denied of signalling rejection. One would think that this denial is insignificant, but I imagine that it’s rather critical. In the former case the starer ignores signals of refusal and is therefore unyielding in soliciting interaction. In this case the starer avoids interaction himself. A discomfort might arise of a different kind, of having one’s sense of privacy being violated. Of course, in public we put ourselves out there, but what we put out is the potential of our privacy being opened to strangers, rather than this opening itself. In a sense we are invisible in public to others until mutual attention is being cast. The hit and run starer puts the staree in his attention, peeling off some of the public privacy-of-disregard, while hedging himself in his own privacy. I think the inability to signal rejection via this denial would render it less threatening and more of a disturbance (and only so long as the starer is at the staree’s attention), but maybe this is just a minor variation of a way to reach the same kind of creepiness.

We can draw a distinction between two “non-communicative starers”. One is causing unease by the means illustrated above; it might be an awkwardly curious or interested but shy and possibly tactless person, or it can be a person of genuinely malign, concealed but not completely hidden intensions. The creepiness is evoked by a sense of being in that person’s attention –– via repetitive eye gazes and special encounters –– without an explicit engagement. On the other hand, there’s the threatening starer. This is the starer who doesn’t (merely) repeatedly glance, but one who relentlessly holds an eye contact without otherwise communicating and potentially with an explicit non-responsiveness. Here the eyes are used in a somewhat paradoxical manner: rather than avoiding communication, they communicate a commitment to non-communication. The former starer avoids communication, might even avoid eye-contact, as he seeks undetected inspection of the staree and/or gathering the nerve to make a move. The latter starer, however, seeks eye contact by which he utilizes the double role of the eye gaze. The eye-contact establishes communication, but the lack of content –––not saying anything, even ignoring the initiation of communication by the staree––– calls attention to the sensory world-inspecting faculty of the eye, sending a message of, “you are an object in my eyes (and not an equal peer to be respected)”. Imagine two gangs sizing up each other, or a seated gang following a passer-by with their stare. 3

Just like in the urban dictionary examples, you might have noticed a “gender bias” above, where the staree is a she and the creepy starer is a he. This is no coincidence. For one thing, it seems safe to say that women are generally more threatened by men than the other way around. But “something more substantial” is at play here, I think. In his book “Sex Signals”, Timothy Perper introduced the idea of “proceptivity”, applying Frank Beache’s ethological idea to humans. I’m willing to be corrected if wrong, as I have read the book several years ago, but I remember the idea like this: while a man at a given time might view himself as the initiator of a ‘courting approach’, he might really be responding to the ‘procepivity’ of the woman he is accosting. In heterosexual settings, this proceptivity is the state of engaging in behaviour that facilitates an initiation of interaction by the man the woman woos, such as signalling interest: glancing, smiling &c, as well as making herself more accessible to him, for example by standing next to the guy. Not finding out anything surprising, scientists doing “fieldwork” at bars had found that guys were more likely to approach a woman if she first made repeated eye contact with them, followed by smiling.
Either way gazing denotes some sort of interest. When a woman glances repeatedly at a man it is regarded as an invitation to engage; a man’s unrequited gazing, however, might be less welcomed ––– being a brash violation of courting protocol? I suppose it might come down to the fact that tête-à-tête, a woman can expect to be physically weaker than a man, and overpowered at a brawl. This might give rise to a (“biologically wired”?) system of courtship in which the “unthreatening” women may stare at men, but men cannot stare at women while preserving their equanimity.
I could spend more time thinking and reading about the topic, but at this point of writing I believe this to be indeed a system emerging from intrinsic biological instincts. Outwardly our societies (in the west?) are not so governed by violence and physical prowess, at least not in public, but I think that our comfort in the public space is upheld by conventionally cultivated habits, while our nature  is wary of the strangers we walk past by or who stand around us in the subway, who we don’t know where they came from nor do we know of their intentions. Regardless of how well our societies are moved by words than by muscle, to a large extent our interaction with other humans is dictated by ancient heirloom heuristics.

I’d like to return to what I have said at the top, that the adjective in “creepy house” is analogous to the creepy I have been discussing. My trustworthy google image test shows that “creepy house” and “haunted house” are equivalent4, that what makes a house “creepy” is the same that makes it “haunted”-looking. Generally speaking, there are two types of haunted houses: the abandoned and the inhabited kind. I’d say that “creepy houses” are exclusively of the former kind, while the inhabited kind ––– a house in which people reside but which also harbours ghosts ––– is not intrinsically creepy, though it might give rise to “creepy moments”.
How can we characterize these creepy houses? They are former well-to-do single-family homes that fell into a dilapidated state. The creepiness arises from them being inside the uncanny valley between regular occupied houses and a truly uninhabited place. When we see the façade of a like house, or enter the interior of such a house, our expectation is that somebody lives there, as virtually all the houses we have encountered in the past were inhabited. If we don’t hear anything, then perhaps nobody is home ––– except, of course, we never enter strangers’ houses when nobody is at home, since somebody must open us the door. So if we enter a “haunted seeming house” (in the case the doors are unlocked or just open), our expectation is that somebody is there. And if we don’t hear that somebody, it is possible that they are ambushing us, waiting for us to draw near. 5
The creepiness is a fear one step removed. We are not being afraid of something that is clearly in front of us that might be dangerous to us, such an aggressive looking barking dog or an uncontrolled fire, but rather wary of a possibly scary thing imminently emerging. This is similar but distinct of fear of the dark,  as here too we have a fear once removed: it is not the darkness which is scary, but the possibly lurking dangers inside of it. The difference is that the darkness and the potential source of danger are two separate and distinct entities, and the nature of the darkness itself is unchanged whether something is lurking or not. This is not the case with the haunted house. The apparition of “danger” in the house is in a sense something that is distinct from the house itself (perhaps the way the son is distinct from the father in the trinity, to use a loaded example), but it transforms the meaning of the house itself6. From an unoccupied or “seemingly haunted” it turns into an occupied or “actually haunted” house, turning us from explorers into intruders. As opposed to the darkness into which we would enter again given the knowledge it was “clear”, we won’t do the same with the house where we have bumped into unexpected company ––– unless we were burglars.
The same type of confusion in regards to the inhabitancy of the house is inherent to the creepy guy. He seems to be a regular person like all the other party or event goers, but he is not “social”: he came into contact with us, but he does not adhere to the meta-communication protocol which negotiates interaction between strangers. At the same time, he didn’t pose nor disclosed the potential for posing any threat to us, and so we have no “justified reason” to flee ––– neither justified to us nor to others to whom we might want to explain our behaviour. And so the significance of this person is elusive. I wander if therefore the best way to confront this situation is to accost the creeper oneself, possibly with a friend as a comforter?
It was this comparison of the “creepy guy” and the “creepy house” that led me to see that the former’s similarity to the scary-creepy is even closer than I first imagined. The former, too, invokes the “uncanny valley” feeling, though in a very different way. The ambiguity is not of any physical form, but rather of behaviour.

Humans are good at detecting stares6; just like one catches one’s mentioned name in the hubbub of a party, one also catches a direct stare at themselves (or that was my personal experience at least, and I extrapolate; I never heard people talking about it). Indeed, the human eye seems to be designed similarly to the deer or antelope’s rump which is intended to visually draw to itself attention of predators. I was about to be speculative, but googling images brought me to discover that there were earlier speculators with stronger credentials than mine, namely two Japanese Zoologists who in a 2001 paper stressed the visual difference between human eyes and other apes’ (and generally animals’ eyes); the human sclera (the white part) is unpigmented and a high ratio of it is exposed.  I quote, “[…] we postulated a hypothesis that only coloration of the human eye is adapted to enhance the gaze signal while eye coloration of other primates is adapted to camouflage the gaze direction against other individuals and/or predators. This hypothesis was examined and supported by analysing relationships among iris coloration, sclera coloration and facial coloration around the eye. Our results suggested that unique features of the human eye started to evolve as adaptations to large body size and terrestrial life and were completed as a device for communication using gaze signal.” Another paper from three years ago examined the faces of different canids and found that specimens of species with strong “eye contrasts” (contrast within the eye and vis-à-vis the surrounding face) were engaged in more frequent social hunting, and stared longer at each other’s faces, i.e., were more social.
This suggests that among homo sapiens the advantage of being able as a species to communicate with each other with the eyes triumphed over the individual advantage of being able to look undetected at a conspecific (and perhaps the stereotypical obfuscating shades of officers is a reappropriation of the anti-social inspection; the eyeliner of women would be antithetical to it, amplifying detection of gaze, facilitating proceptivity).
Fittingly, humans have developed instincts that made them attentive and wary of quiet strangers who seemed to have more than just momentarily glanced at them. In the first paper, elongation of the eye in primates was attributed to the advantage it confers of increased capacity to horizontally scan the visual field without moving the head, but further still in humans –– the hypothesis goes ––– in order to facilitate the eye signal. But I wonder if an additional reason for this extreme elongation to develop might not have been that it allowed to circumvent, as it were, the conspicuousness of gazing, by enabling starting at an angle that greatly diverges from the frontal “default” one.7 8

We know as people that possible observers might also be hidden, and so in settings where privacy is demanded obstacles are raised to hide us and thus free us from the need to monitor our environment for onlookers ––– we have walls to enclose the private indoors from the public outdoors, we have partitions and screens indoors, and we have curtains, blinds and reflective films to offset the degree by which our embrasures allow for ingressions. Otherwise, people are aware of the space around them and the “optical” opportunities it offers, and in public (I might be over generalizing from myself, but I’d imagine it to be rather universal) they would generally seat themselves at a good vantage over the space and its points of entry. You won’t see people standing in the train against the wall, unless there’s a window or a door there; if you do, I suspect that you’d suspect that they are on the phone, but I think even in such cases people would face the inner space.
This is a huge topic by itself, but I’ll say as if I can do so in passing that I think that human beings are rather more of observers of their own lives and of their own actions (a vague and also problematic statement, I know) than the actors leading their lives. This is a problem when certain aspects of life, which were hitherto well taken care of by instincts, change their form in such a way that our instincts fail to detect them as relevant, and it is left for our conscious and intentional will to take care of them. To take a not so good example, there are many poisonous substances that were not available for most humanity’s existence. Thallium is odourless and tasteless, two modifiers that allude not to the properties of this element but to properties of the human sensory faculties which happened to not be “attuned with such a substance in mind”. Humans can consume and continue to consume this element unknowingly, killing themselves in the process. Or, to take another example, libido used to “take care of” the issue of propagating the human species, but with the advent and spread of contraception it became possible to satisfy this drive without pregnancy as a consequence. Since there still is “family planning” (as well as good old “unplanned pregnancies”) the human species is not going to extinct as a consequence of this, but I’d say this technology certainly “took its toll” on birth rates, and I won’t hesitate to say that there are people nowadays who die childless who wouldn’t have died this way had contraception been not as convenient and widespread as it is today.
What I’m concern with, however, is the sense of being observed. Many people would be uncomfortable and object if there was a person, especially a stranger, especially a stranger from a governmental agency, standing constantly over their shoulders while they are on their computer or phone. However, my general impression is that people have done very little to establish security/privacy in their communication, even after the extent of NSA surveillance or the CIA’s spying capabilities were revealed. The sense of privacy, or the drive for privacy, if you will, fires only at the perception or suspicion of actual directed eyes, and is rather dormant in the face of the ambiguous notion of someone remotely monitoring our virtual ramblings.
That people talk about the internet as a place is evident in word choices. You have “cyberspace”, for example. People “go online”  (or at least they used to? Now being online is so common and constant that “going offline” is more of an action then being online, and words like “afk” or “brb” are obsolete outside the sphere of the “virtual realities” of online games) or “meet online” (i.e., as if it was a place  ––– that’s the expected answer to “where did you meet?” rather than the “geographical” first location of “IRL” rendezvous). Nonetheless, on the immediate perception level, say, people experience the internet as a tool rather than a space, and tools don’t “return a gaze”. But here our long-honed perception is deceiving us, as when we are having a 1-on-1 chat with a friend over this or another messenger and we feel “we are two alone in the room”, but really someone might be eavesdropping.
To answer why they take no measures  to surf or communicate online more securely people say that they have nothing to hide, but I suspect it is more of an ad-hoc interpretation of their neglect, or a solution to an otherwise potential cognitive dissonance. This can also be put differently: People use their rationale to plan and to understand, but they are not required to understand their (often advantageous) impulses and emotions in order to act upon them. People don’t need to understand why it is “good” that they have a sexual drive, or why they cry at one moment or become furious at another, or why they fear asking someone out on a date even if “the worst that can happen is that they will get rejected” ––– they just act upon these or, alternatively, use their willpower to overcome these when they deem it necessary. Otherwise, it becomes rather trivial to give an explanation to an action when there was no contrary impulse in the first place. The question “why didn’t you buy that pizza when we were at the pizzeria” is much more easily answered if you weren’t hungry at the time, and it doesn’t matter if the other person says “you should have been hungry” because “you haven’t eaten all day”. Similarly, not having any impulse to take special measures to utilize a secure e-mail service, for example, one easily answers she “doesn’t have anything to hide” –––– or with any other answers, as nothing really contends other than the person saying she should have been hungry when she wasn’t.
I find this crucial, because the world in which our society is embedded has vastly changed (and presumably keeps on radically changing) since the time of the world in which our species had evolved. In the past one could have much more reasonably relied on one’s instincts. Today you might be in danger of becoming unhealthily obese because the once rare sugar is now ubiquitous, develop physical ailments because work and so many other activities can be conducted by while sitting on your ass instead of walking around, develop “psychological problems” because many activities can be done without having a “full interaction” with other human beings but instead with their virtual manifestations or with no humans at all, your property can be stolen if you click the wrong link or insert your card into tinkered-with ATM and so on. I suspect that people who say that they are unconcerned because “they have nothing to hide” would still use curtains on their windows if they face potential onlookers as if “they do”, in the same way that people who feel like they should be working watch hedgehog videos instead (i.e, succumb to impulses).
There are other reasons why people would like to put curtain on their windows, but I think that generally speaking, our need of privacy is there for a good reason. To begin with, one can compare its complementary to certain fears that concern rare but highly damaging phenomena, such as being scared of getting into a dark forest or alley. It might be very unlikely to encounter a fearsome beast or violent person, but one does not want to be the one in the thousand that got eaten or robbed (or worse):
CIA and NSA agents, employees of corporations that hold private data as well as hackers of various sorts, are first of all human beings, and as human beings they occupy the same public space as you do. Imagine you (knowingly or not) cross roads with such a person, and he or she has a reason to want to read through your emails or chats, or through your online purchases list of yesteryear ––– perhaps they are “romantically interested” in you, or perhaps in your partner, or perhaps they have a friend that is a business competitor of yours. Suddenly your emails are not something that you “don’t want to hide”8. You might assume that such people are not so numerous and that they don’t frequent the same localities that you do, but the picture changes when you regard the distanceless internet. What if you post something online that gets shared just enough to come to the attention of such a person on a particularly bad day, and they happen to strongly disagree with you, and now they want to mess with you just out of spite,  and this fantasy of theirs is just that much easier with them having access to your data?10
Such personal considerations aside, I think one should be concerned on more theoretical grounds. This prying is a privilege of the powerful –– both of the governmental and the corporative kind ––– and is, at the same time, a strong tool for keeping them in power. Just keep the ability of spying on your political or business rivals in mind, and I think it seems rather clear. And this becomes dangerous once the extreme possibility is fulfilled, of this power being strong enough to ensure the constant holding of a political party in power or the virtual monopoly of a corporation. The next step is inevitable corruption, as the responsibilities or these organisations diminish (as they can act freely without fear of repercussions to their holding of power) and them becoming an attraction to the psychopathic type of persons who seek power rather than acting within the organisations with their original goals in mind. Given the role of “media” in the contemporary age, this state of affairs would be alternatable only via an almost impossible “offline-incited” and perhaps very underground and violent revolution. The dangers of amassed capital by the few is big by itself, but to me it seems like the dangers of amassed data-collection-powers are further greater an in some sense more “absolute”.
It might seem like I’m describing the power stemming from the suppression of expression/press by rivals rather than the power stemming from mass and deep surveillance. It’s true that my “prediction” is exaggeratedly grim, but I think that a combination of “omniscience” and some executive power can go a long way. I’m thinking about Russia and the way the government controls the oligarchy: the law is such that “everyone is breaking the law”, but only those who do not comply with the agenda of the government is persecuted while the rest is spared. Or Nixon’s war on drugs which was a ploy to arrest people from the anti-war left and the black community. To take the word of an our contemporary Baltimore police officer, drugs are possessed in equal proportions among white and the rest of the population, but the police only stops and searches young male blacks and hispanics who are subsequently arrested. How more convenient would it have been if one didn’t need to go through the hassle of stopping people on the street, but could sit at the comfortable office, sip coffee, and look for online conversations of targeted groups about that weed that was smoked the night before?
In this kind of situation the government can cherry pick whoever it wants to jail or greatly discomfort. If it can read everybody’s communication, it makes it all the more easier to dig up dirt. Further, due to this asymmetry of information, the government can pass laws that would “secretly incriminate” both its own agents and its rivals. For example, some widespread but hard to detect election-campaigning practices might be outlawed, but the instruments of government surveillance would be only directed at opponents. Such laws can be passed also to persecute selectively press outlets, advocacy groups, activists and so on.
The challenge is that, as opposed to visual occlusion, it is hard to tell whether you are “covered” or the television is listening to you or there’s a not so friendly ghost haunting your online (and perhaps offline) conversations. As far as I see it, the only thing to be done, on a personal level, is to adopt best practice with regards to online browsing and communication, convince one’s peers to do likewise, and, on a wider level, to push for legislation that  inhibits companies from collecting and publishing11 private data, and, of course, the government from spying unrestrictedly and without warrant on users12.

1. I want to make it clear in case it seems like a big leap. The reason the design is on the body and not on another non-bodily aspect of Alice is that what the observant eye sees is the world, its matter. In other words, what it sees, as oppose to the communicative eye, is always an object. Therefore it objectifies everything it lays its gaze on, and thus turns humans into bodies. Another case is found in war. If enemy soldiers are looking at you it is not because they want to talk to you, but because they need to see you before they shoot you.

2. This is anecdotal, but curious: looking for google images of monsters and scary things, one finds creatures with prominent eyes and teethy mouths, but their ears are often missing. On the other hand, google imaging “cute creature” immediately brings forth enlarged ears. I was surprise to see my expectations being so well answered.
Also, I wonder now if this is why “troll” –– the internet kind ––– is called a troll: predatory and non-listening (wiktionary puts the etymology with the “fishing” kind of trolling, rather than the one connected to the monster, but I think that origin is one thing and propagation is another. How many people know about the fishing kind of meaning of the word? But this is wholly a different discussion).  Notice that the common troll depiction (“troll meme”)  has a large mouth (here because it speaks, not because it will eat you) but no ears whatsoever.

3. I think phones (especially landlines? And with obscured caller id?) have a clear analogy: imagine someone calls you, but says nothing at all (no content); creepy? Threatening? If such a call is repeated?
The phone analogy of the rejection-denying starer would be a call that is hang up immediately as you answer.

4. It’s certainly problematic to prsent anything that is based on some unfathomable algorithm. Google and google-image in particular would interpret verbal expressions in certain ways that would usually bring the searchers what they are looking for, but it can also over-generalize and be slightly too presumptuous. That being said, I mostly use it as a “sanity check” confirmation, in this case to see that what I imagine to be synonymous expressions are not just confounded in my own mind.

5. One might add that these houses are often depicted at “dark times”, i.e., during the night and weather. This increases the creepiness in two ways. On the one hand it provides obfuscation which makes it harder to tell whether the house is “really unoccupied” or not, therefore increasing the uncertainty aspect of the creepiness. On the other hand, it also increases the danger of the potential lurk, as during such a weather less if any people are roaming outside, making it less likely to reach help if needed.

6. To use another image example. There’s a “meme” kind of image called “when you see it/ you’ll shit brix”. The “generalization” of it includes any images where there’s a “hidden detail”, but I imagine that it stemmed from a more particular type of photo where there’s a person or group of persons taking a selfie, while in the background there’s a creepy face (googling “creepy when you see it” will bring examples). The speed at which one can detect the “hidden face” is significant. However, this might simply arise from the affinity of visual perception towards face detection in general. I tried to find pictures where the background is full of faces where only one is staring at the camera (to check how quickly I pinpoint it), but I didn’t find any.

7. Have you ever stared at anybody that returned back a side-gaze with one eye (the other eye being on the other side of the nose)? I have, and I didn’t get a feeling on the “gut level” that I was being “caught”. My mind said that perhaps I was being looked back at, but otherwise it was hard to tell. Is this a bug? After all, we mostly regard faces looking directly at us or turned completely away from us, and so it might be harder to learn what a side-glance-at-us looks like (i.e, stare detection might be a learned ability, and staring-from-the-corner-of-the-eyes might be a relatively rare example of starting, which precludes us from correctly “learning it” to be an instance of starting).  But I think that if a mechanism by which to observe surreptitiously conspecifics was developed (from the corner of one’s eyes, for example), there must have developed a mechanism by which to detect self-detection (i.e., get a sense of when the self was caught) if the former was in any way effective or significant. And so I think this gut-level-lessness is a feature rather than a bug.
With the more extreme ends of the side-gaze, one can see things quite far from one’s center. However, the very extreme end of the vision is being caught by the corners rather than by the fovea, and I suspect it is difficult/impossible to correctly discern the direction of (others’) eyes caught by the corner of an eye ––– and in this case a staring person might be “aptly” unalarmed by a seeming person-searching side-stare.  That is, if you stare with a sidish look at another person, and that other person is looking back at you with an extreme side look, it’s possible that rather than them catching you staring, you are catching them. But don’t take my word for it.

8. Another more probable in my opinion function of the eye elongation, and which had not been addressed in that one paper as far as I skimmed it, is the expansion of the “gamut”/”repertoire” of the eye signal, rather than just increasing its “strength”. I’d imagine that after a certain low ratio between the dimensions of the iris and the dimensions of the eye rim had been reached, the rate at which the conspicuousness of the gaze increases with the decrease of this ratio, decreases.  In other words, the marginal benefit diminishes at some point –––– and it has to contend with the marginal cost of having more of this vulnerable and conspicuous organ exposed to the outside.
However, this decreased ratio between the iris and the eye, on top of making it more conspicuous, also increases the number of “states” the eye can assume, just as there are more possible first moves on a 19×19 go board than there is on a 9×9 one,  and hence theoretically increases the number of signals the eyes alone can render, both static ones such pointing in a direction, and “dynamic” ones such as an eye-roll. I also imagine that it increases the saliency of signals rendered by changes of the size of the exposed eye such as squinting or widening the eye, as the changing ratio between the iris and the exposed sclera is easier to detect than the absolute change in the size of the exposed eye.

9. You might wonder how many people within Facebook or NSA, for example, actually have the power to access to your data, and one might imagine that low level employees are barred from it. But as the amount of denial the NSA has expressed regarding surveillance seems to suggest it is corrupted as an organization, I would put little faith into assuming that there’s any restrictive discipline within the organisation.

10. Or consider the example of Chris McKinley:
(this link is bound to expire one day, but “Chris McKinley okcupid” should probably bring up the story a long time after it does)
Vis-à-vis potential CIA agents’ abuse of power, his approach was more bottom-up than top-down, but the effect is rather similar and is illustrative, I think, of the potential abuse of “top-downers”. Were his dates duped? They certainly attributed more significance to him as a potential partner than he did to them. It was filtered through an algorithm, but to them he was a unique man who “answered all the questions correctly” (I mean it in a more metaphorical sense, though in this particular example it is rather literally the case), which is what a person could do had she/he had an access to your emails or chat logs. What seems to be a miraculous chance and therefore a significant one to the second party was really a systematic calculation of the first.

11. This is not “publishing” as in “making public” in the old sense of the word, but it is very close to it. Users’ private data is collected and then sold to other companies(!), that is, is put on the market place which is essentially a public realm. Not the whole public has access to this market, but the private users have no say as to to whom this data is offered, and as far as they are concerned, it could be anybody. The fact that this data brokerage market place is not wholly public in the full sense of the word makes it somewhat worse, as the users cannot even know to whom it is offered, what exactly is offered, and who actually bought the data about them.
Honestly, I’m not so versed at current laws regarding data protection and at common practices among giants such as Google and Facebook, but given the state of things, I assume that even if user agreements present users a good sense of what is being done or could be done with their data, most users do not read this “fine print”, and, further, that even if they read it (or rely on some blogger’s digest of it) and find something disagreeable in it, they would still grudgingly acquiesce to it. While Google and Facebook’s rise to power is “fair game”, they now possess of something that can be rightly said to be a major infrastructure of the internet. Sure, one can opt to stop using running water and go draw from the local well instead ––– but how many would?
But this water analogy is not so fitting. The “commodity” concerned here is of a social nature (I’m putting Google’s search engine aside). Of course one can have a social life without using Facebook, but what is lost by abandoning it is not quite replaceable by “walking to the well”. There’s a two-fold problem here. First, being a communication channel, the utility of Facebook (and other “social media” as well as online messengers) increases as more of one’s contacts use it. One can opt out from such a channel and move to another, but for it to be done successfully, other people must do it in parallel as well. But such transitions confers a cost, even if ever so tiny as a little time and willpower/decision, and the coordination of such a transition is difficult to impossible, in particular when such established “institutions” as Facebook are concerned. Because of this, Facebook has a very wide leeway in abusing its power, knowing that it would lose only an insignificant amount of its users by increasing its “abuse”. Second, Facebook’s own infrastructure (servers, staff) is expensive. For this infrastructure to be established and maintained, someone needs to pay for it. What Facebook does is generating money from its user-base via ads. As an Alternative, one could create an ad-less and “data protecting” service ran through direct payments (of a rather miniscule amount) of the users ––– but I’d imagine that most people would find it just cosy to get the service for free by simply sucking up the presented ads and remaining in ignorance as to how’s the data they are generating about themselves is propagating.

12. One might be tempted to say “citizens” or “residents” rather than users, but this case might be problematic. How does one determine whether a user is a resident or not? If “domestic users” (within the country) are protected and everybody else is not, then, first of all, both citizens who use internet abroad and users who use it domestically but router their connection through foreign servers are unjustly exposed. Second, if the government is allowed to unrestrictedly spy on foreign people, then they could probably collect quite a lot of information on domestic users indirectly by spying on foreign users/sources, given the widely distributed nature of the internet.

The Double Role of the Human Gaze