post #8: on pareidolia and hallucinations

     Motivating this post (beyond these topics being things I perennially think about, anyways) is that I finally got around to watching the latest of Jacob Geller's “Fear of ___” series, (here, “Fear of Dark”). In it (“Fear of Dark”), he talks about the movie Skinamarink, drawing attention to how the it uses digital cinematography to accentuate the way shooting in poorly lit environments at a high iso causes the camera to perceive or record a kind of static/noise artifacting in the darkness. (I haven’t seen the movie, but I assume what he’s saying about it holds true). Geller wonders if this can be said to constitute the machine (i.e., the digital camera) hallucinating. After bringing it up, though, he largely drops the question and moves elsewhere. Geller’s posing of the question connects to a larger throughline in his video of what the fear of dark looks like, at a psychological/phenomenological level. That is, that the fear of dark generally breaks down into a dichotomy, fear that there is something there (in the dark) or that there is nothing there. In the former case, Geller talks about various examples of how people hallucinate sounds and images when in darkness, or onto dark spaces.

     I want to return to the ideas Geller turned over as they are things I've thought about myself for some time. In my own audio visual work, when I shoot the footage myself, I have so far always used my phone (a digital camera) to do so. While I don't necessarily have alternatives readily available to me, I appreciate the medium's quirks and generally attempt to accentuate or play with them in my work. While I more frequently play with my phone’s digital zoom and the artifacting produced therein, I’ve also tinkered with my camera’s treatment of darkness. Aside from this interest based in practice, I’m also interested in the topics of hallucination and pareidolia in general.

[^^ drawing of the Hearthstone card "Spore Hallucination" -- called "Hallucination" until 2025-01-21 ^^]

     One thing I often come back to in my thinking on hallucinations and pareidolia is to what extent these phenomena should be considered separate from their medical contexts. I don't personally have experience living with schizophrenia or schizoaffective disorder (etc.), and I'm not completely caught up on the DSM’s taxonomy for conditions which relate to these things (these things = hallucination + pareidolia) / the DSM’s working definitions for what constitutes these things in and of themselves. If possible, I want to avoid reproducing sanist stereotypes or bigotry, especially of the type which I perceive as prevalent in 2025’s internet media ecology. I hope to speak about these subjects in ways that don't enter into derogatory or which enter the use of slurs, but I don't have a solid sense of where the line of acceptability lies. (I.e., I know well enough to avoid using the most common slurs, but I don’t have enough of an education on the best ways of talking about these topics).

     (Part of what I haven't been able to tell, for example, is to what extent Guattari and Deleuze's framework of schizophrenia plays into or against these systems; for whatever it counts for, G+D’s 1000 Plateaus – the book where G+D outline their understanding of Schizophrenia – was a popular text in Japan in the 1980s, and the ways people were talking about it there at that time used what in English would now be considered slurs;; but that’s besides the point, I’ll talk about G+D some other time).

     In saying all of this, part of what I want to move toward thinking about is to what extent these phenomena are literal (and must be understood literally) and to what extent they can productively be understood as metaphorical (as well as to what extent the boundary between the two lies as blurry).

[^^ drawing of the Balatro joker "Pareidolia"^^]

     On a literal level, it may be false to say a camera hallucinates, given that hallucination is understood as a uniquely psychological effect (since in this framework the camera does not have a psychology). However, I think trying to determine where hallucination is located may not be a fully productive act. Motivating my thinking is that in order for someone to declare something a hallucination, the declarative party must possess a psychology, therefore, it is impossible to address hallucination/to think about it in a way distended from psychological bodies. I wonder if it could be useful to think of it as though through declaration of the hallucination, the speaker relocates it, in part, to themselves, given that they are the one who comes to understand it as such.

     Strictly speaking, hallucination and pareidolia are not things in and of themselves, rather they are semiotic ways of being and relating to meaning. They can only be said to exist when signification happens between units. (I say this hesitantly in some regard as it seems like these definitions are inching towards a kind of anthropocentrism – something I am wary of; However, I don’t know how else to frame things at the moment).

     I am not interested in saying if hallucinations are real or not. Rather I want to emphasize that they come from a kind of discrepancy, something is/becomes something else. In this way, I view pareidolia and hallucination as transformative, and therefore generative matrices of interacting with the world and of making meaning.

[^^ drawing of a funny looking guy ^^]

     On the extreme end, where one accepts that non-humans can hallucinate, I would say machines often hallucinate, especially in computer science fields. Most AI image technology, for example, engages in these practices when it generates something (an image) out of noise – hallucinating into random chaos an image with a subject (that's not noise, that's big tit goth girl at the pool, etc). This is more easily seen with earlier versions of AI image technology, where, for example, you could take the spotty output of the technology trying to hallucinate a dog from noise, and feed it back into the machine, but as data and telling it that it's actually a jellyfish (thus doubling or more the hallucinogenic effect, creating something out of nothing). This accelerates reading meaning into chaos in a demonstrably visible manner.

     Continuing, camera hallucinations, like the kind Geller talks about, observe and generate something too. One could say, for example, that they speak to the air of movement nascent in space and darkness themselves. Cameras have always been able to see in a way that humans do not. Most human eyes with with a different focal length than what most cameras use (e.g., generally the human eye lens is locked at around 24mm [iirc], whereas cameras go all over the place, can toggle between focal lengths, and can go up to 600+mm, seeing things almost impossibly far away). Cameras see the world with a fundamentally different kind of distortion. From a kind of objective perspective, digital noise does not inherently contain meaning; it is data read out from a machine / produced by a machine. Human spectators are the ones who assign meaning to it, via watching it. Imagining movement and time means imaging meaning.

     I believe this is what pareidolia fundamentally is – seeing something into something where it is not there in that same capacity for others. When I frame pareidolia this way, I intend to draw a comparison to kinds of literary analysis and interpretation of a text.

[^^ drawing of ,, something ^^]

     Since my current hyperfixation remains FNAF, and I feel that FNAF provides an evocative subject to reflect what I’m gesturing at, I’ll use it here as my example. The games themselves (and the other FNAF content), say things and contain things. They do so, however, in a restrained kind of way where they don’t say what exactly they are saying. While there are general points of consensus within the FNAF community, the texts explicitly state a comparative dearth of what can be said to be lore, story, or narrative. In this way, any kind of narrative that someone reads into the text, especially if it’s a complex one, will likely, almost necessarily, be hallucinating something into it which was not innately present.

     Because we cannot know what other people know, we cannot know in the moment of seeing or thinking if our senses are hallucinations. I say this since, contrary to the concept of digital cameras, where you can cross reference what the camera sees with what you see (*to be annoying this is false, impossibly so; but we pretend otherwise for argumentative convenience), it is relatively difficult if, if not impossible, to compare something approaching the “Truth” to someone’s interpretation, narration, or imagination of an event or moment.

     In the case of FNAF, Scott, the generally agreed upon authorial voice, is fairly cloistered with regard to saying what he’s cooking or what he means when he signifies something. I wonder if this example could help demonstrate the relativity of truth and of hallucination. Under this framework, hallucination is defined as happening when something occupies a minority position with regard to the majority (terms defined not numerically but by proximity to evental origination and to power). If it’s not possible to know what the majority is (much less to verify it in any external kind of way), then it becomes a taller ask to verify something as a hallucination or not.

[^^ drawing of digital camera imagery; sic digital hallucination ^^]

     There's something personal and indulgent in pareidolia and hallucinations. They are phenomena through which orders a chaotic world (sometimes volitionally, sometimes not). While I would say this is not inherently good, I would also say it's not inherently bad. There are limits to fiction and the unreal, but they can't ever be fully opted out of. The world is chaotic, but it's never fully chaotic. Flexing agency is selfish, but it's also self-affirming.

     With all of this idyllicness being said, though, I wonder how much the possibility of theoretical bad actors using the blamelessness afforded by hallucination to smuggle guilt out of their actions is something that should be considered for within this whole calculus. I think it's useful to respect a kind of truth in certain cases, e.g., truths about what being trans is and what trans people, on the whole, get up to. That someone thinks trans people are groomers or rapists or... is divorced from what the lived realities of trans people are (indeed these anti-trans narratives are phantasmatic as Judith Butler would have it; I think this could be a way of demarcating hallucinations that makes what I'm trying to talk about less potentially messy). There's a difference between opinion and hallucination, probably, but I think something like this example could fit what I'm trying to get at.

     I wondered while writing this if power dynamics could be a relevant dimension for litigating / evaluating the question of hallucinations and pareidolia. When hallucinations victimize oppressed peoples, there is a stake in weighing against or dispelling them. I also think there's a difference in hallucinations about people and their characters and hallucinations about fiction (e.g., fnaf and its story). That is, I think there's probably room for separating between fiction and nonfiction. That said, this is always a slippery slope, and ordering nonfiction is a more precarious gesture than ordering fiction.

     At the same time, I don't want for the potential of bad faith actors to encroach on what's at stake for oppressed people who experience these phenomena. Sanism, as I mentioned, is still societally common and sanist slurs are the most common slurs I hear used online (not to say they are the most severe, just the most frequent). (I mean something by this but I maybe can't explain neatly the exact calculus I'm guesstimating in my head).

[^^ drawing of camera vitality ctd. (see also: my post on analog horror) ^^]

     Anyways. I think there's a lot to chew on here. I am excited by the potentials of digital technologies and I think they provide evocative sites for people to play with fundamental phenomena of being people (e.g., digital as a medium for to work through the intricacies of fiction ...)

     Try it out some time (it here being both the use of digital camera technology and the phenomena of hallucination and pareidolia). Or not, up to you 😀. But that's why I use pareidolia as my video essayist moniker last name