A Neo-Formalist Approach to Imperfect Praxis
By Emmanuel Harish Menon // Posted on: Aug 26, 2025
This is a script for a video essay written in the winter of 2025 for my Screen Studies class at AFTRS, in which I explore a potential response to the AI-fication and sanitisation of mainstream cinema through Dr. Alistair Gall’s framework of ‘imperfect praxis’. The video is embedded below but I’ve (very quickly) added citations to the script where I saw fit. It was written and edited under a bit of tight crunch—i.e. the usual post-procrastination caffeine binge and sleep deprivation—and the V/O was all done in a single take on a very shitty mic. Despite all this, this assignment—and corresponding class—were the one thing I absolutely had the most fun in all the three years I spent at AFTRS and I hope you enjoy watching/reading this as much as I enjoyed making it.
Over the past three years, we’ve seen near-exponential progress in the capabilities of AI, with little to no signs that this progress is slowing. In fact, a recent paper led by ex-OpenAI researcher Daniel Kokotajlo forecasts that, by ~September 2026, AI agents will match a professional programmer’s coding abilities while simultaneously being able to “think” at 15x the speed of the average human brain. This has—and should continue to—force a critical re-evaluation of our approach to almost every field of study from the scientific to the artistic—film included.
Every single major film studio has already integrated AI tools into every step of the production pipeline. Disney’s Zurich-based research team has even created FaceDirector—a tool that utilises AI to composite an actor’s facial expressions from across multiple different takes into a single unbroken shot, allowing them to manipulate performances in post-production . As scary as that sounds, it’s important to note that these are assistive tools and that in the right hands, with the right goals, these tools can be incredibly useful. However, in the hands of these major corporations—who already sell us soulless, repackaged nostalgia in an attempt to sate the insatiable appetites of their shareholders—achieving their goal of maximal profit means universal appeal, which in turn means sanitisation and dilution of themes, messaging and stories to ensure broad marketability across cultures. Given AI’s inability to capture cultural subtleties due to the vast amounts of data it’s trained on , the integration of it into this hyper-commercialised, homogenised approach to the creation of art is bound to only accelerate the sanitisation we’re seeing all across the cinematic mainstream.
So how do we respond to this?
We look to the past.
In 1979, a Cuban filmmaker named Julio Garcia Espinosa published an essay titled “For an Imperfect Cinema” . Motivated by what he saw as the culturally imperialist tendencies of the cinematic exports of the first world—which placed much emphasis on bourgeois notions of technical quality—he argued for the establishment of a Third Cinema in direct opposition to the capitalist sensibilities of Hollywood as well as the arthouse lean of European films. Espinosa spends much of the essay laying out the foundational context of this so-called Imperfect Cinema, before finally defining it, stating—
Imperfect cinema is no longer interested in quality or technique. It can be created equally well with a [35mm camera] or with an 8mm camera, in a studio or in a guerrilla camp in the middle of the jungle. Imperfect cinema is no longer interested in predetermined taste, and much less in ‘good taste.’…The only thing it is interested in is how an artist responds to the following question: What are you doing in order to overcome the barrier of the “cultured” elite audience which up to now has conditioned the form of your work?
Far from being dated, Espinosa’s fight against cultural imperialism directly parallels the challenge of cultural homogenisation that we face in this era of cinema. Interestingly, the ideas put forth in his essay are far ahead of his time—he predicts the “universalisation of college-level instruction” as well as the democratisation of film technology and even the impact that the invention of the television would have on cinemas worldwide.
The ideas at the heart of Espinosa’s Third Cinema would later be expanded upon by Dr. Allister Gall in his 2016 thesis titled “Towards a Cinema of Imperfection: Participatory Film as Research”. In it, he defines his central idea of ‘imperfect praxis’—a film-specific participatory research method that resists the often-rigid nature of theoretical academic work by embracing a DIY, disruptive approach towards what is a typically procedural methodology . He then goes on to outline the driving question of the thesis—can this framework of imperfect praxis create emancipatory value by making filmmaking accessible to all through the emerging social space of the internet?
Gall’s work leans very heavily towards exploring the participatory and theoretical dimensions of imperfect praxis. The idea at its core, however, is to deliberately disrupt these default, widely accepted methods of practice or creation in order to free the creator of this burden of technical quality and aesthetic polish that has increasingly become the focal point of the cinematic artform, superseding meaning-making and storytelling as the primary driver of the creative act. As more and more studios begin to replace redundant roles throughout the production pipeline with AI tools, it stands to reason that, when faced with creative decisions, these tools will have in mind the same goal of achieving broad universal appeal that studio executives have, leading to an acceleration of the ongoing sanitisation of screen content.
This idea of deliberate methodological disruption at the center of Gall’s imperfect praxis meshes well with the methodological approach of neo-formalism. While imperfect praxis concerns itself with the methods of creating, neo-formalist film theory deals with the consumption of said creation, exploring the relationship between a spectator and a piece of art through a parametric framework, forcing a viewer to critique each film on the strength of the unique formal elements utilised within the film, rather than through the rote, often-formulaic approach typical of other avenues of criticism.
A key principle of neo-formalism is defamiliarisation—the act of presenting familiar things in an unfamiliar way so as to disrupt the viewer’s habitualised perception of reality. As noted by prominent neo-formalist scholar Kristin Thomson however, the effectiveness of this defamiliarisation is tied directly to the historical context in which a piece is produced—devices that may be new and defamiliarising at the time of creation will slowly decline in effectiveness as their use becomes more and more commonplace . This synergises perfectly with Gall’s definition of imperfection as “methods against methods” and by synthesising the principles at the heart of these two parallel approaches to film, we arrive at a framework that—unlike frameworks created by past movements such as Dogme 95, Nick Zedd’s Cinema of Transgression or even Third Cinema—does not center itself around ideology or specific techniques, but rather the principal philosophy driving the creative act itself.
Nick Murcott’s 2023 short, Gooners is an excellent film to explore this proposed framework with, especially given Murcott’s use of AI-generated elements through the film. Created as part of a university assignment, Murcott originally intended to shoot the entire film on the Mini DV format, was unable to do so due to a requirement that a set amount of the film be shot utilising hi-res cinema cameras. In the edit, however, Murcott settled on using this limitation to create an aesthetic juxtaposition between the “jarringly high-definition…moments of clarity for the main character [and] scrunched-up images of his disintegrating mind”, going on to state that the edit pace of the film attempts evoke the information overload that comes with the over-consumption of the internet. Interestingly, this has similarities to a Brechtian technique known as verfremdungseffekt, a German term roughly translating to “estrangement effect”, a technique where stylistic devices are used to distance the viewers from the characters so as to critically engage the viewer, drawing their attention instead to the formal construction of the piece. Gooners inverts this technique—utilising its frenetic pace to instead situate the viewer deeper into the main character’s perspective—with the achieved effect being the same, drawing the viewer’s attention to the construction of the film. This is a clear example of both defamiliarisation and the deliberate disruption of a method used since early in the 20th century, clearly demonstrating the driving philosophy of my proposed framework in practice.
This philosophy is something that AI will never be able to reproduce by itself. AI thrives on mimesis—if we choose to respond to the AI-driven sanitisation of art by simply chasing aesthetic imperfection, we fall into the same trap we face now—the unfamiliar we create will be rendered familiar in the mere blink of an eye. Instead, if we apply this guiding philosophy of utilising deliberately disruptive, defamiliarising methods of creation to our pursuit of artistic expression and analysis, we will find ourselves fighting this capitalistic, tech-bro driven homogenisation of art on terms dictated by us, the creatives, and us, the consumers. If we do not shift the goalposts set by these mega-corporations—grooming us to endlessly chase the high of systematic optimisation and perfection—we will lose this fight and we will lose it decisively. The only way forward, is for us to bring those goalposts back under our control by embracing our human imperfection with open arms and returning our focus—both as creators and consumers—to the human expression of meaning and emotion that is the core of all art.
Bibliography
Tagged with:
film
/
school