Rachel Irons

Long Live The Toxic Avenger 

 

He was 98 lbs. of solid nerd until he became...

 

The Toxic Avenger is a 1984 low-budget horror/comedy/superhero flick about a nerdy, weak gym janitor who through an exposure to toxic waste transforms into a misshapen town hero. The film is part spoof and part earnest advocate against the corruption and environmental damage that have become normalised in Tromaville. 

I discuss taking up the mantle of toxicity as a weapon against an established system of organisation. Just as the Toxic avenger takes up the ‘waste’, rejected by dominant societal structures, I argue for toxicity as a tactic of illegibility against systems of categorisation and control. 

 

Powerful systems and corporations depend ever more on data extraction for profit, power and control, this data extraction is built on systems of identification, categorization and tagging. Sorting images, objects, consumers and their habits into identifiable categories enhances the prediction power of data systems. 

The acquisition and control of comprehensive data sets mirrors what Deleuze saw in the emergence of the societies of control, the society of control is one where the individual is managed through administration, data modulation and regulation. He describes and foresees a society in which power is no longer industrial but informational.

 

Deleuze draws heavily on Burroughs here and gives conceptual weight to Burroughs ideas, which were at once insightful but visceral and experimental. In the limits on control written by Burroughs in 1978 the author discusses how different control mechanisms operate throughout society. Burroughs’ critique of control has its roots in the original control machine; the word, like a virus that spreads person to person forming perception and creating control through repeated symbolic forms . Language was the most potent manipulative force, not just The Word but the way it existed through image and communication. For Burroughs, resistance requires disruption in the material through which control reproduces itself: language, there is no evading a control system so pervasive by remixing it. Through forgery and falsehoods power is exerted in Burroughs’ work because it does not make a claim to truth based on external references but refers only to itself.

 

Data systems reinforce and reestablish the limits of The Word, they actively ‘tag’ items, identities and objects forcing them into established and accepted categories, it then in  its operation re-established and perpetuates categorisation. For Burroughs systems of control exist to prolong and propagate themselves, like viruses. 

 

The idea of generality is central to Deleuze’s estimation of what defines newness, and the qualifiers of repetition. Generality is linked to a way of seeing the world in terms of equivalences, one object or term can be replaced by another because they share value, they can be substituted or exchanged for one another, in this way Generalities belong to a world of capital. Repetitions on the other hand are non-exchangeable, a reflection cannot be exchanged for its original, its existence reaffirms the unique singular event of the original. 

Anything that seeks to predict reality sits within generalities and concepts, attempting to group things under an umbrella of equivalent value, if this happens, the same will happen next time. It is a sense of generality that groups objects and people together in data and statistical prediction, outliers and unpredictable objects deviate significantly from the rest of a dataset. They can distort statistical metrics like the mean and skew machine learning models. 

 

Opportunity arises in the form of accident, toxicity and pollution that spontaneously interrupt the language of categorisation and capture. 

“The notion of visibility connotes, what can be seen, often drawing on meteorological terminology referring to the distance we can see, the extent of clear vision, which can be potentially restricted by fog, mist, pollution etc”

What is the Fog or Mist that can be used to obscure visibility? Pollution already exists as a term connected with P2P sites and the act of copying. Pollution in P2P refers not only to low resolution links, but to uploads that were mis-tagged, corrupted and misleading. The uploaders of Pollution could be blacklisted from the host site as they obscured access to the desired file. 

The same Toxic Waste that was hidden and shunned by the people of Tromaville gave Toxie the power he needed to disrupt the predicted narrative and become a major destructive force to the established powers and systems of control. It offers a tactic of illegibility that resists capture and categorisation. 

 

Pollution as opposed to the poor image resists generality, it is not generally exchangeable for the original, it cannot be cannibalised into a data set for creating reproductions. Pollution is not just low in resolution but is intrinsically wrong, it exists somewhere beyond the line of a passable copy. Incorporation of pollution, corrupted or misleading artefacts into a data set can cause adverse effects to the machine learning model, this kind of intentional corruption is often referred to as ‘data poisoning’. Data Poisoning is an intentional introduction of misleading or incorrect data to compromise the legibility or coherence of its results. 

 

Let's imagine as we did earlier, through Burroughs, that the word is a virus wielded  by the powerful to present a certain image of the world, producing habits of perception, repeated systems of interpretation and belief. 

An embracing of Pollution, becoming toxic is not just an attempt in data poisoning but a cultivation of a habit of misunderstanding, and frustrates the smooth reproduction of meaning upon which predictive systems depend.