The law needs a culprit. Criminal law has a job to do, and that job requires an individual. It requires a person who intended an outcome, or foresaw an outcome, or whom caused an outcome through negligence. Someone who can therefore be held to account for it. The entire apparatus, actus reus, mens rea, the defences that sit around them, is a machine for locating responsibility inside a single body and then acting on that body.
This is not incidental to the law, rather it is the law’s founding assumption. This is precisely the assumption the Ecosystemic Model of Distress and Wellbeing was built to dismantle. Not that criminal responsibility is a falsehood, but that criminal acts do not occur in isolation. Even those who can not use such a defence as “…by reason of insanity” can still have their criminal behaviour understood within it’s ecosystemic context.
The ecosystemic position is straightforward; distress is an output, rather than an intrinsic property of a person. It emerges from the interaction between a bodymind and the complex and dynamic environments it exists in relation to.
The immediate sensory and relational environment, the institutional environment of school and service and workplace, and the structural environment of policy, economy, and the neuronormative expectations that saturate all of it. Distress is what happens when those environments demand something a bodymind cannot sustainably give. As opposed to having a specific location, it is the sum of the whole arrangement.
If that is true of distress, then it is also true of behaviour. Since psychologically distressed states and criminality are so entwined in the literature, then when behaviour is criminalised, we arrive at a collision. The law wants to know what was in your head. The ecosystemic account says; look at what your head was in. This, it could be argued, begins to move us away from criminal justice and towards a model by which we can actively work to prevent crime in the first place.
What The Defences Will And Will Not Hear
The thing that should trouble anyone reading criminal defences through an ecosystemic lens is that the law is not entirely deaf to the argument that a person’s capacity was compromised: it has a whole vocabulary for it. However, it will only listen if the compromise is relocated inside the person. It fails to acknowledge the complex cultural and ecosystemic variables that create the context for the crime to occur in in the first place.
- Insanity in England and Wales still runs on the M’Naghten Rules of 1843; a defect of reason, arising from disease of the mind.
- Diminished responsibility under the amended Homicide Act requires an abnormality of mental functioning arising from a recognised medical condition.
- Automatism turns on whether the cause was internal or external, which is why a diabetic’s criminal liability can depend on whether they took their insulin or didn’t. The US position varies by state, but the M’Naghten test dominates there too, and since 2020 a state is free to abolish the insanity defence altogether and permit mental illness only to negate the mental element.
Every single one of these demands the same trade; you may have your reduced culpability, but you must accept a pathologising account of yourself to get it. The law will hear “my brain was disordered”, however, It will not hear “my life was disordered”.
This is the bind, and it is a familiar one to anyone who has fought for support in the SEND system or through adult mental health services. Access to accommodation is gated behind diagnosis; access to mercy is gated behind pathology. In both cases you are handed a script written by someone else and told that speaking it is the price of being heard at all.
Fricker’s testimonial injustice is doing a lot of work here, the defendant’s own account of why they did what they did carries no evidential weight until a psychiatrist restates it in the language of disorder. The expert is the credible speaker, the person is a set of symptoms with a mouth.
Foucault would recognise this immediately. The insanity defence is not a mercy mechanism, it is a sorting mechanism. It separates the mad from the bad, and it does so on terms set by the institutions that will subsequently manage whichever category you land in. Prison or hospital order, either way, you are dealt with.
The Exceptions That Prove The Rule
The law is not wholly incapable of recognising that circumstances compel, however it recognises it in a very particular shape. Duress admits that a person can be so constrained by their situation that blame is inappropriate; though it should be noted that case law has historically set the conditions tightly:
- A threat of death or serious injury
- No safe avenue of escape,
Critically, the defence is unavailable to anyone who voluntarily associated with criminals when they could have foreseen coercion. Section 45 of the Modern Slavery Act 2015 does something similar for people compelled to offend through trafficking or slavery.
So the law will accept ecosystemic compulsion when it takes the form of a specific person making a specific threat, but what it will not accept is diffuse structural coercion. There is no defence of poverty or of having been excluded from education at nine, restrained through your adolescence, and left without a single functioning support relationship by the time you were an adult. There is no defence of being a monotropic bodymind in an environment engineered to produce monotropic split, sustained in that state for two decades, and then criminalised for what the resulting distress made of you.
This is the sharpest critique of criminal justice provided by ecological models. The law punishes people for the ecosystem they were placed in. If you grew up somewhere where the only available relational network was a criminal one (because that was what the neighbourhood contained, because that was who would have you, because every legitimate institution had already excluded you) the law treats your presence in that network as a free choice that forfeits the defence of coercion. It reads adaptation to a hostile environment as complicity in it.
The same logic governs intoxication. Some case law holds that voluntary intoxication is no answer to a basic intent crime, with some cases law confirming that a drugged intent is still an intent. In the US, case law upheld a state’s power to exclude intoxication evidence entirely. Addiction is thereby constituted as a series of freely made choices, the first of which contaminates all the rest.
English law will allow alcohol dependence syndrome to assert diminished responsibility, but only for murder, and only once psychiatry has certified the dependency as a disease. Again, the pathology toll has to be paid. Your compulsion is legible only after it has been medicalised.
Where I Would Draw The Line
None of the above is an argument that nobody is responsible for anything. That would be an unserious reading of the ecosystemic model, and it would also be a cruel one, because it disappears the person who was harmed. Any account of criminalised behaviour that cannot look a victim in the face is not an account worth having.
The distinction I would draw is that criminal responsibility is an allocation of state power. It is the question of who the state will act against, by what means, and with what force. It is not a metaphysical judgement about the moral condition of a soul, rather, an institutional decision with material consequences; imprisonment, detention, restraint, the loss of a life’s remainder.
Understood that way, the question is not “did this person really do it” but “is the deployment of this power against this person, in this situation, justified and productive of anything good?”. Ethical responsibility is relationally dependent and it is answerable. It is the question of who owes what to whom in the aftermath of harm, and crucially, how can we learn from this to prevent harm occurring again?
The ecosystemic model does not dissolve ethical responsibility, it multiplies it. If distress is an output of complex and variable relationships between environments (with the bodymind of the aggressor and victim consituting environments in their own rights), then so is the harm that flows from distress; everyone who shaped those environments is in the frame.
- The commissioner who defunded the service.
- The school that used restraint on a child for years and called it behaviour management.
- The panel that refused the EHCP.
- The system that made a diagnosis a prerequisite for help and then imposed a six-year wait for one.
These parties hold real, ethical responsibility for outcomes they made predictable. Currently they face no accountability of the criminal kind, and the person in the dock carries all of it. That is not a balanced allocation. It is tantamount to a laundering operation, in which structural violence is converted into individual criminality and then filed away.
The person in the dock does not vanish from their own life, and this is where I part company with any purely deterministic reading. An ecosystem constrains; it does not author. A person acting within a hostile arrangement is still acting.
Deleuze and Guattari’s line of flight is not something that happens to you; it is taken, from within an ecosystem, by someone. To say the ecosystems produced the conditions for an act is not to say that nobody was present when it happened. Agency under constraint is still agency. It is just smaller, and more expensive, and harder won than the law’s fantasy of the freely choosing individual permits it to acknowledge.
So, I would argue that criminal responsibility is paradoxically considerably narrower than it’s current approach, and immeasurably broader than simply criminalising individuals, ethical responsibility lies within the ecosystem. Fewer people should face the full force of the state for behaviour that was manufactured by conditions imposed on them. Far more parties should be answerable (publicly, materially, and by name) for manufacturing those conditions.
The Risk In The Argument
We should be upfront about where this reasoning can be turned against us, because it can, and it already has been.
If you successfully argue that a person’s behaviour was produced by their ecosystem and their neurocognition, you have also argued that the risk is ongoing and not attributable to a discrete choice. That is exactly the logic of indefinite preventive detention. Autistic and learning disabled people have been held for years in assessment and treatment units on the reasoning that their risk is constitutional and dependent on a lack of community provision and therefore never quite discharged.
This is the terrifying inversion, the very account that ought to generate mercy generates permanent management instead. The way out is not to abandon the ecosystemic model. It is to insist that if the ecosystem produced the risk, the ecosystem is the object of intervention. If distress is an output of complex environments, then changing the environment is the treatment.
Autism + Environment = Outcome
Beardon’s Golden Equation (note that in this respect autism can simply be replaced with “individual”)
Detaining the output changes nothing, it is simply cheaper, and of lesser existential threat to power structures that are complicit in the creation of the very harm they are criminalising.
Desistance from offending, like recovery from addiction, is not linear. It does not proceed by a single decisive act of will, instead proceeding in small, sustainable, multidirectional movements, each conditional on the establishment of an anchor point stable enough from which to move forward to the next step in the journey (the basis of Tanya Adkin’s lilipadding concept).
A sentencing framework built on the assumption of a single moment of moral reform will fail people continuously and then blame them for the failure. A framework that expects non-linearity would look completely different, and would understand relapse as information about the ecosystem rather than evidence of the person’s bad character.
What Justice Reform Would Actually Have To Do
The Law Commission has an insanity defence project queued to begin once its Criminal Appeals and Contempt of Court work concludes in 2026/27, and it has previously recommended replacing M’Naghten with a verdict of “not criminally responsible by reason of recognised medical condition”. That would be an improvement on a test written in 1843, but would still locate the whole story inside the defendant’s head.
An ecosystemically honest reform would not principally be a defence at all. Defences, like distress, are downstream issues. By the time a defence is being argued, the prosecution has already happened, the person has already been constituted as a defendant, and the structural question has already been foreclosed. Real reform would live upstream:
- Charging decisions
- The abolition of the routine criminalisation of distress
- Genuine liability for institutions that produce foreseeable harm
- A diversion infrastructurr that does not require a diagnosis as an entry ticket
- A reduced focus on criminal punishment and an increased focus on the addressing of ecosystemic causes of crime
Until then, the description we have is that our legal system will offer a person reduced culpability on the condition that they accept a description of themselves as broken, while extending complete immunity to every institution that did the breaking.
That is not justice. It is bookkeeping.
