Building the lens

Autism is a difference in how we attend to the world around us and within us.

How I put it together

Monotropism × Predictive Processing × Allostasis

Monotropism describes attention as a limited resource that interests compete for. In a monotropic mind, attention gathers around a few interests at a time and stays there, so whatever holds it is processed deeply and everything outside it fades into the background.

The research behind it

Developed by Dinah Murray, Mike Lesser and Wenn Lawson, and carried forward largely by autistic researchers. Attention is a limited resource. Monotropic minds concentrate it in a few places at a time, deeply; polytropic minds spread it across many. Many autistic traits follow from how attention is distributed.

Predictive processing grew out of two lines of work: predictive coding, a model of how the brain passes predictions down and errors up through its layers (Rao and Ballard, 1999), and Karl Friston’s free-energy principle, which treats the brain as a prediction machine working to minimize surprise (Friston, 2010). In short: the brain predicts, and only the mismatch gets the expensive, bottom-up treatment.

The research behind it

Applied to autism, the major models disagree on where the difference lies: weaker predictions (Pellicano and Burr), atypical weighting of precision (Lawson, Rees and Friston), or prediction errors given high, inflexible weight (Van de Cruys and colleagues). The lens builds on the last.

Allostasis is the body’s way of staying stable by changing ahead of time. The brain predicts what the body will need next (glucose, oxygen, heart rate, attention) and arranges it in advance. The mind’s most important job is to keep the budget balanced.

Allostasis itself is well established in physiology. The idea that the brain’s predictions run the whole body budget is newer, widely influential, and still being tested.

The research behind it

Sterling and Eyer named it: stability through anticipation rather than reaction. Barrett and colleagues tie it to predictive processing, with the brain’s predictions of the body’s needs (interoceptive predictions) doing the budgeting.

HypothesisDeep and narrow attention builds detailed and precise predictions.

Predictions are built from what attention has already processed. Polytropic attention is spread across many things at once, so its predictions are broad and forgiving. Monotropic attention is pulled into a few places, so its predictions are detailed and precise. (Van de Cruys and colleagues, 2014: autistic predictions are too precise, not too weak, and every mismatch is given a high, inflexible weight.)

What are the major predictive-processing theories of autism?

1. Weak predictions (Pellicano and Burr, 2012). Autistic predictions, or “priors,” are weaker than typical, so incoming sensory information isn’t smoothed out by expectation. The world arrives raw and “too real.”

2. Precise, inflexible prediction errors (Van de Cruys and colleagues, 2014; Lawson, Rees and Friston, 2014). Predictions aren’t weak. The difference lies in how much weight, or precision, the brain gives to mismatches. Every prediction error is treated as important and hard to ignore, so the brain can’t settle on “close enough.”

The lens builds on the second: monotropic attention builds detailed, precise predictions, and a precise system treats every mismatch as worth processing.

HypothesisDetailed and precise predictions → more prediction error → more sensory input that requires bottom-up processing.

For a monotropic thinker it can feel like prediction error, like processing demands, are everywhere, and that every one of them demands full processing.

HypothesisThe deep processing style of monotropic thinkers can also result in greater weight being assigned to each prediction error that crosses the threshold.

ConceptBottom-up processing demands are highly energy-consumptive. They stress the body budget.

When the brain meets something it didn’t predict, it has to build it from the bottom up: gathering the details, then adding layers of significance (what does this mean, is it safe, how do I feel about it?). That is some of the most expensive work the brain does, and emotional and sensory significance cost the most.

The brain doesn’t pay the cost of bottom-up processing alone, for two reasons.

The body helps build the experience. The brain constructs each moment in the body as well as the head, sending motor and visceromotor predictions that tense muscles, shift breathing and heart rate, and ready the body to respond. All that consumes cellular energy.
Processing in the mind and sensing and responding in the body draw from the same account. The brain uses about a fifth of the body’s energy at rest, drawn from the same blood, oxygen and glucose as every other organ. When processing demands rise, it rearranges the whole budget to keep up: heart rate climbs, glucose is released, digestion and repair are put on hold.

Neuroscientist Lisa Feldman Barrett calls this the body budget: the brain’s running account of the energy every part of the body needs.

Are you saying that deep focus drains the body budget?

No, not exactly. This is where it gets tricky. Autistic thinkers often find deep focus restful. Inside the tunnel, attention narrows and processing resources pour into one thing, so the stimuli outside it go unnoticed. Fewer channels to process means less prediction error, and less to pay for.

But that only works when what lies outside the tunnel is safe: safe for the senses, and safe emotionally. A child can’t zoom in at school unless their teacher makes it safe to miss what’s happening outside the beam, both in the room and in the world of judgment and emotion. Deep focus in a safe space restores. Deep focus held while bracing for the next interruption doesn’t.

Prediction error isn’t an error, it’s a signal to pay attention.

Something unexpected or unknown requires bottom-up processing.

Inside the lines
the mind relies on predictions to navigate the world
Outside the lines
prediction error significant enough to require bottom-up processing

Prediction style (move the sliders ↓)

Broad and forgiving
(polytropic)Detailed and precise
(monotropic)

Environment

Familiar
(low prediction error)Unfamiliar
(high prediction error)

Is it safe for attention to narrow?

What does narrowing attention look like?

Physical tools like headphones, a hood, or a pressure vest, or psychological tools like attention tunnels, sameness, routines, or saying no.

No
(broad attention required for safety)Yes
(attention is free to dive)

Resources in the bank

Low
(sick, tired, hungry, alone)High
(rested, well, fed, supported)

Body budget

Energy to spareStressed but manageableOver capacity

Understand the limitations of this visual representation.

Accurate enough to show the idea, and inaccurate in one important way. It’s not that predictions leave the mind and sensory data comes up and they meet in the middle. The mind is always a step ahead; it predicts before it senses. Predictions, what the brain thinks it needs the cells to do to navigate the very next moment, have already landed in the cells before that sensory experience arrives. So the interface starts in the periphery, in the body, long before the brain. Once prediction error breaks through, the mind’s interoceptive network (insula, anterior cingulate, amygdala, ventromedial prefrontal cortex) filters it and assigns significance.

This model is still in beta. Tell me what needs adjusting or adding.

HypothesisNonautistic people rely heavily on predictive processing to reduce processing demands. Autistic thinkers need a different strategy.

Managing allostatic load is something every mind has to do to survive. Most minds do it by predicting: broad, forgiving predictions cover most of what happens, so little has to be built from the bottom up. Autistic thinkers just can’t lean on predictive processing in the same way. Their detailed, precise predictions miss more often, so more of the world arrives as something new that has to be processed.

HypothesisMany autistic traits are strategies to reduce processing demands.

Rigidity, sameness, compartmentalization, avoidance: these are strategies to avoid more processing demands, more threads to pull. Not that monotropic thinkers don’t love to pull threads. They tend to pull one at a time. These aren’t deficits. They are a mind protecting its body budget, by lowering prediction error or limiting how much arrives at once:

Sameness and routine make the world predictable. Fewer surprises means less bottom-up processing.
Rigidity and rules keep precise predictions intact. A change in plans isn’t a small thing to a precise system; it is a cascade of mismatches.
Deep interests are well-predicted territory. Inside an attention tunnel, a monotropic mind can go deep at low cost, which is why it can feel restful.
Stimming creates sensory input the mind can predict, a steady signal that helps tune out the unpredictable ones.
Scripts and echolalia turn the most unpredictable demand, conversation, into something rehearsed.
Compartmentalization keeps one context, one set of predictions, active at a time.
Physical tools like headphones, hoods and weighted blankets turn down unpredictable input, or replace it with steady, predictable pressure.
Avoidance and withdrawal cut the input when the budget can’t cover it.

What happens when we take these strategies away?

If we extinguish the strategy without lowering the demand, the cost doesn’t go away. It goes up. Masking is the clearest example: suppressing stims, scripts and avoidance while still processing everything underneath. It looks like coping, but it spends the body budget twice.

Support works the other way around. Lower the demand first, and the strategy is needed less.

What is low demand parenting?

Low demand parenting means dropping every demand that isn’t essential, so a body budget stuck in survival mode can recover. It works by lowering processing demands, especially emotional ones, and by letting kids keep the strategies above instead of taking them away.

Pull on the full thread →

HypothesisWithout the space (and safety) to reduce processing demands, processing capacity is drawn low and the mind quickly shifts to conserve.

Overwhelm is fundamentally a ratio problem. It is the alert that tells us when processing demands have exceeded capacity. That we no longer have enough available resources to manage what comes next.

Attention constricts, locking onto the most immediate demand and dropping everything else. The prefrontal cortex, our most sophisticated and energy-intensive processing center, begins to go offline. The nervous system shifts from flexible, creative, open-ended thinking to something faster and more rigid: react, protect, survive.

And if we don’t listen, the body doesn’t stop sending messages. It sends louder ones.

When energy supply is low, or expected to be low, the body first conserves. And some of the first tasks to stop are those related to cellular clean up. Skip cellular cleanup for too long and cellular waste accumulates. The body reads that waste as damage and the immune system responds with inflammation. Inflammation adds to the expense.

Burnout is the state of a body budget that has been overdrawn one too many times. In this state, even a small demand will feel like more than the system can bear.

The lens

Autism is a difference in how we attend (notice, process and respond) to the world around us and within us.

In environments that keep processing demands high, that difference stresses the body budget. And in that stressed state, inflammatory triggers can push cells into deep conservation mode.

How I think this changes how we approach autism

Protect attention tunnels. An attention tunnel is where a monotropic mind goes deep: where it rests, learns and does its best work. Kids need protected time inside them, and care with the transitions in and out.
Rethink emotional regulation and self-care. Regulation isn’t a skill to drill. It’s what becomes possible when the body budget has room. Self-care means protecting capacity, not just calming down.
Retire the stigma around emotion, empathy and theory of mind. It’s not that autistic people don’t feel. Sometimes they feel too much, and feeling deeply takes capacity.
Look at the whole body. A stressed body budget doesn’t stay in the brain. It shows up as inflammation all around the body.
Lower demands, and learn from autistic adults. In childhood, that looks like low demand parenting. Autistic adults show us what it looks like across a lifetime.

More is coming. Each of these will grow into its own thread on my upcoming page, Applying the lens. In the meantime, pull on threads with me in Exploring the lens.