A hypothesis about binocular vision, the observer who never shuts up, and why I am taking it seriously without yet believing it.
I have been assessed a great deal. Assessed for depression, assessed for bipolar, assessed at length by people who were confident and wrong for the better part of thirty years, and finally assessed at sixty-six for the autism and ADHD that had been sitting there the whole time. I have had my blood taken, my sleep questioned, my childhood reconstructed and my personality inventoried.
I have had my eyes tested too, most recently in April. That examination measured refraction, and it found plenty: marked hyperopia in both eyes with astigmatism, a reading addition of two and a half dioptres, a separate prescription for computer distance, and noticeably more correction needed on the right than on the left.
What it did not measure was whether the two eyes work together. Convergence. Accommodative facility. Saccades and pursuits. Fusion. Visual fields. That is a different examination and nobody has done it.
This is not a complaint about optometrists, who did the job in front of them. It is a point about what refraction measures. It tells you how each eye bends light, one at a time. It says nothing about what it costs to run the pair of them together. The test that answers that is a functional vision evaluation, and it needs a behavioural or neuro-optometrist, a specialty far enough off every referral pathway I have been on that nobody assessing me had reason to think of it.
So I have a folder of precise numbers about my eyes and no information at all about whether they agree with each other. Which is a strange position to be in at sixty-seven, and stranger still given how thoroughly the rest of me has been catalogued.
What put the question in front of me was a post by Lindsey Mackereth, who writes Complexity Edge, and who is doing something in it that I want to describe carefully, because the careful version is the interesting one and the sloppy version is everywhere.
Both of the posts I am drawing on are behind her paywall. I am about to summarise a decent chunk of them, which puts me in the position of a man describing a meal in detail to people standing outside the restaurant. If any of this interests you, go and eat there. She cooked it.
What is actually going on behind your face
Right now your brain is receiving two slightly different pictures, one per eye, and fusing them into a single world that feels like it was simply lying there waiting for you. You do not experience the fusion. Only the result.
Mackereth disclosed in March 2026 that she had been diagnosed with severe binocular vision dysfunction. She had 20/20 eyesight. She had also had migraines for the better part of a decade, headaches since first grade, motion sickness, anxiety in visually busy rooms, and fatigue sleep would not resolve. Nobody across four decades of medical contact had joined those up. I wrote about it in the fourth edition of Understanding AuDHD, which is why her May post landed the way it did.
She has divergence insufficiency and esophoria, both of which her neuro-optometrist believes she was born with. Esophoria means the eyes drift inward when relaxed, and the brain, declining to accept double vision, holds them in line with continuous corrective tension. Divergence insufficiency means they resist moving outward as the gaze goes long, so the far world is harder to reconcile than the near one. She makes the point that this is a signal-processing problem rather than merely a muscular one.
The consequence is that the fusion is unstable, and the seam becomes, now and then, perceptible.
Which is a lot of standing overhead for an upgrade nobody consulted us about. Two eyes evolved so we could judge how far away things are, and the invoice for depth perception turns out to be a permanent reconciliation meeting between two departments that were never quite briefed the same way. Most people’s meeting runs itself and files no minutes. Some people’s meeting runs every waking hour and occasionally puts an item back on the agenda.
Metacognition, taken seriously
Most of us met the word as a study skill, which undersells it badly. The precise version is the capacity of a mind to hold itself as an object of its own observation while continuing to operate. Not reflection after the event. A second layer running alongside the first, in real time, watching.
The live monitor is the part that matters here. Am I talking too loudly? Why did I say it like that? It does not wait to be asked and it does not accept a request to stand down, which anyone who has tried to talk themselves out of self-consciousness will recognise as the whole problem. In neurodivergent people it runs hot, and it is commonly described as an observer sitting slightly behind or to one side of your own experience.
The structural parallel
Here is Mackereth’s move. Where binocular fusion is unstable, brains often resort to suppression: one eye becomes the primary feed, the other is downweighted, kept available but not foregrounded.
Set that beside metacognition, which is holding a process as an object of observation while running it, and the shapes match. A brain that spent its development managing two streams that never fully collapsed into one, privileging one and monitoring the gap, may have been rehearsing self-observation at close to the hardware level.
Onset does the heavy lifting in her argument. A brain that loses stable fusion in adulthood experiences disruption to a baseline it once had. A brain that never had the baseline built unity out of instability instead, and built attention, working memory, executive function and self-monitoring around that negotiation.
Most people never get experiential access to the fact that the unified world is constructed rather than given. They live inside the seam. If Mackereth is right, some of us were handed the working file with the tracked changes still showing, and then spent fifty years being told we overthink things.
What the evidence will and will not carry
The prevalence end is on reasonably firm ground. Perna and colleagues meta-analysed forty-six studies covering fifteen million people and found vision and eye problems occur more often in autistic populations. Bellato and colleagues did the same job on the ADHD side, across thirty-five studies and more than three million participants, and found the same direction of travel.
Narrow it to how the two eyes cooperate and the pattern holds. Grönlund’s group in Göteborg examined forty-two children with ADHD against a reference group of fifty, and turned up some abnormal ophthalmological finding in seventy-six per cent of them. Three in four. That is a hit rate high enough that you would expect eye examination to have become a routine part of ADHD assessment somewhere in the following nineteen years. It has not. Off stimulants, heterophoria showed up in twenty-nine per cent against ten per cent of the controls, and abnormal convergence, meaning a near point beyond six centimetres or no convergence at all, in twenty-four per cent against six. Putting the children back on stimulants did not fix it, which is the finding the paper itself leads with. Whatever the eyes were doing, they were not doing it because the attention was broken.
Note what Grönlund’s sample actually was. Forty-two children, mean age twelve. I am sixty-seven. Nobody has run that study on men my age, and forty-two is a small number even for the age it does cover, so I am extrapolating across five decades and a few dozen participants. That is exactly the move I would flag in somebody else’s argument.
Now the caveat, which is substantial. Clavé and Torrents showed in 2025 that the figure swings from twenty per cent down to almost nothing depending on which of the three diagnostic signs you require, and Granet said himself it might be association rather than cause. The link is real. Its size is partly a decision about where somebody draws a line, which is true of more of medicine than medicine likes to advertise.
None of that tests the hypothesis. It establishes that the two things co-occur more than chance, which is where a hypothesis is allowed to start and not where it is allowed to finish.
What my own numbers might be doing
Here is where it stopped being an interesting idea about somebody else. Majumdar and Tripathy, reviewing hyperopia for StatPearls, set out the sequence without ceremony. The hyperopic eye recruits accommodation to drag the image forward onto the retina, and sustained effort of that kind produces asthenopia: eye fatigue with frontal headache, worse after long stretches of near work. Presbyopia adds to the total, which is what a reading addition compensates for. The lens does that job without complaint for about forty-five years and then stops, having given no notice and filed no intention to retire.
The part I had not known is that accommodation and convergence are yoked. The drive to clear the image also drives the eyes inward. So a hyperopic eye working hard to focus is simultaneously being pushed toward convergence it never asked for, and whether that ends in comfort or in strain depends on how much fusional divergence is available to push back. Lembo and colleagues set out that three-way relationship between hyperopia, accommodative convergence and fusional divergence in their review of accommodative esotropia. It is the eso side of the ledger. Mackereth’s side.
Add the asymmetry. More correction on the right than the left means the two eyes are not being asked to do the same amount of work, which is exactly the arrangement that would make a reconciliation problem more plausible rather than less.
I am a counselling psychologist and not an optometrist, and a man reasoning about his own headaches is the least reliable narrator in medicine. So this is a question rather than a finding. But it is a question with a mechanism attached, and nobody has ever asked it.
Why the honesty is the reason to take it seriously
Mackereth does not overclaim, and says so at length. She is not arguing that binocular vision dysfunction produces metacognition, nor that everyone with unstable fusion is a recursive thinker or the reverse. The causal arrows, she says, are unclear and probably run several ways at once.
Her actual proposal is narrower and better. In people already predisposed toward high-complexity processing, congenital binocular dysfunction may be one of several features of a nervous system organised around duality, with the visual and cognitive systems expressing the same underlying architecture in different rooms rather than producing each other.
That is a hypothesis I can do something with, and the reason is the restraint. A version promising me an explanation would be worth nothing. A version that names which part is unsupported is worth reading twice.
The bill
The corrective work and the observer layer draw on the same pool that also runs language, social processing and executive function. Mackereth’s suggestion is that this is a real contributor to the disproportionate fatigue neurodivergent people report, the kind that looks inexplicable from outside and gets filed as anxiety, poor motivation, or a mood disorder.
I was filed as a mood disorder for thirty years. So that sentence went in somewhere below the ribs.
What I felt reading it was not vindication. It was closer to being tired in a new way, on behalf of a younger version of myself who was doing an enormous amount of unpaid work and being told he was simply difficult. Juster, McEwen and Lupien set out in 2010 what an allostatic load index is actually made of: cortisol, catecholamines, inflammatory markers, blood pressure, lipids, waist-to-hip ratio. It is a good list, assembled by people who took the cumulative cost of chronic strain seriously enough to try to measure it. There is no line on it for holding two images together, and none for the observer who never shuts up. Nobody costs invisible labour, and no assessment I ever sat went looking.
So I have written to my doctor and asked whether the test nobody has ordered is worth ordering. I am content to be told it is a dead end. What I am not content to do is leave it sitting there unexamined, which is what has happened to it for sixty-seven years.
References
Bellato, A., Perna, J., Ganapathy, P. S., Solmi, M., Zampieri, A., Cortese, S., & Faraone, S. V. (2023). Association between ADHD and vision problems. A systematic review and meta-analysis. Molecular Psychiatry, 28(1), 410–422. https://doi.org/10.1038/s41380-022-01699-0
Clavé, L., & Torrents, A. (2025). Convergence insufficiency prevalence in attention deficit and hyperactivity disorder children depends on the diagnosis criteria. Ophthalmic and Physiological Optics, 45(1), 23–30. https://doi.org/10.1111/opo.13411
Granet, D. B., Gomi, C. F., Ventura, R., & Miller-Scholte, A. (2005). The relationship between convergence insufficiency and ADHD. Strabismus, 13(4), 163–168. https://doi.org/10.1080/09273970500455436
Grönlund, M. A., Aring, E., Landgren, M., & Hellström, A. (2007). Visual function and ocular features in children and adolescents with attention deficit hyperactivity disorder, with and without treatment with stimulants. Eye, 21(4), 494–502. https://doi.org/10.1038/sj.eye.6702240
Juster, R.-P., McEwen, B. S., & Lupien, S. J. (2010). Allostatic load biomarkers of chronic stress and impact on health and cognition. Neuroscience and Biobehavioral Reviews, 35(1), 2–16. https://doi.org/10.1016/j.neubiorev.2009.10.002
Lembo, A., Serafino, M., Dello Strologo, M., Saunders, R. A., Trivedi, R. H., Villani, E., & Nucci, P. (2019). Accommodative esotropia: The state of the art. International Ophthalmology, 39(2), 497–505. https://doi.org/10.1007/s10792-018-0821-6
Mackereth, L. (2026a, March 22). Nobody told me to check my eyes. Lindsey Mack’s Complexity Edge. https://lindseymackereth.substack.com/p/burned-out-for-decadesthe-hidden
Mackereth, L. (2026b, May 21). One eye sees. The other watches: Metacognition in your eyes? Lindsey Mack’s Complexity Edge. https://lindseymackereth.substack.com/p/your-eyes-might-be-why-you-cant-stop
Majumdar, S., & Tripathy, K. (2023). Hyperopia. In StatPearls. StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK560716/
Perna, J., Bellato, A., Ganapathy, P. S., Solmi, M., Zampieri, A., Faraone, S. V., & Cortese, S. (2023). Association between autism spectrum disorder (ASD) and vision problems. A systematic review and meta-analysis. Molecular Psychiatry, 28(12), 5011–5023. https://doi.org/10.1038/s41380-023-02143-7
Leave a Reply