Article 2 – There Is No Standard Observer
This is the second article in “The Brain Behind the Behaviour, a three-part series” that began with a lecture by Nancy Kanwisher and a question about how the brain processes perception. The first article established the scientific foundation: the brain contains dedicated hardware regions for specific functions, broader networks for distributed ones, and a layered architecture (hardware, operating system, and application layer) that constructs perception before we are even aware of it. This article makes the argument that follows from that foundation.
In colorimetry (the science of measuring color) there is a concept called the Standard Observer. It is a mathematical model of the average human visual system, used as a reference point for color measurement, display calibration, and design. It is an enormously useful tool. It makes reproducible color communication possible across industries and continents.
It describes almost no actual human being.
Every real observer deviates from the standard. Cone cell distributions vary. Lens pigmentation changes with age. Cultural and linguistic categories carve color space differently. The predictive model each brain has built from its own history of color experience produces its own calibration offset.
The standard observer is a necessary fiction. It is a shared reference point that enables communication precisely because it does not attempt to capture any individual’s actual experience. The standard observer in psychology, in education, and in workplace design is the same kind of fiction. It is a useful abstraction calibrated to a statistical mode; it is the configuration that appeared most frequently in the populations that built the measuring instruments. Unfortunately, the cost of treating the standard observer as a ground truth is not just inaccurate measurements. It is missed potential, misaligned systems, and a systematic inability to see what the instrument was never designed to detect.
What the Architecture Tells Us
The first article in this series described how the brain is organised into two distinct types of processing: dedicated hardware regions that handle specific functions with extraordinary precision, and broad distributed networks through which higher-order cognition emerges from connectivity rather than localisation. Between them sits the operating system: the connectivity patterns that mediate between hardware and the application layer of conscious experience.
This distinction turns out to matter enormously for how we understand neurodivergent difference, because different neurotypes appear to differ along this axis in structurally distinct ways. Not all neurodivergent brains are differently configured in the same direction or at the same level of the stack. And this is where the conversation about neurodiversity tends to lose precision.
Take autism and ADHD. Both are described as neurodivergent. Both involve differences in attention, executive function, and social processing. Both are frequently conflated in popular discussion. But their neural signatures are structurally different in ways that matter.
Neuroimaging studies of large cohorts show that autistic individuals tend to display locally concentrated cortical differences (particularly in the superior temporal cortex, a region involved in language and social processing), alongside a characteristic connectivity pattern: stronger than typical local connections within regions, and weaker than typical long-range connections between them. The brain is, in a meaningful sense, more internally focused, more locally dense, and less integrated across distance.
One specific consequence of this connectivity profile involves the temporoparietal junction: a region Kanwisher’s research has shown to activate selectively when we think about what other people are thinking, as distinct from thinking about physical representations or even about other people’s pain. This is the hardware for what cognitive scientists call mentalising: the ability to infer another person’s beliefs, intentions, and mental states.
Autistic individuals show characteristic differences in how this region represents information about intentional versus accidental actions. It’s not an absence of the region, but a different information-processing profile within it. This is worth naming precisely, because it clarifies something the popular framing of autism often muddies: the difficulty many autistic people experience in social situations is not a failure of social interest or intelligence, and it is not a general-purpose cognitive deficit. It is a specific consequence of how one dedicated region, among several that are differently configured, processes a specific kind of information.
The practical consequence is visible in the autistic software engineer who spots patterns in data that colleagues experience as noise, because local connectivity that amplifies detail within a domain, at the cost of distributing attention broadly across domains, is precisely the configuration that makes deep pattern recognition possible. The mentalising difference is part of the same architecture. They are not separate problems. They come from the same source.
ADHD, by contrast, shows more globally distributed cortical differences, with dopaminergic regulation differences that affect attention and motivation networks broadly rather than locally. Research findings in this area are more varied than for autism or dyslexia, and group-level patterns, including differences in cortical surface characteristics and dopaminergic regulation, should be read as tendencies rather than fixed signatures. What the pattern consistently shows is a dopaminergic system that rewards novelty, movement, and exploration rather than sustained execution. Which is why the same person who cannot finish a report may be the entrepreneur who thrives in the chaotic, high-stimulus early stages of a startup. Not despite their ADHD, but because of the motivational architecture that comes with it.
Dyslexia follows yet another pattern. The phonological processing region in the left temporal-parietal cortex (the specific hardware responsible for mapping visual symbols onto sounds) shows consistent structural differences in dyslexic individuals. This is a hardware-layer variation, localised and specific, producing a specific functional gap: difficulty decoding written language. But the broader distributed network (particularly spatial reasoning, pattern recognition, and three-dimensional processing) frequently shows enhanced connectivity in the same individuals. The architect who cannot produce a clean written report but moves through three-dimensional space with intuitive precision, whose buildings flow in ways that purely verbal thinkers could not have conceived, is not working around a deficit. They are working from a different hardware configuration, one that closes a door in one direction while opening a larger one in another.
Synesthesia presents differently still. The hardware regions are intact, and the sensory processing architecture works normally. What differs is the connectivity between regions that in most brains operate with clear inhibitory boundaries; the grapheme-processing region and the color-processing region, for instance, connect more freely than typical, producing the involuntary color experience that accompanies letters or numbers. The artist who sees music as color, who paints the emotional tone of a piece in ways that other people may not see but certainly feel, is not confused about sensory categories. They are working from an OS that runs connections most brains suppress, and producing something from those connections that a standard-observer brain cannot access in the same way.
Four neurotypes. Four structurally distinct types of difference. None of them reducible to the others. None of them, on examination, straightforwardly a deficit.
The Calibration Is Not the Deviation
This is the argument that the architecture makes possible, and that the standard-observer framing makes precise: the calibration is not the deviation.
A spectrophotometer that reads slightly differently from the reference standard is not a broken spectrophotometer. It is a calibrated instrument with its own characteristic offset. The offset is real, measurable, and consequential for certain applications. But it is not an error in the instrument. It is a property of it.
The same applies to a brain whose processing stack is configured differently from the assumed standard. The autistic brain that weights bottom-up sensory input (the raw signal arriving from the senses) more heavily than top-down prediction (the brain’s model of what it expects to perceive) is not a brain that is failing to predict correctly. It is a brain whose prediction-to-signal balance is set differently, with consequences in both directions. More fidelity to raw sensory input means more perceptual detail, more resistance to certain optical illusions, more sensitivity to pattern and irregularity. It also means more sensory overwhelm in environments designed for a brain that filters more aggressively. The configuration produces both the strength and the difficulty. They are not separable, because they come from the same source.
The dyslexic brain that struggles to map symbols onto phonemes is not a brain that is trying and failing to do what a neurotypical brain does easily. It is a brain whose hardware is configured differently in one specific region, with a characteristic offset that affects reading acquisition, and a characteristic connectivity profile in the distributed network that frequently supports spatial and three-dimensional reasoning in ways the standard-observer brain does not.
The assumption that certain perceptual responses are fixed biological universals extends further than reading. The preference for consonant over dissonant sound combinations — the acoustic quality most Western listeners experience as inherently pleasing — has long been assumed to be hardwired. Research by the neuroscientist Josh McDermott and his colleagues, working with the Tsimane’ people of the Bolivian rainforest, challenged that assumption directly. The Tsimane’, who have minimal exposure to Western music, show no particular preference for consonance over dissonance. What had appeared to be a biological constant turned out to be a calibration shaped by cultural exposure. The standard observer, it seems, is not just a fiction in color science. It is a fiction in auditory science too, and probably in every domain where we have mistaken our own calibration for a universal one.
The synesthetic brain that experiences letters as colors is not a brain that is confused about sensory categories. It is a brain whose OS connects regions that most brains hold apart, producing experiences that are consistent, involuntary, and, for the person having them, entirely real.
In each case, what looks like deviation from a standard is, on closer inspection, a different calibration of the same underlying system. The system is not broken and the standard is not correct. The frameworks we use to assess and describe these differences were built for the standard observer. Which means they are systematically better at detecting deficits than at detecting differences, and almost entirely blind to strengths that emerge from configurations they were not designed to recognise.
The Detection Error
In an earlier piece (“The Lateral Motive”), I described a pattern I had been noticing for years: in my own cognition, in certain clients, in certain colleagues. Our frameworks for understanding motivation consistently failed to register something real. People who were genuinely, powerfully motivated, but whose motivation did not move toward a goal in a linear, deficit-closing way. People who needed to understand the landscape before committing to any path. People whose dopaminergic reward came from the search itself, not from the find. The piece argued that this was not an absence of motivation. It was a detection error: the instrument was calibrated for one type of signal and could not see the other.
The neuroscience in this series gives that observation its structural explanation.
The brain manages the balance between two distinct modes: exploitation, which uses what is known to achieve goals efficiently, and exploration, which gathers information about the environment without immediate payoff. These are not different strategies chosen consciously. They are distinct neurological systems, with different activation patterns and different reward structures. The ‘seeking system’ — the dopaminergic pathway Jaak Panksepp identified as fundamental to motivated behaviour — activates during the search itself, independent of whether that search reaches a goal. Exploration has its own reward. You can be motivated by the act of exploring, entirely apart from where the exploration leads.
Classical motivation frameworks were built around exploitation-style motivation: gap-closing, goal-directed, deficit-driven. They describe a significant portion of human motivation accurately. But when they encounter exploration-style motivation, they register it as absence of motivation, because the person is not moving toward a visible goal. The framework produces a false negative. Not because the person is unmotivated, but because the instrument is not calibrated to detect that type of signal.
This detection error repeats itself at every level of the systems we have built around the standard observer.
Schools that reward sitting still and following instructions will pathologise the child who learns by moving, exploring, and questioning; not because that child cannot learn, but because their learning does not look like the instrument’s expected output. Workplaces that equate productivity with focused execution will overlook the employee whose best ideas emerge from seemingly off-task exploration; the person whose lateral connections between domains produce the insight that months of linear work did not. Research that defines intelligence as speed and accuracy on standardised tasks will miss the kinds of intelligence that do not fit the mould; the pattern recogniser, the spatial thinker, the person for whom depth in one domain comes at the expense of breadth across many.
In each case, the instrument is reading correctly by its own calibration. The problem is not the reading. It is the assumption that the instrument’s calibration describes the thing it is measuring.
What the Standard Observer Costs Us
The practical consequences of building all our frameworks around a standard observer that describes few actual people are worth naming directly.
In education: systems designed for sequential, goal-directed, phoneme-based learning will consistently underserve learners whose hardware is configured differently at the phonological level, while simultaneously failing to recognise or support the distributed-network strengths those learners frequently bring. The dyslexic child who struggles to read but thinks in three dimensions is not receiving a deficit. They are receiving a mismatch between their calibration and the system’s design.
In the workplace: performance management built around linear output, goal achievement, and consistent execution will consistently misread exploratory motivation as underperformance, local over-connectivity as inflexibility, and sensory sensitivity as distraction. The patterns that standard frameworks read as absence are frequently the presence of something different and something the framework has no category for.
In research: if the experimental designs, diagnostic instruments, and outcome measures we use to study human cognition were built for the standard observer, then the domain-specific strengths of differently-calibrated brains are systematically under-researched. We know a great deal about what neurodivergent brains struggle to do in standard-observer environments. We know considerably less about what they do better, or differently, when the environment is designed for their calibration rather than against it.
This last point is where the argument opens toward something larger. Because if the detection error is not just a practical problem (not just a matter of building better workplaces or more inclusive classrooms), but an epistemic one, embedded in the research frameworks themselves, then we are not just failing individuals. We are failing to understand something fundamental about the range of human cognitive possibility.
The Chipset Was Set Before You Arrived
One further layer of the architecture deserves its place in this argument, because it changes the moral framing as much as the scientific one.
Dick Swaab’s research on gender identity showed that a small region of the hypothalamus is differentiated before birth through prenatal hormonal influence. The felt sense of gender, one of the most intimate aspects of human identity, is at least partly encoded in brain structure before a single experience has occurred. It is a BIOS-level determination: made before the operating system loads, before culture applies, before language provides the categories to name it.
The same principle applies, with varying degrees of evidence, across much of what we think of as fundamental human variation. The structural differences between neurotypes are not, in most cases, acquired. They are not failures of development in any neutral sense. They are different outcomes of the same developmental process: different chipsets, produced by the interaction of genetic and prenatal influences that precede any environmental shaping.
This isn’t about fate. It is about design. If the chipset was set before we arrived, the question is not how do we fix it. It is how do we build systems that work with it.
Because calling one configuration the standard, and calibrating all frameworks to it, is not a neutral scientific decision. It is a choice. And like all choices, it has consequences: for every person whose calibration differs from it, and for every question about human potential that remains unasked because we have been measuring with the wrong instrument.
Expanding the Instrument
None of this requires discarding existing frameworks. What it requires is recognising what they were built to detect and what they were not.
The Kanwisher-style question “is the human brain modular or distributed?”, turned out to have a more complex answer than either camp expected: both, at different layers, for different functions, in different proportions across individuals. The same complexity applies to motivation, to perception, to learning, to cognitive strength. The question is not whether the existing frameworks are wrong. It is whether they are complete.
And they are not. And what they are missing may be more valuable than what they measure.
A spectrophotometer is an extraordinarily useful instrument. But if you want to understand color, you eventually have to look at the thing you are measuring, not just at the reading the instrument produces. The reading tells you something real. It does not tell you everything. And if the instrument was calibrated to a standard that does not describe the thing you are measuring, the gap between the reading and the reality is not a problem with the reality.
The standard observer is a useful fiction. The actual observers are the interesting ones.
The third and final article in this series, “The Questions We Haven’t Asked”, moves from the argument to its implications: asking what a research agenda that starts from cognitive difference rather than cognitive deficit might actually look like, and what questions we have been unable to ask because we have been looking through the wrong instrument.ibe human cognition may be systematically missing what they most need to see.
Comments are closed