Glossary

It sits in the temporal lobe, above the ear, inside a fold. The path to it is long: sound becomes vibration in the eardrum, becomes movement of fluid in the cochlea, becomes an electrical impulse, climbs through the brainstem and only then reaches the cortex.

A1 is organised by frequency, like a keyboard: low notes on one side, high notes on the other, in order. That organisation is anatomical and was there before you learned anything about sound.

The detail that interests whoever designs is speed: the auditory path has fewer stops than the visual one, and so sound reaches the brain before the corresponding image. It is the reason a sound alert is more effective at interrupting than a visual one, and more aggressive for the same reason.

Two almond-sized structures, one on each side, at the inner tip of the temporal lobe. That is where the name comes from.

It is the brain’s relevance detector, and it is fast: it receives a coarse version of the sensory signal through a shortcut that bypasses the cortex, and fires before the person knows what they saw. Fear is the most studied response, but it responds to anything that matters, including good things.

In an interface, it is what reacts to the error in red, to the deletion warning, to the countdown. The reaction arrives before the reading, and no later explanation undoes it, only stops it from happening again. That is why the tone of an error message matters more than its content.

A set of nuclei deep in the brain, below the cortex.

It is where repeated behaviour becomes automatic. While a sequence is being learned, it consumes prefrontal cortex; once repeated enough, it migrates here and starts to run at almost no cost, and almost without awareness.

It is the structure behind every habit of use: the path the person takes through your product without looking, the button they hit from muscle memory. It is also why moving an element’s position hurts more than it seems: the automatism does not update on notice, only through new repetition.

The stalk that links the brain to the spinal cord, at the base of the skull. It is the oldest part and the one that cannot be lost: it controls breathing, heartbeat and level of consciousness.

Every auditory signal passes through here before climbing, and it is here that the difference in the arrival of sound between one ear and the other — a matter of microseconds — becomes the perception that something came from the left or from the right.

It is also the origin of the systems that regulate alertness and sleep, which makes it the point where tiredness and attention meet physically. No interface talks to it directly, but every interface is used by someone whose brainstem has already decided how much wakefulness is left for the day.

The wrinkled structure at the back and bottom of the skull. It takes up about ten per cent of the brain’s volume and contains more than half of all its neurons.

It handles the timing and precision of movement, and, according to the more recent literature, also the timing and precision of operations that are not motor. It anticipates: it predicts where the hand will land and corrects before the error happens.

It is what makes a keyboard shortcut faster than a menu, and what makes a well-calibrated drag feel natural. It is also what suffers when the interface’s response is irregular: a delay that varies prevents prediction, and the gesture stops becoming automatic.

A hormone, not a neurotransmitter: it is released by the adrenal glands and circulates in the blood, which makes it slower and longer-lasting than a neural signal.

It is the hormone of the prolonged stress response. In a short peak, it improves attention and memory. Sustained, it does the opposite: it harms the hippocampus, degrades the formation of new memory and narrows attention to what looks like a threat.

An interface that produces constant pressure — a flashing deadline, a balance in red, a warning that will not close — is not keeping the person alert. It is making them worse at what they came to do, and charging the difference to their memory.

A neurotransmitter, that is, a molecule one neuron releases to talk to another. It is not a structure, it is a chemical signal.

The popular association with pleasure is wrong and worth undoing. Dopamine encodes reward prediction error: it rises when something is better than expected, falls when it is worse, and stays neutral when it is exactly as predicted. It moves you towards, more than it makes you like.

That is why the anticipation of a notification is stronger than its content, and why “dopamine” has become a loose word in product conversation. When someone says a feature “releases dopamine”, the useful question is: compared to what expectation?

The upper, outer part of the prefrontal cortex, on the side of the forehead.

It is the region most associated with holding information active while doing something else: keeping the goal in view, resisting distraction, comparing what is on the screen with what was on the previous screen. It works together with the frontoparietal circuits, and it is one of the first to fail when the task tightens.

It is the structure behind the practical limit this site keeps talking about: whatever the interface asks the person to remember between one screen and the next is charged here, and the balance is low.

The auditory equivalent of iconic memory, and more generous: the echo of what was said lasts a few seconds, not fractions.

It is the reason you can answer “what?” and, halfway through the question, have already understood, the sentence was still available for internal re-reading.

The difference in duration between the two channels has a practical consequence: spoken instruction has a wider window for recovery than instruction shown and withdrawn. An alert that appears and vanishes depends on the person looking; a sound does not.

It is not a structure, it is a conversation: the frontal and parietal regions working in sync, linked by long bundles of fibres.

It is this network that supports working memory and directed attention. It keeps active what matters and suppresses what does not, and it does so at a high metabolic cost, which is why the feeling of mental effort is literal, and not a figure of speech.

Every time a screen forces the person to keep something in their head — a code, a comparison, an instruction that vanished — it is this network that is paying, and it does not scale.

A fold on the underside of the temporal lobe, on both sides.

It contains the most face-specialised region in the whole cortex, and “specialised” here is a strong word: lesions in this area produce prosopagnosia, the inability to recognise familiar faces with vision preserved in everything else. The same region takes part in recognising written words, which suggests that reading is, in part, recognising the shape of the word the way one recognises a face.

That is why faces in an interface capture the gaze before any other element, and why familiar words are read as a block while unknown ones are spelled out. An unfamiliar typeface and all caps undo that shape and charge the difference.

A curved structure on each side of the brain, on the inner part of the temporal lobe. The name comes from the shape, which resembles a seahorse.

It is not where memories are kept: it is what decides what deserves to be kept and makes the transfer. New memory passes through here; old memory already lives spread across the cortex. The best-known clinical case in neuroscience, the patient H.M., lost the hippocampus on both sides in surgery and went on to live in a continuous present: he kept everything he had learned before and formed no new declarative memory.

For whoever designs, the consequence is harsh: nothing your interface teaches is retained for free. What is not repeated, contextualised or unnecessary to remember simply does not survive the session, and a good part of what it teaches today is already gone tomorrow.

The visual trace that survives for a fraction of a second after the image is gone. It is why you can mentally “re-read” what just flashed on the screen.

It lasts around a few hundred milliseconds, and holds much more than the person can report. The classic experiments show exactly that: someone who sees a matrix of letters for an instant can name any row asked for right afterwards, but not all of them, the information was there and faded during the report.

It is the reason very fast animations are perceived without being understood: the signal existed, but did not last long enough to become something else.

A set of structures in the core of the brain — amygdala, hippocampus, hypothalamus and neighbours — historically grouped under the name “emotional brain”.

The term is older than the evidence that supports it, and today it is used with a caveat: the structures exist and talk to one another, but they do not form a system separate from the rest, and emotion does not live in one region. Worth knowing before treating “limbic” as a synonym for irrational.

What survives of the idea is useful all the same: there is a fast circuit, tied to memory and to value, that responds before the slow, deliberate circuit. Designing while ignoring the first is designing for a person who does not exist.

It is the memory that survives the session. It includes what the person can declare, facts and episodes, and what they know how to do without declaring, like the path to a menu they have used a hundred times.

Its formation passes through the hippocampus and the medial temporal lobe, and the long-term archive is the neocortex. The dependence on the hippocampus decreases over time, as Squire and Zola-Morgan described in 1991.

For whoever designs, it is the side of memory that the interface does not control in the moment. What it teaches today only stays if it is repeated, contextualised or recognisable tomorrow.

The part of the prefrontal cortex that sits on the inner wall, between the two hemispheres.

It activates when the person thinks about themselves and when they think about other people, and that overlap is no accident of measurement: thinking about what the other person wants uses, in large part, the same machine as thinking about what I want. It is one of the pieces of the network that supports reading other people’s intentions.

It is the substrate of everything in design that depends on “what will they think of me”: social proof, counters of people looking, ratings. It works because the question about the other and the question about oneself run in the same place.

The inner part of the temporal lobe, which includes the hippocampus and the cortex around it.

It is the set that supports declarative memory, what can be put into words. It works in functional opposition to the basal ganglia, which support what one knows how to do without being able to explain.

The distinction matters in an interface: teaching by text and teaching by repetition record in different systems. The tutorial the person read lives in one; the gesture they repeated a hundred times lives in another, and it is the second that survives a change of layout, which explains why redesigns break precisely the oldest users.

A set of regions — medial prefrontal, temporoparietal junction and neighbours — that activate when the person thinks about what another person is thinking.

It is the theory of mind system: attributing intention, predicting reaction, imagining what the other knows and does not know. It works on its own, without conscious effort, and it is one of the brain’s default modes at rest.

In design, it is what makes the person attribute intention to your product. A screen that hides the cancel button is not read as badly designed: it is read as wanting something. The network does not distinguish interface from person, and the suspicion, once formed, contaminates the rest.

Neurons that fire both when the individual performs an action and when they observe another performing the same action. Discovered in monkeys, in the 1990s, by accident.

The existence of the phenomenon is solid; the extent of what it explains is not. The popular reading — that they would be the basis of empathy and culture — went far beyond what the data support, and the field went through a strong correction. Worth citing with care.

What remains applicable is modest and real: seeing someone do partly activates the system of the one who does. It is one of the reasons a video demonstration teaches better than a written description, and an interface that shows the gesture is learned faster than one that describes it.

A small structure at the base of the brain, at the meeting point between the emotional circuit and the motor one.

It is the centre of the reward system, and the finding that changed everything is that it responds more to expectation than to the gain. The response is at its peak when the reward is possible and uncertain; when it is guaranteed, the response drops.

It is the neural basis of the infinite feed, the loot box and “pull to refresh”. Uncertainty is not a side effect of these patterns: it is the mechanism. Knowing that is what separates using it to give a task rhythm from using it to hold someone captive.

It sits just above the eye sockets, at the base of the frontal lobe.

It is where value is assigned. Not the price, the subjective value, the one that changes with context, expectation and what came before. The same offer is worth more after a high anchor and less after a low one, and this is where that recalibration happens.

Anchoring, loss aversion and the endowment effect depend on this region. It is also why these patterns are so effective and so easy to turn into manipulation: they do not convince anyone of anything, they just change the ruler the person measures with.

While the semicircular canals handle rotation, the otoliths handle straight-line displacement and gravity. They are two pouches, the utricle and the saccule, with a gelatinous membrane covered in calcium carbonate crystals.

The crystals are denser than the rest and, because of that, lag behind when the body accelerates and slide when the head tilts. Their displacement bends the cilia underneath. That is how you know the lift has started going up with your eyes closed, and that is how you know which way the floor is.

They do not distinguish acceleration from tilt, because both produce the same force on the membrane. What resolves the ambiguity is vision, crossing the two pieces of information, and it is in that stitching that motion sickness lives.

Neurons in the hippocampus and neighbouring regions that fire when the animal is at a specific point in space. Grid cells do the same in a hexagonal pattern, like internal graph paper. The discovery earned the 2014 Nobel Prize.

What came afterwards is what matters here: the same machinery seems to organise not only physical space but abstract relations, the proximity between ideas, the position of an item in a sequence, the path travelled through a subject.

It is the most likely neural basis of the metaphor design uses all the time without thinking: “where am I”, “back”, “go deeper”. The person is not imagining a place out of poetic licence; they are using the system that evolved for places.

It is the region behind the forehead, and it is the last part of the brain to finish maturing, a process that runs until around the age of twenty-five. It is also the first to lose efficiency under tiredness, hunger, stress or alcohol.

It is where the person holds a goal while doing something else, compares alternatives, inhibits the immediate impulse and picks up where they left off after an interruption. None of this is automatic: all of it costs, and the budget is small and runs out.

Most of what this site calls “load” happens here. Every extra field in the form, every decision the interface hands over to whoever uses it, every instruction that has to stay in the head instead of on the screen: all of it is charged to this region, which arrives at your product already partly spent by the rest of the day.

It is a strip that crosses the top of the head from ear to ear, just behind the strip that commands movement. Each piece of it corresponds to a piece of the body, in order.

The order is preserved, but the proportion is not: the map is drawn by number of sensors, not by the size of the part. Lips, tongue and fingertips take up enormous areas; the back and the thighs take up almost nothing. It is this distorted map that the figure alongside draws.

That is why touch interaction works: the fingertip is one of the regions of highest tactile resolution in the body. And that is why tactile feedback — the short vibration when confirming an action — is perceived even when the eye is elsewhere.

They are three ring-shaped tubes, filled with fluid, mounted at nearly perpendicular angles to one another: one for each axis of head rotation. Inside each one there is a gelatinous gate with hair cells.

When you turn your head, the bone turns with it and the fluid lags behind for an instant, by inertia. That drag pushes the gate, bends the cilia and changes the firing rate of the fibre. That is why they detect rotational acceleration, and not constant rotation: in a chair spinning at a fixed speed, the fluid catches up with the bone and the sensation of spinning disappears.

For whoever designs, what matters is what they do not do: they do not know you are looking at a screen. They report what the skull did, and nothing more.

A small structure in the brainstem, very old in evolutionary terms, earlier than the visual cortex.

It receives a copy of the visual signal through a shortcut that bypasses the cortex, and commands the reflex movement of the eyes towards what moved or appeared suddenly. It is the reason you are already looking at the thing before you know what it is.

It is the structure a flashing banner hijacks. And it is the reason that kind of device works even on those who hate it: the shift of gaze does not pass through your opinion of it.

It sits at the geometric centre of the brain, roughly where you would point if you had to point to the middle of your own head. It is a paired structure: there is one on each side.

Almost everything the senses capture passes through here before reaching the cortex, the known exception being smell, which comes in through another door. The thalamus is not a pass-through cable: it regulates what passes and with what intensity, and it is one of the pieces that decide what gains access to consciousness when there is more going on than fits.

It is the structure that explains why attention is subtraction, not addition. There is no such thing as “paying more attention to everything”: what comes in extra on one channel comes in less on another.

It sits at the back of the head, at the rear pole of the brain, as far from the eyes as possible, which already says something: the signal crosses the entire head before it becomes an image. It leaves the eye through the optic nerve, passes through the thalamus and arrives here.

What V1 does is not seeing: it is decomposing. Each group of neurons responds to something very specific: an edge tilted at 30 degrees, a light–dark contrast, a movement to the right. None of them knows what the object is. The image you think you are seeing does not yet exist at this stage.

For whoever designs, what matters is what V1 delivers for free and what it does not: contrast, edge and orientation are processed before any decision of yours about where to look. That is why an element with a strong edge is found without effort, and why low-contrast text costs dearly even when it is legible.

It sits right around V1, like a ring. It receives the loose edges from it and does the first assembly.

This is where an edge becomes the outline of something. V2 decides what is figure and what is ground, closes shapes that are partly hidden and groups elements that belong together. The Gestalt laws — proximity, similarity, closure — describe the behaviour of this stage, and that is why they seem obvious: you do not choose to follow them, the brain has already grouped before you looked.

The practical consequence is uncomfortable: the grouping you drew by accident is as real as the one you drew on purpose. Two elements close together read as a group, and no caption undoes that.

The region that contains the nucleus accumbens and the tissue around it. The two names appear almost as synonyms in the literature, and the difference is one of anatomical delimitation.

It is where the dopamine signal arrives when something is worth more than expected. What it encodes is not pleasure, it is surprise: the difference between what happened and what was predicted.

In product, this explains why the same reward delivered the same way every time stops working. It is not that the person got tired: it is that they started to predict, and a prediction met in full is a signal at zero.

The lower, inner part of the prefrontal cortex, just above the eye sockets.

It is where the emotional evaluation of an option meets the rational evaluation of it. Classic clinical cases show people with lesions in this region keeping their logical reasoning intact and still deciding badly, because they lost the affective signal that marks an alternative as good or bad before the sums are done.

The lesson for whoever designs is that “rational decision” is a useful fiction and nothing more. The person does not decide in spite of feeling; they decide with what they feel, and the feeling arrives first.

It is not a place, it is a state: the set of things that are active in your head right now and that vanish if you stop sustaining them. It depends mainly on the prefrontal cortex and the frontoparietal circuits.

The capacity is small and well studied. The contemporary estimate, more conservative than Miller’s famous seven, sits at around four simultaneous elements, and drops when there is noise, hurry or a second task. It is not trainable to the point of changing level: what changes is how much fits inside each element, through practice.

It is the most operational number on this entire site. Every field the person has to retain, every code copied from another screen, every comparison between options that are not visible at the same time takes up one of those four places. A good interface does not require the person to hold on to things: it leaves them in view.