PhronesisTry it right now: read this number once, look away, and hold it — 7 4 9 2 6 1 8 3 5. Now say it back. Somewhere around the sixth or seventh digit you felt the floor give way. Not a wall — a cost. The numbers were slipping while you reached, and holding the rest took visible effort. That effort is the whole module.
Working memory is not a storage box. It is an active holding — a state your brain has to keep paying for, moment to moment, or it collapses. Stop rehearsing the digits and they are gone in seconds. That decay is the tell: you weren't storing the number, you were holding it off the ground, and the ground is where unrehearsed information falls.
This is displacement made visible. A held item is a bit of you kept away from rest. The instant you stop spending, it returns to ground — which here means: it leaves. Everything in this module is about the price of that holding and why minds work so hard to lower it.
The architecture of holding
Baddeley and Hitch (1974) cracked working memory into parts, and the split has held up for fifty years. A phonological loop holds sound and words — it's the inner voice rehearsing the digits, which is why the number felt spoken. A visuospatial sketchpad holds images and locations — where your keys are, the shape of a room. And a central executive steers attention between them, deciding what gets held and what gets dropped. Decades later, Baddeley (2000) added an episodic buffer that binds these streams into a single scene. The point for us: holding has channels, and each channel has a separate, payable cost.
How small is the room
Miller's famous "magical number seven, plus or minus two" (1956) was the first measure of the room's size. But the better modern estimate is Cowan's ~4 (2001) — roughly four independent chunks when you strip away rehearsal tricks. Four. That is the working surface of conscious thought. Everything you reason with right now passes through a window about four items wide.
So how do experts hold a chess position or a phone number or a line of code? Chunking. A chunk is a single held unit that has been compressed — "1492" is four digits to a stranger and one chunk to you. Compression is the central move: it lets one slot in the four-wide window carry far more meaning. Chunking doesn't enlarge the room. It shrinks what you're holding, so the same holding costs less.
Why the held tension is expensive
Hold something in mind and try to do anything else — the second task suffers. This is cognitive load, and Sweller's cognitive load theory (1988) names its source: working memory is a narrow channel, and load is what accrues while you keep things in it unresolved. Beyond a few chunks, accuracy falls off a cliff. The load isn't a metaphor; it is metabolic and attentional, and it climbs the longer the tension stays unresolved.
Which is exactly why you reach for paper. You write the number down, count on your fingers, leave the browser tab open, talk through a problem out loud. Every one of these is offloading — moving the held tension out of your skull and into the world, where it sits at no ongoing cost to you. We externalize because holding is expensive, and the cheapest held item is the one you no longer have to hold.
What you'll be able to do
- Recognize when a task has exceeded your ~4-chunk window — and chunk or offload before accuracy collapses.
- Design your own thinking environment (notes, lists, diagrams) so the held tension lives on the page, not in your head.
- Tell apart load that comes from the problem itself versus load you've imposed by holding too much at once.
The precise version
The rigorous layer. Optional — the plain version above already carries the idea.
Let the cognitive ground state $S^0_{cog}$ be the configuration of clearest understanding, least effort. Working memory holds an item by sustaining a displacement $\xi_{cog} > 0$: the item is kept off the ground, and the gap between held-state and rest is the displacement. Release the spend and the system relaxes toward $S^0_{cog}$ — the unrehearsed item decays. Decay is the return path running on its own.
The instantaneous cost $D_{cog}(\xi) \ge 0$ is the load of holding: roughly the number of independent chunks held times the effort per chunk. Capacity is a ceiling on $\xi_{cog}$ — past about four chunks, $D_{cog}$ rises steeply and items begin to drop, because the system cannot fund a displacement that large. Crucially $D_{cog}$ does not vanish even at rest: living ground is not free, $D_{cog}(0) = \theta > 0$, the idling brain still burning its baseline metabolic cost.
Over an episode of thought the accumulated price is $\Phi_{cog} = \int D_{cog}(\xi_{cog})\,dt$ — load integrated over how long you hold. This exposes why duration matters: a small displacement held long can cost as much as a large one held briefly. Chunking lowers $\xi_{cog}$ for the same information content (compression reduces the displacement needed to hold a fixed meaning). Offloading sets $\xi_{cog} \to 0$ for that item by relocating it outside the system — you stop integrating cost entirely, paying a one-time write instead of a continuous hold. Both are strategies on the same quantity: minimize $\Phi_{cog}$ for the understanding you need.
Read this way, the laserbrain inversion fits: the displacement is the holding, and the return — decay, forgetting, putting it on paper — is the release. You don't pay to forget. You pay to keep remembering.
Worked example
Multiply 27 by 14 in your head. Watch what fails: you compute 27 × 10 = 270, then 27 × 4 = 108 — and now you must hold 270 ($\xi_{cog}$ active, $D_{cog}$ accruing the whole time) while computing the second product, then add. Most people lose the 270. The information was fine; the holding was what broke, because two partial products plus an in-progress multiplication exceeds the four-wide window. Now do it on paper: each partial product is written the instant it's formed, $\xi_{cog} \to 0$ for it, and the answer arrives easily. Same arithmetic, radically lower $\Phi_{cog}$ — because you stopped holding and started offloading.
Exercises
- Read a 10-digit number once, then recall it. Now re-group it as a phone number (3-3-4) and try again. You've turned ten chunks into three. Notice the holding got cheaper without the information changing — that's compression lowering $\xi_{cog}$.
- Pick a routine task you do entirely in your head (a recipe, a route, a budget calculation). Externalize it onto paper or a list, then do it again. Time both and rate the effort of each. The gap is the cost of holding you'd been silently paying.
- (Open-ended.) Find one place in your daily life where you habitually hold tension in working memory that you could offload instead — open tabs, a sticky note, a spoken reminder. What is the true cost of keeping it in your head, and what would change in your thinking if you let the world hold it for you?
Sources
- Rincón, D., alice, & clöe (2026). Cognitive Displacement: A Planck Scale for Human Understanding.
- The Displacement Framework.
- Baddeley, A. D., & Hitch, G. (1974). Working memory. In The Psychology of Learning and Motivation (Vol. 8). The phonological loop, visuospatial sketchpad, and central executive.
- Baddeley, A. (2000). The episodic buffer: a new component of working memory? Trends in Cognitive Sciences, 4(11).
- Miller, G. A. (1956). The magical number seven, plus or minus two. Psychological Review, 63(2).
- Cowan, N. (2001). The magical number 4 in short-term memory: a reconsideration of mental storage capacity. Behavioral and Brain Sciences, 24(1).
- Sweller, J. (1988). Cognitive load during problem solving: effects on learning. Cognitive Science, 12(2).