the second hundred · metaphor 160
Where does your attention actually go? Not evenly. A handful of things — a person, a debt, a fear, one recurring name — soak up most of it, and the vast rest of your life gets the crumbs that are left. The distribution is lopsided by nature, not by accident.
Count the day honestly and it rarely divides fairly. One worry gets called to mind again and again; the next two get a good share; a dozen more flicker past; and hundreds of small concerns — people you love, tasks that matter, whole rooms of your life — sit almost untouched. It is not that the neglected things are unimportant. It is that attention, like a budget, is spent — and a spent thing gets spread thin, with most of it landing on a few line items.
There is a shape to this lopsidedness, and it repeats: in the words people use, the cities they build, the customers a shop serves, the songs a country plays. Rank everything by how much use it gets, and the second gets about half the first, the third about a third, the tenth about a tenth. A steep, self-similar staircase where the top few steps carry almost the whole climb. It has a name — Zipf's law — and once you see its curve you cannot unsee it in your own head.
The item names are illustrative labels pinned to rank (rank 1 = the thing you attend to most). Every number — shares, the log–log dots, "to cover half" — is counted live from the actual sampled events, never set by hand.
The shape itself
Take anything you can rank by usage — words in a book, worries in a week, the people you text — and line them up from most-used to least. Zipf's observation is that the amount of use falls in near-lockstep with the rank: frequency times rank stays roughly constant. The most-used item is used about twice as often as the second, three times the third, ten times the tenth. Plot it and you get a smooth hyperbola that plunges from a towering head into an endless, flat tail.
The clean way to see it is to take logarithms of both. On a log–log plot — where each step out multiplies the rank tenfold and each step down divides the use tenfold — that plunging hyperbola becomes a straight line. Its steepness is a single number, the exponent s. At s ≈ 1 you get the textbook 1/rank; push s higher and the line tips steeper, the head grows more tyrannical, and the tail flattens toward nothing. That one slider is the whole personality of the distribution.
Why should attention obey it? Because the things that get used get reinforced. What you thought about yesterday is easier to think about today; the worry you rehearsed is the worry that returns. Use begets use — the same rich-get-richer feedback that builds power laws everywhere — and its natural resting shape is Zipf's staircase.
What to try
Start at the classic s = 1 and let it pour. The dots march down a straight line, and the readouts stabilise: the single most-attended item claims a fat slice on its own, the top three together take most of what's left, and it takes only a small handful of items to cover half of everything. That last number — to cover half — is the honest measure of how concentrated a life is. When it reads "3," three concerns are eating half your days.
Now drag s toward 1.6. The line tips, the top bar swells until rank 1 alone owns a third of the total, and "to cover half" drops to one or two. This is obsession's geometry — nearly all the attention on a single item, the rest of the world reduced to a flat tail of near-zeros. Drag the other way, toward 0.6, and the staircase softens toward fairness: no item dominates, "to cover half" climbs, the tail lifts off the floor. The presets — a scattered week, a settled life, a single fixation — set the scene, but the slider is where you feel it.
The mapping
The metaphor is exact where it counts. Your worries are ranked, not listed. You do not carry fifty concerns equally; you carry one or two that return unbidden, a few that surface when prompted, and a long tail you would have to be reminded even exist. The people, too: a few names take most of your emotional bandwidth — beloved or dreaded — while dozens you would call friends live in the tail, thought of rarely and fondly. And the words: most of what you say is a tiny core vocabulary said again and again, the rare and precise words kept for occasions.
Seeing the slope is useful because it reframes the guilt. You are not failing to attend to everything — no one can; the distribution forbids it. The real question is not whether your attention is concentrated but onto what. A steep slope aimed at what you love is devotion; the same slope aimed at a fear is rumination. The shape is fixed. The one degree of freedom you have is which items sit at rank one.
Read as life lessons
You cannot spread attention evenly any more than a language can use every word equally. Something will sit at rank one. The only choice is whether you placed it there on purpose.
The rarely-visited concerns aren't clutter to delete — a language needs its rare words for the moments only they fit. A rich life keeps a long tail; a flat tail of near-zeros is not the same as an empty one.
Grief, deadlines, new love all steepen the slope until one item swallows the days. Not wrong — but worth noticing, because a slope near-vertical means the rest of your life has gone quiet.
The mapping, exactly
| Mathematics | Life |
|---|---|
| rank of an item | Where a concern sits in your private order of what gets thought about — first, third, or forgotten. |
| frequency ∝ 1/rankˢ | How much of your attention it actually receives — falling off a cliff after the first few. |
| the exponent s | How single-minded you are right now: gentle and spread, or steep and consumed by one thing. |
| the head | The one or two people, fears, or tasks that own most of your inner life. |
| the long tail | The hundreds of things you genuinely care about but rarely reach — real, but starved. |
| rich-get-richer feedback | Rehearsal: the thought you had yesterday is easier to have today, so use compounds into dominance. |
Where the metaphor tears
The law ranks by how often something is used, and attention is a poor proxy for what matters. The child you think of constantly and the will you never revisit sit at opposite ends of the curve, yet the will may be the weightier thing. A worry can dominate rank one precisely because it is unresolved, not because it deserves the room. Read the staircase as a map of where your mind goes — never as a verdict on where it should.
Zipf's law is a robust regularity, but it is not a law of physics compelling you. People redirect the head all the time — therapy, distraction, a change of scene, a deliberate practice of gratitude all reshuffle which item sits on top. The curve keeps its shape; you can still change what occupies its peak. Mistaking "this is the usual shape" for "this cannot be moved" is the metaphor's most seductive lie.
Plenty of quantities are not — heights, blood pressure, the time you leave for work cluster around an average with no towering head. When a distribution really is bell-shaped, forcing a rank-frequency story onto it invents a hierarchy that isn't there. The instrument draws the straight line even for gentle exponents; the discipline is to check whether your data actually falls on it before you trust the tale.