DST and the Class of '64
The trouble with one number
Almost every attempt to rank people collapses into a single figure — net worth, follower count, a position on a list. The number is easy to sort and impossible to argue with, which is precisely the problem: it flattens a physicist, a novelist, a politician and a goalkeeper onto one line and pretends the comparison means something.
DST — the Domain and Scale Taxonomy — is an attempt to keep the dimensions apart. It does not ask how important is this person. It asks two cleaner questions and keeps the answers separate: what do they do, and how far does it reach. A local council leader and a touring stadium act can both be excellent at what they do while operating at completely different scales, and a good taxonomy should say so without ranking one above the other.
Two axes
Domain — what field someone works in
Eight main domains, each with eight sub-domains (64 cells), and every person gets exactly one of each. Each domain carries a colour and a Lucide icon so you can place someone at a glance before reading a word.
| Code | Domain | Colour | Covers |
|---|---|---|---|
| D1 | Physical & Natural Systems | Emerald | Climate, geology, ecology, space, physics, chemistry, energy |
| D2 | Biological, Cognitive & Health | Blue | Neuroscience, genetics, medicine, psychology, public health |
| D3 | Engineered, Digital & Technological | Violet | AI, software, hardware, networks, engineering, manufacturing |
| D4 | Economic & Productive Systems | Amber | Finance, entrepreneurship, retail, logistics, economics, strategy |
| D5 | Political, Legal & Institutional | Red | Governance, law, administration, military, international relations |
| D6 | Cultural, Symbolic & Belief | Teal | Religion, philosophy, ideology, history, narrative, values |
| D7 | Expressive, Aesthetic & Performative | Pink | Visual arts, music, film, literature, media, dance, theatre |
| D8 | Kinetic, Embodied & Explorative | Orange | Sport, exploration, craft, extreme environments, aviation |
The full 64 sub-domains, with their icons and colours, live in the DST spec.
Scale — how far it reaches
Scale is deliberately independent of domain. A scientist can reshape a
discipline (S4) while remaining unknown to the public; a television
personality can be known across borders (S3) without that saying anything
about the domain they occupy.
| Code | Scale | Reach |
|---|---|---|
| S1 | Field / Local | Within a specific field or locality |
| S2 | National / Regional | Across a country or region |
| S3 | International | Across national borders |
| S4 | Civilizational / Global | At the level of the whole culture |
S4 is the only scale with its own colour — amber — so civilizational figures
are visible the instant you open the table.
DST began life with a third, middle axis. It was dropped during development once it became clear the two surviving axes already did most of the work. The name reflects what’s left, not what was planned.
Reading one row
A single classification is meant to be read left to right. Take the actor and director Andy Serkis:
| Scale | Domain | Sub-domain | Person | Contribution |
|---|---|---|---|---|
| S3 — International | D7 — Expressive (pink) | D7.3 — Film & Moving Image | Andy Serkis | Gollum, Caesar, The Batman |
That row says: his influence crosses borders (S3), he works in the
expressive and performative domain (D7), and specifically in film (D7.3).
Three facts, none of them ranking him against anyone — just locating him.
Someone outside the cohort slots in the same way. David Braben — co-creator of
Elite, founder of Frontier Developments — would read D3.2 (Software
Engineering) at S3: a different domain entirely from Serkis, the same reach.
The 1964 experiment
A taxonomy with empty cells proves nothing, so we tested DST against a real group: people born in the United Kingdom in 1964. The year sits at roughly the peak of the post-war baby boom, which should mean a large enough population to find genuine examples for every domain and every scale rather than reaching for whoever springs to mind. The result is 62 people across politics, the arts, the sciences, sport and business.
What the cohort showed
The interesting part is where the hypothesis broke. Sort the explorer by domain
and the skew is immediate: nearly half the cohort (30 of 62) lands in D7 —
the arts — while D2, the entire Biological and Health domain, is empty, D3
(technology) has a single entry, and D1 and D4 have two apiece. Sort by scale
and it’s the same story at the edges: just two S4 figures, and not one S1.
That isn’t a flaw in the taxonomy so much as a finding about the source. A single birth year, filtered through who becomes publicly notable, massively over-represents performers and politicians and barely touches working scientists and engineers — who tend to be famous within a field rather than to the public. Filling every cell evenly would need either a much larger net or a deliberately stratified one. Good to know before building the next cohort.
Try it
The explorer lets you sort any column, filter by region, and search by name or contribution. Sorting by scale is the quickest way to see the shape of the data for yourself.
Update — filling the gaps
The skew above isn’t something to live with — it’s a to-do list. There’s now a small runner in the repo that checks which domains are empty or thin, goes looking for real people born in the UK in 1964 to fill them, verifies the birth year and place, and proposes classified entries for review before anything is merged. The first targets are the obvious holes: the entire Biological & Health domain, plus technology, the physical sciences, and economics. The sample database will grow and even out from here.
An open framework
The taxonomy and the dataset are released under
CC BY 4.0; the explorer code is
MIT. Everything lives in the
DST repository.
It’s v0.1, with a short list of known rough edges — two duplicated
sub-domain labels and an amber that’s currently doing double duty — documented
in the spec’s open-issues section rather than quietly ignored.
Next: more cohorts (other years, other countries), and folding the explorer into a proper Svelte island that reads the dataset as JSON instead of carrying it inline. The data already lives apart from the presentation, so that move is mostly mechanical.
Roll well.