Empiricalism is the study and collection of recurring patterns in human behaviour and institutions, judged by the strength of the evidence supporting them.
People notice things. Deadlines change behaviour, measurements alter what they measure, bureaucracies grow, crowds distort judgement, experts overreach. Some of these observations survive long enough to earn a name: Parkinson’s Law, Goodhart’s Law, the Peter Principle, Chesterton’s Fence. The rest live on as proverbs, professional maxims and things experienced people simply know, scattered across books, trades, disciplines and centuries.
Not every memorable saying is true. Not every famous law is a law. Not every psychological effect survives replication. Yet an observation should not be thrown out for arriving as an aphorism rather than a hypothesis, or for belonging to no department.
The task of empiricalism is to gather these claims and ask the question that matters most: what does the evidence actually show?
IEmpiricalism is not empiricism
Empiricism is a theory of knowledge: it asks how experience can give us knowledge at all. Empiricalism is a programme of collection and evaluation: it asks what lessons experience has actually produced, and how well they stand up to examination.*
Take Parkinson’s Law, that work expands to fill the time available. It began as a joke in The Economist in 1955. The empiricalist neither admires it for its wit nor dismisses it for its origin, but asks what happened when someone tested it. Bryan and Locke did, in 1967: people given twice the time a task needed did work more slowly, not because work expands but because a generous deadline led them to set easier goals. The law survives, narrowed and explained. Confirmation, qualification, rejection and uncertainty are all useful results.
IIThe unit: the empiriceme
A programme of collection needs a name for what it collects, and every available word smuggles in a verdict. Law overclaims for most adages. Finding implies evidence many of these lack. Meme imports a theory of transmission. A theorem is proved, a regularity measured, a model argued for, a maxim urged and an adage merely apt, and no ordinary noun covers all five.
An empiriceme is the smallest unit of noticing that travels: a claim about how people and institutions tend to behave, short enough to quote whole and usually carrying a name. Its truth is a separate question from its existence.
The word is a filing term, not a warrant. Some of the most widely travelled empiricemes in the ledger are among the least supported, which is exactly why they are kept. What an empiriceme is, and whether it is true, are two questions, asked in that order. The full definition explains the word.
IIIA memorable saying is not evidence
Short sentences are easy to remember, easy to repeat and easy to mistake for truth. A maxim becomes a quotation, the quotation gains an attribution, the attribution lends prestige, and nobody checks.
The learning pyramid still circulates in teacher training: we remember a tenth of what we read and nine-tenths of what we teach. Its figures have no traceable source, and the institute that spread them could not find the research. “When the student is ready, the teacher will appear” is sold as Buddhist wisdom and traces to a Theosophical tract of the 1880s. A famous quotation begins an inquiry; it does not end one.
A claim may be well established, probable, plausible, conditional, contested, anecdotal, or false despite its popularity. A weaker claim accurately labelled is more valuable than a stronger claim falsely presented as certainty.
IVExperience is data, but noisy data
Farmers, soldiers, teachers, doctors, engineers and civil servants all notice patterns, and much of what they notice is real. But we remember the dramatic, infer cause from coincidence, generalise from small samples, and prefer the observation that is witty or cynical to the one that is dull and true.
The bystander effect was first studied because of the 1964 killing of Kitty Genovese, and the newspaper story of 38 witnesses who did nothing turned out to be substantially wrong. The effect itself holds on average across many studies, moderated by danger and by who else is present. A false anecdote launched a real finding. Folk wisdom is evidence of what people believed, controlled research is stronger, and replication is stronger still; the task is to weigh each for what it is.
VRegularities, not laws
Newton’s laws describe physical relationships with extraordinary precision. Murphy’s Law does not, and neither does Goodhart’s. Human affairs are probabilistic: people adapt, institutions learn, and the same cause produces different effects in different places. Law, effect, principle and razor are historical labels here, not guarantees of scientific status.
An observation need not hold everywhere to be worth knowing, any more than a weather forecast must be certain to be useful. The empiricalist asks not whether a claim is always true, but how often, under what conditions, and at what cost.
VIContradictions are data
“Look before you leap.” “He who hesitates is lost.” Collections of wisdom hide their contradictions; empiricalism keeps them, because they mark boundaries. The ledger files the worked example effect, which says novices learn best by being shown, beside productive failure, which says struggling first builds deeper understanding. It files the wisdom of crowds beside group polarisation. Each pair is a question about when, and the answer is usually that both are true under different conditions.
VIIEvidence changes standing
An empiricalist catalogue must never become scripture. For decades the marshmallow test stood for self-control as destiny; a 2018 replication found the link half the original size, and a third of that once family background was taken into account. Recordings from the Stanford Prison Experiment show the experimenters coaching guards towards harshness. Reanalysis of the original Hawthorne data found little of the Hawthorne effect. Each entry says so.
Revision is not an embarrassment; it is evidence that the system works. A database that never changes in response to new evidence is not empirical. It is merely archival.
Provenance belongs to the same discipline. Chesterton’s fence really is Chesterton’s. “As simple as possible, but not simpler” is a later paraphrase hung on Einstein. The famous cobra bounty in colonial India is poorly documented, while the 1902 rat-tail bounty in Hanoi, which bred rat farms, is not. Knowing who said something, and when, separates authority from evidence.
VIIIAcross disciplines, against forgetting
Regularities do not respect university departments. Goodhart described monetary targets collapsing in 1975. Campbell described social indicators corrupting in 1976. Hanoi’s colonial government had already paid for rat tails and got rat farming. Three fields, one pattern. The organising question is what happens? rather than which department owns the explanation?
Many of the most valuable entries are warnings: how incentives corrupt, statistics mislead, groups suppress dissent, and survivors hide the failures behind them. This is negative knowledge, and a culture that forgets how it has been fooled will be fooled the same way again. The small observations earn their place beside the grand theories. Brandolini’s Law, that refuting nonsense takes far more effort than producing it, is one sentence long and explains a good deal of public argument.
IXFive principles
1. Observation before theory.
Begin with what appears repeatedly in experience. Explanations may follow.
2. A memorable saying is not evidence.
Elegance, antiquity, popularity and a famous name establish nothing.
3. Repeated experience deserves to be recorded.
Informal observation can hold real knowledge, provided its limits are stated.
4. Contradictory evidence is preserved, not hidden.
Failures, exceptions, replication problems and boundary conditions belong in the entry.
5. The strength of a claim should never exceed the strength of its evidence.
Certainty should be earned, and the empiricalist must be willing to say we do not know.
The catalogue is unfinished by design. It is not a monument to settled wisdom but an instrument for continuous inquiry: an attempt to determine, as carefully as possible, what experience has actually taught us about ourselves.
Every empiriceme in the ledger this describes is dated, attributed, graded for what kind of claim it is, and given a worked case and two examples of it in practice. Open the ledger, or read how to use it.