Socrates, defending a life a jury had already judged worthless, claimed one advantage over the men who had come to condemn him: he did not mistake his own ignorance for knowledge.1 He did not apologize. Apologia is a defense, not a regret, and what he defended was the practice of asking questions he could not answer.
This work claims nothing more than that, and claims it here once, so that the admission need not be repeated on every page that follows. It does not know its subject the way the subject deserves to be known. It knows only as far as its own study has so far carried it. It is published anyway — not because the understanding is finished, but because publishing it is the method by which it is meant to be finished.
What this is
What follows is not a finished body of knowledge, released once its author was qualified to write it. It is the record of an understanding still being built, made public from its first page rather than withheld until some later point of completion: a structure to be added to, not a document waiting to be finished before anyone is permitted to see it.
There is precedent for both halves of that. At the end of his life Augustine went back through everything he had written and issued the Retractationes — a work-by-work correction of his own corpus, made in the open, by the author, on the record.2 It was not a repudiation. It was a revision, published as one. Montaigne revised differently and to the same effect, layering later thought over earlier across three editions without deleting what he had superseded, so that a modern text prints the strata and the reader watches the mind change; he wrote that he did not portray being, but passing.3 A work that grows in public, and that carries the marks of its own revision, is not an invention of the internet. It is an old form, briefly out of fashion.
The subjects taken up here range widely — law, finance, construction, architecture, governance, history — while orbiting a single question about public and private development. The range is not eclecticism. Vitruvius argued that the architect must know letters, drawing, geometry, history, philosophy, music, medicine, law, and astronomy, and justified the demand not as cultivation but as necessity: the work touches all of them, so the one who does the work must reach into all of them.4 Development is the same kind of subject. A parcel does not become a building without passing through the hands of a surveyor, a lender, a zoning board, a court, and a general contractor, and an account of that passage which understands only one of them will misdescribe the rest.
The study draws on what is actually available to it: the open internet, a personal and professional network, and a local library. No claim is made that this is the whole of what is knowable on any subject taken up here — only that it is what has been found, checked, and understood so far, and that more accurate or more complete sources may well exist beyond the reach, or beyond the knowledge, of this work.
In the course of researching, writing, and building this site, tools including AI systems, spell-checking software, digital archives, and translation tools may assist the work. Their use does not relax the standard stated here. Anything produced with with use of a tool is checked against a primary source to the best of this author's ability, or left out if it cannot be.
An admission of ignorance is cheap if it costs nothing. What follows is what it costs.
How sources are cited
Where a claim can be traced to a primary source — a statute, a recorded deed, a published study, a court's own words — it is cited directly and linked to that source wherever the source can be found online, so that it can be checked rather than taken on faith. Where the primary record cannot be reached, the most direct secondary account available is cited, and the fact that the primary record was not accessible is stated plainly rather than left for the reader to discover.
Both halves of that rule are inherited. Thucydides separated the speeches in his history from the deeds, reconstructed the first to what the occasion seemed to demand, checked the second against multiple witnesses, corrected for partiality and imperfect memory — and said so, in the text, so that a reader could discount him accordingly.5 Herodotus, whom it is customary to treat as the looser of the two, supplies the other half of the practice: he held himself obliged to report what he was told and not obliged to believe it.6 Citing a source is not endorsing it. Labeling what kind of source it is, is what makes the difference legible.
There is a nearer precedent, and a more exacting one, inside the subject this work studies. American property law rests on a public record precisely so that a claim to title can be examined by a stranger who has no reason to trust the person asserting it, and record notice binds the world whether or not the world troubles to look. A footnote does for an argument what the register of deeds does for a chain of title. It makes the claim examinable without the author's cooperation.
Where the record is incomplete
Where evidence is incomplete, missing, or simply unavailable, the gap is stated outright in the body of the text rather than smoothed over by confident phrasing. An argument resting on a gap it has not admitted is not rigor; it is the appearance of rigor. Where the work must proceed in spite of a gap, it says so, and says why the gap does not undermine the specific point being made.
Roman procedure made room for this formally. A iudex who found the case genuinely unclear could swear that it was not clear to him — sibi non liquere — and be released from deciding it.7 An institution that licenses “I do not know” as a permitted verdict is rarer than it sounds, and more disciplined than the alternative.
Cartography learned the same lesson later and more publicly. Older maps filled the unmapped interior of Africa with invented rivers and kingdoms, a habit Swift mocked by observing that geographers painted elephants where they had no towns to put.8 In 1749 d'Anville stripped the interior back to what was actually known and left the rest blank.9 The blank was read first as an embarrassment and then as a revolution. It is worth being precise about why: the blank space did not add information, it removed false information, and in doing so it made every marked feature on the map worth trusting. Blank space is not a failure of cartography. It is the point at which cartography became trustworthy. A research project that includes a map should not need this explained twice.
Where sources disagree
A gap in the record and a genuine disagreement about how to read it are not the same problem, and are not treated the same way. Where competent sources differ — a contested legal reading, a disputed historical account — the disagreement is presented as a disagreement, not resolved by silently choosing a side. Where this work does take a position on a live dispute, it says so plainly, and states its reasons, rather than presenting a contested reading as settled fact.
The scholastic disputatio required the objections to be stated first and at full strength before they were answered, on the theory that an argument which has not met the strongest form of its opposition has not been tested at all. The sharper precedent, though, is a rule about the record rather than about argument. The Mishnah preserves rejected opinions alongside the accepted ones, and gives its reason: so that a later court, finding the minority view persuasive, may rely on it.10 The published judicial dissent is the same institution in another jurisdiction, and it exists for the same reason. The record turns. A work that has erased the losing argument leaves nothing for it to turn back to.
How general patterns are tested
Where this work proposes a pattern — that development passes through some fixed number of independent, sovereign functions, for instance — that pattern is treated as a thesis under test, not as a settled fact. It is checked against cases chosen because they could have disproven it, not because they were convenient. Where a case instead reveals a function the thesis did not anticipate, or shows two supposedly separate functions collapsing into one, the thesis is revised to fit the case. The case is not set aside to protect the thesis.
Three sources stand behind that. Bacon diagnosed the failure first: the mind, having once adopted an opinion, draws all other things to support it, and the remedy he prescribed was a table of the cases where the phenomenon fails to appear, compiled with the same care as the cases where it does.11 Maine built a comparative method out of a small number of deliberately contrasting legal systems, chosen for their difference from one another rather than surveyed exhaustively, and induced general patterns from the contrast.12 Popper supplied the standard the other two were reaching for: a general claim earns its standing not from the instances that confirm it but from surviving genuine attempts to destroy it.13
The trades this work studies reached the same conclusion without the philosophy. A crane is not certified by lifting what it is rated to lift; it is proof-loaded beyond capacity, because the only informative test is one it could fail. A surveyor does not trust a traverse because the bearings looked right in the field. He runs it back to the point he started from and publishes the error of closure as a ratio — a measurement of his own failure, stated as a number, so that anyone can judge whether the work is good enough for the use it will be put to. Confidence in both cases is earned by measuring the failure rather than avoiding it.
What this work asks of a reader
Feedback here is not merely tolerated. It is the mechanism by which this work is meant to improve, and the aim is not to permit correction but to solicit it — particularly from readers who know a given subject, a given source, or a given piece of ground better than this work currently does. To that end, annotation tools allowing a reader to comment and argue directly against the text are intended for this site, built as skill and time allow.
That, too, restores something rather than inventing it. For roughly four centuries the standard published form of a legal or scriptural text was the text itself surrounded by the accumulated commentary of its readers, and Accursius's Glossa Ordinaria on the Corpus Iuris Civilis became authoritative enough that courts came to cite the gloss rather than the code beneath it.14 Argument in the margin was the scholarly institution. It is not a concession to one.
And what Socrates actually practiced was not conversation but elenchus: refutation by cross-examination, a claim taken apart by someone determined to locate the place where it fails. That is the reading this work asks for. Where new information or a sound argument shows a claim here to be mistaken, that claim will be revised or removed, by the same standard of evidence stated above. A correction is not a failure of this work. It is the work functioning as intended.
The question
Read this, then, as what it is: not a finished work interrupted, but a small and early body of understanding, growing in public, on the terms stated here from its first day, and expected to look considerably different — and considerably more complete — the longer it is allowed to continue.
The dialogues that ask what a thing is — what is piety, what is justice, what is courage — characteristically end without an answer. The absence is the result. What has been cleared away is the confidence that the question was already settled, which is the only thing standing between a beginner and an education.
What follows does not open with an answer. It opens with the question a beginner would actually ask, asked here honestly and for the first time.
What is development?