²²Why One Building

This site studies one ordinary subject in depth rather than many subjects briefly. It takes one corner and one building, and as many of its occupants, firms, permits and pictures as can be found, and follows each back to a source before keeping it or marking it. Narrowing the subject that far means you can afford to be exhaustive about it. A written working method, applied to every question, keeps that exhaustiveness honest.

The wager is an old one in history writing: the general lives in the particular. Followed all the way down, a single ordinary subject can show how a stretch of a city actually worked, in detail that a survey touching many subjects lightly flattens away. Read in full, this corner is no smaller a history than one of the furniture trade or of the neighbourhood around it. It is a different kind, where every claim can be tied to a document and the texture of an ordinary place survives the telling.

The method runs as a small cycle, repeated for every question the research opens. You state the question and how far you mean to take it, and check what the research already holds, contradictions and open questions included. You search the record and log what came back: the useful hits, the empty searches, the doors that would not open. You preserve what is worth keeping and note where it came from. You read it and transcribe it, keeping the source's own words apart from your reading of them. Only then do you draft a claim, with its support and its uncertainty attached, put any change to accepted knowledge in front of the researcher, and rebuild the views. The cycle closes with the next question, or with a plain account of what was searched and what was not. It is no straight line: a new name, date or address can send it back to an earlier step.

Two of those steps carry most of the honesty. The first is knowing when to stop. The record is never exhausted, so you stop when the search has been reasonable for how much the question matters, when the most relevant record classes and the likely spelling and date variants have been checked, and when more of the same would not change the answer without new access or a new lead. Stopping reasonably is not proof that nothing happened. It ends with a scoped sentence, not found in these named indexes for these years, and never with the flat no record exists. The second is what becomes of an empty search. A search that ran cleanly and returned nothing is a result, and it is kept. A search only partly carried out stands for the part it covered and no more. A search that was blocked or failed says nothing about absence at all, and is kept as a limit. Only the clean empty count can hold up a negative conclusion, and only at its own scope.

The work differs from two things it is often mistaken for. A house-proud local history sets out to celebrate a building; here nothing is kept that cannot be shown from a source. A genealogy that happens to carry an address follows a bloodline wherever it leads and treats the house as a waypoint. Here the place is the subject, and everyone who passed through it is read against it.

A building suits this kind of work for a second reason: it holds still while everything around it moves, and the paperwork of ordinary life keeps piling up against its address. That paperwork reaches only so far. It can say who was registered at the corner and when, what trades a firm advertised for, what a business called itself in the directories, and what its telephone number was. The number 30047 is recorded at Rochussenstraat 169 under three successive firms between 1929 and 1960¹. In June 1900 the earliest labour advertisement traced for the corner asked for skilled cabinet-makers and an upholsterer², while the directories filed the works first as joiners and only later added upholstery³. The paperwork can also say who was registered at each of the address's three sub-numbers in a given year. The 1911 street list puts the manufacturer G. J. van Duffelen at 169a and his partner C. Pelt at 169c, with the firm of Van Duffelen & Pelt between them at 169b⁴. Names, dates, trades, numbers, tenancies: this is what an address leaves behind.

It leaves no trace of what anyone thought. The record cannot say why the founders built where they did, what the men at the benches felt about the wage, or what passed between tenants on a landing. Two rules keep the site inside that boundary. Where it reconstructs an intention instead of reading one off the page, it says so, and marks the claim as an inference or a probable finding. And where a search comes back empty, it prints the emptiness. No nuisance-permit notice naming number 169 has been found in the digitised press, though the same street's notices are plentiful around it, and the site records exactly that scoped absence instead of pretending the question was never asked⁵.

Beneath all of this lies a rule about the order of work: preserve first, claim second, publish last. A grand building has an institution to keep its record. An ordinary corner has none, so whatever record it has is assembled once, deliberately, or not at all. Before a statement is drafted, the bytes it will rest on are retrieved and logged, so the claim points at a kept document and not at a link that may rot. A source can be preserved with no claim attached at all, held against a question nobody has answered yet. The written record and the reading of it are kept apart on purpose.

Publishing is a further, deliberate step. Building this site locally does not publish it; putting it in front of the world is a separate act that has to be confirmed each time. The chapters and deep dives you are reading sit downstream of the underlying record, not beside it, so a correction made there updates every place it is cited, instead of being copied across formats where one copy could quietly stay wrong. That order also lets the work outlast the website. The preserved documents, and the record built from them, are what lasts. These pages are only the current way of showing them, and could be rebuilt in another form with nothing lost. A full count of what the record now holds, the sources, documents and lists behind every page, is set out in the record in numbers.

One standard ties all of this together: every statement on the site is fastened to a registered claim and the source under it, and traced on to the preserved document itself wherever that document survives. The next page follows that chain end to end for a single claim, and measures how far it reaches across the whole record.