Resources — 30 April 2026
THE RESEARCH
PROCESS
Half archive work, half listening to the silences
People assume historical research is a matter of finding documents. It is not. Finding documents is the easy half. The hard half is working out what the documents are not telling you, and why.
A service record will give you a number, a date of attestation, a height and a chest measurement. It will not tell you why a man from East Tyrone walked into a recruiting office. A census return will place a family in a house. It will not tell you what was said in that house the night before someone left it for good.
Where I start
Always with the paper trail, because it is fixed and it can be checked. Service and pension records, the Bureau of Military History witness statements, the Military Service Pensions Collection, census returns, parish and civil registers, newspaper archives, court and police files, ships' manifests. Each of these is a fragment. Laid side by side in date order, the fragments begin to argue with one another, and the argument is where the story is.
Then the family
Family memory is unreliable and indispensable. It gets dates wrong, it flatters, it quietly buries the awkward parts. But it preserves the things no archive holds — the nickname, the grudge, the reason a photograph was cut in half. My rule is simple: family testimony tells me where to look; the record tells me what to write.
The silences
The gaps are evidence too. A man absent from a record he should be in, a year with nothing in it, a name that stops appearing in the local paper — these are not dead ends. They are questions. Some of the most important things I have found came from asking why something was missing rather than what was there.
Writing it down
Research without a reader is a filing cabinet. At some point the boxes have to become a story, and that means making choices: what to keep, what to cut, where to slow down and let a single moment breathe. I try never to invent, and never to bore. Where I have had to speculate, I say so on the page. The reader deserves to know which bits are bone and which bits are flesh.
Seven years went into the first volume. Most of it was reading things that led nowhere. That is not wasted work — it is how you earn the right to be believed on the pages that do go somewhere.