Irish Roots: Six tactics of the successful researcher:

1. Lose the blinkers. You need to keep trimming away your own presumptions, because otherwise they'll just grow back. No, not all Cholmondeleys were Protestant. Yes, some nineteenth-century families moved back to Ireland from the US. No, we're not all descended from Milesius.

2. Burrow. If you can't find what should be there, don't give up. Look at records for adjoining areas, look at earlier and later records, try different spellings, different forenames, different families in the same area. . . You are a dog and this is your bone.

3. Know where the devil is. In the detail, of course. For example, that the date of the offence in your great-grandfather's conviction for public drunkenness was the day after your grandmother's birth. He was celebrating.

4. Turn off the computer and go down to the library or archive. What's online may be wonderful, but it's still only a small fraction of what survives.

5. Stare at records. There is nearly always something more to be learned from a record, no matter how well you think you know it. Last month I noticed on my grandfather's familiar 1901 census return that the head of the household in the shop where he was an assistant had recorded him as a "cuzin". Tracing her family showed he was in fact her second cousin, and revealed a plethora of related lines. Welcome to the extended family, all you Flynns, Shines, McManuses and Seerys.


6. Think sideways. Your family were all small tenant farmers, with no property and hence no reason to leave a will. But what about their uncle, the priest? Maybe he left one. And every testamentary record after 1858 is an open book at

It begins to occur to me that the portrait of the implied genealogist shows a sceptical, stubborn picker of nits. Well, if the hat fits . . .