In Praise of Rootsireland

Rootsireland then

For a long time, rootsireland had a very bad reputation among Irish family historians. It was impossibly expensive and seemed designed to restrict research access to the absolute minimum. Small wonder that a cottage industry sprang up devoted to ways of extracting information from the site without paying.


As a result, vitriolic criticism of the site is still very easy to find online.

Rootsireland now

The problem is that it’s no longer justified. Over the past eighteen months, the site has been transformed out of all recognition, guided by what looks like a very good understanding of¬†research and researchers.

First, the clunky pay-per-view unit-based subscription is gone. Now the site offers the online standard, time-based subscriptions, ranging from a year for ‚ā¨225, to (stock up on the black coffee) ‚ā¨10 for 24 hours. Instead of zeroing in on just a few important records, you can now range up and down collateral branches, picking up the kind of secondary information that sheds important sidelights on a family.

And it’s much, much easier to search. The number of obligatory search fields has been slashed. It’s no longer necessary to enter dates, making it possible to retrieve every single record for a surname (and variants) from every record-set on the site in one go: trawl broad, then winnow.

Show me all deaths in Cappayuse townland 1864-1902

In some circumstances, it’s not even necessary to enter a surname. When in one of the county sections, you can retrieve every record that lists a particular townland or address, irrespective of the family it records. And you can use wild-cards (“%” not “*”) to do that. So it’s possible, for example, to extract every death in a particular townland between 1864 and 1920, or to find place-names that were ignored by the Ordnance Survey but are recorded in rootsireland’s early baptismal registers.

The range of records on the site also continues to expand – it looks as if the threat to its effective monopoly from Ancestry and FindMyPast has galvanised it into action. There are still black spots – Fermanagh, Clare, Wexford, those pesky missing thirteen Catholic parishes in East Galway – but the areas that are good are very, very good indeed. And the chance to test the quality of the Catholic transcripts against Ancestry and FindMyPast has seen rootsireland win again and again.

Even the recent addition of the General Register Office record images to IrishGenealogy has only boosted the usefulness of rootsireland’s database transcripts of local registrars’ records: they are two different records¬†of the same events, each with its own mistakes, but each different to the other’s mistakes. And rootsireland’s version is a searchable full(ish) transcript, not just a name index.

Of course, there are still many flaws. The forename search doesn’t use variants, so searching for a Bridget won’t find you a Brigid. There are undocumented holes in some collections, and unlisted records in others – west Galway appears to have the civil marriages for most of its areas online, for example, but they’re not in the sources-list.

That darn map

The practice of filling in the county on their map when they have any records at all online is also deeply misleading: Clare appears identical to Sligo, when Sligo has completed almost everything but Clare has just seven sets of baptismal registers.

And above all, there is no library subscription, an option that would democratise access for the many people who just can’t afford their full prices.

Still, no Irish genealogy site is without its flaws.

I’ve never been a booster of rootsireland. In fact, for a long time they saw me as one of their chief tormentors. Now, though, if I’m asked what’s the one essential commercial genealogy site for Irish research, the answer is rootsireland.

BTOP and a(nother) rant about ethnicity tests

Genealogists are shy, retiring creatures, averse to daylight. Seeing so many of us at Back To Our Past over the last few days, all out in the open, blinking nervously at each other, was just a tad disturbing.

No genealogists. Not daylight-averse

But, as ever, it  was very worthwhile.

Yes, the event is full of ordinary, decent punters driven a little doolally by FoMO (Fear of Missing Out), as they try to collect every last piece of free paper from every single stand in the RDS. Yes, the corporate pitches can be teeth-grindingly high-powered. Yes, there’s always a vague but persistent sense that it’s all pointless but you still have to be there.


But you do have to be there. Nowhere else is there anything even remotely similar, with the entire gamut of those involved in genealogy in Ireland on display, from the to the humblest of crumbs-from-the-table self-publishers.

This year again the Genetic Genealogy Ireland conference ran as a welcome symbiote, and drew packed houses. I only managed a few of the talks, but got some of the best and worst. One jargon-filled hour interpreted in excruciating detail new methods of identifying chromosome subgroups within subgroups within subgroups. It reminded me of¬† Lord Rutherford’s classic remark: “All science is either physics or stamp collecting.” This was stamp-collecting on steroids.

On the other hand, Paddy Waldron’s talk “The Ups & Downes of atDNA matching” (available here) was complex but a model of clarity, using case studies to show¬†just how useful a tool autosomal DNA matching can be¬†when mixed with existing research sources and trustworthy family trees. It persuaded at least one sceptic (me) that, in the right circumstances, atDNA can be a powerful clarifier of family history beyond the horizon of documentary research.

One aspect of the corporate selling of genealogical DNA tests got no attention at the conference. The ethnicity calculator used (especially by Ancestry) to sell its tests will tell you you’re 12% ¬†Native American, 60% Irish and 38% Viking. When challenged, the corporate PR response is to describe this as a harmless piece of fun. It’s not.

Ethnicity testing has become one of the main selling points of the phenomenally profitable genealogical DNA testing industry. ¬†It is pitched at and attracts many people with no interest in genealogy, who think they’re getting a scientific breakdown of their ethnic makeup. That’s just not true. It is pseudo-science at its worst, running on the unspoken but still queasy implication that race has some scientific basis –¬†see here for a summary of the truth.

Apartheid South Africa used to have a test for distinguishing Blanke from Nie-Blanke. If your hair was curly enough to hold up a pencil, you were Nie-Blanke.  Scientifically and ethically, DNA ethnicity testing is its contemporary equivalent.

Rube Goldberg genealogy

I’m a great fan of the eccentric machines cartooned by Rube Goldberg in the US and Heath Robinson over here. They are wonderful parodies of the supposed benefits of automation, the little man blowing a raspberry at the pomp and self-importance¬†of the Machine Age.

How the wizard was written

They also sum up pretty accurately a fair amount of my working life, at least on the software development side. Most of it still feels like having to stand on my head in order to comb my hair. But there is a genuine satisfaction in seeing one of my software contraptions chug into life like a piece of homemade clockwork.

The reason for going on about this is that I’ve finally put up a step-by-step wizard on this site, taking whatever someone knows about their Irish ancestor and producing an unthreatening summary of sources and links to get research going.

The old Irish Times/Irish Ancestors site had something like this, popular enough, but dating from 2012, the era when dinosaurs roamed the landscape of Irish genealogy. When I started re-coding for this present site, the wizard was high on the to-do list but¬†it’s been in production on and off for more than a¬†year.

The wizard start page

The problem was that records were coming online faster than I could incorporate them into the wizard. And every new set of records changed the relative importance of all the others. Rube Goldberg on steroids.

Over the past week, I’ve found myself tinkering with bits of punctuation and wording and realised it’s time to make the thing public. It will always be a work in progress (a euphemism for Rube Goldberg if ever there was one) but it has to be tried out by real researchers.

So give it a go. Try to break it. Make it come up with ludicrous suggestions. Just let me know when you do.

Wills, wills and more wills

An Irish genealogist digesting the new online records

The digesting of last month’s¬†vast info-dump continues.

Among the most spectacular sections of the new material are the two new collections covering wills. To appreciate what they consist of, some background is needed.

The old PRO

The old Public Record Office of Ireland had a huge collection of original wills, divided into pre- and post- 1858. Before then, the Established Church had responsibility for probate. From that year on (as in the rest of what was then the UK) the state assumed responsibility. In the 1870s, the PRO gathered up all the pre-1858 material they could find Рoriginal wills, transcript books, administration bonds, finding aids, correspondence Рcatalogued it all, created indexes, and stored the lot with fanatical defensiveness in the most secure record vault on the island, the PRO Treasury.

From 1858, by¬†law every probated will or intestacy was¬†a public record. So within a few years of probate, all the original wills and grants were passed to the PRO. Where they were carefully catalogued, indexed and stored with fanatical defensiveness …

And then, on June 30, 1922:

Shelling the Four Courts from Winetavern Street, June 30 1922
Shelling the Four Courts (and the PRO)  from Winetavern Street, June 30 1922

Everything in the Treasury was obliterated. All that survived was Reading Room finding aids or material¬†that hadn’t made it¬†back to the Treasury. So the indexes to the old pre-1858 wills survived (and are now newly¬†online here), along with a few other bits and pieces, but the wills themselves didn’t. You can see a will existed, but that’s the only evidence left.¬†“Frustrating” doesn’t come near.

Lots more survived for the post-1858 period. Every will had been calendared, so there’s at least a basic description of its contents. And the District Will Registries outside Dublin kept their own transcript books, into which they transcribed every will before sending the original to the PRO.¬†These comprise¬†the second collection newly online, more than half-a-million full wills probated between 1858 and 1900. Whoopee.

The one thing that’s missing is NAI’s own card index. Over the years since 1922, the Archives has¬†assembled a mixum-gatherum of things that might make up for the enormous loss of testamentary material ¬†– solicitors’ records, genealogists’ abstracts, papers found in Auntie Gertie’s attic – and included them in the Reading Room card index. The only online version of this index is the one created by Eneclann more than ten¬†years ago and now available at FindMyPast.


Roadmap of the Promised Land

How best to use all the new records we have at They consist of:

  • images of 6.2 million births 1864-1915;
  • images of 1.2 million marriage records, 1882 – 1940;
  • images of 4.2 million death records, 1891 – 1965.
Search with nothing in the boxes …

Why am I harping on about images and not just records? Because they are precisely images only. There is no searchable database transcript that would, for example, allow you to pick out only Presbyterian marriages, or narrow your search for deaths to a particular townland. The only route of access to the images is the index, which simply points to the image of the page on which the indexed event is recorded: for all recorded births, starting in 1864, and for marriages and deaths from 1882 and 1891 respectively.

… and get every record they have, including the non-imaged

(Incidentally, I think the absence of the early deaths and marriage images has produced a perverse increase in the numbers buying printouts; the easy availability of the later images has made the earlier ones all the more desirable.  Download the request form here if you really need that itch scratched. )

The IrishGenealogy index is not the only route into the record images, however. The Mormons’ FamilySearch has a full copy of a (flawed) transcript of the old paper indexes, right up to 1958, very useful for twentieth-century births and marriages. And they also have a part-transcript of the birth registers from 1864 to the first quarter of 1881. Which means you can search on both parents’ names on that site¬†and then use the results to see the full original image on IrishGenealogy.

Rootsireland provides¬†another way in. They have¬†plenty of transcripts of local registrar’s records, of which the IrishGenealogy records are copies, many of them coming right up to 1920. Counties with significant numbers are Armagh, Derry, Donegal (excellent), Galway, Kilkenny, Leitrim, Limerick, Mayo, Roscommon, Sligo, Tipperary, Waterford. As always with Rootsireland, keep a close eye on what you are and aren’t searching and don’t be put off by their categorising records into¬†local Registrars Districts rather than the ¬†larger Superintendent Registrar’s districts used by IrishGenealogy. You can see which local districts are in the Superintendents’ districts via the map here. Logainm has detailed maps showing the locations of the local districts.

And of course Rootsireland has its church baptism, marriage and burial transcripts, useful proxies for a civil registration index, especially where a surname is common. The site, it should be said, has improved out of all recognition over the past year. It remains the only essential Irish genealogy subscription site.

For areas now in Northern Ireland, the General Register Office of Northern Ireland’s ¬† is also¬†extremely useful, especially for periods not yet imaged on IrishGenealogy. For imaged periods, their database transcript is excellent for pinpointing records, though most users will then slink off to IrishGenealogy for the free image rather than pay on the NI site. It wasn’t very neighbourly of our Department to drive a tank through GRONI’s business model without any warning, but the result is wonderful for researchers. Neighbourliness be damned.

If you’ve ever managed to get your hands on the original local registrar’s volumes, you’ll know just how useful it can be to go through a local area page by page, volume by volume, year by year. Patterns emerge that aren’t apparent anywhere else – pairs of siblings marrying each other, local epidemics, occupations clustering in a particular village.

So is it possible to use these new images to reproduce the original volumes? Yes, though with difficulty. This is what you see at the top of¬†a register page-image: ¬†civil_url_barThat final number ‚Äú2253803‚ÄĚ is the one you need to change in order to navigate. Add 1 to move a page forward, subtract 1 to move a page back. By doing this you can browse the local registrar’s records. But only for a single three-month period, because the records were copied to the central GRO every quarter. After that you need to find the next quarter, not always easy because there appears to be no discernible pattern to the way the imaging took place: it’s not alphabetical by SRD nor RD nor county.

So there is a job of work to be done here, mapping the locations of the start pages of each quarter for each RD. And it’s a big job of work. For births alone there are 61 years, multiplied by four¬†quarters, multiplied by 865 Registrar’s Districts, for a total of 211,060 points to map.

Get cracking.