I’ve written a bit before about the online resources available to genealogists and what can (and can’t) be done with them. It seems that every week something new is coming online, and the quality of digitized images and search parameters continues to improve almost palpably. But recently I’ve been working with some much more old-school tools – printed indexes – and have found working with them can be really quick, unstressful, extremely thorough and for a number of reasons much MORE efficient at finding that ancestal needle in a haystack than the slick interfaces of many online search facilities.
While nothing can beat working with original census returns, parish registers and the like for providing as much historical, social, geographical, chronological and cultural contextualization as one can hope to reconstruct – carefully! – from these records of a byegone age or community, the limitations imposed by ONLY having access to these should be obvious: travelling to see an original isn’t always possible, and trawling through heavy tomes, delicate parchment fragments or motion-sickness-inducing microfilms is certainly time-consuming and tiring, plus it can be all the more frustrating if information is illegible, elusive or absent. Online availability and search functions can and do save a lot of time and effort IF you know how to search and, equally importantly, if the resource allows you to search in ways that are most likely to turn up what you are looking for – if it exists – without being either too narrow or too broad in focus. More about this later, but the basic point I wanted to make here is that for all the time and effort you can save through effective online searching, you are of necessity narrowing your focus each time you search and possibly missing out on important context and tangential discoveries that can ultimately provide a crucial missing link.
So what about the published indexes I mentioned earlier – what makes them so good, if they are neither the original source nor equipped with a high-tech search function? What I ordered was indexes to the (approx. 17th-19th century) baptism, marriage and burial registers for two parishes in South Wales, Llantwit Fardre and Pentyrch, produced and distributed in booklet format by the Glamorgan Family History Society and orderable online via e.g. Genfair. They arrived in the post remarkably quickly and I started flicking through them immediately, and with almost instant success. So who’d have thought a booklet could be so user-friendly? It was neither heavy nor unwieldy, very easy on the eye, there was no scrolling, clicking on links, waiting for pages to load, no mechanical typing of names and dates into boxes, and all sorts of details that caught my eye here and there without me having to make a deliberate decision to search for them. (Of course I’m not seriously surprised by any of this, though I’m not sure whether I should be amused or rather appalled that someone else might be…)
But the advantages of this format go much further, and here I can make direct comparisons with online resources, starting with findmypast.co.uk. Much of the information contained in these same booklets has been digitized and made available to search online at that site (I wrote about this a year ago, when that service launched its main Welsh collection), and I must say that the search parameters are pretty flexible: variants in spelling can be taken on board (and are handled sensitively and intelligently, on the whole) or ignored as you choose, and you can search by fore- and/or surname as you like, plus restrict by date and/or parish if relevant; for some of the records, a digital reproduction of the original pages of the registers can be accessed and browsed, so the online site does win in terms of access to something approaching an original (for those records where they have been made available, at least – coverage does seem patchy). However, the booklet index offers two further approaches that are not available as search options on findmypast; and these approaches prove absolutely invaluable (one could almost say, indispensible) in the context of these rural Welsh communities:
1. Patronymic indexes for baptisms
In the original registers, a 17th- or 18th-century baptism might be recorded as something like “David son of Evan Richard”, preceded or followed by the date of baptism (the mother’s name is not always given). The modern way of doing things would be to assume that little David was identified as David Richard, but depending on date, area and specific family tradition, back then he might in fact have been known henceforth as David Richards, David Evan, David Evans, David Bevan, or even David Evan Richard (with David Pri(t)chard and David ap Evan being further, albeit less likely choices).
GFHS approach to patronymic patterns and spelling variants
The booklets take this “systemic inconsistency” (my term: I know it sounds a bit flippant) into account, offering one index following the modern practice for all baptismal entries but for pre-Victorian entries also providing a further index alphabetized according to the father’s (or in e.g. some cases of illegitimacy, the mother’s) first name. All names are transcribed as they appear in the registers, so one does need to take care to look under e.g. Howel AND Howell AND Howells, but you can see immediately whether there are entries for these (or further) variants and it really doesn’t take long to flick through the entire index if you’re on the lookout for more unusual variants (especially those that might not be alphabetically adjacent).
Online search facilities’ approaches to the same issues
The online resources largely ignore (or are unaware of) the existence of the older naming system. On findmypast, if you suspect the individual you are looking for may have been given an older patronymic-style name, you have to leave the surname blank and hope for a manageable number of results, then click your way through to every individual record to see whether the father’s first name is what you might have expected – which can be pretty time-consuming. In addition to this, some records have been attached to the wrong parish name, or the parish name is missing, so even if you restrict by parish (which is perfectly possible), you might not get an entirely reliable set of results (this is a known issue and they are working on it, by the way).
To the best of my knowledge, only familysearch.com gives you the option of searching using a combination of child’s first name and father’s first name only (other combinations are also possible), and this LDS site also has a powerful, apparently pretty reliable function for taking on board variant spellings. A further resource, freereg.org.uk, which in general I find absolutely laudable in its efforts to make church and chapel register entries freely available online, some of which are just not available on the “bigger” or more commercial sites, really falls by the wayside here and becomes extremely hard to use for older Welsh records: not only are you obliged to enter a (modern-style) surname, but e.g. Evan and Evans are considered two separate names, even if you check the “include variants” option, and so there is a lot of doubling-up on searches and making a mental note of your various “+s” and “-s” search history, assuming you even have a semi-workable surname to search for in the first place. Likewise, ancestry.com is not terribly reliable when it comes to soundex other ways of searching for variants. Ultimately, none of these technologies replaces the watchful human eye in that respect (or, indeed, various others).
A practical example of how some of these features play out: I had an elusive ancestor called Mallt Thomas (according to her marriage record) who was born in around 1735 (calculated from the age given on her burial record). I had hoped to find her baptism in Llantwit Fardre (where she married and was buried) but I drew a blank in both of the baptism indexes for that parish. I decided to take a look in the Pentyrch records while I had them immediately to hand, and hey presto, I found her in the patronymic index: she was the daughter of a Thomas William. The Pentyrch baptism records for this date seem not to be available on findmypast or familysearch, so I wouldn’t have found her there even on the basis of a forename-only search. And on freereg, which does have the records for that date, I’d potentially have had to try out all sorts of guesses for the surname box – admittedly, William(s) would have been one of my first choices, but still, the booklets were much quicker in providing the information I was looking for.
2. Details of abode / address
William(s) turns out to be just about one of the most common surnames (by any of the systems used) in this part of South Wales over the entire period covered by the registers, and another headache I’ve had to deal with on a regular basis is the sheer number of individuals with identical names and similar dates of birth. Not surprisingly, back in the day people were often referred to by attributes other than their name – their profession (think “Jones the Voice” or one or two of the characters from Under Milk Wood) or their abode. My ancestor David Morgan (Mallt Thomas’s son, incidentally) had brothers called Thomas and Evan, but then again so did a great many David Morgans in South Wales at the time. I know from passed-down snippets of information, though, that Thomas was better known as “Twmi’r Gedrys” and Evan as “Ifan o’r Waran”, Gedrys and Waran (or Warren) being the names of the farms they managed. Flicking through the Morgan entries in the Llantwit Fardre registers can be a daunting task – based on the names alone, you can’t see the wood for the trees – but the presence of an additional column on each page for any abode noted in the original PRs can REALLY help to pick out a possible spouse, children etc. Again, spelling variants are part and parcel of this – and place names do genuinely evolve over time – but it really is very quick to scan the “abode” column for any forms of names that might look familiar or resemble others.
How do the online resources compare here? Rather poorly. In most cases you can’t search by abode on any finer level than the name of the parish itself, and even if an address, farm or township is specified in the original records, it is not necessarily transcribed in the indexes that have been made searchable. Findmypast has in some cases double, even triple sets of data derived from different “levels” of indexing, one with the residence included and the other not – so you do need to learn which results to look at first to get hold of this information; better still, look at the original image if it is available (and if your mouse-clicking finger isn’t tired by this point).
Who provides the transcriptions in each case is also of relevance. The family history society transcripts and indexes are put together by genealogists (largely unpaid volunteers!) with knowledge of the locality concerned, manuscript and handwriting conventions (including a range of abbreviations), and the language(s) of the region. Even the best transcriber of historical records will have to scratch their head and make a “best guess” from time to time, but the results are plausible and variance is only slight. In contrast, the census transcriptions available on the biggest international genealogy site, ancestry.com, are in some cases WILDLY inaccurate and can only have been done by people without the specific range of knowledge mentioned above. People’s names, addresses, ages, family status and occupations can all be mistranscribed, rendering them unfindable in searches, and Welsh placenames understandably present a further level of challenge. Even the National Library of Wales’ (largely excellent, extremely valuable) probate records search is hard to use from the aspect of place names: it’s great that you can search by parish, but in each case only one spelling – sometimes idiosyncratic or obscure – will turn up any results (e.g. “Pen-tyrch” will work, but the normal spelling “Pentyrch” will draw a blank). Sometimes only the Welsh name / spelling is accepted, sometimes only the English: it seems rather random. This must be quite daunting for someone with little or no knowledge of Wales or the Welsh language looking for their Welsh ancestors. The search function also offers the option of searching by “township” – this might be the name of the hamlet or farm where the testator lived, if such specific detail is included in the will or bond – and this functionality is absolutely brilliant IF said township has a stable spelling OR you are good at guessing variants, but there is no soundex function on any of the search parameters. In some cases I have ended up calling up all the wills for a particular parish in order to be certain not to have missed anything due to a rogue spelling used in the will (in fairness to NLW, it is rarely a case of transcriber error in these cases).
At the risk of repeating myself, I’m certainly not out to diss any of the online resources available – each one has its strengths, as I’ve tried to make a point of showing here, and together they are continuing to revolutionize people’s access to family, local and social history, some of them at no cost to the end user. However, this recent experience of the sheer uncomplicated joy of dealing with more low-tech resources that are so reliable and effective has just reminded me that in so many areas, human skills and knowledge remain the key to handling all the “data” available out there, in whatever form it might exist.