Years ago, in my population genetics days, we discovered a lot of Manx cats in parts of Maine and New Brunswick. By "a lot", I mean countable examples because, on the streets of New England and the Canadian Maritimes, Manx cats are really rare; so even a handful needs some kind of explanation. One of the hypotheses to be tested was that there had been particular and peculiar traffic between these maritime towns and the Isle of Man - where Manx cats are indeed reasonable common. In those pre-WWW days you had to scrabble about for data to analyse and I spent some time in the library looking at the Shipping News columns of local-to-Maine 19thC newspapers. That came up empty. So I turned to telephone directories to see whether there were more-than-expected typically Manx surnames [Clague, Corlett, Kelly, Moore, Quayle, Quirk] in Augusta, Me and St Stephen, N.B. That yielded enough data to analyse . . . and show that settlement from the IoM was not a good explanation for the excess of tailless cats.
Channelling Pat the Salt in his aphorism never move anything twice, it seemed to us sensible to add an empty cabinet at each end of the existing row of seven and start distributing the A's into one set of empty drawers until bored and then start similarly disbursing the X Y Zs to the the other empty cabinet. That way she'll be able to leave the Mc's and O's in the centre of the Irish alphabetical world. So there is an implicit hypothesis: that these patronymic surnames are indeed more common in Ireland.
How to test that? Find a telephone directory [remember them?] and count how many pages were devoted to each initial letter. For reasons inertia, we have retained a copy of the 05 Sunny South East telephone directory from 2010. And here is the data:
|D||78||27||8.3||2.99||9.97||D E F|
|G||126||11||3.4||1.22||13.51||G H I|
|J||158||3||0.9||0.33||16.17||J K L|
|P||270||11||3.4||1.22||29.46||P Q R|
First thing to note is that frequency of Initial letters [M C B D O K H F R W S L G P T N A E J Q I V Y U X Z] is very different from general letter freq for English etaoinshrdlu. Murphy and Mc/MacNames do indeed fill the middle ground of Irish name space, O' not so much: shoved down the list by Connollys Byrnes Dohertys. You might think, and we did consider, that it would be better to audit the actual data in the existing file-cabinets. But that is not really future proofing the system: At the present moment, for example, there is nobody in the City Library Service whose name begins with A. But there may be in the future, especially if the Library succeeds in diversifying its employee base - think Akinrunde, Battacharaya . . . Yong, Zieliński