Appendix 1 Variations in Reaney Reaney(1) gives no indication of how he chose to group variants. How did he decide on a head-form? Was it on etymological grounds, the most occurrences, the earliest form or the latest form? With these uncertainties, a peremptory analysis of variants using Reaney is bound to be somewhat suspect. I spent a week at a hospital bedside, and to pass away the time, I decided to count how many of the variants in Reaney (3rd ed) differed in the lead consonant from the head-form. Most variants are created through a change in a medial vowel sound, or through a doubling of consonants, or through the addition of a suffix. The initial consonant is deemed to carry most ‘information’ about a name, and hence if it changes, that should be significant. In the following table, column 1 is the letter of the head-form, and the subsequent columns indicate the count of variants that do not match that initial letter. (1) Reaney Dictionary of English surnames 3rd ed OUP, 1995 a b c d e f g h i j k l m n o p q r s t u v w x y z Total Spread A 3 4 68 1 53 6 11 9 7 1 4 167 11 B 3 2 2 14 21 4 C 4 7 50 4 1 2 11 79 7 D 2 5 2 4 2 15 5 E 15 1 38 10 2 3 5 1 4 3 80 10 F 4 11 21 58 94 4 G 7 2 1 2 1 59 11 2 2? 20 107 10 H 30 2 5 5 1 11 10 64 7 I 6 15 2 23 3 J 1 30 1 1 33 4 K 16 2 2 4 24 4 L 6 6 1 M 1 1 5 7 3 N 8 2 10 2 O 4 4 14 1 8 1 1 6 39 8 P 2 3 1 6 12 4 Q 1 1 2 2 R 1 1 1 S 9 1 1 2 1 1 6 21 7 T 5 5 1 U 5 1 3 9 3 V 12 1 9 22 3 W 1 4 13 1 4 12 1 36 7 X 0 0 Y 4 4 1 Z 0 0 51 5 32 44 85 29 50 128 18 63 50 61 10 24 33 47 0 7 12 5 38 59 21 0 3 6 881 a b c d e f g h i j k l m n o p q r s t u v w x y z Total Spread Provisos The analysis should have been on phonemes rather than consonants e.g. P’s will cover both a hard ‘P’ and ‘F’ sounds as in Paul and Phillips. There is grapheme-to-phoneme guidance which helps. (Edward Carney A survey of English spelling Routledge 1994.) Even so, names often do not adhere to these rules, and a personal judgment would still have to be made as to pronunciation. Correspondences with D, L, M or O can be explained in some cases with the dropped suffixes -De, Le, Mc and O’. Analysis Names that are most likely to ‘mutate’ are those commencing with A, then G then F, E, C, H and O. L, R, Tand Yare the initial consonants least likely to be altered. In the case of L, all the changed initial consonants are due to the dropping of the prefix De. The resistance of the letter ‘L’ may be due to the fact that phonetically, it belongs in a class of its own, as a lateral (where the sound is uniquely made along the sides of the mouth), and therefore less likely to be confusable with other consonants. If consonants and glides are considered together, then they are not grouped at the top ranks -as might be expected- but throughout the ranking. Voiced and unvoiced consonants e.g.T and D, F and V, P and D do not appear in groups. Variants do not seem to be evenly distributed between similar confusable consonants. There appears to be a movement from one to the other e.g. Under F, Reaney groups 58 names beginning with V, but for V, only 12 F’S are listed. The most common changed-into consonant is ‘H’ – primarily from A and E. Is this a result of Reaney’s ‘system’ – or does it indicate something more significant about Variant inial letters?