Micro scale

Investigating surname distribution and frequency – the micro scale

Small-group surname statistics

• Take a listing of all those in your year.
• Count the number of occurrences of each name.
• Compare the most listed names in your sample with e.g. Smith, Jones, Williams, Patel (UK national frequencies.) (USA national frequencies)
• How many names only appear once? What percentage is this of the total?
• Are singletons the largest group?
• Repeat this exercise with an electoral roll extract – maybe a couple of streets around where you live.
• Are the results comparable with the school-roll exercise?
• Count the number of syllables in each name. Does any number of syllables predominate?
  
  Syllable-length of UK’s top 100 surnames
  (Source: personal count)
• Does the number of syllables in your sample differ from the above pie chart?
  (For more on linguistics and names see Variance: Appendix 2)

Surname length

• A 1968 sample survey of the surnames in the London Telephone directory found that the average length was 6.47. A later survey of a large hospital patient file had similar result (6.38) as indeed did a recent survey (Jan 2004) of a large sample from the 1881 census:

No. of letters	2	3	4	5	6	7	8	9	10	11	12	13	14	15+
Freq.	0.1	1.4	10.1	19.1	24.7	20.3	22.3	6.8	2.7	1.3	0.7	0.2	0.2	0.1

(Source: M Healey The lengths of surnames Journal of the Royal Statistical Society-A, 1968)

• The above sample included hyphenated names.
• How does the frequency distribution compare with your modern-day classlist?
• Do the letters 5-8 still account for at least 85% of your list?
• Are there more longer names (13+ letters) reflecting the growth of hyphenated names given to children?

Initial letters of surnames:

An analysis of the electoral register for 1957-8:

	England	Wales	Scotland	GB
	%	%	%	%
A	3	1.7	2.9	3.0
B	11	6.5	8.0	10.5
C	8.1	4.2	8.4	7.9
D	4.3	7.4	5.3	4.5
E	2.1	5.3	1.0	2.2
F	3.5	1.8	3.9	3.4
G	4.9	4.2	5.4	4.9
H	9.2	8.2	5.5	8.8
I	0.4	0.2	0.6	0.4
J	2.8	11.4	1.7	3.2
K	2.1	1.2	2.5	2.1
L	4.1	4.5	3.8	4.1
M	7.5	6.7	21.1	8.9
N	1.7	1.1	1.3	1.6
O	1.3	2.1	0.9	1.3
P	5.7	6.6	3.3	5.5
Q	0.1	0.0	0.2	0.1
R	4.9	6.7	5.6	5.1
S	9.2	4.8	8.5	8.9
T	4.1	5.4	3.3	4.1
U	0.3	0.1	0.1	0.2
V	0.5	0.6	0.2	0.4
W	8.8	9.1	6.0	8.5
X	0.0	0.0	0.0	0.0
Y	0.4	0.2	0.0	0.0

(Source: abridged from Percy Gray, Initial letters of surnames Applied Statistics, March 1958)

Note that:

• England- B is the most common initial letter (11%).
• Wales – J is the most frequent (11%) – and noticeably higher in rural Wales.
• Scotland – the initial letter M accounts for 21% of the surnames (and within the crofting counties, the proportion rises to almost 40%.
• Compare your list with the above table (but with the proviso that a national snapshot today would probably be subtly different).

• The above plot of each initial letter reveals interesting peaks and troughs.
• Are some initial letters more conducive to names because they can form consonant clusters for strong syllables?
• Or perhaps because those initial letters are rich in suffixes or because they mimic the pattern of initial letters of given names?

As a further analysis, the individual figures have been separated and ranked in decreasing order:

The graphs for England and Wales do seem to be reminiscent of a power-law relationship(?) – though not for Scotland.

Initial Letters forming 50% of surnames:

England	Wales	Scotland
B	J	M
H	W	S
S	H	C
W	D	B
C	M	W
M	R
	P

Probably, at a sub-national level, these patterns are not sustainable. But I am intrigued as to why they exist, and are identifiable in analyses 50 years earlier. None of the frequencies bear any relationship to the general frequencies of initial letters in the English Language, and for England, in particular, the letter B is not a predominantly fertile source of given names.

A survey of the New Oxford Dictionary of English found that the leading initial to form words was S, which was followed at some distance in decreasing order by P, C, D, M, and A. So surnames are not congruent with the background of words from which they are formed: but why not?

Contemporary Spelling

Have you noticed how the spellings of surnames differ from the noun they derive from? For example, Taylor rather than Tailor. Spelling variation seems to be tolerated rather than forbidden, in the case of surnames. For more on this topic, see the topic Modern Spelling on the page on The Spelling of Surnames.