ANACROGRAMS: prolegomenon to acrostic anagrams.

Take the initial letter of each word in a common phrase, saying or longer quote, rearrange them and form a word or phrase that summarises or relates to it. Such anagrams of acrostics I call ANACROGRAMS. Sounds passe? Maybe. Impossible? Usually: those I'll present later are a tiny fraction of all the candidates considered. But since the latter in turn form a very tiny fraction of possible sources, this new game opens virtually unlimited vistas for meaningful word play, perhaps as much as straight anagramming. An example: < No man is an island. Main, I. >

Success depends of course on an adequate supply of vowels. But it's also a lot harder to anagram first letters than whole words. Letter frequencies are quite different in the two. E N R are three of the seven most common letters in text but frustratingly uncommon as first letters. Conversely, T A and especially W are more common in acrostics than in text due to highly repeated words like the, to, a, at and and. W has more of these logological pests than any other letter: with/out, was-were, will-won't, would/n't, who/m/ever, whose, what, which (where/as why when/ce while we went, well, worse?)--not to mention world wars, wide webs and word ways!

As a hint at first letter frequencies I did manual counts on five pages of fiction (1766 words) from four different authors (Dickens, Melville, Twain, Ian Watson) In the table below I compare them to dictionary and overall frequencies. Of course a far larger and broader count is needed and should alter the positions of some letters. Does anyone know of such a study?

letter rank  T  A  S  W I  H O M F P C D B L  E G R N U V J K Y Q Z X
% of 1sts   14 13  9 -8-  -6-  4 -----3-----  ---2--- --1---- --0*---
% o/all **   8  9  7  2 8  4 7 3 2 2  4 4 2 4 12 2 7 8 3 1 0 1 2 0 0 0*
% W3 ***     6  6 12  3 3  4 2 6 4 9 10 5 6 3  4 3 5 2 2 2 1 1 0 0 0 0*
or by rank rather than percent:
letter             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
1st letter rank    2 13 11 12 15 9 16 6 5 21 20 14 8 18 7 10 24 17 3
                     1 19 20 4 26 23 25
overall rank       2 18 12 11 1 15 17 10 4 24 22 9 13 5 6 16 26 7 8 3
                     14 21 20 23 19 25
Dictionary rank    4 6 2 8 12 10 14 11 15 22 21 13 7 18 17 3 23 9 1 5
                     19 20 16 26 24 25.

* 0 = <0.5
** overall frequency of all letters in the Cambridge Encyclopedia
 (Crystal, cited below)
*** % of pp in Webster's 3rd NI Dict., ie, [approximately equal to] %
 of words starting with that letter Note that many words but not so
 many frequent words start with C, P, B, R and D, while the opposite
 is true of I, O, H and of course W.

Loath as I am to being pragmatic, it's worth considering first letter counts as an aid in identifying authorship of disputed writings, as a supplement to and possible improvement on doing so with whole letter counts as illustrated by David Crystal in Cambridge Encyc. English Language (1995).

Many examples of anacrograms will be revealed later. Meanwhile, try your head at it. It's fun. It's addictive.

ANIL

Perth, Australia