The WWP encodes passages printed in mixed-size capital letters so as to capture the effect of relative capitalization which is produced (that is, with larger capitals functioning as uppercase letters, and with smaller capitals functioning as lowercase letters). We term this phenomenon “small capitals” or “smallcaps” even though it includes capital letters of both sizes. This practice allows for the encoding of entire words or text chunks withouth constant highlighting to indicate different sizes of capital letters, and gives an accurate representation of the general effect.
The WWP defines “small capitals” as follows: capital letters of two sizes used together to indicate a difference of case, or (less often) capital letters used in running prose which are clearly of a smaller size than the normal capital letters in the passage. In the former situation, two sizes of capital letters must be present together; a word or passage in capital letters all of the same size cannot be identified as “smallcaps”. In the latter situation, only one size of capitals need be present in the particular word or phrase being encoded as “smallcaps”, but full-size capitals need to be present in the textual surroundings (e.g. at the start of a sentence).
Although printers’ fonts of type do often include a special “small capitals” font, our transcription does not attempt to discern or capture whether or not such a font is in use. Our transcriptional goal is only to capture the meaningful differences in type size (that is, cases where size differences are involved in representing the meaning of the text, for instance by conveying emphasis or setting a word apart from its context).
Small capitals are encoded using the “case” keyword on the rend= attribute: rend=“case(smallcaps)”. The words themselves should be transcribed using uppercase and lowercase letters (uppercase for the larger capitals and lowercase for the smaller). Thus the word London would be fully encoded as: