The history of English spelling is an eventful one, from Old English with an almost one‑to‑one sound‑to‑spelling relationship, to Modern English, notorious for its sound‑to‑spelling unpredictability. In between lies a vast period characterised by immense spelling variability, reflecting the cumulative effect of dialectal variation and lack of uniformity, additionally compounded by the mode of text transmission in the manuscript culture, whose characteristics were adopted in a wholesale fashion into the culture of early print. In effect, early printed books present a rich kaleidoscope of spelling variants, which – not infrequently – co‑occur on the same page or even in the same line of a printed text. This paper addresses the issue of this variability with a view to measuring in mathematical terms the degree of internal spelling variation within a text and showing that much of the spelling variation is associated with compositors as agents in the printing process. The analysis of internal spelling variation is based on George Joye’s 1534 English translation of the Psalms printed in Antwerp and aims at identifying parts of the text which are similar or different in terms of spellings by applying cosine similarity measurements performed on individual quires of the publication.

It is the objective of this paper to analyse selected English Renaissance translations of the Book of Psalms in the light of their reception. In particular, I intend to illustrate how a strong preference for a familiar rendition over a new one (regardless of its quality and status) showed itself in the textual composition of the most important book of the Anglican Church – the Book of Common Prayer. Discussion of the Psalm translation selected for the five successive versions of the Book of Common Prayer against the backdrop of the emergence of new renditions of the Psalms leads on to formulating a desideratum for sound methodology which would express the level of similarities between texts in mathematical terms and in this way objectivise assessments of Psalter renditions. The paper offers a preliminary attempt at such methodology by applying the cosine distance method. The obtained results need to be verified on a larger corpus of data, but they are promising enough to consider this method an important step towards assessing Psalm translations.