There exist in seventeenth-century manuscripts and early editions several versions of the Holy Sonnets, differing in both number and arrangement. Two collections of twelve sonnets – distinct in ordering and content though sharing eight sonnets in common – are now believed to represent Donne’s original and revised sequences. A number of collections of sixteen sonnets appeared, each conflating according to its own plan the two collections of twelve; none, however, is believed to have been fashioned by Donne himself. Finally, the Westmorland manuscript consists of nineteen sonnets: the first twelve follow Donne's original plan; the four found in the revised set of twelve but not the original come next; three sonnets extant in no other source complete the collection. This manuscript had been preserved in the library of the Earl of Westmorland and not discovered until 1892. It was this version of the Holy Sonnets that became known to Britten. From the nineteen, the composer selected nine and made from these an arrangement quite radically different from the ordering of any other version.

Figure 2-1 shows incipits of the nineteen Holy Sonnets as one finds them in modern editions, ordered to reflect Donne's final arrangement of twelve (the “revised sequence,” numbered consecutively in boldface), followed by the four found in the original, but excised from the revised, sequence (numbered 1–4), followed finally by the three peculiar to Westmorland (1–3). Figure 2-2 gives the Westmorland ordering described above – the arrangement that Britten encountered. Beneath that, Figure 2-3 shows the nine sonnets Britten selected and how he organized them.

One of the questions that drove the investigation of manuscripts and seventeenth-century editions was whether Donne had intended that broad themes be addressed by groups of sonnets. If he had, textual authority would of course be an especially critical issue, since ordering varies much more widely across the various editions than does content. Herbert Grierson, whose labors helped stir a burgeoning enthusiasm for the work of the metaphysical poets in the early twentieth century, found no sign of such themes, believing each sonnet to be a “separate ejaculation.” His ordering follows a 1635 edition that is now believed to lack authority, but of course he would have considered this to be of little consequence.