Once one has one's corrected transcription in hand, one can begin to assemble the evidence on which the edition will be based. This is composed of the readings of all the manuscripts, the comparative material by which one's copy-text might be verified – or revealed as requiring correction. The procedure for amassing this material, known as ‘collation’ (literally, bringing the copies together or side by side) – although not analysing it – is extremely straightforward.

One simply reads the copy-text, word by word, against every other relevant version. Wherever one finds a substantive deviation between them, one notes it. This word-by-word comparison will generate a relatively vast amount of data, what is known as a ‘corpus of variants’. From this wad of material, the editor will pass on to assess the value of these diverse readings, word by word – and from that assessment, construct the edited text.

As a procedure, this seems simple enough. But again, just as in transcribing, accuracy is key. Just as with your transcription, you will need to check every manuscript at least twice, a second time to verify that you have noticed everything that might be relevant, and that you have copied all these details correctly.

But recording and keeping track of all the data you unearth may prove difficult. The obvious way to have everything would involve transcribing each of the manuscripts and using highlighters or coloured pens to identify variations. But this strikes me as a monumental amount of work and still leaves one, in the case of Rolle's ‘Super Canticum’, trying to cope with fourteen separate files of materials.

Here Manly and Rickert, in their extensive edition of Chaucer's Canterbury Tales offered a very convenient halfway house and described in extensive detail a procedure useful for a beginner. These editors were dealing, in the main, with Chaucer's relatively short verse-lines. They assigned to each verse an index card and wrote their full copy-text out across the top. Below this, they left the card blank, with space to enter the readings varying from their copy-text – those transmitted otherwise in each individual copy. They compared each card in turn with the individual manuscripts from which they derived readings.