1 With very few exceptions, Tallis's vocal works are edited in Thomas Tallis, Tudor Church Music (hereafter TCM), 6 (London, 1928), and Thomas Tallis English Sacred Music I–II, ed Leonard Elhnwood, rev Paul Doe, Early English Church Music (hereafter EECM), 12–13 (London, 1971, rev 1973–4) For the consort works, see Elizabethan Consort Music, I, ed Paul Doe, Musica Britannica, 44 (London, 1979) No critical edition exists for the keyboard music, and this repertory has not been taken into account in the present study, even though several pieces exist in significantly different forms in the various sourcesGoogle Scholar

2 For a detailed bibliographical study of the Cantiones sacrae, see Byrd, William, Cantiones sacrae (1575), ed Craig Monson, The Byrd Edition, 1 (London, 1977), v–xiii Evidence for corrections made during the print-run is shown in footnotes to the music The only other printed edition containing works by Tallis that may have been prepared under the composer's supervision is Archbishop Matthew Parker's The Whole Psalter Translated into English Metre (printed for private circulation, London, [c)567]), for which Tallis provided nine harmonized tunes There is no evidence that these tunes existed before 1567, and their only manuscript source – Lbl 15166 – is close to the printed edition (Full expansions of the manuscript sigla used in this article are given in Appendix 1)Google Scholar

3 Edmund, H Fellowes, Tudor Church Music Appendix, with Supplementary Notes (London, 1948), 24 For Tallis's contributions to the Cantiones sacrae, see TCM, 6, 180–241, Dum transisset is on pp. 257–61Google Scholar

4 Much of the material in the present article is condensed from a more detailed evaluation of the sources in John Milsom, ‘English Polyphonic Style in Transition A Study of the Sacred Music of Thomas Tallis’ (D.Phil dissertation, University of Oxford, 1983), Chapter 2 ‘The Chronology and Transmission of Tallis's Vocal Music’ This includes full parallel transcriptions of many of the works that survive in more than one versionGoogle Scholar

5 Edition in TCM, 6, 123–43Google Scholar

6 Milsom, John, ‘A New Tallis Contrafactum’, The Musical Times, 123 (1982), 429–31 I am grateful to the President and Fellows of Corpus Christi College, Oxford, for permission to consult and quote from this manuscript For a full transcription of Occ 566, see Milsom, ‘English Polyphonic Style’, Appendix 2.8Google Scholar

7 This interpretation was first put forward in EECM, 12, 129.Google Scholar

8 Edition in TCM, 6, 219–21Google Scholar

9 The eighteenth-century score of this version in Lbl 31226 was evidently copied from Lbl 30480–3; the fifth voice added in this late source is clearly inauthenticGoogle Scholar

10 EECM, 12, 68–72 For a parallel transcription of both versions, see Milsom, ‘English Polyphonic Style’, Appendix 2 14.Google Scholar

11 Milsom, John, ‘A Talhs Fantasia’, The Musical Times, 126 (1985). 658–62. this includes a transcription of the music contained in Lbl 7578 See also the letter of response from M A O Ham in The Musical Times, 127 (1986), 74, in which the anonymous anthem Deliver us, O Lord our God in Lbl 17792–6 is identified as being another early version of O sacrum convwiumCrossRefGoogle Scholar

12 For transcriptions of various versions of this work, see Milsom, ‘English Polyphonic Style’, Appendix 2.13 O sacrum convivium is edited in TCM, 6, 210–12, this largely follows the reading of the 1575 Cantiones sacrae For additional variants, see the Appendix. 24 and 27 For an edition of I call and cry to thee, sec EECM, 12, 60–7Google Scholar

13 The two ‘versions’ are in TCM, 6, 276–8 and 279–81, see also the Appendix, 24 and 27 For a transcription of the variants in early versions of this work, see Milsom, ‘English Polyphonic Style’, Appendix 2 18Google Scholar

14 For a list of manuscripts from the library of Edward Paston, see Brett, Philip, ‘Edward Paston (1550–1630) A Norfolk Gentleman and his Musical Collection’, Transactions of the Cambridge Bibliographical Society, 4 (1964–8), 51–69Google Scholar

15 EECM, 12, 132, the music is edited on pp 106–10, and in The Mulliner Book, ed Denis Stevens, Musica Britannica, 1 (London, 1951, 2nd edn, with further amendments, 1962), 63 For a parallel transcription of various versions of the song, see Milsom, ‘English Polyphonic Style’, Appendix 2.6Google Scholar

16 Although the edition is dated variously 1560 and 1565 in its constituent partbooks, there are good reasons for believing that publication was envisaged before 1552 but delayed by political events. See Aplin, John, ‘The Origins of John Day's “Certaine Notes‘”, Music and Letters, 62 (1981), 295–9Google Scholar

17 EECM, 12, 126–7. For a more detailed discussion of the various versions of this anthem, see Milsom, ‘English Polyphonic Style’, Appendix 2 3Google Scholar

18 Editionen EECM, 12, 111–16Google Scholar

19 Edition in EECM, 12, 43–50 See also the version in The Mulliner Book, 36–7Google Scholar

20 The Taverner quotation was first observed by Richard Abram in a letter to The Musical Times, 118 (1977), 642 For an edition of Dum transisset, see TCM, 6, 257–61 Other variants from the early version are shown in Milsom, ‘English Polyphonic Style’, Appendix 2 10Google Scholar

21 See Milsom, , ‘English Polyphonic Style’, Appendix 2.15 At least part of Absterge domine derives from an earlier, rejected composition See Milsom, ‘A Tallis Fantasia’, 659.Google Scholar

22 There is good reason to believe that Och 979–83 is copied on the printed music paper that Tallis and Byrd issued under the terms of their 1575 monopoly A similar arrangement to that of Och 979–83 – printed music paper bound in front of a copy of the 1575 Cantiones sacrae – is found in the partbook set of which T 389 and the ‘James MS’ are the only surviving members This may imply that such compound volumes were issued for sale in ready-bound form See Iain Fenlon and John Milsom, ‘“Ruled Paper Imprinted” Music Paper and Patents in Sixteenth-Century England’, Journal of the American Musicological Society, 37 (1984), 147–8Google Scholar

23 These include Lbl 22597, Lbl 29247, Lbl 30480–4, Lbl 32377, NY 4180–5, Ob 1–5 and T 341–4; their variant readings, many of which are noted in TCM, 6, and the Appendix, should be regarded as corruptions Also derived from the Cantiones sacrae is the reading of Suscipe quaeso domine in Lbl G 16–20, Lbl G 21–6, and T 341–4, all three of which are Paston manuscripts This motet has sometimes been described as a partner to the Mass ‘Puer natus’, which is copied adjacently in these sources and is scored for a similar seven-part vocal combination However, the fact that the manuscript readings follow the print so closely suggests rather that the copyists relied upon different sources for the two pieces, and brought them together on account of their common composer and similarity of scoring On internal evidence of style, Suscipe quaeso may be placed around 1570, close to Spem in alium, to which it is related in contrapuntal technique, see Milsom, ‘English Polyphonic Style’, 33–4 and 181–5 The Mass ‘Puer natus’ is most likely to have been composed in 1554, see Doe, Paul, Thomas Tallis (2nd edn, London, 1976), 21Google Scholar

24 Tallis was not the only mid-century composer to suffer from such insecurity, though evidence of recomposition among the works of his contemporaries is scarce and generally of a different kind. Most closely related, perhaps, is the case of Robert Johnson's five-part psalm-motet Domine in virtute tua, which exists in two versions, neither of them published in a modern edition Retaining many of the imitative points of the first version, Johnson reassembled them into entirely different lattices to make up a second version, the melodic ideas remain largely similar, but the working of them changes completely, with cell-like alternations of paired voices in the earlier version giving way to less regimented, more consistently five-part imitative expositions in the later one Tallis apparently never went to such an extreme of dismembering and recomposing the entire substance of one of his pieces, but the changes he made to the inner textures of several scores was evidently motivated by a similar interest in discovering new possibilities of imitative combinationsGoogle Scholar