Garrod, or from the versions given in H. E. Rollins’s edition of the Letters (often the single source).

For a full account of the manuscripts and their variants, and for a record of the ‘fingerings and gropings’ of Keats’s creative imagination, the reader is referred to Stillinger’s edition and his book on the texts. Like Garrod, but unlike Professor Allott, I give ‘Woman! when I behold thee flippant, vain’ as a single poem rather than as three sonnets (see p. 562), and ‘God of the meridian’ is treated as a separate poem, not as the last twenty-five lines of ‘Hence Burgundy, Claret, and Port’ (see p. 614). With previous editors, I have followed the earlier manuscript version of La Belle Dame sans Merci, since Keats’s alterations for the Indicator coarsen the delicate texture of the poem. I have also followed the published text of The Eve of St Agnes like all other editors, though Jack Stillinger has argued that Keats’s later alterations, which roused the puritanical objections of his publishers, should be adopted (see pp. 643–4).

In deciding which works to include in the canon, I have sought a critical neutrality. Thus, unlike Professor Allott, but like Garrod, ‘Can death be sleep, when life is but a dream’ and ‘Stay, ruby-breasted warbler, stay’ are given side by side with poems undoubtedly by Keats. Although these are hardly worthy of a major poet, second-rate verse (or worse) is a feature of Keats’s early work, and throughout his career he dashed off indifferent verses with no thought of publication. Badness alone is not a certain sign of Keats’s innocence. Although I have placed the fragment, Gripus, among the Doubtful Attributions since there is no decisive external evidence one way or the other, I think it quite possibly by Keats (see headnote, p. 720).

A very detailed chronology of Keats’s life is supplied by his letters, the transcripts of the poems, H. E. Rollins’s The Keats Circle: Letters and Papers 1816–1879 (1965), and by the biographical work of W. J. Bate and Robert Gittings, among others. I have arranged the poems by their date of completion. The notes aim to give details of each poem’s composition and early publication, to indicate characteristically Keatsian usages, to provide glosses and explanation where necessary, and to record the major critical attitudes to the poetry. Some details of draft readings are also given, though the density of annotation of textual and other matters varies according to the importance of the work – thus the Odes and other major poems are heavily annotated while Otho the Great receives little attention. Classical references, which abound, have been gathered together in a Dictionary of Classical Names (pp. 721–43), which draws on the reference works, particularly Lemprière, known to Keats. To have included this information in the commentary would have involved constant duplication. Allusions are noted, but the frequent echoes of other poets in Keats’s work have been largely ignored in the interests of economy. De Selincourt, C. L. Finney’s The Evolution of Keats’s Poetry (1936) and Miriam Allott’s notes were here especially helpful.

In modernizing the text I have interfered with punctuation and capitalization as little as possible, believing that the early texts present little difficulty of access to the modern reader. Although the punctuation, spelling, and capitalization in Keats’s letters and manuscripts are notoriously irregular, it was common practice for authors at this time to leave such matters to their publishers, and Keats is on record as having asked Taylor to do so. Keats saw the proofs of the three volumes published during his lifetime, and their punctuation follows a loose, but comprehensible pattern.