¹ Milton's Sonnets (London: Macmillan, and New York: St. Martin's, 1966), pp. 62–68. The sonnets are quoted from this text.

² “Implied author” is Wayne Booth's term: See The Rhetoric of Fiction (Chicago: Phoenix Books, 1967), Ch. iii.

³ The chief difference between Poems and MS is that Milton omitted from Poems four sonnets, 15 through 17 and 22, whose politics had become dangerous to him. They are restored to the sequence by John S. Smart (1921) and Honigmann in the order in which they appear in MS. The absence of Sonnet 22 from Poems does not materially affect my argument in this paragraph.

⁴ Translations of Milton's Italian are by Smart, as reprinted in Honigmann.

⁵ For the restricted definition of “persona” which 1 am using here, see Irvin Ehrenpreis, “Personae,” in Literary Meaning and Augustan Values (Charlottesville: Univ. Press of Virginia, 1974), esp. p. 57. Ehrenpreis protests against the fashion of seeing personae indiscriminately, arguing that what is often praised as “artifice” is rather a necessity of expression: “this kind of rhetorical pose is absolutely inseparable from all language and communication. One could never reveal the whole truth about oneself, even supposing one knew it… . One cannot speak without selecting a limited number of remarks from among possible remarks” (p. 51). It is in just this necessity for selection among possible utterances that the concept of “implied author” finds its justification. Although Booth's language (“an ideal … version”) seems to grant the implied author a sort of Platonic excellence, he perhaps means only that good writers use necessity to advantage —in short, select well.

⁶ A sign that Milton wanted to mute the occasions of the sonnets is his treatment of their MS titles: See Honigmann, pp. 71–72, and n. 7 below.

⁷ That Milton did not in 1628 set out to write the series he eventually wrote is not strictly true, for Sonnets 1 through 6 and the Canzone, all composed probably between 1628 and 1630, must surely have been intended as a sequence. Milton first arranged sonnets for print in 1645 (Poems) and in doing so established two principles which he followed more or less closely thereafter. First, the order of the ten sonnets in 1645 is chronological (but whether strictly or loosely so has been debated). Second, four of these sonnets—7 through 10—appear in the Cambridge MS.; here they lack numbers, but 8 and 10 are titled. In 1645 all ten are numbered, and none is titled; with three exceptions, this pattern of converting titles to numbers is maintained in 1673. In composing new sonnets Milton continued the numbering initiated in 1645, but whether higher number always means later date of composition has been often doubted. Since, however, no one doubts that in general the series is chronological, and since my interpretation stands in no need of scholarly latitude regarding individual sonnets, I frankly embrace the straight conservatism of Maurice Kelley (“Milton's Later Sonnets and the Cambridge Manuscript,” Modern Philology, 54. 1956, 20–25), who argues that Milton's MS numbering represents exactly the order in which the later sonnets were composed. Between 1646 and 1652, as Kelley reconstructs it, Milton or his copyist entered Sonnets 11 through 17 and “On the Forcers of Conscience” in what were then two adjacent leaves of the MS. pp. 43-[44] and 47-[48]. These were M ilton's sonnet workshop during those years : The leaves contain nothing but sonnets and drafts of sonnets. We may reasonably imagine, therefore, that as Milton composed new sonnets he had the old not only in view (until his blindness), but also in mind.

In 1655 (Kelley again) Milton began a new transcript of these later sonnets, noting at the top of p. 1 (present p. [46] in MS) that they were to follow the ten in the printed book. Half of this transcript is lost, but Kelley believes it contained all the later sonnets including “Forcers.” The next document is 1673 itself. Here “Forcers” is divorced from the numbered sonnets, 11 and 12 are reversed, and 15 through 17 and 22 are suppressed (see n. 3). Although early MS drafts of Sonnets 11 and 13 through 17 have elaborate titles, the surviving part of the 1655 transcript shows the titles dwindling away, and in 1673 just three sonnets are titled : 12, “On the same” (i.e., as ll); 13. “To Mr. H. Lawes, on his Aires”; and 15 (i.e., 18), “On the late Massacher in Piemont.” I concur with Honigmann (p. 72) in taking these more as footnotes to allusions than as real titles.

From this review it appears that Milton ordered his sonnets chiefly in the process of writing them, that he made most—though not all—of his choices as he went. He may never have asked himself the questions I proceed to ask in this paragraph. But that is only to say that a poet may do by “instinct” things a critic must do by argument.

⁸ Sidney's Poetry: Contexts and Interpretations (1965: rpt. New York: Norton, 1970), p. 133. For a survey of ordering techniques, some of them used perhaps but once, in the very earliest sequences, see William O. Harris, “Early Elizabethan Sonnets in Sequence,” Studies in Philology, 68 (1971), 451–69.

⁹ Milton, Works, ed. Frank A. Patterson, xu (New York: Columbia Univ. Press, 1936), 260. The Holbein woodcut, used more than once as a title-border, allegorically presents the course of human life from birth to redemption ; reproduced in Thomas P. Roche, Jr., The Kindly Flame: A Study of the Third and Fourth Books of Spenser's Faerie Queene (Princeton : Princeton Univ. Press, 1964).

¹⁰ The following description requires many cameo explications of individual sonnets. These are instrumental to its purpose, but, although they contain here and there a new reading, in general I make no claim for their originality. I attend to the parts for the sake of their sum.

¹¹ Latent in these oppositions and in the historical pattern described later is Augustine's dialectic of the Two Cities. I do not think the sonnets are designed to illustrate it; rather, it enters into them as part of their general “set” on the world. But De Civitate Dei xv.4 does have special application to Sonnets 15 and 16, discussed below.

¹² The syntax of Sonnet 16 is described by Taylor Stoehr, “Syntax and Poetic Form in Milton's Sonnets,” English Studies, 45 (1964), 289–301.

¹³ They are four in the many descriptions Northrop Frye has given of Renaissance poetic cosmology. They can also be collapsed into two, as by A. S. P. Woodhouse in “The Argument of Milton's Comus,” University of Toronto Quarterly, II (1941), 46–71, and “Nature and Grace in The Faerie Queene.” ELH. 16 (1949). 194 228. I have gone for my account to two essays by Frye, “Nature and Homer” and “New Directions from Old” (1958 and 1960; rpt. in Fables of Identity, New York: Harcourt, 1963, pp. 39–66), because the fourlevel conception more precisely embraces the sonnets.

¹⁴ The idea of Purgatory as a scene of personal transformation differs from the literal doctrine of Purgatory as the place where souls do penance after death. As a Protestant, Milton of course rejected the doctrine; as a poet, he found Purgatory a useful metaphor of man's aspiration to virtue. For the first point see C. A. Patrides, Milton and the Christian Tradition (Oxford : Clarendon, 1966), p. 264; for the second, Irene Samuel, Dante and Milton (Ithaca: Cornell Univ. Press, 1966), pp. 212, 218. With a change of name, the purgatorial hill can domesticate in Protestant writing: “On a huge hill, / Cragged, and steep, Truth stands, and hee that will / Reach her, about must, and about must goe” (Donne, Satyre iii ) ; “this hill top of sanctity and goodnesse above which there is no higher ascent but to the love of God” (Milton, Church-Government; Works, iii, 261).

¹⁵ Cf. Church-Government: “Nor is there any sociable perfection in this life civill or sacred that can be above discipline, but she is that which with her musicall cords preserves and holds all the parts thereof together” (Works, iii, 185).

¹⁶ Spitzer, “Understanding Milton,” Hopkins Review, 4 (1951), 17–25. Two readings which get the emphases right are Martin Mueller, “The Theme and Imagery of Milton's Last Sonnet,” Archiv, 201 (1964), 267–71, and Marilyn L. Williamson. “A Reading of Milton's Twenty-Third Sonnet,” Milton Studies. 4 (1972), 141–49.

¹⁷ The Poetical Works of John Milton, ed. Henry J. Todd (1809; rpt. New York: AMS, 1970), vi, 438.

¹⁸ An early version of this essay formed part of my dissertation at Rutgers Univ. I acknowledge with pleasure suggestions given me by Carren Kaston, Linda Koenig, and Scott Consigny, and helpful criticism from Irene Samuel and Stanley Fish.