In his Letter to Raleigh, Spenser offers a notoriously frustrating description of The Faerie Queene.¹ Appended to the 1590 edition, and again to the 1609 edition (but omitted in 1596), it appears to be an exegesis of the contents of his epic poem. But the Letter is, as Jan Karel Kouwenhoven says, “bristling with glaringly counter-factual statements” that no ingenuity can (or perhaps should) reconcile with the work itself.² Still, the urge to make use of Spenser’s signposting is strong.

Spenser admits that his work “may happily seeme tedious and confused” (LR 84) because, as he says earlier in the Letter, “the beginning of the whole worke seemeth abrupte and as depending vpon other antecedents” (LR 44-45). It is the Letter itself that causes much of the confusion by implying that some of these antecedents are narrated in the poem; in particular, the Faery Queen’s “Annuall feast” (LR 50-51). As every reader discovers, no account of the feast is offered in the poem—which never gets as far as Book XII, where the feast is promised (LR 50-51). But Spenser offers the time-honored excuse that the methods of a poet differ from those of a historian: the “Historiographer discourseth of affayres orderly as they were donne,” whereas the Poet “thrusteth into the middest” (LR 46-47)—the classical in medias res. This, we think at first, provides the reader with the clue to disentangling parts of the plot that seem “tedious and confused.”

But does it serve the purpose? One reason for doubt is that the in media res explanation comes too late to be of help to the first-time reader—after the text, not before. Another is the reductio ad absurdum: “The beginning therefore of my history, if it were to be told by an Historiographer should be the twelfth booke which is the last” (LR 49-50). This Spenser offers as a corollary of his distinction between the poet and the historian (his “therefore” claims as much). The critic W. J. B. Owen would seem to be justified in complaining “there appears to be no precedent either in practice or theory for placing the preliminaries to an epic action in the last book.”³

Yet it is this conundrum that has potential to open a new line of approach to reading the epic (if it is indeed an epic). Maybe what Spenser is offering, as he elaborates on the distinction to which he has appealed, is a kind of algorithm, in the sense of a compressed instruction manual for how to read. Broadly understood, an algorithm sets out a series of steps by which to address a given task, and solve the conundrums that the task throws up. As information theorist Donald Knuth puts it: “starting from an initial state and initial input…the instructions describe a computation that, when executed, proceeds through a finite number of well-defined successive states, eventually producing ‘output’ and terminating at a final ending state.”⁴ So when Spenser characterizes the poet, in contrast to the historian, as “recoursing to the thinges forepaste, and diuining of thinges to come” (LR 48), and claims that this enables him to make “a pleasing Analysis of all” (LR 49), we could try treating this as an occluded algorithm for arriving at a fuller comprehension of the poem.

The word “recoursing” may be interpreted etymologically as combining “to run,” as in “course,” and “back”: to re-run. Working in the opposite direction, “diuining” suggests anticipating future events. But “recoursing” and “diuining” together may carry a greater load: Spenser could be hinting not just at revisiting or anticipating events, but at a technique more akin to re-enactment, or the traversing of a train of events twice over. He could be hinting at a rule for reading The Faerie Queene that, as poet, he seems also to have deployed himself: to proceed, in Philip Sidney’s phrase, “fore-backwardly.”⁵ In other words, the reader should gradually understand that the poem seems to progress forwards, but is covertly structured to run backwards—or maybe in both directions.

To my delight, I discovered that there is such a thing as a forward-backward algorithm. It involves applying the computation forwards through the material, then backwards; then, by a process of inference known as “smoothing,” melding the two results.⁶ Happily, too, recent critics have prepared the way for the application of this notion to The Faerie Queene. The form of the stanza Spenser invented for his epic, with the rhyme scheme ababbcbcc, has been identified by at least two critics, Jeff Dolven and Clare Kinney, as incorporating a “turn” of some kind. Dolven finds Spenser frequently starting a new and contrary unit of thought in the line of the second b rhyme halfway through his stanza.⁷ He speaks of Spenser constantly “dither[ing] over every threshold, turning forward, then back” (Method 24). Kinney sees a “recursive” pattern, a constant “doubling-back,” the form of the stanza replicating the form of the whole epic.⁸ I argue here that reading the poem aright entails a puzzling journey out, during which everything feels back-to-front, and a more enlightened journey back, when events seem to follow a more logical sequence.

Let us track three imaginary readers, and describe the sense of confusion each of them may feel on first encountering the work. One tackles the 1590 edition, one the 1596, and one the 1609 edition. Their unease about narrative coherence is likely to take a slightly different form in each case.

The first reader finds many specific oddities in her three books. She can explain away the fact that the “Gentle Knight” at the opening of the epic wears armor showing “old dints of deepe woundes,” “The cruell markes of many’ a bloody fielde,” by supposing that this armor is inherited from a battle-worn hero—though Spenser does not say so (I.i.1.1-4). Less explicable is the fact that in the first stanza she learns that “armes till that time did he neuer wield” (I.i.1.5), whereas thirteen lines later, she finds that “nothing did he dread, but euer was ydrad” (I.i.2.9). He seems to instantiate an inherent contradiction—a debutant knight already, and always, “ydrad” [dreaded] by all.

The episode of Redcrosse at Errour’s cave (I.i.11-13) is similarly confusing, as Peter Herman notices in his essay “‘With-hold till further trial’: Spenser’s Letter to Ralegh and Modes of Rereading in the 1590 Faerie Queene.”⁹ When Redcrosse comes to a hollow cave in the woods, Herman points out, he gets off his horse without thinking, and, perhaps even more foolishly, gives “to the Dwarfe a while his needlesse spere” (I.i.11.9). Herman observes that it “is only after he cannot turn back…that Una…tells him where he is,” which is in the Wandering Wood by Error’s den (“With-hold” 199). At this point, the Dwarf utters a warning: “Fly, fly… / this is no place for liuing men” (I.i.13.8-9). Herman puts the whole episode under the rubric of “rereading”—a necessary Protestant skill. As he explains, “Spenser’s technique…echoes the Protestant reevaluation of individual authority in reading and rereading the Bible” (205), in contrast to the Catholics’ seeking for guidance from ecclesiastical authority. But the narrative could equally be clarified by “reading backwards,” as will be tested below.

More causes for puzzlement present themselves to the 1590 reader as she reads on. The seven deadly sins in Book I.iv.18-37, she notices, are in reverse order: Sloth leads the way; Lucifera, representing Pride, comes last. Then there’s the wood-cut of St. George and the Dragon which comes at the end of Book I, though it might have been more appropriate on the page before or even during their deadly battle (I.ix). There often appear characters labelled simply “she,” “he,” or “that man,” remaining nameless until much later. One instance of this befuddling phenomenon is the case of Amavia, who starts as a mere shriek (II.i.35); progresses to a vague “she” in the next stanza; and is not named until II.ii.45, where her story ends. Another is Sir Ferraugh, who appears as “An armed knight” at III.viii.15.3, but is not named until the next book (IV.ii.4.5). And sometimes a character’s entire story is twisted back to front: Kinney observes that Amoret’s rescue occurs in Book III, but her original wooing is in Book IV (“Romance” 125).

With Book IV, we have started to encroach on the journey of the reader of the 1596 Faerie Queene. His concern is more with the overall schema. He feels that there is something back-to-front about the progression from the romance atmosphere of Book IV to the martial feats of Book V to the pastoral of Book VI. But this benighted reader has no appended Letter with the algorithm to look to for instruction.

The third reader is likely to be baffled from the outset of Book VII, the “Mutabilitie Cantos,” comprising only a Canto VI, a Canto VII and two stanzas of a Canto VIII. These may be hedged about with the publisher’s hesitant disclaimer that they “appeare to be parcell of some following Booke of the FAERIE QUEENE” (my italics), but this hardly assuages the puzzlement. She was expecting the seventh book to follow a similar pattern to the previous ones. Yet though the publisher has placed this seventh book “under the legend of Constancie” on its own title page, no knight appears as champion of Constancy. It is left to Cynthia who, as Mutabilitie points out, is as “changefull as the Moone” (VII.vii.50.9), to perform this function.

Further on in the fragmentary book, this reader pauses at the invocation to the greater Muse in the first two stanzas of Canto VII. This is an odd place to make such a plea, she thinks. Every successive title page of the work announced twelve books, so an appeal for a second wind immediately after the completion of six books would be comprehensible. But halfway through Book VII, it seems out of place. She also finds perplexing the word “turne” in the poet’s plea for new inspiration “fit for this turne” (VII.vii.2.3).

As she reads on, she is strongly reminded of Virgil’s Georgics, with the Parade of the Seasons, month by month (VII, stanzas 28 to 43). Reaching the last line of the whole “epic,” she discovers there is still more text—the Letter to Raleigh, dedicatory poems to Spenser, and dedicatory poems by Spenser to various lords and ladies. Belatedly, the Letter prompts her to adopt a strategy of “recoursing” (LR 48). She grasps the fact that it is she who must “turn”: she must turn back and re-traverse the whole poem.

The present-day reader of this essay will immediately see what her backward reading accomplishes. The paratext now comes first, with poems followed by an epistle, as is customary. The “Mutabilitie Cantos” appear in the light of a Council of the Gods, a fit beginning for an epic, in the tradition of the Odyssey, the Aeneid, or Lydgate’s Assembly of the Gods.¹⁰ Pastoral and Georgic (not quite in the right order) anticipate Epic, consonant with the Virgilian schema of a poet’s career.¹¹ The vague deictic references—“that man” and so on—are now clear, since their referents have been identified by name earlier. Amoret’s adventures occur in their chronological order. The seven deadly sins resume the correct hierarchical order. The woodcut of St. George arouses a pleasurable expectation of his battle. At Errour’s cave, the sequence becomes coherent: the dwarf issues a warning first, then Una explains why, but Redcrosse still advances too far without his horse or spear. Finally, the Knight at the beginning (end) understandably looks somewhat battered: he has completed his battles by what looks like the start of the work. The fact that he is both “full iolly” in the first stanza (I.i.1.8), and “solemne sad” by the second (I.i.2.8), is the last pitfall of this labyrinthine journey. Split vision is needed to see him as both a tyro and a battle-worn hero: the back-to-front evolves into the out-and-back, in accordance with the ever-rewarding forward-backward algorithm.

The Elizabethans were perfectly familiar with the notion of the back-to-front. They labeled it the “prae-posterous,” but also knew it in its original Greek—hysteron proteron, meaning “the later first.”¹² Patricia Parker in her chapter on the “hysteron proteron” in Renaissance Figures of Speech cites the definition of Richard Huloet (1552): the “preposterous” is the “out of order, overthwarth, transverted, or last done which by rule have ben first.”¹³ Parker points out that George Puttenham in his Arte of English Poesie (1589) “used ‘Preposterous’ itself as his formal English equivalent for this Greek rhetorical term, ranging ‘Histeron proteron, or the Preposterous’ under ‘Figures Auricular working by disorder’” (“Hysteron” 133). Simultaneously, “preposterous” had the connotation “monstrous,” or “ridiculous,” as it does today.¹⁴

Thomas Nashe’s sending up of Spenser’s sequencing of his poem and its paratextual material is a clear demonstration of the wide-spread understanding of the two meanings. Nashe also provides potential evidence that the forward-backward trajectory of The Faerie Queene was an integral feature of Spenser’s project. In Pierce Penilesse His Supplication to the Divell (1592), Nashe delays his address to his readers until almost the end of his pamphlet, making an imaginary reader expostulate “What? an Epistle to the Readers in the end of thy booke?”¹⁵ Within a page or two, he proceeds to apostrophize Spenser and point out that he (Nashe) likewise had to leap to “the latter end” of The Faerie Queene to find any dedications—“short lynes, to sundry Nobles pend” (Penilesse L4^v). He immediately concedes, however, that he may not have grasped Spenser’s purpose.

This invoking of Spenser’s intentions may indicate that Spenser himself was responsible for the unusual placing of The Faerie Queene’s paratext (though the possibility remains open that the decision was that of the publisher of each edition). What is clear is that the terminal position of the Letter and dedicatory sonnets appeared to some contemporaries as peculiar, “preposterous” in both senses. Jean Brink insists that there is nothing untoward about their position, on the grounds that other publications of the period had terminal dedications.¹⁶ But Andrew Zurcher, in his forensic analysis of Penilesse’s parodic and repeated echoing of The Faerie Queene, argues convincingly that “Pierce” (Nashe) is mocking Spenser, albeit playfully, for this placing of his dedicatory material last.¹⁷ Zurcher highlights the preface to the pamphlet by Nashe’s publisher Richard Jhones (“Getting It” 180). In remarking on his own author’s unwonted placement of his Epistle, Jhones, who is obviously in on the joke, directs attention to the point in the pamphlet where Nashe remarks on the anomalous position of Spenser’s.¹⁸ Jhones actually uses the word “preposterous” as he does so:

...which Title [Pierce Penilesse] though it may seeme
strange, and in it selfe somewhat preposterous, yet if you vouchsafe the
Reading, you shall finde reason, aswell for the Authours uncouth
nomination, as for his unwonted beginning without Epistle, Proeme, or
Dedication: al which he hath inserted conceitedly in the matter; but Ile
be no blab to tell you in what place. 

(Pierce Penilesse, A2r)

Further external evidence is available in Spenser’s and Gabriel Harvey’s correspondence of 1580. In June 1580, the book-seller Henry Bynneman published their three “Familiar” letters and two “Commendable” letters, the former dated 1580, the latter, October 1579.¹⁹ The letters dated 1579 have been appended, rather than placed at the front, because (we are told on their separate title page) they were “More lately deliuered unto the Printer” (633). In other words, what we read is back-to-front, or at least chronologically jumbled: the 1580 ones as printed proceed from April 12, to April 7 to a date in May labeled “Nono Calendas” (630); the 1579 ones go from October 15/16 with an attachment dated October 5, to October 23. And yet the impression one has in reading them is that they follow on in time quite logically one from the other. We are informed we are reading backwards, but feel we are going forwards.

That the Letters do proceed very “fore-backwardly,” however, is indicated by the fact that the Commendable Letters end with fervent farewells between the two men, seemingly providing a fitting closure to the correspondence. Spenser says he is on the point of going abroad within a week of October 5, though the “Harvey” of the exchange expresses his doubts about this in the Letter of October 23. Sure enough, the Familiar Letters show Spenser still in London in April 1580 (612).

The two correspondents make equally confusing remarks regarding The Faerie Queene. On “Quarto Nonas Aprilis 1580” (612), Spenser asks Harvey to send the poem back to him with his judgement of it. Yet on October 15 of the year before, he speaks of some work Harvey has commented on, which he is still hesitating to publish: “The selfe former Title stil liketh me well ynough, and your fine Addition no lesse” (635). It is ostensibly during the following May that Harvey gives an extremely adverse verdict on The Faerie Queene, calling it the “Elvish Queene” (628). The “fine Addition” sounds suspiciously like the mocking title applied by Harvey to the poem. Since the other works for which Spenser had asked for a critical response seem already to have been completed and ready to publish, according to the first Familiar Letter (612), it does seem likely that the work still under discussion is The Faerie Queene, despite the incompatibility of this conclusion with the chronology being foisted upon us. Weaving a forward-backward path, it might be possible to disentangle the chronology.

If Spenser was being as disingenuous with his epic poem and its paratexts as this analysis suggests, it would not be alien to the period. Louise Wilson has shown how a romance author such as Anthony Munday “participate[s] in an intricate, tongue-in-cheek rebuttal of accusations of the unworthiness of the genre.”²⁰ “These pseudo-humanist preliminaries…expose a playful attitude to the function of the peritext,” she writes (“Playful Paratexts” 122). In the same spirit, Spenser mischievously echoes contemporary Italian theorizing about romance. The critic J. W. Draper shows how familiar Spenser was with this, citing his parroting of the Italian theorist Minturno, who in the course of a mere two pages of his Arte Poetica, deals with “starting in the middle or at the end,” “episodes” as opposed to main plot (Spenser’s “Accidents” [LR 79]), and the method of historiographers as opposed to poets.²¹ And significantly, Draper notes, Minturno “declares unequivocally that an epic ought to begin with the ‘last things’—a statement that might be taken to imply its converse, that the poem should end with the first” (“Narrative Technique” 321-22). Spenser’s friend Lodowick Bryskett employs similar concepts and vocabulary in his Discourse of Civill Life:²²

The end in all things that men do in this world, is the first that is
considered, though afterwards it be the last to be put in execution. And
as, when it is brought to perfection, it beareth the name of effect, so
is it the cause that moveth all other to bring it to effect. 

(Discourse E4)

This passage follows a scene in which Bryskett gives his readers a vignette of Spenser in Ireland, reluctant to read his Faerie Queene aloud in company (E1^r- E2^v), which makes it all the more likely that where Spenser speaks of “the generall end…of all the booke” in his Letter (LR 7-8), he too is playing with two senses of “end,” namely “terminus” and “aim.”

The stage-by-stage application to The Faerie Queene of the forward-backward algorithm has offered ways to make sense of some of the poem’s seeming illogicalities. The structure leaps into focus when seen from what looks like the end: dedications, explanatory Letter, council of the gods, georgic, pastoral, martial, and romance episodes occur in (almost) due order. The ricorso trajectory through the adventures, forwards in mystification, backwards with more comprehension, appears to achieve a composite or synoptic vision of the whole—something like the “smoothing” of the information theorists. I am probably stretching their technical terms. Spenser could not have had the modern understanding of “algorithm” in his mind when composing. Nor could he have known Giambattista Vico’s theory of the ricorso.²³ Yet, run as a jeu d’esprit applied to what may be a jeu d’esprit, the theory appears surprisingly fitting.

However, it brings in its train some dangers of its own. Contrary to the spirit of recent critiques such as those of Dolven, Catherine Nicholson, and Gordon Teskey, it tends to unify the work, and to impose one kind of reading—that of the skeptic on a quest for a stable logic. Dolven stresses the poem’s “unmaking” of itself, and how “the poem’s ambition for totality and wholeness might properly be suspended.”²⁴ Nicholson makes the case for the virtues of piecemeal readings (since it is common practice to read only one section, book or adventure at a time), and against the disqualification of any line of interpretation, however wild.²⁵ Teskey feels that the poem succeeds musically and thematically, but argues that “architectonically…The Faerie Queene stands before us as a ruin.”²⁶ My inverted reading is of its nature more deterministic and my proffered explanations sound more positivistic than these. One startling conclusion, for example, is that there are no missing books: contra Teskey, the poem’s architectonike is both coherent, and achieved. The algorithm has effected its smoothing, resulting in output that does indeed “terminate at a final ending state”—the beginning (Knuth Fundamental Algorithms, 5).

The fluid, unsettled readings outlined in the preceding paragraph are attractive. Yet I would like to see future critics imitate the forward-backward reader of the poem, and retrace their steps right back to Harvey’s adverse reaction to the draft he had seen in 1579/80. In the third Familiar Letter, he compares the work to the Italian satires of Bibiena, Machiavel, and Aretine and the comedies of Aristophanes and Menander, Plautus and Terence, rather than to the romance of Ariosto, Spenser’s avowed model (628). “Hobgoblin runne away with the Garland from Apollo” (628) is his final verdict. In other words, as a pretended epic poem, the work is preposterous.