On all available evidence, the critical-theoretical association of allegory with personified abstractions (in modern understandings, a common and generally unchallenged association) is later than Isidore and indeed later than Tyndale. That is: in literary usage from late antiquity to the early modern age—from Prudentius’ Psychomachia in the late fourth century to the Romance of the Rose and beyond—there are innumerable instances of what is now customarily identified as personification-allegory, as there are in the visual art of (especially) the fifteenth and sixteenth centuries; notwithstanding which, there is no theoretical appeal to any such category until, at the very earliest, the closing decades of the sixteenth century. All we find is a few incidental associations, like one in Dante’s commentary on his own poetry in his “Banquet,” the Convivio (1307), where the poet-commentator presents himself as a lover of “Lady Philosophy.” Dante’s “Banquet” is incontestably allegorical (the “food” is spiritual and intellectual “food”), and in his commentary he duly expounds the poem as an allegory. The key passage here is worth close attention. Dante asks: who is this Lady I am now in love with?— and answers: “that Lady of the intellect called Philosophy . . . . It would be appropriate here . . . to say what this thing, called Philosophy, is—i.e. what this name signifies. And once that has been explained, one can discuss the present allegory more effectively.” So: is what we would call the personification of philosophy itself part of the “allegory”?—is the personification allegorical per se? Answer: no, not quite. What is allegorical, for Dante, are the specific attributes, the particular specifications, of this presentation of philosophy. Consider another passage in the text, where he comments on his characterisation of the Lady as “proud” and “disdainful”:
the true meaning of the poem . . . may be quite easily inferred from [my] literal exposition, except insofar as . . . I called this Lady “proud and disdainful” . . . . [The point is that] to begin with, Philosophy herself seemed to me “proud” . . . because I did not yet understand her persuasions; and “disdainful,” because . . . I could not see her proofs . . . . [So] by this [interpretation] and by what is indicated on the literal level, the allegory of [this] closing stanza is made clear.
The representation of Philosophy as an animate figure is not itself allegorical; rather, it is the attributes—pride and disdain—that are allegorical and that need allegoresis to explain them. And if there seems to be any doubt about this conclusion, consider a different passage in the commentary, where Dante focuses on the way he addresses an inanimate abstraction, philosophy, as if it were an animate she: “to address inanimate things like this is a figure that rhetoricians call prosopopeia, which poets use pretty often.” Prosopopeia, from the Greek prosōpopoiía (via Latin), is the medieval, and ancient—and, still, Renaissance—equivalent of “personification.” No one in antiquity or the Middle Ages ever proposes a theoretical connection between prosopopeia and allegoria, nor does Dante here. For him, the personification is one thing (prosopopeia), and its attributes another (allegoria)—even if some modern commentators on Dante’s text conflate the two by alluding, simply, to the “allegorical figure” of Lady Philosophy. Occasionally, similar moments occur in earlier texts, as when Heraclitus comments on Homer’s personified Éris (“Strife” or “Discord”), or when he discusses Homer’s representation of Átē (“Infatuation”) as a crypto-goddess, “strong and fleet of foot”: “Átē,” says Heraclitus, "he represents as 'strong' . . . because the folly that [the goddess] is associated with is full of irrational impulse. . . . Homer attaches to the names of gods something in our experience, allegorically." Once again, it is the specific detail of the representation that counts as allegory. Like Dante, Heraclitus might well be said to be only a step away from identifying the personification tout court as allegory, but, like Dante, he never takes that step.
No-one, to my knowledge, has ever conducted a close study of when, or how, the connection between personification and allegory is first established. Various medievalists correctly insist that in antiquity and the Middle Ages the connection is not made, and one suggests “no earlier than the Renaissance”—but nothing more specific than that. As a fully operative neutral-theoretical category, no doubt, personification-allegory is definitively institutionalised in the twentieth century, by the likes of Wittkower and Panofsky in art criticism, and Curtius and Auerbach in literary interpretation, while in English studies the influential figure of C. S. Lewis went so far as to equate allegory with personified abstractions, but it is Romantic critics and thinkers at the turn of the nineteenth century who first give personification-allegory a definitive significance, identifying and (like Yeats, in later years) deploring it. On the strength of which (as it seems), one medievalist calls it “really an invention of the nineteenth century.” That, however, is demonstrably wrong—as wrong as reading allegory into characterisations by Dante or Heraclitus. As an identified category, personification-allegory seems to come to the fore in the late seventeenth century. “There is indeed a way of writing purely allegorical, as when vices and virtues are introduced as persons”: Richard Blackmore in 1695. In 1675, John Milton’s nephew Edward Phillips associates with allegory “a personated virtue or vice rising out of the ground and uttering a speech” (though he distinguishes this from “proper allegory”), while in the same year the French critic René le Bossu remarks that “of our very passions we may make so many allegorical deities.”
By the early eighteenth century, the category is repeatedly identified and discussed, in England in particular, and often in connection with a celebrated pair of personified entities in Milton’s Paradise Lost, Sin and Death. In one of a series of essays in The Spectator in 1712, Addison focuses on Milton’s usage:
[Milton] has brought into [his poem] two actors of a shadowy and fictitious nature, in the persons of Sin and Death, by which means he has interwoven in the body of his fable a very beautiful and well-invented allegory. . . .
Virgil has indeed admitted Fame [“Fama,” i.e. “Rumour”] as an actress in the Aeneid. . . [and] we find in mock-heroic poems . . . several allegorical persons of this nature
Three years later, John Hughes declares:
An allegory sometimes, for the sake of the moral sense couched under its fictions, . . . gives life to virtues and vices, passions and diseases, to natural and moral qualities, and represents them acting as divine, human, or infernal persons. . . . [O]f this kind are Sin and Death . . . in Milton, and Fame in Virgil.
In 1717, Francis Atterbury, in a letter to Alexander Pope, commends what is now simply called Milton’s “allegory of Sin and Death,” while a few years after that, Pope himself, in “A Poetical Index to Homer’s Iliad,” appended to his Iliad translation of 1715–20, has a list of “Allegorical or Fictitious Persons in Homer,” which includes figures like “Eris or Discord,” before a separate list of “Characters of the Gods” (Jupiter, Juno, and so on). Then, as late as the 1770s, Samuel Johnson is still exercised by the Miltonic instance: “Milton’s allegory of Sin and Death is undoubtedly faulty.”
Before the late seventeenth century, relevant formulations are few and more elusive. The earliest relevant text I have been able to identify that actually proposes, or at least discusses, a relationship between personification and allegory, albeit very briefly, is an unpublished version of a lecture on Prosopopea by the Italian scholar, Francesco Bonciani, in 1578. In a thoughtful, but also awkward discussion, Bonciani distinguishes prosopopea (as a “figure of thought”) from metaphor (as supposedly a “figure of speech”), and prosopopea from continuous allegory (because the former, but not the latter, keeps the “name of the chief subject” unchanged)—but not before pondering a possible relationship between prosopopea and allegoria, and at one point floating the idea that the two “figures” might be thought to be “the same thing” (“une e medesime cose”), seemingly on the grounds that, in both, “actions may be attributed to inanimate things.” In other sixteenth-century sources, I find no trace of any comparable claim, across a wide range of literary-theoretical texts, from Erasmus to Scaliger to (even) Tasso, later in the century—and “even,” because Tasso’s great epic, Gerusalemme Liberata (Jerusalem Delivered, 1575), is conceived and lengthily defended as allegory by its author. The same goes for the enormous mass of—mostly Italian—art criticism of the century. When art-critics like Vasari discuss “emblems” or “invention” or “fantastic things” (“cose fantastiche”), they frequently have occasion to cite personifications and have ample opportunity to invoke allegory, but, seemingly, never do. They may invoke “symbols and other things that are not human figures,” but without ever relating these to allegory.
Personification-allegory presents an instructive case study. The fact that we now associate the two categories, but they for centuries never did, leads to endless critical slippages and misleading formulations. Take a simple example, involving, again, Dante. In both the Convivio (1307) and then, a decade later, a celebrated letter to Can Grande (c.1316), Dante himself offers theoretical accounts of allegory—in the second case, with reference to the Commedia, and its allegorical “journey” from Hell to Heaven, on which the poet meets a range of historical individuals. “Dante” (says one scholarly reader) “would abandon the abstract allegory of Lady Philosophy which he used in the Convivio and return to an allegory ‘grounded in history.’” How neat!—the semblance of a master-conclusion, validated by historicist evidence: but only the semblance.
On theoretical, and not just historical, grounds, one might well think that the relevance of personification to allegory is in any case open to challenge—and credit here to Bonciani for anticipating at least some of the issues. If allegory involves “saying one thing and signifying another,” does that really fit personification? Is there always, or even usually, “hidden meaning” in personification? For Heraclitus, Homer’s Odysseus is “really” an “instrument of every virtue,” whereas Dante’s Lady Philosophy is really . . . philosophy. Does personification work the same way? But this is not the only question that needs an answer. Alongside the uncharted history of precisely when personification comes to be associated with allegory, another unconsidered issue is: how does it come to be so associated? Is it via the metaphor connection? (Personification, in practice, involves subdued metaphor, while allegory is, or can be, a kind of metaphor itself.) Is it assisted by the fact that like personification, allegory commonly presupposes a sequence, and not just a single word?—a fact trivially concealed in English (but not in many languages) by the orthographical habit of indicating personification by singling out the “name of the chief subject” for initial capital letters. Is it, in effect, by transference of the allegorical quality of the attributes of a personification (as in Dante or Heraclitus) to the personification as such? Or, quite differently, does it arise from responses to the spectrum of pagan divine figures?—from “actual” gods with specific spheres (like Venus), to quasi-divinities with comparable spheres (like Cupid), to abstractions with cult worship (like Concordia, who had a temple in Rome), to abstractions without cult (like Discordia: Homer’s Éris, under a Latin name)? Whatever the answer, it does at least remain beyond doubt that the association with personification cannot be said to be built into the notion of allegory itself, and that the practice of what is now called personification-allegory antedates its theoretical identification by many centuries.
Excerpt from “Invoking the Other: Allegory in Theory, from Demetrius to de Man,” in Allegory Studies: Contemporary Perspectives, ed. Vladimir Brljak (Routledge, forthcoming 2018-19)
[The author’s procedure in this chapter is to provide his own translations of passages from non-English works and, wherever possible, to cite original sources without reference to particular editions. For the reader’s convenience, and by agreement with the author, suggested editions are listed in square brackets in the notes and/or bibliography, along with suggested translation of the non-English works involved. This excludes works available in the Loeb Classical Library. Here and throughout, items in square brackets are editorial additions.]
 At Convivio, 2.1, Dante also offers a formal, if somewhat unorthodox, account of the medieval theory of allegoria as such: cf. e.g. Medieval Literary Theory and Criticism, c.1100–c.1375: The Commentary-Tradition, ed. A. J. Minnis and A. B. Scott with David Wallace, rev. ed. (Oxford, 1911), 382–86.
 Convivio, 3.11: “quella donna de lo ’ntelletto che filosofia si chiama . . . conviensi qui . . . dire che è questo che si chiama filosofia, cioè quello che questo nome significa. E poi dimostrata essa, più efficamente si tratterà la presente allegoria.” The translation above, and (unless otherwise noted) all other quoted translations are my own. [See editorial headnote; for a translation of the Convivio, see n. 6.]
 Convivio, 3.15: “la vera sentenza de la presente canzone . . . per la litterale esposizione assai leggermente qua si può ridurre, salvo in tanto quanto dice che io sì chiamai questa donna ‘fera e disdegnosa’ . . . dal principio essa filosofia pareva a me . . . fiera . . . in quanto le sue persuasioni ancora non intendea; e disdegnosa, chè . . . io non potea vedere le sue dimostrazioni . . . . E per questo, e per quello che ne la sentenza litterale è dato, è manifesto l’allegoria de la tornata.”
 Convivio, 3.9: “è una figura questa, quando a le cose inanimate si parla, che si chiama da li rettorici prosopopeia; e usanla molto spesso li poeti.”
 The French “personnification,” along with related terms, like the English verb “personate” (as in Edward Phillips’s 1675 Theatrum Poetarum: see below), is so used from the seventeenth century.
 So e.g. Lansing, in Dante’s “Il Convivio” (“The Banquet”), trans. Richard H. Lansing (New York, 1900), xv, and Teodolinda Bartolini, Dante’s Poets: Textuality and Truth in the “Comedy” (Princeton, 1984), 21–22.
 Heraclitus, Problems, 29 (Iliad, 4.2–3), and 37 (Iliad, 9.505) [trans. in Heraclitus the Allegorist, Homeric Problems, ed. and trans. Donald A. Russell and David Konstan (Atlanta, 2005)]. Cf., in Byzantine scholarship, Eustathius (twelfth century) on Homeric Éris in his commentary on the Iliad: Commentarii ad Homeri Iliadem pertinentes, ed. Marchinus van der Valk, with indices by Helena Maria Keizer, 5 vols (Leiden, 1971–95), 1.785—where Éris, “touching the sky with her head” (Iliad 4.443), prompts Eustathius to add “allegorically,” likewise.
 Pace Reinhart Hahn, Die Allegorie in der antiken Rhetorik . . . (Tübingen, 1967), 51. In 1295, Dante himself discusses a personification in a different passage—Vita Nuova, 16 (25)—defending his “speaking of Love [Amor] as if it were a thing in itself,” but without any reference to allegory. Minnis and Scott, Medieval Literary Theory, 377, allow themselves the loose claim that here “Dante does discuss the use of personification allegory”; likewise, Frisardi, in Dante Alighieri, Vita Nova, trans. Andrew Frisardi (Evanston, 2012), 222–29. Contrast, rightly: Cecil Grayson, “Dante’s Theory and Practice of Poetry,” in The World of Dante: Essays on Dante and his Times, ed. Cecil Grayson, 146–65 (Oxford, 1980), 151; Zdenko Zlatar, “Allegoresis and the Western Epic Tradition from Homer to Tasso,” Sydney Studies in Society and Culture 10 (1993): 47–180, pp. 61–62. A millennium earlier, in his Types of Style (Perì Ideôn: c.200 AD), 333–34, the Greek rhetorician Hermogenes provides a comparable (or converse) instance, in discussion of stylistic “sweetness,” with reference to Plato, Phaedrus, 230d: “Trees and open country are not willing to teach me anything [oudén m’ ethélei didáskein], but the people in the city are.” Hermogenes comments: “there is a trope [tropḗ] in the phrase ‘are not willing,’ . . . involving metaphor . . . . So one might argue that to say ‘willing to teach me nothing’ . . . in effect allegorises [ēllēgorêsthai].” [Trans. in Hermogenes’ “Types of Style,” trans. Cecil W. Wooten (Chapel Hill, 1987).] We might well decide that the phrase involves personification, and that, therefore, Hermogenes connects personification with allegory—cf. e.g. Véronique Montagne, “La notion de prosopopée au XVIesiècle,” Seizième Siècle 4 (2008): 217–36, pp. 217–22, esp. 220n16—but Hermogenes himself makes no mention of personification here, nor is this actually the kind of usage that ancient theory most commonly connects with prosōpopoiía. Cf. e.g. Quintilian, Institutio, 9.2.29–37, and the representative citations in Heinrich Lausberg, Handbuch der literarischen Rhetorik: Eine Grundlegung der Literaturwissenschaft, 2 vols (Munich, 1960), 1:411–13 [trans. in Handbook of Literary Rhetoric: A Foundation for Literary Study, ed. David E. Orton and R. Dean Anderson, trans. Matthew T. Bliss, Annemiek Jansen, and David E. Orton (Leiden, 1998)].
 As Jean-Claude Fredouille, “Réflexions de Tertullien sur l’allégorie,” in Allégorie des poètes, allégorie des philosophes . . . , ed. Gilbert Dahan and Richard Goulet, 133–48 (Paris, 2005), 147, and Grayson and Zlatar, n. 8 above.
 Jean Pépin, La Tradition de l’allégorie de Philon d’Alexandrie a Dante (Paris, 1987), 255: “non antérieure à la Renaissance.”
 See esp. Rudolf Wittkower, Allegory and the Migration of Symbols (London, 1977); Erwin Panofsky, Studies in Iconology: Humanistic Themes in the Art of the Renaissance (Oxford, 1939); Ernst Robert Curtius, Europäische Literatur und lateinisches Mittelalter (Bern, 1948), trans. in European Literature and the Latin Middle Ages, trans. Willard R. Trask (London, 1953); Eric Auerbach, Mimesis: Dargestellte Wirklichkeit in der abendländischen Literatur (Bern, 1946), trans. in Mimesis: The Representation of Reality in Western Literature, trans. Willard R. Trask (Princeton, 1953).
 The Allegory of Love: A Study in Medieval Tradition (Oxford, 1936), 1, 45, and passim.
 Suzanne Reynolds, Medieval Reading: Grammar, Rhetoric, and the Classical Text (Cambridge, 1996), 135.
 Preface to Prince Arthur [Prince Arthur. An Heroick Poem (London, 1695), sig. b2r]. Here and elsewhere, quotations from premodern English works are given in modernized spelling and orthography, wherever possible.
 Preface to Theatrum Poetarum [Theatrum Poetarum, or A Compleat Collection of the Poets . . . . (London, 1675), sigs **5v–6r].
 Traité du poème epique, 5.3 (tr. “W.J.,” 1695) [René Le Bossu, André Dacier, and Bernard Le Bovier de Fontenelle, Monsieur Bossv’s Treatise of the Epick Poem . . . An Essay upon Satyr, by Monsieur D’Acier; and A Treatise upon Pastorals, by Monsieur Fontanelle, trans. W. J. (London, 1695), sig. P7v].
 The primary passage is Paradise Lost, 2.648–883.
 The Spectator 273, 12 January 1712. “Fama”: Virgil, Aeneid, 4.173–97.
 “An Essay on Allegorical Poetry” (1715) [in The Works of Mr. Edmund Spenser . . . , ed. John Hughes, 6 vols, 1:xxiv–lvii (London, 1715), 1:xli–xlii].
 Letter of 8 November 1717 [The Epistolary Correspondence, Visitation Charges, Speeches, and Miscellanies . . . , 4 vols (London, 1784–87), 1:40]; The Iliad of Homer, trans. Alexander Pope, 6 vols (London, 1715–20).
 Lives of the English Poets, ed. George Birkbeck Hill, 3 vols (Oxford, 1905), 1:185–86.
 Bonciani, “Lettione della prosopopea,” Bibl. Ricc. 1539, fols 132r–44r, at 142r–3r. The idea is floated at 142v, and glossed at 143r: “la figura di prosopopea sarebbe forse diventata allegoria” (meaning extended but discontinuous allegory—“allegoria mista,” 142v—as discussed by Quintilian at Institutio, 8.6.48–49). These comments are absent from the only published version of the lecture, Bibl. Ricc. 2237, printed in Trattati di poetica e retorica del Cinquecento, ed. Bernard Weinberg, 4 vols (Bari, 1970–74), 3:235–53. There is a third—also unpublished—version in Bibl. Ricc. 2942. My thanks to Vanessa Cazzato for help in securing access to Ricc. 1539 itself.
 Thus (e.g.) “Goffredo . . . is none other than the understanding” (“Allegoria del poema,” printed with the 1581 edition of Gerusalemme Liberata) [trans. in “Tasso’s Allegory of the Gerusalemme liberata,” in The Genesis of Tasso’s Narrative Theory . . . , trans. Lawrence F. Rhu, 155–62 (Detroit, 1993)] and “my poem is full of allegorical significations” (letter to Curzio Ardizio, 25 February 1585).
 Annibal Caro, letter “al padre fra Onofrio Panvinio,” 15 May 1565 (“di simboli e d’altre cose che non siano figure umane”) [Lettere familiari, ed. Aulo Greco, 3 vols in 1 (Florence, 1957–61), 3:237].
 One representative instance: Luca Contile, Ragionamento . . . sopra la proprietà delle imprese . . . (Pavia, 1574), 29–30, lists the figurative modes involved in “imprese” (heraldic devices). These include “le metafore” and “le prosopopeie,” alongside such diverse categories as “le parasiopesi” and “le omeosi,” and then “altre ch’ allegoricamente, moralmente, istoricamente, divinamente s’interpretano,” without any phraseological encouragement, or other hint, that “prosopopeie” and “allegoricamente” might be connected or connectible.
 The authorship of Dante’s letter is often (questionably?) disputed; for my purposes, the issue is irrelevant. [Trans. in “The Letter to Can Grande,” in Literary Criticism of Dante Alighieri, ed. and trans. Robert S. Haller, 95–111 (Lincoln, 1973).]
 Mindele Anne Treip, Allegorical Poetics and the Epic: The Renaissance Tradition to “Paradise Lost” (Lexington, 1994), 17, following Charles Singleton, “Commedia”: Elements of Structure (Cambridge, MA, 1954), 93.
 In the sphere of visual representation, compare Antonella Fenech Kroke, Giorgio Vasari: La fabrique de l’allégorie . . . (Florence, 2011), who offers a detailed and sophisticated discussion of sixteenth-century artistic uses and art-critical descriptions of what we would call “personification,” and what we would call “allegory,” but without ever engaging with the crucial issue, and, though well aware of the general absence of relevant sixteenth-century theoretical texts (389), preempts debate at the outset with the pseudo-historicist claim: “La personnification . . . n’est qu’une modalité, une variante de l’allégorie” (2).
 Cf. Bonciani, “Lettione,” fol. 142r, where prosopopea is “figura della sentenza,” and likewise, e.g., Antoine Fouquelin’s Rhétorique Françoise . . . (Paris, 1555), I.B.2.b, where “prosopopée” is “une figure de sentence.” Cf. Montagne, “Notion de prosopopée,” 235.