¹ The text of Book 3 of Irenaeus’s Adversus haereses is quoted here from the edition Contre les hérésies. Livre III (ed. and trans. Adelin Rousseau and Louis Doutreleau; 2 vols.; 2nd ed.; SC 210–11; Paris: Cerf, 2002); I refer to the volumes of this edition as SC 210 (Introduction and notes) and SC 211 (Latin text, Greek back-translation, critical apparatus, and French translation).

² The parallel between Eve and Mary had already been set out by Justin Martyr, Dial. 100.4–6; Irenaeus himself has it again at Haer. 5.19 and Epid. 33.

³ Irenaeus, Haer. 3.22.4, lines 56–72, slightly modified from the ANF translation (The Ante- Nicene Fathers [ed. Alexander Roberts and James Donaldson; 10 vols.; 1885–1887; repr., Peabody, MA: Hendrickson, 1994] 1:455). Latin text: “Maria Virgo obaudiens inuenitur dicens: Ecce ancilla tua, Domine, fiat mihi secundum uerbum tuum. Eua uero inobaudiens: non obaudiuit enim adhuc cum esset uirgo. Quemadmodum illa uirum quidem habens Adam, uirgo tamen adhuc exsistens, … inobaudiens facta, et sibi et uniuerso generi humano causa facta est mortis, sic et Maria habens praedestinatum uirum, et tamen uirgo, obaudiens, et sibi et uniuerso generi humano causa facta est salutis. Et propter hoc lex eam quae desponsata erat uiro, licet uirgo sit adhuc, uxorem eius qui desponsauerat uocat, eam quae est a Maria in Euam recirculationem significans.”

⁴ The bibliography on Irenaeus’s Marian doctrine (or specific aspects of it) is substantial; see, e.g., José Antonio de Aldama, María en la patrística de los siglos I y II (BAC 300; Madrid: Editorial Católica, 1970); Jean Plagnieux, “La doctrine mariale de saint Irénée,” RevScRel 44 (1970) 179–89; Luigi Gambero, Mary and the Fathers of the Church: The Blessed Virgin Mary in Patristic Thought (trans. Thomas Buffer; San Francisco: Ignatius Press, 1999) 51–58. Mary’s role in the process of recapitulation has been discussed by several scholars in the past fifteen years; see, e.g., M. C. Steenberg, “The Role of Mary as Co-recapitulator in St Irenaeus of Lyons,” VC 58 (2004) 117–37; Benjamin H. Dunning, “Virgin Earth, Virgin Birth: Creation, Sexual Difference, and Recapitulation in Irenaeus of Lyons,” JR 89 (2009) 57–88; Maria Del Fiat Miola, “Mary as Un-tier and Tier of Knots: Irenaeus Reinterpreted,” JECS 24 (2016) 337–61.

⁵ I will not consider this aspect here, but it should be remembered that, beside problems of translation, an ancient version reflects a particular state of the text, since the translator uses a particular manuscript and thus depends on the faithfulness of the preceding textual transmission.

⁶ On errors in ancient Latin translations of Greek Fathers, see Sven Lundström, Übersetzungstechnische Untersuchungen auf dem Gebiete der christlichen Latinizät (LUÅ N.F. Avd. 1 51.3; Lund: Gleerup, 1955).

⁷ On this aspect, a systematic study is still lacking, but one can consult Sven Lundström, Studien zur lateinischen Irenäusübersetzung (Lund: Gleerup, 1943) and idem, Neue Studien zur lateinischen Irenäusübersetzung (LUÂ N.F. Avd. 1 44.8; Lund: Gleerup, 1948).

⁸ On this point, the lexicon compiled by Bruno Reynders (Lexique comparé du texte grec et des versions latine, arménienne et syriaque de l’Adversus haereses de Saint Irénée [2 vols.; CSCO 141–42, Subsidia 5–6; Leuven: Durbecq, 1954]) is indispensable.

⁹ The volumes containing these books (SC 211 [2002²], 100** [1965], and 153 [1969], respectively) include a back translation of the parts of the text no longer extant in Greek.

¹⁰ The edition of book 3 by Rousseau and Doutreleau (SC 210–11) has replaced that of François Sagnard (Irenaeus of Lyons, Contre les hérésies. Mise en lumière et réputation de la prétendue “connaissance”, livre III. Texte latin, fragments grecs, introduction, traduction et notes [ed. and trans. François Sagnard; SC 34; Paris: Cerf, 1952]) in the Sources chrétiennes series.

¹¹ See Pierre Nautin, “L’Adversus haereses d’Irénée, livre III. Notes d’exégèse,” RTAM 20 (1953) 185–202, at 197; cf. SC 210: 373–76.

¹² Strangely enough, άναπεριφορά was proposed as a translation for recirculatio (see Sancti Irenaei Episcopi Lugdunensis Libros Quinque Adversus Haereses [ed. W. Wigan Harvey; 2 vols.; Cambridge: Typis academicis, 1857] 2:124 n. 1), but it suits recircumlatio even better; the same is true of “back-reference” as an English equivalent, which is a valid translation.

¹³ See SC 34:380, line 10, and 456; see also Nautin, “L’Adversus haereses,” 197.

¹⁴ A particle seems to be missing to connect this phrase with the preceding one. I have added it in the English translation.

¹⁵ This typographical presentation is inspired by that of Miola, “Mary as Un-tier,” 343.

¹⁶ The fact that many scholars, prior to Sagnard (see below), were silent on the issue is not easy to interpret: did they regard these lines as unproblematic? Did they simply avoid engaging with them? Unlike Massuet (see below), older editors such as Grabe (S. Irenaei Episcopi Lugdunensis Contra Omnes Haereses Libri Quinque [ed. Joann Ernst Grabe; Oxford: E Theatro Sheldoniano, 1702] 261) or Stieren (Sancti Irenaei Episcopi Lugdunensis quae Supersunt Omnia [ed. Adolph Stieren; 2 vols.; Leipzig: Weigel, 1853] 2:545) limit their notes to textual variants. As for Harvey, he provides only a back translation into Greek, without further explanation (Adversus Haereses [ed. Harvey], 2:124 n. 2).

¹⁷ Sancti Irenæi Episcopi Lugdunensis et Martyris, Detectionis et Eversionis Falso Cognominatæ Agnitionis, seu Contra Hæreses Libri Quinque (ed. René Massuet; Paris: Coignard, 1710) 464 n. a (repr. PG 7:959–60 n. 78).

¹⁸ ANF 1:455 n. 6.

¹⁹ “Car on ne peut délier ce qui a été lié qu’en défaisant en sens inverse l’assemblage des nœuds, en sorte que les premiers soient déliés grâce aux seconds, ou qu’en d’autres termes les seconds libèrent les premiers. Il arrive donc que les premiers réseaux soient déliés par les seconds, et que les seconds servent à libérer les premiers” (SC 34:381).

²⁰ See SC 34:379 n. 1: “Marie, nouvelle Ève, remonte jusqu’à Ève (recircumlatio) en dénouant par son obéissance ces mêmes générations nouées par Ève.”

²¹ “Le P. Sagnard a compris à tort qu’il y avait plusieurs nœuds à défaire, comme si Irénée voulait dire que le retour de Marie en Ève se fait à travers toutes les générations intermédiaires, et que Marie délie au passage chacune de ces générations…. En réalité, Irénée parle d’un seul nœud…. Quel est cet unique nœud dont parle Irénée ? Il le dit lui-même en clair un peu plus loin : Sic autem et evae inobaudientiae nodus solutionem accepit per obaudientiam Mariae [lines 88–89]. Le nœud à délier, c’est donc le péché d’Ève (en vertu de la comparaison courante du péché avec un nœud). Pour défaire un nœud, observe Irénée, il faut replier les brins sur eux-mêmes, refaire les mêmes boucles et le même tressage en sens inverse, comme si on faisait le même nœud à l’envers. Ce second nœud, inverse du premier et qui lui sert de dénouement, c’est celui de l’obéissance de Marie…. Irénée n’a pas dans l’esprit l’image d’une ‘remontée’ de Marie jusqu’à Ève à travers toutes les générations, qu’elle délierait l’une après l’autre. C’est à la phrase précédente qu’il a décrit le retour de Marie en Ève…. Maintenant, par la comparaison du nœud, il cherche seulement à nous expliquer pourquoi il était nécessaire que Marie fût ainsi reportée dans la situation d’Ève: de même que pour défaire un nœud il faut un nouveau nœud qui soit inverse du premier, de même pour réparer la désobéissance d’Ève il fallait une nouvelle Ève qui fût inverse de la première, c’est-à-dire obéissante” (Nautin, “L’Adversus haereses,” 198–99). Recently, our passage has been the object of an in-depth study by Miola, “Mary as Un-tier.” In particular, she has sought to place the image of knots in the cultural context of Irenaeus’s time, since knots were much more present in everyday life and symbolically significant than they are today. However, although that article has many useful insights, its main conclusions are doubtful, since they are based on conjectural connections with realia. Since Irenaeus does not indicate that he has a precise kind of knot in mind (and, if so, which), it seems risky to interpret the image beyond the general symbolism of tying and untying.

²² This is true regardless of which verse of the Pentateuch Irenaeus has in mind, and irrespective of whether one opts for recirculatio or recircumlatio, the underlying Greek term, and its meaning.

²³ At first sight, one might consider solving this issue by regarding lines 68–72 as a parenthesis and connecting the clause that begins with quia directly with the Eve-Mary parallel, but this would be no more than a desperate solution. Besides being rather unconvincing in itself, it would leave the problem of the connection with the following lines (see hereinafter) unresolved.

²⁴ According to the standard interpretation, having stated that the second knot (which would represent Mary’s obedience) has served as the undoing of the first (which would represent Eve’s disobedience), Irenaeus would claim that this is why (et propter hoc, line 78) the Lord said that the first would be the last, and the last, the first (Matt 19:30). Since he hardly thought that Jesus was speaking about knots, one might consider that Irenaeus did not pay much attention to the fact that primi and nouissimi are plural and interpreted the reversal of position in Jesus’s saying as applying to Mary and Eve. But this would not solve the problem, rather it would merely displace it: the logical link between Jesus’s saying and the following biblical quotation (Ps 44:17: pro patribus nati sunt tibi filii, line 80–81) would be incomprehensible, since Irenaeus states that the Psalmist says the “same thing” (hoc idem, line 80) as Jesus.

²⁵ As for the Latin text, the only open question is whether the clause that begins with et euenit is part of the subordinate clause (as the punctation in SC 211 suggests) or forms an independent sentence that expresses the same idea (as Sagnard, Nautin [“L’Adversus haereses,” 198], the translation in the same SC volume, and Miola [“Mary as Un-tier,” 343] all imply).

²⁶ Reynders, Lexique comparé, 2:268.

²⁷ Like Rousseau and Doutreleau (who have chosen this verb in their back translation), I regard it as very likely that the underlying Greek verb was συμβαίνω. Admittedly, evenio can translate a variety of Greek verbs, but if one looks closely at the few cases in which the Greek text is available, it appears that only two provide the same construction as in our passage (evenio with an infinitive): one with γίγνομαι (1.3.5, line 86 = Greek fragment 1, l. 337), the other with συμβαίνω. The closest parallel is the latter, while the former is somewhat different (έμοί … μή γένοιτο [Gal 6:14] rendered as mihi … non eveniat).

²⁸ See Reynders, Lexique comparé, 2:307.

²⁹ Haer. 1.pr.1: Έπεί την ἀλήθειαν παραπεμπόμενοί τινες …

³⁰ Haer. 3.22.2 (Greek fragment 34, line 2) (slightly uncertain, since the preceding lines of the Greek text are missing); 4.20.5, end (= Greek fragment 10, lines 11–14).

³¹ See the similar case at the end of 4.20.5 (= Greek fragment 10, lines 11–14). See also, shortly before the passage under examination, 3.22.2, l. 28, which is comparable from the point of view of the syntax, but where έπεί means “for otherwise” (see LSJ s.v. έπεί, B.1). The Latin translator has, incidentally, rendered έπεί there as ceterum.

³² See John D. Denniston, The Greek Particles (2nd ed.; Oxford: Clarendon, 1954) 316–21; Jean Humbert, Syntaxe grecque (3rd ed.; Collection de philologie classique 2; Paris: Klincksieck, 1960) 414 (§ 728); see also Christophe Guignard, La lettre de Julius Africanus à Aristide sur la généalogie du Christ. Analyse de la tradition textuelle, édition, traduction et étude critique (TUGAL 167; Berlin: de Gruyter, 2011) 390 n. 11.

³³ Or, in Greek, a perfect with a present sense. This was probably Rousseau and Doutreleau’s understanding, since they print συμβέβηκε in their back-translation, but translate it as a present (“il se trouve que”), likely inspired by the fact that evenit (as a present) is attested as a rendering of συμβέβηκεν at 1.21.1 (line 2/856).

³⁴ Translation from ANF 1:455 (slightly modified).

³⁵ Or, as Miola, “Mary as Un-tier,” 340 succinctly puts it in the context of Irenaeus’s “recapitulative typology”: “those who are chronologically last have ontological priority.”

³⁶ As Rousseau, SC 210, 362 (n. 2 on p. 427) rightly observes, this section (which in my opinion includes 22.3) is essentially about the recapitulation of Adam; see n. 46 below.

³⁷ Cf. consequenter, line 56.

³⁸ But still without using the word “recapitulation” to describe the relationship between Mary and Eve; only later, at Epid. 33, will Irenaeus explicitly include Mary in the recapitulative process, though in a rather indirect way; see n. 51 below.

³⁹ The emphasis that Irenaeus puts on the fact that Adam himself is among those regenerated (cf. 22.3, in a different context, i.e., that of recapitulation) aims to prepare the reader for the following account of Adam’s salvation (23).

⁴⁰ As stated above, the first quotation (Matt 19:30) still concerns Jesus and Mary, as opposed to Adam and Eve. As for the second (Ps 44:17), Irenaeus says that it conveys the same message (hoc idem significat), yet the Psalmist’s saying is specifically addressed to the Lord (cf. tibi). It thus has a transitional function, while the last two biblical references concern only the Lord.

⁴¹ I.e., the parallel between Eve’s temptation and the Annunciation to Mary and a passage of the Pentateuch (probably Deut 22:23–24).

⁴² Jean Daniélou, From Shadows to Reality. Studies in the Biblical Typology of the Fathers (trans. Wulstan Hibberd; London: Burns & Oates, 1960) 30. French original: “Il s’agit d’un nouveau commencement (κεφαλή) qui est une reprise du premier, à la fois en tant qu’il rétablit l’ordre violé (c’est l’aspect de réparation du péché) et qu’il dépasse l’ébauche commencée (c’est l’aspect d’accomplissement)” (Jean Daniélou, Sacramentum futuri. Étude sur les origines de la typologie biblique [Études de théologie historique; Paris: Beauchesne, 1950] 21).

⁴³ As Irenaeus states, “the second act of tying served as the undoing of the first (secundam … colligationem primae solutionis habere locum).” Therefore, it consists only in untying, without being a knot in its own right. There is no reason to regard it as a true act of tying, as Miola, “Mary as Un-tier,” 345 does—although she rightly observes that “the end result is a slack, knotless rope.”

⁴⁴ That is, her personal and moral role prior to—and in some respect independently of—the virgin birth, which is what she does, and in which she also takes on an essential role thanks to what she is, both as a virgin and as a descendant of Adam (see 21.10–22.1).

⁴⁵ Ireneaus will express the same idea, even more clearly, at 5.19.1 with a formula that seems to be inspired by the image of tying and untying: “quemadmodum adstrictum est morti genus humanum per virginem, solutum est per virginem” (lines 15–16).

⁴⁶ This could include, for example, the interpretation of Sagnard that was rightly criticized by Nautin (see the excerpt quoted above in n. 21) or, along similar lines, that of Orbe, who sees the recirculatio as a stream of life (or of salvation) originating from Mary and extending back through the generations secundum carnem (according to the flesh) (Antonio Orbe, “La ‘recirculación’ de la Virgen María en san Ireneo (Adv. Haer. III,22,4,71),” in La mariologia nella catechesi dei Padri (età prenicena). Convegno di studio e aggiornamento, Facoltà di Lettere Cristiane e Classiche (Pontificium Institutum Altioris Latinitatis), Roma, 18–1- marzo 1988 [ed. Sergio Felici; Biblioteca di scienze religiose 88; Rome: Libreria Ateneo Salesiano, 1989] 101–20, esp. 106). To my mind, Miola also goes too far when, in her recent article, she claims that “Mary is not only part of the ‘omnia’ that Christ restores in the fullness of time, to put it in Pauline terms, but she herself also recapitulates all things in Christ in the fullness of time” (Miola, “Mary as Un-tier,” 341 [italics mine]; cf. 360). We should recall that at Haer. 3.22.4, Irenaeus does not speak (and possibly avoids speaking) of recapitulation when Mary’s role vis-à-vis Eve is the issue, and that, when he eventually does so in Epid. 33, Mary’s part in the recapitulation does not stretch beyond the recapitulation of Eve: “It was necessary/fitting that Adam be recapitulated in Christ, ‘so that mortality might be swallowed up by immortality,’ and Eve [be recapitulated] in Mary, so that the virgin might become the advocate of a virgin and untie virginal disobedience with virginal obedience” (quoted by Miola, ibid.; italics added).

⁴⁷ “Quia quemadmodum per inobaudientiam unius hominis introitum peccatum habuit et per peccatum mors obtinuit, sic et per obaudientiam unius hominis iustitia introducta uitam fructificat his qui olim mortui erant hominibus” (Haer. 3.21.10, lines 216–218, trans. ANF 1:454). Irenaeus expressed a similar thought in Epid. 31; see below and n. 51.

⁴⁸ “Luctatus est enim et uicit; erat enim homo pro patribus certans et per obaudientiam inobaudientiam persoluens; adligauit autem fortem et soluit infirmos et salutem donauit plasmati suo, destruens peccatum” (3.18.6, lines 156–162, trans. ANF 1:447–48). In these lines, Irenaeus is opposing a docetic reading of the Passion.

⁴⁹ Cf. “et sibi et uniuerso generi humano causa facta est salutis” (Haer. 3.22.4, lines 67–68).

⁵⁰ For the parallel between Eve and Mary, see also Haer. 5.19.1, lines 3–5.

⁵¹ Perhaps more precisely “recapitulated”; see the Latin translation of the Armenian text by Adelin Rousseau, (Irenaeus of Lyons, Démonstration de la prédication apostolique [ed. and trans. Adelin Rousseau; SC 406; Paris: Cerf, 1995] 130): “oportebat-et-conveniebat enim recapitulari (άνακεφαλαιόομαι) Adam in Christum.” The recapitulative role conceded to Mary appears more clearly in his translation than in that of Smith: “Car il fallait qu’Adam fût récapitulé dans le Christ … et il fallait qu’Ève le fût aussi en Marie” (ibid., 131; italics mine). The same is true of the translation in Miola, “Mary as Un-tier,” 341, quoted in n. 46 above.

⁵² Irenaeus of Lyons, Proof of The Apostolic Preaching (trans. Joseph P. Smith; ACW 16; Westminster, MD: Newman, 1952) 68–69.

⁵³ On this topic, see, e.g. Joseph Ratzinger, God and the World: A Conversation with Peter Seewald (trans. Henry Taylor; San Francisco: Ignatius Press, 2002) 306.