Chapter 2
Seven commitments for the long road that follows
Two patristic scholars sit down at the same desk, open the same Greek texts, work through the same passages of Origen, Gregory of Nyssa, and Isaac of Nineveh. They come out with mutually exclusive conclusions about what those passages mean.
This is not a thought experiment. In 2013, Ilaria Ramelli published The Christian Doctrine of Apokatastasis: A Critical Assessment from the New Testament to Eriugena with Brill — nearly nine hundred pages of dense argument that universal restoration was a major strand of patristic theology, taught with confidence by Origen, Gregory of Nyssa, Maximus the Confessor, Eriugena, and a long list of less famous figures.1 Five years later, Michael McClymond's The Devil's Redemption: A New History and Interpretation of Christian Universalism appeared from Baker Academic — roughly fifteen hundred pages across two volumes, engaging many of the same primary texts but reaching a very different verdict.2 Where Ramelli sees a thriving universalist mainstream that the Augustinian West later marginalized, McClymond sees a recurring heterodox impulse the Church wisely declined to endorse.
Same texts. Same languages. Same scholarly tools. Different answers.
How does this happen? The short answer is that historical theology is much harder than it looks. Reading Christians across two thousand years is not like reading the morning newspaper. The texts are in dead languages. The conversational background is buried. The categories the figures used are not quite our categories. Their controversies were aimed at opponents whose books we sometimes barely have. Their key terms — aiōn, aiōnios, apokatastasis, gehenna, hades, misericordes — sit at the intersection of philosophical vocabulary, biblical translation history, and pastoral practice. Lift a sentence out of context and you can make almost any patristic figure sound like he taught almost anything.
So before walking into the historical material, I want to lay my own methodological cards on the table. The seven commitments that follow are the rules I have tried to follow in writing every chapter of this book. They are not original. They are the working methodology of anyone who has been reading Christian thought across centuries for very long. But naming them up front lets you hold me accountable — and lets you see, when scholars like Ramelli and McClymond reach opposite verdicts on the same texts, where their methods most likely diverged.
Most Christians who can tell you what "Augustine taught about hell" learned it not from Augustine but from a textbook, a sermon, or a study Bible. The same goes for Origen on apokatastasis, Calvin on the descent of Christ, and Wesley on the wider hope. This is normal and largely unavoidable — life is short, the originals are long, and we all stand on the shoulders of teachers. But the result is a chain of citation in which the original text grows dimmer with each retelling.
The chain can break in either direction. A confident summary of Origen on the salvation of the devil may turn out to be a paraphrase of a paraphrase whose first link was already a polemical caricature. Or it may turn out to be exactly right. You cannot know without reading Origen.3 When I am told that Augustine "settled" the question of postmortem mercy, the responsible move is to open the Enchiridion and the City of God and see what Augustine actually wrote — and what he was responding to.
This book commits to going back to the texts. Every chapter on a major figure will contain substantial direct quotations, with citations the reader can check. When I summarize, I will note that I am summarizing. When I paraphrase, the original is one footnote away. The goal is not to claim that primary-source reading is easy — it isn't — but to make my readings auditable. If you think I have got Augustine wrong, you can find out by reading the same pages I read.
Modern categories are useful. "Inclusivism," "exclusivism," "annihilationism," "conditional immortality," "universalism," "postmortem opportunity" — these terms let a contemporary reader compare positions across centuries without inventing new vocabulary every time. I will use them throughout this book.
But they were not the categories Origen or Aquinas or Calvin used. When we project our terms backward, we sometimes find figures conveniently teaching exactly what we expected — because we framed the question in our terms before reading them. Was Gregory of Nyssa an "inclusivist"? The honest answer is that Gregory was not asking the modern Vatican II question about the salvation of non-Christians. He was asking whether God's pedagogical love would, in the end, succeed in restoring all rational creatures to communion with their Maker. The two questions overlap, but they are not the same question.
Jaroslav Pelikan put the warning sharply in his classic introduction to The Emergence of the Catholic Tradition: the historian's first task is to recover what the figure thought he was doing, before asking what we want him to have done.4 Carlos Eire makes a similar point about reading premodern Christians on eternity — that what we today call "the doctrine of hell" or "the doctrine of heaven" is itself an artifact of a particular history, not a stable category we can simply impose on Origen or Aquinas.5
So in this book, when I apply a modern category to a premodern figure — and sometimes I must, for clarity — I will flag what I am doing. Gregory of Nyssa was a universalist in something close to the modern sense. Origen was a universalist in a different and more philosophically loaded sense. Augustine was not. Aquinas was something more complicated. The labels are useful shorthand; they are not magic spells that conjure premodern figures into modern parties.
Reception history is its own discipline. "What Augustine taught" depends partly on whether you mean the young Augustine of the early 390s, the mature Augustine of the 410s and 420s, the Augustinian tradition of the seventh century, the high-medieval Augustinianism that gave Aquinas one of his syntheses, or the Calvinist Augustinianism that came to dominate the Reformed tradition. The same goes for Origen, whose reception history is so contested that "Origenism" can mean four or five different things depending on which century and which controversy is in view.6
This is one of the places where popular polemic most often goes wrong. A preacher confidently quotes "Augustine" against postmortem mercy. The quoted view is a real position — held by Augustine in a real text. But it is sometimes held more tentatively by Augustine than by the Augustinian tradition that grew up around him. The mature Augustine of the Enchiridion was, in his own context, arguing against fellow Christians, many of them, who held softer views. The Augustinian tradition of the next thousand years often forgot that those fellow Christians had ever existed.
The same problem cuts the other way. A contemporary advocate of universal restoration sometimes quotes "Origen" in defense of a view that the Origen of the Commentary on Matthew or the homilies on Jeremiah was, in fact, more reticent about. Chapter 8 takes this question apart at length. The point here is just that "what Origen taught" and "what later Origenism became" are two different historical inquiries, and they need to be done separately before being put back together.
One of the earliest models of this kind of careful reading is John Cassian, the fifth-century monk whose Conferences drew deeply on Evagrian and Origenist patterns of spiritual theology while declining to commit to the more speculative Origenist eschatology. Cassian appropriated what was edifying and left aside what was contested — and he did this before the sixth-century controversies hardened the lines. His example is worth keeping in mind whenever a later writer claims to read Origen straight off the page.14
Origen's exegesis of the Sermon on the Mount in his Commentary on Matthew is not on the same level, in his own mind, as his speculative reflections about pre-existent souls in De Principiis. He says as much, more than once. He is willing to be confidently dogmatic about what the Rule of Faith requires. He is exploratory and tentative when he turns to questions on which the Rule of Faith leaves room for inquiry.7
Augustine is similar. His confident, settled, central teaching on grace and predestination is one thing. His polemical responses to particular opponents in particular years are another. His pastoral letters of consolation, written for grieving Christians, are yet another. Reading all three as if they were a single doctrinal statement misses how Augustine himself thought about the levels of his own teaching.
I will try to mark these differences as I go. When Origen is reflecting on a question he himself flags as open, I will say so. When Augustine is in heated polemic against a specific opponent, I will name the opponent and notice that the heat of the dispute may be shaping the formulation. When Calvin is responding to Anabaptist soul-sleep, I will note the polemical context. The goal is not to relativize anyone's teaching but to read each figure at the level he intended to be read.
Some historical questions have settled answers. Gregory of Nyssa taught universal restoration. No serious scholar of Gregory disputes this; it is simply what the texts say. Augustine taught the eternal conscious torment of the lost. No serious Augustinian scholar disputes that either; it is the plain meaning of book 21 of the City of God.
Other questions remain genuinely contested. Did Origen teach the eventual salvation of the devil? The Latin De Principiis as Rufinus translated it points one way; the Greek fragments preserved by hostile witnesses point another. Did Augustine ever waver on the eternity of hell, in an early work or a private letter? A handful of passages have been read both ways. What does Maximus the Confessor's quiet appropriation of Origenist categories amount to? Brilliant scholars have read Maximus in opposite ways on this question, and the dispute is not silly.
Where the scholarship is genuinely divided, I will name the disputants and let the reader follow the argument. Ramelli is one of the most learned recent advocates of a robust universalist reading of the patristic record; McClymond is one of the most learned recent skeptics. Brian Daley's Hope of the Early Church tries to hold the middle, calling some figures hopeful and others explicit and acknowledging that the line between is sometimes blurry.8 I will draw on all three, and several others, without pretending that the question is settled where it is not.
Patristic prayers for the dead are not the same kind of speech as the Summa Theologiae. The Dies Irae's plea on behalf of the trembling sinner is a different act than a doctrinal treatise. The prayers of the Saints for the lost — Isaac of Nineveh's longing prayers, the prayer of Macrina in Gregory of Nyssa's dialogue, the prayer of the desert mothers and fathers — sit in a different register than the polemical works that come down to us with the same authors' names attached.
Treating devotional speech as if it were dogmatic confession is a category error in either direction. It is not honest, on the one hand, to read every patristic prayer for the dead as covert universalism. It is also not honest, on the other hand, to dismiss the longing of saintly people for the salvation of the lost as merely "sentimental" speech that "doesn't count" against the systematic teaching they were also producing. These prayers were prayed by the same people who wrote the treatises. They tell us something about how those people held their doctrines — what they hoped for, what they feared, what they expected God to do.
I will try to handle each kind of speech in its own register. A patristic prayer that pleads for mercy on the departed unbeliever is evidence of something — but evidence of what, exactly, requires care. The same is true of the great medieval intercessory liturgies, the prayers of the eastern Saints, and the hymnody of the Wesleyan revival.
The last commitment is the hardest. There are two ways to misrepresent a historical figure, and they are mirror images of each other.
The first is apologetic distortion. This happens when an interpreter, finding a figure who seems to support his position, makes the figure say more than he actually said. Origen is more confident in his apokatastasis than the texts will bear. Wesley is more universalist than his sermons will bear. Barth is openly universalist when the careful passages in Church Dogmatics II/2 are more careful than that. The wish to find a friend in the past leads us to tidy up the friend's actual record.
The second is polemical distortion. This is the inverse: a figure whose views one dislikes is read as more extreme, more incoherent, or more obviously wrong than the texts actually warrant. Augustine becomes a cartoonish monster of double predestination. Calvin becomes a heartless determinist with no pastoral instincts. The medieval doctrine of purgatory becomes a tale of moneygrubbing indulgence-sellers. None of these caricatures survives careful reading of the actual texts.
I want to be charitable equally to figures I find theologically attractive (Origen, Gregory of Nyssa, Isaac of Nineveh, Wesley, MacDonald, Barth) and to figures I find theologically harder to embrace (Augustine on hell, Calvin on double predestination, the strict scholastic Augustinianism that came to dominate the Reformed tradition). The same care is owed to all of them. Chapter 12 will read Augustine as fairly as I can manage. Chapter 8 will read Origen the same way. The reader is entitled to assume that the same standards apply across the board.
Methodological commitments are easier to state than to apply. Here are four places in the historical record where these rules make a concrete difference — places this book will return to in later chapters, but worth previewing here so the reader can see what disciplined reading actually looks like in practice.
The notoriously difficult passage about Christ "preaching to the spirits in prison" became, in the second through fifth centuries, a fixed point in Christian discussion of what happened to the righteous and unrighteous dead between the death of Christ and the Last Day. The Apostolic Fathers, the Apologists, Clement of Alexandria, Origen, the descent hymns of the Syriac tradition, and the eastern liturgies all return to this passage.
The temptation today is to retroject either modern postmortem-opportunity arguments or modern arguments against PO into the patristic descent tradition. Both retrojections distort the record. The Fathers were not, by and large, debating "postmortem opportunity" in our sense. They were working out what the descent meant within a shared assumption that the dead in Hades were conscious, that Christ encountered them, and that something happened in that encounter that mattered for the cosmic scope of redemption. Some Fathers go further than others. Some explicitly affirm postmortem conversion in the descent narrative. Others stop short. Chapter 4 and Chapter 6 will take the descent tradition apart text by text. The point here is that reading the early Fathers as either contemporary PO advocates or contemporary PO opponents misses what they were actually doing.
In Enchiridion 112, near the end of his great compact handbook of faith, Augustine writes what may be the most quoted line in the history of the eternal-punishment debate:
"It is in vain, then, that some — indeed very many — make moan over the eternal punishment, and perpetual, unintermitted torments of the lost, and say they do not believe it shall be so; not, indeed, that they directly oppose themselves to Holy Scripture, but, at the suggestion of their own feelings, they soften down everything that seems hard, and give a milder turn to statements which they think are rather designed to terrify than to be received as literally true."
— Augustine, Enchiridion 112 (trans. NPNF, 1st ser., vol. 3)Augustinians have cited this passage for fifteen hundred years as proof that Augustine demolished postmortem hope. Read carefully, the passage says something more interesting than that.9
Augustine is not arguing against pagans. He is arguing against fellow Christians. By his own account there are "very many" of them. They are not in his judgment heretics — he does not call them so; he describes their motive as misericordia, mercy. In the City of God 21.17–27, he expands the catalogue: there are Christians who believe all the baptized will be saved, Christians who believe all who participate in the Eucharist will be saved, Christians who believe the prayers of the saints can rescue any human soul, Christians who believe punishment is purgatorial rather than perpetual, and so on.10 Augustine refutes them at length. But the very existence of the catalogue is evidence for what this book will document at length: in the late fourth and early fifth centuries, what we now call "the historic Christian view" was a contested view. Augustine had to argue for it, against many fellow Christians, because it was not yet the settled consensus the later Augustinian tradition assumed it had always been. Chapter 12 will work through the entire catalogue with care. For now: notice how much you can miss by reading only the polemical conclusion and not the inventory of opponents.
Popular evangelical citation runs like this: "an ecumenical council in 553 definitively condemned universalism, so the Church has always rejected it." This is a sentence I have heard from pulpits and seen in study Bible footnotes more times than I can count. It is also a sentence whose every clause is more complicated than the speaker usually realizes.
Was there a council in 553? Yes — the Second Council of Constantinople. Did it condemn universalism? It condemned a list of fifteen anti-Origenist propositions whose precise text and conciliar status are debated. Were those propositions equivalent to "universal restoration" as such, or to a specific sixth-century Origenist construction associated with Evagrius Ponticus and certain Palestinian monks? The latter, in the careful judgment of most recent scholarship.11 Did the council's conciliar authority extend to those anathemas, or were they appended to the council's main acts by Justinian's prior synod of 543? That is also debated.12
None of this proves that universalism is correct. It does, however, demolish the rhetorical move that "an ecumenical council settled this." The move is too quick. Chapter 11 will go through the historical record carefully. The methodological point here is simple: a confident appeal to conciliar authority that depends on more historical claims than the speaker realizes is precisely the kind of move our seven commitments are designed to slow down.
Calvin's reading of the Apostles' Creed's "He descended into hell" in Institutes 2.16.8–12 is one of the most debated passages in Reformed theology. Calvin reinterprets the descent as a description of Christ's experience of divine abandonment on the cross — a powerful and pastorally rich reading that, however, departs from a long Christian tradition that had read the descent more literally.13
Did Calvin's metaphorical reading leave any room for an actual encounter of Christ with the dead? Did he, by reinterpreting the descent, intend to foreclose any patristic-style PO reading of 1 Peter 3? Or was he simply addressing the specific pastoral question of Christ's experience on the cross, leaving the broader descent tradition alone? Reformed commentators have split on this. Chapter 15 will return to it. The methodological point here is that Calvin's intent has to be reconstructed carefully from the texts themselves, not assumed from what the later Reformed tradition concluded he must have meant.
One more methodological observation before turning the page. As I mentioned in the introduction, I write from a particular theological position — conditional immortality, with a postmortem opportunity for those who never had a clear chance in this life, and an honest openness to the possibility that the universalists may yet turn out to be right about the very end. My position will sometimes show in the framing of the chapters and in the selection of which voices receive sustained attention.
But the chapters themselves are descriptive, not prescriptive. When Origen taught the eventual restoration of the devil, that is what Origen taught — not a softer position I might prefer to find. When Augustine taught the eternal conscious torment of the lost, that is what Augustine taught — not a more hopeful position later interpreters have wished for. The book's contribution, if it has one, is the honest map. Faithful description must come before any theological assessment.
The seven commitments above are tools. They do not guarantee right answers. Ramelli and McClymond both believe themselves to be following something like these rules, and they still reach opposite verdicts. Method is not magic.
But method does something important. It exposes the joints of the argument. When two careful readers reach different conclusions, the method helps us see where they diverged — at which inference, on which translation, in which assessment of contested scholarship. The reader of this book may, in the end, reach conclusions different from mine. That is fine. What I hope to give the reader is not my conclusions but the actual evidence, presented as fairly as I can manage, with the joints visible.
The deeper hope is that this book will model a kind of historical theology that the contemporary church badly needs. Too much of our public eschatological argument runs on confident assertions about "what the Church has always taught" without any actual engagement with what the Church has actually said. Too much of it traffics in inherited summaries of figures the speaker has never read. Too much of it treats one of the most disputed questions in the history of Christian thought as if it were obvious. It is not obvious. It never was.
When the early Christians prayed for the dead — and they did, in liturgies we still have — they were not doing one thing. They were holding together a cluster of hopes and uncertainties that the later tradition would sort into competing camps.15 When Augustine catalogued the misericordes, he was acknowledging — without meaning to, perhaps — that the question of postmortem mercy was alive and contested in the late fourth century. When Wesley preached on the wider hope, he was drawing on a tradition older and broader than the strict Augustinianism his Calvinist contemporaries assumed was the only Christian option. When Karl Barth wrote that we may not deny the possibility of universal salvation, he was retrieving voices that had never quite gone away.
Those voices are the subject of this book. The next chapter sketches the conceptual map that lets us tell one set of voices from another — what "postmortem salvation" has meant in different eras, why the modern reader needs to disambiguate before classifying figures across centuries. From there, we begin in the second century, with the writings we call the Apostolic Fathers, and follow the conversation forward.
The journey is long. Two thousand years is a long time. But the destination is worth the road: a more accurate, more honest, more humble picture of what Christians have actually believed about the fate of the dead, the scope of God's saving will, and the long hope that has accompanied the Church from its earliest days.
↑ 1. Ilaria L. E. Ramelli, The Christian Doctrine of Apokatastasis: A Critical Assessment from the New Testament to Eriugena, Supplements to Vigiliae Christianae 120 (Leiden: Brill, 2013). Ramelli's argument is sustained across nearly nine hundred pages of close patristic reading. The book has been reviewed extensively, including substantive critical reviews questioning some of her readings; see especially Mark Edwards's review in Journal of Theological Studies 66.2 (2015).
↑ 2. Michael J. McClymond, The Devil's Redemption: A New History and Interpretation of Christian Universalism, 2 vols. (Grand Rapids: Baker Academic, 2018). McClymond's project is openly framed in part as a response to Ramelli; he engages her readings at numerous points and reaches very different conclusions about the cumulative shape of the tradition.
↑ 3. The locus classicus for the disputed reading of Origen is De Principiis 1.6, especially 1.6.1–3, where Origen reflects on the eventual subjection of all rational creatures to Christ. The Latin translation by Rufinus and the Greek fragments preserved by hostile witnesses (Jerome, Justinian, the anti-Origenist florilegia) give somewhat different impressions of Origen's precise claim about the devil. See Chapter 8 for sustained engagement.
↑ 4. Jaroslav Pelikan, The Christian Tradition: A History of the Development of Doctrine. Vol. 1: The Emergence of the Catholic Tradition (100–600) (Chicago: University of Chicago Press, 1971). Pelikan's prefatory remarks on method and his careful distinction between "what the Church believed, taught, and confessed" run throughout the volume; the book also functions as a sustained demonstration of how to do this kind of historical theology well.
↑ 5. Carlos Eire, A Very Brief History of Eternity (Princeton: Princeton University Press, 2010). Eire's book traces five major Western conceptions of eternity and is especially helpful on the methodological discipline of letting premodern conceptions stand in their own categories before translating them into modern ones.
↑ 6. The classic study of the multiple senses of "Origenism" remains Antoine Guillaumont, Les "Képhalaia Gnostica" d'Évagre le Pontique et l'histoire de l'origénisme chez les Grecs et chez les Syriens, Patristica Sorbonensia 5 (Paris: Seuil, 1962). Henri de Lubac, Histoire et esprit: L'intelligence de l'Écriture d'après Origène (Paris: Aubier, 1950), develops the related distinction between Origen and Origenism. See also Chapter 11.
↑ 7. Origen states the principle in the preface to De Principiis, where he distinguishes between the items the apostolic preaching has clearly handed down and the questions on which the inquiring mind may legitimately speculate. See De Principiis, pref. 3, in the Butterworth translation (Harper, 1966; Christian Classics reprint, 1973), as well as Ramelli's discussion in Apokatastasis, chapter 3.
↑ 8. Brian E. Daley, S.J., The Hope of the Early Church: A Handbook of Patristic Eschatology (Cambridge: Cambridge University Press, 1991; Hendrickson reprint, 2003). Daley's book remains the standard handbook on patristic eschatology in English. He distinguishes between figures who teach apokatastasis explicitly and figures whose work is hopeful but stops short, and he resists both the maximalist universalist reading of the patristic record and the minimalist Augustinian one.
↑ 9. Augustine, Enchiridion ad Laurentium, sive de fide, spe et caritate 112, in A Select Library of the Nicene and Post-Nicene Fathers of the Christian Church, ed. Philip Schaff, 1st ser., vol. 3 (Buffalo: Christian Literature Co., 1887), 274. For a careful contemporary reading of the passage in context, see Richard Bauckham, "Universalism: A Historical Survey," Themelios 4.2 (1979): 48–54.
↑ 10. Augustine, De civitate Dei 21.17–27, in NPNF, 1st ser., vol. 2, 464–478. The full catalogue and Augustine's responses are unpacked at length in Chapter 12. Augustine does not himself use the substantive misericordes as a technical label in the Enchiridion; the term has become a convenient shorthand in modern scholarship for the cluster of "merciful" positions he is opposing.
↑ 11. The standard recent treatment is Daniël Hombergen, The Second Origenist Controversy: A New Perspective on Cyril of Scythopolis' Monastic Biographies as Historical Sources for Sixth-Century Origenism, Studia Anselmiana 132 (Rome: Sant'Anselmo, 2001). On the conciliar question, see Norman P. Tanner, ed., Decrees of the Ecumenical Councils, 2 vols. (London: Sheed & Ward; Washington, DC: Georgetown University Press, 1990), 1:105–122.
↑ 12. The fifteen anti-Origenist anathemas frequently cited as products of the 553 Council are by most recent scholars assigned instead (or in addition) to Justinian's prior synod of 543; whether they were ratified by the 553 Council itself is contested. See again Hombergen, The Second Origenist Controversy; and the careful summary in Ramelli, Apokatastasis, chapter 7. McClymond, The Devil's Redemption, 1:587–632, reads the conciliar record more strictly.
↑ 13. John Calvin, Institutes of the Christian Religion 2.16.8–12, ed. John T. McNeill, trans. Ford Lewis Battles, Library of Christian Classics 20–21 (Philadelphia: Westminster, 1960), 1:512–520. For the range of Reformed readings of Calvin's metaphorical descent, see Randall Zachman, John Calvin as Teacher, Pastor, and Theologian (Grand Rapids: Baker Academic, 2006); and the older but still useful summary in Heiko Oberman, "The 'Extra' Dimension in the Theology of Calvin," Journal of Ecclesiastical History 21 (1970): 43–64. Chapter 15 returns to the passage.
↑ 14. The reception of Origen in the early monastic tradition is itself a methodological case study. John Cassian, in his Conferences (especially Coll. 13 and 14), draws on Evagrian-Origenist categories without committing himself to explicit apokatastasis; his careful selective appropriation models one way of reading Origen well. See Owen Chadwick, John Cassian, 2nd ed. (Cambridge: Cambridge University Press, 1968); Columba Stewart, Cassian the Monk (New York: Oxford University Press, 1998).
↑ 15. The practice of intercession for the dead is documented in the inscriptions of the Roman catacombs and in liturgical sources from the second century onward. For an accessible historical overview, see Daley, The Hope of the Early Church, especially chs. 1–4; on the relationship between liturgical prayer for the dead and patristic eschatological reflection, see also Geoffrey Rowell, The Liturgy of Christian Burial (London: Alcuin Club / SPCK, 1977).