Two Cities and Two Loves

What Augustine actually said

Augustine wrote The City of God against the Pagans — the full Latin title is De Civitate Dei contra Paganos — between 413 and 426 CE. The immediate occasion was the sack of Rome by the Visigoths in 410, an event that produced a crisis in the Roman intellectual imagination because Rome had been held to be eternal. Pagan critics charged that the catastrophe was the consequence of Rome's abandonment of its traditional gods in favor of Christianity. Augustine's response, written over more than a decade, became the longest theological work of late antiquity.

The work organizes itself around a contrast between two communities, which Augustine calls cities. The earthly city (civitas terrena) is the community of those whose loves are ordered toward themselves; the heavenly city (civitas Dei) is the community of those whose loves are ordered toward God. The two cities are not earthly Rome and heavenly New Jerusalem in any geographically simple sense. They are interwoven communities, present in every time and place, distinguished by the orientation of their members' loves rather than by visible institutional boundaries. A given person, on Augustine's account, can be in either city, and the choice is the choice of what one most deeply loves.

Book XIV is the book in which Augustine names this distinction directly. The full passage from chapter 28, in standard English translation, reads: “Two cities, then, have been formed by two loves: the earthly by the love of self extending even to the contempt of God; the heavenly by the love of God extending even to the contempt of self. The one, therefore, glories in itself, the other in the Lord; the one seeks glory from men, but the greatest glory of the other is God, the witness of conscience.”

The structure of the formulation matters. Each love “extends even to” an extreme. The love of self extends even to contempt for God; the love of God extends even to contempt for self. Augustine is not contrasting healthy self-love with healthy God-love. He is contrasting two complete orientations of the heart, each of which, when fully realized, displaces the other. The two cities are produced not by behavioral differences but by what their members would, in the limiting case, give up.

The encyclical's translation tightens this slightly, in the form that appears at paragraph 130: “Two loves have built two cities: the earthly city, the love of self even to the contempt of God; the heavenly city, the love of God even to the contempt of self.” The truncation drops the second sentence about glory and conscience but preserves the load-bearing structure: two loves, two cities, the “even to” that names what each love is willing to relinquish.

What Pope Leo XIV does with it

The encyclical's use of Augustine at paragraph 130 is a particular kind of magisterial move. The pope is not arguing for Augustine's position. He is using Augustine's frame to organize a contemporary analysis. The distinction matters because it explains why the frame works for non-Catholic readers as well as for Catholic readers.

The full text of paragraph 130 reads: “Questioning this alternative path of progress and how we interpret and live it is ultimately a matter of examining our own hearts. The way we understand and shape relationships, work and institutions, in practice reveals our fundamental values. In the end, it all stems from what we hold most dear. This is a love that guides us as to what we truly cherish, both as individuals and as a society, and directs our lives and actions. Saint Augustine described human history as a struggle between two loves, which give rise to two ways of inhabiting the world and living together — or two 'cities,' as it were: on the one hand, the love of God and neighbor; on the other, the exclusive love of self. 'Two loves have built two cities: the earthly city, the love of self even to the contempt of God; the heavenly city, the love of God even to the contempt of self.' As throughout history, these two loves continue to contend for dominance in our hearts today. The age of AI is no exception: the construction of Babel or the rebuilding of Jerusalem begins within each one of us.”

Three moves are being made in this paragraph.

First, the location of the AI question. The standard AI ethics conversation treats the question as primarily political and institutional. Who controls the models? Who regulates the deployment? Who benefits and who is harmed? The encyclical does not dismiss these questions. They occupy much of Chapter 3 and Chapter 4. But paragraph 130 places them inside a larger question: what loves are operating in the persons who build, deploy, and regulate AI? The political and institutional questions cannot be answered without engaging the prior question of what their actors love.

Second, the diagnosis of what the technocratic paradigm gets wrong. The encyclical has spent Chapter 3 critiquing what it calls the technocratic paradigm: the disposition that treats every problem as solvable by technical means and every value as quantifiable. The Augustinian frame names what is missing from that disposition. The technocratic paradigm cannot ask what one loves because it has no vocabulary for the question. The question requires a different mode of self-examination than the technocratic disposition can sustain.

Third, the connection to the Babel and Nehemiah frame. The biblical frame the encyclical introduced in the opening paragraphs returns at the close of Chapter 3 by way of Augustine. The construction of Babel is not just a different building project from the rebuilding of Jerusalem. It is the building project that emerges when the love of self even to the contempt of God is operating in those who build. The rebuilding of Jerusalem is the building project that emerges when the love of God even to the contempt of self is operating. The biblical frame names two outcomes; the Augustinian frame names what produces them.

The closing line of paragraph 130 — “the construction of Babel or the rebuilding of Jerusalem begins within each one of us” — carries the weight of all three moves at once. The reader is being told that the AI question is, in the end, a question about the heart of the reader.

Why this pope made this move

Pope Leo XIV is an Augustinian friar. He entered the Order of Saint Augustine in 1977, made his solemn profession in 1981, and served as prior general of the Augustinians from 2001 to 2013. His theological formation is Augustinian in a way that is not incidental to his magisterium.

This is biographical detail, but it becomes magisterial method at paragraph 130. Popes routinely draw on their formation; the formation shapes the documents they produce. The encyclical's use of Augustine in this passage is not an outsider's deployment of a useful frame. It is a member of the Augustinian tradition engaging that tradition in a magisterial document, where the question of what one is doing with Augustine is itself a theological question.

The choice of City of God Book XIV specifically also matters. Augustine's City of God has been used by popes before, most famously by John Paul II, but the use has typically drawn on the work's political theology or its philosophy of history. Pope Leo XIV draws on the work's anthropology — the account of what human beings are and what makes them what they are. The anthropological reading of Augustine is, in some sense, the deeper reading. It is also the reading most appropriate to the question paragraph 130 is asking.

What is significant about the move is not that it is theologically sophisticated. Popes are theologically sophisticated as a baseline. What is significant is that the move is made at the structural pivot between the AI ethics analysis of Chapter 3 and the social analysis of Chapter 4. The placement is deliberate. The encyclical is saying: before we go from the AI question to the social question, we have to stop and ask what the AI question really is. The Augustinian frame is the answer to that question.

For readers attending closely to the encyclical's argument, paragraph 130 is the place where the document's depth becomes visible. The chapter could have ended without it. The argument could have been completed without it. The frame is not load-bearing for the encyclical's policy recommendations. It is load-bearing for the encyclical's account of what kind of question AI is.

What the frame asks of readers

The Augustinian frame is not principally a political frame. It is an anthropological one. It asks readers to do interior work that the institutional and regulatory work cannot substitute for.

This is a demanding request, and it is worth being honest about why. The contemporary AI policy conversation has structural advantages that the Augustinian frame disrupts. Policy questions can be addressed by institutional actors, can be measured by outcomes, can be improved through iteration. The question of what one loves cannot be addressed in those ways. It can only be addressed by the person asking the question, in the form of self-examination that produces, if it produces anything, conversion.

For Catholic readers, the frame is familiar. The Catholic spiritual tradition has developed practices of self-examination — the examen of conscience, spiritual direction, the sacrament of reconciliation — that the Augustinian frame fits into directly. The encyclical is asking Catholic readers to bring these practices to bear on questions about technology that the practices have not typically been used for.

For non-Catholic readers, including the secular AI policy audience the encyclical is also addressed to, the frame is less familiar but still tractable. The question of what one loves does not require Catholic theology to engage. It can be engaged by anyone willing to take the question seriously. The frame is offered, not imposed. Readers who decline the Augustinian theological background can still take the anthropological question on its own terms: what loves are present in the persons building the systems that will shape the next several decades of human life, and are those loves the right ones for the work?

What the frame refuses is the strategy of locating the AI question entirely outside the self. Industry actors who attribute the structural problems of AI development to competitor pressure are using that strategy. Critics who attribute the problems entirely to industry are using a version of it. Regulators who attribute the problems to insufficient legal frameworks are using another version. The Augustinian frame insists that the question is also — not only — inside the persons engaging the question. This is uncomfortable. It is also, on the encyclical's account, where the analysis has to begin.

How the frame pairs with disarming AI

The two cities frame at paragraph 130 and the disarming AI cornerstone at paragraph 110 are doing complementary work in Chapter 3. They are not alternatives.

Disarming AI names what has to change at the level of institutions, governance, and industry practice. The phrase is operational. It identifies the competitive race dynamics of AI development as part of the problem and calls for the technology to be freed from monopolistic control and restored to the plurality of human cultures. The work it asks for is structural: legal frameworks, treaty regimes, oversight institutions, accountability mechanisms.

The two cities frame names what has to change at the level of the human heart. The phrase is anthropological. It identifies the orientation of love as what produces the building projects, and calls for the loves operating in those who build to be examined and converted. The work it asks for is interior: self-examination, the disciplines that produce changes in what one most deeply cares about.

The encyclical's argument is that neither move succeeds without the other. A disarmed AI built by hearts still ordered to the love of self even to the contempt of God will rearm. The competitive dynamics will reassert themselves; the institutional safeguards will be circumvented; the structural reforms will be hollowed out by actors who never engaged the anthropological question. Conversely, hearts ordered to the love of God even to the contempt of self, operating inside institutional structures that continue to reward the love of self, will be defeated by those structures. The interior work and the institutional work are mutually dependent.

This is not a new claim in Catholic social teaching. The Church's social tradition has long held that institutional reform and personal conversion are inseparable. Magnifica Humanitas applies the claim to AI. Paragraph 110 gives the institutional work its sharpest language. Paragraph 130 gives the personal work its sharpest language. Read together, they are the encyclical's most concentrated account of what has to change.

Where this sits in the document

Paragraph 130 is the final paragraph of Chapter 3. The chapter has done its conceptual work on AI and the technocratic paradigm; it has introduced disarming AI; it has analyzed transhumanism and posthumanism. Paragraph 130 is the chapter's closing move, and its placement is significant.

The chapter does not need paragraph 130 to complete its argument in the analytical sense. The analysis of the technocratic paradigm is complete by paragraph 129. The closing paragraph could have been a brief recapitulation or a transition to Chapter 4. It is not. It is the chapter's deepest move, placed at the structural pivot between the AI ethics core and the social analysis that follows.

This placement tells us something about how the encyclical is built. The depth of paragraph 130 carries the rest of the document. Chapter 4's analysis of truth, work, and freedom, and Chapter 5's analysis of war, depend on the anthropological frame established at the close of Chapter 3. The reader who reaches Chapter 4 having missed the move at paragraph 130 will read Chapter 4 less deeply. The reader who reaches Chapter 5 having missed the move at paragraph 130 will read the war chapter as a separate concern rather than as the place where the loves identified by Augustine produce their most lethal outcomes.

For readers attending to the encyclical's structure, paragraph 130 is the document's theological hinge. The biblical frame from the introduction connects to the Augustinian frame at paragraph 130. The Augustinian frame then connects forward to the chapters that follow. The whole document holds together because of the move at this paragraph.

The day-after analysis identified the Augustinian frame as one of six moves pre-release coverage did not anticipate. For the full set, see Magnifica Humanitas: What Was Surprising. For the pairing partner that operates at the institutional level, see Disarming AI. For the chapter where the loves identified here produce their most lethal effects, see Magnifica Humanitas on War.