Known Knowns and Known Unknowns: Ellen Muehlberger’s Things Unseen: Essays on Evidence, Knowledge, and the Late Ancient World

This essay was part of a 2026 NAPS Conference Book Review Panel devoted to Things Unseen Essays on Evidence, Knowledge, and the Late Ancient World by Ellen Muehlberger. Read the full forum here. 

Ellen Muehlberger’s latest monograph, Things Unseen, reminds readers that knowledge is a product, the output of an intricate machine. The “Knowledge Machine” (to borrow the title of Michael Strevens 2020 book) is sophisticated, with as many cogs, regulators, pinions, gears, gimbals, switches, widgets, and valves as variation in the knowledge-products it creates. Its complexity makes it so that no single person can operate it. Rather, it relies on the participation and activity of an incalculable number of individuals working together—sometimes with intentional collaboration, other times unaware of each other; sometimes with chronological and geographical proximity, other times joined together by nothing other than an imagined tradition. Who are these individuals? The answer is regrettably tautological. Those endowed with power to operate the Knowledge Machine are those endowed with power to operate the Knowledge Machine, be it by (a claimed) right of birth, education, rank, marriage, effort, merit, wealth, influence, position, prestige, or even luck. Just as knowledge itself is boundless, so too is the Knowledge Machine. It can produce local or universal knowledge; it can serve a variety of clients, ranging from a small group of elites to sprawling government agencies and global organizations. Cultures throughout every historical period and every geographical region have had their own iteration of the Knowledge Machine.

In Things Unseen, Muehlberger stands before the late ancient Knowledge Machine and takes in its vastness, evaluating what it produces and how it works. Indeed, the act of producing and what is produced may be observably distinct, but neither can exist without the other. Epistemological content and process are concomitant. Yet, Muehlberger confronts the Knowledge Machine in full humility, cognizant that she cannot grasp it all herself. Its output is, admittedly, boundless (there can be knowledge of anything!), and therefore the Knowledge Machine itself must be boundless. To borrow one of Muehlberger’s favorite phrases, the late ancient Knowledge Machine exists in a state of “perpetual adjustment.” It is ceaselessly creating, essentializing, totalizing, idealizing, revising, and contradicting. It produces one bit of knowledge, then deletes it, only to re-produce it again in a modified form—over and over and over again, in a slew of contexts. In front of such a colossus, Muehlberger’s humility serves her well. Things Unseen does not provide a sketch of the unsketchable but invites readers to observe the Knowledge Machine at work in discrete situations—in classrooms, in private homes, in church councils, and in the wonder with which modern scholars have regarded “portraits” of the dead.

The first chapter looks at the late ancient Knowledge Machine in the context of rhetorical education. “Speech-in-character” was a popular pedagogical exercise in late antiquity that asked students to imagine a character undergoing a counterfactual reversal of fortune (that is, something that would never actually happen) and then give first-person voice to that character’s experience. The particular exercise on which Muehlberger lingers is this: what would a prostitute (pornē) who has come to her senses (sōphronēsasa) say? Here the Knowledge Machine starts humming: the student must transform a known unknown into a known known by creating knowledge about a subject he knows very little if anything about (women generally, prostitutes in particular). The success of that transformation hinges on how closely the student’s rhetorical fabrication—voice, feeling, emotional depth, character—hews to the instructor’s understanding of realness.

Muehlberger provocatively suggests that the same exercise occurs in a monastic context, where male authors compose hagiographies of impossible women like Mary of Egypt, a reformed prostitute who outperforms male monks in virtue (a counterfactual if there ever was one from the perspective of a late ancient monk!). While one might see speech-in-character, whether rhetorical or hagiographical, as an act of empathy—isn’t the male writer trying to adopt the perspective of someone marginalized and stigmatized?—Muehlberger sagely reminds us that, no, empathy has nothing to do with the exercise. After all, no change in late ancient social hierarchies or gender relations came about because of these exercises. Rather, they were performances of power: “Instead of empathy, then, the practice of speaking this way teaches a boy the disposition of mastery: that his thoughts and assumptions, however inexpert, are sufficient to the task of knowing others” (p. 39; my emphasis). Put differently, the elite culture of rhetorical education entrusted these young boys with the power to operate the Knowledge Machine. What they produced is not an accurate representation of another human being’s lived experience, but—ironically—a “resilient ignorance” (p. 51) that refuses to permit real encounters with these human beings to affect their fabricated representations.

The second chapter tracks the late ancient Knowledge Machine creating certainty about religious identity. Which markers can reliably identify someone as Christian? Sure, a simple confession might do, but late ancient Christian leaders knew that actions speak louder than words. So, which actions mark someone as a Christian? Regular church attendance? Conversion? Submission to the catechumenate and baptism? Meaningful performances, sure, but they can be performed without authentic, sincere faithfulness. Just ask Augustine, who reminds his congregation that, among them, were vere conversi and falso conversi; just ask Eusebius of Caesarea and Socrates Scholasticus, each of whom notes the self-interested motivations that lay behind individuals’ decisions to get baptized (pp. 57–59). It turns out that late ancient church leaders didn’t really regard actions as unimpeachable evidence of interior dispositions. Where can the Knowledge Machine produce something known, something undeniable, when there are so many unknowns?

Christian leaders had an answer: the private house. For the Christians of late antiquity, you are where you live (or more precisely, you are what you keep in your house). The early Christian imagination understood that, within the confines of their private residences, folks kept objects that indubitably evidenced authentic religious identity. A true Christian will have Christian objects, but a crypto-pagan, or a crypto-Jew, or a half-Christian, will stockpile pagan or Jewish or even demonic objects. This axiom is what informs the regulation of religious identity in imperial legislation through the regulation of religious property and building use; it’s what authorizes the archimandrite Shenoute to break into Gesios’s house and out him as a crypto-pagan in the fifth century, or what authorizes John of Ephesus’s forcible pop-in at the houses of Rufinus and Anatolius to identify them as crypto-pagans. From the perspective of those empowered to use the Knowledge Machine to monitor and police interior commitments, the private house contains a trove of evidence that confirms or refutes someone’s public claim of identity.

The third chapter finds the Knowledge Machine operating in the context of the Second Council of Constantinople (553 CE). Sure, this council falls short of the other “important” ecumenical councils like Nicaea (325 CE) or Chalcedon (451 CE) because of its pitiable decrees and methods, motivated in part by a desire to re-open cases against figures who had long been regarded as orthodox. Yet, Muehlberger argues that Constantinople II was just as generative as its illustrious ecumenical predecessors. The clergy members gathered there used the Knowledge Machine to create a new type of idealized figure; Muehlberger calls it “the superfather,” someone who existed as a “transtemporal guarantor of orthodoxy . . . without will or substance” (p. 92), someone who was not “subject to the constraints of time” (p. 115). Muehlberger’s term refers to an idea connected to a very real historical person, but whose identity came to both transcend the living person altogether and exert a new gravitational force on memory and tradition: “Once created, a superfather changes how other Christians act going forward, but his presence also changes the past that has already happened” (p. 115).

The first one was Cyril of Alexandria, the controversial fifth-century bishop, whose superfather status was created to resolve a conflict of contradictory orthodoxies. On the one hand, Cyril adopted the Apollinarian phrase (which he understood to be an Athanasian phrase) “one incarnate nature of the Word” (First Letter to Succensus 6). On the other hand, Ibas of Edessa condemned Cyril’s use of that phrase in a letter deemed orthodox by the bishops gathered at the ecumenical Council of Chalcedon in 451. The bishops at Constantinople II—a century after Chalcedon—confronted a potential brain-breaker: how could the orthodox Council of Chalcedon have affirmed the orthodoxy of a text that condemns the orthodox Cyril? The solution lay in Cyril’s elevation to superfather status, which authorized attendees at the Constantinople II to bend over backwards to protect Cyril’s orthodoxy (and Chalcedon’s too) against any accusation of heterodoxy by alleging that Ibas’s letter was, in fact, a forgery. That is, using Cyril’s status as a superfather, Constantinople II rewrote the past by sacrificing Ibas’s orthodoxy in order to preserve Chalcedon and Cyril’s orthodoxy. The alignment of a father’s statements (however varied) into a single position, along with a commitment to rework a tradition’s past determinations to conform to the perceived legacy of that father (the condemnation of the “Three Chapters,” Muehlberger concludes, has less to do with condemning theological ideas than condemning explicit or implicit challenges to Cyril) is the origin story of the superfather.

The final chapter of Things Unseen examines the Knowledge Machine at work in the context of modern scholarly study of the so-called Fayyum Portraits. Since their “discovery” in the 1880s, they have been treated as evidence for direct knowledge of real human beings from antiquity. Yet, that’s not what they are. They are the funerary depictions installed within mummies by ancient people that were extracted, distributed, and sold for profit by European antiquity collectors like Theodor Graf, Flinders Petrie, Albert-Jean Gayet, and others in the late nineteenth and early twentieth centuries. The attraction of these objects was the facial depictions, which both collectors and viewers beheld as an ancient version of a photographic portrait. Indeed, photography, invented in 1839 and popularized over the course of the later nineteenth century, was being lauded as a tool that could statically and permanently capture a physical representation, which could be analyzed with “race science” and the nascent discipline of psychology. So, when collectors and viewers looked upon the Fayyum Portraits, they saw not simply ancient portrayals of dead people, but poignant expressions of individuality, personal presence, genetic essence, and essentialized ethnic identity (be it Romanness, Greekness, Egyptianness, or some combination thereof). The thought was that viewers could actually see regular ancient people in these objects.

Of course, Muehlberger points out, that conclusion was merely the illusion created by the modern Knowledge Machine. What viewers see when they look at the Fayyum Portraits is the product of profit-driven malfeasance, unauthorized exhumation, mistreatment of human remains, and ultimately vandalism of funerary objects. But the deracinated exhibition of the facial depictions in western museums (divorced from their funerary provenance, displayed without any sort of interpretive framework) has left viewers to generate the most challenging knowledge of all: known knowns from unknown unknowns. The result is a rebranding of the Knowledge Machine’s output as something new, something strikingly mysterious—wonder. Yet, as Muehlberger points out with characteristic incisiveness, contriving wonder as the scholarly response to these funerary objects is less the result of critical analysis than it is a, dare I say, lazy copout: “When knowing depends on not knowing, wonder is a manufactured response, and predictable” (157).

I will conclude with praise for the innovative structure of Things Unseen. No single monograph could comprehensively capture the protean nature, magnitude, pervasiveness, nebulousness, and elusiveness of the Knowledge Machine in any period, and so Muehlberger’s episodic approach is fitting. In the process, she invents a new genre of historical writing—the fake invited lecture series. The invited lecture of a senior scholar is smart and authoritative, yet loose and almost conversational, filled with confident assertions earned from decades of study; sometimes a senior scholar gets invited to give a series of such lectures and then publish them together in a book. Well, no one invited Muehlberger to give these lectures, but here they are, informed and casual, provocative and convincing, the perfect blend of insight and waggishness. It is my hope, after reading this delightful volume, that other scholars take inspiration from Muehlberger’s creation in their own work.

Bradley K. Storin is Associate Professor of the History of Christianity in the Department of Philosophy and Religious Studies at Louisiana State University.