I am happy to have been invited to participate in this forum on oaths—these powerful declarations—because I have been wrestling for some time with the question of how scholars in religious studies and across the academy might better identify and analyze one particular type of powerful declaration: martyrdom. Martyrdom (whether we mean the act of becoming a martyr oneself or the discursive configuration of an exemplary figure as a martyr) encodes values and ideals, declaring these foundational truths to an anticipated audience that is expected to understand and be affected by the martyrdom. Martyrdom connotes the ultimate “all-in”—the truth you would stake your life on and whose expression outweighs all else. Nonetheless, a definition of martyrdom remains elusive—like pornography and religion, we “know it when [we] see it” (see Sullivan, 2018).

This definitional attitude, as it pertains to martyrdom, is functional in its own way, allowing us to avoid a singular, fixed definition that would elide the vast differences in theologies of martyrdom that exist both between and within (and without) traditions. But “I know it when I see it” is also stymieing—it does little to prevent our preconceived notions about martyrdom from dictating what we register as martyrdom. Perhaps more dangerously, it provides no analytical framework for identifying martyrdom when the precise terminology is lacking (what Bryan van Norden terms “the lexical fallacy”). This is dangerous to historians because it conveys a false sense of the phenomenon’s etiology, and dangerous to observers of current events, who miss a substantial undertone of global discourses when we fail to note their martyrial inflections.

In this essay, I want to propose that thinking about martyrdom along the axis of oath’s quasi-cognate “witness” holds the key to a fruitful and functional understanding of martyrdom as a transhistorical phenomenon. Martyrdom, functionally speaking, can best be understood as sacrificial witness.

Why Martyrdom Is Not Simply “Dying for a Cause”

Martyrdom is colloquially and habitually defined as “dying for a cause.” Scholarly enhancements of this understanding attend to martyrdom’s discursive character, its reliance on narrative communities to construct and receive those identified as martyrs, the historical variety in (and disagreement over) who counts as a martyr, the origins of the terminology of “martyrdom” in the Greek concept of judicial witness (μάρτυς), and the concept’s development through various eras and cultural contexts. But despite their nuance and their salutary correctives to the colloquial understanding, all of these treatments operate on a definition of martyrdom that centers on death. Death is the necessary object of spectacle, discourse, narrative, and witness; it is martyrdom’s sine qua non.

But the martyrs I study do not die. Or rather, their martyrdoms are represented as having been earned without death. This is the reason for my wrestling with this definitional question and my impetus for attempting to develop a new analytical tool for identifying martyrdom in history and discourse. If we look at how various Christian communities in history have understood martyrdom, we see that martyrdom is, quite often, not configured around death.

Indeed, the third century North African poet Commodian chided his contemporaries for thinking that one could become a martyr by dying: “You offer so many empty words, who lazily seek in one moment to raise a martyrdom to Christ!” (Instructiones, 2.18.16-17).1 Instead, the Christian seeking martyrdom must engage in a daily, all-consuming war against sin (Instructiones, 2.18). If you purge yourself of envy, restrain your tongue, make yourself humble, use force against no one and have instead a patient mind, Commodian says, “Understand that you are a martyr” (Instructiones, 2.14.14-18). The Gallo-Hispanic poet Prudentius, writing in the late fourth century, used his poetry to train his contemporaries to be living martyrs by interpreting the world around them typologically; he also configured the virgin martyr Encratis as a powerful intercessor precisely because she had survived her martyrdom and lived to tell the tale (Peristephanon 4). Meanwhile, Paulinus of Nola, Prudentius’s contemporary, defended the martyr-status of his saintly patron Felix (who had survived persecution through miraculous subterfuge and died quietly of old age years later) by claiming that God “remits punishment of the flesh on account of proper piety” (Carmen, 14.9). Martyrdom by death, Paulinus implies, is a consolation prize for those servants of God whose piety was somehow lacking. Augustine of Hippo similarly eschewed death as a criterion for martyrdom. For Augustine, martyrdom indicated “witness”—he repeatedly reminded those attending his sermons of the term’s etymology and at one point declared: “Anyone who preaches wherever he can is, indeed, a martyr” (Sermo260E, 2). We can see the rejection of death as well in Augustine’s famous (and oft-repeated) formula for distinguishing between true and false martyrs: non poena sed causa, or, “it is not the punishment but the cause” that makes a martyr. If it is not the punishment but the cause that makes the martyr, would it not be possible for a Christian to be a martyr without any “punishment” at all? For each of these influential late ancient Christians, martyrdom is not defined by death.

Scholars cannot dismiss these martyrologists as “speaking figuratively” (as did Thomas Aquinas [Summa Theologiae, 2a2ae. 124, 4]); they all seem quite invested in their claims. And martyrdom without death is a tenacious idea: even after Aquinas’s declaration that martyrs must die, many have nonetheless claimed the title without expiring. The fourteenth-century English pilgrim Margery Kempe, for example, in characterizing her continuous public humiliations as martyrdom, truly understood herself to be surpassing the feats of martyrs who died. This opinion is confirmed for Margery by a vision she has of Jesus himself, who tells her, “Daughter, it is more pleasing to me that you suffer disdain and scorns, shames and censures, wrongs and distress, than if your head were struck off three times a day every day for seven years” (Kempe, Ch. 54).

Whatever Aquinas’s objections, it is the duty of scholars working outside or across confessional boundaries to make our definitions as inclusive, accurate, and functional as possible. The question, then, is how we might come to such a definition.

For this, I think, we can turn to the concept of witness.

Sacrificial Witness

To witness is to observe—to sense and register something as both true and known. To witness is to testify—to communicate in some fashion that which you know to be true. To witness is to enact—to be forever changed by the known and to embody it such that others, seeing you, see your truth made manifest. Witness aligns and commits one to a truth, to its performance, profession, and communication. The interplay between observing, testifying, and enacting is unsettled (as are the boundaries between them), shifting as the focus of our analysis shifts, but consistently results in constructions of reality that are simultaneously reflective and generative.

Using this polyvalent witness alongside an understanding of “sacrifice” that incorporates both forfeiture and sanctification (that is, we give something up and by giving it up render it significant), we can identify a stable conceptual phenomenon across historical contexts that can be labelled “martyrdom.”

Martyrdom does not consist in dying or sacrificing oneself for a cause; rather, it comprises having one’s representation subsumed into a larger “truth,” the realities, multiplicities, and infelicities smoothed away and subordinated in the service of representing that truth. Thus martyrdom necessitates sacrifice, but not in the literal, death-centered ways we commonly assume. It is, in effect, the narrative subordination of a character to a cause, accomplished through various manipulations of observing, testifying, and enacting witness.

This might seem too vague to be useful. How is this subordination of “reality” to narrative different from any other? How does it differ, for example, from the stories we tell ourselves about ourselves, or from the narrativization of experience that gets us through our days? This is where witness once again helps us: martyrdom requires an audience to whom the martyr or martyrologist is attempting to convey something they have identified as true.

But, then, is all persuasion-oriented communication martyrdom? How can we distinguish between a narrative about a hero and a narrative about a martyr? This is where sacrifice comes in. In a heroic narrative, whatever reality the hero has is subordinated to her representation in narrative, just as it would be for the martyr. But the subject of the narrative is the hero: the truth being conveyed and witnessed or witnessed to is the hero herself. The martyr’s witness points beyond the martyr to a larger truth. Sacrificial witness involves the subsuming of the person witnessing into the truth witnessed.

Understanding “martyrdom” as sacrificial witness helps us foreground the communicative and communal aspects of martyrdom, and to ask what values and what truths are being encoded and endorsed in the choice and representation of a martyr.

Any definition must, of course, survive repeated challenge and testing to be considered true. But in one area, at least, this definition of martyrdom has already had demonstrable success: my teaching. Students in classes where I introduce the idea of martyrdom as sacrificial witness come up with far better paper topics on martyrdom than students in classes where I have not. They more clearly see the declarative and communicative aspects of martyrdom, they ask far more sophisticated questions of their primary sources, they are more judicious in their use of secondary sources, and we have richer class discussions. I can think of few more powerful declarations of “truth” than that.


  1. All translations in this piece are my own.