This book argues that conventional understandings of Roman law – as an impersonal decisional paradigm under which disputes were resolved through casuistically elaborated rules – are not quite right. They are not entirely wrong, either; certain sources, especially the legal compilations authorized by Justinian I (r. 527–565 CE), describe a “Roman law” that looks very much like what I have described and that came to exercise enormous influence over later legal systems. But the law that would eventually be taught in Europe’s first law school in Bologna, or that men like Irnerius,Footnote ¹ Henry de Bracton,Footnote ² and Hugo GrotiusFootnote ³ took as a historical exemplar,Footnote ⁴ is surprisingly hard to find in the period we imagine it to have governed.

Instead, Romans of the “Classical period”Footnote ⁵ experienced a less rulebound political order that featured many different ways of thinking about state power. Law was present in the world but was not the only way to solve problems, or even always the most authoritative and final. Rome at this time combined an ancient legal tradition, which had been understood as inherently just and linked to the aristocratic power-sharing arrangements of the Roman Republic, with a newer autocracy. These two orders coexisted but did not really compete, since even at their weakest emperors had more power to enforce their judgments than did legal experts. Accordingly, imperial Romans understood law in somewhat symbolic terms: as an older way of doing things that worked as a metaphor for other older ways of doing things. Emperors who worked through legal channels (as judges, as legislators, and even as litigants) demonstrated respect for Republican political tradition, while those who did not appeared as more novel or troubling figures.

This pattern arose haltingly and slowly, but its most consequential evolution dates to the Severan dynasty (193–235 CE). The Severan period saw rulers represented in state communications as quasi-functionaries: high-ranking officials whose decisionmaking followed the prerogatives and logics of an overarching institutional structure. That structure was organized according to legal rules, and accordingly it is in the Severan period that Romans began to tell stories about an imperial state that was headed not by a man who follows or does not follow the law, but by the law itself. This tendency continued, and even the more centralized late antique (by which I mean the mid third to mid sixth centuries) autocracy described itself in similarly bureaucratic terms.Footnote ⁶ Severan legal documents come down to us – mostly – as part of Justinian’s grand regularizing project, but that project was not a historical chronicle and attributes to these texts a social role that they only came to exercise far later. The set of principles and institutional relationships that we now understand to make up “Roman law” were creatures of ideal theory, which became socially powerful – in the way that we now understand law to be socially powerful – with time.

Those principles and institutional relationships were rarely stated out loud, but are strongly implied in Roman legal discourse. That discourse represents law as a technical discipline that governs judicial outcomes according to predictable and impersonal rules. This vision of legal sovereignty was compelling to Roman legal authors: to emperors who benefited from appearing to be institutionally legitimated agents, to bureaucrats who understood Roman power in terms of their particular specialties, and to subjects who found law a more appealing or useful sovereign than the one they actually had. All of these people discussed a legal regime that never quite existed, but in doing so they laid the groundwork for contemporary legal regimes, which mostly do.

Two Kinds of Thinking About Hay

Justinian’s compilation, the Corpus Iuris Ciuilis or “Body of the Civil Law,” presents Roman law as a set of rules. Roman law appears in these documents as a mass of dazzlingly precise technical arguments, in which jurists – trained scholars who served as advisers in courts throughout the empire – applied (or advised others on how they should apply) systematic and impersonal principles to specific fact patterns. By contrast actual court records, generally from Roman Egypt, show litigants ignoring the rules and invoking the pity of a benevolent sovereign; literary historians describe tyrants upending legal systems on a whim; and inscriptions thank enlightened rulers for handing justice down to their grateful subjects.

The difference is best illustrated via example. One can imagine specialists in all of these different corpora of evidence confronting the question “what remedies were available for a farmer in the Roman empire whose crops were stolen?” The Corpus (more specifically its collection of juristic writing, called the Digest) answers the question with an excerpt of writing from the jurist Paul, who focuses on the maturity of the crops and the nuances of legal ownership.

Other documents tell a different story. For example, the papyrus P. Mich. VI.423 preserves a petition to the strategos (a high provincial official) from the Greek Egyptian Gemellus-called-Horion. Gemellus-called-Horion de-emphasizes the elements of theft Paul treats as legally salient in favor of a pathetic narrative involving blindness and witchcraft.

These sources preserve two very different views of, or modes of engagement with, Roman law. One describes legal order as formal, systematic, rational, and coherent: the other merciful, ethical, and sublime. That difference has been heavily theorized for more than a century, since papyri from Roman Egypt (like P. Mich. VI.423) clearly record very different legal processes than those found in the Corpus Iuris Ciuilis. Following Mitteis we sometimes refer to these two legal paradigms as Reichsrecht (law of the kingdom, law on the books) and Volksrecht (law of the people, law on the ground).Footnote ⁹

More recent scholarship has analyzed this duality through the lens of legal pluralism.Footnote ¹⁰ Legal pluralism identifies “bargaining and conflict between contending normative orders” in much of what we imagine as legal history; it argues that many legal regimes feature overlapping paradigms with their own claims to legal force, and that people living under those regimes are often quite sophisticated in seeking out advantage under different sets of rules.Footnote ¹¹ Whether conceptualized in weak terms (as two recognizably legal systems imposed by the same sovereign upon different populations)Footnote ¹² or strong (where “the law” interacts with other paradigms that work to structure a community’s normative beliefs),Footnote ¹³ legal pluralism gives historians a theoretical framework with which to bridge the gulf between what a law-code says should happen and what people actually do.

The problem at the core of this book – the difference between Roman law as imagined by Gemellus-called-Horion, and that imagined by jurists like Paul – is almost classically pluralist, but most legal pluralists care more about what happens in a courtroom than I do. The pluralistic tradition emerged from legal sociology, and in the Roman context has been taken up by scholars of ancient judicial practice. Accordingly, much pluralist work takes legal or quasi-legal process as its object of study, with codified law serving as a foil.Footnote ¹⁴ By contrast, I am primarily engaged with the questions and assumptions of Roman Reichsrecht. I agree with the pluralists that law on the books did not serve the purpose it has historically been taken to serve in subjects’ lives, but instead of asking what those subjects really used, I ask what law on the books was really doing. In an influential article on the value of doctrinal sources for critical historians, Gordon notes that “mandarin materials are among the richest artifacts of a society’s legal consciousness.”Footnote ¹⁵ The Corpus Iuris Ciuilis preserves an astonishing volume of “mandarin texts,” produced by mandarins whose hopes and fears give Roman Reichsrecht its distinctive texture. Accordingly, the Corpus can help us better understand how Romans thought about the nature and function of law in their world – a set of beliefs that I will refer to hereinafter as Roman “legality.”

That legality is my focus. This book is not a history of Roman law as the subject has traditionally been understood – I do not discuss doctrinal evolution at all, and while I consider innovations in legal argumentation and recordkeeping I use them primarily to show change in what people thought they were arguing about and keeping records of. I hope historians of Roman law in that more traditional sense will still find the book useful, since it is a history of the conditions and assumptions that make Roman law coherent, and thus Roman legal historiography possible. But it remains a political history, of how law worked in Roman political thought.

By that, I mean that I am interested in how law imagines Roman power. Roman legal discourse describes a system operating according to prospective and clear rules, which derive their authority from broadly legitimated institutions like the Senate; the epigraph to this book is a claim by Ulpian that imperial power was only valid because it was bestowed by a “kingly law” (lex regia). In practice imperial power came from a broad variety of sources, and Rome’s idiosyncratically weak autocrats relied upon buy-in from other powerful social actors like the Roman aristocracy, wealthy provincials, or the army.Footnote ¹⁶ That buy-in was not guaranteed, and the unwritten rules that governed imperial politics were complex. Roman subjects – particularly the elites whose views are overrepresented in literary sources – sometimes understood those rules as a system of virtues. Good emperors act one way and live, bad emperors act another way and die.Footnote ¹⁷ At other times, imperial selection might be determined by military strength, palace intrigues, or even the result of an auction.Footnote ¹⁸ At all times, however, imperial politics were unpredictable and potentially violent.

That chaos is hard to find within the Corpus Iuris Ciuilis, whose structure and logics tell a very different story about imperial power. The result is an odd bifurcation within imperial historiography, which this book seeks to address. Many historians of the Principate and DominateFootnote ¹⁹ find law forbidding and artificial, since it describes a different world from the one in which they are interested, and the Digest in particular is tough to square with the realities of the emperor’s social and judicial role.Footnote ²⁰ By contrast, scholars of the Corpus (who frequently work in law faculties) may be more interested in the fictive and influential constitutionalism that survives within those books than in the world of their authors. Accordingly, the majesty they sometimes attribute to Roman legal thinking can seem more like a misapprehension of imperial history than the aesthetic or philosophical judgment it actually is.Footnote ²¹

I aim to resolve the problem by understanding Roman Reichsrecht as an archive of political imagination. The Corpus Iuris Ciuilis tells stories about the world that did not so much govern readers’ encounters with law – the experience of litigants, as recorded in other media, suggests that they often did not – as help readers make sense of or accept those encounters. Imperial and legal sovereignty could not easily coexist, but Romans frequently used the structures and institutions of “law talk”Footnote ²² to think through problems of imperial power. This process was influenced by the conventions of Roman legal thought, and influenced those conventions in turn. The political crises of the Severan period and beyond necessitated new stories about imperial legality, but that legality is neither as straightforward nor as old as it appears.

Slouching Towards Bologna

Since I am arguing that the Corpus Iuris Ciuilis is part of the history of imperial representation, I should first clarify how I understand legal texts to engage in representation. While this particular capacity of law has not traditionally featured within doctrinal scholarship, it has been an object of intense interest within other branches of legal studies. This interest began in the early twentieth century, in a “legal realist” movement that understood adjudication as politically charged and often nondoctrinal, but the realists nevertheless treated law as a coherent object of study that they saw as affected by outside forces.Footnote ²³ It was only in the 1970s and 1980s that scholars began to examine law using the broader toolset of critical theory and to situate it within larger discourses of cultural power.

The work that came out of this engagement – commonly known as Critical Legal Studies, or CLS – can be thought of not as a rival to realism but as a fusing of realist ideas about law with poststructuralist analysis of knowledge and society. Accordingly, while CLS is sprawling in its approach and critiques, it orbits around a few core descriptive and social commitments.Footnote ²⁴ The most important of these for my project is Unger’s critique of “formalism,” defined as “[a] commitment to, and therefore also a belief in the possibility of, a method of legal justification that can be clearly contrasted to open-ended disputes about the basic terms of social life, disputes that people call ideological, philosophical, or visionary.”Footnote ²⁵ Like Unger, I see law as one of many discourses people use to contest social questions, and the separation between them as one of genre more than anything else.Footnote ²⁶ Law is one language with which to tell stories about the world, which plays a particular role in organizing both Roman and contemporary society but which cannot be separated, or studied in isolation, from those societies. In more recent years, a growing strand of scholarship has considered how law and other forms of narrative work together to structure the symbolic order undergirding social life.Footnote ²⁷ For specialists in contemporary law like Cover (who developed the concept of legal narrative studies), legal sociologists like Fitzpatrick, or more explicitly normative theorists like Matsuda, legal narratology offers a new angle on the world in which we live.Footnote ²⁸

Romanists do not have it so easy. Laws – or texts produced and transmitted in explicitly legalizing milieux – make up an enormous part of our evidentiary corpus, and we often rely on their testimony in imagining the world to which they responded and over which they purported to govern. As a result we often take that testimony as fact, without quite meaning to. Nobody thinks that when Constantine I (r. 306–337) claims that he will personally hear every complaint lodged against his administration he is speaking the literal truth, but we often imagine imperial or juristic writing as regularized, general, and socially powerful because it represents itself in such a fashion.Footnote ²⁹ This book argues that we can enrich the history of Roman law by approaching these narratives critically, and by adopting a less legalistic understanding of how legal texts worked in the Roman world.

That shift in perspective requires reckoning with what we mean when we talk about narrative. Law is a narrative mode of speaking, in the sense that it operates using paradigmatically narrative forms where actions lead to consequences.Footnote ³⁰ When Section 18–4-102 of the Colorado Revised Statutes states that “[a] person who knowingly sets fire to, burns, causes to be burned, or by the use of any explosive damages or destroys, or causes to be damaged or destroyed, any building or occupied structure of another without his consent commits first degree arson” it presents its prohibition as a description of events, and so do the cases that interpret it. Legislation speaks in a general and prospective mode, and adjudication retrospective and specific, but both involve people doing things and thus include material explaining the factual predicate of a given judicial process.Footnote ³¹

These narratives are an important part of how law structures a world. But law’s influence on people often extends beyond the claims made within these narratives to other murkier or less factual commitments. Roman law is a good example of this extension. Legal texts do not explicitly assert the legality that seems to mark Classical Roman culture, but the reader nonetheless comes away thinking that laws are the things that mattered there. This idea is transmitted through broader features of the world surrounding the text – through what Genette would call its epitextFootnote ³² – that often postdate its creation in a narrow sense. The passages of Paul I cited above do not look like law only because they describe legal processes; they look like law because they are collated with other similar texts that we call law, because other people engaged with those texts in the ways that we expect people to engage with law, and because those texts have been transmitted and read and taught as part of later histories of law. None of that work is narrative, exactly, but it helps represent Rome’s legal world – to us and to Romans themselves.

So what should we call it? When I try to wrap my head around this particular kind of slippery-but-potent ideological messaging, I have found it useful to look to people writing in less formalist or theoretical modes. In particular, Joan Didion’s writing on journalism and memoir (other genres that are explicitly narrative, but that do a great deal to shape readers’ worldviews outside of that narrative process) offers an account of representation-through-storytelling that emphasizes function over form.

The difference is small, but meaningful. Didion downplays the mechanics of representation and focuses instead on its effects. A story, here, is something people tell in order to accomplish a goal. Didion’s stories help the people who hear them, and perhaps more importantly the people who write them, reconcile to their conditions; they make contingency legible, and life fair.Footnote ³⁴ Didion’s own discussion of stories precedes an account of late-sixties Los Angeles marked by paranoia and dissociation, and what makes the world of “The White Album” so scary is the lack of a coherent story that can structure or explain it.Footnote ³⁵

Rome was scary too. This book examines the legal discourse of a brutal autocracy, a frightening and nonsensical world; most of the writers I discuss died violently, and the ones who survived had very good reason to worry. Roman legal literature talks around that worry by describing a political order that made more sense than the ones its authors inhabited, and by offering technical guides that have the imaginative and fabulative qualities of Didion’s captive princess.Footnote ³⁶ In other words, Roman law did not so much offer narratives that described Roman reality as tell stories that refashioned reality into something more palatable. These texts have come down to us as a law-code, but should not be understood as separate from other stories that Romans told to themselves and each other.

That said, law is not the same as poetry. Roman law worked differently than other kinds of writing, on an audience that had been acculturated to different kinds of reading; because this book is so concerned with how Romans read legal stories, it is worth taking a moment to clarify what is so special about the stories that law tells.

Law talks in multiple registers. This point is basically uncontested within legal historiography, since any close reading of a legal text reveals different modes of speech that interact with the world in ontologically distinct ways, but scholars have tended to imagine these modes in dichotomy. For example, Austin contraposes the “performative” commands of law with “constative” assertions about the state of the world, and Constable notes that actual legal documents like statutes or case dispositions frequently employ “nounlike” statements of fact and “verblike” statements of law in tandem.Footnote ³⁷ By contrast, I find it more helpful to think of law as communicating in three distinct modes. While command and description are both important to how law works in the world, they are not the whole story. In particular, both Roman and modern legal texts often employ a third register. Their audience walks away with certain ideas about the world that these texts never constate, but that they strongly imply to anyone who reads them in the way they have been taught to read law. I call this register “constructive,” and thinking through its operations helps us understand not only what Roman law did for Romans but what it appears to have done for us. This triad is not perfect – categories always blur at the margins, and apply differently in different legal systems – but it introduces terminology I will use going forward and may be helpful in contexts outside of this book.

The normative register is probably the one that first springs to mind when you read a legal text. It is a story in the jussive or potential, telling the reader how they ought to act in order to stay on the good side of a sovereign or what they can do to alter their relationships with that sovereign. Examples might include do not set houses on fire, and once you have reached the age of eighteen you may enter into the kind of relationship that might change your obligations to the state, and pay the government 37% of all income over $578,125.Footnote ³⁸ If there is any kind of cross-cultural definition of law, what separates legal texts from others is their capacity to speak in this register and to do so with authority. Positivistic definitions of law (like Shapiro’s Planning Thesis, or Hart’s secondary rules) treat law’s ability to structure the world as its constitutive feature,Footnote ³⁹ but natural lawyers like Fuller or Dworkin also invest law with a special moral force that derives from its normative capacities.Footnote ⁴⁰ In particular, law’s normative stories alter the world in ways that other kinds of commands do not. I might tell my child not to litter, for them to follow or ignore; by contrast, a law imposing a fine for littering meaningfully changes the world at the moment of its enactment. Law’s normative stories act; they “verb” in Constable’s formulation and perform in Austin’s.

One might expect these kinds of legal stories to become less interesting when the sovereign who tells them becomes less powerful, and accordingly when they become less salient as social agents. Because of Roman law’s influence on later systems, however, legal scholars remain concerned with how Roman doctrines developed over time, and how they might have come together to form a coherent system. These questions are not my questions, but they are the questions that law trains its readers to ask, and that the Corpus Iuris Ciuilis contemplates answering. The texts that come down to us appear as a system of interlocked commands, before they appear as anything else.

That said, while law acculturates its readers to look for and interpret its normative claims those are not the only claims it makes. Law also speaks descriptively, telling factual (or apparently factual) stories about the world. These claims can take a variety of forms. An American court might issue an opinion that relies in part on the assertion that certain practices were or were not widespread in earlier historical periods,Footnote ⁴¹ or that records certain fact claims about litigants’ lives in its prefatory material,Footnote ⁴² or a Roman emperor might describe a prior state of affairs in the context of his own legal production.Footnote ⁴³ This feature of legal speech is less distinctive; obviously, other genres of writing contain assertions of fact. But law trains us to read those assertions differently. The organizational structure of legal writing subordinates fact claims to normative reasoning, presenting them as explanatory or prefatory to the speech that matters. To give an example: while Sallust might deploy his factual account of the conspiracy of Catiline in the service of a broader project of moral instruction or political philosophy (as he states at Cat. 4), he still writes in a primarily descriptive register for an audience that is presumed to be interested in learning the facts as he describes them. By contrast, legal writing does not appear to contemplate being read for its facts. A reader of an imperial rescript presumes its descriptive elements to be true (in a different way than might, say, a reader of Propertian love elegy) but also unimportant, since they appear mainly as aids to understand the legal holding.

This odd hermeneutic helps law engage in different kinds of representation than do other texts. Modern Romanists often read legal texts for their descriptive claims, and those claims contain valuable information. For examples of this work one thinks of administrative historians like Corcoran and Coriat who reconstruct imperial legal production on the bases of references to that production that survive in the law, or of legal sociologists like Czajkowski, Connolly, Eckhardt, Kelly, and many others.Footnote ⁴⁴ But one of the reasons why these texts seem so useful is that their fact claims seem uncontroversial. In certain cases, that allows legal documents to make highly contestable claims about the world in the context of table-setting for a legal argument. These claims can be productively studied within histories of imperial representation but serve a unique function within those representational programs because they recede into the background of the texts in which they appear.

Finally, law tells stories in a third register. Because law speaks as if it governs reality, it primes its reader to look to law to better predict how their reality will unfold and accordingly to look at their reality as something that law actually governs. This priming is implicit – laws do not begin with “as a law, I exercise control over social life” – but can be very persuasive, particularly when a law is read by people who do not inhabit a given culture and who use law to understand how it worked. I call this register constructive, since law in this mode operates less by making assertions than by constructing a world within the imagination of its reader in which those unasserted claims appear to be true.

For example, suppose that tomorrow the Boulder City Council were to pass a law forbidding hail on Fridays. That law is obviously silly, but should also strike you as in some important way mistaken. It offers an account of the world that feels inaccurate, even though the Council never explicitly claims to control the weather. That is because the law in question constructs a world in which it works the way we expect law to work (i.e. in which the Boulder City Council is able to prevent hail), and because that constructed world is meaningfully different from the one in which we live. This law does not say anything false but strongly implies it, and does so using a rhetorical strategy that is unique to performative texts like law.

Of course, the Boulder City Council will never pass this law. While the above example might clarify the mechanics of constructive storytelling it does not explain the stakes; real people use lawlike utterance to construct worlds all the time, in ways that are subtler but speak to more legibly human concerns. Here, a useful (if somewhat grim) example comes from international human rights law. Savvy observers understand documents like the Universal Declaration of Human Rights, adopted by the General Assembly of the United Nations in 1948, as effectively hortatory and as opening new rhetorical or organizational possibilities to actors hoping to make change on the ground.Footnote ⁴⁵ But the Declaration is not written as a hortation, and represents itself as something like a rule. For example, Article 9 reads in its entirety “No one shall be subjected to arbitrary arrest, detention or exile.”

Both of the examples above issue commands over a field that the writer does not have the practical authority to alter in the ways that the document makes it appear. But both imply that the writer does possess that authority; while few Boulderites or later historians would read the law against hail and picture an omnipotent municipal government, the Declaration makes certain kinds of conduct appear to be illegal rather than simply wrong, and thus the General Assembly of the United Nations appear to exercise powers that it does not possess in the quite the way it seems to. We can see why the Assembly might wish to be seen as exercising such powers, and why it might further human welfare for them to be seen as exercising such powers; it creates that impression, however, through the rhetorical device of construction.

That device works only because these texts look like texts that perform in the world, and thus posit a world in which they could. To be more specific, while I have referred to these lawlike utterances as misleading neither one makes false assertions per se. Instead, they are what Austin would call “infelicitous.”Footnote ⁴⁶ They do not affect the world as do other similar utterances, and thus do not succeed in the regulatory work they seem to attempt. There are a number of ways for a performative utterance to be infelicitous; any performative utterance depends on the existence of an institutional framework that the utterance might serve to alter, and a predictable mechanism by which the alteration occurs. For example, a rabbi saying “I now pronounce you man and wife” has a legal effect only if the rabbi inhabits a jurisdiction in which rabbis are allowed to marry couples, and where that marriage is understood as legally significant, and furthermore if they are addressing a couple that is otherwise permitted to marry. If any of those three conditions fail, the rabbi’s statement “I now pronounce you man and wife” is infelicitous and has no more effect than it would for me to say the same in an audio version of this book.

The laws I described above – the law against hail and the Universal Declaration – work similarly. They have no performative effect because there is no “accepted conventional procedure” by which they could do what they appear to do;Footnote ⁴⁷ there is no mechanism by which the Boulder City Council can prevent hail or by which the General Assembly of the United Nations can exercise legitimate force against those who contravene the Declaration. Accordingly, the laws are not false so much as absurd. That said, most people read texts as if they are not absurd. Since these utterances depend on certain conditions to perform in the way that they look like they should, the reader assumes (often subconsciously) that such conditions hold. These assumptions are an important part of how law works in the world, but are difficult to find in a legal text; they instead inhabit worlds that the text builds around itself, in which it works the way it says it does.

This unreal quality can sometimes be hard to see, especially if laws are our main piece of evidence for dispute resolution in a given time and place. But when the world that a legal text constructs is obviously inapt, that accordingly opens up productive avenues for historical inquiry. For Romanists, this phenomenon is most evident in how legal communications change from the third to the fourth centuries. From about the fourth century on, imperial legal writing becomes much more voluminous and, frankly, histrionic. Late-antique law makes wildly inflated claims about the emperor’s power, importance, and relationship to his subjects.Footnote ⁴⁸ These claims help us think about the rhetorical force of later law, and recent work shows how law shaped perceptions of emperors and administrators during this period. For example, Dillon considers how Constantine fashioned his own image as a transformational ruler through legal rhetoric, whereas Lenski describes the same ruler but focuses more on the dialogic representational process that occurred between petitioner and sovereign.Footnote ⁴⁹ For the later fourth century, Drijvers considers law as an important aspect of the emperor Jovian (r. 363–364 CE)’s representational program, and – the first and likely progenitor of these studies – Schmidt-Hofner’s analysis of the legislation of Valentinian I (r. 364–375 CE) argues that Valentinian’s relative productivity in lawmaking was part of a broader attempt to portray the emperor as an active and responsive administrator.Footnote ⁵⁰ Legal texts from earlier periods do not employ this same rhetorical strategy, and do not exist in nearly the same volume; accordingly, historians like Tuori and Christoforou who consider emperors-as-lawmakers in earlier periods analyze more obviously ideological media of communication like portraiture.Footnote ⁵¹ I stand within that methodological tradition – this book begins by analyzing legal discussion in other genres of communication before moving onto more specific discussion of legal documents in its later chapters – but even seemingly bland or straightforward legal communication can do representational work.

That matters, because Romanists have a great deal of seemingly bland legal communications and they allow us to tell stories that might otherwise remain obscure. For example, histories of Roman political theory, like Clifford Ando’s Imperial Ideology and Provincial Loyalty in the Roman Empire or Carlos Noreña’s Imperial Ideals in the Roman West, often focus on state communications because those communications are so comparatively well preserved.Footnote ⁵² It was rare for subjects of the empire to openly discuss political desiderata, but understanding the political valence of legal talk opens up new archives from which we can perceive that discussion. Bryen’s work finds these conversations in the legal documentary record,Footnote ⁵³ but codified law preserves a different set of actors with different, and influential, concerns. My methodology ventilates the unreality at the heart of Roman law, but also shows how central law can be to histories of Roman thinking about the imperial state.

To summarize that methodology, then. In order to read law “properly” one must assimilate a number of paratextual assumptions about the legal regime within which that law operates, and those assumptions did not necessarily hold when “Roman law” was written. Roman legal texts seem to presume a coherent and socially powerful law, which was itself an object of political imagination throughout the imperial period. As a result, laws of the imperial period construct a world around themselves that was perhaps more influential than it was ever quite real. In his discussion of Constans (r. 337–350 CE) Woudhuysen mentions a danger in the use of legislation as historical archive: “It has always been worryingly easy to see government being carried on as normal in the Code and thus conclude that such and such an emperor, condemned by our other sources, was a good administrator and cannot have been so bad really.”Footnote ⁵⁴ The Corpus Iuris Ciuilis tells us what it meant for Roman government to “carry on as normal,” and constructs the administrative paradigm within which we understand Roman law to have operated. This book treats that construction as a rhetorical project, and (hopefully) helps us better understand why that project was so world-changingly successful.

Cicero and the Talking Law

The project I describe took the shape it did because Romans of the imperial period found law extremely helpful to think with, and one reason for that is that Romans of earlier periods had invested law with tremendous symbolic weight. In particular, Romans imagined law as inherently moral and antiautocratic – these features of Roman legal thinking predate the documents I discuss in this book, but do a great deal to inform the context in which those documents were produced and consumed.

The most famous example of this view of law is the story Rome told about its first written law-code, the Twelve Tables. According to the accounts of the historian Livy (3.31, 34, 57) and the jurist Pomponius (D.1.2.2.4: Enchiridion), soon after the establishment of the Roman Republic a committee of ten men was charged with writing down the laws of the Roman state for public display.Footnote ⁵⁵ Previously Roman law had been viewed as a kind of sacral arcanum, the province of a college of priests, but the Tables made law into something visible and spectacular.Footnote ⁵⁶ The Tables survive only in quotation (as at, e.g. Gell. NA 20.1.45) but are described by later sources as central to Romans’ understanding of their own state. They were posted in the central Forum of the city, were at one point memorized by schoolchildren,Footnote ⁵⁷ and were the subject of juristic commentary long after they had ceased to function as statutory law.Footnote ⁵⁸ The Tables – or more broadly the codified, prospective legal order that traced its genealogy thereto – functioned as a near-metonym for Republican political arrangements. For example, Livy transitions from his account of the monarchic period to the early Republic with “I will now describe peacetime and wartime deeds of a free Roman people, their annual magistracies, and the majesty of their laws which was greater than that of their men” (2.1.1: Liberi iam hinc populi Romani res pace belloque gestas, annuos magistratus, imperiaque legum potentiora quam hominum peragam). These arrangements were nominally democratic and marked by frequent invocations of the “Roman people” (populus Romanus),Footnote ⁵⁹ but should be better understood as an aristocratic system in which major decisions were informally negotiated by leading families. Accordingly, Republican nostalgia – a recurring theme in the early parts of this book, since that nostalgia frequently engaged with legal concepts – is mostly an elite trope. But notions of the ideal state, or more precisely of the ideal sovereign, guided how elite Romans thought about law. Those thoughts left their mark on the Corpus Iuris Ciuilis, and thus on our modern legal world.

The clearest explanation of how law worked in Republican political theory comes from Cicero. This fact is not shocking; Cicero was a politician and jurist who wrote quite a bit, and we should not take him as representative of median Roman values.Footnote ⁶⁰ But Cicero’s account of political goods spends more time on law than do others from the Republican period,Footnote ⁶¹ and it treats law differently than a modern reader might expect. For one thing, while most ancient and modern descriptions of law treat its value as essentially instrumental Cicero views law as a bonum in se, which good men ought to seek out and cherish for its own sake (1.48: per se igitur ius est expetendum et colendum).Footnote ⁶² This vision differs sharply from modern positivistic justifications of law, in which law is a good thing because it serves human flourishing,Footnote ⁶³ but also from other ancient accounts. Compare Aristotle’s consequentialist argument for legal sovereignty at Rhet. 1354b:

For Aristotle lawmaking is a superior mode of decisionmaking, but nevertheless a better means to a desirable end. Cicero, by contrast, sees it as necessarily moral and thus predicts Lon Fuller’s (much later) account of law’s “inner morality.”Footnote ⁶⁴ Cicero also predicts the (much later) jurisprudence of Ronald Dworkin, by linking ideal law to ideal governance and ideal governors. Dworkin sees law as a negotiation between several different registers of public good, and adjudication as occasionally requiring judges to evaluate several competing rationales that might influence the disposition of a hard case.Footnote ⁶⁵ Dworkin acknowledges that these requirements are demanding, and that most people cannot meet them; for a legal system to reliably instantiate Dworkin’s constitutional values it would need judges of superhuman wisdom and deliberation, if not necessarily modesty.Footnote ⁶⁶ Cicero agrees that judges are critically important, but his ideal decisionmakers are less super- than inhuman.

Other Romans had drawn this link between the character of the state and the character of its representatives. For example, Ennius, one of the earliest Latin writers to survive, claimed that “Rome stands on ancient customs and on her men” (Moribus antiquis res stat Romana virisque).Footnote ⁶⁷ But Cicero’s account focuses less on the magistrate’s moral imagination than on his obedience to broader quasi-ethical constraints. The magistrate is not a talking law (legem loquentem) because he internalizes the law’s virtue, or because his thinking reflects virtuous qualities, but because he stays in line. The relationship between law and the magistrate is supervisory (praesunt), with the ideal magistrate subordinating his individual decisionmaking to the communal will as embodied by the laws. This subordination might not seem all that controversial, since it is arguably central to our understanding of what it means to have a legal system. But given the political breakdowns and extralegal violence of the late Republic, we can see why leges loquentes might be so desirable. Cicero frames his desire for orderly public life as a rule-of-law theory, and (as Edelstein and Straumann argue) he conceives of libertas in terms of legal regularity, as opposed to participation or consent.Footnote ⁶⁸

That focus makes Cicero’s libertas, more than other glosses of the term, into the kind of good an emperor can provide. This book focuses on imperial Rome for a reason; the sources of Roman law, which come almost entirely from the imperial period, use that law to think through the distinctive paradoxes of Roman imperial sovereignty. Emperors of the Principate existed in a kind of permanent state of exception. Imperial powers were broad but undefined, and coexisted with an older set of political conventions that governed in name only.Footnote ⁶⁹ That disjunct between appearance and reality was a problem for many Romans living under the Principate, especially the elite Romans who had exercised power under the earlier regime. If the “restored republic” of AugustusFootnote ⁷⁰ was defined by meaningful elite participation in political decisionmaking, then it was a lie – other ways of thinking about libertas opened up more room for negotiation, or at least for plausible flattery. Emperors could promise to act like leges loquentes, subordinating their decisionmaking to higher normative orders, and because law had historically occupied such an important role in the Roman political imaginary it served as a potent metaphor for that subordination.

Romans of this period used law to think about other problems, and that process gave Roman law its distinctive and unreal character. The figure of the talking law has never existed; laws do not talk, and judges are people. But its shadow looms large over Roman law, however hard it might be to find in Roman history.

* * * * *

Now, at long last, it is time to say what this book actually does. It consists of seven chapters in three parts and proceeds roughly chronologically. That schema has costs and benefits. I am tracking discursive tendencies that coexisted for much of Rome’s history, and while they wax and wane over time a chronological approach may make things seem more directional than they are. It is, however, worth the risk. For one thing legal systems, particularly ones as historicizing and nostalgic as the current object of study, are strongly path-dependent. Quirks of legal writing in the third century CE take on outsized importance in the sixth, and those causal links can be hard to isolate in a more thematic argument. More broadly, the history of Roman legality is a history of crises, of “eventful temporalities” that shifted Roman thinking.Footnote ⁷¹ Those shifts are more visible in a diachronic history, even if they worked upon an ethical and social substrate that persisted over time. I also hope such a structure is less confusing. Not everyone who cares about Roman law comes to it with a knowledge of imperial history, and the political evolutions of the Principate and Dominate are easier to explain when placed on a timeline.

The book argues one basic point: that “Roman law” is better understood as a metaphorical language in which Romans talked about their state than as a law-code in the way it now appears or in the way law now more or less works. I also hope to validate some core methodological commitments, which are not unheard of in Roman legal history but which have perhaps found less purchase there than in other subfields: that Roman administration worked in tandem with state and imperial representation,Footnote ⁷² that people experienced Roman government in largely symbolic terms,Footnote ⁷³ and that those symbolics exercised world-changing power independent of changes in people’s material conditions.

Part I of this book discusses Roman law in the Principate. While we have relatively few records of Roman legal practice or education before Gaius’ Institutes in the late second century, a substantial epigraphic, literary, and artistic corpus survives that records elite concerns; that corpus records major changes in how elites understood law to relate to their lives, as well as actual regulatory shifts that were in many cases of only secondary importance to those changes of understanding.

Chapter 1 focuses on the role of law in the accession and politic of Augustus né Octavian, Rome’s first emperor (r. 31 BCE–14 CE). Since Romans tended to associate legal supremacy with Republican political virtue, Augustus had to be carefully positioned as both an enforcer and abrogator of Roman laws in order to maintain his position. Accordingly, Augustan Rome saw a complex messaging program that included Augustus sacrificing his own daughter to an ascendant moral/legal order.

Chapter 2 considers how later emperors and their interlocutors developed a new discursive type: the “emperor who obeys the law.” Both Vespasian (r. 69–79; the first “successful” emperor outside of the Julio-Claudian dynasty) and Titus (r. 79–81; the first successor to an emperor outside of the Julio-Claudian line, as well as the first emperor to succeed his biological father) presided over enormous changes to the legal framework of the Principate. Vespasian’s “authorization” by the Senate was inscribed in bronze, taking the fiction of senatorial rule that originated in the Augustan period and expressing it in the physical medium and generic form most closely associated with Republican legislation. Similarly the Edict of Titus (which renewed, en masse, all privileges granted by his late father) regularized a system of bequests that had previously been understood through the symbolic language of beneficence, while also assimilating dynastic succession to ancient principles in the Roman law of obligations and persons. These changes would become standard tropes of imperial representation, as shown by Pliny’s panegyric to Trajan (r. 98–117). Emperors could represent themselves as obeying law in order to comply with ethical notions of ideal governance even in the absence of real constraint.

Chapter 3 considers the rise of the rescript system over the second century CE and beyond. The emperor Hadrian (r. 117–134) initiated a series of legal reforms designed to centralize adjudicative power in the imperial bureaucracy. I argue that Hadrian inaugurated a model of lawmaking that derived its legitimacy from the distinctive personal qualities of the lawgiving sovereign, and which replaced the elaborate ritual praxis of earlier Roman law with a new symbolic system centered on different goods (like meritocracy and efficiency) that were broadly valued within political rhetorics of the early second century.

The system of enlightened judgeship Hadrian inaugurated would last for another seventy-five years or so, but the Severan period (193–235) posed new challenges and made Hadrianic legality untenable. Part II of the book discusses how both emperors and jurists of that time responded to its particular dynamics, and how those responses imprinted themselves upon the Corpus Iuris Ciuilis (whose constitutive documents largely date from this period and after).

Chapter 4 focuses on how imperial legal writing developed in the late Severan period, and particularly how it came to derive its force less from the personal qualities of a lawmaker than from that lawmaker’s position within suprapersonal frameworks that had their own claims to legitimacy. I focus on two features of constitutions associated with the child ruler Severus Alexander (r. 222–235): his apparent reissuance of legal opinions issued by his disgraced predecessor Elagabalus (r. 218–222) under his own name, and the unusual tendency of Alexander’s rescripts to describe him as following the opinions of his dynastic forebears. I argue that both of these practices depict the emperor not as “obeying the laws” (like Augustus and Titus) or granting attention to grateful subjects (like Hadrian), but as fulfilling an institutionally specified role in creating and interpreting Severan legal order.

Chapter 5 examines juristic writing of the same period. Nearly all of the juristic writing that makes up the Digest comes from the Severan jurists, in particular from Papinian, Ulpian, and Paul. These three men held high positions in the imperial bureaucracy and even – in one case – likely exercised imperial power. That melding of imperial and juristic careers gave rise to a distinctively “Severan” jurisprudence, which I argue has influenced our understanding of juristic work more broadly. I discuss these jurists’ tendency to treat the consequences of specific legal interpretations as integral to evaluating their goodness, which appears quite suddenly in Severan juristic writing and suggests that jurisprudence changed to match the evaluative frameworks applied by other actors within imperial bureaucracy. I also note how jurists of this period cited “sacred constitutions” or other anonymized imperial authorities in support of their claims, rather than to specific documents with specific authors. This practice reflects a similar depersonalizing process to that visible in Alexander’s writing, but here jurists subsume the imperial lawgiver into institutional, rather than dynastic, continuities.

The Principate Severus Alexander ruled would never return, and Part III of this book discusses how the analytic moves of the early third century were received over the course of the late-antique period. In addition to the imperial administration organizing along more frankly autocratic lines, the late-antique world was marked by a growing and increasingly powerful imperial bureaucracy. These changes drove shifts in legal culture, and led to the codification of a very specific version of “Roman law,” which differed in important ways from that of earlier periods but which relied for its legitimacy on imagined connection with those periods.

Chapter 6 considers how legal thinking adapted to the realities of the fourth and fifth centuries. In particular, constitutions from this period were more institutionalizing in how they described imperial legal authority, and those constitutions were then compiled into novel archival forms (multiauthorial codices, organized by subject matter) and interpreted by a growing corps of experts. These changes looked back to the juristic treatises of the Severan period, and made the bureaucratic legality for which Roman law now appears to stand tenable, while laying the groundwork for Justinian’s final codificatory project.

That project is the focus of Chapter 7. I argue there that the Corpus Iuris Ciuilis reflects Justinianic prerogatives, at a deeper level than textual redaction. Justinian presents a coherent, knowable, and static law-code as a backdrop for his own extraordinary munificence, and accordingly his compilation highlights texts that are well suited to that image. I focus specifically on the dearth of explicit conflict between rescripts within the Codex Justinianus, and on the comparative isolation of legal principles from other kinds of literary endeavor that might originally have been represented in juristic corpora.

Finally, I conclude the work by discussing Roman law’s later reception. The bureaucratic impersonality of “Roman law” has come to stand for a set of values that would have likely read as foreign to its original writers and experts. This metonymic process is perhaps most apparent in Roman legal scholarship of the mid twentieth century. Nazis and their victims, particularly refugee scholars like Fritz Schulz, used “Roman legal order” as a barely suppressed metaphor for liberal conceptions of the rule of law. The Roman law we have was embodied in the pronouncements of autocrats and the philosophical debates of their functionaries, which themselves became symbols of a liberty that none of their authors had ever known.

As I hope the summary above has made clear, this book deals with a fairly nightmarish textual tradition; as a result, editions matter. All citations to the Latin of the Digest refer to Watson et al. (1985), and the Codex Frier et al. (2016). It happens that one of our most thorough and reliable historiographical accounts of both the Augustan and Severan periods, that of Cassius Dio, has also had a difficult transmission. I rely on Boissevain (1955)’s and Cary (1927)’s divisions of books, and (following current scholarly practice) indicate the source of individual fragments in brackets when needed. All English translations are my own, unless otherwise specified. All errors are my own, or are loved as if they were.