You are here

Arcana imperii Reconsidered: Tacitus and the Ethics of State Secrecy

Matthew Taylor

This paper presents a critical reconsideration of the Latin term arcanus in the works of Tacitus, in order to offer some new insights into the historian's attitude towards the Principate as a mode of government. It will focus in particular on the four occasions where the neuter substantive “arcana” is used by Tacitus to refer to secrets related either to the state or to imperial government (Hist. 1.4.2; Ann. 1.6.3, 2.36.1, 2.59.3), twice as part of the famous phrase “arcana imperii.” This phrase was adopted from Tacitus by early modern political theorists such as Botero and Clapmar (Donaldson, Gaida), and has since enjoyed a rich history in political thought down to the present day (Agamben). Modern Tacitean scholars have thus generally taken some note of the importance of the term (Syme, Benario, Haynes), and even acknowledged that a relationship may exist between these instances of “arcana” (Damon), but have yet to treat them systematically in order to see whether they represent a consistent theoretical approach to the exercise of imperial power. While Bérard has recently considered what “arcana” meant to Tacitus in his role as a researcher of imperial history, scholars generally seem to follow Mellor in assuming that, for Tacitus, “secrecy and its attendant paranoia were a disease gnawing at the very vitals of the Roman state.”

Since the concept of arcana has proved so valuable for the Western tradition of political thought in conceptualizing the ethics of government, particularly in relation to autocratic power and utilitarian philosophy, I propose that a fuller appraisal of its meaning and bearing in Tacitus' own writings is a productive way to approach, once more, the question of how the historian may have felt about the Principate. In some sense I am arguing that the early modern writers may have read in Tacitus a more nuanced attitude towards political secrecy than our modern sensibilities commonly permit.

The paper begins by offering two sets of criteria for interpreting Tacitus' “arcana.” First, I present an abbreviated selection of representative examples from other Latin authors (Vergil Aen. 1.261–2, Horace 2.1.30–34, Livy 23.22.9, and Suetonius Tib. 43.1), gleaned from an extensive survey of all surviving instances of the term, in order to establish a general semantics of arcanus within Latin literature. Chief among these is what I call the ethical imperative of arcanus, in that it seems clear that secrets described using this word are things the authors suggest ought to be kept secret. I supplement these with a schema for approaching the concept of state secrecy outlined by the modern philosopher Sissela Bok, who has observed that secrecy in this context denotes not just the things that are kept secret, but also the principle of secrecy itself and the methods by which such secrecy is pursued. I argue that these same three classifications are in play within the semantic field of Tacitus’ writings, and that his attitude towards state secrecy therefore needs to be considered against them.

I then proceed to analyze each of the four paradigmatic instances of “arcana” in Tacitus, considering them in their individual context, but also establishing progressively a systematic intellectual and philosophical relationship between them. For example, I observe how Tacitus supplements the inherent ethical imperative of “arcana imperii” by claiming that such secrets preserve what he calls the “vis imperii” (Ann. 2.36.1), moving us towards a normative discourse of state secrecy that echoes the passage from Livy especially. Similarly, the use of “arcana” to characterize the self-serving speech of Crispus (Ann. 1.6.3.) presupposes an otherwise positive discourse of secrecy. Finally, I offer some propositions regarding the historian's attitudes to state secrecy and the Principate, which I believe complicate the prevailing picture. Whatever Tacitus' personal feelings about the emperors, this paper argues that he could acknowledge that secrecy was a necessary evil for a monarch to govern securely, and thus that it ultimately served the interest of the state.

Session/Panel Title

Arms, Secrecy, Citizenship, and the Law: State Security in the Ancient World

Session/Paper Number

60.5

© 2020, Society for Classical Studies Privacy Policy