Bryn Mawr Classical Review

Bryn Mawr Classical Review 2005.07.24

Alexander Weiss, Sklave der Stadt: Untersuchungen zur öffentlichen Sklaverei in den Städten des Römischen Reiches. Historia Einzelschriften no. 173.   Stuttgart:  Franz Steiner Verlag, 2004.  Pp. 265.  ISBN 3-515-08383-9.  €46.00.  

Reviewed by Noel Lenski, University of Colorado at Boulder (
Word count: 3101 words

This book represents the publication of W's 2002 Bonn dissertation. Despite the fact that this version has been revised only moderately, it represents an extremely mature and fully developed piece of scholarship. The subject is satisfyingly well proportioned to a first book and was crying out to be investigated. Servi publici / δημόσιοι in the Roman empire had been previously dealt with comprehensively only in the 1897 monograph of L. Halkin and, despite some important articles and book chapters on the subject, had been largely ignored since.1 W. Eder's Servitus Publica (1980), though itself a fine work, focuses only on servi publici in the city of Rome and trails off markedly in the imperial period when, Eder assumes, public slaves were replaced by servi and liberti Caesaris.2 Because, to W's mind, Eder had already covered Rome adequately, W largely avoids the capital except where it impinges on the provinces or where he takes issue with Eder's conclusions. Rome aside, W was well aware that the hundred-year hiatus in scholarly attention to servi publici has witnessed the publication of scores of relevant inscriptions, not least the lex Irnitana, and a considerable advance in scholarship on ancient slavery.3 W takes account of both developments, though his focus is primarily trained on the epigraphy rather than new theoretical and comparative approaches to slavery. W has assembled a catalog of 397 inscriptions (305 of slaves, 92 of freedmen); these include 136 not yet included in Halkin, among which are 81 in Greek as compared to Halkin's exiguous 13; moreover, he judiciously rejects ten of Halkin's inscriptions (and a handful of others) as irrelevant or unusable. W has thus sifted and augmented the material introduced by Halkin and thereby doubled the epigraphic source base. This has allowed him to bring many issues into much sharper focus and to create the storied "standard work" on this important question.

After briefly reviewing previous scholarship, W divides the book into four chapters: 1. the acquisition of public slaves; 2. the activities of public slaves; 3. the position of public slaves in Roman society; 4. the historical development of public slavery. A brief appendix and the catalog of inscriptions follows, and the book closes with an extremely meticulous and useful quadripartite index locorum.

In the first chapter W articulates four categories of acquisition: purchase, confiscation, gift and inheritance, and natural reproduction. Evidence for each is produced, including evidence that allows W to elaborate and even rebut the extensive arguments already marshaled by Eder. Thus on purchase, where Eder, following Mommsen, had assumed that slaves could be bought by the city of Rome only through a legal fiction via locatio operis rather than emptio venditio, W shows that rubric 79 of the lex Irnitana. now proves the contrary: western colonies, and Rome probably as well, used emptiones to acquire slaves. W also uses his extensive epigraphic database to demonstrate that natural reproduction played a much larger -- probably the most important role -- in the acquisition of servi publici than Eder allowed.4 Indeed, W goes so far as to assert that the reason cities maintained servae publicae -- well attested in the epigraphy -- was primarily to ensure the natural reproduction of the familia publica.

The second, and far and away longest chapter, on the tasks of public slaves, is broken into six sub-categories. The first treats servi publici as personal retainers of magistrates. W opens with another refutation of Eder, again based on the new evidence of the lex Irnitana. Following Isidore of Seville, Eder had argued that the term limocinctus (clad in a purple-striped girdle) -- attested already in the lex Ursonensis of 44 BC (CIL 2:5439 = ILS 6087 rubric 62) -- was a generic term for servi publici. W shows, however, that the term was in fact much more circumscribed in its use, being limited to the retainers of the Duumviri and Aediles. Particularly striking in this section is W's demonstration that the Classical Greek city -- if Athens can be taken as representative -- had no such category of public slave retainer, but that Greek cities developed these in the Roman period, presumably in imitation of Roman municipal structures.

W's second category of public slave employment, urban administrators, is further divided under public finance, actores, archives, market inspection, grain supply, administration of alimenta schemes, and the so-called vilicus kalendarii Septimiani. The most commonly attested form of public slave administrator was the arcarius (treasurer), 25 of whom are known by name. We also know of six dispensatores (accountants) and three vilici summarum whose functions were closely related. W wisely resists the temptation to see in these designations some hierarchy rather than simple variation in nomenclature. Where slaves played a crucial role in managing public finances in the west, W is able to demonstrate that the cities of the east sometimes used slaves, but more regularly relied on free citizens -- often men of quite high standing -- to serve as οἰκονόμοι τῆς πόλεως. Indeed he argues that the shift to unfree οἰκονόμοι occurred only under the empire and, once again, in imitation of Roman civic structures. For servile actores publici W assembles six named instances, twice as many as Halkin, and also deploys the rich font of references from the legal and literary sources to show how prevalent they were, especially in their function as bursars. Regarding archival work, where public slaves are attested in Rome already by the second century BC, W collects not only four named slave tabularii but also six liberti rei publicae who continued to serve as tabularii after manumission. Regarding their eastern equivalent (γραμματεῖς) W moots the possibility that some slave archivists from Thyatira even adopted patronymics to highlight their status (TAM V.5.1075, 1084, 1142; cf. IGR V.4.1281, 1284). On market oversight we have only one epigraphically attested s(ervus) vil(icus) macelli (from Piacenza), but our knowledge is supplemented by a number of literary and epigraphic testimonia to the employment of public slaves in the service of aediles and ἀγορανόμοι. The one specifically named horrearius (from Beneventum) is matched by comparatively little other evidence for the management of civic grain supplies by public slave. By contrast, four named servi publici are attested in connection with alimentary programs, evidence supplemented by W to demonstrate that the brunt of administrative responsibility for these programs was born by slaves rather than the elective quaestores alimentorum. Finally, W treats an interesting inscription regarding the public slave vilicus kalendarii Septimiani from Savaria (CIL 3:4152 = ILS 7119) in which he offers important insights on the broader administration of kalendarii.

The third major category in chapter two treats the maintenance of public order. In this arena W finds no equivalent of the classical Athenian police force of Scythian archers. Indeed, though ample evidence of police activity exists from martyr acts and epigraphic testmonia, very little of it points to the use of public slaves as police. Instead they seem to have been deployed by police agents as servants, especially for the conduct of house searches, a job attested from the early Republic through the Great Persecution. Public slaves are also attested as prison personnel, most clearly in Pliny's famous letters 10.19-20. Nevertheless, where previous scholars have assumed empire-wide use of servi publici as prison guards, W -- building on the work of J.-U. Krause5 -- has argued that it was a more restricted and regional phenomenon. Again following Krause, W shows that slaves also served as executioners and, for that matter, that executioners and prison wardens were often synonymous, and often free.

In a fourth section on manual and technical labor, W points to fundamental differences between Rome and the cities in as far as Rome rarely used servi publici in banausic and menial jobs prior to Augustus and really only in aqueduct maintenance thereafter, while the cities relied more heavily on public slaves for manual labor all along. The unfree aquarii who maintained Rome's aqueducts are obviously well attested by Frontinus, but there are only three inscriptions attesting to municipal aquarii, from Philippi, Brundisium, and Venusium, a dearth of sources that W argues brings into question any serious role for public slaves in this arena. Much more commonly attested (eleven examples) are public slave plumbarii, whose job it was to produce, lay, and maintain lead water pipes -- again a job dominated by free men. Public slaves are also attested as personnel in public baths, where we even have two epigraphic testimonia of public slaves serving as surveyors (mensores) from Sipontum and Comum.

In a long section on public slaves in cult activities, W continues to break new ground by demonstrating both that Eder was wrong to assume that slaves could play no direct role in public ritual (as opposed to performing menial tasks for priests) and that Halkin was wrong to assume that they played a significant role in cultic activity in western cities outside Rome. There are in fact only two certain epigraphic attestations to public slaves operating in municipal cult activities, but both (a victimarius from Brundisium and an a sacris from Capua) indicate some direct involvement with ritual (AE 1964:138; CIL 10:3941). From the east, which had its own tradition of temple slaves quite independent of public slaves, we have much more abundant evidence for the cultic use of δημόσιοι. Evidence from Hellenistic Athens, Lagina, Lindos, Tymnos, Labraunda/Mylasa, and above all Delos is marshaled that shows slaves overseeing temple funds, maintaining buildings and grounds, keeping records of temple property and dedications, and even coordinating some cult activity. On this last, W argues convincingly that there was even a period on Delos -- after the Athenian takeover of 167 BC -- during which Athens several times imposed public slaves to serve as priests of the oriental cults in order to diffuse unrest. The realm of cult thus represents a departure from the normal pattern on two counts: first because in the cultic realm the east saw a more developed use of public slaves than the west, and second because the western cities did not here follow the example of Rome, where public slaves involved in cult activity outnumbered those in profane professions four to one.

W closes the second, long chapter with a section on the size of a familia publica. He begins with the observation, confirmed repeatedly in what precedes, that servi publici worked side by side with free laborers in virtually every task to which they were assigned. They thus never constituted the entirety of a city's public workforce. On the question of total numbers, W interprets the seemingly most useful inscription, a list recording the familia publica of Ostia (CIL 14:255 = ILS 6153), as incomplete --perhaps reflecting members of a collegium rather than the entire familia. Its 21 identifiable slaves and 35 freedmen must then be the starting point for Ostia's total. W also treats the relative numbers of occupations: his catalog lists a heavy preponderance of arcarii, followed by vilici, dispensatores, actores, and tabularii. Even here, however, many public slaves who occupied more lowly positions may have left no inscribed record of their existence, equally skewing our conclusions. W ultimately hazards a hypothetical average of c. 40 public slaves per city, but this is little more than speculation.

The third, much shorter chapter is devoted to the status of public slaves in Roman society. Here W is very much in concord with previous work which identified servi publici as quite privileged relative to other slaves and even to many ingenui, though he does modify previous arguments based on his own findings. Thus public slaves, though not generally allowed to wear the limus (as previously assumed), did receive their clothing at the expense of the state. This has become clear from rubric 79 of the lex Irnitana, which also confirms, contrary to Eder's argument, that liberti publici were expected to undertake operae. W also calculates the salary of aquarii at c. 1000HS per annum (based on figures given in Frontinus), a strikingly high figure that further confirms their status. The epigraphic testimonia also make it clear that public slaves regularly married freeborn women, albeit not under matrimonium iustum. Unlike Eder, and Mommsen before him, W does not see the regular practice of joining collegia as any particular privilege of servi publici over against private slaves, who often did the same. Above all, public slaves enjoyed the right to testamentary disposition over half of their peculium, not a negligible privilege in light of W's demonstration of the (often enviable) size of their fortunes.

The fourth and final chapter covers the historical development of public slavery. Here W highlights the quite striking differences between the Greek and Roman versions of the institution. Where the municipalities of the Latin west formulated their civic structures and the organization of their familia publica after the model of Rome, those in the east had their own structures prior to Roman encroachment. This meant that many eastern cities (particularly in Egypt and the near east) had no demosioi and those that did deployed them differently. Where eastern cities did use public slaves -- in general Greece and Asia Minor -- these already had fully developed structures for their employment (as administrative aids in finance and archiving, court servants, market watchmen, prison guards) by the time the Romans arrived. Roman involvement had only minor effects on this system -- like the development of the οἰκονόμος τῆς πόλεως on analogy with the arcarius. In the west, where more testimony is available, W builds from the assumption that servi publici adoopted new functions in the municipalities in imitation of (and therefore chronologically subsequent to) developments in Rome. As in Rome, they probably began serving in strictly religious roles, moved first into the profane realm as assistants to magistrates for conducting house searches, and then expanded from there. The choice of emperors from Augustus onward to turn over fields of responsibility formerly occupied by servi publici to servi Caesaris had a sort of trickle-down effect on the cities as well. Just as the familia Caesaris was modeled on the organization of the private household, so too that of servitus publica was modeled on the growing development of the familia Caesaris. W argues that this organization further changed as free and freed personnel gradually replaced public slaves in the civic workforce -- just as they did in the imperial bureaucracy.

There is very little to object to in a so well-researched, well-organized, and well-written book.6 On the contrary, there is very much to be praised, for W has taken a circumscribed but quite extensive problem and brought all of its aspects to life in admirable fashion. This book not only updates Halkin, it brings the scholarship on the question into an entirely new era, both in terms of meticulousness and comprehensiveness and in terms of sophistication and insight. Its wide-ranging learning and judiciousness will make it the standard work for the next century.

The only real problem worth mentioning results from the classic temptation to read too much into gaps in the source record. This is doubly true of epigraphy, where the very circumstantial specificity and reliability of inscriptions can seduce us into drawing larger conclusions than warranted about trends and patterns. The discussion of plumbarii on pages 122-5 provides a good example. W points out that the names of slave plumbarii survive in surprisingly large numbers whereas those of aquarii are relatively scanty; he thus concludes that the latter were relatively uncommon and the former relatively common. More likely this difference stems from the fact that the lead pipes that plumbarii produced and inscribed survive in much greater quantities (because mass produced and quickly inhumed in their ancient context) than the stone dedicatory inscriptions left above ground (and in much smaller numbers) by other public slaves like aquarii.

This same problem of reading too much into the gaps in our epigraphic data has also led to the single biggest oversight in the book, the general failure to take account of late antique source material, and especially to examine the Codices systematically. W is right to note that the last securely datable inscription to mention servi publici traces to AD 250.7 Despite observing, however, that the subsequent silence should not be over-interpreted because it corresponds precisely with the decline in the epigraphic habit, W allows himself to be seduced by it into believing that other sources for the institution also dropped off in the early fourth century. W would have it that the last textual references to municipal public slaves date to the early fourth century and that public slavery must have disappeared in the years to come (189-90; cf. 13; 107-9; 112-3). This conclusion is demonstrably false as I show in a forthcoming article which offers a wealth of sources for the survival of servitus publica in both Rome and the municipalities through the fourth century and down to as late as the sixth century.8 Apart from extending the timeline for public slavery, these later sources would generally have only enhanced what we can learn from earlier material. There are, however, times when they could even shed new light on issues that W must leave unresolved. At pp. 67-9, for example, W rightly points out that there were both free and slave actores. Because earlier testimonia indicate that the latter generally acted as treasurers, the former as legal representatives of the city, W concludes that this distinction was firm. Two laws of Diocletian make it clear, however, that unfree actores publici could and did act as legal representatives for cities as well (CJ 8.47.2; 8.50.3; cf. already CJ 11.37.1-2). So too, unfamiliarity with the later sources leads W to rank the παῖδας mentioned in a letter of Constantine permitting Chrestus of Syracuse to use the cursus publicus for travel to the Council of Arles as servi publici (W p. 36 n. 28 with reference to Eus. HE 10.5.23). These were, rather, imperial slaves in the employ of the cursus publicus (as W might have surmised based on his reference at p. 13 to Eder's discussion of the use of the term publicus to refer to all things associated with the public post system in late antiquity). To be fair to W, the late antique sources were simply beyond his research horizon at the time he completed the project nor does he make overstated or dogmatic claims about the decline of public slavery, which he sees simply as going extinct in some foggy period after the early fourth century (p. 190). His book should, however, be read in conjunction with my article to fill out the picture for the fourth century and beyond.


1.   L. Halkin, Les esclaves publics chez les romains (Brussels: Société belge de librairie, 1897; reprint New York: Arno Press, 1979). See also R. H. Barrow, Slavery in the Roman Empire (London: Methuen, 1928): 130-43; N. Rouland, 'A propos des servi publici populi romani'. Chiron 7 (1977): 261-78.
2.   W. Eder, Servitus Publica: Untersuchungen zur Entstehung, Entwicklung und Funktion der öffentlichen Sklaverei in Rom. Forschungen zur antiken Sklaverei no. 13 (Wiesbaden: Franz Steiner Verlag, 1980).
3.   For the lex Irnitana see J. González, 'The Lex Irnitana: A New Copy of the Flavian Municipal Law'. JRS 76 (1986): 147-241 and the relevant articles of T. Giménez-Candela, 'Una contribución al estudio de la ley Irnitana: la manumisión de esclavos municipales'. Iura 32 (1981): 37-56 and A. Weiss, 'Limocincti in Irni: Zur Ergänzgung des Duumvirnparagraphen 18 der lex Irnitana'. ZPE 135 (2001): 284-6.
4.   As already argued in the excellent study of E. Herrmann-Otto, Ex Ancilla Natus: Untersuchungen zu den 'Hausgeborenen' Sklaven und Sklavinnen im Westen des römischen Kaiserreiches. Forschungen zur antiken Sklaverei no. 24 (Stuttgart: Franz Steiner Verlag, 1994): 196-205.
5.   J.-U. Krause, Gefängnisse im Römischen Reich (Stuttgart: Franz Steiner Verlag, 1996): 255-56; 263; 305-308.
6.   Orthographical problems are sparse and usually minor, a major feat in a book so rich in detail. I note only p. 69 l. 14 missing a "wir" between "begegnen" and "in diesem"; p. 99 l. 32 for "kerkömmlichen" read "herkömmlichen"; p. 165 . 1 for "werde dürfe" read "werden dürfe"; p. 170 l. 32 for "bzeugt" read "bezeugt".
7.   AE 1933:113. W also notes that I. Aquileia 129 is likely datable to the fourth century based on letter forms. See also I. Aquileia 553 (not mentioned by W) and possibly Supplementa Italica 5.253 no. 3 = ILS 9420 (a. 323).
8.   N. Lenski, 'Servi Publici in Late Antiquity' in Die Stadt in der Spätantike, eds. J.-U. Krause and C. Witschel (Stuttgart: Franz Steiner Verlag, 2005) forthcoming.

Read Latest
Index for 2005
Change Greek Display
Books Available for Review

HTML generated at 13:30:12, Friday, 03 April 2009