Slaves, Prisons, Paul, Philemon, and Papyri

I. Introduction

Paul’s letter to Philemon seems to suggest Onesimus is imprisoned with Paul. One question that has preoccupied my time lately has been how common was it for a slave, i.e., a piece of property and a living tool in the ancient world, to be thrown into a prison? What do our ancient sources say? Surprisingly, both the ancient literary and non-literary records on slaves and prisons are scare. Like the subject of crucifixion for which our canonical gospels are the most detailed discussion of crucifixion in the Roman world, it seems ancient, educated people did not desire to talk about such grim topics. Even the archaeological record does not help us in our quest. According to Laura Nasrallah from Harvard Divinity School, no public prisons from the Roman world have been excavated.


II. Slave Prisons

The commonest manner in which slaves were confined was the “slave prison” (ergastulum), which was located on the grounds of the large estates of the wealthy. An ergastulum was “a room in which tied-up slaves had to spend the night.”[1] These were typically underground and had high windows from which slaves could not escape.[2]

According to the Stoic Seneca, these prisons were the proper place for a rebellious slave: “We do a fine thing, indeed, when we send a wretched slave to the ergastulum. Why are we in such a hurry to flog him at once, to break his legs straightway?” (Seneca, On Anger 3:32; LCL). Suetonius records Tiberius had to investigate the contents of the slave prisons because non-slaves were occasionally imprisoned in them: “In the meantime [Tiberius] undertook two public charges: that of the grain supply, which, as it happened, was deficient; and the investigation of the slave-prisons throughout Italy, the owners of which had gained a bad reputation; for they were charged with holding in durance not only travellers, but also those whom dread of military service had driven to such places of concealment” (Suetonius, Tib. 8; LCL).[4]

While this information is well and good, it is clear from the NT Paul is not a literal slave, though he is a slave of Christ, and thus there is no way he could have met Onesimus in one of these prisons.

III. Slaves and “Public” Prisons

Both our literary and non-literary sources on the imprisonment of slaves in some form of a “public” prison are difficult to interpret. When incarceration or arrest is discussed, the sources are vague and it is unclear where the slave is imprisoned. Moreover, the majority of our sources are later than the first century CE and it is unclear if the matters discussed were pertinent for life in the first century CE.

First, according to the church father Cyprian of Carthage, if a slave does not serve a master well, he should be imprisoned: “If you are not served [by your slave] at your whim, if he does not yield subserviently to your will, as the dictatorial and excessive enforcer of his status as a slave, you scourge him, you whip him, you afflict him frequently with hunger, with thirst, with nakedness, and with the sword and with prison” (ferro etiam frequenter et carcere affligis et crucias) (To Demetrius 8) (translation from Brent, On the Church: Select Treaties: St. Cyprian of Carthage (Crestwood, NY: St. Vladimir’s Seminary Press, 2006). Into what prison should the slave be confined? Cyprian does not use the Latin term ergastulum, slave prison. Rather, his use of the word carcer suggests Cyprian has a public prison in mind.[7] It is unclear for how long the slave was imprisoned.

Second, the fourth century CE rhetorician Libanius berated the deplorable conditions of prisons in his day. In the process, he noted some people were left in “chains” for their entire lives: “Others besides who get there by other ways, live their lives in chains (ζῶσι τὸν ἐν δεσμοῖς βίον) (Or. 45.7; LCL). Libanius also indicates slaves were confined with free people: “And among them slave and free die alike (ἀποθνήσκουσι δὲ ἐν τούτοις δοῦλοί τε ἐν ἴσῳ καὶ ἐλεύθεροι), some guilty of no offense at all, others of offenses that do not deserve death” (Or. 45.11; LCL). These references are likely to a “public” prison of some sort. At least in the fourth century CE, slaves were imprisoned with free peoples in public prisons. However, the fact remains Libanius is depicting the inhumane treatment of prisoners and it is possible that these occurrences were rare for the Roman world.

Third, an inscription from Dephi notes: “Eisias is to remain with Kleomantis for the rest of his life and do everything that she is ordered as if she were a slave (ὡς δούλα). And if Eisias does not remain or does not do what she is ordered, Kleomantis is to have the right to punish her in whatever way he wishes—by beating her or imprisoning her or selling her (ψο]φέων καὶ διδέ[ων] καὶ πωλέων)” (Fouilles de Delphes, 3, 3, No. 329, lines 4–7; translation from Wiedemann, Greek and Roman Slavery (1989)). Although Eisias is not a slave, for all practical purposes she was treated like one. One of the ways this occurred was the punishment Kleomantis could mete out. Among the various things mentioned, Eisias could be imprisoned. This seems to suggest that in Delphi slave imprisonment was as common as being beaten or sold.

Fourth, according to the following papyrus dated c. 150 CE, slaves who were put in chains were ineligible for bequests: “Bequests made to one who as a slave was put in chains and was afterward freed or who was freed when not yet thirty years old are confiscated” (Sel. Papyri 2.206=BGU 5.1210; translation from LCL). It is unclear what being put “in chains” means. Was it in an ergastulum? Or, was it in a “public” prison? Either are possible.

Fifth, the following papyrus dated c. 298 CE is a deed giving permission for the friend of the owner of a fugitive slave to go to Alexandria, find the slave, and punish him: “Aurelius Sarapammon also called Didymus, . . . to Aurelius . . . and as you are styled, of the said city of Oxyrhynchus, greeting (sic). I appoint you by this instruction as my representative to journey to the most illustrious Alexandria and search for my slave called . . . aged about 35 years, with whom you too are acquainted . . .; and when you find him you are to deliver him up, having the same powers as I should have myself, if present, to . . . ; imprison him (εἴργιν), chastise him, and to make an accusation before the proper authorities against those who harboured him, demand satisfaction. This instruction I have issued to you in a single copy, which is to be valid wherever produced, and in answer to the formal question I gave my consent (translation from P.Oxy 14.1643). Literally, this friend who functions as a bounty hunter is told to “shut up the slave.”[5] It is unclear where the slave is to be confined. Since this slave has escaped from his master’s estate in Oxyrhynchus and journeyed to Alexandria, which is at least several hundred miles apart, he is most likely to have been jailed in some form of a “public” prison and not his master’s ergastulum.

What is interesting about this papyrus is the owner of the slave calls for the punishment for those who harbored the slave. Could this be a possible reason Paul sent a letter to Philemon? That is, Philemon somehow knows Onesimus is with Paul and Paul does not want to be found guilty of harboring a slave?

Finally, another papyrus from Egypt dated c. 325 CE indicates a fugitive slave was to be apprehended for his actions. A certain Flavius Ammonas requested a man named Flavius Dorotheus find his fugitive slave who had also stolen some items: “Flavius Ammonas, officialis on the staff of the praefect of Egypt, to Flavius Dorotheus, officialis, greeting (sic). I order and depute you to arrest (ἐπιτρέπω) my slave called Magnus, who has run away and is staying at Hermopolis and has carried off certain articles belonging to me, and to bring him as a prisoner (δέσμιος) together with the head-man of Sesphtha. This order is valid, and in answer to the formal question I gave my consent. I, Flavius Ammonas, officialis on the staff of the praefect of Egypt, have made this order” (translation from P.Oxy 12.1423).

The word the translators have rendered “arrest” may not be the best translation for ἐπιτρέπω. According to LSJ, the term means, “to transfer or to turn over.”[6] However, context best determines the definition of a word and the translators of P.Oxy have read a lot more papyri and classical sources than I ever will. Notwithstanding, Flavius is invested with power to take the slave into custody. Where would the slave have stayed during the journey from Alexandria to Oxyrhynchus? Most likely in some form of a “public” prison.

This papyrus could explain Onesimus’s situation and how he makes into to the prison in which Paul is staying. Onesimus could have stolen some items from Philemon, i.e., “if he has harmed you in any way or owes you anything” (v. 17), and, because of a letter from Philemon to a local official, Onesimus could have been arrested (cf. v. 15). Coincidentally, Onesimus could have been taken to the same “public” prison in which Paul is staying.

Notwithstanding the above evidence, it seems the majority of the time slaves were beaten, tortured, sold, or sent to the mines: “Slaves should be sentenced according to the rules applying to the punishment of the lower orders. For those crimes for which a free man [of the lower orders] is thrashed with rods, a slave must be sentenced to be whipped and returned to his owner; for those crimes for which a free man is first thrashed and then condemned to hard labor, a slave must be sentenced to be whipped and then to be returned to his owner to be kept in chains for the same period of time. If a slave who has been sentenced to be returned to his owner to be kept in chains is not in fact taken by him, he is to be sold; and if he can find no one who is prepared to buy him, he should be sentenced to hard labor for life” (Digest 48.19.10; translation from Wiedemann, Greek and Roman Slavery, 30).

IV: Conclusion

Our scant, extant evidence indicates slaves were imprisoned outside of private slave prisons in the Roman world. Save for being a fugitive slave and/or stealing items from one’s master, we have no idea for what other reasons, for how long, etc. slaves were imprisoned. As noted in the introduction, this paucity of evidence is probably due to the fact of the social status of those from antiquity who composed our extant documents. The relative silence of our sources could indicate the imprisonment of slaves in public prisons was axiomatic in the Roman world and ancient authors were not compelled to discuss such a common occurrence. Nevertheless, the little papyrological evidence I was able to marshal, along with the commonest function of prisons in the Roman world, indicates a slave’s imprisonment was short term and only until the owner either came and repossessed his/her slave or the slave was punished.[8]

This could be one of the reasons for the composition of Philemon. Thus from the papyri my reconstruction is as follows: Onesimus stole certain items from his master and either ran way or went AWOL while on a trip for Philemon (“If he has wronged you or owes you anything” (v. 18)). Philemon sent a representative to the city where Onesimus fled to retrieve him. The representative had Onesimus arrested and coincidentally (“Perhaps for this reason he was separated for this moment” (v. 15)) he ends up in the same prison as Paul (stranger things have happened). Paul, therefore, converts Onesimus and writes a letter to Philemon on Onesimus’s behalf. While there are difficulties with this reconstruction, given the papyrological evidence, it seems to me the most plausible.

[1]. Werner Eck, “Ergastulum,” Brill’s New Pauly (Accessed Online). Eck, “Ergastulum,” further notes: “With the acquisition of larger numbers of slaves during the Roman expansion in the 2nd and 1st cents. BC, the Roman slave owners were more frequently confronted with the fact that slaves fled or became violent against their owners. The result was an increased occurrence of slaves being tied up, who also had to do their work in that condition (compediti or vincti).”

[2]. Ibid.

[3]. Pliny the Elder notes of Roman farms: “But at the present day these same lands are tilled by slaves whose legs are in chains” (Pliny, NH 18.4).

[4]. Similarly, Augustus had to take actions against the wrong persons being imprisoned: “Many pernicious practices militating against public security had survived as a result of the lawless habits of the civil wars, or had even arisen in time of peace. Gangs of footpads openly went about with swords by their sides, ostensibly to protect themselves, and travellers in the country, freemen and slaves alike, were seized and kept in confinement in the workhouses (ergastula) of the land owners; numerous leagues, too, were formed for the commission of crimes of every kind, assuming the title of some new guild. Therefore to put a stop to brigandage, he stationed guards of soldiers wherever it seemed advisable, inspected the workhouses, and disbanded all guilds, except such as were of long standing and formed for legitimate purposes” (Suetonius, Aug. 32).

[5]. LSJ, “ἔργω.”

[6]. LSJ, “ἐπιτρέπω.”

[7]. “Prison,” OCD 1212. Sallust describes the public prison in Rome as follows: “In the prison, when you have gone up a little way towards the left, there is a place called the Tullianum, about twelve feet below the surface of the ground. It is enclosed on all sides by walls, and above it is a chamber with a vaulted roof of stone. Neglect, darkness, and stench make it hideous and fearsome to behold” (Cato 55).

[8]. Prisons were meant to be a short-term solution until either a magistrate was bribed or the prisoner was executed. “Prison,” OCD 1212; Walter Eder, “Prison Sentence,” Brill’s New Pauly (accessed online). Digest 48.18.9, “Prison indeed out to be employed for confining men, not for punishing them.” See William Buckland, The Roman Law of Slavery: The Condition of the Slave in the Private Law from Augustus to Justinian (Union, NJ: Lawbook Exchange, 1908), 404.

Review of Baden’s The Promise to the Patriarchs

Anyone concerned with Paul’s use of Gen 12:3 (Gal 3:8) and his employment of Abraham as the litmus test for a Gentile mission free from the boundary markers of Jewish/Judean identity for Gentiles should spend some time with Joel Baden’s new work: The Promise to the Patriarchs (Oxford: Oxford University Press, 2013) in order to understand the promise to which Paul refers in its literary context. This concise (it is c. 200 pages) and clear publication is a literary investigation into the promise texts of the Pentateuch from a source-critical perspective. The type of questions with which he wrestles are: (1) what is the promise; (2) what does the promise mean theologically; (3) why was it crafted; and (4) how does it affect the way we read the Pentateuch. Baden concludes that the promise texts naturally fall into the classical sources J, E, and P, each with their own emphases and perspectives. It was this diversity that the compiler of the Pentateuch attempted to maintain as he structured his sources. The promise to the Patriarchs is the glue that holds the Pentateuch together: “The story of the promise is not one among many in the Pentateuch. It is the sole story of the Pentateuch” (p. 158).

It is beyond the scope of this blog post to provide a thorough review of Baden’s book. However, I will point out a few laudable observations. Baden notes that the promise to the Patriarchs is constant throughout the entirety of the Pentateuch appearing in numerous chapters in Genesis (Genesis 12, 13, 15, 16, 17, 21, 22, 24, 26, 28, 32, 35, 46, 48, 50) and at critical moments in the story of Israel, i.e., the call of Abraham, the sacrifice of Isaac, the descent to Egypt, the death of Joseph, the call of Moses, the golden calf incident, the condemnation of the exodus generation, and the death of Moses. This promise, therefore, is the “guiding force” for the narratives of the Pentateuch and is always either explicitly or implicitly in the text, and should be interpreted as YHWH’s promise to make Israel a nation, which includes land and progeny.

As Baden delves into the promise texts within J, E, and P, he discusses their versions of the promise texts complete with distinct perspectives, themes, and emphases. For example, P’s promise texts each use ברית; contain the phrase be fruitful and multiply; the patriarchs are called גוי; they contain a connection between family and promise; they note that the fathers’ progeny will be a multitude of nations and kings; and the giving of the promise results in the change of the patriarch’s name (p. 104). J’s are directed to the patriarch alone, are always contextual (p. 119), emphasis is on the land, and there is a concern about the nations that surround Israel (Gen 26:3, 12, 24, 28, 29; 33:1; 34:30; 46:31-34) (p. 114). Finally, E’s promise contains the phrase “fear not” as the introduction to the promise (Gen 15:1; 21:17; 46:3), and is refers to Egypt (Gen 15:13-16; 46:3-4), for the promise texts in E anticipate and predict Israel’s time there (Genesis 15; 46), which functions to give identity to God’s people (Exod 20:23-23:33; 19:5; 22:20; 23:9; 23:15) (pp. 124-25).

Like the author of Hebrews, time would fail me to tell you more about this fantastic work. In sum, it is careful, thorough, concise, well articulated, and Baden’s method is economic, taking into account the most data with the fewest outliers. Most laudable is his close attention to the text of the Pentateuch. Nevertheless, there are a few aspects of the promise that I wish Baden would have addressed. First, what is the socio-historical context for the formation and accentuation of the promise texts? Was their political influence in their formation? What is the relationship, if any, of the promise texts to the northern kingdom of Israel and Judah in the south? Second, how do these promise texts related to other literature within the HB? For example, how does J’s promise text relate to Second Isaiah? How does P’s promise text relate to Chronicles and Ezra-Nehemiah? Although Baden’s methodology precludes him from addressing these questions, his work is a must read for any one working on the promise texts of the Pentateuch, the narrative of Genesis, and I would say Paul.















Going Through Hell is Finally Published

I am happy to say that after three long years, my latest article, “Going Through Hell; ΤΑΡΤΑΡΟΣ in the Greco-Roman Culture, Second Temple Judaism, and Philo of Alexandria,” Journal of Ancient Judaism 4 (2013): 352-78 [click the hyperlinked title to gain access the article], has finally been published.

In the article, I question the long-standing assumption that a divided afterlife with rewards for the righteous and punishment for wicked developed during the Persian period. On the contrary, my thesis is that Second Temple Jews, especially Philo of Alexandria, are indebted to Graeco-Roman culture and not the Bible for the theological concept of postmortem punishment for the wicked.

Scripted Prayers in Civic Religion and the Gospel of Matthew

While this blog is dedicated to studying Paul in his Graeco-Roman context, since I have lived all semester with Matthew, I felt the need to address one grave error that is popular among Matthean scholarship.

One of the things Matthew constantly disparages throughout his Gospel is pagan mores. Readers of the First Gospel are encouraged to love their enemies and thus not be like Gentiles (Matt 5:47), they are not to seek after the necessities of life like the non-Jewish nations (Matt 6:31-21), and anyone excommunicated from the churches of which Matthew was a part was considered a Gentile (Matt 18:17). Not only does Matthew belittle Gentile customs, but also he disdains the way in which the non-Jewish nations pray. Hence, Matthew’s audience is specifically instructed not pray like them: “When you pray, do not βατταλογέω like the Gentiles, for they suppose that with their many words they will be heard” (Matt 6:7; translation mine).

Most commentators suggest the Greek term βατταλογέω means to stammer non-sensically and they surmise that Matthew condemns pagans for their unintelligible prayers to their deities (cf. Sim, “Attitude to Gentiles,” 177; Luz, Matthew 1-7, 1.305; France, Matthew, 240-41; Davies and Allison, Matthew 1-7, 1.588)

The difficulty with interpreting this verse is that not only is βατταλογέω a hapax legomenon (i.e., it appears only one time) in the NT, but also βατταλογέω is a rare term in the broader Graeco-Roman world, with the majority of its subsequent uses by authors dependent upon the NT (BDAG, 172; LSJ, 311; MM, 107). As a result, I am suspect of the definition provided by the majority of commentators.

Contrary to popular opinion, I suggest that Matthew’s critique of Gentile prayer accurately reflects the manner in which prayer was offered in civic religion of Matthew’s day.

With regard to prayer in Graeco-Roman religion, Valerie Warrior, Roman Religion (Cambridge: Cambridge University Press, 2006), 18 notes, “In any kind of prayer, precautions had to be taken to ensure that no error (vitium) might prevent its fulfillment. Formulaic invocations were like passwords that opened communication with the gods. If any mistake in the prayers or the ritual should occur, the gods would not respond to the request.”

This accords well with the polymath Pliny the Elder’s description of prayer in the first century CE:

“In fact the sacrifice of victims without a prayer is supposed to be of no effect; without it too the gods are not thought to be properly consulted. Moreover, there is one form of words for getting favourable omens, another for averting evil, and yet another for a commendation. We see also that chief magistrates have adopted fixed formulas for their prayers (certis precationibus); that to prevent a word’s being omitted or out of place a reader dictates beforehand the prayer from a script; that another attendant is appointed as a guard to keep watch, and yet another is put in charge to maintain a strict silence; that a piper plays so that nothing but the prayer is heard. Remarkable instances of both kinds of interference are on record: cases when the noise of actual ill omens has ruined the prayer, or when a mistaken has been made in the prayer itself; then suddenly the head of the liver, or the heart, has disappeared from the entrails, or these have been doubled, while the victim was standing” (Pliny Hist. Nat. 28.10-11; translation from LCL).

One of the reasons for scripted prayers was the fact that the gods were not mind readers. Because Graeco-Roman religion was founded on the principle of orthopraxy and not orthodoxy, one had to spell out word for word what one wanted from them. It was not that you said something to the gods, but what you said, how you said it, and when you said it that mattered. As Pliny indicates above, there were devastating effects if prayers were not properly offered: “when a mistaken has been made in the prayer itself; then suddenly the head of the liver, or the heart, has disappeared from the entrails, or these have been doubled, while the victim was standing” (For more information see Jean-Louis Durand and John Scheid, “‘Rites’ et ‘religion’. Remarques sur certains préjudgés des historiens de la religion des Grecs et des Romains,” Archives de sciences sociales des religions 85 (1994): 23-43).

In sum, the necessity of the right words at the right time accounts for scripting prayers, which I suggest is what Matthew has in mind with his use of the term βατταλογέω and that Gentiles use numerous words (πολυλογία) as they petition their deities (Matt 6:7).

Contrary to popular Christianizing assumptions not only about Matthew, but also Graeco-Roman religion (cf. Craig Keener, The Gospel of Matthew: A Socio-Rhetorical Commentary (Grand Rapids: Eerdmans, 2009), 210, who suggests Matthew means you do not pray “to manipulate an answer from their deities”), by actually examining the way in which prayer was offered in Matthew’s day, something about which Matthew was indeed familiar, we are able to arrive at a better explanation for Matthew’s condemnation of pagan prayers.



One of the many ways in which the grassroots phenomenon of emperor worship manifested itself in the Roman world was the incorporation of the emperor into the sacred spaces of the Roman world. This inclusion of a living ruler into temples of the gods is predicated upon the manner in which the denizens of the Roman world treated their deities. Since the gods of civic religion were the first citizens of the city, it was only natural for the gods, like the other citizens of the city, to dwell together. As a result, Plutarch (Quaest. conv. 7.6) says that people frequently not only pray to the deity for whom the temple, altar, or shrine was constructed but also they petition the gods who were sharing that particular sacred space with said deity. Consequently, this incorporation of the emperor into the sacred space is evidence to the extent that the denizens treated the emperor as a god and thus believed the living ruler to be divine.

For example, on the Greek island of Cyprus a sanctuary of the island’s patron goddess Aphrodite was re-dedicated in 79/80 CE to include the worship of the living emperor Titus:

Αὐτοκράτορι Τίτωι
Καίσαρι Οὐεσπασιανῷ
Σεβαστῷ καὶ Ἀφροδείτηι
Τόπον ἱερὸν ἀπο-
κατέστησεν τὸν ἐν-
τὸς τῶν στηλῶν
ὄντα Λούκιος Βρούττιος
Μάξιμος ἀνθύματος
ἔτους δευτέρου

“To Imperator Titus Caesar Vespasianus and to the great goddess of Cyprus Aphrodite. In the second year, Lucius Bruttius Maximus, proconsul, restored [this] sacred place confined by the steles” (Amathous no. 3).

From this inscriptions a few observations are evident. First, the emperor was worshipped alongside of the patron goddess of Cyprus. While there may have been a hierarchy in the temple with placement of the cult images of Titus and Aphrodite and Aphrodite may have remained (and probably did remain) the chief deity of the temple, that the emperor was included in the inscription in the same manner as Aphrodite (both Aphrodite’s and Titus’s names are in the dative case) indicates that Titus is in an exalted position and was probably given the same honors as Aphrodite. Thus the living Titus was treated as a god and in the same manner as Aphrodite. Second, although Titus was treated as a god, he is not called one in the inscription. It is possible that since the dedicator is a high-ranking Roman official he has followed Roman protocol and intentionally not referred to the living Titus as a god (cf. Cassius Dio, Rom. Hist. 51.20.6-8). Third, contrary to the manner in which many NT scholars interpret emperor worship, this honor given to the emperor was at the behest of Lucius Bruttius Maximus and not the emperor Titus. Therefore this is evidence of the political, religious, and monetary allegiance that people within the Roman empire gave to the emperor.