Wednesday, September 21, 2011

Slavery Language in the Bible: How should it be translated?

*This is a longish post for me, so I apologize in advance. Hopefully it will be of interest to some.

I discovered through facebook that the BBC recently aired a program titled: When God spoke English: The making of the King James Bible. The program is part of the international celebration of the KJV’s 400th anniversary. Unfortunately, the program is not yet available in North America.

Part of the program features a 5 minute clip of the ESV committee working at Tyndale House. The topic of the committee’s discussion is how to translate the word slave in the Bible. Pete Williams, warden of Tyndale House, was kind enough to direct me to a youtube clip of the committee’s discussion. It provides a brief look at how translation committees work on difficult issues. I have placed the clip below and then would like to make a few remarks in response to some of the comments made.

There are few things that I would like to say in response to the way slavery language appears and is translated in the Bible.

First I want to acknowledge that this is a difficult if not near impossible task for any translator/interpreter. Not only do we have to contend with the understanding and practice of slavery in antiquity, we also must grapple with the world’s most recent experience of institutional slavery in America and the way that the Bible was used by some to support the enslaving of human beings. It is near impossible for the one not to influence the other as was noted in the clip.

Second, it is clear that the youtube clip is an edited compilation of what was clearly a longer meeting and certainly involved much more conversation than is being presented. The way the discussion moves between Hebrew Bible and New Testament makes this clear. I have every confidence that the scholars in this clip took longer than five minutes to debate how they would translate “slave” in the Bible.

With the above two caveats in mind, here are some of my thoughts on what is being said here.

  1. While it is difficult to tell from the context, it seems some are advocating that the word “servant” be used for the Hebrew word ebed rather than "slave" since otherwise one would be "slaving for God." But I wonder why this is a problem. In Exodus, the story is of Israel being released from slavery to Pharaoh. But it is not a story of freedom as much as it is a story of being transfered from slavery to Pharaoh to slavery to God. In Exodus 4:23; 7:16; 8:20; 9:1, 13; and 10:3 God, through Moses, commands Pharaoh “let my people go that they might serve me.” The term for “serve” here is the Hebrew ebed which in the context of the Exodus story is probably better understood and translated as slave like service. It seems that the Israelites were released from Egyptian slavery to be God’s slaves. Indeed, the major reason that Israelites cannot own fellow Israelites as permanent slaves is because they are God’s slaves (Lev 25:39-46). There are many more examples of this we could point to in the Hebrew Bible (e.g. 2 Chr 12:8), but the theme is also found in the New Testament. In Romans 6 the overall contrast that is set up is that believers were at one time slaves of sin but are now slaves of God (6:22). So I am not sure that there is a problem with the phrase “slave of God” or slaving for God since it seems to be what the Bible has in mind. The Bible holds up slavery to God as positive. We may feel uncomfortable with that, but to mitigate it to "service" or being a "servant" seems to weaken the contrasts being made. I would still vote for using "slave" or "slavery."

  2. I am a bit confused by the statement that the difference between a servant and a slave is whether they are owned by the master or whether they are paid by him. I am not sure that we can say that calling someone an ebed means that the individual was owned by a master rather than paid as an employee. I think of examples like 2 Samuel 2:12-13 where Abner and Joab, commanders of the armies of Ishbosheth and David, are both referred to as the ebed of their master. The context certainly implies that these men are not owned by their masters, but the term used to describe them is the same one used elsewhere in the Hebrew Bible to describe chattel slaves. Thus, while I agree with the decision by translators to label Abner and Joab as “servants,” the language is not as clear as we would like it. It is possible that aspects of institutional slavery do play into the understanding of Abner and Joab’s position, but they are also not people bought and sold. I am also not sure that once can make the argument that the Old Testament was trying to improve the life of slaves (cf. Ex 21:7-11; 20-21). The presentation of slavery in the Old Testament is pretty bleak and if we are not careful we can read in positive aspects where there weren’t any. The fact is, being a slave was rarely, if ever, a good thing.

  3. My biggest hesitation is reserved for the comments about slavery in the first century or New Testament era. It is sometimes said that the type of slavery practiced in Rome was different than that of North America in the 16th – 19th centuries. In one sense this is correct since Roman slavery was not based on race and there were more opportunities available for slaves to become free. But caution should be exercised. At times Roman slavery can be presented as a harmless institution that provided security and economic benefits to the enslaved. But it is important to remember that slavery, in whatever form or time period, is not a positive experience for the enslaved. Moreover, the notion that in the first century it was “temporary,” “often voluntary,” and “provided status and legal protection” is problematic. In response I draw on the Epilogue from my Recent Research on Paul and slavery.

    A) Most NT scholars are familiar with the thesis that individuals would sell themselves into slavery as a way to relieve themselves of debt, improve their quality of life or even as a means of social improvement. This has also been sometimes suggested as the background for Paul’s discussion of slavery to sin and God in Rom. 6:16–22 and his understanding of slavery to Christ. However, how frequently this form of enslavement was practiced is not clear. References to self-sale in the Roman jurists indicate that individuals who sold themselves into slavery had not only given up their inalienable right to freedom, but also brought shame upon themselves and their family (Digest; 40.12.1). Apart from two references in the Jurists, references to self-sale are few and obscure. Keith Bradley has noted that: “It is generally agreed that self-sale as a mode of enslavement was of negligible importance in the central period of Roman history.” (Roman Slavery, Roman Law, 483)
    B) The only clear allusion to the practice in Christian literature is found in 1 Clement 55.2. The reference is enigmatic, however, and seems to be referring more to those who were willing to suffer for others rather than suggesting that the author was familiar with the practice of self-sale. Another allusion is found in Petronius’s Satyricon where the freedman Trimalchio claims to have sold himself into slavery in order to improve his social standing. However, satire was intended to distort common social values for the purpose of comedy while reinforcing those values at the same time. The limited evidence for the practice of self-sale should serve as a caution to NT scholars. The suggestion that it was “voluntary” or even temporary has little support and creates a positive view of slavery in antiquity that probably should not be perpetuated.
    C) The belief that slaves were upwardly mobile individuals has been a common assumption among both classical and NT scholarship. However, there have been some studies in the last thirty years that have questioned not only the social mobility of slaves and freed persons, but also the social mobility of the free poor. When NT scholars focus on the social mobility that did occur among the very few who were members of familia Caesaris, they are analyzing an abnormal pattern, not one that would have been recognized by the slave population as an opportunity for social mobility. The experience of the vast majority of slaves cannot be mitigated by focusing on the unusual influence or atypical mobility of a select few. Moreover, social mobility among slaves suggests the presence of class consciousness among slaves. Bradley points out, however, that the idea of slaves having a ‘class consciousness’ of their own never developed in antiquity and that rather than admire the master’s ‘slave representative,’ all slaves, regardless of their position, would have been competing for the support and favor of the master. While some slaves were of higher rank and influence, this did not exempt them from the same type of abuse and maltreatment other slaves received. (Slavery and Society, 72-73; 152).

  4. Finally, I am a bit perplexed about the decision to render the four occurrences of doulos in 1 Corinthians 7 as “bond servant.” I am not sure what they mean by that term, but the inclusion of “servant” could suggest something less than a slave. I would argue however, that of the many places in the New Testament where doulos occurs 1 Corinthians 7 is a good place to translate it as slave. My reason is based on the fact that this is the only place in the Bible where the technical term “freedman” (a former slave) appears. Paul is purposely contrasting a slave with a former slave and he means someone owned, not paid. Interestingly, this is also the only time where, in the undisputed epistles, Paul mixes his usage of slavery language. He is clearly addressing real slaves sitting in the congregation at Corinth and he is attempting to mitigate their status by interpreting it metaphorically and referring to them as a “slave of the Lord” and “freedman of the Lord.” This suggests to me, then, that in those places where Paul calls himself a “slave of Christ” he means a slave that is “owned” like a chattel slave, not a servant.

Again, I want to emphasize that I know that the clip has been edited, and Pete Williams informs me that some of the material is out of order. I am not surprised. However, anyone watching this clip might get the wrong impression about slavery and the Bible. I think it is important to note that the topic and the challenge of translating the terminology are more complicated than is sometimes appreciated. Our experiences with slavery, particularly in America, may tempt us to sanitize the Bible of an institution that makes us uncomfortable. But if we understand the Bible in its historical setting, we see that slavery was a very real part of the world. The challenge for translators and interpreters is not to mitigate that reality while at the same time learning from it. It is ok to be uncomfortable with the Bible. It is what we do with it that is important.


  1. Dr. John,
    I've heard some folks state that a bond slave, or servant, was someone who willingly became a slave. The example cited is in Ex. 21:6 and Deut. 15:17. In these texts the volunteer slave had an awl pushed through his or her ear as a sign of the voluntary servitude. This is then equated with a person's voluntary servitude to God through Christ.
    From what you seem to be stating, this would not necessarily be accurate. Yes?

  2. Mike,

    Actually that is the case, but as in everything else, context matters. Those passages refer specifically to the practice of enslaving a fellow-Israelite permanently. Such a practice was forbidden. However, the law did make allowance for the odd occasion when a Hebrew slave wanted to stay with his master. But this does not apply to foreign slaves. So connecting this to any Roman or NT mention of slavery doesn't work since the goal of the law was to make sure that slaves, for the most part, came outside of the community rather than from within.

  3. So, for those of us who were enslaved to sin...the foreign master, cannot become bond slaves in the context of the 'awl' process. We are, in fact, redeemed from the master, (sin); purchased to become the slaves of Christ. I guess that the idea of a person voluntarily accepting the freedom price paid by Jesus confuses the matter. Folks think that because it's voluntary it equates to the 'awl' thing.

  4. Yes, one does not become free from sin for the purposes of self-autonomy but free to be the of God.

  5. Great, John. As you might expect, I agree with all of it! In particular, I think too little theological attention has been paid to the idea that Israel came out of Egypt as slaves of one owner, Pharaoh, to become slaves to their original owner, God. When I put this to a Rabbi, she was shocked and told me in no uncertain terms that these texts referred to Israel as 'servants' rather than slaves. However, I'm not sure that the ancient writers saw this distinction within 'ebed, which is why they needed to qualify it (e.g. 'Hebrew' slave = one in debt bondage). If this is right then, as you say, there are implications for the NT, and understanding terms like 'slave of Christ'.

    Interestingly though, the new version of the ESV has 'bondservants' in Ephesians 6:5-6. Although I would prefer 'slaves', this is an improvement on the previous version which has 'slaves' in verse 5, and 'servants of Christ' in 6 - ruining the parallel.

    Many thanks,


  6. John, excellent post! Thanks for taking the time to write this. We've just started a bible study on James at our place and the first night we focused on Js 1:1 and how James, an important brother in the faith, referred to himself as a doulos of God and Jesus.

    It's clear to me from that discussion that the idea of being a slave to God is challenging for a lot of people. It's loaded with so much imagery and it runs counter to how we value individualism, liberty and willfulness.

    Thanks again - I'm going to link to this post.


  7. Thanks, John. Of course the clips are from a half day discussion, preceded by a number of submitted written papers. My concern is not the NT, but the OT, where I would maintain that one should either translate all 800 occurrences of 'ebed as 'slave' (e.g. 'what does my Lord say to his slave?') or all as 'servant', and that if one chooses 'slave' you should also do the same with the verb! I think the case for 'servant' and 'serve' is far stonger, especially since there was (pre-Athenian democracy) no consistent language of human freedom to which we could oppose the status of slave and the term is always to do with subordination and isn't connected strongly with ownership or social alienation (as in the two common modern definitions of slavery). There are no 'marked/bad' and 'neutral' occurrences of 'ebed. All occurrences can be accounted for within one framework without needing to posit different meanings in different contexts.

    Exodus implies that the Israelites were (by our terms) slaves in Egypt, but it doesn't say that they were. In fact it studiously avoids calling the Israelites 'abadim and regularly applies the term to Pharaoh's henchmen. The Israelites once (5:16) call themselves 'abadim when they're trying to stress their close relationship to Pharaoh.

    To say that the narrative presents the Israelites as Pharaoh's slaves or servants is therefore to assert exactly the opposite of what the narrative is saying.

    The beth 'abadim, often rendered 'house of slavery', continues to be a 'house of subordinates/servants' when the Israelites leave. If the Israelites were 'abadim of Pharaoh, they weren't the only ones or the main ones.

    What implies that (in our terms) the Israelites were slaves is that they did 'abodah qashah 'hard work' and the many other terms describing their plight (Ex. 1, etc.). But we must not impose our categories of slavery on the narrative when it has sought to phrase things otherwise.

    Of course Israelites can look back and remember that they were 'abadim in Egypt, but they can equally look back and say that they were gerim 'sojourners'. Even outside the book of Exodus the root 'bd receives less prominence in relation to the Israelites' time in Egypt than one would expect based on modern rhetoric.

  8. Pete,

    Thanks for this. I understand your point about our perceptions, but I am not sure that we cannot assume that the Israelites were not slaves in Egypt since they are directly referred to as such only once. In addition to Exodus 5:16 I would appeal to Exodus 6:6 where God is going to redeem them from abedh. Could we translate this as "work"? I suppose, but it makes for a rather odd redemption. Freedom from working for the Egyptians as opposed to being their slaves.

    But I wonder to what degree the LXX can help us here? Since in the LXX 5:16 is translated with paides and 6:6 with douleias, is it possible that the translators understand ebed as slavery rather than simply work or service? In fact, of the 13 times douleia appears in the Hexateuch, only two do not relate to Israel in Egypt. So I wonder if Israel's slavery in Egypt is more than just an implication.

  9. John,

    I'm not saying that the Israelites weren't (by our definition 'slaves'. I'm just saying that's not the language of the narrative.

    I think that Ex. 6:6 illustrates my point you have two terms 'burdens' and 'work/servitude'. We know that the work was hard, but the ayin beth daleth root is not asked to do all the work. Lots of other terms of oppression come in (in this verse: burdens). The one term noticeable for its absence is the personal noun 'ebed.

    The Old Greek Pentateuch is a good witness to the 3rd century BC perception of 'ebed. The doul* route is relatively infrequent and other terms like therapon, pais and oiketes predominate. Later books of the OT to be rendered into Greek had the word doulos more and modern Greek has it even more.

    You can trace the occurrence of a term like doulos (in Linear B do-e-ro may even be a wealthy land-owner, though that's not certain) and it's gradually becoming more negative through time. I'm keen that we don't impose later definitions once a full slave-free contrast has developed on a pre-Classical text in Hebrew. To my mind the author did not even possess a word which reasonably lined up with our word 'slave', though the word 'ebed might line up in post-biblical Hebrew. My views are in line with those of Ingrid Riesener, who wrote a monograph on the root 'bd.

  10. Pete,

    Thanks. I take your point about reading back later perceptions.

    How then should we understand Lev 25:39-46 where the prohibition against enslaving fellow Israelites is based on God's claim that they are his slaves whom he led out of Egypt? Is there an understanding of slavery to God here in contrast to other kinds of slavery if not the kind they experienced in Egypt?

  11. I would like it if the ESV and other translations kept the same translation for 'ebed throughout this passage. This is the only way to keep to the logic of v. 42 where God says that they are my 'abadim. I think the words which really do the work and indicate that what we call 'slavery' could be involved here are not the word 'ebed, but the language of selling (vv. 39, 42 2x) and acquiring (vv. 44 and 45) and the phrase 'rule over with force' (vv. 43 and 46) which of course reminds us of Exodus 1:14). Just as in the early 19th century a US slave-owner might refer to their slave as their 'boy' or talk of buying a 'boy', we don't have reason to assume that because language of buying or selling is used here the word 'ebed itself = 'slave'. The logical link between vv. 39 and 42 suggests that the terms in these two verses are meant to have the same meaning.

    If you want to make the meaning in v. 42 'slave' then you should do it everywhere where God has an 'ebed in the OT (e.g. Josh 1:1). But if you're going to do that, why would you not call Pharaoh's henchmen his slaves?

    I'd take it that in Lev. 25 the referent may be slaves, but the meaning is not. Again, I would come back to the absence of any prominence to a contrast with 'free' -- at that stage in the ancient world no one was talked of as totally free; one could merely be free in respect of a particular thing. The abstract concept of 'freedom' probably didn't exist, but we know that the concept of 'serving' did!

    Actually the habit of using 'slave' to render 'ebed really generally arose in the 20th century. You can trace this in Romance and Germanic language Bible translations, as well as a progressive negativity in terminology (for select passages) in other European languages.

    I believe that this rose when hired servitude generally in Europe was waning and the terms for servant in many languages were perceived as archaic or belonging to a bygone era of social stratification. Consequently the number of occurrences of 'ebed rendered 'slave' shot up from zero to 100-150 in some popular English translations, Hebrew dictionaries switched 'servant' from first meaning of 'ebed to second (the Sheffield Dictionary of Classical Hebrew is particularly heinous since its first two meanings are headed by 'slave' but there are far fewer occurrences listed than the abundant attestations of 'servant' for meanings 3 and 4) and Bible dictionaries (e.g. ABD) started running articles on slavery, but didn't have articles on servants or servitude despite the fact that translations in European languages overwhelmingly felt that the most common rendering of 'ebed should be 'servant'.

    So I think the case that there has been terminological slippage during the twentieth century is strong. This has created the perception amongst people who have studied Hebrew that 'slave' is a good equivalent of 'ebed -- it may even have been the first meaning they learned as they studied Hebrew. But in order to make the case that 'ebed means slave (rather than merely may refer to slaves) I need to see more evidence.

    By contrast 'servant' brings out the relational nature of the term (a superior is always implied), the subordination, and work-agency, which is involved in all 800 occurrences of 'ebed in the OT.

    I think that all 800 occurrences could be rendered 'subordinate agent', but that's a bit of a mouthful. 'Servant' is close enough to that.

    'Slave' by contrast is less relational and focuses on status and ownership.