I first started to think about the status of satan when considering how it should be rendered in English in NT translations. I have no doubt that ho diabolos should be translated as the accuser in the NT, and that this would help eliminate the problems associated with the loaded word "devil" in English (even though it is ultimately derived from a transliteration of diabolos). I am equally convinced that hassatan/satan in the Hebrew OT should be translation as the adversary/adversary.
More recently, involvement in the diabolos debate led to additional study and thought, which in turn have led me to propose the following.
In sum, here is the argument:
In Hebrew, as in English, Proper Names (PNs) don't take the definite article. Hebrew satan usually takes the article, and in the cases where it doesn't, the usage is indefinite and should be translated as such ("an adversary").
PNs don't appear in the plural.
Hebrew satan appears once in the plural in the Old Testament, and several times in the plural in intertestamental literature, including the Dead Sea Scrolls.
The LXX translators usually chose to translate, rather than transliterate, satan/hassatan (most often as ho diabolos).
Conclusion: in the OT satan/hassatan is a technical term that does not refer to a single individual and thus should be translated, not transliterated.
This is the easy part of the argument, since many non-Christadelphian scholars would agree. The only ones who don't, are those who for theological reasons demand that the 1 Chronicles 21 example of satan must be a PN.
The NT of course is more difficult because satan is transliterated from the Hebrew, not translated as, for example, is the common noun for angel (malak > angelos).
But it seems to me that there is a good chance, given the precedent in intertestamental literature, that satan is being lifted into the NT as a (transliterated) technical term.
Here are the arguments in favour of this possibility:
1. The word satan is often used in similar ways in the NT as it is in the OT (of human adversaries, etc.).
2. Usage and parallel passages show that Greek satanas is roughly equivalent to ho diabolos, which is not a PN, but rather a substantivized adjective.
3. (And this is the big one) Greek satanas is not used in the same way NT PNs are used. To explain: while Hebrew usage is very close to that of English with respect to PNs not taking definitive articles, Greek usage is only somewhat similar to English. In NT Greek PNs *usually* don't take articles. This much can be shown from examples and the standard NT Greek grammars.
To back this up, here is some statistical evidence for this that I have accumulated:
Now contrast this with satanas:
A quick glance shows that the article usage profile is reverse in the case of ho satanas/satanas. I thought it best to test the closest parallel, which is OT PNs used in the NT, but the pattern seems to hold up with purely NT names, like John (the Apostle):
Curiously (and here I am being honest with the data from my experiment, unlike some scientists), the pattern breaks down with ho Petros (Peter), since of the 92 occurrences 59 have the article. This seems to be explained by the fact that Peter often heads the lists of the 12 disciples and because his name occurs more often on its own in constructions that give him (and hence his name) prominence.
Nevertheless, the fact remains that satan in the NT is not being used like most other PNs--and certainly not like any OT PN.
So, why is it transliterated? Possibly because satan (pronounced "sahtan" with the stress on the second syllable) is in the NT a foreign technical term, like raca and corban.
NT satan should thus be transliterated in English, but without capitalization and with the article when it appears with the article in Greek (i.e "the throne of the satan," again, pronouncing it differently to distance the term from the traditional and incorrect conception). This particular proposal I worked out with Brethren John Adey and George Booker a few years ago, and I would like to thank them for their input.
I would also like to thank Brother Steven Cox for feedback on the above this week, and for forwarding the basics of the idea to NT Greek discussion list, from which he received a response offering partial confirmation.
As we all know, those who believe in the satanic fallen angel theory see the Bible teaching that there is a single fallen angel named Satan--even in the OT. It is not difficult at all to show that this is patently untrue of the Hebrew portion of the Bible. The fact that the article is used with satanas in the NT much more like the article is used with a common, as opposed to proper, noun, coupled with the fact that there is no single satan in the NT, may suggest that the OT and NT usages are consistent or *at least* that there is no radical break between how satan is used in the Hebrew and Greek portions of God's Word.
(Articulate form: hassatan)
Translated in KJV as “adversary” (7):
Translated in KJV as “to withstand” (1):
In the Bible Hebrew proper names rarely if ever take the definite article. The noun satan appears twenty-four times in the Old Testament, ten times without the article and fourteen times with it. Of the anarthrous examples, only one is ascribed to the Satan of tradition (1 Chronicles 21:1).
All other examples that are ascribed to the Satan of tradition (such as those in the Book of Job) are articulate. In other examples of the anarthrous use of satan in the Old Testament, it is accepted that an indefinite article should be added in the English translation or another construction be used that emphasises that the word is not being used as a proper noun (Psalm 109:6; 1 Kings 5:4, 11:14,23,25). This pattern suggests that the word Satan is never a proper name in the Old Testament and that satan in 1 Chronicles 21:1 should be translated as a common noun with an indefinite article in English.
Satan in the NT:
Non-OT PNs in the NT:
Breakdown of usage of (ho) Paulos.
In Acts of the Apostles:
In Pauline and Petrine Epistles:
Breakdown of usage of (ho) Petros.
In Gospel accounts:
In Acts of the Apostles: