There is relatively little written in the grammars on the idiomatic use of εἰ μή. I spent some time working through all the examples in Mark this evening; here’s my summary. It’s limited to Mark and to the use as an idiom (i.e., it does not include use in a conditional statement, etc.).
What have I missed? Or botched?!
εἰ μὴ is an idiom which means “except.” It is used in slightly different ways grammatically. One common construction is seen in Mark 2:7 in which a noun (or sometimes a string of nouns) is appended in either nominative or accusative case which functions as the subject or direct object of an elliptical statement restating the preceding clause and which assumes the same verb, but which states an exception to that first clause. In this verse, the elliptical statement is, “One is able to forgive sin—God.” For other similar statements using a nominative, see 10:18 and 13:32; for an accusative, 5:37; 6:8; 11:13. Other times, but less commonly, the exception is fully stated (e.g., 6:5; 9:9; 13:20), or a prepositional phrase is used (e.g., 6:4), which sometimes repeats the same preposition as the first clause, but specifies a subset excluded from the original statement (e.g., 9:29). (There are other uses of εἰ μή that do not occur in Mark; see Smyth, §2346; BDF, §376; MHT 1:171.)
For reference, here’s Mark 2:7, Τί οὗτος οὕτως λαλεῖ; βλασφημεῖ· τίς δύναται ἀφιέναι ἁμαρτίας εἰ μὴ εἷς ὁ θεός;
Following up on Bill Comb’s comment (see comments below), here is an extract from Boyer’s article on εἰ μή in GTJ.
I think Thrall’s book on particles deals with this.
And I could scan and e-mail you some pages from Gerry Wakker’s _Conditions and Conditionals in Ancient Greek_, if you’d like.
I don’t see anything in Thrall, but there is no Greek index, so I may have missed it. (There is one note on a much longer construction that occurs 5x in Luke, εἰ δὲ μή γε.) I can get Wakker from our library after spring break. If you can give me a page ref, that would save some time. And if you have spare time and want to scan a relevant pg or 2, I won’t complain!
Wakker’s got a great Greek index, but off hand it looks like pages 283-85 or so.
I wouldn’t be able to scan it until Monday when I get back to the library and the scanner. If that’s still of help, let me know.
If you can find a copy at a decent price the book is worth owning (I got mine for something like $65) — but then I don’t have access to as nice of a Greek grammar library as you. Trinity Western is slightly lacking in monographs on Classical Greek grammar.
Boyer had an article in the Grace Journal in 1983 where he discussed ei mn.
Thanks Bill. Problem with working at home is that all my index files are at the seminary as well as my back run of periodicals. But Boyer’s articles have all been posted online (if in an awkward font format).
http://faculty.gordon.edu/hu/bi/Ted_Hildebrandt/New_Testament_Greek/Text/00-GreekArticlesWebBib.htm
I’ve converted the relevant excerpt on εἰ μή and appended it to the main post above. Boyer appears to want to treat most instances as elliptical conditional statements. I’m not sure about that, but I just now got the article into readable format, so I’ll have to spend some time on it.
In most instances, the meaning of the text is accurately conveyed if one sticks to the literal meaning of EI MH, which is “if not”.
No, “if not” is not a “literal meaning” of εἰ μή, it is merely a wooden gloss of two separate words without recognizing the idiom involved. That it may sometimes manage to communicate the correct meaning in a particular context does not mean that it has therefore been understood or defined correctly. Read BDAG’s entry carefully, s.v., εἰ, 277–79, particularly division 6 where several relevant combinations are discussed.
An example or two of EI MH not rendered accurately by “if not” would be appreciated.
How about Gal 1:7, ὃ οὐκ ἔστιν ἄλλο, εἰ μή τινές εἰσιν οἱ ταράσσοντες ὑμᾶς καὶ θέλοντες μεταστρέψαι τὸ εὐαγγέλιον τοῦ Χριστοῦ. Here “if not” makes no sense; this is the “exceptive” sense in which εἰ μή means “but.”