Stephen Carlson wrote:Well, οὐδέ basically means "and not," which is what we'd expect from both members of the compound: οὐ "not" + δέ "but/and". Now, there may be some specialization in terms of the kinds of contexts it prefers out of competition with καὶ οὐ and ὰλλὰ ού, but that issue seems to me to lean more toward pragmatics rather than lexis. One also has to take into account the pragmatics of placing οὐ first in the sentence versus later positions.
Yes I agree that this makes it harder to analyze.
Stephen Carlson wrote:LSJ says that δέ is not placed immediately after ού to avoid confusion with οὐδέ, which was my initial thinking too, but if οὐδέ historically derives from οὐ δέ and there is little appreciable difference in meaning, despite its strong preference for negative contexts, who is to say that οὐδέ isn't just an orthographic convention for writing for οὐ δέ?
So, at this moment, I'm not convinced that δέ is strongly avoided immediately after a sentence-initial οὐ. We just may spell that οὐδέ. And I'm not sure how distinguish ού δέ from ούδέ by sense in negative contexts, where we could expect either to be found.
The data that I have from the LXX+NT is merely that every of the 133 occurrences of "ουδε V" were preceded by an "ουκ" within the same verse, which I assume begins an immediately preceding negative clause, except for 9 cases which I manually checked and found the negative clause in the preceding verse, including Matt 9:17. At the same time, there are about 70 occurrences of verbal clauses of the form "X δε ... ου V" (there were a few false positives to my regular expression), and almost all of them do not have any immediately preceding negative clauses. I manually checked those in Gen-Exo and they are Gen 2:20, 15:10, 37:24, 39:8, 42:4,13, Exo 3:2, 9:32, 33:11,23. The only exception is Exo 24:2.
In short, if "ουδε" was merely a spelling for "ου δε", then I cannot see why there is no occurrence of a negated verbal clause using "ουδε" that has no immediately preceding negative clause, while there are many such clauses of the form "X δε ... ου V". This suggests that "ουδε" has a specialized meaning and thus perhaps "ου δε" isn't used because it would otherwise be confused with "ουδε".
On the other hand, it does seem that δέ is strongly avoided after (adverbial) καί in classical and Koine Greek. I can find examples of καὶ δέ in Homer and then starting in 4th century CE Greek, but between them I could only find the combination in the immensely prolific Galen and a couple of editorially restored manuscripts on TLG. So there are cases where δέ does avoid the second position with certain words in first position, but I'm not sure at this point that an initial οὐ is one of them.
I don't know about "και δε" but since there isn't a word "καιδε" avoidance of "και δε" cannot be due to the same reason that I propose for "ου δε". I can conjecture that in this case it is just a semantic matter, since "δε" denotes a switch to a different entity (a bit like "however" in English) but adverbial "και" denotes that the entity is the same in some way, and these two don't seem to quite go together. Is it really common in Homer or later?
Off-topic, I searched Google for "και δε" and found that "δε" is no longer a post-positive in Modern Greek. How did that happen?