Thank you for the Forum – it’s great to find somewhere to discuss these things .
Apologies for the long post – and thanks in advance for any replies as this is a fairly long post and a problem I have been fretting about for about 3 years or so.
My understanding of the current thinking is that the words κενός and καινός are probably homophones in 1stcentury pronunciation of Greek. I understand the frequent interchanges of the graphemes αι/ε in the papyri, however – I find it difficult to understand how κενός and καινός could be homophones. This is similar to the difficulty seeing how ὑμεῖς and ἡμεῖς in the first century could be homophones if both υ and η were to have undergone ioticisation but he end of the first century.
κενός and καινός are quite frequently used words by New Testament writers and there are important changes in meaning if they are mistaken for each other.
In the following examples the meaning of the following verses are quite changed by substituting one meaning for the other:
John 13:34Ἐντολὴν καινὴν δίδωμι ὑμῖν, ἵνα ἀγαπᾶτε ἀλλήλους, καθὼς ἠγάπησα ὑμᾶς ἵνα καὶ ὑμεῖς ἀγαπᾶτε ἀλλήλους.
2Cor. 3:6ὃς καὶ ἱκάνωσεν ἡμᾶς διακόνους καινῆς διαθήκης, οὐ γράμματος ἀλλὰ πνεύματος· τὸ γὰρ γράμμα ἀποκτέννει, τὸ δὲ πνεῦμα ζῳοποιεῖ.
James 2:20¶ Θέλεις δὲ γνῶναι, ὦ ἄνθρωπε κενέ, ὅτι ἡ πίστις χωρὶς τῶν ἔργων ἀργή ἐστιν;
This problem of confused pronunciation of κενός and καινόςwould be particularly difficult for “Καινῆ Διαθήκη” – particularly when early Christians were discussing the New Testament with pagan critics.
Equally, there are synonyms that could be used, such as νέος or μωρός/ ἄσοφοςand do not seem to have been used..
Do you think the pronunciation of κενός and καινός would have been?
1) Would these have been Homophones?
· Context determining the understanding. This seems to be what happens in Modern Greek. However, this seems to only be for the single case of “Καινῆ Διαθήκη”. Elsewhere the terms “νέος” and “καινούργιος” are preferred for “new”. καινός meaning new doesn’t seem to occur elsewhere in Modern Greek.κενός seems to better preserve its meaning.
2) Would these have been pronounced differently?
· This could be in the context of similar pronunciation of unstressed syllables. However, when clearly enunciated the meanings are different. Such as accept/except or affect/effect etc. For example the sentence “We want them all, except this” needs to be carefully pronounced to make it intelligible.
· These could represent different and irregular pronunciation of these graphemes in different words - such as the “o” in women and wombat in English.
· There could be different syllable stresses in κενός and καινός (such as invalid/invalid in English) – these stress differences could have been wrongly accent marked in later traditions
My personal guess would be that the ε and αι graphemes were pronounced similarly but sufficiently differently so as to be heard as distinct by the writers of the New Testament – such as /ε/and /e̞/. These could be distinguished when spoken carefully - such as for κενός and καινός – but frequently mistaken in unstressed syllables of long words- such as ποιησηται/ ποιήσητε etc. This is a part of the slow drift together of these originally distinct phonemes.
If this is the case - one would predict increasing interchange of αι for ε and ε for αι over time. Three will be decreasing use of καινός over the same time (with complementary increase in the use of synonyms such as νέος in this case). An exception would be for quotes of the LXX or New testament – or the use of “Καινῆ Διαθήκη”.
What do you all think about this? - also is there anything in either Teodorsson or Gignac I can’t find these books despite a long search for both.
Many thanks again in advance and any comments would be really gratefully received
Also, I just finished the initial completion of my database and can provide very very rough initial statistics on this in the Palestinian epigraphy and documentary material.
Before 200 CE:
αι → αι = 1431
αι → ε = 32
ε → ε = 3943
ε → αι = 5
After 200 CE:
αι → αι = 439
αι → ε = 144
ε → ε = 2750
ε → αι = 6
So there is definitely an increase in the αι → ε interchange from the Middle Roman period to the Late Roman and Byzantine periods.
This database is so exciting to me. I'm really looking forward to being able to do many many more queries and get into it! If you have any other specific searches you would like me to do, just shoot me a message :)
I have been reading James today in the Codex Sinaiticus and noticed in 2:20 there is a substitution of καινέ for κενέ
It could argued, both, here that καινέ and κενέ are homophones in the 4ht century when this was written
and
the copyist of the Codex Sinaiticus did not differentiate between καινέ and κενέ on the basis of context in this instance.
Checked the Centre for New Testament Textual Studies apparatus and the only other instance seems to be the Codex Alexandrinus.
I would note a large number of later manuscripts use κενέ - though these are nearly al 9th-14 th Century and could represent the loss of καινέ from commonly spoken greek.
Wow, you write fast and thanks for the advice – I really look forward to reading your statistical analysis of Palestinian Koine Greek.
Also interesting to find out, in terms of ideolects that Epictetus doesn’t use γινωσκω but rather uses οιδα. This may be similar to Mark and Revelation never using πορεύομαι and only usingἔρχομαι.
Thanks for the steer on looking at specific contexts - I think you are right -
I initially had thought that the sheer numbers would mean that we would still statistically detect changes. But - I had not reckoned on a number of difficulties - such as καινός being a proper name. So I like your idea of finding a specific context and will cogitate on that over the next few days or weeks.
Thanks for the reply and your time - liked the reference from John Philoponus and also a pun with a point in [fə-ˈne-tik nəʊˈteɪʃ(ə)n] !
Just to push this discussion a little further –
I think the course of the changes for the αι grapheme seem to be well laid out by Horrocks [1] - simplistically, starting from the 6th Century BC to 6h Century AD the pronunciation of the αι grapheme is initially monophthongised then, along with the switch from pitch accent to primary stress accent, drifts from /æ/ to /ε(:)/ to /e(:)/.
I guess my questions are how far along phonetic process was when the authors of the New Testament were writing.
I have been particularly thinking of the work of Depauw and Stolk [2]and their analysis of the Papyrological Navigator database for the frequency of ε/αι interchangebetween 300 B.C. to A.D. 800. This seems to show a two phase pattern of increased interchanges of ε/αι in the papyri (Graph 2, p 208) – with peak interchanges happening in the 4th Century AD. They postulate the two phase pattern possibly relates to an initial monophthongisation of αι then subsequent phonological changes. This to me implies that ε/αι graphemes had not yet become completely homophonous in the 1st Century.
I was also wondering about analysing lexeme use over time - for the reason that, as you point out:-
This would imply a couple of things
· If users are not making any attempts to modify the pragmatic usage of language to avoid confusing semantics produced by he phonological changes to homophonous pronunciation – then these changes probably have not yet happened and thus the graphemes were still being pronounced differently . For example, in the context of a reading to a 1st or 2nd Century audience of listeners, it is difficult to see how the language usage clarifies a potential semantic confusion due identical pronunciation in the following examples (to name a few). John 13:34 Ἐντολὴν καινὴν δίδωμι ὑμῖν……. James 2:20 Θέλεις δὲ γνῶναι, ὦ ἄνθρωπε κενέ……… This would imply, to me, that ε and αι are likely to be phonologically different in the 1st/2nd Century.
Is it possible to use an analysis of the lexeme use over time that become homophones after a phonological shift to decide on when the phonological change happened? The most obvious example of this to me is the use of the ημεις/ εμεις and υμεις/ εσεις pairs. This lexical change seems to happen in the early/ middle Byzantine period [3] this implies υ and η were distinct phonemes before this time. I have been looking at Papyrological Navigator and thePerseus Digital Library to count the frequency of καινός and νέος over time. Though this is taking a little more time and is a wee bit more complicated than I had initially hoped – but does look as though καινός is being used in the sense of “New” in the 4th and 5th Century. κενὸς is also being used in this period.
My specific questions to you all would be:-
Would the graphemes ε and αι been pronounced distinctly (though similarly) in the 1st / 2nd Centuries A.D. at least in some circumstances?
Does analysing the use of lexemes that become homophones to determine the timing of phonological shifts look like a fruitful line of analysis? Particularly in the case of ε/αι.
Thanks again in advance – for your time reading this and any replies and opinions gratefully received.
1. Horrocks, G., Spoken Koine in the Roman Period, in Greek: A History of the Language and Its Speakers. 2014, Wiley-Blackwell. p. 161-162.
2. Depauw, M. and J. Stolk, Linguistic Variation in Greek Papyri: Towards a New Tool for Quantitative Study.Greek Roman and Byzantine Studies, 2015. 55(1): p. 196-220.
3. Horrocks, G., Spoken Greek in the Byzantine Empire, in Greek: A History of the Language and Its Speakers. 2014, Wiley-Blackwell. p. 296.
Because you wrote such a nice long post, I want to be [kɒmpləˈmɛnt(ə)ri] here.
We tolerate homophone adjectives in English as well. Which adjective did I intend here? "Complimentary" or "complementary"? This is one of those rare context where either would work. However, it might be an unusual use of either one and I think that is really the key. If phonological changes result in confusing semantics in some contexts, the language usually shifts its pragmatic usage or innovates new forms depending on how large the contexts for confusion are. In English, I think we might just use other words to make sure we avoid the confusion in these areas of overlap between homophones. That is really where the explanation lies, in my opinion.
The evidence from both Egypt and Palestine suggests that αι and ε had come to be identified. It is likely, then, that in the small number of places where their usage would overlap, an ancient Greek would have avoided the confusion some other way.
There really are not that many situations where context is not sufficient to determine between homophones.
Your examples from modern Greek are very good.
After ημεις and υμεις came to be identified as [imis], eventually new pronouns were innovated to avoid the confusion: εμεις and εσεις, with /m/ and /s/ contrasting instead of the vowels.
So I think you are right about choosing other lexemes where possible: e.g., νεος.
We have explicit evidence that these words were viewed as homophones at least in the 6th century CE from John Philoponus in his treatise on word only differentiated by accent/stress (that doesn't necessarily mean all words in this treatise follow that pattern):
e.kappa.
(1)
καθαίρω· τὸ ἀφανίζω,
καθαιρῶ· δὲ τὸ καταστρέφω.
καθαίρω· τὸ καθαρίζω.
(2)
Καίνη· πόλις τις τῆς Πελοποννήσου καὶ ἀκρωτήριον τῆς Αἰολίδος,
καινὴ δὲ ἡ νέα.
(3)
κάκη· νόσος τις,
κακὴ δὲ ἡ πονηρά.
(4)
καινός· ὁ νέος,
κενὸς δὲ ὁ ἐστερημένος.
There is no reason to mention καινος alongside κενος if the initial vowel had not been pronounced the same.