Nobody would be shocked to hear that native speakers know their language really well. They speak it natively after all. Everything that you could imagine to put in a reference grammar, they already have all of that knowledge in their heads.
So when a bible professor decides to write up an argument online for why μονογενής should be understood as meaning “only begotten” rather than “unique” and he appeals to native speaker knowledge of the language, surely we should all stop and listen to it, right?
Well, no. That’s probably not a good idea–or, at least, there are a number of caveats we need to think about first.
That’s because there is a significant difference between someone knowing a language and someone knowing about a language. To put it another way: there is a significant gap between one’s ability to speak a language and one’s ability to explain to another person why they spoke the way they did and why they used the grammatical structures they used. Now, of course, there is a sense in which the limit there is terminological: if I asked you to tell me whether the verb ‘ask’ is an object-raising predicate or an object-controlling predicate, your ability to answer would depend primarily on how much language of syntactic theory you have (for those interested, you can read about raising and control predicates on their Wikipedia page).
In linguistics, the term that gets thrown around is native speaker intuition. And in biblical studies, some who have heard the phrase have jumped on it and found it a useful concept to appeal to when they find a dead Greek speaker who wrote something that agrees with their interpretation. That sort of view of things, however, misses the key word: intuition. The moment a native speaker is talking about their language, they’re not using their intuition, they’re just talking. Now, what they’re saying is still going to be useful and important, but probably not for whatever the question at hand is. Native speakers talking about their language are providing insights about their own perceptions and biases, not insights into their language itself. The insight there is sociolinguistic. It isn’t lexical, phonological, morphological, or syntactic.
For example, if you encounter an English speaker who rebukes someone else because they used the word decimate to mean destroy, saying, “Decimate can only mean to reduce by 10%” (see John McIntyre’s recent video the topic at the Baltimore Sun, for example), such a statement says very little about the meaning of the word decimate, but a whole lot about the speaker own attitudes toward their language.
Another great example is the rule about passive voice in English. Don’t use it, they say. It makes for weak writing. But again, did we just learn something about English voice? Or did we just learn something about language education in North America and the hierarchical structure of sociolinguistic authority in English? Probably the latter. The fascinating thing about the passive voice example, is just how often the people pontificating on the topic either cannot actually recognize what a passive is or how often they, fully unaware, break their own rules (see, Geoffrey Pullum’s excellent essay in the Chronicle of Higher Education: 50 Years of Stupid Grammar Advice).
So how do we get around these difficulties? Well, the first answer is simply: Be suspicious about language statements. Do not take them at face value. In the context of Ancient Greek, such statements could be taken as an opportunity to do research. Test the claim on the actual language and see if it’s correct.
The linguist William Labov was the 20th century trailblazer for establishing principles for eliciting linguistic data in a way that did not disrupt the native speakers’ language use or introduce bad data. He was studying how the pronunciation of r varies among English speakers (Labov 1966). Rather than asking directly how they pronounced certain words (which would immediately place the speakers in a situation where their speech is non-natural), he instead went to the mall and asked various people throughout the mall where various department stores were in the building–all of which he knew ahead of time were on the fourth floor. This gave him a means of getting unaffected linguistic data. He also pretended to have misheard so that he could get a repeated statement of the response. The result? One response at natural speed and one response at a slower speed.
This is the challenge of eliciting data. It is difficult to do naturally. Field linguists studying undocumented minority languages have an even more difficult time, since they cannot elicit natural data until they’ve learned the language. So usually, the majority of work is done in the context of asking some form of How do you say… and then transcribing the answer and asking for a translation of it. This can be highly risky for getting trustworthy dat. Claire Bowern (2015) has an entire section dedicated to how to make sure one’s elicited data is reliable in this context and since most of my audience is interested in ancient languages, I won’t go into the details.
Other field linguists will say that the answer is to simply not elicit data at all: only work for corpus texts. Dixon’s (2009) view is corpus work is the only acceptable method for studying language structure and elicitation is only acceptable for the purpose of filling in morphological paradigms that have gaps from the corpus.
So what can we take away from this in terms of Ancient Greek?
- Explicit statements about lexical meaning or grammatical structure shouldn’t be trusted outright. They might be correct, but you need to do the corpus research to make sure–and you needed to do that research anyway.
- Polemic debates about the meaning of words or sentences are even less reliable. If there’s a debate, that means some other native speaker holds the opposite view. Neither can be trusted. Do the corpus research yourself.
- Assumptions about the meaning of words and sentences without argument are more interesting, particularly if there’s significant modern debate. This is more like Labov’s work–unelicited, intuitional information. If a native Ancient Greek speaker simply assumes one view with no awareness that there’s an alternative, that’s something that could be important, but negative data is still not as good as positive data. Check your corpus research and see if it coheres.
Moises Silva (2005, 27) actually provides a good summary of these points in the introduction to his Philippians commentary:
Strange as it may sound, Chrysostom, along with other Greek fathers, can be particularly helpful when he does not offer an opinion on an exegetical problem. As a native Greek speaker, his innate sense of the language—but not necessarily his conscious reflection on it—provides an important bridge between the modern commentator and the Pauline writings (with the qualification that Paul’s Greek was of course not identical to Chrysostom’s). Educated speakers are notoriously unreliable in analyzing their own language. If Chrysostom weighs two competing interpretations, his conclusion should be valued as an important opinion and no more. If, on the other hand, he fails to address a linguistic problem because he does not appear to perceive a possible ambiguity, his silence is of the greatest value in helping us determine how Paul’s first readers were likely to have interpreted the text.
Now, I cannot say that I’m comfortable with how far Silva goes here. I would emphasize that any source of linguistic information ought to be corroborated from other sources. So when Silva says, “of the greatest value,” my fear is that readers will take that to mean that such linguistic intuitions are enough to move on without confirming the conclusion with other data. That would be an unwise move.
Now then, coming back to μονογενής, the fact that the Nicene Creed’s authors chose to say γεννηθέντα is certainly interesting. Here’s the larger context (the 325 CE text):
Πιστεύομεν εἰς ἕνα Κύριον Ἰησοῦν Χριστόν, τὸν υἱὸν τοῦ Θεοῦ, γεννηθέντα ἐκ τοῦ Πατρὸς μονογενῆ, τοὐτέστιν ἐκ τῆς οὐσίας τοῦ Πατρός, Θεὸν ἐκ Θεοῦ ἀληθινοῦ, γεννηθέντα, οὐ ποιηθέντα, ὁμοούσιον τῷ Πατρί
I suppose the main question for me would be: Is it interesting in the context of John 3:16? Or it is interesting in the context of the debates about the nature of orthodoxy in the 3rd and 4th centuries? I hope that everything above makes it clear the methodologically responsible answer. Now, that doesn’t mean Denny Burk is right or wrong about the meaning of μονογενής. It just means that the way he is appealing to native speaker’s knowledge is wrongheaded.
As for myself, I can’t say my view of μονογενής. I haven’t studied it personally. The most striking point in Burk’s article was the fact that BDAG continues to include it as the first entry. But without digging more, I couldn’t say whether that’s a result of traditions of continuity in lexicography from one edition to the next or because of the data that Danker worked through.