The majority of the writing I’ve done about Ancient Greek has focused on grammatical mismatches with English. Whether it is connectives, information structure, or aspects of the verb system( wa wa wa), my work has most struck a cord is helping English-speaking exegetes develop a typologically-informed framework for better understanding Greek on its own terms instead of through an Anglo-centric lens. All was happy and fun until I happened to watch a TED Talk on linguistic relativity that absolutely shattered my understanding of mismatches, leading me down a rabbit hole of cognitive literature in an effort to rebuild my mental representation of the problem in light of this new information.#fn1

The concept of linguistic relativity was popularized by the other Whorf (not pictured at right), Benjamin, who argued that the structure and features of a speaker’s language influenced how that speaker conceptualized the world around them. One of Whorf’s most famous examples of the impact of linguistic relativity, one that took on a life of its own, was the lack of explicit markers of time in the Native American language of Hopi. See here for a quick overview of the controversy, but suffice it to say that his claims caused quite the stir, especially within popular-level misapplications of his claims. Generally speaking, it seems that Whorf’s claims went beyond what his mentor Edward Sapir would have advocated, whereas the popular-level applications of the Whorfian Hypothesis essentially out-Whorfed Whorf, as often happens with over-zealous adherents. Linguists of all stripes vehemently disagreed with these claims and sought to debunk them, with Noam Chomsky being one of the most vocal. Ekkehart Malotki did his bit to definitively refute Whorf’s claims by publishing a 600+ page tome entitled Hopi Time: A Linguistic Analysis of the Temporal Concepts in the Hopi Language that includes what he describes as the future tense-marker –ni in the verb system.#fn2 Needless to say, refutations like this debunked uncritical applications of the Whorfian Hypothesis and sidelined serious work on linguistic relativity until the late 1990s.

Then along came a growing number of cognitive scientists doing empirically-grounded research using instruments like the fMRI and non-linguistic tests to compare how speakers of different languages perform certain tasks. It’s best to stop here and go watch Boroditsky’s TED Talk, but I will summarize a couple examples for the TL;DR crowd. She and others have analyzed non-linguistic tasks like the temporal sequencing of pictures (e.g, like a boy kicking a ball or a man aging). Speakers of English organized the pictures horizontally from left-to-right, whereas speakers of Hebrew ordered them in the same order—but from right-to-left. Contrast this with speakers of Mandarin who used a different strategy with the initial image on the bottom and ascending vertically. Such studies have been replicated and expanded across a variety of factors and languages.

Another aspect of this research concerns how the presence or absence of certain features in a language affects speakers’ sensitivity to such considerations compared to speakers of languages that lacked (or made less use of ) the feature. Boroditsky’s mother tongue is Russian, an aspect-prominent language like Greek. She notes that to say something like, “The elephant ate the peanuts,” she would need to make certain decisions about that event speaking in Russian that would be left unmarked in English. You see, Russian has separate verb forms for perfective and imperfective, and use of the perfective carries implicatures of completion (i.e., the elephant eating all of the peanuts instead of just some of them with the imperfective form) that are lacking in the English version. Similarly, the presence of middle voice in a language, as in Spanish, seems to make speakers much more sensitive to agency and causality compared to English. Generally speaking, it seems that the English default is to specify an agent unless there is some reason not ot, even if it is not meaningful (e.g., I broke my arm, even if it happened accidentally falling out of a tree). Contrast this with Spanish where prototypically the agent is not specified unless it is meaningful. Hence, using an active voice in Spanish to say “I broke my arm” would evoke connotations like Col. Flagg from MASH intentionally breaking his own arm. Greek middle voice has some idiosyncrasies in its alternation with the active that are simply are not as common in English and that are lost on most of us without having had to make the kinds of decisions (and mistakes!) that actual speakers of the language would make, like the causative/anticausative alternation in Greek. We are quite used to active-voice lemmas like φοβέω that we only find in the middle/passive in the NT, which BDAG glosses as “to be in an apprehensive state, be afraid.” Every instance in the GNT is middle or passive, so the gloss makes sense until you look at the active-voice instances like Wisd of Sol 17:9, where the causative sense of the active voice is clear:

καὶ γὰρ εἰ μηδὲν αὐτοὺς ταραχῶδες ἐφόβει, κνωδάλων παρόδοις καὶ ἑρπετῶν συριγμοῖς ἐκσεσοβημένοι

For even if nothing terrifying was frightening them, yet, scared by the passing by of wild animals and by the hissing of reptiles (LES)

The active voice is transitive, meaning the verb indicates that fear is caused by the agent (‘nothing’) in the patient (‘them’) compared to the intransitive middle/passive. I direct you to Rachel Aubrey’s contribution to the Greek Verb Revisited that describes features like this in more detail. The salient point here is that not having to make same kind of nuanced decisions about voice in English about alternations like causality as languages featuring middle voice leaves us with blind spots. We don’t know what we don’t know because of how English predisposes us to think about such things in an Anglo-centric way.

So what’s to be done to overcome this deficit that monolingual English-speakers face? Research has shown that second-language learners can acquire these sensitivities through learning to speak the language. In this sense, adopting a living language approach link Biblingo or the BLC would not primarily be about learning vocabulary or mastering morphological forms. Rather, it would be to gain the sensitivity to features of language like aspect and voice in Greek that operate in decidedly non-English ways. An old anecdote from Randall Buth has now taken on new meaning for me. He tells the story of wanting to buy a bottle of water in Greece, and telling the shop keeper as much using a present-tense form. From an Anglo-centric perspective, this made perfect sense due to the temporal context: present time. However he was taken aback with the shop keeper asked him how much water he wanted, since use of the present carried the implicature of wanting to initiate a water-purchasing relationship, or something along these lines. Instead he should have used the aorist.

There is absolutely no substitute for needing to make these kinds of decisions (or making such mistakes) in language learning. Another, less-satisfactory, alternative is the synthetic approach—thoroughly immersing yourself in the typological literature. This is not to say that the Spanish middle voice behaves precisely like Greek, but studying the middle voice across the languages that feature it will get us a damn sight closer to understanding Greek than Anglo-centric labels like “deponent.”

The long-and-short of it appears to be this: speakers of, say, an aspect-prominent language, are much more sensitive to aspectual considerations due to the constant decisions they need to make about such matters compared to speakers of tense-prominent or mood-prominent languages. So rather than being unable to do or conceptualize X, as a strict Whorfian Hypothesis might claim, it seems that the constant (lack of) decision-making about the feature affects competence and performance. This notion is bolstered by comparative research into bilinguals and their experience moving between two or more linguistic worlds. The experience in the second-language environment has the effect of building capacity to operate outside of the first-language skill set.

1. Hat-tip to my friend Dr. Fred Long for posting the talk that destroyed my life for a few weeks, can’t thank you enough!
2. There is a supreme irony here if you have read D.N.S. Bhat’s description of mood-prominent languages, but I will save an extended discussion for a future post (pun intended). Here is the punch-line suggesting that -ni might be better understood as an irrealis modal marker, and that Whorf might actually have been on to something: “As it turns out from among the numerous suffixes that the Hopi verb can select to mark the grammatical categories of aspect, mode and tense, one is specifically reserved to refer to time, or rather the sequential ordering of events or states. This temporal marker is -ni whose referential force is futurity. Its temporal function is primary; however, in many contexts i-ni also takes on a number of secondary, atemporal functions which essentially belong to the modal category (imperative, hortative, desiderative, etc.). Since no markers exist to point out present or past time, Hopi, like many other languages, can be said to be endowed with a future-nonfuture tense system.”^[52]