It has only been in the past couple years that I have realized that for the most part, the vast majority of Greek reference grammars have a significant lack in terms of the claims they make. This is especially true of grammars written after the reign of the neo-grammarians. I would say most grammatical works that appeared before and during the time of the neo-grammarians are slight better on this front (depending, of course, on which grammar you’re looking, Kuhner is better in my opinion, particularly before it was revised by Blass & Gerth).
But what do I mean by this?
Well, in linguistics, going as far back as Chomsky’s 1965 monogragh, Aspects of the Theory of Syntax, we find this concept of levels of linguistic adequacy (pages 18-27 are most relevant, though the entirety of chapter one is very much worth reading. Chomsky lays out three levels of adequacy that grammatical theory should strive form: observational adequacy, descriptive adequacy, and explanatory adequacy. I phrase these below in a manner slightly different than how Chomsky himself did. This is partially because while the concepts are still significantly relevant for the our goals in grammar, a number of frameworks (including the ones that I like most) have moved away from the conception of language as a rule-based system and view language as an emergent, bio-adaptive system (that’s a mix of Bybee 2010, Givón 1999, and Hopper 1988).
Observational adequacy is the lowest level and least sufficient for a grammar. Within the realm of observational adequacy, it is sufficient to collect all the relevant data points and give them label. And there is really no way to talk about a theory of grammar as being observationally adequate at all, at least not in a practical sense. Observational adequancy, ‘observes the data correctly’ (Chomsky 1964, 29); it doesn’t not, however, attempt to account for the linguistic intuitions of the native speaker or provide a principled account for why the data is the way it is.
Descriptive adequacy is the next level up. For a grammar to be descriptively adequate, not only collects the data and categorizes them, but also constructions a principled account of why the data is the way it is. What are the rules of grammar? What are the motivations for the structure of the language? Descriptive adequacy must address these questions. A theory of grammar that is descriptively adequate should sufficiently account for all of the grammars that satisfy descriptive adequacy (at least in theory).
Explanatory adequacy is the highest level of adequacy and functions are the primary goal of linguistic theory in Chomsky’s conception. This marked a major shift in how linguists approached their craft before Chomsky. Explanatory adequacy seeks to provide a principled account of competing grammatical descriptions. As Chomsky himself explained it: “A linguistic theory that aims for explanatory adequacy is concerned with the internal structure of the device [i.e. grammar]; that is, it aims to provide a principled basis, independent of any particular language, for the selection of the descriptively adequate grammar of each language” (Chomsky 1965, 63).
This last level is essentially where all debate in the field of linguistics exists. Different linguistic frameworks assume this when they argue why their approach to the nature of human language should be preferred over some other approach–often with different fundamental assumptions behind those frameworks. For example, where mainstream generative theory takes as a basic assumption that meaning is derivative from structure (interpretive semantics), Systemic Functional Grammar begins from the extreme opposite perspective: language is first and foremost a social semiotic and as such meaning is logically prior to structure. This fact is one of the reasons why so much linguistic work feels so foreign to many people–the ivory tower of linguistics, as Carl Conrad has called it a few times. The issue is that if you’re primarily interested in the grammatical structure of a language, you’re looking thinking about observational and descriptive adequacy, while large portions of the linguistic literature do not concern themselves with those things. Many times when a piece of linguistic research deals with language data or a descriptive analysis, they’re seeking to arguing a point about explanatory adequacy. This, in turn, can make seeing the relevance of the research for the grammar of the individual language more difficult.
Why is this relevant to individual grammars of Ancient Greek? Well, one question that needs to be dealt with as we look toward the production of a new full reference grammar of the Hellenistic & Early Roman Koine, is this: what is the goal of such a volume? What degree of adequacy should we be aiming for? Most existing grammars do little more than observational adequacy. Is that sufficient for us and what we do with Greek? Perhaps it is or perhaps it isn’t.
In the next couple posts, we will be examining grammatical discussions taken from a variety of grammars new and old: BDF, Robertson & Wallace. In each case, we will evaluate the level of adequacy achieved by each grammar (whether consistently or inconsistently).
Don’t worry, I wrote the other posts before I published this one so that I wouldn’t have to worry about forgetting to get the following posts up, too. I had originally planned on this being one post, but then it got too large so I split it up. Even still, I’m going to stagger these posts over the next couple weeks simply for spacing reasons. It’s been so long since I’ve written substantive content that I want to ease back into things.
In the meantime, I would be interested in any thoughts from whatever audience I still have left on this blog in terms of what sorts of research questions a fresh reference grammar of the Hellenistic and Early Roman Koine should be aiming to answer.