Somehow this post ended up being 2000 words long. I’ve broken it into three smaller parts which are scheduled to be posted every other day for the next week. The ‘works cited’ list at the end is comprehensive for the entire three part series).
Part one: State predicates and the Greek perfect, Pt 1
Part two: State predicates and the Greek perfect, Pt 2
Part three: State predicates and the Greek perfect, Pt 3
I concluded the previous post with the following statement and in part 2, I want to provide an answer:
Fundamentally, the question is this: if the completive gram is inherently telic in nature, referring to an event as being brought to a complete conclusion, then how can it function as a state predicate in this manner? This usage seems to be contradictory to the very nature of completives.
From one perspective the answer actually pretty simple. And I’m a little at a loss as to why this question has come up as many times as it has. The claim about completives with state predicate simply isn’t my claim. It’s from one of my central pieces of secondary literature: Bybee, Perkins, and Pagliuca’s (1994, 74). All I did was cite it and then effectively say, “This happens in Post-Classical Greek, too.” Still, that is by no means a sufficient answer as to why this issue isn’t a problem. So the more involved answer is that this usage with state predicates is very clearly and easily motivated analogically via the metaphor. It is perhaps best illustrated like this.
- A change of state is a process from one state to another state.
A fully completed change of state is the the strongest version of the final state.
By analogy, the strongest version of a state (i.e. intensive) is a more completed state.
Now, we can break that down with a few English examples. Consider the follow sentences.
- The towel is wet.
The towel is drying.
The towel is dry.
These three sentences present a simple (to the extent that anything is simply) process with a beginning middle and end–a more or less prototypical change of state. You have two states and you have the change from one to the other. A normal completive formed from the middle process clause would be:
- The towel is completely dried.
You can see the regular prototypical completive both includes reference to (1) the change of state itself (dried vs. drying) and (2) the total affectedness of the patient argument (the drying is finished entirely). The English verb ‘to dry’ defaults to an accomplishment predicate type, but it also has a stative usage in the form of a predicate adjective, which we see above (and repeated here momentarily). Moreover, the semantics of this state are scalar following the development of the process of drying.
- The towel is not dry.
The towel is a little dry.
The towel is partially dry.
The towel is mostly dry
The towel is almost dry.
The towel is completely dry.
Only one of these sentences is comparable to “the towel is dry” in terms of its propositional content: the final version. In turn, the normal completive version and the stative version are also propositionally similar:
- The towel is completely dried.
The towel is completely dry.
The difference between them is the asymmetrical nature of their reference. The first clause (the completive one) must necessarily refer only to a towel that was previously wet, whereas the second clause (the purely stative one) can be used to refer to any dry towel regardless of whether the towel was previously wet or not. Basically what happens with state predicates is that you have an overlap in usage that already exists and the extension of completely to refer to states that do not involve any sort of process or change is both regular and predictable. The extensions of the English adverb completely and the completive usages of the Greek perfect are parallel in how they grammaticalized with state predicates in their respective languages.
This usage of the completive perfect with state predicates is fundamentally predictable from the nature of the grammatical category itself. In turn, the flaw in the critique is the very assumption that a the contradiction in meaning (an atelic usage from a form that is otherwise inherently telic) is a problem to begin with. Nobody would protest the use of the English adverb completely when it gets used with a state (e.g. completely exhausted or completely drunk to mean extremely exhausted or extremely drunk). They would not protest it because intuitively they are already aware that the principles of language change a fundamentally distinct from the laws of logic. The latter does not government the former.
Keep that in mind as we move to part three of this series.
To be continued…
Works Cited:
Dahl, Osten. 1985. Tense and Aspect Systems. London: Basil Blackwell Press.
Rosch, Eleanor. 1975. Cognitive Representations of Semantic Categories. Journal of Experimental Psychology 104 (3): 192–233.
Vendler, Zeno 1957. “Verbs and times”. The Philosophical Review 66 (2): 143–160.)
Koine-Greek 