Tests for tense and aspect

In a brief discussion published earlier this year, I noted, “Not all Greek verbs inflect as perfects.” I expanded more on this in my paper at the Southeastern Baptist Theological Seminary Linguistics and New Testament Greek Conference in April of 2019. Distributional factors such as this have often played an important role in evaluating the category of aspect in spoken languages.

One of the popular diagnostic tests for determining whether a given aspect is a progressive (as in English) or an imperfective (as in Greek) is the state predicate test: Can the aspectual form be used with state predicates? For English, where we have the more semantically specific progressive aspect, the tendency is usually, “No.” Thus for example, we would tend to say in English, The refrigerator contains a bottle of milk, using the simple present, rather than the rather odd sounding, The refrigerator is containing a bottle of milk, using the progressive present. The latter sentence would tend to receive an asterisk in the linguistics literature as a signal of ungrammaticality, as below:

a.         The refrigerator contains a bottle of milk.
b.         *The refrigerator is containing a bottle of milk

Still, we English speakers certainly continue to recognize that is containing is a natural English verbal construction. It is certainly a possible inflectional formation of the verb contain. And we could likely conceive of situations where the progressive could be used with this verb. There’s a crisis context, where contain quite naturally allows the progressive aspect: We finally got troops on the ground, and they are containing the situation. But note that this is no longer a state predicate. Here, the progressive are containing necessarily signals a degree of dynamicity that is not available for the refrigerator example in (1). Other state predicates, particularly those that exist on the boundary between predicate adjective and stative verb, are less inclined to allow such formations.* Consider, hungry, for example:

a.         Mary is quite hungry.
b.         *Mary is being quite hungry.

* For an interesting class of exceptions, see this discussion on twitter.

Since these sorts of predicating constructions do not allow an alternative dynamic reading, the result is that there is no progressive aspect for such verbs. Of course, there are other state predicates that are perfectly comfortable with the progressive aspect. These are fully verbal (rather than adjectival predicates) and they tend to be in some way temporary. The sentence Jane is sitting on the chair seems natural, but the city is sitting at the foot of the mountain is much more unusual, the simple present sits would be the natural expression. Linguists and English grammarians have worked hard over the past several decades working to understand the semantic requirements for why or why not a given stative verb might allow the progressive aspect (see Bache 1997,103-132).

Yet these sorts of questions have received far less attention for Ancient Greek. Moisés Silva (1993) mentions briefly questions of distribution from one verb to another in his response to Fanning and Porter.* More recently, Francis Pang (2016) has, to some degree, taken up this question with regard to the perfective aspect (the aorist) and motion verbs, focusing on the question of the degree telicity correlates with the perfective aspect.

*Of course, Silva perhaps goes too far also in assuming distributional patterns are somehow in opposition to semantic motivations, when he comments, “These are significant patterns of usage that may be far more determinative than the desire to convey a semantic point” (Silva 1993, 80-81).

But we do not often consider the question of whether a given verb forms a perfect in its paradigm or whether there is a paradigm gap. Of course, we are generally aware that perfects are less common than the other aspects. Indeed, every New Testament Greek student recently having purchased Wallace’s grammar for their Greek exegesis class will likely read at some point (Wallace 1997, 497), “The specific breakdown of each tense is as follows: Present-11,583; Aorist-11,606; Imperfect-1682; Future-1623; Perfect-1571; Pluperfect-86.” While the precision of these numbers invariably diverges slightly from one morphological database to the next and also depend on the Greek text used—I count 1,663 perfects and pluperfects in the SBLGNT and 1,658 in the NA28, both according to Logos Bible Software’s morphology.* Still, the point remains: perfect verbs are less common than imperfective and perfective verbs.

* I also find: 1,666 in Tischendorf, 1,654 in Westcott & Hort, and 1,693 in Robinson-Pierpont’s text.

This necessarily means that there are substantial number of verb lexemes that simply do not appear in the New Testament in the perfect. Of course, many of these do appear as perfects in other texts. The verb ἀποδέχομαι is never found as a perfect in the New Testament, but we do find it in the Literary Koine and documentary papyri (Strabo’s Geography 9.5.17 and P.Stras.131, for example). The availability of large-scale digital corpora of Greek texts means that we can more easily than ever learn new things about how tense and aspect are distributed across lexical items. Human-tagged Greek morphology databases are readily available for several million of words of text with a respectable degree of accuracy, while computer-assistant morphology and lemma tagging in Perseus Digital Library and the Thesaurae Lingua Graecae are becoming accurate enough for analysis of higher and higher degrees of precision. And their vast size means we can speak more confidently than we ever have in the past about what verbs do form perfects and what verbs do not.

It is time that we begin using our substantial corpus resources for devising syntactic and semantic tests of our own.

Works cited

Bache, Carl. 1997. The study of aspect, tense and action: Towards a theory of the semantics of grammatical categories. New York: Peter Lang.
Silva, Moisés. 1993. “A Response to Fanning and Porter on Verbal Aspect.” Pages 74-82. In Biblical Greek language and linguistics: open questions in current research. Edited by D. A. Carson and Stanley Porter. Sheffield: Sheffield Academic Press.
Pang, Francis. 2016. Revisiting Aspect and Aktionsart: A Corpus Approach to Koine Greek Event Typology. Leiden: Brill.
Wallace, Daniel. 1996. Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament. Grand Rapids, Mich.: Zondervan.