Occasional Surveys in the History of Greek Grammar: The Stoics

1. Ancient Grammar

Charting the history of the study of tense and aspect in Greek in the Ancient and Byzantine grammars is a difficult task simply because of the paucity of information about context, the small number of extant manuscripts, issues of dating & authorship, and general access to the requisite materials. All of this has resulted in the contemporary discussions of ancient grammar for New Testament Greek being rather polemic. The ancient voices on the subject include the Stoic philosophers, Dionysius Thrax, Apollonius Dyscolus, and the various scholia on them.

There is an unfortunate situation when we examine New Testament scholarship that has discussed ancient scholarship on the Greek verb. There has been a failure to sufficiently discuss questions of history and authorship surrounding the grammatical treatise attributed to Dionysius Thrax. For example, contrary to what we find in the work of New Testament scholars, the fact of the matter is that the consensus in linguistic historiography rejects Dionysius’ authorship for the main body of the work and the majority of the text is probably better dated to the 2nd century CE rather than BCE.[1] Neither Porter nor Caragounis, the two New Testament scholars who discuss the ancient grammarians most thoroughly, deal satisfactorily with the issue. Both fail to adequately place the Stoics and Dionysius in context. In this light, it is extremely important for us to take seriously the observation by P. H. Matthew:

“Given the deficiencies in our sources, and the discrepancies between ancient categories of inquiry and our own, no survey of ancient linguistics can entirely escape a charge of distortion. A balanced history is not possible, or possible only at a general level.”[2]

Because of difficulties in accessing sources, much of the discussion of ancient grammars focusing as much on contemporary discussions of their work as they do on the ancient works themselves.

1.1 The Stoics

The Stoics were a philosophical movement originating in the 3rd century BCE. While they were not focused on grammar, their work was influential in the history of linguistics. The challenge is that we are limited to secondary sources for knowledge of their approach.[3] They are an essential component for Porter’s conception of the history of research alongside the brief grammar of Dionysius Thrax, particularly for his understanding of 19th century grammars.

The Stoics conception of the verb can be broken down into the diagram in figure 1.[4]

Figure 1 Past Present Future
Incomplete παρῳχειμένος παρατατικός ἐνεστώς παρατατικός
Complete παρῳχειμένος συντελικός ἐνεστώς συντελικός
Undefined ἀόριστος μέλλων

It is clear that the Stoics had both tense and aspect in mind with the distinctions between a predicate’s temporal state (as complete, incomplete, or undefined) and its temporal location (past, present, future). The former are, of course, aspectual categories and the latter are tense categories. There seems to be a small consensus that for the Stoics the tense categories are the primary ones.[5] The basic divisions are generally clear. The difference between the Present and the Perfect (and the Imperfect and Pluperfect) is one of telicity and the difference between the Perfect and Present versus the Pluperfect and Imperfect is one of temporal location. The basic claim of the Stoics with regard to the aorist and the future is not entirely clear, particularly how closely we can connect the Stoic’s ἀόριστος to the modern category of perfectivity, which involves the presentation of a given state-of-affairs as a whole. As Carlotta Smith writes, “The span of the perfective includes the initial and final endpoints of the situation: it is closed informationally.”[6] The first half of this definition might suggest that the modern conception of perfectivity is somehow distinct from the Stoic’s ἀόριστος. Alternatively, the second half of the definition (i.e. “closed informationally”) may very well share much in common with ἀόριστος, with its conception that the internal temporal makeup of the situation is inaccessible and thus undefined. At the very least, it seems that the Stoics view neither of the aorist or the future as making any reference to telicity one way or the other.[7]

One possible explanation might be that because of the telicity contrast between the present and the perfect, the Stoics (and later ancient grammarians) saw a need to place an emphasis not the inclusion of initial and final endpoints for the aorist, but on the informationally closed status. Under this reading of the Stoics, it is possible to still maintain our contemporary understanding of the aorist as perfective, while also accepting the distinction between these inflectional forms emphasized by these ancient grammarians.

Those familiar with Porter might note some major differences in terminology. Porter presents the Stoics as follows.[8]

Action Time Past (παρῳχειμένος) Present (ἐνεστώς) Future (μέλλων)
Incomplete (παρατατικός) παρατατικός ἐνεστώς
Complete (συντελικός) ὑπερσυντέλικος παρακείμενος
Undefined (ἀόριστος) ἀόριστος μέλλων

It appears that Porter has conflated the terminology of Dionysius Thrax with that of the Stoics (while discussing them separately). According to both Robins and Matthews the latter do not use the terms παρακείμενος (Perfect) and ὑπερσυντέλικος (Pluperfect), which have clearly originated from Dionysius. Porter’s scheme is based on the scholia on Dionysius Thraxe’s grammar compiled in I. Bekker, Anecdota Graeca. (Berlin: G. Reimer, 1816), 2.891-2. It is possible that this connection with Dionysius’ grammar has influenced the terminological differences between Porter as compared with Robins and Matthews. As it stands, I have not yet had an opportunity to examine the scholia in Bekker for the Stoics.

In terms of Porter’s critique of the Stoics, he makes a few points, including questioning the general completeness and usefulness of their system. He writes,

“Most obviously, the Stoics have failed to develop a complete system that elucidates all the verbal forms and functions … since they are bound within a temporal framework similar to Dionysius’s. For example, the Aorist like the Future is left undefined, and it is not compared to the Present and Imperfect, but only to the Perfect and Pluperfect. And their categories make no reference to past-referring Presents or Perfects, present referring Imperfects and Pluperfects, as well as non-Indicative usage.”[9]

As a whole, this is an odd criticism. We do not begin to see anything like the modern medium-scale grammatical treatise until after the Renaissance, if not the Enlightenment. Fundamentally, such comprehensive grammars of the language did not begin to appear until there was a greater need for non-native speakers to learn the language. And for that reason, it is likely that the Stoics saw little need to write anything remotely close to being comprehensive. It is notable that aspect of the Stoic’s system Porter views as lacking is far less relevant to the native speaker of the language than it is for linguistic researchers. In fact, considering the nature of linguistic categorization, we should expect that the native speakers would not discuss individual usages that he brings up. When native speakers talk about the meaning of grammatical items, they consistently go directly to the most basic, prototypical senses of that grammatical item.[10]

His other two criticisms continue in the same vein: noting negatively again that their conception is “temporally bound” and states that they do not adequately define their terms.[11] The latter statement continues in the same vein as before, with unnecessary expectations being placed on the Stoics. He writes, “The Stoics evidence terminological difficulty in defining the other tense forms, as seen in the repetition of terms within their conceptual framework, sometimes according to temporal reference, sometimes according to kind of action.”[12] It is not clear whether Porter just honestly does not recognize that the Stoics view both tense and aspect as relevant to the verb and that the two categories are closely integrated in their system or whether his hypothesis of Greek as tenseless has predetermined his view of the Stoics.[13]

The former criticism regarding the system being “temporally bound” marks the beginning of an ongoing mistake, in my view, that Porter makes: that tense is temporal and aspect is non-temporal.[14] Perhaps one of the most profound disconnects in Porter’s literature review is that Greek grammarians who talk about temporality and the verb are constantly condemned, but no comment is ever given when linguists discuss the temporality of aspect.[15]

[1] See, for example: Roy Harris and Talbot Taylor, Landmarks in Linguistic Thought: The Western Tradition from Socrates to Saussure (London: Routledge, 1997), 48-53; Vincenzo Di Benedetto, “Dionysius Thrax and the Tékhnē Grammatik” in History of the Language Sciences, Volume 1 (Berlin: Mouton de Gruyter, 2000), 394-400; David Blank, “The Organization of Grammar in Ancient Greek” in History of the Language Sciences, Volume 1 (Berlin: Mouton de Gruyter, 2000), 400-417. Also relevant is P. H. Matthews, “Greek and Latin Linguistics,” in History of Linguistics, ed. Guilio Lepschy, English ed., 4 vols (New York: Longman, 1994), 2:67, 1-133.

[2] Matthews, “Greek and Latin Linguistics,” 7.

[3] Ibid., 6.

[4] Adapted from R. H. Robins, The Byzantine Grammarians: Their Place in History (Berlin: Mouton de Gruypter, 1993), 27. Robins does not include the Future and Aorist in his chart, but his broader discussion shows where they fit in relation to the others. Matthews (“Greek and Latin Linguistics,” 46) is in perfect agreement with Robins, but he only provides a prose description for the Stoics rather than a chart.

[5] See Porter, Verbal Aspect, 20-21; Matthews, “Greek and Latin Linguistics,” 46.

[6] Carlota Smith The Parameter of Aspect (2nd ed.; Dordrecht: Kluwer Academic, 1997), 66.

[7] The relationship is also likely related to the fact that both have stems terminating in a sigma before their respective person/number endings. This interpretation of the Stoic’s perspective is rather clear. However, whether it is the correct perspective is a completely other issue. Most scholarship views the sigmatic future as historically and semantically distinct from the sigmatic aorist.

[8] Porter, Verbal Aspect, 21.

[9] Porter, Verbal Aspect, 21.

[10] Consider for example, how English speakers conceive of the meaning of prepositions. When we talk about the meaning of “in,” we only think of its basic prototypical use: “Of place or position in space or anything having material extension: Within the limits or bounds of, within (any place or thing).” (OED, loc. cit.). And yet the vast majority of instances of “in” in this very paper are far more abstract than that definition. An instance from footnote 13 above: “ [Robin’s] broader discussion shows where they fit in relation to the others.” The fact that all use of language involves a prototypical exemplar of usage and a variety of (polysemic) extensions that more or less align themselves to that usage is a fundamental fact about the nature of language. For a brief summary of prototypicality and polysemy as related to Greek, see Clarifying Allan & Kemmer on Middle Voice: Cognitive Linguistics. John Taylor (Linguistic Categorization [3rd ed.; OTL; Oxford: Oxford University Press, 2004], 1-18, 41ff.) provides an excellent survey of the research on prototypicality, linguistics, and human cognition.

[11] Porter, Verbal Aspect, 20-21.

[12] Ibid., 20.

[13] But also, it is quite strange that he uses the phrase “kind of action” rather than tense in referring to the complete vs. incomplete distinction. It almost feels as if in choosing these words, he is stacking the deck in his favor right from the start against the neogrammarian Aktionsart—a point to be discussed at length in later sections of this survey.

[14] Porter seems to have worked hard on finding a definition of aspect to avoid time: “a morphologically-based semantic category which grammaticalizes the author/speaker’s reasoned subjective choice of conception of a process” (Ibid., 1). One wonders what a non-temporal process is. Carlota Smith’s statement marks a striking contrast: “Aspectual meaning contributes temporal information and point of view to sentences” (Parameter of Aspect, 1).

[15] For example, Bernard Comrie defines aspect as, “[A]spects are different ways of viewing the internal temporal constituency of a situation” (Aspect [CTL; Cambridge: Cambridge University Press, 1976], 3), a definition that Porter refers to (Verbal Aspect, 45) without a single negative word on this fact.