ΕΝ ΕΦΕΣΩ

Studies in Greek Language & Linguistics…

Category Archives: Greek

Some notes on Ἐργάζομαι & Middle Voice

Ἐργάζομαι is a bit of a difficult verb to deal with in terms of voice. It’s perhaps the only verb that causes problems (at least at face value) after the rejection of deponency as a valid category for the Greek Voice System. Rutgar Allan categorizes it as an indirect middle (Allan 200, 54). That has always seemed a bit forced to me.

However, perhaps there’s some credence to it.

So here are some general notes about this particular verb:

  1. It’s primary sense “to work” is syntactically intransitive, so the traditional definition of deponency as “active in meaning middle-passive in form” is just wrong. There’s nothing “active” about intransitive verbs. Active as a label only applies to verbs that can be passivized. Intransitive verbs can’t. So from an English perspective (which is what the deponency perspective essentially is) ἐργάζομαι is has no voice.
  2. The secondary sense “to accomplish/do [something],” while transitive, is less transitive than it could be. All instances of ἐργάζομαι with an explicit object involve situations where the object is non-referential. That is, the object does not refer to a specific entity within the world of the text. So even this secondary sense that takes an object is not prototypically transitive. In line with that, when ἐργάζομαι takes an object, that object cannot be syntactically passivized and promoted to the position of subject in the clause. If ἐργάζομαι with an object ruly filled the role of “active in meaning” then that should in principle be possible, but that does not appear to be the case.
  3. When one examines the individual instances of ἐργάζομαι, its seems reasonably clear that the verb involves some the performance of an activity (this is, after all, an activity predicate) with a clear benefactive (or malafactive) sense for the subject. This is in line with the semantic category of subject affectedness that is involved in middle systems and also with the indirect middle usage for Greek.
  4. Lastly the fact that in the aorist ἐργάζομαι only allows sigmatic middle morphology rather than a θη middle suggests that it is necessarily either an indirect middle or direct middle. Those are the only middle categories in Koine Greek that still disallow θη morphology (the θη has make sporatic in roads into these two categories, but its not consistent or systematic in any way).

A few comments on grammatical terminology old and new

I know its regularly said by those whose background is in classics and the tradition of classical philology that much of linguistics today and its own set terminology creates a sort of ivory tower situation, where it is almost impossible to follow a discussion as an outsider. And I certainly do not doubt that this is true to some extant for many of the descriptive and theoretical frameworks commonly used in linguistics, particularly in the more structuralist traditions. This problem is one that I’ve tried to fight against where I can, attempting to sufficiently define terminology as often as possible and linking to accessible sources for definitions, such as the SIL glossary of linguistic terms or the Leipzig Glossing Rules and its “appendix” of category labels.

On the other hand, whenever I open up a standard Greek grammar that’s written in the classical tradition, I cannot help but recognize the exact opposite problem. For example, for a contract project I’ve been working on, I’ve spent a little time examining the difference between ού and μή, Greek’s two negative particles. Here’s an example of the kind of obtuse, generally useless description of these two:

The two principle negative adverbs are ού and μή. Οὐ is the objective negative and μή is the subjective. It is hard to say which is stronger. Ού is stronger because it is simple, straightforward and uncompromised. Μή is stronger because it has an underlying sense of passion and rejection. The fervour and unction of μή suits the tone of much expression in the literature of the Hellenistic and Roman ages. So μή tends to supplant ού in many of its uses at writing. Some irregular uses of μή in Classical times, which do not fit into the idiomatic pattern then prevailing, are doubtless to be explained by the touch of passion beginning to make μή  win out even in the earlier period. Another important facet of the meaning of μή is its appropriateness as a generic and so hypothetical negative. The idea of rejection it retains always, at least subliminally, can prevent an idea that is to say specifically, and so the suspended idea passes into the sphere of the typical or conditional.

Gary L. Cooper, Attic Greek Prose Syntax Vol  2., 1997.

The degree is nonsense in that quote is simply overwhelming. But I do not mean “nonsense” in that Cooper is wrong. No, most of the paragraph is fine, assuming one parses the actual proposition accurately (which may or may not be feasible). The nonsense is created by the sheer lack of well-defined terminology at all. What in the world does it mean for a negative to be “straightforward and uncompromised,” anyway? So much of this paragraph reads like filler solely designed to increase one’s word count. It certainly doesn’t help contribute to understanding the difference between ού and μή.

Here’s that same paragraph again with some commentary from me in bold bracketed text:

The two principle negative adverbs are ού and μή. Οὐ is the objective negative and μή is the subjective. [This is good and helpful. Now could we get a technical definition of what these two terms mean in relationship to negation?] It is hard to say which is stronger. [Stronger in what way?] Ού is stronger because it is simple, straightforward and uncompromised. [Yeah, no idea what that means…] Μή is stronger because it has an underlying sense of passion and rejection. [Well, “rejection” is something useful, at least etymologically, but passion? Other grammarians talk about μή being used to denote caution assertion, how does that relate to “passion”?]The fervour and unction of μή suits the tone of much expression in the literature of the Hellenistic and Roman ages. [This may be the first time I’ve see language change attributed to “unction.” Nice.] So μή tends to supplant ού in many of its uses at writing. Some irregular uses of μή in Classical times, which do not fit into the idiomatic pattern then prevailing, are doubtless to be explained by the touch of passion beginning to make μή win out even in the earlier period. [Which “irregular” uses are these? They’re never explicitly marked in the following detailed discussion?] Another important facet of the meaning of μή is its appropriateness as a generic and so hypothetical negative. [The term “hypothetical,” I can understand, but “generic”?  Exactly what makes μή “generic” and does that mean that ού, by implication is *not* “generic”?]The idea of rejection it retains always, at least subliminally [Yeah, I don’t but it], can prevent an idea that is to say specifically, and so the suspended idea passes into the sphere of the typical or conditional.

I hope that gives you a sense of the kind of things that go through my head when I’m reading traditional grammars. The nearly total lack of specificity  in terminology, definitions, and explanation is incredibly frustrating.

A brief comment on telicity and boundedness

I find it difficult to accept the analysis of the Ancient Greek perfect as proposed by Gero and Stechow (2003; LINK: 2002 prepub version), aside from the fact that I think they rely too heavily on the traditional categories of the English perfect, the larger issue is that they successfully survey thousands of years of data without any reference to telicity. I’m at a complete loss how one could do that, especially with a grammatical morpheme like the Greek perfect.

Their only mention of telicity at all is made in defining perfective aspect, where they claim that perfective aspect denotes the completion of an event. And even this does not hold. In as much as the perfective aspect refers to an event as a single entity without reference to duration, or iteration, the perfective aspect is bounded, but it is not inherently telic. Consider the sentences below in terms of how the first entails (or doesn’t) the second.

(1) John was walking in the park this morning –> John walked in the park this morning.
(2) John was walking to the park this morning –/–> John walked to the park this morning.

In sentence pair (a) The imperfective predicate in entails the perfective predicate. This is possible because while the perfective predicate is bound, it is not telic. Conversely, in sentence pair (b) the imperfective predicate does not entail the perfective predicate. There is no way of knowing from the imperfective predicate whether or not John successfully arrived at the park. An endpoint in an imperfective clause is only a potential endpoint with no entailment of its achievement.

All that to say, boundedness needs to be kept distinction from telicity. Boundedness is a feature of the perfective aspect. Telicity is not.*

*To be fair, there are a number of other aspectologists** that fail to make this distinction.
**Incidentally, the correct collective noun for referring to a group of aspectuologists is: “an iteration of aspectologists.”***
***Yes, I just made that up.

A brief comment on historiographical issues surrounding aspect

I just wanted to make this passing note. A few posts back, I presented my analysis of the 1882 edition of William Moulton’s translation of Winer’s Greek grammar. In that discussion, I made the observation that Porter had treated Moulton-Winer unfairly in his criticism of how certain usages of the Greek present imperfective are explained in the grammar. I won’t go back over the question here. My analysis is available in the link above (be sure to read the comments on the post, too).

All I want to say here is that I have now come to the conclusion that this is not an intentionally misrepresentation of Moulton-Winer on Porter’s part. That was driven home for me while rereading some of the papers in the JSNTSupp volumes from the 1990s. In his critique of Fanning’s position, he writes:

Fanning presents a revisionist view of the history of verbal aspect, going to great lengths to preserve the traditional categories and terminology. He stresses his belief that the comparative philologists of the nineteenth century were in actual fact discussing verbal aspect, even if they did not call it this or recognize it as such (Porter 1993, 36).

I hope it goes without saying that it is Porter’s view of the history that I view as revisionist, rather than Fanning. But the fact that Porter views Fanning’s literature review as “revisionist” goes an incredibly long way in explaining how Porter’s reading of Moulton-Winer (1882) arose.

So on that basis I do view Porter’s discussion of that grammar as an honest one. I just view it as wrong and poorly argued…and he would like say the same about my own. At this point, our “world-views” (for lack of a better term) are just so dramatically different.

Works cite:

Porter, S. E. 1993. “In Defence of Verbal Aspect.” Page 26-45 in Biblical Greek language and linguistics: Open questions in current research. Sheffield: Sheffield Academic Press.

Winer, Georg B. and William F. Moulton. 1882. A Treatise on the Grammar of New Testament Greek. 3rd ed. Edinbugh: T & T Clark.

On critiquing scholarship…

For the past couple years now, I’ve generally held back on criticizing other published works of others (with a few small exceptions). This has mainly because of pushback that I’ve gotten from people like Rod Decker on the question, who has pointed out a couple of times that perhaps my lack of experience with the publishing of articles and books (and for that matter, my lack of a doctorate–Dr. Decker has pointed that one out to me, too) makes me slightly more negative in my criticism than I should be. And I’ll be the first to admit that I’ve probably gone too far on a couple of occasions in the past.

That’s fair. I’ve mostly accepted it, while also feeling as if much of what has been published about Koine and New Testament Greek under the name “modern linguistics” wouldn’t past peer-review in were it published in the field of linguistics rather than biblical studies. But I’ve refrained from articulating 99% those criticism here. And lately, I’ve been focused on actually producing substantive and useful discussions on this blog. That’s been one of my goals in surveying how old grammars deal with tense and aspect on their own terms and my work on enclitic pronouns as well. And it’s why I haven’t written much commenting or reviewing recent publications on tense and aspect in Greek for quite some time. I’d rather be known for positive contributions. I’m trying…really.

With all that in mind, I finally got around to reading Porter’s contribution to D. A. Carson’s festschrift: Understanding the times. He really, really goes after Con Campbell. And I mean really go after him.* It’s nearly as bad as the “uncharitable rant” against Kurt and Barabara Aland in Biblical Greek language and lexicography.

Is this standard I’m supposed to live up to in my non-Ph.D. state?

*And that’s actually really strange. Campbell is one of the few published scholars who has expressed significant agreement with Porter’s tenseless view of the Greek verbal system. Surely he would want to coordinate with his allies rather than tear them down–a few of the criticism Porter makes involve issues where Campbell was attempting to justify the tenseless view as theoretically plausible. A very odd state of affairs, indeed.

Logos Bible Software 5 is Here

Logos Bible Software has released the new version of their flagship project: Logos 5.

It incorporates some massive changes in datasets and built in resources while keeping much of the user experience quite similar to Logos 4. I had originally intended to have a useful post ready this evening surveying how tools for studying Greek have changed and have been improved, but prior commitments in thesis writing and a couple other undisclosed Greek projects have held me back. So that’s still in process right now (sorry). Ideally, it will be up by tomorrow. We’ll see.

The most exciting thing about Logos 5, in my opinion, is all the effort they have put into meaningful access to content. It isn’t about searching a massive number of books any more. It’s about finding useful information for specific questions. It’s about making the semantic web real within Logos. The Bible Sense Lexicon is an incredibly exciting project and a great example of creating structured and meaningful information and making it easily accessible.

I know a lot of people that work with languages at the same level I do (or higher) tend to be rather cynical about how tools for using Greek and Hebrew are present in Bible Software packages, but Logos also put significant effort into academic projects and they’re only able to do it because of the pastoral and layperson user base that they have to support such projects. Steve Runge’s Lexham Discourse Greek New Testament is an important example of this. Another is the SBLGNT. Personally, I’d rather work with them to improve how language study is done in the future than merely dismiss them. And, well, that’s what I have been doing and will continue to do so.

I’ll have more thoughts on Greek databases and changes and advancements in the coming few days (because there are some important ones), which is probably just fine. It looks like users are swamping Logos’ servers right now anyway…

Occasional Surveys in the History of Greek Grammar: G. B. Winer & William Moulton (1882)

Despite the production of other New Testament grammars, Winer’s work continued to hold the greatest influence throughout the century and by 1882, it was in its ninth English edition.[1] William Moulton’s contributions to the grammar grew consistently with each of his own editions/translations, though the basic organization continues to be the same. Most importantly, by the time of Moulton, the grammar is no longer the mere supplement that it was in 1825, and it is not merely a grammar of the New Testament, despite the title. The edition of 1882 is full of references, not only to the New Testament, but also to contemporary literary authors such as Appian & Pausanias and Classical authors like Xenophon. Moreover, the tenses are discussed in far more detail with a clear aim toward comprehensiveness. The goal of the grammar is no longer to merely be a supplement functioning along side a Classical Greek grammar.

It is also apparent that Moulton had access to the later German grammars such as Raphael Kühner’s based on some of the terms he uses. The aorist is a simple past and the narrative tense that contrasts with the imperfect and the pluperfect which “always have reference to subordinate events which stood related, in respect of time, with the principal event (as relative tenses).”[2] The perfect functions as a relative tense “and represents an action as a complete one, in relation to the present time.[3] The addition of the concept of relative tenses here suggests that, at least to some degree, Moulton viewed Kühner’s model as an advancement in their understanding the interaction of temporal location and temporal constituency—to borrow Comrie’s turn of phrase for the categories of tense and aspect, respectively.

In terms of usage, Moulton drives home the point of Winer that tenses do not stand in for one another and his examples reflect similar patterns we found in other grammars: the so-called peculiar usages reflect extra verbal linguistic factors, such as the “procedural characteristics” described by Fanning.[4] Of course this is not how Moulton expresses it. Rather, he writes,

Strictly and properly speaking, no one of these tenses can ever stand for another … where such an interchange seems to exist, either it exists in appearance only, there being in point of fact some assignable reason (especially of a rhetorical kind) why this tense is used and no other; or else it must be ascribed to a certain inexactness belonging to the popular language, through which the relation of time was not conceived and expressed with perfect precision.[5]

When we recognize the dramatic difference in context terminologically between 1882 and the present, a few things become apparent. It seems reasonable to take a statement “reason[s] … of a rhetorical kind” as being roughly to modern day pragmatics.[6] And the observation that spoken language is not particularly precise has only been confirmed and emphasized in contemporary linguistic research.[7]

Nonetheless, this is precisely the point Porter chooses to take issue with in his own survey.[8] To Porter’s credit, he chooses an excellent example: Matthew 26:2, provided below.

(1) οἴδατε ὅτι μετὰ δύο ἡμέρας τὸ πάσχα γίνεται, καὶ ὁ υἱὸς τοῦ ἀνθρώπου παραδίδοται εἰς τὸ σταυρωθῆναι.
You know that after two days the Passover takes place and the Son of Man will be handed over to be crucified.

That the present only appears to function as a future confuses form and function, says Porter, because, “he does not say how that would differ from actually functioning futuristically.”[9] On the one hand, this is an extremely good point. Moulton-Winer says nothing about the difference between appearance and actual replacement, but Porter would do well to at least comment on the fact that Moulton provides a reason for the usage:

“It is used for the future in appearance only, when an action still future is to be represented as being as good as already present, either because it is already firmly resolved on, or because it must ensue in virtue of some unalterable law.”[10]

It may very well be true that Moulton does his readers a disservice by not explaining the mechanisms for the apparent replacement of forms. That is not an unreasonable criticism. At the same time, however, Porter does his own readers a disservice by not telling the entire story on this particular point. Moulton’s discussion is far more nuanced than Porter lets on in his critique of Moulton-Winer. Porter selectively quotes Moulton-Winer here, proving only the beginning the statement above, “It is used for the future in appearance only…”[11]. He does not quote Moulton’s actual explanation of the usage at all. Even worse, Porter does not even provide the ellipsis marking that he’s quoting an incomplete sentence. At best this is a horribly unfortunate editorial error, at its worse, this could possibly be viewed as a somewhat manipulative move to slant Porter’s readers toward a particular view of the Greek verb.

It is precisely this kind of discussion of the old grammarians that motivated these occasional studies: to give the old grammarians a fair hearing of their views in this historical contexts. A grammar is a communicative act to a particular audience. Minimally, Moulton would likely be able to defend himself on this point by saying that he lived one hundred years before any sort of adequate theory of the mechanisms of semantics had been developed, but it also is not unreasonable to assume that Moulton had full capacity to give sufficient explanation, even without the kinds of explicit theoretical frameworks that linguistics work within today. William Moulton and G. B. Winer may not have articulated their views of the Greek tenses in the same manner that is done today, but the reasons they give for their grammatical claims make it quite clear that they understood the language extremely well.


[1] Georg B Winer and William F. Moulton, A Treatise on the Grammar of New Testament Greek, 3rd ed. (Edinburgh: T & T Clark, 1882). The 3rd edition in the title refers to Moulton’s third edition—the spine of the volume lists both numbers: 3rd edition and 9th English Edition.

[2] Ibid., 331.

[3] Ibid.

[4] Fanning, Verbal Aspect, 126ff.

[5] Winer and Moulton, Treatise , 331.

[6] Pragmatics, as opposed to semantics: how language is used rather than how language means, though where one ends and the other begins is far from clear, particularly since, in a sense, all meaning in language arises from usage of language. See Taylor, Linguistic Categorization, 132-134.

[7] Indeed, the idea of perfectly precise semantics likely disappeared with the end of logical positivism.

[8] That is, this is the issue Porter takes issue with beyond his standard disparaging remarks about how these old grammars take an out of date “time based” view of the tenses—a criticism that is, itself, quite unfair.

[9] Porter, Verbal Aspect, 51.

[10] Winer and Moulton, Treatise , 331.

[11] Porter, Verbal Aspect, 51.

Occasional Surveys in the History of Greek Grammar: Raphael Kühner & William Jelf (1866)

The translation of Raphael Kühner’s German grammar into English by William Jelf marks the closest we get to a comprehensive grammar of Classical and Hellenistic Greek.[1] And while it does not provide a distinct discussion of New Testament Greek by itself, it does provide references to New Testament usage as it relates to broader usage. The grammar marks a substantial change from the previous grammars examined thus far in that it seeks a comprehensive description rather than assuming that basic knowledge is already known to the reader.[2]

The section on the verb begins at §394 and continues through §409 with nearly thirty pages of discussion. Kühner-Jelf also appears to be the first English grammarian to provide a meta-theory of tense-aspect applicable to all language before attempting to deal with Greek itself.[3] This system is grounded primarily upon Aristotle’s On Interpretation.

Ῥῆμα δέ ἐστι τὸ προσσημαῖνον χρόνον, […] λέγω δὲ ὅτι προσσημαίνει χρόνον, οἷον ὑγίεια μὲν ὄνομα, τὸ δὲ ὑγιαίναι ῥῆμα. Προσσημαίνει γὰρ τὸ νῦν ὑπάρχειν.[4]

A verb is that which connotes temporality. … Now I say it connotes temporality [and this is what I mean]: Ὑγίεια (healthy) is a noun, but ὑγιαίναι (be healthy) is a verb; for it connotes the present [νῦν] existence of the state [in question].[5]

For Kühner-Jelf, time is involved in verbal semantics in two ways. There is what Kühner-Jelf terms the “definition notion of time,” which takes the speaker as the temporal reference point (Absolute tenses). This stands in contrast with tenses which express some other relationship relative to another point:

An action may not only be thus defined by its reference, whether as past, present, or future, to the time present to the speaker, but may also have a reference to some other action expressed by some other predicate, whether it be antecedent to, coincident with, or consequent on this action; that is, whether it be ended before this other action is going on, finished, or intended; whether it is not yet begun, but only conceived as about to happen when the other shall be going on, or finished, or intended. For these also the Greek has forms, which are called the Relative Tenses.[6]

Today when we talk about absolute tense and relative tense, we limit ourselves to tenses such as the English perfect.[7] The term relative tense does not tend to refer to all of the categories that Kühner-Jelf seeks to place in it. We have both the idea of temporal location established by a point other than that of the speaker (the modern definition of relative tense), but we also have the idea of telicity (“action going on, finish, or intended”) that was visible in the ancient grammarians. The modern definition of relative tense is involved here in that the present may function either with reference to the speaker or another reference point within the discourse. The imperfect differs from it in that it is consistently used primarily for the latter of these. As Winer and Levinsohn noted above, the imperfect is used for subsidiary or secondary events connected to the main narrative. In this sense, being a relative tense versus being an absolute tense also involves the modern concepts of backgrounding and foregrounding of information in a discourse.[8]

If this understanding of Kühner-Jelf’s view is accurate, we are left with the lingering question of why he would organize it in this manner with three modern categorical distinctions merged into one bipartite division. Is there a relationship between telicity, temporal reference points, and backgrounded and foregrounded information that has becomes less than clear today than it might have been in the mid-19th century? Or is Kühner-Jelf simply making the system far more complicated than it needs to be? There does not seem to be an easy answer to these questions.

John Lyons provides a potential connection, though it is nothing more than that, “It has been pointed out … that the distinction between tense and aspect is hard to draw with respect to what is sometimes described as relative, or secondary, tense. We will say no more about this.”[9] Rather frustratingly, Lyons provides no reference as to who it was that pointed this fact out. Without any citation, we are prevented from following up the observation elsewhere. Lyon’s refusal to say more on the subject leaves us floundering without any direction as to where this connection might lead.

Porter deals with Kühner-Jelf in a rather different manner than what we have done here. Kühner-Jelf functions as representative evidence of two threads: those who follow the Stoics and those who follow Dionysius Thrax. For Porter, Kühner-Jelf is one of the latter.[10] For reference, figure 2 provides Kühner-Jelf’s chart of the tense-aspect categories.[11]

Present. Past. Future.
I. Absolute γράφω ἔγραψα. γράψω.
II. Relative
a. Coincidence.Action yet going on.Imperfect.
γράφω. ἔγραφον. γράψω.
b. Antecedence.Action past.

Preterite.

Γέγραφα ἐγεγράφειν γεγραφὼς ἔσομαι.
c. Consequence.Action yet to come.

Future.

Μέλλω γράφειν. ἔμελλον γράφειν. μελλήσω γράφειν.

If we interpret the words “action past” as not merely have the sense of action in the past, but also the sense of completed, we find Kühner-Jelf has a system virtually identical to that of Dionysius Thrax.[12] Thus, despite the fact Porter does not think Dionysius had a concept of aspect, he is still generally accurate in saying that Kühner-Jelf’s grammar follows Dionysius in this conception of the verbal system, but the problem is that Kühner-Jelf is also very much like the Stoics as well.[13] That Porter wants Kühner-Jelf to stand with Dionysius in contrast with the Stoics is problematic since the defining line in his view is their conception of “kind of action” (again, refusing to call their conceptualization “aspect).”[14] And we have already seen that both tense and aspect are central to the Stoics and Dionysius.

The irony here comes to view when Porter looks at the German edition of Kühner. Porter used the 2nd edition of Jelf,[15] but the 4rd German edition of Kühner, edited by Blass and Gerth after Kühner’s death.[16] He then describes Kühner more positively as if the observations about aktionsart and what he terms a “neo-Stoic scheme” were original to Kühner. But when we look at the German edition that Jelf translated from, we find that the Absolute/Relative distinction with its complex temporal-aspectual framework is taken from it entirely, including the chart above.[17]

This is not to say that the conception of verbal semantics put forward by Kühner-Jelf is necessarily ideal or should be preferred. There are clear problems with it. In some sense, we could say that the problems more involve issues of elegance and simplicity than they do problems of description, though whether the perfect is best described as a past tense can be (and still is) debated. Likewise, if it is correct that Jelf begins this section talking about the nature of language in general, then there are even more problems, since it is clear that the system could only go so far in explaining tense and aspect beyond a small set of Indo-European languages. Despite these problems, the description put forward in Kühner-Jelf is extremely important in the develop of our understanding of the Greek verb and holds great influence over grammatical description from 1835 when Kühner’s first edition was published through the turn of the century. Even more importantly, potentially valuable elements of his description have been disregarded concerning telicity, temporal reference, and backgrounding and foregrounding of information in narrative.


[1] William Jelf, A Grammar of the Greek Language, 4th ed. (Oxford: James Parker, 1866), vol. 2. By the time Jelf edited the fourth edition, he apparently viewed the work as sufficiently his own to remove Kühner’s name from the title and even the fact that the text was originally a translation. However, the discussion here still refers to the work as Kühner-Jelf because there is little change between the views of Kühner and those of Jelf, even in this fourth edition.

[2] In this sense, this is the first true English reference grammar of the modern era, providing the broader coverage not found in Winer or Buttmann.

[3] This is clear from the fact that Kühner-Jelf uses two distinct sets of terminology: the standard set for Greek verbal forms (present, aorist, imperfect, etc.) and an additional set representing temporal relationships in a given predication. Porter views Kühner-Jelf’s description as only relevant to Greek, criticizing the grammar for using English-only examples for explaining a number of points

[4] Cited in Jelf, Grammar, 54-5. Kühner-Jelf cut a section out of this quote. The ellipsis is mine. See H. Tredennick, ed., Aristotle: The Categories, On Interpretation (Cambridge, Mass.: Harvard University Press, 1973), 118.

[5] My translation.

[6] Jelf, Grammar, 55. Note that the final statement, “For these also the Greek has forms…” also suggests our understanding of Jelf is correct as having a broader theory tense not limited to Greek, but (ideally) applicable to language in general.

[7] E.g. D. N. S. Bhat, The Prominence of Tense, Aspect, and Mood (Amsterdam: John Benjamins, 1999).

[8] See section 3.3 above on Georg Benedikt Winer for discussion.

[9] John Lyons, Semantics (Cambridge: Cambridge University Press, 1977), 2:705.

[10] It should be noted that Porter treats Kühner-Jelf as primarily Jelf’s grammatical work, but even a cursory examination of the German edition Jelf translated from makes it clear that these ideas are first and foremost Kühner. It is only to the extent that Jelf has not revised the discussion—and he does take liberty to do so on a number of occasions—that we can talk about Jelf’s view of the verbal system. This situation is complicated since later on in Porter’s survey, he talks about Kühner views as very different from those of Jelf, but in that case he is referring to Gerth and Blass’ revision of Kühner’s German edition, which drastically changes the discussion of tense-aspect so that it is essentially unrecognizable as Kühner’s. What Porter describes as Kühner’s views are more accurately described as those of Gerth and Blass (see esp. Porter, Verbal Aspect, 25). This fact complicates the situation of Porter’s discussion of Jelf, as we will see below.

[11] Jelf, Grammar, 56.

[12] That this is the correct interpretation of “action past” is made explicitly clear in the section that follows where Kühner-Jelf provides examples of relative tenses, as well as the discussion above.

[13] There is actually a second problem as well. Does Kühner-Jelf truly rely on Dionysius for his view or do they share a common source? We know for a fact that Kühner-Jelf , to some degree, relies Aristotle for the basic nature of the verb. And according to Robins (Byzantine Grammarians, 228), the ancient grammarians, including Dionysius, also took Aristotle as their starting point. For that reasons, there is really no way of conclusively determining the nature of the \relationship between Kühner-Jelf and Dionysius Thrax.

[14] He goes on to critique Kühner-Jelf’s view of individual uses of the tenses, stating, for example, “[C]oncerning Present verbs with Perfect meaning h dismisses this as arising from the ‘sense of the verb’ rather than the ‘force of the tense…’” (Verbal Aspect, 23). This seems an incredible objection. Does Porter truly believe that the sense of the verb has no impact at all on the temporal semantics of a situation?

[15] William Jelf, A Grammar of the Greek Language, 2nd ed. (London: James Parker, 1855).

[16] Raphael Kühner, Ausführliche Grammatik der griechischen Sprache: Satzlehre., ed. B. Gerth and F. Blass, 4th ed. (Leverkusen: Gottschalksche, 1955).

[17] Raphael Kühner, Ausführliche Grammatik der greicheischen Sprache (Hannover: Hahnschen Hofbuchhandlung, 1835), 62.

Tense and Conceptual Reality

I have said before on a number of other occasions that the fact that a mismatch between a particular location in time and a particular grammatical form does not, in itself, constitute sufficient evidence that the language does not have tense. Thus, though not all Greek aorists refer exclusively to past time situations or not all Greek presents refer to present time situation, this is not adequate for arguing that Greek is a tenseless language or that the Greek verbal system does not grammaticalize temporal deixis (again, not time). In other posts, (for example, this one about Huddleston’s analysis of the English past tense) I have suggested that polysemy is a perfectly acceptable way of dealing with such mismatches.

The polysemic approach accepts the mismatch as real, but denies its relevancy for the existence of tense on the grounds the polysemy is the far normal phenomenon in language. That is to say, because multiple meanings should be expected from the outset (Goldberg 1995), the mismatch itself does not prove that the language is tenseless, it proves that the language has polysemic forms.

The alternative, which we will call the cognitive approach, denies the reality of the mismatch itself. The cognitive approach points out that grammatical forms, as linguistic signs, do not refer to anything within the external world. Rather, they refer to mental representations or conceptual structures within the mind of the language user, which in turn involve the language user’s perception, interpretation, and chosen presentation of a given event, process, or state and its temporal deixis. There has been much talk about the subjective nature of aspect in the contemporary literature on the Koine Greek verb, but little has been said about the subjective nature of tense. Though, in a rather limited sense, tense is simpler in its structure than aspect, it is, nevertheless, still subjective simply because it is constrained by physiological and cognitive limits of the language user. And for this reason, it can also be manipulated by the language user.

I should emphasize that such ideas are not unique to me. They represent a major stream in the literature on tense, aspect, and—in fact—all grammatical categories. Carl Bache (1996) makes this extremely clear, in the following long quote (I’ve started it a bit early for a little extra context):

We have just noted evidence that the rule sets central to semantics all operate on linguistic entities which are subject to recognition and hence to conceptual computation. Once plausible implication of this is that all components and elements in semantics are not only present at conceptual structure but the effects of all major rule applications are immediately accessible at this level and available to mental processes, such as recognition. By itself this is of course not proof that semantic and conceptual structure are in fact the same (cf. Jackendoff’s attractive hypothesis of the structual identity between semantics and cognition in e.g. his 1983 book on semantics and cognition; see also Lakoff & Johnson 1980, Lakoff 1987, Taylor 1989, Langacker 1991) but t is strong evidence that they are similar enough to be treated as if they are the same for all practical purposes. So this is in fact what I will do in this book: meanings assigned to grammatical categories and to the forms serving as members of these categories are to be understood as conceptual units or elements. Moreover, the rules involving these units and their interaction will be claimed to have a conceptual rationale.

One important consequence of this approach to semantics is that it is the ‘conceptual reality’ rather than the ‘real’ reality’ which matters in our definitions of categories and members categories …. Form-meaning relationships are not to be understood as relationship between language and the world but rather as relationship between language and the world as conceived by human beings, i.e. the ‘projected world’ in Jackendoff’s terminology (cf. Jackendoff 1983:23ff). Reference is accordingly redefined to be a relation between language expression and projected entities (projected things, events etc.), i.e. entities in the world as conceived by us.
(Bache 1995:53ff).

Now, the cognitive approach to dealing with form-meaning mismatch does not stand opposed to the polysemic approach. They may very well go hand in hand, in different parts of the grammatical system. And, considering that language is an organic and emergent entity, we would precisely expect this to be the case. Personally, I view both polysemic and cognitive reality to be at work in many of the so-call mismatches that we find within the Greek verbal system. For example, I would view what we see with the historical present as choices of conceptualization by a the language user rather than polysemy, but much of variation in usage in the perfect and the future as more likely involving polysemy. With both polysemy and conceptual reality, I reject the idea of Koine Greek as a tenseless language because I reject the methodological assumptions upon with the claim is based.

Incidentally, if anyone comes across a copy of Carl Bache’s book The study of aspect, tense and action for sale. Please, please let me know. I’ve been looking for a copy for months now. It’s out of print and there are no copies anywhere on the online used market. I’ve had it out from the library all summer long and I think the library is getting a little weary…

Works cited:

Bache, Carl. 1995. The study of aspect, tense and action. New York: Peter Lang.

Goldberg, Adele E. 1995. Constructions: A construction grammar approach to argument structure. Chicago: University of Chicago Press.

Jackendoff, Ray. 1983. Semantics and cognition. Cambridge: Mass.: MIT Press.

Lakoff, George, and Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press.

Lakoff, George. 1987. Women, fire, and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press.

Langacker, Ronald. 1991. Foundations of cognitive grammar: Descriptive application. Stanford, Calif.: Stanford University Press.

Taylor, John. 1989. Linguistic categorization. Oxford: Oxford University Press. (<–This book is now in its 3rd edition)

Read this! Imperatives and the countability of events

I noted in my previous post (Once and twice: The countability of events) that I’ve been reading David Armstrong’s (1981) article, “The Ancient Greek Aorist as the Aspect of Countable Action.” As I said before, there really aren’t any surprises. Armstrong’s claims hold for the Koine period quite well. This wouldn’t have been a surprise to Armstrong either, since he made occasional reference to Modern Greek along side Classical and Homeric. When a phenomenon functions the same in Classical and Modern, its pretty likely that it’ll be that way in the Koine Greek as well. But my own digging through the data has been worth the time. Armstrong’s corpus was limited to published lexicons, concordances, and grammars. He didn’t have access to digital texts. Our modern ability to quickly look at massive amounts of data provides for a useful opportunity to test linguistic claims more consistently.

Beyond that, Armstrong also doesn’t look at some specific issues that are important. How does aspect interact with mood and modality? The relationship between aspect and the Greek imperative has attracted a lot of attention from scholars without a whole lot of agreement. I don’t have time at the moment to get into the details of various views (e.g. Porter vs. Fanning). The difference between their views is also documented rather well elsewhere.

The imperative is particularly difficult, though, for the interpretation of aspect. Often times the choice of an imperfective imperative over a perfective imperative appears inexplicable, leaving many struggling to understand what’s going on. This was quite visible in a B-Greek* discussion a few months ago on a thread dealing with the imperative with verbs of speaking.

But what I’m seeing with adverbial numbers here makes it quite clear that aspectual distinctions continue consistently in the imperative mood. Thus, the aorist (perfective) imperative co-occurs with cardinal count:

ἔκχεον εἰς τὴν κεφαλὴν τρὶς ὕδωρ εἰς ὄνομα πατρὸς καὶ υἱοῦ καῖ ἁγίου πνεύματος.
Pour water on his head three times in the name of the Father and the Son and the Holy Spirit (Didache 7.3).

You can count events presented in the perfective aspect. This is not the case with the imperfective imperative, however. The only times you find cardinal count adverbs with imperfective imperatives are when they’re presenting a indefinitely repeatable event. That is to say, while the event itself is internally countable, there is no end to its repetition. As Armstrong puts it,

Such situations of distinguished frequency count occur regularly when what in question is actions, themselves internally capable of count, indefinitely repeated; in such situations it is not the internal count (‘doing this three times’) but the frequency of the multiple action itself (‘always doing this three times,’, doing this three times every year’, and so on) that dictates the choice of the aspect” (8).

This point is also consistent across the Koine period. Daniel prayers three times a day. His prayers are individually countable. But the aspect of “pray” is determined by the atelic nature of the activity. He never stops praying three times a day. In such cases, imperfective aspect has scope, not merely over the verb, but over the whole of the predication, including the number of times it is done. The same is true for the numerous sacrifices the Jewish priests perform in the LXX. The sacrifice itself is countable, but because the priests perform the “two sacrifices a day” indefinitely, the predication itself is not bound and thus it necessitates an imperfective verb.

This situation is maintained for imperfective (i.e. present) imperatives:

Τρὶς τῆς ἡμέρας οὕτω προσεύχεσθε.
Pray in this manner three times a day (Didache 8.3).

Here again, in the imperative, the imperfective is used with τρίς, but the imperfective aspect is necessary because the event is still atelic. The author is instructing regular prayer that must not stop after just a single day.

So again, aspect doesn’t change in the imperative. It’s functions in just the same way. And I think that viewing the perfective vs. imperfective distinction in terms of countability may very well be helpful in understanding other more difficult to interpret imperatives as well.

————————————–

*Incidentally, on the subject of B-Greek, the new forum format recently past its one year anniversary. And I have to say, the change has been very exciting. To those of you who have refrained from participating in the e-mail list because of the elementary level of the content, you might want to look at how some of the more technical sub-forums have developed. There has been some excellent discussion over the past year.

Cited:

Armstrong, David. 1981. “The Ancient Greek Aorist as the Aspect of Countable Action.” Pages 1-12 in Syntax and semantics, vol. 14: Tense and aspect, ed. P. Tedeschi and A. Zaenen. New York: Academic.

Follow

Get every new post delivered to your Inbox.

Join 391 other followers