Lemmatizing Greek Verbs

Biblical Greek morphology and syntax, aspect, linguistics, discourse analysis, and related topics
MAubrey
Posts: 1090
Joined: May 6th, 2011, 8:52 pm
Contact:

Re: Lemmatizing Greek Verbs

Post by MAubrey »

Jonathan Robie wrote:Very helpful article. For teaching, I like this table from the paper, it's much easier to follow than the equivalent in most grammars:
infinitive-bagl.png
But wouldn't you have to learn more than this for some verbs like φέρω (oops, φέρειν...), where you would need to know the future stem?

We could, of course, teach more forms as needed for a particular verb ...
infinitive-micheal.png
I should probably jump in here being that I'm one of the primary contributors to BibleMesh's approach to the Greek verb here. We do much more for individual verbs. the distinctions presented in the first chart is just supposed to be a basic schema for students. And I should emphasize that this chart is a little complicated, too. It sort of gives the appearance that the aorist & future infinitives are treated as past vs. nonpast. But that's more a convenience of the chart more than anything because the the aorist/future infinitive is just a big irritating mess more than anything else. We do, however, treat the future as perfective aspect, however (typological justification of such a grammaticalization can be found in Thelin 1990).

Lemmatizing the infinitive is something we've talked about a lot, but haven't acted on. It's a very good idea, but it's still a hard sell to most intro Greek teachers who don't know anything beyond, well, Mounce & Wallace.
Mike Aubrey, Linguist
SIL International
Koine-Greek.com
cwconrad
Posts: 2112
Joined: May 5th, 2011, 5:52 pm
Location: Burnsville, NC 28714
Contact:

Re: Lemmatizing Greek Verbs

Post by cwconrad »

In response to several questions raised by Stephen Hughes in his post of December 24th, 2014, 11:33 pm:

1. Lemmatizing by verb-forms that do not appear in extant ancient texts: one complaint has been that contract verbs in -εω, αω, and οω are not found in Greek texts; they are used as a convenient way to indicate that these are contract-verbs of one of these three types (what about ηω verbs?). Jonathan raised the issue about whether some of the infinitives suggested as alternative listings for the contract verbs (ποιεῖν, τιμᾶν, δηλοῦν) may perhaps not appear in extant texts either — in “our corpus” (but I think we agree that “our corpus” should extend well beyond the Greek biblical text of GNT and LXX into Hellenistic Greek literature).

I'd respond: Let’s put it this way: how often do you find uncontracted verb-forms of alpha-, epsilon-, or omicron-contract verbs in your reading in the GNT or Hellenistic literature. Herodotus has things like δοκέων and ἑπιδέεται. Do you read Heraclitus’ fragments frequently? What about Pseudo-Lucian’s Syrian Goddess? It may well be that some of these infinitives aren’t found, but (a) we’re only using the first one as a lemma for a verb, and (b) as principal parts, it’s as easy to derive a tense-stem from an infinitive as from a first-singular, considering the fact that tense-stems don’t carry an augment: compare
δηλοῦν, δηλώσειν, δηλῶσαι, δεδηλωκέναι, δεδήλωσθαι, δηλωθῆναι
δηλόω, δηλώσω, ἐδήλωσα, δεδήλωκα, δεδήλωμαι, ἐδηλώθην

2. A somewhat different question is, if I read him aright, what sort of lexical information does a user of the lexical reference work want or need?
Stephen Hughes wrote:I think it depends whether you are creating a work about the language or the texts - a grammar or an index. Different information in the grammar has different value, but most of it is unmarked. I don't think it should worry you, per se, but you may like to be aware of it. My Greek is not great, but over the past 4 or 5 years, my thinking / processing needs have changed. Initially, I struggled to form or identify forms of verbs, then I was confident and reached 95+% skill in associating forms with particular verbs. Now lately in processing Greek, I'm looking for information that blanket terms based on the form like infinitive - can't give me. I want information like, that verb is used as an infinitive of purpose in such and so a construction, or with a certain verb. After the learning forms is complete, the level of desire to learn is still there, just I'm looking for "idiomatic" and syntactic information to associate with particular verbs and verb-forms.
The original topic was lemmatization: should we refer to a verb and look it up in a lexicon by the first-person singular present indicative active (or middle) — or by the present active (or middle) infinitive.

It seems to me that what Stephen is discussing is what kind of information about a verb (or other word) and how much ought to be included in a lexical entry. That’s a different question. I think that the best lexical entries are those that indicate, in one way or another, all the tense-stems and variant forms found in our literature, and that catalogue as intelligently as possible and illustrate with examples all observed usages of a verb (or other word)

3. Is it true that the contract-verbs never appear in uncontracted form?
cwconrad wrote:Of course, it never bothered us for one moment that uncontracted contract-verbs do not occur in or beyond "the corpus." We all learned ἀγαπάω and φιλέω and δηλόω -- forms that never appear!
Stephen Hughes wrote:
Could you reassert that last statement or circumlocute it please.
Well, let me just say that I learned these verbs in those first-person-singular present indicative active forms, and my guess is that most B-Greekers learned them that way. I said these were “forms that never appear” — my question would be: where have you encountered these forms apart from lemmas, lexical entries and paradigms or as reference-forms? Where have you encountered these forms in the literature you’ve read?
Stephen Hughes wrote:[I can't remember in which authours, but I think I remember encountering uncontracted forms during my (few brief) years reading classical Greek, and also short vowel omega verbs conjugated as -μι verbs in a couple of the dialects. Am I mixing up grammar books and actual texts in my memory.{/quote]
Ionic dialect uses uncontracted forms; those of us who’ve read Herodotus have seen them there — or in Heraclitus or in Ionic inscriptions. In Homer what we see is not uncontracted forms but something called “diectasis” where, for metrical reasons, a verb-form that is contracted in regular speech is re-opened or stretched out into two syllables, e.g. ὁροωσι from ὁρῶσι from an uncontracted ὁρά-ουσι. And yes, there are dialectal variants to be encountered, e.g. Ionic contraction of -εο- into -ευ- as in ποιεῦντι, or Lesbian Aeolic ποιημμι for Attic ποιέω.]
4. the form of the perfect active imperative:
Smyth §466 c. wrote:λῦσ-ον aor. act, and λῦσ-αι aor. mid. are obscure in origin.
The -σον of the sigmatic aorist active imperative seems unrelated to the -σο 2 sg. middle imperative ending (which looks, in fact, more like the secondary 2 sg. middle ending.

It looks like the “stative” imperative form cited from BibleMesh’s chart in Kraeger’s article is an artificial composition based upon perfect stem λελυ- and the aorist active imperatival ending -ον.
οὔτοι ἀπ’ ἀρχῆς πάντα θεοὶ θνητοῖς ὑπέδειξαν,
ἀλλὰ χρόνῳ ζητέοντες ἐφευρίσκουσιν ἄμεινον. (Xenophanes, Fragment 16)

Carl W. Conrad
Department of Classics, Washington University (Retired)
Jonathan Robie
Posts: 4159
Joined: May 5th, 2011, 5:34 pm
Location: Durham, NC
Contact:

Re: Lemmatizing Greek Verbs

Post by Jonathan Robie »

I didn't see this to approve it before several people responded in this thread, so let me quote it again to bring it to the top of the thread.
S. M. Kraeger wrote:Thank you all for your reflections on my research. If I may, I would like to respond to some of your questions/concerns:
Eeli Kaikkonen: I presently don't use SLA categories for my pedagogy. Part of the problem with SLA for our present discussion is that there are no native speakers of the language. I would be interested in seeing how those who apply SLA research to Greek would approach this subject. I suspect, though, that using the infinitive as a basic form would only lend itself more naturally to SLA pegagogy. My research was trying (in part) to situate the verb within the verbal system both pedagogically and lexicographically. One must have a taxonomy of some sort, and my suggestions (I attempted to demonstrate at the end) are, I think, better than the alternative λόγος based model.
Jonathan Robie: The list was taken from the BibleMesh Greek Project (http://www.biblemesh.com/languages). It was not meant to suggest that a trinary aspectual approach (leaving out the future tense) is the way I would necessarily taxonomize the verb. How does one fit the future tense into any discussion of verbal aspect? I was trying to leave this question open. As far as your concerns about irregular verbs, this is of course a problem for those verbs whose stems differ between the aorist and future tenses. I think one's taxonomy (include the future tense as an alternate form or not) will probably depend on how helpful it really is to insist upon another tense-form for that discussion of irregular verbs. Finally, your thoughts about using prompts is instructive, and a way that I never actually considered using the infinitive pedagogically. Thank you!
Carl Conrad: I had the same struggle with that imperative form. I inserted it because that was how BibleMesh had it in their grammar. I left that issue alone, but I think a good footnote might have begun to address the issue. Is there a form that better represents an imperative in the perfect tense? Would that I had an answer! But I feel compelled to say that the chart was meant to illustrate a possible way (one that is actually used) to approach the issue. If I had to recreate this for my own grammar, I would probably do some revising (for example, why the 1st person plural form, and why the non-standard participial forms?), though I would probably also keep the same taxonomy.
Stephen Carlson: I appreciate your comments on my work. As far as aspectual terminology, I'm actually at a loss as to what sort of terminology to use for verbal aspect. I think Porter's timeless model is not quite right. But I appreciate the terminology because it keeps me from using, e.g., "Present Tense" or some such thing, where "Present" is not very helpful for discussing the tense in general, much less in the infinitive. I still need to think this over and I'm open to suggestions on aspectual terminology. As far as failing to consider the benefit of principal parts in the first-person form, perhaps you are right. Can we see the principal parts in terms of stems and work our way through the principal parts by way of the infinitive? Add an augment here, add a person-number suffix there, and it wouldn't be too hard, I think, to mentally transfer from infinitive forms or stem forms to the various finite endings. But I might also be responding like this because my focus is to draw attention away from such a model in favor of an alternative. But as a pedagogical feature, I might actually be favorable to using principal parts in their finite forms. Your critique is well taken.
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/
RandallButh
Posts: 1105
Joined: May 13th, 2011, 4:01 am

Re: Lemmatizing Greek Verbs

Post by RandallButh »

In our Morphologia book (BLC) we use the aorist infinitive as the lemma for the charts and with the both the aorist and continuative infinitive (as a second form) in the English-Greek index.

We call the continuative infinitivethe παρατατική in Greek because it does service for both the present indicative and the imperfect (past imperfective) indicative.

In terms of philosophy and pedagogy, we make it a priniciple to use real words. It is not so much a question of whether a particular form is found in a particular corpus, but in filling out what that form would have been if the context were to produce such a form. Naturally, some forms are restricted idiomatically and would not occur, so we leave those blank, and other forms are just too irregular to know for sure, e.g. all of the irregular forms for 'to fly' [aor?] πέτεσθαι. Most of the time things are an easy call. ἀγαπῆσαι ἀγαπᾶν is easy.

Why real words? Because they can occur in a person's mind without needing to go through a conscious algorithm first. That algorithm takes people momentarily out of the language and is not recommend for developing true automaticity in the language. See Buth, in the Danker Festschrift.
Stephen Carlson
Posts: 3351
Joined: May 11th, 2011, 10:51 am
Location: Melbourne
Contact:

Re: Lemmatizing Greek Verbs

Post by Stephen Carlson »

S. M. Kraeger wrote:Stephen Carlson: I appreciate your comments on my work. As far as aspectual terminology, I'm actually at a loss as to what sort of terminology to use for verbal aspect. I think Porter's timeless model is not quite right. But I appreciate the terminology because it keeps me from using, e.g., "Present Tense" or some such thing, where "Present" is not very helpful for discussing the tense in general, much less in the infinitive. I still need to think this over and I'm open to suggestions on aspectual terminology. As far as failing to consider the benefit of principal parts in the first-person form, perhaps you are right. Can we see the principal parts in terms of stems and work our way through the principal parts by way of the infinitive? Add an augment here, add a person-number suffix there, and it wouldn't be too hard, I think, to mentally transfer from infinitive forms or stem forms to the various finite endings. But I might also be responding like this because my focus is to draw attention away from such a model in favor of an alternative. But as a pedagogical feature, I might actually be favorable to using principal parts in their finite forms. Your critique is well taken.
It's the characterization of the perfect as "stative" that shows a strong influence from Porter. By contrast, the terms of perfective and imperfective are standard and not problematic.

As for the other issue, I would just pointing that I would have preferred the article to have discussed the pros-and-cons of the indicative forms, or at least flagged the issue for further discussion. In my experience, asking students to compute inflections (here, the augment, especially when it lengthens the initial vowel) is unrealistic and not good instruction. The best forms to lemmatize upon and the best forms to present principal parts may well be different.
Stephen C. Carlson, Ph.D.
Melbourne, Australia
S. M. Kraeger
Posts: 2
Joined: December 24th, 2014, 9:41 pm

Re: Lemmatizing Greek Verbs

Post by S. M. Kraeger »

It's the characterization of the perfect as "stative" that shows a strong influence from Porter. By contrast, the terms of perfective and imperfective are standard and not problematic.
I don't have a dog in that fight, but your point is appreciated. I don't know what I would use other than "perfect tense" for a term. I don't know if "stative" is much worse than "perfect," particularly if so many insist on the practice of using "perfective aspect" for a form other than the perfect (which is fine by me; I'm not against the practice per se, but I am noting the problem from the angle of tense terminology). I did have a discussion with Porter on this issue about 4 years ago and it seems that he would agree with using the infinitive-based system in principle--though I had not yet worked out my research in detail at that point. His major complaint was not with the approach I'm advocating, not with its principle, but with the tremendous work it would take to reverse an already established trend. I think that argument had more merit in the paper-bound days of grammars and lexica, and significantly less merit in the electronic age.
As for the other issue, I would just pointing that I would have preferred the article to have discussed the pros-and-cons of the indicative forms, or at least flagged the issue for further discussion. In my experience, asking students to compute inflections (here, the augment, especially when it lengthens the initial vowel) is unrealistic and not good instruction. The best forms to lemmatize upon and the best forms to present principal parts may well be different.
I agree. It was unfortunate that I did not include that. Still, I think if we are able to help students conceive of the basic verb from the infinitive rather than the indicative in terms of its aspectual meaning, it would not take much to help them conceive of the verb's morphological indicators. I need some statistical data first. I wonder if I could find any takers on that!
SK
nicholasj.ellis
Posts: 25
Joined: June 22nd, 2011, 7:01 pm

Re: Lemmatizing Greek Verbs

Post by nicholasj.ellis »

It's the characterization of the perfect as "stative" that shows a strong influence from Porter. By contrast, the terms of perfective and imperfective are standard and not problematic.
As the managing editor and one of the primary contributors to the BibleMesh Greek project, I can add a little clarity to this particular discussion.

I've had long discussions with Mike, Randall, and others on how to characterize the traditionally "perfect" forms. We've been reticent to simply call it "perfect", due to the overlap with "perfective". Internally to our team, and especially in discussion with Mark Dubis (our primary Greek editor), Randall, Stephen Levinsohn and Steve Runge, and Chris Fresch (at Cambridge) we've taken to calling this aspect "combinative", reflecting the combination of both perfective (sigmatic) and imperfective (reduplicative) morphology and semantic meaning. I believe this understanding was also represented in Randall's paper at SBL this past November. For now, we've settled on Stative terminology, though with reservations. I think we'll either end up with "combinative" if we're feeling brave, or simply "perfect" to identify the form and "combinative" to describe its aspectual semantics.

In short: theoretically, our understanding of tense/aspect in general, and perfect in particular, diverges substantially with Porter's approach (indeed, the starting points are quite different). This is reflected in the paedagogical table which Kraeger reproduced in his paper (in conversation with me), which does see tense reflected in the augment. On the so-called "statives", we are not intending to communicate that the "perfect" forms communicate only "stative aspect", at least insofar as Porter would identify stativity (i.e., a purely ongoing state of affairs). Rather, we describe this aspect as perfective in the event, but imperfective in its relevance (though some middle forms do seem to deemphasize the perfective event, hence the lack of a sigmatic aspect marker).

We find it quite helpful to teach verbal semantics and morphology from the non-indicative forms; I personally prefer the infinitives. We have also built out a lexical list based on the infinitive, though I'm afraid that we still continue to refer to "traditional" lexicographical categories ("lexical forms", etc.) due to the needs of our academic partners. I hope that in the electronic world this will cease to be a problem, and we can move towards a more semantically oriented lexicography.

I should add that when I wrote that chart for Kraeger in an email I erroneously typed λελυον rather than λελυκε. The error was entirely my own, and should have been copy-edited more carefully by me before dissemination.
RandallButh
Posts: 1105
Joined: May 13th, 2011, 4:01 am

Re: Lemmatizing Greek Verbs

Post by RandallButh »

Thanks for the background, Nick.

One addendum: Since my 2014 SBL paper was cited, let me clarify. In the paper I rejected the term 'stative' in passing, because it was already a semantic category of verbs. It would be unfortunate to use the term as a semantic label and also as an aspectual category.

In other words, "stative" should not be used for the παρακείμενος χρόνος/παρακειμένη ὄψις because 'stative' is already widely used as a lexicographical category, stativity refers to the meaning of particular lexemes, an aksionart, a subset of semantic states of affairs -- and thus should not be used as an aspect category.

Morphosyntactic aspect categories, of which Greek has three (ἀόριστος, παρατατική, παρακειμένη, calling the future ὁ μέλλων a tense and a mood), need to be kept distinct from the lexical meanings of verbs. Otherwise we will have students twisted in knots. What happens when a dynamic verb is placed in the perfect (aka 'stative'?) aspect? Or, what happens when a stative verb is placed in an aorist or imperfective aspect? Do we want people saying that δέδωκεν and πεποίηκεν are stative and also that ἔκειτο and ἐκοιμήθη and ἠδυνήθην are statives?
nicholasj.ellis
Posts: 25
Joined: June 22nd, 2011, 7:01 pm

Re: Lemmatizing Greek Verbs

Post by nicholasj.ellis »

One addendum: Since my 2014 SBL paper was cited, let me clarify. In the paper I rejected the term 'stative' in passing, because it was already a semantic category of verbs. It would be unfortunate to use the term as a semantic label and also as an aspectual category. In other words, "stative" should not be used for the παρακείμενος χρόνος/παρακειμένη ὄψις because 'stative' is already widely used as a lexicographical category, stativity refers to the meaning of particular lexemes, an aksionart, a subset of semantic states of affairs -- and thus should not be used as an aspect category.
Thanks Randall. That's basically the conclusion we're reaching as well.

Historically speaking, we've disliked the ambiguity of the aspectual category "perfect" when juxtaposed against the tense-based category "perfect" (noting also the "pluperfect"). Thus, influenced by the state of scholarly terminology a few years ago (probably more by Peter Gentry than by Stan Porter, to tell the truth), we went with the category "stative". So, in attempting to avoid the particular ambiguity of the aspect/tense label ambiguity, we ran into the ambiguity you note above. Rock... hard place.

My current thinking: with reference to the indicative "perfect/pluperfect", rather than labeling the tense/aspect conglomerate as "past/non-past stative" as we've done, and avoiding the perhaps the more accurate but quite unwieldy label "past/non-past combinative", we return then to the label "past/non-past perfect". This allow for "perfect" to become a purely aspectual category, and the tense label to reside as a simple binary. This gives us in the indicative:

past/non-past perfective (the aorist and future tenses; caveats to the future's irrealis complexities notwithstanding)
past/non-past imperfective (the so-called "imperfect" and "present" tenses)
past/non-past perfect (the so-called "perfect" and "pluperfect).

Do I like the confusing redundancy of "perfective" and "perfect"? No, I don't. But it seems to be a simple solution. In an ideal world, though, I could still be convinced to use "past/non-past combinative" for the "perfect" and "pluperfect" tense-forms....

Thanks for the clarifying words Randall, as always.


Back to the topic of lemmatization: I have fought tooth and nail for the use of the perfective, imperfective, and combinative infinitive forms as the basic lemma. This provides the basic morphological data to identify most forms, as well as the most basic semantic values. We went so far as to create a global vocabulary/gloss list for the NT based on these forms. Unfortunately, we've struggled mightily to get this accepted in academic usage I still have hope, though, as I think the benefits far outweigh the standard model. This approach still doesn't solve the question of when to use a dominant middle form, however. I'd be very tempted to list the active and middle infinitive forms of "ιστημι" as different lemmas.
RandallButh
Posts: 1105
Joined: May 13th, 2011, 4:01 am

Re: Lemmatizing Greek Verbs

Post by RandallButh »

Glad to hear of your wanting to separate στῆσαι from στῆναι. These need partitioning in anyone's brain.

And using the aorist, continuative, and perfect infinitives is also what we do in the Morphologia book. Yes, we can live with the multidimensionality that occurs when these aspects occur in the tensed indicatives. The augment really does carry past reference.

And an encouragement: I wouldn't shed a tear for the term "combinative", because it didn't name what it combined. If you want an -ative name
you could coin "resultative." Linguists won't like it for two reasons, a. they already have "perfect," and b., some use 'continuing result' as a sub-category of perfects. we would need to ask, οἴδαμεν if such a term 'resultative᾽ is able ἑστάναι?
Post Reply

Return to “Greek Language and Linguistics”