[B-Greek] Is Greek Present a tense? Was Verbal Aspect theory -- misgivings

Kimmo Huovila kimmo.huovila at kolumbus.fi
Wed Nov 19 15:03:52 EST 2008

I realize I go perhaps off-topic to linguistic methodology, but that is 
crucial to building any theory of the Greek tense.

Perhaps our exchange is an illustration of Carl's frustration about 
terminology among linguists... But I would say that for the most part 
linguistic terminology is not very confusing, in my opinion. More confusing 
is some (not all) theories of the Greek verbal system and their formulation. 
But perhaps that's just me and shows how I get confused. :-)

On keskiviikko 19 marraskuu 2008, Rolf Furuli wrote:
> >But we do know a whole lot since we have quite a bit of text written using
> >these linguistic conventions! It is just applying corpus linguistics to the
> >study of a dead language. It applies to dead languages just as well as 
> >ones. And by the way, it is very interesting to compare results of corpus
> >linguistic studies to intuitions of several natives (and your own)! That
> >teaches something about both native intuition and corpus linguistics, and 
> >to best use them.
> RF:
> When I approach a dead language in order to find the meaning of the 
> properties of its verbal system, I need no special theories, and 
> corpus linguistics and the intuitions of modern informants will only 
> be disturbing and may lead me astray.

If you do not use corpus linguistics, how much do you have left, since 
we know these languages mainly because we have corpora of them?

What I suggested where modern informants are important is for the testing of 
general methodology. If your methodology only applies to certain languages 
(for example dead) but not others (for example living), you have made already 
huge assumptions in your approach to a dead language. Analyzing living 
languages is very important for the development of linguistic skill. How 
could that lead you astray? Or corpus linguistics? Or perhaps I just did not 
understand you.

> I have to study the dead  
> language in its own right. The only baggage I have, that could 
> resemble a small part of a theory, is that I know that the basic 
> parts of the verbal systems of many modern languages, are tense, 
> mood, and aspect. I do not assume that these must be found in the 
> dead language, but that is my question. The research has two sides:
> THE QUANTITATIVE TEST: The temporal  reference of ALL the finite 
> verbs are mapped on the basis of the deictic center and reference 
> time of each verb.
> THE QUALITATIVE TEST: To find whether particular time references are 
> cause by the context or by the verb form itself (a distinction 
> between semantics and pragmatics).

Perhaps I do not completely understand your qualitative test. How can you know 
by which the time reference is caused? Is the time reference caused, or is it 
just there and signaled somehow to the audience? I'm asking just to make sure 
I understand your terminology and methodology.

If you assume nothing of the verb form, then every time you know that it has a 
particular time reference, it has to be the context, because otherwise you 
would not know, right? Would that mean that you will never know that it was 
caused byt the verb form itself, if you did not know it to begin with? I 
would assume that in a normal, coherent text, the context and the verb forms 
would agree.

How would your methodology be capable of finding out whether, for a 
hypothetical example, a verb form is used in two senses, one being always a 
past reference with an imperfective aspect and the other always a future 
reference with a perfective aspect? Or would you conclude that since both 
occur in perfective and imperfective contexts and in past and future 
contexts, neither has anything to do with time or aspect?

> >I don't think I have implied that we should study a dead language in the 
> >of modern descendant languages. However, I do think that the methodology we
> >use in studying a dead language must yield right results also when applied 
> >a living language. If it does not, the methodology is suspect.
> RF:
> The term "right results" is problematic, because it could imply a 
> pre-judgement (prejudice). We simply do no know what a dead language  
> should look like.

It may not be very problematic in the case of a living language. That's the 
arena where methodologies should be tested, especially if, as you say, we do 
not know what a dead language should look like. In the case of a living 
language, just test the results with a good sample of natives. If there is a 
huge mismatch, then back to the drawing table to figure out whether the 
methodology was incorrect or incorrectly applied.

> RF:
> To *account for* the non-use of a form is rather problematic- we 
> cannot account for something that do not exist.

We need to account for the fact that something does not exist. I take a very 
simple English example (with plenty of natives on the list to call foul if 
the example does not work). The word 'children' is used to refer to several 
instances of 'child'. The normal plural is to add '+s' to a noun (this is 
somewhat simplified, but enough for the purposes of this illustration). So, 
the word 'child' has two plurals: children and childs? Is our linguistic 
theory supposed to be able to explain the fact that 'childs' does not exist 
as a plural in the language? This is a lexical example, but the same 
principle works with grammar, too. A form (like +s) may be used in many 
contexts, but be absent in some. For example, a historical present may be 
used in many past contexts, but be absent in other past contexts.

> But I suppose you  
> mean that we should consider why a particular form (here present) is 
> used in some places and not in others. But in this connection we need 
> to be cautious. For example, more than 50% of past-referring present 
> forms are verbs of speaking, and the percentage varies in different 
> books (Mark 48%, Matthew 73%, Luke 73%, and Acts 85%). This is a 
> matter of style and choice of subject, and there is no linguistic 
> reason why we should *account* for the differences. But of course, 
> the distribution of a form should be carefully considered.
> >How about a reformulation: The present does not always have tense? That 
> >you can safely deduce. But then again, the impefect is not always a tense,
> >either, as there is a secondary meaning of unreality.
> RF:
> Your suggestion regarding a reformulation shows explicitly that we 
> use the word "tense" in very different ways, because "The present 
> does not always have tense." is a contradiction of terms in my 
> linguistic system. We are allowed to use different linguistic systems 
> and different definitions, but it is extremely important that we 
> define our terms. It seems to me that "tense," "time", and 
> "timelssness" have been used on the list in a rather loose and 
> unspecific way, so I think that those who are interested in this 
> subject will profit when I now give a precise definition of what I 
> mean by "tense".
> In this connection I agree completely with Mari Broman Olsen, and I 
> use her description- interested persons can find it in her book.

I do not believe this dichotomy is clear cut. There are conventional 
implicatures, which are somewhat in between. You cannot use conversational 
pragmatic implicature to make entirely correct predictions about usage and 
meaning, even though they help. For example, how would you explain the 
difference between "could you pass me the salt" and "would you be capable of 
passing me the salt" just by using conversational pragmatic implicature with 
no recourse to conventionalization?

We are dealing with a scalar phenomenon, not a strict dichotomy. To make it a 
strict dichotomy requires, in my opinion, a somewhat arbitrary dividing line. 
It is an important theoretical point not to assume a strict dichotomy and to 
be open to the scalar nature of the distinction.

How does an extreme case of very strong implicational conventionalization 
differ from semantic meaning? I think it is just a matter of degree, with 
semantic meaning at one extreme and a non-conventional pragmatic implicature 
at the other?

> The basic point here is that some meanings connected with words, word 
> forms, and clauses cannot be canceled, because they are an intrinsic 
> part - these represent semantic meaning. Other meanings do change 
> according to the context-these represent conversational pragmatic 
> implicature.

Some meanings cannot be canceled, and to cancel others takes more effort than 
to cancel others. This illustrates the scalar nature of the phenomenon.

Also, practical linguistic work with living languages and native informants 
will show that natives may not always agree on these kinds of questions. What 
is uncancelable to someone (and thus they may not accept a sentence that 
explicitly denies an implication or a meaning) may be cancelable to another. 
If this works like this with living languages, could dead languages work 
similarly? Of course natives will agree on a great number of cases, but 
sometimes one native may feel that no way any native will accept a sentence 
that another native, in fact, accepts.

> Examples of semantic meaning are words that are marked for 
> durativity, telicity, or dynamicity; these properties cannot be 
> canceled. But stativity and punctiliarity are changeable.
> The word "time" is linguistically speaking rather vague, and can be 
> used in different senses, but that is not true with "tense". Comrie 
> defines it as "grammaticalized location in time," and this indicates 
> that its time reference is uncancelable.

OK. For example, the imperfect is uncancelably past except in counterfactual 
conditionals? And the present is uncancelably non-past except in 1) 
historical presents 2) etc...?

> A diachronic study of a  
> language can reveal grammaticalization processes: a linguistic entity 
> gradually looses its multiple meanings, and is used more and more 
> uniformly until it is fully grammaticalized and has one uncancelable 
> semantic meaning.
> Thus, a verb form that represent past tense, always  
> signals past tense; this tense cannot be canceled by the context.
> A  
> verb form that is not a tense may have different time references. (I 
> have already shown that special cases, such as hypothetical 
> conditional clauses, do exist. But the English word "went", for 
> example, which has an uncancelable past tense, does not loose this 
> property even in the special cases.)

If I went there tomorrow, everyone would be glad?

> The advantage of the distinction described above and a scrupulous 
> distinction between semantics and pragmatics, is that clear-cut 
> definitions can be given, so we all can know what we are discussing. 
> Without clear-cut definitions -then panta rei! For me, therefore, it 
> is completely unacceptable (and a contradiction of terms) to say that 
> a verb form some times is a tense and sometimes not.

OK. How then would you describe the imperfect? It is clearly not a past tense 
in counterfactual conditionals, but is clearly a past tense in other 
contexts. But if that formulation is completely unacceptable, then suggest a 
better way of putting it.

> RF:
> Your explanations to my examples from Mark 11 may or may not be 
> correct. But they are all tied to one thing, namely, to discourse 
> factors. And there is a very big difference between hypothetical 
> conditional clauses and theme, rheme, and stylistic factors. These 
> are discourse factors which hardly can be "special cases" from a 
> grammatical point of view. I would rather say that such a claim is 
> special pleading. Let me illustrate the case. No person would use 
> English futures in a narrative account for discourse reasons; that 
> would be confusing. And no person would use English simple past in 
> future descriptions. And similarly, when Greek present is used with 
> past reference, this is good grammatical language, and a supposed 
> non-past meaning is thereby canceled, and it is shown that present 
> does not have any tense "grammaticalized location in time".

The supposed non-past meaning is canceled by the past narrative frame.

> Greek aspects are often viewed differently by different people, but 
> the terms are still in use, and they give us much insight. So why 
> should we not be satisfied with the definition of Greek present as 
> "the imperfective aspect" and nothing else? This meaning may be 
> viewed as too general, but it is better to have one general meaning 
> that covers all the examples, than to use a specific meaning that 
> contradicts some examples. Inside the imperfective definition we can 
> describe all the different uses of Greek present.

What examples does my explanation of the present contradict?

> And lastly, for the record: Saying that Greek present does not have 
> tense, is not the same as to say that it is timeless. It is not! But 
> the point is that the past, present and future references is caused 
> by the context (being pragmatic) and not by the verb form itself 
> (being semantic).

I am not sure what it means that a reference is caused by the context. The 
reference is what it is, but it is signaled to the audience by different 
means, including context and grammatical forms etc. Grammatical forms may be 
part of that signaling even if the signal is cancelable.

Kimmo Huovila

More information about the B-Greek mailing list