Biblical Hebrew Forum

[Kimmo Huovila <kimmo.huovila AT kolumbus.fi>]

On Sun, 01 Sep 2013 18:23:21 +0200
"Rolf" <rolf.furuli AT sf-nett.no> wrote:

> Kimmo raised  methodological questions, whose answer is extremely important 
> for the study of Hebrew verbs: Is there a a difference between "tense" and 
> temporal reference? Can a verb in one context be a tense, and in another 
> context be an aspect, thus loosing its tense? If I understood Kimmo 
> correctly, his opinion is that in no language do we find a verb form that 
> always have past reference. The best example of the opposite that I am 
> aware of is Greek imperfect, which consistently has past reference. If this 
> is true, one cannot argue that the non-past references of WAYYIQTOLs do not 
> prove that WAYYIQTOL in not past tense, and that the future references of 
> QATAL do not prove that QATAL is not a past tense, because in all languages 
> there are such exceptions. Linguistic theory and methodology with examples 
> from different languages can be discussed if they are relevant for the 
> study of Hebrew.
> 
> In the work with my doctoral dissertation, I found 997 WAYYIQTOLs with past 
> reference, and 956 QATALs with future reference. These were found in normal 
> contexts. I also found that YIQTOLs, WEYIQTOLs, and WEQATALs have past, 
> present, and future reference.  All this show  that tense (grammaticlized 
> location in time) is non-existent in Hebrew.
 

Dear Rolf,

You recognize that my argument is methodological, and you have understood me 
fairly well. (The only quibble I have is that I did not make a claim that in 
NO language do we find a verb form that always has past reference – I only 
implied that I don't expect it to be common for grammaticalized past tense. 
But this is irrelevant for the purposes of our discussion.) I attempt to 
address the methodological concerns using a hypothetical case.

(Another good candidate besides the Greek imperfect would be the Finnish 
imperfect. In neither language is the past reference 100 %, but Finnish might 
be even better, since it is not used for counter-factuals. Finnish uses the 
conditional mood instead. But this is also quite irrelevant for the purposes 
of this discussion, but enlarges the database for your argument a little bit.)

As for Hebrew, I think you rightly recognize some of the implications my view 
may have for your study of Hebrew, but I admit I have not read your 
dissertation, so I do not know for sure if the implications stand. At least 
it is something to think about.

On Sun, 01 Sep 2013 09:11:57 +0200
"Rolf" <rolf.furuli AT sf-nett.no> wrote:

> In all kinds of scientific research we need clear definitions and we need 
> controls. If it is true that "a form represents tense only in certain 
> aspects or aspect only in certain temporal references," we have linguistic 
> anarchy. Then we need another definition of tense than "grammaticalization 
> of location in time." And further, there are no controls, because we cannot 
> test the interpretation of the scholar on the basis of clear definitions. 
> So, in that case, everything goes.

I agree with your call for clear definitions and controls, but do not agree 
with your inference of linguistic anarchy. I see the need another definition 
for tense only if we assume that grammaticalization means 100 % 
correspondence. This is the very point I do not assume.

There is no need for linguistic anarchy in my framework, in neither sense: 
anarchy (1) for the language user or (2) for the linguist.

(1) I think we both aree that there are languages that do not grammaticalize 
aspect, and languages that don't grammaticalize tense. Yet they are capable 
of making the same distinctions that other languages make. They use context 
and lexical material for this purpose.

Now suppose we have (in a hypothetical language) a verb form that is used in 
the past context only when the perfective aspect is appropriate, and in 
future contexts only when the imperfective aspect is appropriate and another 
verb form that is used freely for the future, but in the past its use is 
limited to the imperfective aspect. In such a hypothetical language, 
communication would be no problem. The context will tell you how to interpret 
the verb, once you can infer either tense or aspect from the context. 
Basically if languages can function without grammaticalized tense or aspect, 
they can handle this for communicative purposes. Our hypothetical language 
offers some grammaticalized help for the listener to figure out the intended 
temporal and aspectual reference. No linguistic anarchy is implied in sense 
(1).

(2) How would a linguist go about analyzing such a hypothetical language? You 
would classify a corpus along parameters of aspect and tense, and you would 
see the pattern. Again, no linguistic anarchy. Would the two hypothetical 
forms grammaticalize aspect? Or tense? If we require 100% correspondence, it 
would be neither. Actually the correspondence might be poor. It would not 
follow that we need to throw TA categories out the window and seek something 
else or need to redefine tense or aspect, or end up in desperation. All we 
need to do is to look at TA together.

If our methodology does not allow for this, then it is suspect and we may 
well miss out some real linguistic phenomena. Östen Dahl has argued that 
tense and aspect are often intertwined in real languages.

> 
> It is true that language cannot be treated the same way that we treat 
> mathematics. But still there are clear patterns. Your words about "not 
> 100%" relates to two different situations, 1) normal use, and 2) special 
> situations.  For example, the English forms "bought" and "walked" represent 
> past tense. But under special circumstances, the forms may appear to some 
> not to refer to the past, although, even in hypothetical situations one can 
> argue in favor of past reference, if the deictic center is taken into 
> account.
> 
> Challenge 1: Can you find a normal non-hypothetical English clause where 
> "bought" or "walked" do not represent past tense?

I don't think I need to in order to make the point I am making. I am saying 
there is no 100 % correspondence. You rightly recognize this as a special 
case. You raise the idea of normal use and special situations. Rightly so, in 
this case. How special does the special case need to be? I think we are on 
the right track if we try to see in what situations a form may be used in the 
past~present~future~any other tense and imperfective~perfective~habitual~any 
other aspect and indicative~any other mood. In a language like Hebrew, let's 
look at the whole TMA palette together before saying that T, M or A is 
irrelevant.

> 
>  The default reference of the Greek aorist is past.  But aorist can have 
> present and future reference as well. Therefore I claim that Greek aorist 
> does not represent past tense, only the perfective aspect. But my claim is 
> that the Greek imperfect represent past tense and the imperfective aspect.
> 
> Challenge 2: Can you give one or more examples of Greek imperfect with 
> non-past reference?

Counterfactual conditions. I think you will find examples in good reference 
grammars. But perhaps on a list of Hebrew, we should leave discussion of 
specific Greek texts off-list. If you want to discuss Greek grammar, feel 
free to write to me off-list.

Kimmo Huovila