Biblical Hebrew Forum

[Rolf Furuli <furuli AT online.no>]

Dear Stephen,

Please see below.


> Dear Rolf,
>
> You haven't understood my question. I know you don't believe 
> WAYYIQTOL is ever past tense, and I believe I understand the terms 
> we are working with here. I'll try to rephrase both my question(s) 
> and my example.
>
> Actually, I'll start with the example. Obviously, the English 
> (so-called) past indicative form (went / thought / swam / walked) is 
> usually a tense (past grammaticalization of location in time), so a 
> name like "simple past" or "past indicative" fits. But what I am 
> trying to show is that in some cases, this form is NOT a tense, but 
> rather a mood (subjunctive). The following conversation may 
> illustrate this better:
>
> A: I'm going to drive to work.
> B: If you walked, you would be helping the environment.
>
> For "walked", the deictic centre is the present moment, and the 
> reference time is future - though strictly speaking, it's a future 
> moment in one of several possible "worlds" or mental spaces (one 
> which seems not very likely).

RF:

In B. Comrie (1985) "Tense" p. 9, we find the definition of tense as 
"grammaticalized location in time". On pp. 18-19 Comrie discusses 
examples of the use of past tense that seemingly contradicts the view 
that in past tense reference time comes before the deictic center. He 
mentions counterfactuals and polite expressions. But as Comrie said, 
and we all know, there can be special uses of grammatical categories. 
But that does not invalidate the fact that tenses exist, and that 
they are expressed by particular forms.  The reason why 1) and 2) do 
not make sense is the existence of tenses, and the reason why 3) is 
ungrammatical is the existence of aspects.

1) #Matt ran as I am speaking.

2) #Matt ran tomorrow.

3) #Matt has run last week.


The fact that tenses may have special uses is of course a 
complicating factor in the analysis of a dead language.

People who are native speakers know their mother tongue, and 
intelligent communication can be performed. Please look at 4) below. 
On the basis of a knowledge of the meaning of the imperfective 
aspect, and because subject and object are singular, a native speaker 
would take the clause to mean that Rita was on the point of reaching 
the peak (A person who did not fully know the nature of the 
participle would perhaps give it a different interpretation.)

4) Rita was reaching the peak.

I use this example in order to show that  to understand the 
subtleties of a language, we must have an accurate understanding of 
all its different parts. And that is the problem with a dead 
language, because we can only get an understanding of its different 
parts on the basis of induction, and induction is not always reliable.

I started my study of the BH verb with the assumption that a 
difference in morphology suggests a difference in meaning: The 
participles has a meaning different from the infinitives, and the 
prefix forms have a  meaning different from the suffix forms. My 
first test of the verbal system was for tense, and it consisted of 
finding whether one or more of the verb forms had a uniform temporal 
reference. (I would like to stress that in my view there are some 
parts of any language that represent semantic meaning that do not 
change, but by far, most of the parts of a language are pragmatic; 
meaning and communication are dependent on the context.)  A test 
should not only be quantitative but be qualitative as well. So my 
test for tense did not only consist in counting examples. The first 
step was to look for an unfinished grammaticalization process. If 
such a process could be shown, a particular form could have one 
meaning in an old text and another meaning in a younger text. My 
conclusion was that no such process was discernible, so the basic 
assumption that forms with the same morphology have the same meaning, 
could be upheld.

The only candidate for tense was the WAYYIQTOL form, because I found 
that 93.1% of the cases had past reference. Two possibilities had to 
be assessed: The 997 forms with non-past reference could be 
exceptions, as Comrie has shown exist in all languages, or they could 
represent real evidence against WAYYIQTOL being past tense. Several 
factors were considered in order to answer these questions, such as: 
Could verb meaning be different in prose and poetry? Would a 
diachronic mapping of the forms show a pattern that could illuminate 
their use? In both cases the answer was No. And because the forms 
could not be viewed as special cases, the conclusion was that their 
existence falsified any view that WAYYIQTOL was past tense.



> Anyway, the point is that the form which grammaticalizes "past" in 
> the majority of constructions/contexts does NOT do so in this type 
> of construction (and others): in this construction it is a 
> subjunctive verb, often with future reference. However, this more 
> rare modal use does not automatically disprove that, most of the 
> time, the form is indeed a past tense.
>
> Hence my conclusions: (1) the presence or absence of tense 
> (grammaticalized time) is not inextricably tied to a particular verb 
> form; that is, one form can have two mutually exclusive TAM 
> functions, and (2) the "semantic value" of a verb form must take 
> into account wider contextual considerations (or better, wider 
> constructional considerations).

RF:

I disagree with both (1) and (2).  The very definition of tense as 
"grammaticalization of location in time" excludes the view that tense 
is pragmatic. One way a grammaticalization process proceeds is that a 
word form looses meanings or uses and becomes more and more 
specialized, until it has only one meaning and is fully 
grammaticalized. Special uses of the form can occur, but do not 
contradict its grammaticalization.

I suggest the following test: Start with a group of ten person and 
expand it to one thousand persons. Show them the word  "walked" and 
"went" without a context. and ask for an analysis. I am quite certain 
that all the native speakers would say that both are past tense. A 
context is not needed for such a conclusion.


>
My question, then, relates to the assumptions in your methodology. Do 
you assume, in your analysis, that a given verbal form (in this case, 
WAYYIQTOL) must always have the same core TAM function/meaning? Or do 
you consider the possibility that its "meaning" could be different in 
different grammatical contexts, just like the -ed form in English? 
For instance, Waltke and O'Connor's analysis of YIQTOL has it as 
sometimes imperfective and sometimes modal.

> Actually, I'd like to broaden the question and open it to anybody, 
> because even if I'm raising a valid methodological point, it may 
> have no practical relevance here. Is there any value in posing this 
> question with respect to WAYYIQTOL? Is anyone aware of studies which 
> have considered the possibility? And if the answer is "no value", I 
> humbly bow out.


RF:

Your question, as I now understand you, does have value and is 
relevant. The assumption I started with that there must be a 
difference in meaning between the suffix forms and prefix forms, is 
general, and does not say anything regarding the nature of these 
meanings. The advantage of using the fundamental parameters deictic 
center, event time, and reference time, is that no assumption 
regarding tense, mood, or aspect at the outset is necessary. Having 
found that the prefix forms (WAYYIQTOL, WEYIQTOL, and YIQTOL) have 
the same meaning ( imperfective meaning), and that the suffix forms 
(QATAL and WEQATAL) also have the same meaning ( perfective meaning), 
I will expect that the forms behave as the imperfective and 
perfective aspects respectively. But I will not expect that they 
always have the same modality or temporal reference.


> Best regards,
> Stephen Shead
> Centro de Estudios Pastorales
> Santiago, Chile


Best regards,

Rolf Furuli
University of Oslo