[b-hebrew] tenses; frequency

Rolf Furuli furuli at online.no
Wed Aug 10 09:40:39 EDT 2005

Dear Peter,

In order to be polite, I use to answer questions that are posed to me. You 
asked a question. I answered it, and I see no purpose in any further 

Best regards

Rolf Furuli
University of Oslo

----- Original Message ----- 
From: "Peter Kirk" <peterkirk at qaya.org>
To: "Rolf Furuli" <furuli at online.no>
Cc: <b-hebrew at lists.ibiblio.org>
Sent: Wednesday, August 10, 2005 11:40 AM
Subject: Re: [b-hebrew] tenses; frequency

> On 10/08/2005 06:39, Rolf Furuli wrote:
>>Dear Peter
>>A very good example of my claim that when you start with four (presume 
>>four), you end up
>>with four, is the study of W. Th van Peursen (2004) "The Verbal system of 
>>the Hebrew Text of Ben Sira".  On p. 7 he writes:
>>In the French edition Joüon had already indicated the importance of 
>>viewing the Hebrew conjugations together as a system. This means that 'the 
>>value of a verbal form is brought out by its contrast with the other 
>>forms. In Hebrew, as in any other language, verbal forms limit each other 
>>reciprocally'. Once we have acknowledged that the conjugations, like any 
>>other element of the language, are interrelated terms of a system, we 
>>should not search for the meaning of, for example, the perfect, but ask 
>>ourselves what in a given context where we find a perf. the meaning of the 
>>impf. or the ptcp. would be. (Italics his.)
>>The words above close the case: There are four conjugations! ...
> Sorry, but I don't see it. Instead there is a very reasonable assumption 
> that different forms are likely to have different meanings. In fact he 
> doesn't even say this, because he doesn't rule out that two forms will be 
> discovered to be synonymous. He mentions only what we would call QATAL, 
> YIQTOL and the participle, and I think we all agree that these are 
> semantically distinct. There is no mention here of prefixed vav changing 
> the meaning.
>>... Particularly was I sceptical to the claim that an element which seemed 
>>to be the conjunction WAW could turn the meaning of a verb form to the 
>>very opposite, because a parallel to this is lacking in any other 
>>language, including the Semitic ones.
> I understand that you would want to examine such a claim critically; on 
> the other hand, in natural sciences very many things which seem much odder 
> than this turn out to be true. In fact other Semitic languages do in a 
> roundabout way support the idea that WAYYIQTOL is semantically very 
> different from YIQTOL, because they indicate that the shortened form of 
> YIQTOL, which is the basis of WAYYIQTOL in the great majority of cases in 
> which the distinction survives, is an originally distinct verb form from 
> the regular long YIQTOL, although the two came to coincide in form in most 
> verbs in Hebrew. Thus we would expect WAYYIQTOL (short) to be semantically 
> quite different from YIQTOL (long).
>>The mentioned scepticism was one reason why I started the work with the 
>>dissertation, and the working hypothesis was that the traditional view is 
>>wrong. However, to have a working hypothesis does not close the case. To 
>>the contrary, a working hypothesis should be modified or even changed on 
>>the basis of data. ...
> The problem here comes when the data is insufficient to decide between the 
> working hypothesis and another hypothesis, which for the sake of argument 
> can be the four or five verb form system. Am I right in understanding that 
> you started with a working hypothesis, analysed a lot of data, found 
> nothing to disprove your working hypothesis, and so concluded that your 
> working hypothesis is correct? Unfortunately this method is logically 
> incorrect. For it is equally possible that I might start with a different 
> working hypothesis, analyse the same data, and find nothing to contradict 
> my working hypothesis. So we have no way of choosing between the two 
> hypotheses, and can conclude only that, unless more data or new methods of 
> analysis can be found, it is not possible to decide which of these two 
> hypotheses is correct.
> Your method sounds rather like the method of induction for which you 
> rightly criticised Vadim yesterday. You start with a working hypothesis 
> that all swans are white. You then examine a limited corpus of data and 
> find nothing to contradict that hypothesis, and then proclaim that it is 
> proved, and that my alternative hypothesis that swans may be either black 
> or white is disproved. Well, the difference is that you claim to have 
> examined all of the verb forms in a corpus. But a corpus is not the whole 
> of a language, it is simply a large body of sample data. And in fact the 
> more serious problem with using a corpus from a dead language is that you 
> have no way to determine which alternative sentences would in fact be 
> ungrammatical or have a distinct meaning. That is, you have no access to 
> native speaker insight. I would suggest that because of this your method 
> is incapable in principle of distinguishing real semantic distinctions 
> from pragmatic ones, and so incapable in principle of falsifying your 
> working hypothesis.
> In other words, my suggestion is that you started with a working 
> hypothesis, tested against a body of data which although large is in 
> principle incapable of falsifying your hypothesis, failed, of course, to 
> falsify your hypothesis, and then claimed that your hypothesis had been 
> proved. Do you have an answer to this suggestion?
> ...
>>Any researcher will be influenced by his or her beliefs, philosophy, and 
>>biases,  A balanced scholar tries to curtail these as much as possible, 
>>but objective research is non-existent. However, to be sceptical to 
>>traditional explanations and to use a working hypothesis indicating that 
>>they probably are wrong does not close the case in favor of two 
>>conjugations. This is so because the existence of just two conjugations 
>>are not used as an axiom.
> Fair enough. I accept that your working hypothesis has not been disproved 
> either, and so that no one should assume that it is incorrect. But I think 
> you might have done better to work from the traditional model as a working 
> hypothesis, as then you could have attempted to falsify that. I think you 
> would have failed, not because the traditional model is correct but 
> because your method is incapable of falsification.
> -- 
> Peter Kirk
> peter at qaya.org (personal)
> peterkirk at qaya.org (work)
> http://www.qaya.org/

More information about the b-hebrew mailing list