whatisaword? (dictionary and beyond)

yochanan bitan ButhFam at compuserve.com
Sat Dec 25 11:48:18 EST 1999

PK:>2) Mainly for intermediate (and advanced) language learners as you 
>mention below:  ... I am sure that it is helpful 
>for these learners, on coming across an occasional form of hishliaH, 
>to be able to relate it to shalaH and shilleH.

agreed. as you mentioned i mentioned.

>Also, more theoretically, the form hishliaH does not occur, so it is 
>even more a construct than the root is! 

Here we may part on views of what a language is. 

for me, if yashliaH exists, then hishliaH and any other predictable forms
"exist" in the competence/langue of the language user. the decoder is one
side of the communication equation.

on the other hand, if nimlaT 'escape' exists, one CANNOT postulate *malaT. 
that was not part of anyone's competence, as far as we know. 

as you can see, i am asking that we treat BH like a real human language.
when a person uses a dictionary of any language they do not run out to
check attestations, they process according to langue.

incidently, the contextual forms are shillaH, and yidda` (as pi`el), 
forms with "e" would be 'pausals' [shilleaH, yiddea,] which i treat as
  but maybe you've got a point:
shilleaH.  hu shillaH oto. 
(do you like that?)
on hitvadda`. reveal himself. i wouldn't  join to the hif`il, even if
semantically close. there are quite a few hif`il that have
reflexive/passives in either nif`al or hitpa`el. but i treat them as
unpredictable and therefore needing separate listing. 

>The basic disagreement is that I do not even agree with the first 
>statement of yours that I have retained below. It is certainly 
>over-simplistic. For example, surely the niphal of a verb like yada` 
>acts as the passive and/or middle of the qal form rather than as a 
>different verb, and its hiphil acts as the causative of the qal. To 
>list the niphal and hiphil as separate words would be most confusing, 
>especially if there is no reference back to the qal to which they 

i think you've over-read, or under-read, my one-liner, which hadimplicit
assumptions (notice my statistics on 'qal' included nifal numbers!). so we
almost agree on this. i would not list separate nif`al if they were only
the passive of qal. i would list nif`al that were reflexive/passive of a
hif`il or had no qal.

on hodiia`, though, i would list it separately. the fact that it is the 
causative of a qal is irrelevant to the fact that it exists and its 
existence was unpredictable, even if expected.

>We are used to drawing a clear line between infectional and 
>derivational morphology and deciding on that basis what merits a 
>separate dictionary entry, but unfortunately that sort of division 
>simply does not work with Hebrew.

here we disagree. prefix/suffix are inflectional. 
binyanim, ignoring passives, are derivational. listable. 
(by the way, in arabic the passives are NOT given binyanim status. 
but 'nif`al' [which did not take over the qal passive in arabic] 
is given status, as well as reflexive 't' forms.) 
another rule of thumb on the 'passives': when in doubt, list it.

a parting note: i do notice that at least you list each binyan under a root
separately. you may be tacitly acknowledging their 'word' status, as
opposed to "inflected roots". 
like i mentioned, that is what hebrew-hebrew dictionaries do without any
the point i am trying to isolate and excise is the 'inflected root' idea
widely generated among students learning hebrew. you may be close to
agreeing on this.

randall buth

More information about the b-hebrew mailing list