Biblical Hebrew Forum

["Yitzhak Sapir" <yitzhaksapir AT gmail.com>]

On 1/16/07, Peter Kirk <wrote:
> On 16/01/2007 18:57, Yitzhak Sapir wrote:
> > ...
> >
> > When graphed, the data clearly show that there is a nice upwards curve.
> > Resh's data point appears way out there because there are so many
> > number of roots that have Resh in Peter's list.  The question is: why is
> > there such a nice curve?  why do the data for the letters cluster around
> > a curved line? ...
>
> I suspect that the graph follows something like a 2/3 power curve
> passing through the origin. This is because of how I think the
> statistics work, but it does fit the graph reasonably well. Passing
> through the origin is essential because zero roots implies zero
> homonyms. Indeed the data point for he with mappiq is also a good one:
> just eight roots implies zero homonyms. But the calculation needs to be
> done separately for first, second and third root letters. The figures in
> column X of my spreadsheet are based on this hypothesis. These figures
> are fairly constant, as would be expected. The high figure for sin on
> its own is because of the four homonyms for sin as the second root
> letter of which three are (alleged) homonyms of (&H.

Having reconsidered, I think that a power curve is simply too weird.  I find
a hard time explaining why the data would model a power curve.  While a
curve does appear to be present in the data, and a power curve does
model it well, a linear analysis can be more easily explained.  For example,
my own explanation of the mergers of letters explained how linearity was
maintained in such a process.  Also, if the (8,0) mappiq heh point is
dropped (and the reasoning of its exceptionality have been noted), we still
get 97% correlation.  It is much easier to explain why it is not legitimate
to analyze the mappiq heh together with the rest of the data, than it is to
explain the behavior in terms of a power curve.  In light of that, I prefer a
linear analysis.

Issues like three roots for "(&H" are not extremely problematic.  Because
this type of methodology is applied to all roots equally, regardless of which
phoneme is present, it is ok.  On the other hand, it illustrates especially
why no one else but Karl can perform the statistical study for him and why
if someone else would have performed a statistical study, such a study
might still not convince Karl -- no one else will use his own methodology
for identifying distinct meanings.

What the statistics do show is that viewing all words as homonyms is
in line with what we would expect, and viewing none or very few as
homonyms is quite unlikely.  The less words that are considered
homonyms from among Shin and Sin, the further the Shin/Sin datum
moves from the predicted location on the graph.  If even half of them
are considered homonyms, that is quite inconsistent with Karl's
original phrasing of the issue as "lack of homonyms" or "how seldom
Shin and Sin appeared in homonyms."  One may not like the graph
or disagree with the very fact that there is such a graph and there
appears a definite relationship between number of roots and number
of homonyms, but as the roots were not identified as separate
homonyms with these statistics in mind, the deviation of viewing
none (or very few) of the Shin/Sin pairs as homonyms has to  be
explained.

I do not know if Karl can "rescue" his theory still.  I think there are
problems with your suggestion ("resolution of homonym pairs") because
there are equally as many "unresolved homonyms" and also because of
the appearance of Shin and Sin in different forms in cognate languages.
Any such "rescue" or restatement however has to be analyzed in the way
that it is restated.  For now, I haven't seen Karl restate his theory.  It may
be that in light of the numbers and the graphs, he is taking time to review
and perhaps reconsider it.

Yitzhak Sapir