Biblical Hebrew Forum

[James Christian <jc.bhebrew AT googlemail.com>]

Hi all,

I've been written to off list and this has stimulated me to feel the need to
clarify some misconceptions. I will attempt to make this visual to make what
I am saying more clearly. Imagine that a basic 'b' phoneme's wave looks like
this:

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\

In reality it would be much different but I'm just intending to show that
you might see in a speech signal a length of waveforms which look similar
(with very slight differences) that repeat for a length of time. Now let's
add a vowel to the speech signal. Let's say that ba, bi and bu look like the
following:

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\-+//////////////////////

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\=_''''''''''''''''''''''''''''''''''''''''

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\"?]]]]]]]]]]]]]]]]]]]]]]]

OK. Now we have a visual we can look at we can see that it is fairly routine
to segment the 'b', 'a', 'i', and 'u' sounds. But where do they begin? And
where do they end? Looking closely in the intersection of the speech signals
we see that the speech signal our perception associates with a 'b' sound and
a vowel sound each intersection is unique. The reason for this is simple. We
form speech with an almost constant stream of air (except for stops and such
like). We make the different sounds by articulating our tonges, lips, teeth
etc. such that the desired sound is formed. But when we move from one sound
to another there is a brief moment in time where a unique sound is formed
when moving from the position required for one sound to the next. Now, if we
were to make a speech signal which ignores that then we would end up with
the following speech signals:


/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\//////////////////////

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\''''''''''''''''''''''''''''''''''''''''

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\]]]]]]]]]]]]]]]]]]]]]]]

Can you spot the difference visually? Yes! That's right! The small
intersections are missing. When we build speech synthesis systems with this
kind of methodology it is quite easy for a human to perceive something is
wrong with the speech signal. This shows a level of perception. And yet, the
IPA doesn't have names or symbols for intersections. This is because of
different levels of perception. We may hypothesis a ladder of perception.

Words-->Morphs-->Syllables-->Consonant and vowel phonemes-->phoneme
intersections

We can see that the different levels of perception have different
importance. Even though a speech signal will sound wrong to a listener if we
don't handle the intersections it doesn't impede the listener from
understanding. But at the same time the level of perception is not strong
enough for us to consider it worth incorporating in our alphabets.

And so, the question is 'What level of perception were the implementers of
the Hebrew alphabet using?'. The structure of the alphabet would suggest
that they were working at the level of perception of the syllable. That is
to say just as we do not feel the need to encode the differences in the
intersections between our vowels and consonants they did not feel the need
to encode the differences in the intersections between syllables (i.e.
vowels).

Observations like how Ugaratic had three symbols for aleph only give further
evidence to people working at this level of perception. Notice that they
made three different symbols for 3 different syllables NOT 3 different
syllables for 3 different vowels.

James Christian