Biblical Hebrew Transcription (was: Jeremiah 23:2)

Kirk Lowery KirkLowery at
Fri Sep 18 12:12:39 EDT 1998

Paul, Ben (and others),

Could I give some historical perspective on this issue?

Back in the late 1970s, just before the dawn of the Information Age, I was
working on my dissertation and became involved in the Univ. of Michigan project
of creating a machine readable text of the Hebrew Bible, led at that time by H.
Van Parunak. I remember clearly the discussions at the time of how to represent
the Hebrew text in transcription for electronic access.

We were limited at that time by the constraints of a 7-bit ASCII set of codes.
Another issue was the fact that any true phonetic transcription inevitably
involves the subjective judgment of the coder, with all the problems that
brings. Finally, we could not predict how future researchers might want to
use/access/modify the data in the future. Since the text itself is so basic, it
was decided to encode only the surface, graphic representation of the text. Our
rule was "Code what is WRITTEN, not what is MEANT."

This is why qamets and qamets chatuf was not distinguished in the coding, and
only one symbol, "F" was used. That was because the same symbol was used by the
Massoretes when they wrote the text. The same rule applied to vocal and silent
shewa. The colon was used for both.

Ok, so much for ancient history.

When this list was first founded on the old <> server, this same
issue came up.

At that time I suggested that we use the Michigan-Claremont-Westminster BHS
transliteration scheme for the following reasons:

1. The typing of Hebrew is a complex, error-prone task. With an easily available
electronic BHS, we could simply cut and paste the appropriate selections of text
under discussion.

2. We *must* assume that readers of our postings have the Hebrew Bible at their
side to consult as they are reading the text. Therefore, scientific accuracy of
transliteration isn't important. This list is an informal discussion. As long as
folks understand what is intended, precision isn't mission critical.

3. There was (and is, IMO) no need to enforce absolute conformity for the
transliteration. As long as the writer is clear, the membership can usually
figure out what is meant, identify the word(s) under discussion, etc.

Over time, we've found that speakers of modern Hebrew prefer the "v" for waw,
some capitalize the the consonants, leaving the vowels in lower case, mark vocal
shewa, etc., or use variants for various letters. I've even come to prefer the
apostrophes " ' " and " ` " for alef and `ayin instead of the Michigan's use of
parentheses, " ) " and " ( ", simply because it is visually easier to read. Some
don't like " $ " for shin and " & " for sin, and I personally have found using
" + " for tet to confuse various programs, including my own! :-(

Since our list membership is world-wide, we have to operate on the lowest common
denominator, i.e., 7-bit ASCII. Until we can be sure that 99% of the world uses
XML-aware email clients, we won't be able to use Hebrew fonts or fancy
transcription and transliteration schemes. So our situation is basically the
same as it was in 1980. What to do?

Why, nothing at all! It isn't possible to enforce conformity, and really isn't
necessary. Generally, all transliteration systems I've seen used here are
variants of the Michigan system. (Yours, Paul, distinguishes some sounds at the
phoneme level, e.g., qamets and qamets chatuf.) If we just say that generally
some variant of the Michigan system is used, that ought to be (and has been, up
to now) enough for folks to understand each other. I can read and understand
yours, Paul, but I have no intention of retyping the Hebrew, when a couple of
clicks of the mouse will do all the work for me. Yes, if the discussion drops
from the text level to a discussion of Hebrew morphology or phonology, then we
need a consensus worked out for that. But that can be done by the participants.
There are electronic (ASCII) versions of IPA that could be used. Or we could
easily adapt the transcription system of the Society of Biblical Literature,
which would be adequate for most discussions of Hebrew phonemes. I suggest these
because anyone who has had taken courses in linguistics or Hebrew are likely to
be already familiar with these systems.

As for myself, I will continue to use the basic Michigan system. I already have
the text in electronic form; I use it daily. A graphic system like this is also
the least confusing for the beginner and won't confuse the trained academic or
modern Hebrew speaker--most on this list will be able to understand me. 

Here is the official Michigan transliteration system:

Consonants          Vowels
----------          ------
Aleph    )          Patah         A
Beth     B          Qamets        F
Gimel    G          Segol         E
Daleth   D          Tsere         "
Heh      H          Hireq         I
Waw      W          Holem         O
Zayin    Z          Qamets Chatuf F
Cheth    X          Qibbuts       U
Teth     +          Shureq        W.
Yod      Y          Shewa        :
Kaph     K          Hatef Patah  :A
Lamedh   L          Hatef Segol  :E
Mem      M          Hatef Qamets :F
Nun      N
Samek    S
Pe       P          Miscellaneous
Ayin     (
Tsade    C          Ketiv        *
Qoph     Q          Qere         **
Resh     R          Dagesh       .
Sin      &          Meqqeph       -
Shin     $
Tav      T

Here is 1 Kings 1:1 in encoded form:


If this is the "ball park" system, I don't think we'll have any problems with
how people will vary from this. For newcomers to our list, I recommend at least
learning the Michigan system. Whether or not you choose to use it, you will
certainly be able to figure out anyone's transliteration.


Kirk E. Lowery, Ph.D.                              voice: (215) 572-3845
Associate Director of Academic Computing             fax: (215) 887-5404
Westminster Theological Seminary              email: <KirkLowery at>
Philadelphia, PA 19118

More information about the b-hebrew mailing list