This help file is organised as a set of questions and answers,
in the style of a "frequently asked questions" document.
- What are the browser and system requirements to view
the Hebrew Bible? (Or, why do I see gnarly ASCII characters
instead of Hebrew letters?)
- One of the principal goals in preparing this on-line
Bible was to enable anybody to read it on the Web
without the need to download and install special
fonts (which differ from one operating system to
another), run a platform-specific application, or
require a Java applet or other non-HTML mechanism.
At the present state of the art, the only way to achieve
this was to employ the "TrueDoc dynamic font" mechanism developed
by Bitstream and
supported in the latest versions (4.0 and above) of
Netscape Communicator and Microsoft Internet Explorer.
Browsers which support dynamic fonts are able to download a
compressed description of the characters in a font and then
use it to render portions of an HTML document marked up as
requiring that font. If your browser does not support dynamic
fonts (or if you have disabled this feature in your browser's
preferences configuration), Hebrew letters will appear as the
ASCII characters with the same codes. When you first display
a document, your browser may also momentarily show the text in
ASCII until the Hebrew font downloads.
To reduce the size of the documents and reduce the
time required to download them, the "Cascading Style
Sheet" feature is used to specify font family and
size for various components of the documents. Your
browser must therefore also support this facility
(in fact, all browsers which include dynamic fonts also
provide cascading style sheets, or at least the very
conservative subset of style sheet features used by
these documents). Make sure you haven't disabled
Style Sheets in your browser's preferences setup.
When operating systems and Web browsers which support
Unicode become
widely available, all of this complexity will evaporate
back into the clouds of chaos whence it originated,
and there will be no difficulty whatsoever including
characters from any number of different languages in
a single document.
- Where did you get the text for the Bible in machine-readable form?
- The Hebrew Bible on-line edition at this site
was based on a transliterated version of the Bible
prepared by Steve Gross. You can download that
document from the
archive directory
at the Shamash site.
Steve Gross employed a somewhat different transliteration
scheme than is used here; a description of the
transliteration is included in the archive.
- Which version of the Hebrew Bible are these documents based?
- The Koren Tanach, the traditional Hebrew
Bible based on the Masoretic texts. The other
principal version of the Hebrew Bible, the
Biblia Hebraica Stuttgartensis
(BHS), is substantially different from the
Koren edition. A machine-readable BHS text is
available from the Center
for the Computer Analysis of Texts at the University
of Pennsylvania, but requires signing a user agreement
and is thus not totally free of strings.
- What happened to the vowels and points?
- The machine readable database upon which these documents were
based follows the usual convention for computer-based Hebrew
documents in omitting all vowel signs and points (dagesh)
within letters. Other than rewriting the entire text by hand,
there is no way to restore this information. Since most
E-mail systems and Web browsers lack the ability, at present,
to render Hebrew letters along with these marks, computer
users have become accustomed to doing without them. I know of
no freely available edition of the Koren text which includes
vowels and points. The BHS text mentioned above does
contain markup for these signs, but it is not available
without restrictions and is regarded as a scholarly text
rather than an authentic Hebrew Bible.
- How accurate is the text?
- One can never be sure of the accuracy of any
machine readable document (or, for that matter, a printed
one) without exhaustively comparing it against
an authoritative edition. In preparing this HTML
edition of the Hebrew Bible, I have used the Steve
Gross text as published, and have not personally
verified its accuracy either exhaustively or statistically.
I did, however, perform a "reality check" which
persuaded me that the text is, if not perfectly
accurate, very close to perfection, at least for
the portion I was able to verify. The test consisted
of comparing the five books of the Torah with an
independently-compiled Torah database which is
considered accurate. That database, however, consists
of just a string of letters--there are no spaces
between words, punctuation marks, or final forms at
the ends of words. Comparing just the letters, then,
the Steve Gross Torah agreed perfectly with the
other database. Erroneous word breaks, etc. may
still exist, but at least the letter sequence is
correct. I have, to date, located no reliable
comparison document for checking the balance of
the books of the Bible.
- So what transliteration scheme do you use, anyway?
- The documents and the corresponding Hebrew font employ
an extended version of the Michigan-Claremont
transliteration scheme. The extensions are the use
of lower case letters to represent final forms of the letters
Kaf, Mem, Nun, Pe, and Tsadi. The transliteration table is
given below, with extensions to the Michigan-Claremont
standard highlighted in blue. Names of Hebrew letters are as
used in the Unicode
standard. (Obviously, if your browser cannot show Hebrew
letters in the other documents, you won't see them in the
"Glyph" column of the following table either.) Many Hebrew
documents on the Web employ a "Web Hebrew" character set which
replaces the ISO 8859-1 accented characters from 0xE0 through
0xFA in the character set with the Hebrew letters,
distinguishing final forms but not Sin from Shin. This
encoding is not used on this site, but is included as
the last column of the table for completeness, since you'll
undoubtedly eventually encounter documents which use it.
| Hebrew Letter | Glyph | Transliteration | Web Hebrew
|
|---|
| Alef | ) | ) | à
|
| Bet | B | B | á
|
| Gimel | G | G | â
|
| Dalet | D | D | ã
|
| He | H | H | ä
|
| Vav | W | W | å
|
| Zayin | Z | Z | æ
|
| Het | X | X | ç
|
| Tet | + | + | è
|
| Yod | Y | Y | é
|
| Final Kaf | k | k | ê
|
| Kaf | K | K | ë
|
| Lamed | L | L | ì
|
| Final Mem | m | m | í
|
| Mem | M | M | î
|
| Final Nun | n | n | ï
|
| Nun | N | N | ð
|
| Samekh | S | S | ñ
|
| Ayin | ( | ( | ò
|
| Final Pe | p | p | ó
|
| Pe | P | P | ô
|
| Final Tsadi | c | c | õ
|
| Tsadi | C | C | ö
|
| Qof | Q | Q | ÷
|
| Resh | R | R | ø
|
| Sin/Shin | # | # | ù
|
| Sin | & | & |
|
| Shin | $ | $ |
|
| Tav | T | T | ú
|
- I'm puzzled by the Hebrew chapter and verse numbers. Could you
refresh my memory how numbers work in Hebrew?
- Certainly. First, let me clarify that what we're talking
about is numerals as opposed to cardinal or
ordinal numbers. In English, for example, "2" is a
numeral, "two" is a cardinal number, and "second" is
an ordinal number. In Hebrew, letters are used as numerals,
thankfully in a much simpler fashion than Roman numerals.
Letters are assigned numerical values as follows:
| Units | Tens | Hundreds
|
|---|
| Name | Letter | Value
| Name | Letter | Value
| Name | Letter | Value
|
|---|
| Alef | ) | 1
| Yod | Y | 10
| Qof | Q | 100
|
| Bet | B | 2
| Kaf | K | 20
| Resh | R | 200
|
| Gimel | G | 3
| Lamed | L | 30
| Sin | # | 300
|
| Dalet | D | 4
| Mem | M | 40
| Tav | T | 400
|
| He | H | 5
| Nun | N | 50
|
| Vav | W | 6
| Samekh | S | 60
|
| Zayin | Z | 7
| Ayin | ( | 70
|
| Het | X | 8
| Pe | P | 80
|
| Tet | + | 9
| Tsadi | C | 90
|
Numbers are then written as a sequence of letters which,
when their numerical values are summed, produce the required
value. Letters in the sum appear in descending
order of magnitude (since Hebrew is written right to left,
the largest valued letter will be rightmost in a number).
Note that this is not a positional number system
and that there is no digit for zero. Hence, the number of
letters needed to express a given quantity bears no simple
relationship to its magnitude. One writes 100 as
Q, 101 as )Q,
110 as YQ, and 111 as
)YQ.
The numbers 15 and 16 would, if written as the sum of ten and 5 or
6, form fragments of the Divine Name. Consequently,
these numbers depart from the norm and are written as the
sum of nine plus 6 or 7. Thus fifteen is written
as W+ and sixteen as Z+.
- I've copied all the files from your site into a directory
on mine in order to establish a mirror. Why does the Hebrew
font work on your site but not on mine?
- The dynamic font mechanism which permits viewing
documents in Hebrew
was designed by
Bitstream, a
company which, among other products, designs and sells
a variety of fonts. As a font manufacturer, the
prospect of copies of their fonts, and those of
other vendors, freely flitting hither and yon on the Web was
disturbing--if one site used a font in one of its
documents, in principle any site could download a
copy of the font and use it (in violation of
copyright and license, perhaps, but it would still
be possible technically and difficult to enforce in
practice).
As a result, the dynamic font architecture implemented
by Bitstream does not make portable font files
directly available on the Web. Instead, fonts are compiled into
"portable font resources" or ".pfr" files, which
are locked to the domain name of the server which makes
them available. Hence, when you downloaded the Hebrew
Bible document tree, you obtained a copy of the file
containing the Hebrew fonts used by the various HTML
files, mcext.pfr. But browsers incorporating
Bitstream dynamic font technology will only accept
this file when it is downloaded from the domain for
which it was created, fourmilab.ch. Even though
the font upon which the file was based in in the public
domain, the tool which creates the .pfr file still
locks it to the site of origin.
As a result, your only option is to download a copy
of the original
TrueType® font,
then re-make a .pfr file for your own site with one of
the Web authoring tools which provide this capability:
an up-to-date list can be found on the
Bitstream site.
The mcext.pfr file was created with the
"Typograph 2.0" package from HexMac
Software Systems, which is available for Windows 95 and
NT, and the Macintosh, and can be used as a stand-alone
font compiler without forcing you into using some horridly
overcomplicated and flaky Web authoring environment.