This page is about the transliteration, transcription or Romanization of Hebrew -- the rendering of Hebrew words in the Latin alphabet.
[Note: In some circles, "transliteration" means to render a language's native orthography in some other symbols. Here, that would mean to use Latin letters in a one-to-one correspondence with the original Hebrew spellings, such that you could perfectly convert back to the original if needed. On this page, we generally take transliteration to mean a Romanized phonetic transcription.]
All transliteration schemes involve compromises. There are six issues to consider and one law.
If your goal is a phonetic transliteration, you first have to decide how you think the Hebrew is pronounced. Ashkenazic, Sephardic, Galitzianer. Israeli sephardic vs. Yemenite sephardic. If Yemenite, which dialect? Ancient Hebrew or Modern Hebrew? [See the "Pronunciation" article in the Jewish Encyclopedia. The onlne version appears to be down 1-3-06.)
Do you think the ayin is pronounced or silent? Should the kof be distinguished from the kaf? Are the samech, sin, and sav three different sounds? Two? Or one? Is the sixth letter of the Hebrew alphabet pronounced /v/ or /w/?
Israelis sound their segol much like the tseirei, but American speakers of Israeli-style Hebrew distinguish them. Which will you use?
And goodness only knows how we plan to pronounce the Aramaic, the Old French loan words, and the proper names from neighboring cultures.
Romanization would be easier if only we could agree on how to spell things in English. Do we render a long A as <ei> (freight) or <ay> (hay)?
You also have to watch out for ambiguities in your transliteration. You might represent a shin with a segol as <she>, but would the reader pronounce it as /shee/? Maybe <sheh> would make it clearer.
While English pronunciation has a variety of national and regional dialects, most transliterators presume that they pronounce English in a standard way. Even if that were true, many of the readers who use their transliteration system don't. The alt.usage.english FAQ says:
Beware of using ad hoc methods to indicate pronunciation. The problem with ad hoc methods is that they often wrongly assume your dialect to have certain features in common with the readers' dialect. You may pronounce "bother" to rhyme with "father"; some of the readers here don't. You may pronounce "cot" and "caught" alike; some of the readers here don't. You may pronounce "caught" and "court" alike; some of the readers here don't.
Of course, we wouldn't have this problem if everyone pronounced English the way I do.
One could avoid the difficulties of English pronunciation entirely, by using the International Phonetic Alphabet (IPA), which has a distinct symbol for virtually every sound possible in human speech -- but normal people wouldn't know how to pronounce any of the symbols.
Such as the voiceless velar fricative. An <x> would be great because Greek uses the X for the same sound and there's no other Hebrew sound that corresponds to the Latin X. But only the IPA uses an <x>. Others use a <ch> (Chanukah), or <h> (Hanukah). The Conservative Movement uses an academic convention that distinguishes the chet (<h> with a dot under it), from the chaf (<kh>).
On the other hand, English does have the schwa (sheva na) sound -- it's the <o> in lemon -- but no standard way of writing it. The dictionary uses an inverted e to represent it. In Hebrew transliteration, I've seen an apostrophe, a lower case <e> or a superscripted <e> used.
Do you want your transcription to preserve the distinction between kof and kaf, tav and tet, etc? Do you want to indicate a dagesh forte (dagesh chazak)? It's important if your readers will be discussing the triconsonantal roots of the Hebrew words or other matters of grammar.
Scholars need perfect correspondence in their transliterations, so the system found in the SBL Handbook of Style (published by the Society of Biblical Literature) is generally used. Of course, non-scholars can't read it at all.
Let's say the system you've settled on doesn't indicate the dagesh forte by doubling. You lay out the cover of your new book and it says "Shabat for Today." You show it to your publisher . . . or client . . . or wife. And they say "It's spelled wrong!" Now you have to decide how to handle "common" spellings such as Shabbat, Maariv, Kabbalah, Leah.
One way out is to use the familiar spellings when in an "English" context such as a book title, headings or English text; and use the transliterated form when rendering a Hebrew passage phonetically. This can get funky if you write a passage about the matriarch Leah and the reader tries to find the word in your transliteration where you've rendered it as <Lei-ah>.
The IPA and the SBL systems (for phonology and orthography wonks, respectively) depend on specialized characters. These are difficult for the layperson to learn, provoke much unhappiness for your typographer, and are problematic for websites and email.
On the other hand, the Latin alphabet -- here let's extend that to mean the ASCII character set -- can be easily typeset, browsed on the web or emailed.
Those who want total correspondence to the Hebrew spelling in their ASCII email, can use the Michigan-Claremont-Westminster (MCW) scheme. However, it takes time to learn and doesn't help with pronunciation.
Those who want total clarity of pronunciation in email could use X-SAMPA, a 7-bit ASCII representation of the IPA character set.
But whatever compromises you make, you'll sacrifice ease-of-use or precision . . . or both.
The One Immutable Law of Transliteration: No matter what system you adopt, someone will come up to you and say, "I showed your transliteration to six different people and not one of them could pronounce it correctly. Here's a much better system that I've developed."