Given the great linguistic similarities between Hebrew and Arabic, it is noteworthy that their definite articles are completely different.

Wikipedia's Arabic definite article talks about the theory that the Hebrew definite article ha- (ה־) and the Arabic definite article al- (ال־) both evolved from a common proto-Semitic particle. It says:

There are three major possibilities regarding the form of the proto-Semitic particle that is the putative antecedent of al-:

  • hal;
  • ha;
  • ‘a;

David Testen and Jacob Weingreen state that [hal] هل۔/הל־ hal is the correct antecedent.

I'm wondering, why put the "ha" and the "al" in that order, if we reverse it we get very close to the word for "a god" in both languages: Hebrew אֱלוֹהַּ (elóah) or Arabic إِلٰه (ʾilāh). Vowels being relatively fluid...

What could be more definite than a god? Why not suppose that the Hebrew and Arabic definite articles both evolved from the concept of deity, with Hebrew dropping the first syllable and Arabic dropping the second? Is there any less support for this hypothesis than, say, for the "hal" hypothesis?

  • 4
    I think GOD idea that GOD word "god" somehow transformed to one of GOD most frequently used words in GOD language is rather questionable - especially since GOD word in question doesn't even have a concrete meaning.
    – jick
    Commented Feb 21, 2019 at 5:15
  • 2
    I don't think the Hebraic ha is related to the Arabic al, because the morpheme ha exists in Arabic language. en.m.wiktionary.org/wiki/ه
    – amegnunsen
    Commented Feb 21, 2019 at 9:17
  • 1
    @amegnunsen All I see on that page with ha is the name of a letter
    – b a
    Commented Feb 21, 2019 at 10:48
  • 8
    If you have to reverse a word, then you're almost always looking at a coincidence.
    – curiousdannii
    Commented Feb 21, 2019 at 12:08
  • 1
    For more information on the definite article in Arabic and Hebrew, look at this question and its answers: linguistics.stackexchange.com/questions/23162/… Commented Feb 21, 2019 at 12:13

3 Answers 3


The present answers are in principle correct, but do not explain the fundamental issues with this idea. In short:

  • The "God" lexeme is relatively infrequent to develop into a definite article.
  • Comparative evidence from Ugaritic and Aramaic suggest that the developments must have taken place largely independently, which makes such an unlikely derivation even more unlikely.
  • An origin in deictic markers (common cross-linguistically) can also account for:

    • The Arabic indefinite marker -n;
    • The Aramaic definite marker (the "emphatic state").

As is mentioned, it is cross-linguistically common for definite articles to develop from deictic markers, which are eventually reanalysed as markers of definiteness. The idea is that the (noun+deictic marker) combination is common; at some point sound changes occur which obscures the distinct building blocks; later the combination is reanalysed as (noun+definite marker+deictic marker); this allows the new definite marker to spread to other configurations. (Of course, the relative positions of these markers can differ.)

This process is much more likely to occur with frequent lexemes. Although "God" is common in the Ancient Hebrew and Arabic texts we have (though surely still not as frequent as deictic markers), this is a feature of the types of texts that were considered worth copying; it is unreasonable to assume that "God" really occurred that frequently in common spoken language.

If that is still not enough evidence, here is the main issue as far as I'm concerned: Arabic and Hebrew split relatively early (after Central Semitic). After Arabic split off, the Hebrew branch (Northwest Semitic) developed several other languages: Ugaritic, Aramaic, and Canaanite (which contains Hebrew, Phoenician, and some other small-corpus languages). Since Aramaic has suffixed , for which it would be rather far-fetched to assume an origin in 'ilāh, and Ugaritic has no definite article, a large part of the development of the Hebrew and Arabic articles must have been independent. This is reasonable when the origin is common cross-linguistically, but an already unlikely development from "God" happening twice makes this even less likely. We would also expect the Aramaic article to be closer to the Hebrew one, since it split off later.

The common derivations of the definite article in Hebrew and Arabic do account for the Aramaic one as well. A good source is Thomas Lambdin, 'The Junctural Origin of the West Semitic Definite Article' (Near Eastern Studies in Honor of William Foxwell Albright, 1971, pp. 315–333). He posits the following stages for Hebrew:

  1. malku 'a king'; malku ð-(baḥarū) 'the king whom they chose'.
  2. Close juncture between the noun and postposed words: malku + ð > malkuðð.
  3. Loss of final short vowels before open juncture leads to malk 'a king', but malkuðð remains.
  4. h syncope allows this to be analysed as malk 'a king'; malakīm 'kings'; malk hazzi 'this king'; malakīm ha´´illay 'these kings'. In the latter two we recognize ha- prefixed to the deictic markers zi and ´illay.
  5. This spreads to adjectives, ordinal numbers, etc., and eventually to nouns as well.

In Arabic, the development is similar. Arabic retained final m/n (i.e. malikun instead of malku in stage 1). This final nasal assimilates to the first consonant of the deictic marker in the close juncture. This explains why definite words in Arabic have no final m/n (i.e., why -n is the indefinite article). The definite article becomes analysed as al- because n becomes l before some consonants.

Aramaic differs from Hebrew in stage 4. Lambdin posits that its speakers shortened "this-x and this-y" to "this-x and -y", i.e. malkaððin(āh) wagabraððin(āh) becomes malkā´ wagabraððin(āh). This prevents the reanalysis in Hebrew stage 4 and leads to the postfixed definite article (the "emphatic state").

This leaves Ugaritic to be explained: why no definite article here? The timespan of the Ugaritic corpus is relatively short, maybe simply not long enough to develop a definite article. This may also have stylistic reasons: the definite article is very infrequent in Hebrew poetry which may indicate it was considered not good style for in literature. But also see Na'ama Pat-El, 'The Development of the Semitic Definite Article: A Syntactic Approach' (2009: Journal of Semitic Studies 54:1, 19–50), who argues that Ugaritic displays a premature state of definiteness marking.

  • 1
    Lambdin's article of 1971 was not very plausible in the first place and is now completely outdated. A new investigation of the history of the definite article would now have to consider the evidence of Ancient South Arabian and Ancient North Arabian, both badly known in 1971.
    – fdb
    Commented Feb 21, 2019 at 20:51
  • Thanks that was very interesting. (1) I just asked a separate question about deictic markers evolving into definite articles. (2) I couldn't figure out how to download the article you referenced; maybe that's just a trip to the library. Commented Feb 23, 2019 at 7:12

A few things.

  1. Definite articles are usually drawn from deictics or anaphorics. It's the case in Romance: Latin ille, illa > French le, la, in English this, that > the, etc.

  2. the Semitic article is derived from *hal, whereas "god" is *'il. These two words have nothing in common, except *l

I'm afraid your question is motivated by some esoteric or kabbalistic nonsense. In all cases, from the linguistic, typological, etymological, semantic and phonetic points of view, your question makes about no sense at all.

  • 5
    The reconstruction of the proto-Semitic article (if PS even had an article) is debated. Life is not so simple.
    – fdb
    Commented Feb 21, 2019 at 15:50
  • Can you edit your question to support "Definite articles are usually drawn from deictics or anaphorics" (and not e.g. the other way around)? Or I could ask another question here for that, either way. And what do you mean "'god' is *'il"? And where do Kabbalists ask my question? All interesting points but would like more info. Commented Feb 22, 2019 at 7:14
  • I mean that the Proto-Semitic word for "god" is *'il(u), while the definite article is derived from Proto-Semitic *hal. As I wrote before, these two words only have l in common.
    – user23769
    Commented Feb 24, 2019 at 5:48

The Arabic divine name allāhu does indeed contain the article al-, it being (according to the generally held view) a contraction of al + ʼilāhu. The Biblical divine names Yahweh and Elohim do not contain an article, though (Orthodox) Jews of the present time do refer to their god usually as hash-shem “the name”, which does contain the Hebrew article, of course. But in both cases the divine name is formed from an article plus a common noun. The article is not abstracted from the name.

(Sorry about the non-scientific transcription of the Hebrew. There is limit to the number of diacritics one can type in any given day.)

  • Thanks ... but I was asking about lower-case "god" and not "God". Implicit was the assumption that the word "a god" came first, i.e. "ilah" came before "Allah = al-ilah". I know that "al" does not come from "Allah", that would indeed not make sense. Commented Feb 22, 2019 at 7:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.