The current answers on Definite/indefinite articles vs. inflections agree that (definite) articles are acquired by languages, not lost.

I'm wondering what Eastern Aramaic has to say about this. Semitic nouns can be in the absolute (default) or the construct state (for genitive constructions). Aramaic develops an emphatic state to mark definiteness. However, in Eastern Aramaic this emphatic state becomes the default, so the definite article effectively gets lost. This is, at least, how I interpret Gzella (A Cultural History of Aramaic, 2015, p. 28):

The “absolute” state (or “unbound form”) acts as the unmarked form; when the emphatic state emerged as a postpositive definite article, the absolute state came to signal indefiniteness. It is generally used with the quantifier /koll/ ‘all’, adverbial and numerical constructions, and predicative adjectives. Morphological definiteness marking spread gradually during the opening centuries of the first millennium b.c.e., following a common tendency in Northwest Semitic, but the “emphatic state” lost this function in later Eastern Aramaic (for instance in Syriac) again and became the unmarked form, whereas new definiteness markers then evolved from demonstratives in Neo-Aramaic languages.

Is "(definite) articles are acquired, not lost" a universal, or should I read the linked question in the context of Indo-European only?

  • The formulation in the linked question is: "Do proto-languages (proto-Indo-European for my examples) have articles or do they vary as well in this respect?" Apparently the articles in Indo-European languages are an acquired feature, but whether it is a general rule for all language families remains a part of the question.
    – Roger V.
    Mar 5, 2020 at 8:21
  • 1
    @curiousdannii Some languages acquire articles, but are there any that lose them? And more specifically: are there any language families where the proto-language has articles, but (some of) its descendants do not?
    – Roger V.
    Mar 5, 2020 at 8:34
  • 1
    @Vadim I've found references to Oceanic, Berber, some branches of Romani having lost articles. Greenberg proposed the "cycle of the definite article" where demonstratives turn into articles which turn into gender or class markers.
    – curiousdannii
    Mar 5, 2020 at 9:10
  • 1
    @curiousdannii re. your first comment: this universal would leave open the possibility that some languages have not yet acquired them - the key falsifiable claim is that languages do not lose articles. Re. your second comment, if you could give those references I think that would make a great answer!
    – Keelan
    Mar 5, 2020 at 10:28
  • 1
    Aramaic has had areal overlap with Iranic languages. Literary Persian and informal Persian differ on definiteness, as do Kurmanji and Sorani. Perhaps a place to dig. Mar 5, 2020 at 14:03

2 Answers 2


Yes, Aramaic through the ages has had a more-or-less complete cycle attested, thanks to its long documented history.

Yaudic Aramaic as attested on the inscriptions at Zencirli appears not to have any articles.

Imperial Aramaic, which has a large corpus, has a suffix א () as its definite article, which has been linked to the prefixed ה (h-) in Biblical Hebrew. This is when the three "states" of Aramaic, construct state (also pronominal state), absolute state and emphatic state arise. The emphatic state corresponds to the one with the definite article suffix.

The split into Western and Eastern Aramaic dates from the 4th to 1st centuries BCE, with the Western Aramaic dialects showing the continuation of the definite vs indefinite distinction, including emphatic state with its definite article suffix. There is some debate about how it all works in the Targums.

On the other hand, the Eastern dialects appear to have weakened the distinction, to the point where it was no longer a functional feature of the language: in Classical Syriac, one of the most widely attested Eastern dialects, the emphatic state with the a suffix is simply the unmarked state, the construct state is maintained, and the absolute state is generally indefinite, where it is encountered.

There is also a use of the third person pronoun ܗܘ hw as a definite article, although this seems to be in imitation of translation from the Greek.

Ṭuroyo is one of the Eastern Neo-Aramaic varieties that has developed a definite article prefix, from demonstrative pronouns.


It’s not a universal. In fact, it’s most probably a cycle, as with other constructions undergoing grammaticalisation. In many Slavic languages, the definite or indefinite forms became the only ones used so the explicit distinction got lost. In Abkhaz, the definite article is now used by default even when the noun is indefinite and I bet one can find many more examples in other languages. In the question you refer to Indo-European was explicitly mentioned where the pathway was “acquiring”.

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.