16

I read that PIE, Latin, old English, and even old German did not use articles, yet current English, German and Romance languages all use articles.

Is it true that articles developed in all these languages independently from the word for "one"? What is the reason for their development? Were they useful in any way, such as helping pronunciation?

Martin Konicek
  • 263
  • 2
  • 7
  • 3
    I believe there were two Proto-Indo-European roots meaning "one". The root sem-* led to words like Greek heis "one", and words like similar and simple through Latin. The root(s) oi-no-* led to one through Proto-Germanic, and to the Romance articles and unique through Latin unus. Neither Latin nor Greek had an indefinite article; they must have developed later, and at least partly independently (cross-pollination is always a possibility). – Cerberus Apr 21 '13 at 21:56
  • 4
    Definite articles came from various demonstrative stems, which only makes sense, if you consider that definite articles are really just a special kind of demonstrative pronouns. Demonstrative pronouns being linked to deixis, it seems probably that deictic use is the ultimate source of definite articles. Proto-Indo-European to-* and so-* were demonstrative/deictic stems; *so- led to the Greek definite article ho, but also through Proto-Germanic to English he/her/him/here, and through Proto-Italic to Latin hic "this, he" etc. – Cerberus Apr 21 '13 at 22:01
  • 2
    Some comparative PIE numerals here. European IE languages with indefinite articles usually proceed from one; definite articles in Romance come from Latin ille, and in German come from various forms of PIE *to (which weakens to an initial TH by Grimm's Law). – jlawler Apr 21 '13 at 22:12
  • ... And in many languages, definite articles used both stems. Most forms of the Greek article were based on to-, like to, tou, tês* &c., besides ho, hê &c. Modern Germanic definite articles seem to have evolved mostly from to-, like English the, German der, and Dutch de, but some forms are from so-, like Dutch het, "the, it" (English it is no doubt also from so-). This -(e/i)t* is probably just a neuter ending. The Romance definite articles are from the Latin demonstrative pronoun ille, "that, him" (etymology uncertain, probably deictic particle -i- + -le-*). – Cerberus Apr 21 '13 at 22:20
  • 1
    @jlawler: Nice comparison; the only thing is that heis doesn't appear to be a reflex of oino-*, but rather of sem-*. – Cerberus Apr 21 '13 at 22:32
  • 1
    Bulgarian/Macedonian, alone among Slavonic languages, have acquired a definite article (again from *to-) and, like their very distantly related neighbours Romanian and Albanian, their article is suffixed to the word. – Colin Fine Apr 29 '13 at 17:38

3 Answers3

11

Indefinite articles developed from numerals, and the definite articles developed from demonstratives. This is a very well known process called grammaticalization.

MGN
  • 1,033
  • 7
  • 16
  • Grammaticalization indeed seems to be the answer. Is it possible that articles appear to take over the grammatical function of inflections as they are disappearing? As for the reason why inflections tend to disappear over time, is it because they are quite complex and people to tend to simplify, especially when learning a language as a second language? – Martin Konicek May 09 '13 at 12:35
  • 1
  • Yes, inflections can disappear but the grammatical function remains in the articles. This is what's happening right now in German, where you almost only have declension on the articles and adjectives, and only few nouns have declensions besides genitive and plural dative. 2) I don't think so, it is also a product of phonetic erosion. Words tend to lose their ending, and if the grammatical functions is already expressed by the article, then there is no need to also make it explicit on the noun, so it starts to "decay".
    • – MGN May 12 '13 at 17:24
  • I'll accept this answer. Grammaticalization makes a lot of sense to me, although it is only a theory because there is too little historical data. This short article explains the concept very well: http://www-personal.umich.edu/~jlawler/TheGrammaticalizationCycle.pdf Thanks! – Martin Konicek May 14 '13 at 09:30
  • @MGN - similar thing happens in popular Portuguese regarding number: As casa é azul with plural inflexion only in the article, instead of standard As casas são azuis with four different inflexions marked for plural. – Luís Henrique Oct 06 '17 at 13:37