8

Did PIE put the adjective behind the noun (like Romance languages usually do) or before the noun (like Germanic languages)?

Sir Cornflakes
  • 30,154
  • 3
  • 65
  • 128
The Thin Whistler
  • 212
  • 1
  • 6
  • 8
    I don’t think there’s any way of knowing that. Most of the oldest languages attested have fairly free word order, so if there was a preferred order in PIE, it’s probably obscured by subsequent loosening of word orders. – Janus Bahs Jacquet Jan 10 '21 at 16:47
  • It is fairly free in Russian. In Germanic you can also find it after (Vater unser...) – Adam Bittlingmayer Apr 13 '21 at 21:24

3 Answers3

7

In Hittite, which is an archaic IE language with the oldest record, the order seems to be fairly free: both Adj+Noun or Noun+Adj are attested.

Miztli
  • 1,085
  • 1
  • 10
  • 21
5

Here are some reconstructed phrases in PIE. It seems, the adjective could go both before and after the noun. Examples:

  1. Adjective before

h₁ōḱéwes h₁éḱwoes "swift horses"

dus menes "bad mind" (> "bandit, enemy")

dus dius "bad sky"

  1. Adjective after

ḱléwos wéru "wide fame"

ḱléwos meǵh₂ "big fame"

ḱléwos ń̥dʰgʷʰitom "imperishable fame"

So it seems,

  1. The adjective dus "bad" was always put before the noun, but it can be interpreted as the prefix of the noun rather than a separate word. The same with the adjective h₁su "good". Again, it may be seen as a prefix.

  2. There were other prefixes that are similar to adjectives, adverbs or prepositions, such as sm(i)-, twi-, tri- (one, two, three), those pointing location etc, such as pri-, peri-, h₂nti-, h₂ntbhi-, proti- etc.

  3. Possessive participles were put before the noun, as in

sh₂uens kʷekʷlos "Sun's wheel"

dems potis "house master"

Other adjectives, possibly could go the both ways, or depending on context.

Anixx
  • 6,643
  • 1
  • 26
  • 38
  • 5
    **dus-* and ** h₁su-* were unambiguously prefixes, not adjectives: they don't have endings and they don't inflect. Outside of Greek's nounified ἐΰς (with ending), their reflexes also only occur as prefixes in the daughter languages. – Cairnarvon Apr 12 '21 at 23:46
  • I think you meant possessive nouns, not participles. – TKR Apr 14 '21 at 00:08
5

PIE had a rich inflection system, as is echoed in the oldest attested daughter languages. Owing to this, if adjective and noun were each appropriately declined, the order could be either way.

As to the actual order, there is not enough evidence to support an absolute trend either way in PIE.

Remember that word order is more important in modern Germanic and Romance languages due to a simplified inflection system. Some other Indo-European languages, such as Marathi, are still relatively free in their word order in general.

007
  • 164
  • 1
  • 1
  • I don't think I agree that the complexity of the inflection system has much to do with it. Some French adjectives can go on either side of the noun with no increase in confusion or sentence structure ambiguity, and highly inflected (and conservative!) Ancient Greek is very rigid about where adjectives can go. – Cairnarvon Apr 13 '21 at 20:31
  • @Cairnarvon Ancient Greek is very free about whether adjectives precede or follow their noun. See e.g. http://atticgreek.org/downloads/WordOrder.pdf – TKR Apr 13 '21 at 23:56
  • @TKR There's a consistent semantic difference between adjectives in the attributive position vs. adjectives in the predicative position, and I'd even say that that and the behaviour of the definite article w.r.t. the attributive position show that it's a distinct grammatical construction altogether, not just a matter of the adjective going before or after the noun. This isn't freedom the way e.g. Latin word order is free. – Cairnarvon Apr 14 '21 at 00:49
  • @Cairnarvon I don't mean attributive vs. predicative position (since the question isn't about predicative constructions), but attributive adjectives preceding or following the noun -- ἀγαθὸς ἀνήρ / ἀνὴρ ἀγαθός, ὁ ἀγαθὸς ἀνήρ / ὁ ἀνὴρ ὁ ἀγαθός are semantically equivalent and all common. – TKR Apr 14 '21 at 02:12
  • @TKR They're not semantically equivalent in Classical Greek, and ἀνὴρ ἀγαθός and ὁ ἀνὴρ ὁ ἀγαθός both have the adjective in a predicative position. Only (ὁ) ἀγαθὸς ἀνήρ has the article in an attributive position. Your own PDF actually says as much. – Cairnarvon Apr 14 '21 at 02:39
  • @Cairnarvon I'm not sure if we're talking past each other in some way. Predicative position would be ὁ ἀνὴρ ἀγαθός -- see e.g. Smyth 1168. All the phrases I gave are attributive and mean "a/the good man". (The indefinite ones could additionally be predicative, "a man is good", but this is less common.) – TKR Apr 14 '21 at 02:50
  • @Cairnarvon The position of French adjectives is still limited. Usually those which go on either side of the noun have different nuances depending on position. And the majority have a fixed position, either before or after. – 007 Apr 14 '21 at 03:42