Based on numerous sources, it seems clear that Proto-Indo-European was

  1. Productively agglutinative with non-root morphemes (and perhaps some specific roots that are also able to act like bound morphemes), and

  2. Derivationally agglutinative with roots in deriving new words.

What I haven't been able to find a clear answer on is whether PIE was productively agglutinative with roots, as is found in many modern Germanic languages (i.e. German). Is this still an unanswered question?

  • 2
    It's hard to see what kind of evidence could answer this question, as that kind of compounding verges on being syntactic and we can't reconstruct much syntax. I suppose if we could reconstruct some words that were then analysable into compounds of known roots, that would be evidence for the process, but I can't see what could be evidence against.
    – Colin Fine
    Jan 27, 2012 at 0:09
  • 2
    Hmm. Perhaps one possible indirect line of evidence would be how broad the genitive case is in Proto-Indo-European. The productive use of compounding or agglutination frees up potential functions of the genitive such that the genitive can have a very narrow range of meanings (e.g. it's almost synonymous with possession in English and German). Alternately, if a language lacks such mechanisms the genitive case is likely to be more broad, e.g. in Latin. Jan 31, 2012 at 20:34
  • 2
    Is the reason you mention German that you do not find English productively agglutinative with roots? Apart from the fact that English doesn't always write compound nouns as one orthographic word, I'm not convinced that English is less productive in this respect than German.
    – dainichi
    Aug 7, 2012 at 11:06
  • 2
    German has productive noun compounding but is not generally described as an agglutinating language. It doesn't really go for verb compounding though. Japanese and Chinese have both noun compounding and verb compounding for contrast... Apr 15, 2014 at 10:39
  • 1
    "agglutinative with roots" - usually that is not named "agglutination", because main term for that is "compound", and they are mostly agglutinative in all languages, "agglutination" is usually used only about grammatic morphemes like case ending morphemes and verb ending morphemes, and they are fusional in many languages, and fusional in many languages.
    – qdinar
    Jul 23, 2016 at 16:59

1 Answer 1


Yes, it was very much. The roots would be either just concatenated, or the thematic vowel -o- or the glue suffix -i- would be used.

Examples include:

  • a̯entio̯q̆os "face" (thing in front of eyes)
  • a̯rĝipta̯i̯os swift-winged
  • cleumntom good reputation
  • demspotis head of family
  • u̯icpotis settlement elder
  • dmpedos house floor
  • du̯eiplos double
  • dusdius bad weather (literally, bad sky)
  • ğeo̯uq̆olos cowherd
  • q̆tu̯rpedos tetrapod
  • sma̯ecsia̯ carriot (having one axis)
  • somoĝne̯i̯os blood relative

One also should note that the most of PIE personal names were compounds of two roots, similar to Greek Cleopatra, Alexander, Theophilos, Patrocleus or Slavic Yaroslav, Vladimir, Yaropolk.

  • Was this type of agglutination productive, or primarily in derivation? Jan 13, 2014 at 17:38
  • Both Greek and Sanskrit demonstrate such features, even in verbs, but is it true/full agglutination? As far as I know, only Lydian has true agglutinative qualities and that could be because it came in close contact with agglutinative languages e.g. Hattic.
    – Midas
    Feb 14, 2014 at 7:52
  • 6
    I would not call this agglutination. I would call it compounding: compounds of more than two elements are rare to non-existent.
    – Colin Fine
    Apr 13, 2014 at 15:55
  • 4
    @Midas: If you're referring to Annalic Hittite syntax (NU-pron-pron-pron NP NP NP ..), that's clearly an artificial style, not agglutination. Adn I agree with Colin that PIE compounds were clearly possible, and that this is not evidence of agglutination. PIE inflectional morphology is very clearly fusional, with multi-dimensional paradigms; agglutinative languages tend to have one-dimensional paradigms.
    – jlawler
    Apr 13, 2014 at 15:59
  • 1
    @jlawler: No, I was not referring to Hittite. I was referring strictly to Lydian.
    – Midas
    Apr 13, 2014 at 20:16

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.