Baltic and Slavic infinitive suffix -ti

Discussion in 'Etymology, History of languages, and Linguistics (EHL)' started by Gavril, Apr 6, 2013.

  1. Gavril Senior Member

    English, USA

    What is thought to be the origin of the infinitive suffix -ti seen in Baltic and Slavic languages (e.g., the infinitive of the verb "to be" is Lithuanian ti, Russian byt', Slovenian biti, Polish być, etc.)?

  2. Triginta Septem Member

    Las Vegas, Nevada, USA
    English - America
    It comes from *-tei, which is originally a case suffix, probably locative. So, verb-ti basically means "in doing verb".

    Interestingly, the gerund (also used as something like a locative, sometimes) is formed in Japanese by adding -te/-de. This is obviously unrelated, and I'd assume it's from the postposition de, meaning by means of, but it's an interesting coincidence. ^^
    Last edited: Apr 6, 2013
  3. Gavril Senior Member

    English, USA
    The suffix -i is considered to be locative, but the -te- would have to come from somewhere else, correct? Is it thought to be from a verbal noun formation, like the Latin supine suffix -tu-?

    Also, is it expected that *-ei would reduce to -i in this position? I recall that *ei was generally monophthongized to *e: in Baltic and Slavic, but maybe the development was different in word-final (unstressed) position.

    It seems quite common (cross-linguistically) for infinitives to be based on frozen locative case forms. I'm thinking of starting a thread on this when I've gathered a little more data.
    Last edited: Apr 7, 2013
  4. Triginta Septem Member

    Las Vegas, Nevada, USA
    English - America
    I don't know much about this, so I can't tell you why, but it would appear to be an exception, if that's the case... (

    Actually, is what he's talking about there what you meant about -tu-?
  5. Gavril Senior Member

    English, USA
    Yes, in the section you linked to, he derives the *-tei ending from a verbal abstract noun in *-ti-, perhaps related to Greek -sis (as seen in words like thesis, praxis, parataxis, etc.), if -si- reflects earlier *-ti-.

    The author references an earlier section of the book when he discusses this verbal abstract noun formation, but this section isn't currently viewable on Google.
    Last edited: Apr 7, 2013
  6. LilianaB Banned

    US New York
    Hi, Gavril. What do you mean by "come from"? Wouldn't an infinitive be the basic form of any verb. What were the infinitive endings in PIE?

    Added: I have actually found out that it is unclear what the infinitive endings in the PIE were (possibly -ti as well, but not confirmed). Infinitives, as verbal nouns, were formed through the use of suffixes in the PIE.

    The third person singular ending was reconstructed as *ti, in the PIE.
    Last edited: Apr 7, 2013
  7. Gavril Senior Member

    English, USA
    Triginta Septem wrote that "[-ti] comes from *-tei, which is originally a case suffix, probably locative." Since the *-t- can't have been part of the locative case suffix, I wondered what its origin was.

    It's not clear that PIE had an infinitive in the sense of a default, undeclined nominal form for each verb. Not every IE branch has one: e.g., the Celtic languages have various verbal noun formations, but no regular infinitives.

    As far as the verbal noun suffixes reconstructed for PIE, I recall reading that the suffix -tu- (which appears in the Latin supines: e.g., conspectus "view" < spec- "look") is found in a number of different IE branches.
    Last edited: Apr 7, 2013
  8. DenisBiH

    DenisBiH Senior Member

    This is what Frederik Kortlandt has to say on the matter:

    I have no idea what the dot under the e is supposed to indicate, perhaps he explains it somewhere else in the article. Also, it is not clear to me when and how ī was shortened, but he says this:
    Last edited: Apr 8, 2013
  9. LilianaB Banned

    US New York
    Since the infinitive is treated as a verbal noun, wouldn't it be more feasible that it would take an Accusative ending, rather than a Locative? I would like an apple (Acc. -- in Slavic and Baltic languages) I would like to eat (Infinitive).
  10. Gavril Senior Member

    English, USA
    I don't think that the accusative singular ending of a noun in *-ti- (which is the proposed origin of the Baltic and Slavic infinitives) would be expected to give *-ti in Baltic or Slavic. In Baltic, if I'm not mistaken, the accusative singular contains at least a nasalized vowel rather than a plain -i: cf. Lithuanian širdį, the accusative sg. of širdis "heart". In Slavic languages, as far as I know, the acc. sg. of older i-stems is endingless. For example, Slovenian pot "path, journey" is thought to be from an earlier i-stem, and its accusative singular is identical to the nom. sg., whereas its locative sg. is poti.

    More generally, I don't think that a verbal noun is necessarily likelier to be "frozen" in the accusative than in the locative: cf. English I'm a-doing [X] < I'm at doing [X]. (I'm hoping to discuss this question in more depth on a thread about the development of infinitives.)
    Last edited: Apr 8, 2013
  11. Ben Jamin Senior Member

    There is no reason to believe that the infinitive came before the finite forms of the verb. One must not confuse the infinitive with the primitive, pre-finite form of a verb. The infinitive, at least in IE languages is truly a newer construction than the finite forms.
  12. muhahaa Member

    I remember reading that lacking infinitives is typical to languages with active-stative alignment (just like the gender system with no active case in neuter gender).
  13. Gavril Senior Member

    English, USA
    Hi Muhahaa,

    What do you mean by active/stative alignment?

    Most of the older (attested) IE languages have a neuter gender category in which the accusative is formally identical to the nominative, but only a few branches lack what we would think of as an infinitive.
  14. LilianaB Banned

    US New York
    Yes,, the Older Baltic languages had an active neuter gender. The present time Lithuanian does not have it (in the active -- only partial passive use). Most Slavic languages have a neuter gender.
  15. Dhira Simha Senior Member

    This is a very interesting question. It is generally agreed that I-E infinitives are "derived NOUNS"
    Robert J. Jeffers. "Remarks on Indo-European Infinitives".Language , Vol. 51, No. 1 (Mar., 1975), pp. 133-148
    Published by: Linguistic Society of America
    Article Stable URL:

    They are believed to be 'fossilised' case forms (accusative, dative, locative etc.) of verbal nouns structured as root + nominal suffix + case ending i.e. formed from verbal roots or bases by means of suffixes. You may want to read Jeffers pp. 142-144 where he treats Slavonic and Baltic infinitives. This does not answer the question of what was the origin of the suffix -ti- which is for me the most interesting part. It was quite productive in Vedic e.g.: iṣ- ‘to endeavour to obtain, strive, seek for’ > iṣ-ti ‘desire’; dā- ‘to give’ > dā-ti ‘gift’. Slavonic and Baltic infinitives in -ti look strikingly similar to the bare (non-inflected) Vedic noun base in -ti. It is also interesting that the other productive Vedic suffix -tu also used for making infinitives (e.g. in -tum) coincides with the old Slavonic supine ending -tŭ (-тъ) which was very close in function to the infinitive.
    Last edited: Apr 20, 2013
  16. neonrider Member

    Vilnius (Lazdynai)
    Lithuanian / Lietúvüü
    I found other "coincidental similarities" between Japanese and Northern European languages, Lithuanian among them. To me, Japanese, if we remove Chinese loan words from it, is more like Northern European or Pacific language rather than Asian. Here's an example: "oshiri kajiri mushi" (butt biting bugs), in Lithuanian it will be "subine kandancios muses". Where muse is rather a fly, but it is also a bug. Now the root kand- is the same in Japanese and in Lithuanian, so is mus-.

    There are other coincidences, I will mention only a few that I remember (JP - LT - EN):

    Kawa = kavoja, slepia (hiding)
    Oppai = papai (tits, breasts)
    Kande = kanda (bite, bites)
    Mushi = muse (bug, fly)
    Oshiri = uzhpakalis (butt, behind) (Lit. often pron, as "oshpakalis")
    Kochi kochi = kuti kuti (tickling, tickle-tickle)
    Katai = kietai, kieta, kietas (hard)
    Konki = kantri (patience) (not very similar)
    Tairaka = taikinga, taika (peaceful, peace)
    Tooi = toli (far) (very similar because our "o" is pronounced long as Japanese "oo")
    Wakai = vaikai (kids, young) (wakaa - in West Lithuanian dialect)
    Ikimas(u) = ejimas (going)
    Ikimasho = eikime (let's go)
    Ate = ate (end, bye)
    Aishiteru = aistra (Jp: I love you; Lt: passion)
    Te = te (JP: hand; LT: take it (command))
    Matane = iki pasimatymo (adios)

    Some entertaining ones:

    arimas (JP, grammatical form) = arimas (LT, a ploughed field)
    simas (JP, grammatical form) = Simas (LT, male name)
    wakarimas (JP, I understand) = vakarejimas (LT, evening is approaching; its becoming dark(er)) or dusk

    This could be the lost Ainu connection.
  17. neonrider Member

    Vilnius (Lazdynai)
    Lithuanian / Lietúvüü
    I think -ti, -tj is similar to English to: to do = daryti, delatj (Lit., Rus.).

    For example:

    Lithuanian: grandyti
    English: to grind

    Yet English also has it at the end of words sometrimes, for instance:

    Lithuanian: persekioti (very often Lithuanian -ti is written, spoken and spelled as persekiot, without "i" at the end)
    English: to persecute

    Lithuanian: laukt(i)
    English: to wait
  18. Gavril Senior Member

    English, USA
    There might be something to the comparison involving English wait: the English word is thought to be related to Dutch wachten "wait", which is in turn thought to come from the same stem as Eng. wake, awake etc. It's possible that the -t- suffix used to derive wacht- from earlier *wak- is the same as the -tu- of the Latin supines, the -si- of Greek verbal nouns, and the -ti of Balto-Slavic infinitives. However, there are other possibilities: the -t- of wait could be participial (from earlier *-to- rather than *-tu-/-ti-), for example.
  19. itreius Senior Member


    Why is coincidental similarities in quotes?

    Could you clarify what you mean by "is more like", Northern European and Asian?
  20. ahvalj Senior Member

    So many fruitful ideas, why not just to check the relevant literature where everything was discussed to death back in the 19th century?

    Both infinitive forms are related and continue the former deverbal nouns on -t-, but the Lithuanian Infinitive continues the original Locative of t-stems (the IE Locative ending was -i or zero), while the Slavic form continues the Locative of ti-stems (since it has an acute "i" that may come only from the i-stem Locative -e:i). If Lithuanian had the same form, it would have sounded -ti in plain and -ties in Reflexive verbs and would have been stressed in verbs with non-acute roots (e. g. *vestì/vestíes like Praes. Sg. 1 vedù/vedúosi instead of the actual Infinitive vesti/vestis with the stem-stress).

    The situation in Latvian is ambiguous: this language actually has vest/vesties, but since -ie- in Latvian is a universal fill vowel between the consonant and the reflexive marker -s (e. g. -ties in Imperative Pl. 2), it is unknown, which stem, -t- or -ti- was original there.

    By the way, Prussian had different Infinitive forms (e. g., -twei), so the actual Infinitive was a result of choice between various deverbal nouns made in individual Balto-Slavic branches.
    Last edited: Sep 29, 2013
  21. berndf Moderator

    German (Germany)
    The most widely accepted theory about the development of the PIE gender system is that it developed out of a active-stative language that had "animate" nouns that could appear as agent and patient in a sentence and where marked for nominative and accusative and "inanimate" nouns than such a marking because they could only appear as parameters of stative verbs. These "inanimate" nouns had two distinct plurals, an individual and a collective one. The hypothesis is that the animate gender developed into the masculine, the inanimate into the neuter and the collective plural of the inanimate into the feminine gender of PIE. The explains the lack of distinct accusative forms for neuter nouns in many IE languages.
  22. Ben Jamin Senior Member

    Polish shows remnants of animate/inanimate nouns. The masculine gender has three subgenders: personal animate (human), non personal animate (animal) and inanimate.
  23. berndf Moderator

    German (Germany)
    That is probably a re-invention rather than a remnant of this old gender system. Hittite, a language that probably split from PIE very early, had this two gender system (animate vs. inanimate).
  24. Ben Jamin Senior Member

    Yes of course, I agree, it is not a feature inherited directly from PIE, but the idea of animate vs inanimate must have been somehow preserved in the language.

Share This Page