Feminine "-a" suffix in Semitic and Romance Languages

Discussion in 'Etymology, History of languages and Linguistics (EHL)' started by tFighterPilot, Jun 24, 2012.

  1. tFighterPilot Senior Member

    Israel - Hebrew
    Is it a coincidence that this suffix in both these unrelated language families? Is there some deep psychological reason maybe?
     
  2. perevoditel Junior Member

    It is present also in Scandinavian and Slavic languages. Probably something in PIE? And you can't say those languages are not unrelated, all come from PIE.
     
  3. tFighterPilot Senior Member

    Israel - Hebrew
    Ok, but Semitic languages are still not related to them.
     
  4. Konanen

    Konanen Junior Member

    Germany, Stuttgart
    Turkish; German
    Although, one must say, that Arabic verbs express muliebrity with "(-)t(-)", "-i" and "-n-" respectively.
    But, that is a whole different story.
    It is still curious, why feminine nouns mostly have the suffix "-a".

    EDIT: <Moderator note: Moved here>
     
    Last edited by a moderator: Jun 26, 2012
  5. fdb Senior Member

    Cambridge, UK
    French (France)
    In Semitic the primary marker of the feminine gender is –t- or –at-, compare Arabic ibn-un ‘son’, and bin-t-un or ibn-at-un ‘daughter’. The feminine marker –a results from the loss of –t- and the case ending in pausal position (ibnatun > ibna). In Indo-European there is a contrast –o- vs. –ā- for masculine and feminine. Seen in this way, the two families are actually very different.
     
  6. berndf Moderator

    Geneva
    German (Germany)
    As a Hebrew native speaker, tFigterPilot is certainly aware that "t" was the original Semitic feminine marker. I think his question was why several languages (Hebrew/Arabic and Romance languages) developed (seemingly independently) -a as a feminine marker.

    @tFigterPilot: Are you aware that many linguists think that the PIE feminine has developed out of a collective inanimate form? I.e. the use of the -a suffix as neuter plural and as feminine singular marker is probably not by accident and the neuter plural meaning is the original one.
     
  7. perevoditel Junior Member

    Sorry, that was a bit faux pas. Thought that Semitic languages belong to Indo-European... Nevermind.

    I think you must check when -a suffix appeared in Semitic languages first. In Western Europe it can be Latin with it's -a suffix for feminine adjectives and nouns. In Sanskrit there are -a, -is, -os suffixes, in Old Church Slavonic it was -a, -i and -'/-ь (soft sign). So we can assume first appearing of -a suffix in about 1000-500 BC. How about Semitic languages?

    EDIT:

    @Konanen: The same with Norwegian, names of places are feminine, if they don't have masculine root (ie. High Hill is masculine, because hill is masculine in Norw.).
     
  8. perevoditel Junior Member

    Why death have different sex in different cultures, then? ^^
     
  9. merquiades

    merquiades Senior Member

    France
    USA Northeast
    One thing I've noticed in Russian (it may be present in other slavic languages too) is the masculine genitive and accusative singular have the ending -a. I found it interesting coming from a Romance language background where for example Robert / Roberta reveal gender differences. In Russian "of/ from Robert" and "I see Robert" become "Roberta".
     
  10. tFighterPilot Senior Member

    Israel - Hebrew
    I do know that it was originally "t", but not because I'm a native speaker, rather because I study linguistics. But as you said, in both Hebrew and Arabic it turned into "-a" seemingly independently. Pretty sure Aramaic doesn't have it at all. I don't know much about the southern Semitic languages.

    My knowledge of Indo European languages is very limited tbh.
     
  11. arielipi Senior Member

    Israel
    Hebrew
    Aramaic has -a suffix for hell sure(at least jewish one), maya,haga leyisrael - though its actually for male, ata tura ve shata lemaya dkava lenura dehika lekalba de nashakh keshunra deachla legadya dizabin aba bitrey zuzey had gadya had gadya.
    You could expand your question to that: why most languages have the sound of /m/ in the word mom?
    Just think about it - babies use it because its the easiest one, so it became mom, ima, mamme, etc etc.
    so, -a is the easiest vowel of them all
     
  12. tFighterPilot Senior Member

    Israel - Hebrew
    Of course, that's the definite suffix. It indicates neither gender nor quantity.
     
  13. berndf Moderator

    Geneva
    German (Germany)
    I thought because the <t> is revived in status constructus and in suffixed forms, it was still transparent to modern speakers that <t> is the "true" feminine suffix.
     
  14. tFighterPilot Senior Member

    Israel - Hebrew
    I think most speakers suppose it's the other way around, that "-a" is the original form and that "-t" is an allomorph.
     
  15. berndf Moderator

    Geneva
    German (Germany)
    I see.
     
  16. francisgranada Senior Member

    Slovakia
    Hungarian
    Interestingly, the -a appears in the IE languages also in masculine animated nouns: sluga, scriba, etc...
     
  17. Konanen

    Konanen Junior Member

    Germany, Stuttgart
    Turkish; German
    I have done some researches about that, also because I had to do a presentation about first language acquisition of children.
    I came to the humble conclusion, that one of the first sounds babies experiment with, are bilabials, and thus parents conclude their children naming them with "mamamama" or "papapapap" or "bababababa".
    Ex.:
    Mama(D, HR, ...)/Mamma(IT)/Maman(FR)/Mom(EN) - mother
    baba (TR)/Papa(D)/... - father
    baba/baka (HR) - grandmother
    dad (EN)

    BUT: mama in Turkish means baby food.

    I ask myself, why feminina and not masculina or neutra?
    Why formed the "simplest" vowel, enunciated by merely opening the mouth (a), presumably overall (coexisting with some different forms and also forming other declensions) a feminine marker, whereas masculina had a complete diverse and unpredictable development?
     
  18. francisgranada Senior Member

    Slovakia
    Hungarian
    I don't think that the IE feminines would be much more "predictable" than the masculines. Nouns like mater, soror, domus, virgo, turris, aetas ... are feminines and belong to various declensions. The same is valid for the Slavic and other IE languages.
     
    Last edited: Jun 25, 2012
  19. arielipi Senior Member

    Israel
    Hebrew
    Excuse me tfp, but the -a suffix is actually for [single] male.
     
  20. tFighterPilot Senior Member

    Israel - Hebrew
    uh, the word for city, מדינתא (pronounced in Syriac mditta) is feminine.
     
  21. rayloom Senior Member

    Paris, France
    Arabic (Hijazi Arabic)
    The common view is a parallel development or areal diffusion phenomenon (as suggested by Huehnergard, see p267 & p166 & ) in Central Semitic, so technically not independently.
    But even this is arguable, since it's based on the absence of this feature in Ugaritic (a dead language, unvocalized) and OSA (another dead language, unvocalized and not commonly considered a Central Semitic language).
    The spelling of a -t in all positions in Ugaritic, does that really tell us that the -t was always pronounced in all positions. Especially given the fact that Ugaritic was an inflected language, which means regarding the feminine singular, the -t would have been pronounced always in mid-speech like in Classical Arabic, unless the word occurs in the end of the sentence (well a feature of Classical Arabic, unknown situation for Ugaritic).

    If you take Arabic, Aramaic and Hebrew (which are the only surviving languages of Central Semitic), this is the situation for the feminine singular:
    - Classical Arabic pausal forms (same for Colloquial Arabic):
    Indefinite state: CCC-a
    Construct state: CCC-at
    Definite state: al-CCC-a

    Aramaic (see here):
    Indefinite state: CCC-a
    Construct state: CCC-at
    Definite state: CCC-ta (or CCC-to in Neo-Aramaic Suryoyo)

    Hebrew
    Indefinite state: CCC-a
    Construct state: CCC-at
    Definite state: ha-CCC-a

    So I don't see why it can't be a feature of Proto-Central-Semitic!
     
  22. fdb Senior Member

    Cambridge, UK
    French (France)
    There are three problems with this:

    1. the concept “Central Semitic” is far from being universally recognised.
    2. There is a difference between long /ā/ as in the Aramaic fem. sing. abs. and in Hebrew, and short /a/ as in Arabic fem. sing. pausal.
    3. If you really want to reconstruct a proto-language you have to take the oldest form of each daughter language as your point of departure, in the case of Arabic the context form –atun (like Akkadian –atum) and not the pausal –a(h).
     
  23. fdb Senior Member

    Cambridge, UK
    French (France)
    And while we are on the subject of vowel length: the IE ending for neut. pl. nom./acc. is short /a/, while the fem. sing. nom. of many (of course not all) words is /ā/, or rather the stem ending /-ā-/ plus the case ending zero. But besides: what is actually the logical connection between "collective inanimate" and feminine?
     
    Last edited by a moderator: Jun 26, 2012
  24. berndf Moderator

    Geneva
    German (Germany)
    He are talking about different development stages here. The lengthening of the feminine suffix probably resulted from the lost -h2 which did not happen when the neuter plural developed; pre-PIE inanimate probably didn't have dual and plural forms but only distinguished individual and collective, both with singular verb forms. The use of singular verb forms with neuter plural in Greek is assumed to be a reflex of this stage.
    Many nouns relating to females were probably inanimate which may have caused the re-interpretation. But this is indeed something this theory can't explain very well. The explanation sounds a bit ad-hoc. Here is a summary of the theory.
     
  25. rayloom Senior Member

    Paris, France
    Arabic (Hijazi Arabic)
    2. There might be a difference in the quality or length of the (same) vowel, but they do lose the feminine -t marker.
    3. Not to construct, but at least not to rule out a genetic inheritance of this phenomenon.
    Also as far as Arabic is concerned, the pausal form -a(h) is as old as the "context" or sandhi form (-atun/-atan/-atin), both features of Classical Arabic.
     
  26. berndf Moderator

    Geneva
    German (Germany)
    If you read the question carefully you will find that this is not the purpose here. The question was if there is a "natural tendency" towards -a as a feminine marker. If so, it can develop at any stage of a language.
     
  27. Ironicus Senior Member

    English & Swahili - East Africa
    True, by virtue of the tav, and the following alap (since we're speaking Aramaic) makes it definite, so it's not merely city, it's the city.
     
  28. tFighterPilot Senior Member

    Israel - Hebrew
    Not in modern Aramaic (Syriac), though. There you always add the א in the end regardless of whether it's "a city" or "the city".
     
  29. Ironicus Senior Member

    English & Swahili - East Africa
    Well, there's only one literate Aramaic speaker in my neighborhood, tFighter, so I'll take your word for it until I can consult with him....
     
  30. fdb Senior Member

    Cambridge, UK
    French (France)
    I was replying to rayloom: "So I don't see why it can't be a feature of Proto-Central-Semitic!"


     
  31. berndf Moderator

    Geneva
    German (Germany)
    I see.
     
  32. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    I don't know about the Aramaic & Hebrew usages, but my understanding of the Arabic usage is that the -t is merely softened in pausal form, it does not really become -a, in effect only half the sound is made, which sounds similar to -a(h).
     
  33. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    Most societies considered females to be part and parcel of the belongings of men (chattel) until not that long ago, so it would've made sense to lump them grammatically with all the other [virtually] inanimate possessions they had, no?
     
  34. francisgranada Senior Member

    Slovakia
    Hungarian
    I don't think so. The feminine and masculine declension in the IE languages is quite similar, the paradigm depends rather on the stem and not on the gender (various stems, including in -a, do exist both in masc. and. fem.). The neuter nouns, instead, behave differently in the sense, that the accusative has the same form as the nominative (both in sg. and pl.).

    P.S. See also #16 and #18
     
    Last edited: Jun 29, 2012
  35. fdb Senior Member

    Cambridge, UK
    French (France)
    It is a delusion (held mainly by monolingual English speakers) to think that there is an inherent connection between (grammatical) gender and (biological) sex. In the classic Indo-European and Semitic languages the gender of nouns designating animate beings is (normally) governed by their sex, but the nouns designating inanimate beings are distributed without any obvious logic between the three (in Semitic: two) genders. French speakers (for example) do not really believe that their foot (le pied) is male and their hand (la main) is female. It is purely a question of grammar.
     
    Last edited: Jun 29, 2012
  36. berndf Moderator

    Geneva
    German (Germany)
    You are again mixing up development stages. In this theory we are speaking about a postulated state before the PIE case system developed and the animate agent marker -s and the animate patient marker -m out of which the nominative -s and the accusative -m developed was anything there was in terms of a case system. Inanimate nouns weren't case marked as they didn't appear as agents, they could only be patients of active and subject of stative verbs (Pre-PIE at this development stage is postulated to have been an active-stative language). Please read the text linked to in #24.
     
  37. berndf Moderator

    Geneva
    German (Germany)
    You are absolutely right, but I don't think anybody here fell into that trap. There is still an interaction between gender and sex which led to the interpretation of the three genders as male, female and neuter.
     
  38. fdb Senior Member

    Cambridge, UK
    French (France)
    So the poor proto-Indo-Europeans could not say "The stone crushed him" ?
     
  39. berndf Moderator

    Geneva
    German (Germany)
    I largely agree with you. There is one thing, one has to be careful about. Like gender should not be confused with sex, animate/inanimate should also not take as purely biological. Animate means that a noun can occur as an agent of a transitive verb.
     
  40. fdb Senior Member

    Cambridge, UK
    French (France)
    I think this is a totally arbitrary definition of the word “animate”. Anyway, the article to which you have supplied the link says explicitly: “only living things can act upon other things, so only animate nouns could take the *-s.” The same link also says that Hittite “had only masculine and neuter genders”. This is wrong. Hittite did not have any “genders”, but only a distinction between animate (also called “common”) and inanimate nouns (like Sumerian, Tamil, Elamite and lots of other languages).
     
  41. berndf Moderator

    Geneva
    German (Germany)
    The arbitrariness lies mainly in the choice of labels (masculine, animate, living thing). In an animistic society as the pre-PIE with some degree of likelihood was as many ethnologists would argue the animate-inanimate distinction would be meaningless anyhow. Needless to say that all this is very speculative. The main reason why I build up this case here is to show there are other ways to approach all this than through the glasses of our modern weltbild.
     
  42. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    I think we effectively meant the same thing there. My point was women were not even considered agents within society itself, they were things acted upon, likewise [virtually] inanimate property had no agency with which to act either, they were acted upon, and therefore were virtually inanimate.
     
  43. Ironicus Senior Member

    English & Swahili - East Africa
    Do you have a source for this extraordinary statement? There is no trace in either the Romance or the Semitic family of grammatical restrictions on the capability of females.
     
  44. fdb Senior Member

    Cambridge, UK
    French (France)
    Indeed, there is no source for this, it is pure speculation. As I mentioned already, there are lots of real languages which have the semantic catagories "animate" vs "inanimate", and, in all of these, nouns denoting female beings belong to the animate group. This is the case in Sumerian and Elamite, two of the three oldest attested languages in the world.
     
  45. berndf Moderator

    Geneva
    German (Germany)
    No. Some grammarians call the Sumerian genders "animate/inanimate". But these are purely labels. But only because you read these labels in modern texts books it does not mean that Sumerian genders had this semantic. Less misleading labels for the Sumerian genders could be individual vs. collective. But also this interpretation remains ultimately speculative.
     
  46. francisgranada Senior Member

    Slovakia
    Hungarian
    Sorry, but don’t understand your reaction … Aren't we speaking about the -a suffix in existing languages? Where from could we arrive to whatever conclusion/opinion if not analyzing development stages?

    As reaction to the Abu Rashid's post, I only wanted to say that the IE declension system does not support/suggest the idea that women “belonged” to men ("they were their property") and therefore these nouns were “treated” as inanimate objects (ending in -a, as the plural of the later neutra or collective nouns). To the contrary, “women” and “men” present similar (or equal) paradigmata suggesting that they “philosofically” belonged to the same cathegory (animated human beengs and not inanimated objects) probably even before the development of the proper declension. But this is only my opinion …

    An other question is the origin of the relatively diffused feminine –a suffix in the IE languages ...
     
  47. berndf Moderator

    Geneva
    German (Germany)
    Of course not, because the animate/inanimate distinction belonged to a different development stage (if this theory is true). AR contemplated in how the IE feminine could have developed out of an earlier inanimate collective.
     
  48. Angelo di fuoco Senior Member

    Germany
    Russian & German (GER) bilingual
    Yes, but there's a difference between Ро́берта (masculine singular Genitive and Accusative) and Робе́рта (feminine singular Nominative). Stress usually isn't marked in Russian except in dictionaries and when the context doesn't make it clear.
     
  49. merquiades

    merquiades Senior Member

    France
    USA Northeast
    Are you meaning to say that the stress of masculine accusative-genitive nouns in -a consistently switches stress to the first syllable/ stem? Or is it just the case with Robert(a)
     
  50. tFighterPilot Senior Member

    Israel - Hebrew
    I think it's pretty clear that this suffix simply leaves the stress of the word in the same place (in this case the original word is Ро́берт), while the feminine version of the name pushes the stress forward.
     

Share This Page