Letter of Recommendation: Uzbek.

A NY Times Magazine piece by Lydia Kiesling about her experiences with the Uzbek language begins:

Four years ago, the federal government paid me a large sum — a year of graduate-school tuition, plus a stipend — to study Uzbek at the University of Chicago. Uzbek is among the least commonly taught of the so-called Less Commonly Taught Languages, or L.C.T.L.s. So uncommonly is it taught, in fact, that without federal largess it would hardly be taught at all. Because I happened to speak decent Turkish, a cousin of Uzbek, and because I spent a week in Uzbekistan when I was 22, and because life is nothing if not a sequence of odd choices vaguely considered, for two years I sat in a room with two other students and produced some extremely literal translations.

It’s a charming reminiscence, but I’m bringing it here for this brief section:

The grammar is simple, but the history is complex. National borders can be risibly at odds with reality, especially in Central Asia, where Turks, Mongols, Persians and others roved and mingled, where “Uzbek” was, for a time, more of a descriptive antonym of “Tajik” — no­­madic versus settled — than an ethnic classification. Later, the Soviets complicated things with mass reorganizations of their Central Asian subjects. The question of whether there is mutual intelligibility among Turkic languages is not simply a linguistic matter but an ideological one, at the core of nationalist movements that have formed and reformed across time and empires.

There is more actual, verifiable, sensible information about language and history packed into those few sentences than in the entirety of most Times “news” articles on linguistic topics. Well done, Ms. Kiesling!

Comments

  1. Greg Pandatshang says:

    Was “Tajik” really applied to settled Turkics? I always thought it was a catch-all term for non-Turk locals in Central Asia, esp. Iranians and/or esp. Muslims (i.e. an outsider, but not a religious outsider).

  2. The saddest thing about the article is that she never gets to actually use her Uzbek.

    Her comments on grammar also largely apply to Japanese and Mongolian (which I’m acquainted with) and probably Korean, Hungarian, and Finnish (which I’m not). I’ve always wondered what it’s like to translate texts between these languages. Do translations just follow the contours of the original — something that is often impossible when translating into English. Or are there other issues (especially written expressions and vocabulary) that make this difficult?

  3. Settled population of Turkestan were called either Tajiks or Sarts (Turkic speaking settled farmers).

    Soviet nationalities commissions studied the issue, went for the linguistic criteria and decided that there is no need for separate Sart nationality and they were lumped together with nomads as Uzbeks.

    The problem was that the Sarts and Tajiks were bilingual in Uzbek and Tajik, had same religion – Sunni Islam- and had no concept of nationality.

    Partition into Uzbekistan and Tajikistan was therefore pretty artificial. To this day, Uzbeks complain that half of Tajikistan is populated by Uzbeks and Tajiks maintain that half of Uzbekistan is actually Tajik.

  4. Uzbek language reflects this heritage bilingualism. It is quite possible to make Uzbek phrases or even whole speeches where every single word is Tajik loanword.

  5. We discussed Tajik and Sart (as well as Tat) a few years ago. I’ve always been fascinated by both the term Sart and the people it denotes.

  6. A minor quibble: qa is not nasal – it’s uvular.

  7. Also, afsuski, qaerda.

  8. A charming little essay.

    Some twenty years ago I chanced to have a conversation with an Uzbek in Moscow’s Ярославский вокзал (Yaroslavl [train] Station). My Russian is limited to maybe 100 words and there’s no way I can so much as string a subject and verb together. My interlocutor spoke no English, or perhaps his English was at the level of my Russian. Fortunately, as a conscripted soldier he had served for two years in East Germany. So his halting German and the stumbling Yiddish I picked up from my grandmother — plus lots of gestures and the occasional resort to drawing on paper — sufficed to let us chat for fifteen minutes or so. All I recall of the conversation is that he taught school, drove an elderly BMW, and was headed back home.

  9. “Partition into Uzbekistan and Tajikistan was therefore pretty artificial. To this day, Uzbeks complain that half of Tajikistan is populated by Uzbeks and Tajiks maintain that half of Uzbekistan is actually Tajik.”

    Wait a few more years and they will be partitioning Moskvabad 🙂

  10. marie-lucie says:

    qa is not nasal – it’s uvular

    The two are not incompatible: it could be a uvular nasal, which speakers of other languages might interpret differently according to their own linguistic background.

  11. The two are not incompatible: it could be a uvular nasal, which speakers of other languages might interpret differently according to their own linguistic background. But it’s a voiceless uvular stop. I don’t know where Ms Kiesling found the nasality.
    On language and ethnicity – the native inhabitants of the old cities of Samarkand and Bukhara identify themselves as Uzbek, but speak Persian (Tajik) among themselves.
    Also, afsuski, qaerda. And chiroyli. The NYT is in urgent need of an Uzbek copy editor. 😉

  12. Charles Perry says:

    Well, Uzbek is somewhat studied. In 1969 the Indiana Uralic and Altaic series published a course in Uzbek by Alo Raun. Uniquely in this series (so far as I know), it has a conversation in which an American asks a local girl for a kiss (Bir opish bermaysizni?) and tells her “It’s impossible not to love you” (Sizni sewmaslik mumkin emas). Possibly an attempt to interest college-age guys to study Uzbek.

  13. bermaysizni

    I hate to be a spoilsport today, but the question particle is -mi, which has btw been borrowed into Tajik.

  14. Greg Pandatshang says:

    If they put me in charge of re-doing early 20th century Soviet anthropology (which, come to think, is a great idea. I’ll start drafting a letter to the United Nations), I would call the Uzbeks “West Chagatais” and the Uyghurs “East Chagatais”.

  15. China is having enough trouble with “East Turkestan”. Why would they agree to “East Chagatai”?

  16. Fortunately, we haven’t yet reached the point where we all have to have China’s agreement before coining terminology.

  17. Moghulistani would do as well….

  18. This brought me back to my wanderings around the eastern Ukraine eight years ago. Completely frustrated with the catered “Russian” food we received while on tour there, i discovered the open markets in cities where Uzbeks run the dried fruit and nuts stands, and there is always a halal Uzbek lunch shack and bakery. I didn’t speak any Russian, but my bad tourist Turkish kept me filled with lamb kebab and lagman.

  19. lamb kebab and lagman

    Now I’m hungry!

  20. Still, the notion that Turkic languages are mutually intelligible should be taken with a pinch of salt.
    In 1935 my grandmother and mother had to flee their home village in Ryazan oblast after the great-grandfather had been arrested and sent to a labor camp. (He died there in 1942.)
    After arriving in Tashkent Mother couldn’t understand anything. She was eleven going on for twelve, had no Russian to speak of, and could only speak the local dialect of Tatar. Still, she could make no sense of Uzbek at all. Grandmother, I think, could, because she had lived in Kokand sometime in the early 20s and had a daughter there, who died of scarlet fever or typhoid at the age of one or so. (Grandfather had been a railway worker; he had gone missing before Mother was born.)
    Of course, Mother learned both Russian and Uzbek in time, but it was neither quick nor easy.

  21. Very interesting, and a useful thing to keep in mind when one is tempted to overemphasize the similarity of Turkic languages.

  22. marie-lucie says:

    Just because some languages are closely related does not mean that they are mutually intelligible, especially orally. Speakers of one may find it relatively easy to learn another, but that does not mean that learning will be quick or effortless. A French person finds it much easier to learn Spanish than German, but both are incomprehensible at first hearing. Ease and speed of learning will also depend on a person’s general “aptitude for languages”: some people have trouble even with obvious cognates.

  23. Right, but the popular image of Turkic languages among those who know something about language families is that they’re particularly closely related, much more so than French and Spanish.

  24. I wonder why this sense of the close relation of Turkic languages persists. (I know a handful of Turkish words, and the meaning of Almaty [Alma Ata], but that’s it.) It would seem that the Turkish of Turkey has been drifting away from the other Turkic languages for pretty much 1,000 years. What’s going on?

  25. Turkish, Tatar or Uzbek are about as different from each other as French and Spanish.

    But pairs like Turkish and Azeri or Tatar and Bashkir or Uzbek and Uighur or Kazakh and Karakalpak are much closer (like Spanish and Portuguese or even less)

  26. As far as I know, Chuvash is the most divergent.
    Grandmother told me that after a relative was married off and had to move to the village in Chuvashia where her husband was a mulla, she was bitter that she could not understand the locals.

    On a tangent somewhat, I’ve been wondering if the Turkic ta:γ – variously realized as to’g, dag, taw, to: – meaning “mountain”, the Japanese 岳take, and the Inuit nunatak are related, however remotely. The form is suggestive, but it may not prove anything.

  27. Greenlandic nuna-taq, from nuna- ‘land’ and -taq ‘pertaining to’, so something like ‘land-like’, ‘of earth’. The relation to the Turkic or Japanese words is coincidental.

  28. marie-lucie says:

    This example shows one of the pitfalls of word comparison if one does not know a language well enough to analyze its words. The problem is compounded when dealing with several languages that one knows only fragmentarily or not at all.

  29. Sorry for my english. I am Azerbaijan turk, and I did understand what you’ve written ” it has a conversation in which an American asks a local girl for a kiss (Bir opish bermaysizni?) and tells her “It’s impossible not to love you” (Sizni sewmaslik mumkin emas)” we would say: bir opüş vermezsız? Sizi sevmemek mümkün deyıl (olmaz). So as a turk I think turkıc languages are really very similar. And azerbaijan.and turkey turkish aren’t like spanish and portuguese, they’re like britain, american.and australian english. I am writing frım the hotel in Turkey now, there are a lot of gazakhs, and when they speak ı can understand what they’re speaking about a little bit, but my and their turkish is from different groups – oğuz and qıpçaq. I dont think that any german can understand spanish even so little. And I know russian, like my 2nd language, but belarus and ukraine are very different, but they are much closer than we and qazaxstan. I’m not joking or lying I can even understand china uyghur songs))) so, please dont say that turkish languages are not similar

  30. Shebnem, how do you find Tabriz Azeri–is it difficult to understand for you? It has had some dramatic sound changes that Istanbul Turks usually have lots of trouble with, but someone from Bakı might have more luck.

  31. David Marjanović says:

    On a tangent somewhat, I’ve been wondering if the Turkic ta:γ – variously realized as to’g, dag, taw, to: – meaning “mountain”, the Japanese 岳take […] are related, however remotely.

    Other people have wondered, too. The state of the art at a glance.

  32. As Marie-Lucie said, mutual intellegibility has also something to do with language learning aptitude. I find that Kazakh, Uzbek, and Turkish are sufficiently similar that knowledge of one of them often carries over to the other ones and helps me to understand them. As is frequently the case, it works better on the level of the written language than on the spoken language level.

  33. Also, some Turkic languages consist of a literary standard superimposed over several widely differing dialects, some of which might be closer to literary standard of another Turkic language.

    Crimean Tatar is case in point. One Crimean Tatar dialect is very close to standard Turkish (Oghuz branch of Turkic languages), another dialect is close to Nogay language (Kypchak branch which includes Kazakh) and the literary standard is based on a third dialect which is somewhat in the middle.

  34. So when someone says that differences between, say, Turkish and Kazakh, are within range of differences between dialects of one language, this is what is actually meant.

  35. Uzbek, for example, includes dialects very close to Turkmen, dialects relatively close to Kazakh and dialects close to Uighur (most dialects and majority of speakers).

    They belong to three different branches of Turkic languages.

    It’s like a Romance language which has some dialects very close to French, some to Spanish and some to Italian.

    Or using Slavic analogy, it would be a very strange Slavic language which includes some dialects close to Russian, some to Czech and some to Serbian.

  36. So are Tatar and Uzbek mutually unintelligible?

    The correct answer is yes, unintelligible, but no more so than different dialects within Uzbek itself.

  37. marie-lucie says:

    SFR: It’s like a Romance language which has some dialects very close to French, some to Spanish and some to Italian.

    Add Catalan to the mix and you have described “Occitan” (which, depending on your criteria, can be considered one language or four or five).

  38. Could the answer be zero?

    I mean could we reassign all dialects of Occitan to other Romance languages, ending up with no Occitan proper?

  39. In principle, yes. But is it useful to do so? There is or was a chain of mutually intelligible varieties that leads from Standard English through Sranan and Saramaccan to Standard European Portuguese, but English is not Portuguese.

    “Though no man can draw a stroke between the confines of day and night, still light and darkness are on the whole tolerably distinct.” —Edmund Burke

  40. Marie-Lucie: actually, the so-called “dialects” of Italian would be a better example than Occitan: Occitan does have a core of shared innovations (chiefly phonological in nature: which is why, SFReader, you couldn’t reassign all Occitan dialects to neighboring Romance varieties on purely structural grounds), unlike the Romance “dialects” of the Italian peninsula.

    The unfortunate ambiguity of the term “dialect” and the resulting miscommunication which this ambiguity causes, even among linguists, is very nicely analyzed (using Romance data, but it is quite relevant for Turkic or indeed any language family) in this fine (and short!) article (in French) which I recently found and heartily recommend:

    http://www.romanistik.uni-muenchen.de/downloads/links_personen/krefeld/krefeld-iasi.pdf

  41. marie-lucie says:

    Merci, Etienne! I agree with your characterization of Occitan. I am not very familiar with Italian dialects.

    The linked article is indeed very interesting both theoretically and practically.

  42. Etienne’s point is that the Romance varieties spoken in northern Italy (excluding Ladin and Friulian and maybe Venetian) pattern with French and Occitan rather than Standard Italian. In particular, they are not pro-drop and have lost final vowels other than /a/ (French -e), though the vocabulary is much more Italian-like. The linked page shows a single sentence ‘She always closes the window before dining’ in many varieties of northern Italy, plus Standard Latin, French, Italian, Romansh, and Romanian.

  43. 1-Marie-Lucie: I gave some indications on the diversity of Italian dialects at the end of this thread:

    http://languagehat.com/written-language-quiz/

    2-John Cowan: actually, while you are quite right that Gallo-Italian is in many respects more French- than Italian-like, many of the common features arose at a demonstrably late date, long after French and Gallo-Italian had turned into separate varieties.

    Such as the case for the obligatory use of cliticized subject pronouns (i.e. the so-called loss of Pro-Drop; I say so-called because a linguist has shown that there is no sharp diving line between Pro-Drop and non-Pro-Drop Romance varieties and thus that there is no evidence for a single “Pro-drop” parameter in human, or at any event Romance-speakers’, minds), which thus is quite irrelevant for purposes of Romance genealogical (as opposed to typological) subclassification.

    The same is possibly true of the more Italian-looking vocabulary of Gallo-Italian. This is a purely personal impression: long ago, in my wild and misspent youth, I examined writings in medieval Romance languages, including Old French, early Tuscan and contemporary varieties of Gallo-Italian, and I remember being struck by how “un-Italian” Gallo-Italian texts were.

  44. Yes, I assumed the vocabulary was superstratal. I don’t believe in P & P pro-drop either, but it’s a useful way to describe the normal state of languages (though it would be better if English, French, German were called pro-required (or perhaps anti-drop).

  45. David Marjanović says:

    Fun fact: my dialect is pro-drop in the second person, where the verb endings are unambiguous and – from a historical point of view – contain remnants of the pronouns, but in the other persons it’s just as anti-drop as the standard.

    Further fun fact: this does not apply to clauses with verb-final word order, where a clitic is obligatory – to the extent an unreduced pronoun can be added to it (for emphasis), but cannot replace it.

  46. David Marjanović says:

    Speaking of clitics… it has been suggested that the French unstressed pronouns are actually just verb prefixes anymore, so French is actually pro-drop again. I’m quite inclined to agree.

  47. And as a consequence French now has free word order:

    ʒdetɛst mari (VO)
    ʒdetɛst mari mwa (VOS)
    mwa ʒdetɛst mari (SVO)
    mari ʒladetɛst (OV)
    mari ʒladetɛst mwa (OVS)

    The “la” prefix signals that the object is not directly after the verb, a very non-IE idea.

  48. David Marjanović says:

    And as a consequence French now has free word order:

    More like Chinese word order: topic first.

  49. it has been suggested that the French unstressed pronouns are actually just verb prefixes anymore
    How? Doesn’t that mean that everything else in between — negators, auxiliaries, object pronouns — is morphologically part of the verb?
    That’s why God invented clitics, so that one wouldn’t have to make such extreme commitments.

  50. Sure, why not? Negative verbs are often found outside IE, and so are object agreement affixes.

  51. marie-lucie says:

    DM: it has been suggested that the French unstressed pronouns are actually just verb prefixes anymore, so French is actually pro-drop again. I’m quite inclined to agree.

    I do too. I would call the “prefixes” ‘clitics’ (less closely bound to the verb), as Y suggests. (This is not quite a new idea!).

    JC: X ways to say “I hate Mary”, with varying emphasis:

    ʒdetɛst mari (VO)
    ʒdetɛst mari mwa (VOS)
    mwa ʒdetɛst mari (SVO)
    mari ʒladetɛst (OV)
    mari ʒladetɛst mwa (OVS)

    The “la” prefix signals that the object is not directly after the verb, a very non-IE idea.

    But that possibility (“la” announcing an O after the verb) is also grammatical (in conversational register, which all the examples with “mwa” belong to):

    mari mwa ʒladetɛst (OSV)

    mwa mari ʒladetɛst (SOV)

    ʒladetɛst mari mwa (VOS)

    ʒladetɛst mwa mari (VSO)

  52. marie-lucie says:

    Someone (whose name I don’t remember) years ago wrote about “the Chinook structure of French”. The Chinook language (formerly spoken along the Columbia River) builds verbs with a series of clitics preceding the verb stem, in very similar fashion to colloquial French.

  53. David Marjanović says:

    I would call the “prefixes” ‘clitics’ (less closely bound to the verb)

    In the written standard they’re of course less closely bound than prefixes; they can for example go on the other side, as in the thoroughly unidiomatic question déteste-je Marie ?. But are they still clitics in modern colloquial language?

  54. David: you and I had already discussed French pronominal pre-verbal elements and the issue as to whether they are best considered clitics or prefixes right here at Casa Hat:

    http://languagehat.com/is-french-losing-gender/#comments

    Also, it should be “Détesté-je Marie?”, not *”Déteste-je Marie?”: French verbs which, in the present indicative and subjunctive, end in schwa in the first person singular, regularly change this schwa to /e/ when pronominal inversion occurs with the first person singular doohicky (whose nature we may argue about) ‘je’.

    Mark you, inversion (with any personal pronoun) is so marginal in spoken French today that I am certain this rule is quite unknown to many native speakers…

    Y: The pre-verbal French negator element “ne” is dying, leaving spoken French with a purely post-verbal negator (pas). As a result, between a subject “clitic” and its verb (which can be an auxiliary verb, of course) the only possible element is an object or adverbial “clitic”, which in colloquial registers can be accompanied by the noun supposedly replaced by said clitic.

    To exemplify: whereas in Standard French you’ve either “Je parle aux élèves de leur examen” or “Je leur en parle”, Colloquial French would have “(moi) (je leur en parle) (aux élèves) (de leur examen)”, where all of the elements between parentheses may be scrambled according to emphasis and/or taste, and all except “je leur en parle” may be deleted.

    The one constraint: the preposition is deleted if the NP it qualifies is pre-verbal. Thus “Je leur en parle de leur examen aux élèves, moi” versus “Les élèves, leur examen, je leur en parle, moi” (“moi” needn’t be sentence-final in either instance of course).

    The important point is that in “Je leur en parle”, in Colloquial French, the first three elements are in a fixed order and cannot be separated from the verb, which itself can change tense/mood or be negated without the three initial elements changing: “Je leur en parle/Je leur en ai parlé/Je leur en parlais/Je leur en parlerai/Je leur en aurais parlé/Que je leur en parle/Je ne leur en parlerais pas…” . Thus overall these three elements behave like prefixes far more than like pronominal clitics.

  55. Correction: my last example of Colloquial French should be “Je leur en parlerais pas”, since (as I had myself pointed out…) “ne” is nearly extinct in Colloquial French…

  56. So my question is, is there a reason to consider the French verb and its attachments, from the subject pronoun to the root to the optional pas, a single phonological word, with a root surrounded by affixes, a run-of the-mill verb frame?

  57. marie-lucie says:

    Y: I would use “stem” rather than “root”, but overall, yes.

    In this type of structure (not typical SAE but common in some parts of the world), the pronominal affixes are obligatory parts of the verb rather than simply “agreement markers” referring to nouns. In turn, nouns (or NPs) and stressed pronouns (eg “moi”) are considered adjuncts to the verb, and do not necessarily occupy a fixed position.

  58. So what is the relationship between colloquial French and the written language? How do people connect the two? Two different languages? Two varieties with regular correspondences?

    Is there a divide between people who use colloquial French and those who don’t (i.e., are there people who don’t use ‘colloquial French’, or do so only sparingly?)

  59. David Marjanović says:

    a single phonological word

    Hm. Does French have such a thing as phonological words?

    German does, and they don’t line up with what are words by other criteria. The components of compound nouns remain separate phonological words, which is visible in the fact that each of them retains its own stress – there’s never such vowel reduction as in English postman, mailman, policeman, Newfoundland; and yet, many such compounds contain a connecting element (usually -s-) that has no other function than to mark the compound as a single word. (More specifically, it marks the end of the first component, which doesn’t happen to independent words.)

  60. David Marjanović says:

    So what is the relationship between colloquial French and the written language? How do people connect the two? Two different languages? Two varieties with regular correspondences?

    I’m sure what’s going on in the conscious minds of literate adults is the latter (two – or more – registers with regular correspondences), especially now that everyone in the Hexagone speaks either Parisian or slower Parisian. Whether that’s the description that makes the most sense is a separate question, though.

    For comparison, the Viennese believe that their mesolect is the colloquial register of Standard German. Germans tend to believe instead that this is the Viennese dialect.

  61. Here’s a thought. Consider these French sentences (which I can’t judge for grammaticality, since I’m unfamiliar with colloquial French):

    Il chant.
    Il chant pas.
    Il a chanté.
    Il a pas chanté.

    Although each one consists of one phonological word, you can’t fit all of them into a standard verbal frame with fixed slots. Since pas occurs at different positions with regard to the verb, it cannot be considered an affix, but rather a clitic, and so nothing else here can be said to be an affix to the stem.
    I haven’t read the article comparing French to Chinook, but this seems a looser affair than what I would call a textbook case of synthesis.

  62. David Marjanović says:

    each one consists of one phonological word

    On which criteria?

    pas occurs at different positions with regard to the verb

    The verb, for morphological purposes, isn’t the participle chanté in your 3rd and 4th examples. It’s a; you need to compare chante pas to a pas.

    Basque does this, too; affixes are piled on auxiliary verbs, while the content word stands next to this pile all isolated.

  63. Aidan Kehoe, of course we understand Tebriz azerbaijanians, except for persian words they use, and they dont understand if we use russian words. I dont know how in İran, but in Azerbaijan we call Tebriz, Marağa, Erdebil and neighboring places- South Azerbaijan))) , and here in Az.-n we think that we are same nations, only they have a dıfferent intonation, I think a kind of persian and it’s like our Nakhchıvan people’s intonation

  64. David: ‘one morphological word’: feel free to show me wrong. I’m thinking il can’t be a separeate word, because je in the same role can’t be a separate word on phonotactic grounds; each of these phrases has one primary stress, and no secondary stresses; and because, I presume, a speaker would never break this up into individual morphemes in slow deliberate speech. Again, I’m guessing here, and I may be wrong on any of these points.

    I agree that pas attaches to a, not chanté. What I am saying is that because of this, this scheme is very different from what you’d see in most (poly-) synthetic languages of the Americas, where the verb stem is surrounded by TAM markers, negators and pronominal markers all in fixed relative positions.

    Because of this, I’d call the French subject marker a clitic, not a prefix: it attaches to words of different classes.

  65. David Marjanović says:

    each of these phrases has one primary stress

    But French doesn’t have word-level stress to begin with. Stress exists, but it goes on the last syllable of the whole utterance with no respect for words.

    Because of this, I’d call the French subject marker a clitic, not a prefix: it attaches to words of different classes.

    Only if you count main and auxiliary verbs as different classes.

    Pffft: “One of the remarkable characteristics of the Basque verb is the fact that only a very few verbs can be conjugated synthetically (i.e. have morphological finite forms); the rest only have non-finite forms, which can enter into a wide variety of compound tense structures (consisting of a non-finite verb form combined with a finite auxiliary) and are conjugated in this way (periphrastically). For example, ‘I come’ is nator (a synthetic finite form), but ‘I arrive’ is iristen naiz (a periphrastic form, literally ‘arriving I-am’).”

  66. Y: your forms “Il chant/Il chant pas” should be “Il chante/Il chante pas” (“Chant” is perfectly good French, but it’s a masculine noun).

    I frankly am baffled by your claim that because the subject/object prefixes are attached to auxiliary verbs and not to the main verb they ought to be considered clitics. To compare: the second person singular /i/ suffix of Italian (DIC-I” you say”, CRED-I “you believe) is universally considered to be a suffix, despite its also being attached to auxiliaries and not to main verbs (cf. L’AVEV-I CREDUTO “You had believed it”. The French subject prefixes/clitics thus strike me as being more akin to prefixes than to clitics, because their behavior is so similar to that of what in other Romance languages are unambiguously bound morphemes (in this instance suffixes).

    Bathrobe: the connection between colloquial and standard written French is not that different from the relationship between spoken and standard written English. In the case of the pronominal prefixes/clitics French speakers learn very quickly that, in the standard, you may use the pronominal prefixes or the nouns which they replace, but you cannot combine the two, unlike what takes place in spoken colloquial French, where the two are frequently combined.

    And in this instance this is the only real difference between both registers: the pre-verbal pronominal prefixes must appear in a strict order (subject + indirect object + direct object + adverbial clitic (i.e. “en” or “y”) before the verb), with one twist (if two object pronouns with initial /l/ appear after one another, the relative order becomes direct object + indirect object.Cf. “Je te le donne”/”I give it to you”, “Je t’en parle”/”I speak to you about it”, “Il se le donne”/”He gives it to himself” versus “Je le lui donne”/”I give it to him/her”).

    Crucially, these rules of clitic/prefix ordering are nearly identical in standard and colloquial French, so that there really isn’t that big a gap between those two registers.

  67. Etienne, thanks for correcting my French, a language I haven’t studied or used in many years. I see your point with Italian. However, l’avevi creduto are two separate words. Would you say that /ilapaʃɑ̃te/ is one or two phonological words? If /ʃɑ̃te/ is separate, then yes, I see no reason not to call il- and -pa- prefixes.

  68. I would say that it is two morphological words, and that French does not have phonological words.

Trackbacks

  1. […] Hat links to the story of a woman who learned […]

Speak Your Mind

*