Evolution of Language

Discussion in 'Etymology, History of languages and Linguistics (EHL)' started by sethmachine, Jun 7, 2008.

  1. sethmachine Senior Member

    English-US
    A trend in many Western European languages is to gradually become more analytic and less synthetic over time. Why is that? Many modern-day languages are descended from much more heavily inflected ancestors (English to Old English). However, in some places, languages still remain heavily inflected just like their ancestor(s). Can a moderately analytic language, like English, ever become more synthetic over time and re-develop case endings (taking into account all the communication technology available today)? If so, how (where would we get case endings, etc.) . If not, is there some constant/law which would say that a slow shift from a language being analytic to synthetic in this modern era can never happen?
     
  2. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    New development of case endings certainly is possible and has happened many times in recent history (in the last two millenia). Many European languages developped new future tenses; in Serbian and Croatian, for example, there's one (of two) future tenses where still the 'new ending' could be written separately (e. g. vidjet će, in Croatian) or as one word (e. g. videće in Serbian = видеће in Cyrillic script; the only difference between the two in pronunciation is /e/ vs. /je/ which has nothing to do with the ending, meaning for both is the same).

    So it happens, and it has happened - also in Romance and Germanic languages. And it may happen again.

    As for why the Indoeuropean languages mostly had gone through huge simplifications (only some, like most Slavic and all Baltic languages, have preserved much of the original complexity) and are on its way to becoming predominantly analytic languages - there are loads of theories about that one.

    To dispute them would take years (in fact this is has been disputed for centuries already).
    My personal favourite is that languages develop towards simplification if there are external factors (severe economical and/or cultural changes in society) - if there aren't then they should remain relatively stable - and that they differentiate again to make up for the loss of linguistic means of expressing certain meanings; this could be a cycle like 'synthetic > analytic > synthetic' as also proposed by generative grammar, but personally I do not support the generative grammar viewpoint at all because I think the reason for change is not grammatical at all, but mainly cultural and/or economical.

    For my viewpoint there is (far as I know) no authority in the linguistic world supporting it; as for generative (and universalist) grammar you may find lots of references by googling.
     
  3. palomnik Senior Member

    Vietnam
    English
    There's a subtext to this argument that bothers me.

    It's the inherent assumption that analytic languages are "easier" than synthetic languages are.

    I don't necessarily think that this is true. Classical Chinese is the epitome of an analytic language. And after you set aside the hurdle of the writing system, it is still formidably difficult. Things like plurality are indicated by adverbs as often as not (this happens in modern Chinese as well) and sentence structure can be very complicated.
     
  4. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    I agree.
    Even English is by far not as 'easy' as it is perceived by a great many people.
    (They even bear with my faulty English. ;-)

    It's just relatively easy to communicate in English even if it is not your mother tongue because if you have learned some basics English native speakers understand you even if you do make mistakes, and they patiently bear with your faults in idiomatic use.

    If a language develops from synthetic towards analytic then the loss of complexity is being balanced with new developments (new rules in sentence structures, newly developped idioms, etc.). Wether a language is 'easy' or 'difficult' does not necessarily correlate with being analytic or synthetic (though it may), and with Chinese you seem to have found the perfect example to proof the contrary.
     
  5. Outsider Senior Member

    Portuguese (Portugal)
    I'm usually skeptical about claims concerning language complexity, but I'd like to look for some middle ground in this thread.

    Let's not use the words "complex" and "simple", as they are loaded terms. Let's speak instead of "synthesis" versus "analysis". There does appear to be a secular trend towards less synthesis and more analysis in all Indo-European languages, as well as in other languages that have been heavily influenced by Indo-European languages (I'm thinking of Basque, for example). Whether that corresponds to a decrease in "complexity" is perhaps best not discussed here.

    But what of this trend towards analysis? Is it also noticeable, statistically, in other language families? And if so, how can it be explained?
     
  6. palomnik Senior Member

    Vietnam
    English
    I agree that there is a marked tendency toward this in some other language families too. In Bantu languages there is an ongoing tendency toward analysis, mainly in losing noun classes. However, when dealing with third world languages one must be careful about drawing conclusions; to some extent the change in the language is due to a large influx of non-native elements into the language, which often means a wholesale importation of words that don't fit the standard linguistic paradigms of the language and an increase in simplified sentence structures. Swahili has a much simpler noun structure than the Bantu languages it derived from, but this could be due to the large number of Arabic and English words it has inherited, as well as an influx of people using the language that are not native speakers, including a sizable contingent that do not speak Bantu languages.

    On a third world language that has had a relatively small influx of foreign items, such as Tamil, there have been changes over time to the language, but I don't think I would say that it has become more notably analytic.
     
  7. modus.irrealis Senior Member

    Toronto
    English, Canada
    Not specifically with case-endings, but there are certain features in Western European languages that could be analyzed as being new synthetic features.

    For example in English, I've seen it argued that English auxiliary verbs are inflected for negation (e.g. will > won't, can > can't). One reason for arguing this is that these forms occur in places where verb + not cannot occur, e.g. you can say "can't you see I'm busy" but not "can not you see I'm busy" suggesting that these are inflected forms rather than contractions. Historically of course, they were contractions, and that's probably one possibility of forming new inflections, you have a separate word that marks something, it becomes phonetically weak and a clitic, and eventually speakers reanalyze it as an inflectional affix.

    A more controversial analysis on these lines is to see the French verb as being an extremely complicated form which can be inflected to mark subject, direct object, indirect object, etc. -- basically I've seen it argued that the weak pronouns in French (je, tu, .., le, la, ...) are no longer clitics but have become inflectional affixes in Spoken French. For example you have equations between French and Latin of the type j'aime = amo and moi j'aime = ego amo, where j' seems to function more like the Latin ending than it does the independent pronoun.

    As for case-endings, I think one good candidate would be adpositions, which are often phonetically weak. I don't much about Persian, but the definite accusative marker -ra may be an example of such a case-ending (if it is a case-ending -- I'm not sure how it's analyzed).
     
  8. Joannes Senior Member

    Antwerp
    Belgian Dutch
  9. vince Senior Member

    Los Angeles, CA
    English
    Are there any examples where Modern English has become more synthetic over Middle English and Old English?

    In Mandarin/Standard Chinese, I can think of the evolution of wo (I/me) vs. wo-men (we/us), or Cantonese ngo (I/me) vs. ngo-dei (we/us) as somewhat synthetic in that Mandarin/Standard Chinese -men and Cantonese -dei are somewhat like plural-marking suffixes in a limited context. The same with the Mandarin/ Standard Chinese diminuitive "-er", which is used as a suffix.
     
    Last edited: Jun 15, 2008
  10. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    How about contractions of auxiliary verbs and their negations? Modern English is very fond of them, and I would guess that at least some of them appeared long after the Middle English period.
     
  11. darnil Junior Member

    Madrid (España - Spain)
    Castillian Spanish
    I completely agree. I've often read that less morphology implies more syntax and viceversa, and the usual way in which things go seems to be that when a certain feature disappears, another one(s) appear(s) to fill in the gap. In the process leading from Latin to Romance languages, for instance, we lost the morphological case endings, but they were replaced by word order and/or prepositions, a system which appears to be sometimes more "complex" than the old one. Are Romance languages less "complex" than Mother Latin because their case endings have been lost (Roumanian apart, of course)? Well, we added new features, like a prepositional accusative, clitics, 3rd person personal pronouns, articles... What about our verb system? The Latin one was so "simple" in comparation!
    Analysis vs. synthesis: that is the real issue.

    Maybe English just swapped the places of the personal endings in verbs from suffixal to prefixal: cf. what has been said by modus.irrealis about French:

    ("the weak pronouns in French (je, tu, .., le, la, ...) are no longer clitics but have become inflectional affixes in Spoken French. For example you have equations between French and Latin of the type j'aime = amo and moi j'aime = ego amo, where j' seems to function more like the Latin ending than it does the independent pronoun").
    I know very little (very, very little, indeed) about the history of English, but it seems to me that the trend is the same. Of course, I'm not denying the obvious tendency from synthesis to analysis in English, but sometimes it seems that it is just a question of using new categories instead of the old ones.
     
    Last edited: Jun 19, 2008
  12. malaka_malaka Junior Member

    United States
    Ingles de Nueva Jersey
    Well, as the development of English came along, sentence structure must have changed. "Can not you see I'm busy" sounds awfully archaic. But "can you not see I'm busy" is completely normal to hear. I find it interesting how even though the former is archaic, because during that period of time in English that the contraction, the change, occured, it is still normal to say it. :p
     
    Last edited: Jun 22, 2008
  13. trance0 Senior Member

    Slovene
    I would say that contractions like j`aime, je t`aime and the likes in French could only be considered the beginning of some form of new synthesis and not exactly the completed form of synthesis. In French such contractions of "weak" pronouns only occur before verbs starting with vowels and therefore one cannot apply this rule to all verbs and pronouns. In English also almost all cases of contractions can be changed so that etymology can afterwards still clearly be seen. The only example I can think of, that cannot be rephrased by using the same root, is "aren`t I". For if one tries to transform the last example, a word with a different stem has to be used, namely: "am I not".
     
  14. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    I've never learned French, but I'm pretty sure that even in the case of verbs starting with consonants, clitic pronouns can be analyzed as a part of the verb. They are just pronounced differently in this case, just like, for example, the English past tense suffix -ed is pronounced differently depending on the last sound of the verb root. This different pronunciation doesn't make the clitic pronouns somehow "more separate" from the verb, since the point isn't in the contraction, but in the fact that they are always pronounced as a single word with the verb and can't occur as standalone words. It is only an artificial orthographic convention that they are spelled separately.

    Could someone more knowledgeable about French confirm or correct this?

    This is generally not true for English contractions. For most, if not all English contractions, there are syntactic roles where they can't be expanded back into their full form without rephrasing the sentence. See the example in post #7 above; you'll easily think of many other similar examples too.
     
  15. trance0 Senior Member

    Slovene
    This is not true for French. The pronunciation of a pronoun+verb is different for each case, that is if a verb starts with a vowel like "aime", then one would pronounce the cluster "j`aime" as one word. However if a verb starts with a consonant like for example "decider" one would inherently pronounce the combination "je decide" as two words NOT one(j`decide is NOT possible!).


    I won`t argue with you about that, but read a few other examples yourself. It is possible that in the future ONLY contractions will be allowed in English and it will not be possible to use long forms (verb+not). If that happens, then I would say a new form of synthesis appeared.
     
  16. Outsider Senior Member

    Portuguese (Portugal)
    I'm not sure I would agree with what you've said about French clitics. Here are the French subject pronouns:

    je (j' before words that start with a vowel) IPA: [ʒ(ə)]
    tu [ty]
    il / elle / on; [il] / [ɛl(ə)] / [õ]
    nous [nu] ([nuz] in liaison)
    vous [vu] ([vuz] in liaison)
    ils / elles; [il] / [ɛl(ə)] ([ilz] / [ɛlz] in liaison)

    I agree that je/j' is a clitic (i.e. normally pronounced together with the word that's next to it (which is either a verb, or another personal pronoun). But I doubt very much that the other subjective pronouns are clitics. It may be so, but it doesn't seem obvious to me.

    You say that it's only an orthographic convention that they're written separately from the verb. Well, one can turn that around, and claim that it's only an orthographic convention that clitics ought to be written together with the words that support them.
     
  17. modus.irrealis Senior Member

    Toronto
    English, Canada
    About the French verb, I should emphasize that the arguments I've seen restrict themselves to informal Spoken French, rather than the more formal language used in formal contexts and reflected in writing.

    For example, je is often pronounced [ʒ] even before consonants, through the rules for dropping e muet, and in fact it becomes [ʃ] before voiceless consonants (and I believe can merge with a following s as when je sais pas is pronounced [ʃεpa]).

    The other pronouns also have various forms that aren't reflected in the orthography: tu becomes t' before vowels (e.g. t'as fait quoi?). Il is usually before consonants and often just [l] before vowels, whereas ils is before consonants and can be just [z] before vowels.

    But there are other things about the French pronouns that might make them more than clitics. For example, nothing can go separate the pronouns from the verb they are attached too -- nothing like English I only just finished it is possible in French. Things like je l'ai lu ce livre make it seem like the verb agrees with the direct object. Now I don't think the weak pronouns are affixes (at least not yet) since they just seem to be very tightly bound to the verb, but I do think it's a nice example of something in a Western European language that might be analyzed as a new synthetic feature or as something that's on its way to becoming one.
     
    Last edited: Jun 23, 2008
  18. Outsider Senior Member

    Portuguese (Portugal)
    Although it could be argued that the English sentence is itself already a reduction of I've only just finished it, in which "I" is clearly attached to the verb.
     
  19. modus.irrealis Senior Member

    Toronto
    English, Canada
    For me, even "awfully archaic" might be stretching it :D. Now I wonder if that's what happened though, that the contraction took place there originally and only the possibility of the contraction survived. I had assumed at first that these questions were simply the result of turning e.g. "you can't see" into a question, and English speakers having decided that can't now is a single word and so it had to be moved as a whole.

    This, though, is also true of the future and conditional tenses in the Romance languages, where one can see how they developed out of infinitive + to-have combinations. Although they've clearly gone farther along in development since the "uncontracted" forms are not used as far as I know -- although I believe that Portuguese can place pronouns inside these combinations.

    But all sorts of things can go between the subject pronoun in English and the verb: I, on the other hand, am staying or you always did the right thing or we sincerely hope that....
     
    Last edited: Jun 24, 2008
  20. Outsider Senior Member

    Portuguese (Portugal)
    I think that is a valid point. I believe in French you would use the disjunctive pronoun in the first sentence (moi). In the second sentence, French syntax would require that the word for "always" be placed after the verb, so you'd get Tu as toujours fait...
     
  21. berndf Moderator

    Geneva
    German (Germany)
    But this still doesn't mean that the nominative pronoun is always immediately in front of the verb. Example: "Je t'a vu" = "I saw you".
     
  22. Outsider Senior Member

    Portuguese (Portugal)
    No, but if both pronouns that precede the verb can be analysed as inflectional prefixes... Jetai vu = I saw you.
     
  23. berndf Moderator

    Geneva
    German (Germany)
    Ok, what about "Je ne le lui ai pas donné" ("I didn't give it to him"). Would you still consider "Jeneleluiaipas" an inflectional prefix?
     
  24. Outsider Senior Member

    Portuguese (Portugal)
    Excellent example, and a good question for Athaulf and Modus! ;)

    Ne is traditionally analysed as a negative adverb (although it is a clitic). But, moreover, lui is definitely separable from the verb. So, je, ne and le do not seem to be attached to the verb in that case.

    For those who do not speak Frech, here's my attempt at a breakdown of this sentence:

    Je ne le lui ai pas donné (I did not give it to him), literally "I not it to-him have not given" (remember that French has double negation):

    NOMINATIVE 1ST PERS. SG. (CLITIC) - NEGATIVE PARTICLE "NE" (CLITIC) - ACCUSATIVE 3RD. PERS. SG. MASC. (CLITIC) - DATIVE 3RD. PERS. SG. (DISJUNCTIVE PRONOUN!) - AUXILIARY VERB "TO HAVE" IN 1ST PERS. SG. PRESENT INDICATIVE - NEGATIVE PARTICLE "PAS" (NOT A CLITIC) - PAST PARTICIPLE OF MAIN VERB.
     
    Last edited: Jun 23, 2008
  25. modus.irrealis Senior Member

    Toronto
    English, Canada
    About je ne lui ai pas donné, I would imagine that an analysis as an inflected form would be something like this:

    First the inflected part would be je ne lui ai. I don't see why donné would not still be a participle. And I don't see any reason to not treat pas as a separate word: it can be used on its own without a verb and receive stress, unlike the weak pronouns, and even with a verb, it can be separated from it, e.g. il n'a vraiment pas de chance.

    So we're left with je ne lui ai, which, I believe, would be pronounced (dropping the ne) something like [ʒlɥiε], and such a short form looks more like it could be an inflectional form. But as I said before, if the French verb is inflected in this manner, then it's very complicated, being marked not only for subject, but for direct and indirect objects of various kinds -- although there are such verbs in other languages as well (Bantu languages, I believe).

    That's a good point, that lui is more independent than the subject/direct object pronouns. The only counter-point I can think of is to say that one could argue that indirect pronoun lui is a different word from the disjunctive pronoun lui, because the latter is masculine while the former is both masculine and feminine -- so it's just a coincidence that unlike moi ~ me/je, for lui ~ lui/il, the disjunctive pronoun happens to be identical in form to one of the weak pronouns. I don't know :D.
     
  26. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    Just to avoid any confusion, I don't claim absolutely any expertise on French! I read about this analysis of verbs in spoken French as agglutinative a while ago, so I just wanted to use the opportunity to summarize my understanding of the topic to see if it makes any sense.

    I guess the relevant criteria should be (a) whether the suspected clitic and the adjoining word can be stressed separately, and (b) whether the suspected clitic can occur in any other syntactic role except attached to the host word (especially whether it can occur alone as the answer to a suitably formulated question). Croatian is rich in clitics, and these criteria can be used to identify them very clearly; as far as I know, the same criteria work in Spanish too.

    Now as for this question:

    Actually, according to an article about this topic I read a while ago, but whose exact reference eludes me at the moment, the answer is yes! The author of the said article (a real linguist) claimed that in modern spoken French, or at least some of its dialects, the verb can indeed be analyzed as this large agglutinative structure. Thus, the agglutination pattern for the present tense would be: subject pronoun + (ne) + (object pronoun(s)) + verb root + (negation particle), where the elements in parentheses are optional.

    Please note that since I don't speak French, some details above might be incorrect -- I'll try finding the exact reference later.

    And to make things really interesting, the same article claimed that this structure of the French verb closely parallels that of verbs in Swahili. Also, it claimed that if analyzed this way, French has very free word order, since if you just keep clitics together with verbs and the other host words, you can reshuffle these agglutinative "super-words" quite freely. This might be great example of a language moving from an analytic, rigid syntax stage to a synthetic, free word order one.

    I should note that the author was discussing certain varieties of modern spoken French, so all of this might not hold for the standard literary language.
     
    Last edited: Jun 23, 2008
  27. trance0 Senior Member

    Slovene
    "Can`t I live without air? Can I not live without air?" As long as both options exist(even though the first one is more common) I cannot say that this is some new form of synthesis already. But it could be a beginning of one. The same goes for French. Not at present, but possible in the future.
     
  28. Outsider Senior Member

    Portuguese (Portugal)
    That actually made me think of another reason why I find those kinds of analyses you read a bit dubious (at first glance).

    Oh, it's a neat trick when you think only of sentences composed of clitic pronouns. But it clearly breaks down in sentences where the subject, or one of the objects, or all of them, are regular nouns:

    Jean embrace Marie. John hugs Mary.
    Marie embrace Jean. Mary hugs John.
    There is no free order whatsoever in these kinds of sentences. And plenty of sentences are like this in French (including colloquial French).

    Unless the author argued that "Jean" and "Marie" -- along with every other noun in the French language -- are clitics acting as ridiculously specific inflectional affixes!
     
  29. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    But it is! Notice that in this example, you can't just switch back and forth between "can not" and "can't". If you want to expand the latter into the former, you have to change the word order:

    :tick: Can't you come with us?
    :cross: Can not you come with us?
    - ungrammatical!
    :tick: Can you not come with us?

    Therefore, it's impossible to claim that "can't" is just "can not" with a couple of swallowed sounds (as it undoubtedly was when this contraction was first developing). If that were the case, you could just literally replace one with the other in any sentence and it would still be grammatical.

    What you might be confused about is that the inflected form "can't" and the analytic form with a separate word for negation exist in parallel in modern English. This doesn't mean that the short form isn't inflected; what it goes through satisfies every reasonable definition of "inflection". Therefore, it does represent a new form of synthesis, even though the older analytic form is still in use.

    I'd say such an approach could be reasonable, even though it might sound like stretching the argument. To take an example from a language that I'm more familiar with, Croatian has a complicated system of personal pronouns, in which the genitive, dative, and accusative cases of each pronoun exist in two forms: full and clitic. These forms are usually distinct, but they happen to coincide in certain persons and cases (you can see the table in Section 2.1.3.1 of this grammar book). Yet even in cases where they coincide, i.e. where the only difference between the full and clitic form is how you place the stress, the rules for choosing between the full and clitic forms and the difference in meaning depending on which one you use are completely analogous to the other ones.

    For example, if you say vidiš ga "you see him", the clitic is pronounced as a single word with the verb, but if you use the full form vidiš njega, the full accusative form njega is pronounced as a clearly distinct word. The difference in meaning (as you might have guessed) is that the latter form specially emphasizes him, i.e. "it is him that you see, not someone else". However, if you say vidiš nas "you see us", the only difference between the full and clitic form is whether you pronounce it as a single word or stress nas as a separate word. Yet, the difference in meaning is completely analogous to the previous example.

    As a non-speaker of French, of course, I don't know if it's justified to carry these conclusions over to French, but I think it shows that it at least might make sense to treat these clitic and non-clitic instances of lui as distinct words.

    Of course not. :) However, if memory serves me well, the author argued that in sentences with nouns as subjects and/or objects, the subject and object clitic pronouns still can (or, in some cases, perhaps even must?) be attached to the verb, even though they are logically redundant, which then enables the subject and object nouns to move relatively freely around the sentence. I know that things can work this way with objects (though not with subjects) in Spanish: (la) quiero a esta chica vs. a esta chica la quiero*, where the clitic la is duplicating the object and marks the verb to agree with the object (again, it's only a convention that Spanish clitics preceding the verb are spelled separately, but the ones following it are spelled together). Of course, in Spanish it's a complicated topic that can heavily depend on the dialect, and I'm sure the same is the case in French too.

    * - My Spanish is rudimentary, but if I'm not mistaken, the clitic doubling is optional with the usual SVO word order, but required if you want the object to precede the verb.

    Now it would be nice if a speaker of French clarified if any of this makes sense. I find this topic extremely interesting because if these theories are valid, then this might be a wonderful example of a language going towards a more synthetic grammar in a remarkably short period of time and with a well-documented history of this change.
     
    Last edited: Jun 24, 2008
  30. Outsider Senior Member

    Portuguese (Portugal)
    Your example from English, above, seemed persuasive to me. :cool:
     
  31. trance0 Senior Member

    Slovene
    It depends on your point of view. I suppose someone with a degree in historical linguistics would be able to give his/her proper opinion as to when languages get transformed from analytic to synthetic.
     
  32. Outsider Senior Member

    Portuguese (Portugal)
    He might have been referring to constructions such as the ones discussed in this thread. ;)
     
  33. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    Yes, that's exactly what I had in mind. Judging by what a native speaker wrote in this post, the spoken language might indeed be moving towards obligatory verb marking with subject pronouns, even when the subject is explicitly present as a noun or a noun phrase. I found another example on slide #16 of this lecture presentation.

    Now if only a native speaker could comment on how free the word order really is in colloquial French if you consider the verb together with the clitics as a single word.

    UPDATE: I found this page with a very interesting example. Could a native speaker confirm whether these sentences really sound OK in the modern spoken language?

    Yes, but in English, we're still dealing with isolated instances of contractions developing into inflections here and there. In French, it seems like the language is going through a much more radical transformation in the synthetic/agglutinative direction.
     
    Last edited: Jun 23, 2008
  34. Hulalessar Senior Member

    Andalucía
    English - England
    I think the idea that the way most of us think about French is heavily influenced by the way it is written and the fact that we know it comes from Latin is a fascinating one. In another thread I said this (slightly edited):

    For the very reason French is written down it is perhaps overlooked how far away from Latin it has in fact moved. We only need to consider the word "est" - spelled exactly as in Latin, but pronounced in a way that would surely be unrecognisable to Cicero. If the way words are pronounced has changed so much to what extent is it legitimate to consider that the grammar has only moved in the direction from synthesis to analysis?

    It is of course difficult to stand outside a language one knows so well, but here are a few thoughts about what a linguist, finding French spoken by an isolated tribe in the jungle and not knowing its relation to any other language, may come up with:

    1. Questions are formed by putting the word "εskë" at the begining of a sentence; it is shortened to "εsk" if the first word of the sentence begins with a vowel.

    2. Not finding "je" occurring in isolation, he might conclude that the verbal form of the first person singular is a front inflected form and would (I use French orthography for convenience) write "I love" as "jaime" and "I know" as "jesais".

    3. When it comes to negatives of the first person singular he would formulate this rule:
    (a) if the second letter is "e" insert "ne" after the "e" and add "pas" at the end; e.g. "jenesaispas".
    (b) in any other case insert "en" after the "j" and add "pas" at the end; e.g. jenaimepas"

    4. When it comes to direct and indirect pronouns the rules would start to become quite complicated and he would talk about inserting them into the verb and would be coming to the conclusion that French had distinct agglutinating tendencies.

    5. We know that "chien" is singular and that "chiens" is plural, but both are pronounced the same; most French nouns behave this way. We also know that French nouns rarely occur without the degree of definiteness being specified, a notable exception being when ownership is invloved. Accordingly, he may conclude that the French noun is varied by prefixes and state it thus (I again use French orthography but leave out unpronounced "s's" but use an accent to show the sound).

    unchien
    lechien
    déchien
    léchien
    monchien
    méchien
    tonchien
    téchien
    and so on

    There would be a different paradigm for feminine nouns and yet another for nouns that we know being with a vowel and peculiarities such as "mon amie" would have to be accounted for.

    6. He would note that adjectives usally come after, but also quite often come before nouns. Since he has decided that "un chien" should be written as "unchien" it will follow that he will have to write "unbonchien" and when he has done that he really has to write "unchienbon". All this is going to strengthen his opinion that French has agglutinating tendencies. He is going to have a hard time when he realises that when talking about female friends the prefix showing first person ownership will vary according whether the adjective comes before or after the noun and will also depend on whether the adjective begins with a vowel or consonant! Students trying to learn the language according to these complicated rules will give up in despair!

    7. Hearing forms such as "L'état, c'est moi" as well as "Il est sur la table" will lead him to conclude that are two forms for "is" which he may write as "se" and "e"; when liaison is taken into account he will have to double them.

    8. Indeed, when he encounters the phenonemon of liason he will conclude that many words have two forms. We do not think this way since they only have one written form and we just say that the word is pronounced in two different ways according to whether or not it is followed by a word that begins with a vowel. Once again you have a situation where if you were to start with the spoken language you would probably end up with complex rules. Just to take the plural of the definite article, you would not come up with a spelling "les" and a rule that the "s" was not pronounced before a consonant; you would have two different spellings.

    I admit the above may not be terribly well thought through, but I hope it gives an idea of what I am getting at.
     
  35. berndf Moderator

    Geneva
    German (Germany)
    This is certainly true. The fact that feminine dative lui and isolate elle are different should be sufficient proof that the masculine dative lui and isolate lui are different forms which just happen to be identical.
     
  36. Outsider Senior Member

    Portuguese (Portugal)
    Couldn't we say the same about English, though?

    adog
    thedog
    somedogs
    mydog
    mydogs
    yourdog
    yourdogs
    agooddog

    Still, good point about spelling influencing how we analyse language. We seem to be touching on the complicated semantics of the notion of "word".

    Another language that has left me wondering about this is Vietnamese. Here's an article written in Vietnamese. Notice how all syllables are written separately. Does this mean that each one of them is a word?
     
  37. berndf Moderator

    Geneva
    German (Germany)
    I would regard sentences like "L'état, c'est moi" more as an example of French having developed some topic-prominence (see http://en.wikipedia.org/wiki/Topic-prominent_language). In particular in spoken language you often hear constructs like "L'argent, j'en ai jamais assez." (note that "ne" is normally omitted in spoken French).
     
  38. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    Hardly, because of the stress. English has preserved the old Germanic tendency to have a strong stress on the root of each word, and English phonological word boundaries are intuitively very clear for this reason. In contrast, French is pronounced in such a way that these boundaries are nowhere as intuitive.

    For this reason, English has very few, if any clitics for which it would be unclear whether they constitute inflectional affixes or independent words. English contractions tend to fuse into single words as soon as the particle loses its independent stress, as evidenced by the fact that native speakers often misspell e.g. "they're" as "their".
     
  39. modus.irrealis Senior Member

    Toronto
    English, Canada
    I think so. Here the key to me seems to be that le, la, les, de, des, etc. can be separated from a following noun by an adjective. They just seem to be clitics that attach phonologically to the following word, but not affixes like the pronouns on verbs may be.

    Plus as far as I know (and I'll await for native speaker verification) they can be stressed. For example if someone says something to you and you're not sure if he said la mère or le maire (hopefully someone can come up with a better pair), I would think to clarify you would say LA mère ou LE maire? with stress on the article.

    I agree that a lot of this is about the different meaning of words, and how phonological and grammatical factors can divide up a sentence into different words, and it's clitics that seem to straddle this divide. The question about the status of the French verb has always made me wonder about what exactly distinguishes a clitic and an affix but there doesn't seem to be many clear answers out there.

    I'd also like to hear a native speaker's opinion on some of those sentences. One of my first posts here was about having both a noun phrase direct object and the pronoun on the verb. Since then, the reading I've done suggest that these sort of dislocations have to do with topic-status and so on, like berndf mentioned and it seems strange to see so much material be dislocated in some of those sentences.

    About Outsider's example, I guess it might be theoretically possible to say something like Marie, il l'embrasse, Jean with OVS order, but I wonder if anyone actually would say that.

    In fact, it's very similar to Greek though where you can also dislocate subjects, objects, etc. to the left or to the right and use a clitic pronoun on the verb, but there there's also the possibility of simply rearranging the sentence order without using any clitic pronouns, where the noun-phrase is then more closely tied to the verb and has a different status in terms of topic and information-structure. But I'm not sure this reflects anything on the status of clitic pronouns in Greek, which like French are very tightly bound to the verb (in terms of order and separability), but it's unlikely that they're affixes because those very same pronouns occur with nouns and adverbs/prepositions.

    Yes, I'm now convinced that it is better to treat them as different words.
     
    Last edited: Jun 24, 2008
  40. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    You might find the discussion in this article interesting:

    http://www.stanford.edu/~zwicky/ZPCliticsInfl.pdf
     
  41. Alijsh Senior Member

    Tehran
    Persian - Iran
    rā is a postposition.

    You're right, adpositions are good candidates. The evolution of Persian has been very interesting. New Persian (and Middle Persian) does not have noun declension any longer but marks cases and as a result, it continues to have free word order. For example, the ablative case is marked with the preposition az e.g. pedar az doxtar porsid (the father asked the daughter), which has six permutations (3 x 2 x 1):

    pedar az doxtar porsid (normal order: SOV)
    pedar porsid az doxtar
    az doxtar porsid pedar (major emphasis is on the object)
    az doxtar pedar porsid
    porsid pedar az doxtar (major emphasis is on the verb)
    porsid az doxtar pedar

    In English, if you change the order e.g. "the daughter asked the father" the meaning completely changes but it's not so in Persian.
     
    Last edited: Jun 24, 2008
  42. Outsider Senior Member

    Portuguese (Portugal)
    That's hard to believe, because French has fixed stress, while in modern English stress is unpredictable.

    Well, that's another problem. Any word can be stressed, if you really want to. In English, words like "the" and "a" are usually unstressed, but if you happen to have some reason to emphasize them, you can always say "thee" and "ay".
     
  43. berndf Moderator

    Geneva
    German (Germany)
    Stress is all but irrelevant in French. Native speaker largely ignore it. This is why it so hard for French speakers to learn the stress pattern of a foreign language. French rearranges the sentence structure using semantically unnecessary relative clauses to convey emphasis (see http://forum.wordreference.com/showpost.php?p=5279726&postcount=37).
     
  44. Outsider Senior Member

    Portuguese (Portugal)
    Pronounce French words with the stress in the wrong place, and I suspect you will sound very strange to native speakers.

    I think it's more likely that they have trouble with stress in other languages because they're used to a very rigid stress pattern: always stress the last syllable that does not contain a schwa. (A good pattern to mark word boundaries, by the way!) In languages with the same last-syllable-stress pattern as French, they probably don't have much problem pronouncing words right.

    Many languages do, if I do say so myself. Where are you going with that?
     
    Last edited: Jun 24, 2008
  45. berndf Moderator

    Geneva
    German (Germany)
    I think you have to distinguish between things which make a language sound natural and those which carry actual information. And stress does not contribute to the meaning of a sentence in French. You have similar issues with vowel lengths and quality. E.g. French doesn't distinguish between open and closed "i" and between long and short vowels. I have repeatedly heard native speakers complain that they cannot hear any difference whatsoever between the English words "to live" and "to leave". I am almost sure (I can research that on Thursday when I am back in Geneva) that French speakers have a hard time hearing the difference between "arithmetic" (noun) and "arithmetic" (adjective).

    Besides, there are entire sentence, like "Qu'est-ce qu'il y a?" which sound perfectly natural, if you stress only the final syllable (though you would normally stress "Qu'est" as well, but not any of the words in the middle).

    This is a question of shades of gray, not of black and white. If stress is important in a language (like in all Germanic languages) you would do more with stess and less with sentence structure (in spoken language, but that is what we are takling about).
     
    Last edited: Jun 24, 2008
  46. Outsider Senior Member

    Portuguese (Portugal)
    Of a sentence? Do you mean a word?

    I cannot agree that "whatever is not phonemic is not important". Non-phonemic features can make the difference between a good pronunciation, and a foreign-sounding (and in extreme cases even difficult to understand) one. Non-phonemic variations are usually what distinguishes the various accents within a language. You don't speak a language properly until you get those details right.
     
  47. berndf Moderator

    Geneva
    German (Germany)
    If you re-read your post #42, Athaulf's argument relies on the attention English speakers pay to stress pattern in both, word and sentence. This you failed to address in your response. In Germanic languages, stress is not only an issue of accent and proper pronunciation. Shifts in stress significantly modify the meaning of sentences.
     
  48. Outsider Senior Member

    Portuguese (Portugal)
    Really?! I don't see any mention of sentence stress in Athaulf's post.

    I'll wait for his confirmation.
     
  49. berndf Moderator

    Geneva
    German (Germany)
    My apologies for my unclear wording. Correct would have been "Athaulf's argument relies on the attention English speakers pay to stress pattern". The addition "in both, word and sentence" is my own to indicate that I don't think the difference is important to the argument (that English pays attention to stress while French doesn't).

    A good idea, let's wait for Athaulf's feedback.:)
     
    Last edited: Jun 24, 2008
  50. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    I wasn't saying anything about sentence stress and intonation. My point was that the Germanic word-root stress in English gives native speakers, as well as non-native listeners, a pretty clear idea of phonological word boundaries, with a very few exceptions. In contrast, I have the impression that in French, whole phrases or even sentences can be pronounced as a more or less uniform stream of syllables, with the stress coming only on the last or penultimate syllable of the last word. Again, as a non-speaker of French, I might be wrong about this -- my ear is very badly tuned for listening to French, since my native language has dramatically different phonology and prosody.

    Therefore, my impression is that if a linguist encountered two isolated jungle tribes speaking modern English and French, he would immediately notice the very prominent root stress in English, and likely use it as a basis for separating words from the very start. In contrast, he would have a harder time discerning the word boundaries in French, and might easily end up with an analysis along the lines of post #34. (Although probably not that drastically -- I'd say adjectives can move around much too freely to be analyzed as bound morphemes.)


    An interesting test of my hypothesis has just occurred to me: we could look at how native first grade kids misspell things. If they tend to misspell certain words together, it's a pretty good indication that an outside linguist would also likely analyze the same utterances as single agglutinated words. For example, in Croatian, not just kids, but even adults often misspell verb negation particles together with the verb -- and indeed, it's a mere artificial convention that these particles are considered as separate words; Czechs, for example, treat them as verb prefixes.

    So, another question for native French speakers: do French primary school kids have a tendency to lump together these suspected agglutinative structures in writing?
     

Share This Page