WHY ADJECTIVES?

Lameen has an interesting post about adjectives at Jabal al-Lughat. He points out that “often, a concept expressed using an adjective in one language is expressed only by a verb or a noun in another”; for instance, “there is no adjective ‘happy’ in Kwarandzyəy; instead, you use a verb, yəfṛəħ ‘be happy, rejoice’. And to say ‘the happy people’, you say ‘the people who are happy/have rejoiced’: bạ γ i-ba-yəfṛəħ person who they-PF-happy.” And yet there are few if any languages without adjectives. He ends up:

So clearly people can do without some adjectives, and clearly the behaviour of adjectives tends to be very similar to the behaviour of some other word class. Why not do without them altogether? It would be easy enough to construct a language where no morphological or syntactic tests could distinguish adjectives from verbs, or from nouns. So if practically every language does take the trouble to distinguish them, there must be some pretty powerful cognitive motivation for it – and some pretty powerful historical tendencies acting to separate adjectives from verbs and/or nouns.

Interesting, no?

Comments

  1. marie-lucie says

    Not all languages need to have exactly the same parts of speech to express the same concepts. “Adjective” in particular is not found everywhere as a separate, distinctive part of speech. Many native languages of the American Northwest (US and Canada along the North Pacific) express the concepts expressed by “adjectives” through intransitive verbs, just as in Lameen’s example, or (in the case of very frequent basic concepts such as “large” or “small”, for instance), as a kind of adverb or as a prefix to a noun. In the Salishan language family (spoken from the Columbia River northwards into British Columbia), there has been a continuing controversy among linguists as to whether the languages actually distinguish nouns from verbs (not in terms of meanings, but of behaviour in the sentence), since for instance one could say “I woman” just like “I speak” (but there are usually other tests to tell a verb from a noun). Many of these languages do not have a separate category of “adverbs” either, for instance one doesn’t say literally “you did well” but “what you did was good”.
    In general, adjectives and adverbs as grammatical categories are not as basic as nouns and verbs: adjectives concentrate on a property of a concept represented by a noun, adverbs on a property of a concept represented by a verb, and they “piggyback” on these categories, occurring next to them in a sentence but often optional and able to be omitted without destroying the sentence: the (red) book, to run (fast), the (brown) fox jumped (quickly) over the (fat, lazy) dog. A language with adjectives distinct from verbs requires a “copula” (the otherwise meaningless verb “to be”) in order to consider the property conveyed by an adjective separately from the noun which has that property: the book is red, but there is no need for a copula if the adjective works like a verb, literally the book reds.

  2. michael farris says

    marie-lucie says pretty much what I would, there’s no real evidence that the statement “practically every language does take the trouble to distinguish them”. Words that are translated as adjectives in English are at most a sub-class of nominals or verbals in very many languages.

  3. According to George Kennedy, classical Chinese did not distinguish verbs and adjective except by placement in the sentence. Later linguists (Harbsmeier, Cikoski, I think) found that there was a distinction in some cases.
    I also found a book in Taiwan explaining what we call propositions as coverbs. I have heard this: the Chinese equivalent of “I will with you [gen ni]”.

  4. What ML said. There are two kinds of classical Chinese sentence, stative and verbal. “The dog runs” and “The dog is brown” are both verbal sentences — no copula. Literally “Dog run” and “Dog brown”. But “The dog is an animal” needs a copula, the equivalent of “Dog animal is”.
    I totally fell in love with classical Chinese right from the beginning. No obligatory inflections, articles, genders, or plurals, and and in most cases, no copulae. It was so beautifully economical, like a computer language.

  5. Funny; 18-19c. language theorists were busy trying to reduce verbs to copulae + adjectives.

  6. And the copula isn’t exactly a copula; it’s just a sentence-final, clause-final, or phrase-final particle that in certain structures functions as a copula.

  7. Also, as a syllable-tonal language Chinese doesn’t have sentence-tone (e.g. the questioning inflection), so they pronounce many of their punctuation marks.

  8. A language with adjectives distinct from verbs requires a “copula” (the otherwise meaningless verb “to be”) in order to consider the property conveyed by an adjective separately from the noun which has that property: the book is red, but there is no need for a copula if the adjective works like a verb, literally the book reds.

    marice-lucie, how could adjectives come to “work like verbs” in a context where it is suggested that in general … adjectives concentrate on a property of a concept represented by a noun, … and ‘piggyback’ on [that category]?
    Russian has no “copula”, though it has adjectives distinct from verbs. Perhaps that has something to do with the fact that there has never been a widely-known Russian Philosopher of Being. The major work of such a person, ? и время, would probably not have wowed the literate community. I bet Heidegger goes off in Russia like a damp squib.
    Come to think of it, copulation itself requires no copula, not even braces (suspenders). You just put two people next to each other, and off they go. I can cite Burroughs in support of this: “Me Tarzan, you Jane”.
    “To be” is not meaningless, in English say, since it is needed to bring things together that are conceived of as ontologically inert. I wonder whether evidence exists that a particular kind of “static world-view” is associated, in a statistically significant way, with languages in which a copula plays a prominent part. This could take the form of a comparative frequency chart of schools of ontology, and the languages in which the philosophers involved publish.

  9. Russian has a copula, it’s just not stated explicitly in the present tense. “On — horoshy student” vs. “On byl horoshy student.” In writing it’s even acknowledged that there /should/ be something there.

  10. In writing it’s even acknowledged that there /should/ be something there.

    You mean one says “on horoshy student”, but writes “on /zdyes nado buit/ horoshy student”?
    How essential can /buit/ be, when it’s not needed? Your comment nicely illustrates my speculation about a correlation between world view and language. You want to make actually spoken Russian fit your favorite ontology, which is probably that of your mother tongue. You want to fill the ontological hole with a word.
    But you can’t derive an ought from nought.

  11. Furthermore, the past tense forms of /buit/ are structurally more like adjectives than conjugated verbs (in other languages), and this is true of all Russian verbs. But there’s no copula needed to link up /on/ and /buil/, no more than /on/ and /govoril/. It suffices to juxtapose them.

  12. Ahh, the adjective flamewar continues. How I love it so.

  13. Actually, I’m just waiting for someone to reach down for the wallet on the sidewalk – at which point I will jerk on my string.

  14. It is usually said that Japanese adjectives are actually verbals. They can be put in past tense, etc.
    On the other hand, historically the past tense of Japanese adjectives is derived by using a verb. E.g., akakatta (‘was red’) was formed from akaku atta, where atta is the past tense of the verb aru ‘to be’. Aru also occurs in other forms of the adjectival “conjugation” (concessive akaku wa aru, negative akaku nai).

  15. mollymooly says

    Supplementary question: which language is the most adjective-y? English uses adjectives for emotions and physical conditions (though not illnesses) where e.g. French uses have + noun.

  16. I agree that “Words that are translated as adjectives in English are… a sub-class of nominals or verbals in very many languages”. Even in English they share some syntactic characteristics (eg the predication construction) with nominals. But I’d replace “at most” with “at least”; the point I find more interesting is the fact that, in just about every case examined in detail, there still turn out to be morphosyntactic criteria that force you to define at least one class of adjectives (sub- or otherwise.) If you can come up with a convincing example of a language where no such criteria exist, then you will have refuted Dixon’s conjecture – but this would surprise me, given that such criteria can be identified even in Korean or Wolof or Coptic. (Japanese is easy – -i adjectives have different polite forms than verbs (akai desu / *akakimasu), and only -i adjectives, not verbs, can form adverbs in -ku. Not so sure about Classical Chinese offhand.)
    I know very little about languages of the Pacific Northwest, and would be interested to hear more – but the predicative construction cannot be taken as the defining criterion for parts of speech. In fact, Stassen claims that every language handles simple adjectival predication the same way as either nominal, verbal, or locational predication (though if negation is included Siwi is an exception.) In predicative contexts, Kwarandzyey could be said to express all adjectives by intransitive verbs (apart from a couple of recent loans). What justifies a distinction between adjectives and verbs there is their behaviour in attributive contexts (“the red car”).

  17. I would argue Russia does have an underlying copula – it’s just not explicit. If you need to be explicit you can say “On yest’ khoroshim studentom”, or even “Ya yest’!” – I am!. In many cases “eto” (this) functions as a copula, e.g. “Pytor – eto tot paren’ o kotorom ya tebe rasskazival.” – “Peter is the guy I was telling you about.” “Sein und Zeit” is Бытие и время in Russian – it’s not difficult translating these German concepts into Russian, unlike Japanese or Chinese.
    No one has brought up Japanese adjectives yet – they also are arguably not adjectives in the English sense.

  18. I retract that last sentence, I missed bathrobe’s post entirely. I see they have been dealt with.

  19. Furthermore, the past tense forms of /buit/ are structurally more like adjectives than conjugated verbs (in other languages), and this is true of all Russian verbs.
    True – they are past participles. And in all the Western Slavic languages you still need to use the copula with them – whether independently as in Serbian – “bio sam”, or conjoined as in Polish “bylem”. In Russian “был есмь” long ago got shortened just to “был”, but to the Russian mind I think it functions as a verb today.
    And if Russian really has no underlying copula it’s hard to explain the existence of this Tsvetaeva poem:
    Я — есмь. Ты — будешь. Между нами — бездна.
    Я пью. Ты жаждешь. Сговориться — тщетно.
    Нас десять лет, нас сто тысячелетий
    Разъединяют.— Бог мостов не строит.
    Будь!— это заповедь моя. Дай — мимо
    Пройти, дыханьем не нарушив роста.
    Я — есмь. Ты будешь. Через десять весен
    Ты скажешь: — есмь!— а я скажу: — когда-то…

  20. Vanya, thank you for stating that far more correctly and competently than I ever could.

  21. marie-lucie says

    the predicative construction cannot be taken as the defining criterion for parts of speech
    I agree with everything Lameen said, and the above sentence agrees with what I was implying – you have to consider all of the contexts in which a word is used in order to establish that it belongs to a given category. It is the attributive construction (where the adjective is an optional attribute of the noun) which is the typical one for identifying an adjective, even if in a particular language the adjective can also be used predicatively and take verbal morphology in that context.
    On using a verb “to be”:
    A part of speech which has very little meaning and is predictable in specific contexts can often be omitted, for instance “that” is usually omitted in casual speech in sentences like I know (that) I am right. In Latin mottoes and other traditional, often repeated sayings the verb meaning “to be” is usually omitted, as in ars longa, vita brevis instead of the “correct” full sentences ars longa est, vita brevis est (“art is long, life is short”). The usual omission of the verb in equivalent Russian sentences follows a similar logic.

  22. Doug Sundseth says

    Grumbly Stu: “But you can’t derive an ought from nought.”
    Correct. That takes integration.

    Which comment is somewhere in the range of profoundly meaningless to meaninglessly profound. Almost zen-like, what? 😎

  23. There’s a series of linguistics monographs on the word “to be” in various languages, published by Reidel in Dordrecht and edited by Verhaar.

  24. A bit of bibliography:
    Dixon, Robert M. W. 1977. Where have all the adjectives gone? Studies in Language 1.1.19-80.
    In addition to languages with no syntactic category Adjective distinct from Noun and Verb, Dixon surveys languages with very small numbers of adjectives.

  25. There was also a language of Tlön that consisted only of adjectives.

  26. “But you can’t derive an ought from nought.” “Which comment is somewhere in the range of profoundly meaningless to meaninglessly profound. Almost zen-like, what?”
    Gumby-Stu likes Heidegger, that’s what. If he’s that good with German and is that sympathetic, maybe we can get a translation out of him that makes Heidegger crystal clear once and for all. Kind of like what Kaufmann did for Nietzsche.
    the word “to be” in various languages
    I have never been able to discover a word for “to be” in Arabic. Useful substitutes are fee “there is” and kahn “there was” and bukara fee “tomorrow there is”. No wonder Emerson’s link says the Arabic/Sufi concept of “being” is completely different from the West. Maybe that Wharf dude was onto something.

  27. Also, as a syllable-tonal language Chinese doesn’t have sentence-tone
    You mean sentence intonation, John? I believe Chinese does have it. See here, here, here, and many other places. The tones of a tonal language do not preclude its having prosodic pitch patterns, any more than the word stresses of English preclude its having prosodic stress patterns.
    Whether Classical Chinese had sentence intonation would be hard to investigate, I suppose; but is there any reason to think it did not?
    And did you mean preposition earlier, when you wrote proposition?

  28. David Marjanović says

    Russian has a copula, it’s just not stated explicitly in the present tense. “On — horoshy student” vs. “On byl horoshy student.” In writing it’s even acknowledged that there /should/ be something there.

    Usually it isn’t, but it does seem to be obligatory for sentences like “A is B, and B C” to become “A B, and B — C” even when the equivalence of B and C isn’t meant to be specially emphasized.

    According to George Kennedy, classical Chinese did not distinguish verbs and adjective except by placement in the sentence. Later linguists (Harbsmeier, Cikoski, I think) found that there was a distinction in some cases.

    Does Chinese distinguish anything by any other means than word order? (Apart from the plural suffix -men, I mean, which is almost only used with personal pronouns.) For example, word order is the only way to tell that Chinese is a nominative-accusative language (rather than an ergative-absolutive one like Tibetan and many other Sino-Tibetan languages).

    I also found a book in Taiwan explaining what we call propositions as coverbs. I have heard this: the Chinese equivalent of “I will with you [gen ni]”.

    Also, at first glance, Chinese seems to have both pre- and postpositions.
    I wrap my mind around this by interpreting the prepositions as verbs and the postpositions as nounds. So, your example sentence is best considered “I will accompany you”; there is no “with”, there’s only “accompanying”, just like there’s no “to”, only a “go”, and no “by car” but only “sitting car” (“sitting in car” would be redundant by Chinese measures). Note my silent assumption that Chinese has present participles that just happen to look exactly like all other verb forms…

    I totally fell in love with classical Chinese right from the beginning. […] It was so beautifully economical, like a computer language.

    That’s deliberate. Classical Chinese is a highly abbreviated style used for cramming as much information as possible into as few characters as possible. It’s not probable that anyone ever really spoke like that.

    Also, as a syllable-tonal language Chinese doesn’t have sentence-tone (e.g. the questioning inflection)

    Scarily for the tone-deaf learner, it actually does. The tones are modulations on top of that. Marking questions by intonation alone appears to be astonishingly common… there was a post on the blog of http://www.pinyin.info about it years ago.

    Russian has no “copula”, though it has adjectives distinct from verbs. Perhaps that has something to do with the fact that there has never been a widely-known Russian Philosopher of Being.

    I was once told by a linguist that all the leading authorities on aspect in linguistics are Germans; if true, that must be precisely because aspect is an utterly exotic phenomenon for native German speakers…

    Come to think of it, copulation itself requires no copula, not even braces (suspenders). You just put two people next to each other, and off they go. I can cite Burroughs in support of this: “Me Tarzan, you Jane”.

    😀 😀 😀 😀 😀 😀 😀 😀 😀 😀 😀 😀 😀
    ROTFL! My day is saved. 😀

    But you can’t derive an ought from nought.

    What I just said.

    Correct. That takes integration.

    Good point. Very good point.

  29. David Marjanović says

    that Wharf dude

    Sapir and Whorf.

  30. There’s an extensive literature about the translation of Greek “being” to Arab equivalents to Latin. The gist is that Arabic required distinguishing essence-being and existence-being, and that this distinction entered Latin via Arabic, making scholastic philosophy significantly different than Greek philosophy.
    I spent some time on the question decades ago, and don’t vouch either for the accuracy of my recollection or the accuracy of the articles I was reading.

  31. Yes, John Emerson, you are right about the Arabic, with words derived from the k-w-n root referring to essence-being and words derived from the w-j-d root referring to existence-being.
    And Nijma, of course there is a word for “to be” in Arabic–you just don’t use it in the present-tense, predicative sense, but you use it in the past, the subjunctive, and the future: kaana/yakuun.

  32. That’s interesting about Chinese sentence-tone. I never got beyond trying unsuccessfully to control my compulsion to end my declarative sentence in the fourth tone, and my interrogative sentences in the ?second or ?third tone, thus grossly mispronouncing whichever syllable it was.

  33. كان
    kahn, ma kahn “there was or there was not” was how the fairy tales expressed “once upon a time”. No wonder they were so confused about existence. I always had the sense this was being used more like a modal, although I seem to recall this can be conjugated in MSA.
    كون
    form of كان?
    “words derived from the w-j-d root referring to existence-being”
    وجد-wahada meaning “one”? Wehr says this means “to be alone, unique, singular, unmatched…” also figures in declaration of monotheistic deity.
    At first I though this meant it’s like the Spanish “to be” ser/estar,

    I am (soy) Nijma.
    I am (estoy) happy.

    but this وجد seems to be a different word altogether.
        

  34. Further to Arnold Zwicky’s note:
    According to Bill Foley’s (1991) The Yimas Language of New Guinea (Stanford), there are “only three true unambiguous” adjectives: kpa ‘big’, yua ‘good’, and ma ‘other’. These are the only words denoting qualities of objects that may occur uninflected and tightly bound to a following noun. The rest are either “adjectival verbs” obligatorily inflected for tense, or “adjectival nouns” that appear as the object of the postposition kantk.
    Yimas is a wonderfully complex polysynthetic Papuan language with noun classes determined partly by semantics and partly by phonology. No wonder its 200 or speakers are giving it up for Tok Pisin!

  35. Charles Perry says

    W-j-d, to find: wujida “be found”, wujuud “being there.”
    On the phone, you ask whether Mr. So-and-so is there by saying, “Al-sayyid Fulan mawjuud?” “Is Mr. So-and-So found?”

  36. Clarifying Zwicky’s comment, and Dixon’s claims: In 1977 Dixon believed there were languages with no syntactic category Adjective distinct from Noun and Verb, but he changed his mind later:
    “In an earlier study (Dixon 1977a: 20, 1982:2) I opined that ‘some languages have no adjective class at all.’ The present chapter – building on a further quarter-century of research – puts forward the hypothesis that an adjective class can be recognised for every language, although sometimes the criteria for distinguishing adjectives from nouns, or adjectives from verbs, are rather subtle.” (Dixon 2004:12)…
    “It can sometimes be a tricky matter finding criteria to distinguish ‘verb-like’ adjectives from verbs, or ‘noun-like’ adjectives from nouns. I believe that in every language which is studied in detail, such criteria can be found. (Dixon 2004:45)
    http://books.google.co.uk/books?id=TqdGIRencmUC&pg=PP1&dq=dixon+adjectives&ei=nRVpSdz0Dp-aMpWToKsF&client=firefox-a

  37. وجد found
    Yes the “Al-sayyid Fulan mawjuud?” telephone inquiry is very familiar. I think I’ve heard “honak” (there) as well.

  38. I think I managed to get on the wrong page of the Wehr. وجد has a range of meanings from find to strong emotion, to create, accomplish, be existent.. the telephone inquiry variety listed is موجود.

  39. For classical Chinese, IIRC, the conclusion was that a large proportion of adjectives were indistinguishable from verbs, but not all. Can’t dig up the book, alas.

  40. marie-lucie says

    According to Bill Foley’s (1991) The Yimas Language of New Guinea (Stanford), there are “only three true unambiguous” adjectives: kpa ‘big’, yua ‘good’, and ma ‘other’. These are the only words denoting qualities of objects that may occur uninflected and tightly bound to a following noun.
    If those “adjectives” are the only ones, and they are both uninflected and tightly bound to a noun, they seem to be prefixes or proclitics rather than full words, as do similar forms in some of the Pacific Northwest languages I was referring to earlier.

  41. David Marjanović says

    I never got beyond trying unsuccessfully to control my compulsion to end my declarative sentence in the fourth tone

    I came back from China believing that the “dry cups” exclamation associated to drinking alcohol, Japanese kanpai, was gānbèi. It isn’t. It’s gānbēi — I had mistaken the exclamatory intonation for the fourth tone.

    If those “adjectives” are the only ones, and they are both uninflected and tightly bound to a noun, they seem to be prefixes or proclitics rather than full words, as do similar forms in

    …German. Either German has a class of indeclinable adjectives (which have to occur immediately in front of their noun and are then written as prefixes), or they are prefixes. I’ll elaborate later if someone is interested.

  42. the “dry cups” exclamation associated to drinking alcohol
    God, how I came to dread that when I was teaching in Taiwan. I don’t mind drinking, but I like to do it at my own pace. The pressure to constantly be emptying one’s alcohol container takes the fun out of it (until it puts too much fun into it, which you can’t allow it to do if you’re a professor in a foreign land with a social position to maintain and a job you don’t want to lose).

  43. There is a class of graphs in Chinese which have two pronunciations/meanings, one in the fourth tone, and one in some other tone, where the fourth-tone reading has in some way a more active meaning. And twenty years ago I could have given you several example pairs off the top of my head.

  44. Either German has a class of indeclinable adjectives (which have to occur immediately in front of their noun and are then written as prefixes), or they are prefixes

    The “prefix” examples will be the standard ones, I assume, such as Großmut. But what would an “indeclinable adjective” be that “is written as a prefix”?? You must be using “adjective” here in a way I find it hard to understand. Do you mean something like Stups in Stupsnase? But there is no adjective *stups… , so there is no such indeclinable adjective.
    But wait – if there is no adjective *stups…, then what is “Stups” in Stupsnase? It’s not a noun or associated with / derived from a noun, so there’s nothing else to call it but “an adjective”. Is this the line of reasoning? How about calling it an adverb, since it describes the way in which the nose goes into its curve?
    Fehlgriff is another example, and perhaps Außenseiter. Mißmut is something different, oder?

  45. I take it back, since I just remembered “stupsig”. maybe Fehlgriff is an example of what you mean, David – though the whole “indeclinable adjective prefixed to a noun” business still seems to me rather gekünstelt

  46. akaku wa aru, *akakimasu
    i’m not sure, but aren’t the correct forms akaku de aru are red, akaku natteiru becoming red, no? but i don’t know anything about old sounding Japanese and NHK taiga dramas were the most difficult thing to watch
    can’t say i’ve mastered all things honorific in Japanese too, so i’d be glad to learn what forms are those and their usage if those are correct
    On using a verb “to be”:
    i’m sure our usage of be is the most difficult to learn for a foreigner
    ‘he would still be there’ sounds like ter(s/he) tend(there) baij l baikh baisan baikhgui yu (be -baikh), in informal speech, transformations of be only, i never could think of its equivalent in Russian or English, in Japanese – maybe ‘kare ga asoko ni ima de mo itteita kamo shirimashen (6 words to convey the same meaning)

  47. marie-lucie says

    David M: Either German has a class of indeclinable adjectives…, or they are prefixes. I’ll elaborate later if someone is interested.
    Please, David, do elaborate.
    I think that a word like Großmutter must be a type of compound, where the adjective remains in its bare form and does not take inflection. But are there words with adjectival meaning which only occur uninflected? and in that case, wouldn’t they be better described as adverbs?

  48. Exactly the point I made above, marie-lucie!

  49. marie-lucie says

    Yes indeed, Grumbly, but sometimes it is good to have the same point made in slightly different ways.

  50. Especially for those of us who aren’t that familiar with the languages and want to understand the points being made!

  51. “For classical Chinese, IIRC, the conclusion was that a large proportion of adjectives were indistinguishable from verbs, but not all. Can’t dig up the book, alas.”
    Not only adjectives. Most nouns act like verbs when the subject is the thing the verb refers to. There may be exceptions that I can recall. Anyway, they take nominalizers just the same as any other verb. They do not need to be followed by ‘ye3’ at the end of the phrase either – ‘ye3’ is not really a copula particle.
    In Mandarin I don’t think there are any adjectives that don’t function exactly like verbs under all conditions. In fact I haven’t seen “adjective” as a part of speech in any grammar of Mandarin.

  52. “What justifies a distinction between adjectives and verbs there is their behaviour in attributive contexts (“the red car”).”
    That test doesn’t apply in any variety of Chinese I know of. It is not even restricted to intransitive verbs So you get:
    1) “red car”
    2) “that from Beijing come person”
    3) “eat cheese that foreigner”
    An objection may be made that the subordinating particle ‘de’ normally goes before “person” in 3). Well, it can but it doesn’t have to, and it can equally well go in front of the head being modified in 1) – so that’s not a distinction.

  53. Is someone nursing trodden toes again? Is it I who is supposed to have put his foot down?
    All I said was that I had made the same point about adverb vs. adjective. I did not say “I made that point first, it’s my idea, now nobody else is allowed to have that idea or a similar one” – which is how marie-lucie and LH seem to have taken my comment. It merely expressed surprise (and a certain layman’s gratification) that I had had an idea which a linguist, namely marie-lucie, also had.
    As I never tire of reminding people, I never tire of reminding people that I am not a linguist.

  54. Most nouns act like verbs when the subject is the thing the verb refers to

    Jim, I have inspected that sentence from as many angles as I could think of, without being able to understand it. Could you please explain?

  55. marie-lucie says

    Stu,
    I think I understood you as you just indicated. I am gratified when non-linguists (eg my students) do have the same idea as I have (often the “idea” is not really mine but general knowledge or belief among linguists). You may not be a linguist (= have received formal training in linguistics), but you certainly could be (ie you seem to have the type of analytical mind that characterizes linguists). (I do not mean to imply that non-linguistically oriented people do not have analytical minds, only that they might not focus their analytical skills or propensities on the same things).

  56. which is how marie-lucie and LH seem to have taken my comment
    Not at all! I was just seconding m-l’s point that it can be good to have things stated more than once. No toes were trodden—please reset the safety catch on your phaser!

  57. michael farris says

    “I never tire of reminding people that I am not a linguist”
    Okay, it’s one thing to be lucky and blessed by fate, it’s something altogether different to rub it in.

  58. David Marjanović says

    Here goes. German has word-root accent, which means that, in compound nouns, each component keeps its own stress pattern, and the first component gets the strongest stress. Like in English, more or less. Now, off the top of my head, I can think of two exceptions: the prefixes Riesen-, which is about the same as the English adjective “giant”, and Scheiß-, which is a very productive way of expressing anger about the rest of the compound noun. The stress can go on Riesen- or on the next component, and it never goes on Scheiß-. For example, Riesenreich stressed on the first syllable is as ambiguous as its literal translation “giant empire”: it can designate a gigantic empire as well as an empire of giants (built/ruled/populated by giants). If the last syllable is stressed instead, only the first sense is possible. Adjectives in front of nouns are not stressed (again like the English default: compare “a black bird” and “a blackbird”). So I wonder…
    There is an adjective riesig “huge”. There is no adjective “shitty” except in a couple of idiolects; instead, the past participle geschissen is used.
    Marie-Lucie has hinted at a complication: firstly, in German, “to be” requires the adverb, not the adjective; secondly, adverbs are zero-marked, so they look like English adjectives and are used as the dictionary form of adjectives (that’s what I just did above with riesig; the actual adjective, nominative singular, is riesiger/riesige/riesiges with the indefinite article and riesige with the definite one; same for geschissen). The answer to “how is it” questions is an adverb in German.
    The other phenomenon I was thinking of is that German is very reluctant to newly derive adjectives from nouns. Take the scientific names Mammalia and Aves: English simply reaches into Latin and comes up with mammalian and avian; German gets a “does not compute” reaction and ends up rephrasing the whole sentence, usually making compound nouns with the prefixes Säugetier- or Vogel-, and those compounds are all stressed on the first syllable. This applies to all names that are deemed sufficiently exotic. For example, there is no German adjective to Zulu-; all you can do is use Zulu- as part of a compound noun which, again, gets stressed on the first syllable. The German version of the computer game Civilization II didn’t know about this fact, acted as if German had an adjective zulu, and auto-generated things like *das zulue Reich; and that’s simply wrong. You have only one option which you can spell in two ways: das Zulu-Reich or das Zulureich.
    I’m an autodidact in linguistics, so it’s entirely possible that grammarians of German solved this problem decades ago and I just haven’t seen the relevant literature. So, just my 0.02 €.

  59. David Marjanović says

    I should have written *zulu with an asterisk.

    in German, “to be” requires the adverb, not the adjective

    Except, that is, the Highest Alemannic dialects, like what’s spoken in Zürich (not Berne). Pretty much any generalization about German is wrong.

  60. marie-lucie says

    David,
    I have not consulted a German grammar lately, but it seems to me that all the compound words you describe are made up of noun + noun, not adjective + noun (although such compounds exist, like Grossmutter as mentioned above). There is no reason to think that just because English favours adjective – noun sequences (not compounds), that the German translations of such sequences should have the same structure.
    Now that I see your examples, I have to say that there is no reason to consider forms like riesig as adverbs rather than adjectives just because they are not inflected after the verb meaning “to be”: you can just say that adjectives are inflected when they are part of a noun phrase (eg a group article – noun with an optional adjective between the two) but not otherwise, as when they are on their own, after the verb (or part of a compound, as in Grossmutter). Otherwise you would have to say that “adverbs” like riesig are inflected when they act as adjectives! Agreeing with nouns is not part of the definition of an adverb, which is an optional part of a verb phrase and therefore has no noun to agree with. The fact that some adjectives can be used as adverbs (as in English) does not contradict what I just said: it is its use in a particular sentence that determines which part of speech a particular word is playing.

  61. marie-lucie says

    which part of speech a particular word is playing.
    I think that I mixed metaphors here: I meant which part of speech a particular word belongs to.
    For instance, there are numerous cases of the “same” English word being sometimes a noun and sometimes a verb, for instance “cook”, and which one it is is made clear by the context: there are actually two words, not just one (as you can see from separate dictionary entries).

  62. I have to say that there is no reason to consider forms like riesig as adverbs rather than adjectives just because they are not inflected after the verb meaning “to be” … The fact that some adjectives can be used as adverbs (as in English) does not contradict what I just said: it is its use in a particular sentence that determines which part of speech a particular word is playing.

    marie-lucie,
    here are concrete examples, using “riesig”, of what you are saying. To put it as neutrally as I can: the sound “riesig” sometimes functions as an adjective, in that it indicates that something indicated by a noun has the property of large size – and sometimes as an adverb, in that it indicates “extreme intensity” in the way something is done. Consider the sentence template
    Er hat sich riesig gefreut über … (He was enormously happy about …)
    “riesig” here indicates in what way he was happy, namely enormously. “riesig” is functioning as an adverb. In the sentence
    Das Gebäude ist riesig (The/that building is enormous)
    we see “riesig” functioning as an adjective.

    Scheiß-, which is a very productive way of expressing anger about the rest of the compound noun. The stress can go on Riesen- or on the next component, and it never goes on Scheiß-.

    You amaze me, David. Is that how Bavarians stress Schieß- words all the time? Only today I heard someone say “DAS ist ein SCHEIßLaden” (intonation: high-low-low-low-low-[high-epsilon])- after receiving poor service there. This is the stress pattern I hear in and around Köln – Scheiß- is almost always stressed, and strongly. But of course you also hear things like “Dieses Schieß AUTO springt schon wieder nicht an”.

    which part of speech a particular word is playing.

    marie-lucie,
    I saw your last post, just quoted, as I was about to send off the above. I think that is a very instructive way to describe how the German “riesig” is actually used – sometimes it plays adjective, sometimes adverb. But it is also instructive to say

    there are actually two words, not just one (as you can see from separate dictionary entries)

    when talking about the verb “cook” and the noun “cook” in English.
    I don’t think about adjectives and adverbs when using “riesig” in the one way or the other. “riesig” is “riesig”, that’s all there is to it. “Stepping back” to deploy the old Latin grammarian categories of adjective and adverb doesn’t add anything to my understanding of German. On the contrary, I find it distracting.

  63. The occurrences of “Schieß” (shoot) above should be “Scheiß”

  64. The Mandarin Chinese case is addressed by Dixon, citing Xu 1988, who apparently reports that adjectives and verbs differ in Mandarin in at least the following respects:
    * Reduplication of verbs yields “do a little bit” (dongdong “move a little bit”); of adjectives intensifies them (honghong: “vividly red”).
    * Different derivational suffixes: only verbs take -jia agentive, only adjectives take -hua verbaliser.
    * -de is optional with attributive adjectives (with a couple of exceptions, like “curious”), but not with verbs in relative clauses.
    Jim above denies the third point, but if he’s correct that would still leave at least two tests. I also suspect the use of “hen” allows a distinction – with predicative adjectives I think it has no real semantic value – but I’d need to check.

  65. Is that how Bavarians stress Scheiß- words all the time?
    Yes, that struck me too.

  66. marie-lucie says

    Stu: I don’t think about adjectives and adverbs when using “riesig” in the one way or the other. “riesig” is “riesig”, that’s all there is to it. “Stepping back” to deploy the old Latin grammarian categories of adjective and adverb doesn’t add anything to my understanding of German. On the contrary, I find it distracting.
    Using and analyzing are not the same thing, and are not often best done at the same time. For instance, you probably don’t think of the differing meanings and roles of “pretty” in “pretty good” and “pretty girl” when you use those expressions in normal conversation, but if you were teaching English to a class of immigrants new to the language, it would be helpful to be aware (and to make them aware) of those differences. Similarly for when you use riesig in speaking German, which I guess you do extremely fluently. In linguistics we would say that you have internalized German grammar, just as you internalized English grammar in the course of growing up, so you can use these languages for daily communication without having to think about all the details.
    Another simile that is often used is that of learning to drive a car: if you are a good driver you rarely have to think about what you are doing except when something out of the ordinary happens, and it would be very distracting to have to consciously think of every gesture you make while driving. But if you are a driving instructor, you need to pay attention to such things. Similarly, if you are a professional musician you can only give your best performance if you are able to forget the details of finger movements, etc, even though you may have spent a lot of time concentrating on how to execute these movements until you could do them perfectly and fast.
    About the terminology, “adjective” and “adverb” have been around since the Romans, but that does not make them old, the general definitions of these words have stood the test of time, otherwise they would have been replaced. Specific details of their use may vary from language to language, but there is no reason to discard these useful terms.

  67. David Marjanović says

    Now that I see your examples, I have to say that there is no reason to consider forms like riesig as adverbs rather than adjectives just because they are not inflected after the verb meaning “to be”:

    Cringe-inducingly boring classical examples because I should have gone to bed long ago:
    Sie ist schön “she is beautiful”
    Sie singt schön “she sings beautifully
    das schöne Mädchen “the beautiful girl”
    (There actually are people who use sich riesig freuen? I had only encountered that phrase in writing so far and thought it might have died out.)

    Only today I heard someone say “DAS ist ein SCHEIßLaden” (intonation: high-low-low-low-low-[high-epsilon])- after receiving poor service there.

    Agreed, I’d stress that the same way — except that this usage hardly ever occurs where I come from. Try “SO ein SCHEIßgeSCHÄFT“…

    “Stepping back” to deploy the old Latin grammarian categories of adjective and adverb doesn’t add anything to my understanding of German. On the contrary, I find it distracting.

    Quite so. German is very unlike the rest of Standard Average European in that respect. It took years till I got confident in choosing adjective vs adverb in English, and only then did I manage to get it right in French (which I had actually started learning earlier). When I started with Russian, I had already grasped it, but I was still relieved when Mandarin turned out to lack the difference between adjectives and adverbs completely.

    Reduplication of verbs yields “do a little bit” (dongdong “move a little bit”); of adjectives intensifies them (honghong: “vividly red”).

    Let me just confirm that. Xiūxi “to rest”, xiūxixiuxi “have a little break”; cháng “long/eternal/often” (I may have confused homophones with different characters here), chángcháng “all the time”.

  68. marie-lucie says

    Sie ist schön “she is beautiful”
    Sie singt schön “she sings beautifully”
    das schöne Mädchen “the beautiful girl”

    Looking at these examples, I conclude that in 1 and 3, schön works as an adjective, but in 2 as an adverb:
    – in 1 it refers to a semantic property of the noun phrase signified by the pronoun sie, which could be replaced by meine Freundin “my girlfriend” or other such phrase, including a feminine name; it does not refer to a property of the verb ist;
    – in 3, it works as an adjective, since it occurs within a noun phrase and takes an inflection showing that it agrees with the noun Mädchen “girl” which is the “head” of that phrase; only adjectives can do that;
    – but in example 2, schön works as an adverb, referring to a semantic property of the verb singt, not to the pronoun sie or equivalent noun phrase.
    So: two adjectival roles (reference to a noun), one adverbial role (reference to a verb).
    David, you say you don’t know what German grammarians (or linguists) say: try to find out. It is quite possible (I don’t know enough German to say for sure) that as a general rule “all German adjectives can be used as adverbs”: that would explain why you were having trouble with French or English which generally distinguish the two parts of speech; but that statement is more likely than “all German adverbs can be used as adjectives” (since there are adverbs, such as those referring to place or time, which are not so used). Adjectives are a more basic category than adverbs (adverbs are typically formed from adjectives, but not the reverse).

  69. marie-lucie says

    Here is from a very recent (and good) article on Wiki:
    Many languages, including English, distinguish between adjectives, which describe nouns and pronouns, and adverbs, which modify verbs, adjectives, and other adverbs. Not all languages have exactly this distinction, however, and in many languages (including English) there are words that can function as both. For example, English fast is an adjective in “a fast car” (where it modifies the noun car), but an adverb in “he drove fast” (where it modifies the verb drove).
    And “fast” is still an adjective in “this car is fast”.

  70. I had already grasped it, but I was still relieved when Mandarin turned out to lack the difference between adjectives and adverbs completely.
    That made me big happy when I found out, too.

  71. shireru da, so shiremasen, sorry

  72. michael farris says

    “adjectives and verbs differ in Mandarin in at least the following respects:
    * Reduplication of verbs yields “do a little bit” (dongdong “move a little bit”); of adjectives intensifies them (honghong: “vividly red”).
    * Different derivational suffixes: only verbs take -jia agentive, only adjectives take -hua verbaliser”
    Leaving aside the final disputed criteria, the first example doesn’t work for me at all. It’s just as economical to say that reduplication has one effect on active verbs and another on stative verbs. I’d want more than a purely semantic criteria for adjectives as a separate word class (rather than a class of verbs).
    The second example is maybe more valid but again, what’s gained by calling them adjectives instead of stative verbs?
    If you start off looking for adjectives, I imagine you can always find them. The question for me is which makes more sense in terms of the overall structure of the language in question: subcategories of verbs (of course there may be more than two categories) or separate categories for verbs and adjectives. I don’t know enough about Mandarin to answer that particular question to my own satisfaction. For Vietnamese, I think the verb categories option makes more sense, but that’s a different language family.

  73. marie-lucie says

    If you start off looking for adjectives, I imagine you can always find them.
    To paraphrase: If you start off thinking that adjectives in your own language are translatable by adjectives in the other language, you will find “adjectives” there, but there might be contradictions (eg several quite distinct types, or structures that don’t fit) which should lead you to reconsider your first opinion in the light of “the overall structure of the language in question”.

  74. A word class is a word class; if it shares enough properties with another class to consider them both as sub-classes of some super-class, great, but you still have to distinguish them, whether you call them stative verbs or adjectives or whatever. In the language for which the term “adjective” was originally invented – Latin – adjectives are fairly obviously analysable as a sub-class of nouns, so it seems a bit revisionist to decide to apply the term only to word classes not analysable as sub-classes of some major word class. I prefer the more semantic definition Dixon goes for: any morphosyntactically definable word class whose members primarily express properties of nouns can be termed “adjective”. But that’s just a matter of definition: the interesting question is whether every language does in fact have at least one word class of such a type. If you start off looking for adjectives, can you always find them?

  75. marie-lucie says

    Lameen, I don’t see why “adjective” has to be a sub-class of “noun”: a sub-class is part of a class, so are you saying that an adjective is a kind of noun? it “expresses properties of nouns” but that does not make it a “kind” of noun, even if in some contexts (eg “the rich and the poor”) it can be used as a noun (by what could be considered a deletion of a predictable noun or pronoun from the noun phrase: “the rich ones and the poor ones”). If Dixon’s “semantic” definition includes “morphosyntactically definable”, then it is not just semantic, because not all words which may “express properties of nouns” in terms of their semantics might also fit the morphology and syntax of the main word class expressing those properties. And if those “properties” are expressed by two or more word classes which differ sharply in their morphology and syntactic use, those word classes cannot both be referred to by the same technical term.

  76. marie-lucie says

    Latin adjectives as a subclass of nouns:
    this can only be said if the crucial clue is morphology, not if the clue is syntactic use. Adjectives are similar to nouns in terms of formal morphological classes, but they have to agree in terms of inflection with the noun they (optionally) modify, which indicates that their role is subordinate to that of the noun, not that they are a subclass of noun. As I said before, an adjective can only be used as a noun when the modified noun would have been semantically predictable (either from common knowledge, or because the noun has already been mentioned in recent discourse and does not need to be repeated).

  77. I am under the impression that akai desu is quite a modern form in Japanese (possibly from southern dialects and not native to Tokyo Japanese), and although foreigners are taught to use it, the Japanese themselves originally regarded it as rather jarring, and perhaps still do to some extent. Akain’desu is much more acceptable, although it is somewhat different in nuance. Still, akai desu does fill a gap in the “conjugation”….
    I think akaku de wa aru is ungrammatical in Japanese. Unless I have it wrong, you could say: Akaku wa aru ga, futsū no hito ga suki na aka de wa nai. (It is red, but it’s not the kind of red that ordinary people like.) Of course, it might be better to say Akai koto wa akai ga,….
    Re Chinese word pairs with different meanings according to tone, how about 好 = hǎo (good) and hào (to like)?

  78. akaku de aru is grammatical i believe, akaku wa aru i was asking about, because can’t recall people saying that and akakimasu also seems to me not a proper verb
    akai desu is a polite form, akain’desu would sound informal, in my understanding
    thanks for responding

  79. Bathrobe: Yes, that’s one of them. As I remember there’s a considerable number of them; it may even be an analytic category.

  80. ToussianMuso says

    In Northern Toussian, the West African (Gur or Voltaic) language that I am learning, thoughts and emotions are mostly idiomatic and mostly reside in the stomach. So there are no adjectives meaning “happy” or “angry”; rather, one would say, respectively, “my stomach is chilled” and “my stomach is hot”. To think is to build one’s stomach. To be afraid is to “do fear”, or one can say “fear is on me”. I have not thus far found any way to speak of being sad other than the negative of being happy (“my stomach is not chilled”).

  81. David Marjanović says

    Looking at these examples, I conclude that in 1 and 3, schön works as an adjective, but in 2 as an adverb:
    – in 1 it refers to a semantic property of the noun phrase signified by the pronoun
    sie, which could be replaced by meine Freundin “my girlfriend” or other such phrase, including a feminine name; it does not refer to a property of the verb ist;

    But we act as if it did. We (zero-)mark it like an adverb; the gender/number/case endings only appear in attributive use.
    I do remember having seen a reference to “predicative adjectives” (and presumably “attributive adjectives”), of which the former are undeclined and therefore happen to look like adverbs. But isn’t that a distinction without a difference?
    I’ve also seen a claim that German lacks monosyllabic adjectives because all endings happen to contain a vowel and zero-marking yields the adverb. That was from a rather prescriptivist book on style, though, not from a linguistic paper, and I don’t remember if the term “adverb” was actually used (though I can’t imagine an alternative).
    In fact, after I had understood the basic difference between adjective and adverb in English, I always wanted to add -ly to adjectives that were used as predicates. I think I never actually did it, because I was already used to seeing (ending-less) adjectives behind to be and therefore thought again, but I think I’ve seen other people doing it. In other words, I do think that I unconsciously treat German sein like every other verb in this respect: that I don’t use adjectives or adverbs depending on whether I’m talking about a property of a noun or a verb, but depending on whether a verb is present.
    That must also be why I wouldn’t have guessed that it’s to smell good; it had to be explicitly taught to us. In German, “to smell good” is gut riechen, while “to smell well” is gut riechen können… “to look good” has an intransitive verb of its own, gut ausschauen, while “to see well” is gut sehen können.
    Maybe the best solution would be to claim that German lacks adjectives and adverbs wholesale, and instead uses one form when a verb is present and the other otherwise.
    Of course, the answer to all “why” questions is “everything is the way it is because it got that way”.* I’ve read a fascinating paper in Diachronica (in Vienna I have free online access to it) on how the adjective/adverb markers of English and German arose. First there was a good old Indo-European adverb ending (-e in Old English, -o in Old High German), then it got syncopated, and confusion followed. In fact, in Early New High German (16th century), some German speakers marked adverbs the English way, with -lich, and there were even prescriptivists who considered this meanwhile completely extinct solution the only correct one! At the same time, neuter nominative singular adjectives often dropped their ending — this is preserved in poetic language, old church songs and the like –, and apparently this was also sometimes done to masculine and feminine ones; Standard German could have ended up like English if another dialect had had greater influence or something.
    Some of the -lich constructions survived, and, to my surprise, I found that they are only used with verbs to this day. (Can’t remember any, alas!) Others, however, have become ordinary adjectives, leading — I think — to synonym pairs like dumm “stupid” and dümmlich (applied to fake folk music, but occasionally also silly people or stupid approaches to a problem or… well, maybe the only difference is that dumm is more common… no, I think dümmlich adds ridicule).

    It is quite possible […] that as a general rule “all German adjectives can be used as adverbs”

    It looks like adverbs can be derived from all adjectives (by not adding an ending to the stem), with complications. The example that comes to mind is that the superlative of adverbs is constructed by a kind of nominalization instead of directly: adjective with definite article: gute — bessere — beste; adverb: gut — besser — am besten — the expected form *best does not exist, except as a prefix (bestmöglich “optimal”, “as good/well as possible”).

    but that statement is more likely than “all German adverbs can be used as adjectives” (since there are adverbs, such as those referring to place or time, which are not so used).

    Yes, of course. There are also others: genug works like its English cognate enough in that it can only be used with verbs (which means, in German, that it never gets an ending).

    And “fast” is still an adjective in “this car is fast”.

    There are no such irregularities in German, though I’m not sure what your point is here.
    * And therefore the ancient wisdom that “science answers ‘how’ questions, philosophy (or even religion) answers ‘why’ questions” is myopic all the way to abject, pungent stupidity. But I digress. 🙂
    ————–

    Re Chinese word pairs with different meanings according to tone, how about 好 = hǎo (good) and hào (to like)?

    The textbook example that comes to mind is mǎi “buy” vs mài “sell”. But of course these are both verbs, and — for whatever reasons — they are written with different characters. Maybe there once was a causative ending -/s/.

    In the language for which the term “adjective” was originally invented – Latin – adjectives are fairly obviously analysable as a sub-class of nouns

    But that’s not what they are inside the head of any speaker of Standard Average European.

  82. marie-lucie says

    ML: And “fast” is still an adjective in “this car is fast”.
    DM: – There are no such irregularities in German, though I’m not sure what your point is here.
    I don’t know what you mean by “irregularities” here, but here is my point: English adjectives are always uninflected as far as agreement with a noun goes, but even then it is possible to distinguish adjectival from adverbial roles of the same adjective, using purely syntactic criteria.

  83. English adjectives are always uninflected as far as agreement with a noun goes, but even then it is possible to distinguish adjectival from adverbial roles of the same adjective, using purely syntactic criteria.

    Are you sure? English adverbials can stand in the same predicative position, as in “The girl is well”, “The meeting is today”, “The window is over there”, or “The reason I went is because I had nothing else to do”.
    If you really want to strike a distinction between adjective and adverb in the sentence “this car is fast”, you would have to rely on semantics: If it is used to talk about a car that merely has the capability to go fast, “fast” is purely adjectival. But if it describes a car that is going fast on a race track, “fast” becomes adverb-ish.

  84. David Marjanović says

    But if it describes a car that is going fast on a race track, “fast” becomes adverb-ish.

    No, only in German. That’s because it goes fast; it is not in a fast way.

  85. David Marjanović says

    My point is that English distinguishes these cases for all other adjective-adverb pairs except “hard”, so “fast” and “hard” are most easily interpreted as pairs of adjectives and irregularly formed adverbs, while (non-Highest-Alemannic) German never makes this distinction.

  86. michael farris says

    Getting to Japanese, another reason for not establishing a separate category of “adjectives” is that some purely verb forms are not readly distinguisably morphologically from “adjectives” (a stronger argument from adjectives not being distinguishable from verbs).
    Many years ago, I read a wonderful short and elegant descriptive grammar of Japanese (morphology/syntax alone IIRC) though I can’t remember the title or author (except that the last name seemed French).
    Anyway, the author generally abstained from traditional labels in the interests of presenting Japanese structures on their own terms. Early on, the author positited three verb classes; 1. da (a single member class) 2. u-verbs (like tabera, iku etc) 3. i-verbs (akai, tabenai, ikitai)
    Also nominal forms can be used to describe attributes of nouns but have to be verbalized with da (predicatively) or dependent forms of da ( X no for free nominals and X na for bound nominals).
    I’ve yet to come across any data or counter arguments strong enough to motivate me to revise this.

  87. Grumbly Stu says

    @michael farris
    For some reason I can’t post this email directly to your mail address, so I’m putting it here.
    re your last post at LH:

    a wonderful short and elegant descriptive grammar of Japanese … the author generally abstained from traditional labels in the interests of presenting Japanese structures on their own terms

    In the past, I have wondered whether the Latin grammatical categories are as useful in describing individual languages as is generally assumed – and whether, after a certain point in their application, they become more of a hindrance than a help to understanding languages as a whole.
    The relatively trivial business about “adjective or adverb in German” in the LH thread got me thinking about this again. I feel it’s pointless to argue whether “riesig” is “really an adjective” here or “really an adverb” there. But there is a crude linearity and dubious “ontology” in the associated argument along the lines of “which is more basic”.
    Many years ago, when I looked at commonly available Arabic or Chinese grammars, I had the feeling that the apparent complexity must be due almost entirely to the inappropriatenes of European categories. After all, as the saying goes, there are billions of Chinese kids who learn Chinese without these grammmars, so it can’t be that “hard”. A standard reply to this is something like: “this is a scientific issue of taxonomy and historical placement of the language. It should not subject to considerations of whether 3-year-olds can understand or profit from it”.
    But I am not satisfied by this answer. There are a number of issues here that are simply being sidestepped. Do you know of internet sites where such issues are / have been seriously discussed, with possible implications for linguistic science?

  88. michael farris says

    Stu, the addresses I use in posting responses are generally not … accurate (old habits die hard). If you really want to contact me off-hat, then let me know and I’ll work something out.
    “I have wondered whether the Latin grammatical categories are as useful in describing individual languages as is generally assumed”
    Well, they’ve certainly wreaked havoc on the traditional grammatical tradition in English. Non-native users usually have no idea how absurd and stupid the models of English grammar taught to native speakers is (cause for another rant). The grammar taught to ESL students is usually about a hundred times more reasonable.
    It’s my understanding that traditional Arabic grammar has lots of arcane complications of its own without any help from Latin. I have seen more descriptive grammars of colloquial Arabics that seem pretty reasonable. You need to remember that the full on literary language (complete with dual everything and case endings) is spoken by approximately no one as a first language and even highly educated Arabic speakers find it difficult to use according to the traditional norms.
    I don’t know enough about written Chinese grammars to say anything reasonable except that Classical Chinese is rather like Classical Arabic (most Chinese no matter how educated can’t use it) but maybe for different reasons (and I’m thinking that explatory description has never been a priority in the Chinese intellectual tradition). And again, I’ve seen descriptive grammars of Mandarin that don’t seem that complicated.

  89. michael farris says

    Getting to your final question, the American structuralist tradition (pre-Chomsky) had a mania for describing languages on their own terms and not set against any kind of theoretical model of anything.
    The theoretical assumption was that human languages were capable of displaying an infinite degree of variation. Now, no one completely believed that, but as a heuristic approach it makes it more likely that a linguist _can_ discover what variation there is.
    From Chomsky on, the assumption is that linguistic variation exists only to a trivial degree and all languages are basically the same and the task of the linguist is to map the features of any given language against an accepted model of possible inguistic structures that may then be tweaked (but not challenged or overthrown).

  90. Michael,
    The only reason I wanted to contact you off-hat was just to get some internet sites to start with, on my own. I didn’t want the thread folks to think I was “raising an issue”. I have simply had some ideas for a long time, which have surfaced again in connection with my current Luhmann reading, and the German adjective and adverb business in this thread.
    At LH, everybody gets to put his/her oar in and splash around a lot, and no harm done. But I suppose most contributors have their own, more serious irons in the fire. I have mine as well. Though I’m not surprised that you want to hold your email address close to your chest. There are certainly enough crazies running around the edges of mathematics and philosophy, subjects on which I know a thing or two, and I bet it’s no different in linguistics.
    There’s no way to tell, from what I write here, whether or not I might be an obsessive, closet ranter about things linguistic or anything else. Again, that’s why I hoped you could just name me some sites – I’ll take it from there. That you could do in a comment in this thread.
    I don’t want to learn Arabic or Chinese, by the way. German, French, English and Spanish are mutually different enough for me to be going on with, in connection with this subject.
    <pause >
    Just read your second comment, in which the last paragraph addresses what I meant by “possible implications for linguistic science”. To be perfectly clear about this – I have not the slightest ambition to overthrow anything. What I want to do is follow my thoughts through for myself – zunächst mal. I just don’t think that proof theory, model theory and formal grammars have to have the last or the first word on classification and semantics (last I night I read a piece on languagelog by Mark Liberman called “Formality and interpretation”).
    It’s interesting that, among the people who contribute to this site, I’ve never read a word that suggested “from Chomsky onwards”. But then everyone here does phonology and historical linguistics, right? That doesn’t involve semantics, right? Wrong, maybe? But I don’t want to go on about that here.

  91. Marie-Lucie:
    “Latin adjectives as a subclass of nouns”:
    The similarities between adjectives and nouns in Latin go further than the fact that they share the same inflectional morphology (though adjectives can still be distinguished morphologically by taking the comparative.) It’s also that all adjectives can stand alone as noun heads, so how do you distinguish a noun-noun apposition from a noun-adjective phrase? Gender agreement is the key there – you expect number and case agreement in appositions anyway. By taking adjectives to be a subclass of nouns in Latin, you economically account for both their inflectional morphology and their ability to stand alone.
    “an adjective can only be used as a noun when the modified noun would have been semantically predictable”: surely you can use any Latin adjective on its own as if its head were “thing”, “man”, or “woman” (depending on gender), eg pulchrum “a beautiful thing”? Or was that an oversimplification for high school students? If the latter, I’d love to be pointed to a more accurate description.

  92. Michael Farris:
    That description of Japanese sounds reasonable to me; and by Dixon’s definition the i-verbs, and the “bound” nominals, would each count as a distinct adjectival word class. But bear in mind that the argument for treating Latin adjectives as nouns is far stronger than the argument for treating Japanese i-adjectives as verbs: at least Japanese “pure” verbs are only morphologically indistinguishable in some of their forms! (I assume you’re alluding to desiderative -tai forms.)
    Stu: Some of the older Arabic grammars in English (Wright’s, for example) describe (Classical) Arabic largely in the terms used by Arab grammarians (or calques thereof.) Have a look and let me know if you find it makes things any easier… Personally, I never understood the Classical Arabic case system until I read descriptions of it in English using the Latin grammatical terms, and most educated Arabs I know have at least some difficulties with it.

  93. (My last post seems to have disappeared, so trying again):
    Michael Farris: That description of Japanese sounds reasonable, and in Dixon’s terms would mean that “i-verbs” and “bound nouns” would each separately count as an adjective class. But bear in mind the arguments for calling Latin adjectives nouns are much stronger than for calling Japanese i-adjectives verbs – at least Japanese verbs have forms that are _not_ morphologically just like adjectives, whereas all Latin adjectives and nouns share the same case and number marking morphology.
    Stu: Wright’s Arabic Grammar describes Classical Arabic in the terms traditionally used by Arab grammarians (or calques on them.) If you find it easier, let me know – personally, I never understood Arabic case marking until I read descriptions of it in English using Latinate terminology, and most educated Arabs I know still have at least some difficulties with case marking.

  94. michael farris says

    Lameen, cobbling an ‘adjective’ superclass out of i-verbs and na-nouns seems to me like establishing a word class on the grounds of semantics alone, not a part of the school of linguistics I come from.
    And since the Latin term is so strongly associated with nouns, I’m uncomfortable using it in a language where words with similar semantics align with verbs. And syntactically u-verbs can modify nouns in just exactly the same way that i-verbs do.
    akai hon (the red book)
    katta hon (the book X bought)
    I know of one ‘descriptive’ model of Japanese that posits katta in the second example as a participle (that just so happens to be identical with the predicate form). That seems sloppy and adding on categories to the language for no good reason.
    Some years ago I was involved in trying to help a new Arabic speaking arrival in Poland learn some basic Polish (they arrived with none and their English was pretty basic). At one point trying to explain an accusative form in Polish I hit upon the bright idea of writing Polish and Arabic equivalents (that I was able to dredge up out of my memory):
    pies – kalbun, alkalbu
    psa – kalban, alkalba / kalbin, alkalbin
    (I forget if the example was dog or not but the principle is the same)
    Unfortunately that didn’t help, the point the Arab learner took from my examples were that the difference between pies and psa applied only to the literary language and could easily be ignored in everyday speech (as they explained later some time after they had figured out the difference in Polish was crucial even in the most informal usage).
    Speaking of Polish there are subclasses of what might be called ‘pure’ adjectives (which are in the nominative in predicate position) and ‘nominal’ adjectives which behave syntactically like pure nouns (appearing in the instrumental in predicate position).
    On jest leniwy. He’s lazy (nominative).
    On jest myśliwym. He’s a hunter (instrumental from nominative myśliwy).
    IIRC these are all (a)vocations and do not agree in gender like real nouns.
    Ona jest leniwa. She’s lazy (veminine form of leniwy)
    Ona jest myśliwym. She’s a hunter. (same form as masculine).

  95. ToussianMuso says

    “From Chomsky on, the assumption is that linguistic variation exists only to a trivial degree and all languages are basically the same and the task of the linguist is to map the features of any given language against an accepted model of possible inguistic structures that may then be tweaked (but not challenged or overthrown).”
    I find that very surprising. I would have assumed most major schools of thought to be moving in the other direction. Isn’t there a happy medium somewhere?

  96. Dixon wouldn’t postulate an adjective superclass – just two separate word classes which are both adjectival. He inveighs for some time against the idea that adjectives have to be nominal; of course terminology is just terminology, but I do agree with him that it’s not very symmetrical to have a separate term for noun-like classes of property words but none for equally distinct verb-like ones.
    That Polish example is interesting – how do you know the “hunter” ones are actually adjectives?

  97. michael farris says

    Okay, I have no problem with the idea of a conceptual functional class (property words) and then mapping out how functions and structures interact in particular languages.
    I would include u-verbs in Japanese too as fucntional adjectives since they can be used (with no morphological change) both attributively and predicatively to describe nouns.
    katta hon (the book X bought)
    hon wa katta (the book was bought /as for the book, X bought it).
    I anticipate disagreement about whether ‘bought’ is a property or not, but we can leave that to another day.
    As for the ‘hunter’ ones. Well, that’s the point, morphologically it’s inflected just like anadjective but syntactically it acts like a noun (if that makes sense).
    Apparent mismatches between morphological properties and syntactic function are not at all rare in Polish (and I assume other Slavic languages).

  98. In his short story “Tlon, Uqbar, Orbis Tertius” (http://interglacial.com/~sburke/pub/Borges_-_Tlon,_Uqbar,_Orbis_Tertius.html), Jorge Luis Borges imagines and describes fictional languages without adjectives or in which adjectives play the main role… It is a very interesting story.

  99. David Marjanović says

    Maybe German is to Standard Average European like how ergative-absolutive languages are to nominative-accusative ones: there are three syntactic categories, and everyone (except, sometimes, the Nez Percé) treats two of them the same way morphologically, just not the same two.

  100. David M writes:
    Does Chinese distinguish anything by any other means than word order?
    Oh yes. Chinese verbs have about the same amount of inflectional morphology as English verbs do, though I don’t have the list of affixes on tap at the moment.
    For example, word order is the only way to tell that Chinese is a nominative-accusative language (rather than an ergative-absolutive one.
    I argue here (following Randy LaPolla) that it’s really neither one, and that pragmatic considerations dictate whether conjoined VPs get the nom-acc or the erg-abs interpretation:

    Chinese, on the other hand, is neither ergative nor accusative. Sentences that literally translate to “Cthulhu dropped the watermelon and burst” and “Cthulhu dropped the watermelon and was embarrassed” are equally valid, and Chinese speakers interpret them according to common sense. That is, “…the watermelon burst” (ergative) and “…Cthulhu was embarrassed” (accusative). Only the latter one reads properly in the English translation because English isn’t ergative.

  101. some German speakers marked adverbs the English way, with -lich, and there were even prescriptivists who considered this meanwhile completely extinct solution the only correct one

    I remember when I was taking my one and only formal German class (my mother taught it, which made it easy to audit), and the very first sentence of the very first narrative in the book was “Nun bin ich endlich in München”, and I saw after a bit that endlich not only meant the same as finally but had the same structure: end-lich, final-ly, with the Latin word final replacing the native one in English. Little did I know then how wrong that was, but now I know why. Very satisfying after half a century!

    (Wikt says that endly ‘finally’ is now rare and non-standard (which in dictionary-speak means ‘occurs maybe once in some obscure place’), though it was current in Middle English. The OED just calls it obsolete, and cites it last in 1475.)

  102. Strange that I didn’t respond to read in 2009. read is now long gone as a commenter so there’s not much point in issuing a correction. But just for the record:

    akaku de aru is definitely ungrammatical. Akaku wa aru is definitely grammatical and used in Japanese.

  103. Always good to mention these things; you never know who might be reading these threads and picking up misinformation.

  104. Actually, aka de aru (or aka de wa aru) is grammatical, but aka here is not an adjective; it’s a noun.

    akaku de aru is simply wrong.

  105. David Marjanović says

    Back in the Dreamtime…

    Scheiß-, which is a very productive way of expressing anger about the rest of the compound noun. The stress can go on Riesen- or on the next component, and it never goes on Scheiß-.

    You amaze me, David. Is that how Bavarians stress Schieß- words all the time? Only today I heard someone say “DAS ist ein SCHEIßLaden” (intonation: high-low-low-low-low-[high-epsilon])- after receiving poor service there. This is the stress pattern I hear in and around Köln – Scheiß- is almost always stressed, and strongly. But of course you also hear things like “Dieses Schieß AUTO springt schon wieder nicht an”.

    Scheiß- can receive contrastive stress like everything else. Absent that, it’s unstressed.

    Unrelatedly, Züridüütsch is not Highest Alemannic; to find that, you have to go higher, like to Luzern.

  106. David Eddyshaw says

    I wonder if ToussianMuso is still out there somewhere? Fans of Gur languages should stick together …

    I know nothing of Tusya; WP seems to think it’s no longer classified as “Gur”, which puts it in good company: Manessy’s “Gur/Voltaic” was a real ragbag without much in the way of evidence for its validity. I think his criteria boiled down to: “Niger-Congo, noun class suffixes instead of prefixes, spoken near other Voltaic languages.”

    The phenomenon that ToussianMuso describes is by no means confined to “Gur” languages, of course, and it’s perfectly compatible with the existence of adjectives as a word class. Kusaal, too, expresses emotions via body-part expressions, (M sunf pɛligya “I’m angry”/”My heart has gone white”) but it has an unequivocal adjective class. A potentially open class, at that: there aren’t a huge number of basic adjectives (though there are dozens at least), but there are regular processes for deriving adjectives from verbs.

    Western Oti-Volta in general is the same as Kusaal in this, but it’s not altogether clear whether this is true of Oti-Volta as a whole: you run into problems not so much of fact as of definition in languages like Nateni and Mbelime, in which all stems usable as adjectives can also be conjugated as stative verbs without stem changes, and in which deverbal adjectives from dynamic verbs also simply show the verb stem unchanged without any derivational suffixes. Even there, though, you at least have to distinguish verbal from adjectival stem use somehow.

Speak Your Mind

*