Skip to main content

View Diary: ÍslensKos: The Icelandic Language, Or, What's So Scary About Super-Long Words? (109 comments)

Comment Preferences

  •  I don't think it's the most declined, (4+ / 0-)

    or even in the running.  Icelandic has only four cases compared to e.g. Polish and Czech (seven), Finnish (fifteen!) and I'm not even touching the craziness of Hungarian, which has the standard set of cases and genders and numbers plus personal and possessive suffixes.  

    Saint, n. A dead sinner revised and edited. - Ambrose Bierce

    by pico on Mon Sep 24, 2012 at 07:48:01 PM PDT

    [ Parent ]

    •  A former Estonian girlfriend showed me declination (5+ / 0-)

      15 or 16 cases!

      It's fun walking through Helsinki or Tallinn or Budapest and see these monsterously long words in the window for something that would be 6-7 letters long in English.

      "now this is not the end, it is not even the beginning of the end. But it is, perhaps, the end of the beginning." W. Churchill

      by Thor Heyerdahl on Mon Sep 24, 2012 at 11:49:18 PM PDT

      [ Parent ]

    •  Yep, in terms of noun cases, Icelandic is pretty (4+ / 0-)

      low.  On the other hand,  declensions of both nouns and adjectives change whether or not the definitive article is attached, there's 120 declension forms of adjectives, and verbs are a special hell   ;)  Also, Icelandic is famous for its declensions being "bendy" and irregular.  It's not just endings (which is basically the same as a postposition) like declensions in a lot of languages.  The stem itself often changes in all sorts of crazy ways.  For example, cat:





      That's not an irregular noun - it's actually a whole declension pattern.  And for all nouns and verbs there's stuff like the u-ö shift.

      •  Though to be fair... (1+ / 0-)
        Recommended by:

        there's no real distinction between "pattern" and "exception".  For example, the above "pattern" probably represents about 10-15 base words, and there's probably 100 or so compounds that end with such a base word and are thus declined as such.  At what point of rarity do you declare something to no longer be a pattern and be an exception?  There's no clear line in Icelandic.

        Let's show how blurry it gets.  Do you group örn (eagle) with that pattern, or is it its own thing?  That's örn, örn, erni, arnar / ernir, erni, örnum, arna || örninn, örninn, erninum, arnarins / ernirnir, ernina, örnunum, arnanna.   Okay, then do you group, for example, björn (bear) with the same big group, or is it it's own thing, or is it a group with örn?  That is, björn, björn, birni, bjarnar / birnir, birni, björnum, bjarna || björninn, björninn, birninum, bjarnarins / birnirnir, birnina, björnunum, bjarnanna.

        And then you start getting into patterns that have some characteristics of one declension but then other characteristics of a totally different one!  That's especially common in verbs; they're sort of mix-and-match.

        It's kind of hard to categorize Icelandic noun and verbs sometimes.  It's easy with the big groups but then it gets harder and harder with no clear cutoffs.

      •  Answered my question, above. (1+ / 0-)
        Recommended by:

        Of course.

      •  Yeah, but that's how other inflected languages (1+ / 0-)
        Recommended by:

        work, too.  I was just responding to the comment above - it sounded strange (because I speak a few inflected languages) to hear that claim made about Icelandic, and I looked it up, and it's not the case. (pun!)

        Hungarian I think takes the cake for most needlessly complicated grammar, and I doff my cap to people who can speak it with anything like fluency.  Some of these languages have a few alternate case endings that are more aesthetic than strictly grammatical, which for non-native speakers, can be a real pain.

        Saint, n. A dead sinner revised and edited. - Ambrose Bierce

        by pico on Tue Sep 25, 2012 at 11:17:14 AM PDT

        [ Parent ]

        •  Hmm, I just picked a couple random Hungarian (1+ / 0-)
          Recommended by:

          declension tables and they don't bend the stem at all - it just appears to be suffixes.  What languages are you thinking of that have significant stem-bending in declension?

          I took Latin in high school, and it had very little stem bending.  The only non-irregular case I remember is the pattern of rex/regis - and one could just argue that the x/g is part of the suffix, there's no deep-stem changes like in Icelandic.  The entire stem can get twisted - for example,  For example, nominative-singular-strong-noncomparative-masculine of "pressurized" is "þjappaður".  The same but dative is "þjöppuðum".  And that's not an exception, either.

          Anyway, there's "only" sixteen forms of each noun (and half of those are usually fairly easy to form if you know the other half), but there's 120 for adjectives, and for verbs, it depends on what you count as a form, but it's quite sizeable.  And they're just so danged irregular!  :Þ.  At least for adjectives, the 120 aren't usually too hard and there's not too high of a rate of exceptions  - the main challenge is just memorizing the differences between the 10 or so patterns.

          When it comes to noun genders, it could be worse.  I hear that there's no way to recognize gender in German, for example.  Icelandic has three genders, and you can only kind-of deduce them.  You can basically get a probability matrix out of looking at an unknown word if you know what to look for.  Nominative ends in -a?  Easy one - Over 99% chance it's weak feminine.  Ends in -i?  Hmm, 70% chance it's weak masculine, 30% chance it's strong neuter.  But if it's something like "hótel" that you've never seen before?  Haha, good luck.  If it doesn't fit into any obvious rules and there's no obvious suffix, odds are probably 20% it's strong masculine, 35% that it's strong feminine, and 45% that it's strong neuter. .  

          Also, even once you know the gender, if it's not weak and it's not neuter, that's not enough to know how to decline it.  You need to know at a minimum the genitive singular ending and the nominative plural ending.  You also need to know if it's an exception (of which there are tons) or fits one of the dozens of rarer declension patterns.  For strong masculine, common exceptions (or sshould they be called patterns?) is the loss of the -i in the dative singular.  For strong feminine, the equivalent common irregularity is -na in the genitive plural instead of -a.

          Only once you know all that can you accurately decline it.

          •  Not sure how much experience you have (0+ / 0-)

            in other contemporary inflected languages, but stem-bending (?) happens all over the Slavic languages, for sure.  The Russian for "child" (дитя, nominative) morphs into two separate stems (for a total of three), дит- -> дитят- and деть- in oblique and plural cases; it's part of the template of nouns most of which describe young animals, including children (heh).   Czech has a set of words whose stem not only changes, but is declined either as an adjective or as a noun depending on ...  well, I'm not exactly sure when or why one or the other is used (týž or tentýž -> touž or tutéž, etc.)  These aren't irregulars but paradigmatic.  

            Ditto with gender: both languages have three, both have certain 'typical' signals, and both have deep slates of words that are entirely unrecognizable as any particular gender... you just have to "know".  Throw numbers into the mix and sheer chaos results.  In Russian, if you're looking at "three beautiful women", "three" is accusative singular, "beautiful" is genitive plural, and "women" is genitive singular, so literally none of the words share a common declension pattern, despite being a single phrase.   This is only the case in the numbers 2, 3, and 4, because anything higher has a different set of declensions, and anytime a large number ends in 1 (e.g. 1,001) the whole number is treated as a singular instead of a small or large plural.   Bah.

            Anyway.  I'm not questioning the beautiful complexity of Icelandic, and you've written quite a good diary about it.  But it's certainly not the most inflected European language.  

            For what it's worth, it's probably Basque.

            Saint, n. A dead sinner revised and edited. - Ambrose Bierce

            by pico on Tue Sep 25, 2012 at 10:39:47 PM PDT

            [ Parent ]

            •  Other languages (1+ / 0-)
              Recommended by:

              I also speak some Japanese, which doesn't do stem-bendings for any form of word.  And there's obviously English too  ;)  The only nouns I can come up with in English with stem bendings are archaic words with rare plurals, like brother -> brethren, and one common exception: man->men.  Now, English has strong verbs, which use a single stem-bend of a single vowel to form the past tense - for example, run -> ran, sit -> sat, dig -> dug, etc.  Of course, Icelandic strong verbs have three separate vowel stem bends plus the universal u-umlaut rule plus you need to memorize two suffix patterns and the past participle for each  ;)  There are some rules for strong vowel stem bends, but they're so complicated and so full of exceptions, they don't really help much.  It's always nice to learn that a new verb you're learning is weak, or at least one of the -ti, -di, or -ði past pattern verbs.  :)  A looooot less to memorize for it.

              Not trying to argue that Icelandic is the hardest  :)  Just describing it.

              I'm curious about the Russian stem shifts.  I looked it up and the only example I saw of it was the one you gave, with a difference between singular and plural for child.  A grammar page I found told me that "hard nouns" add -ы and "soft nouns" replace their nominative suffix according to a couple specific rules, but none of them involved stem morphs.  But дитя is not an exception?

              Hmm, interesting abount bending a noun as an adjecctive - how does that work?  Like, what would be the adjective form of "spoon" mean - "spoon-ish"?  Icelandic lets you bend verbs as adjectives (analogous to, run -> running), so that adds the 120 adjective forms to each verb.  Thankfully not every adjective-bending pattern is represented among verbs!  :)  The past participle (which you have to memorize) forms the basis of the adjective bending (in fact, the literal Icelandic translation of the word for past particple is "adjectivemood of past"

              Funny the thing about low numbers in indoeuropean languages being declined but high numbers not.  In Iceclandic, it's 1-4 that are declined.  For cardinal numbers of things that aren't "pairs" (they get their own set, as do cardinal numbers), there's 12 forms (okay, well, there's 24 for "1", but that's special).  But 5-19 are undeclined (after that, it depends on what singles-digit it ends in).  There's actually a joke here about an immigrant going into a store trying to buy four of something who keeps guessing at what form of "four" to use, to the confusion of the shopowner, until finally giving up and buying five of it  ;)

              •  дитя isn't an exception: (1+ / 0-)
                Recommended by:

                It's the same paradigm is котенок, where the stem morphs to котенк- for singular oblique cases and котят- for plural declensions.  Like I said, most of the nouns in this paradigm refer to young animals.

                Not sure how it works in Hungarian per se, but Russian does 'bend' nouns into adjectives: e.g. the stem for "milk" is молок- (noun) -> молочн- (adj).  As for verbs to adjectives: from the infinitive читать (to read) -> the stems читающ- ([who] is reading), читавш- ([who] was reading), читаем- ([which] is being read), (про)читанн- ([which] was being read), plus all the attendant declensions of each, and the thankfully non-declinable читая for the adverbial participle.  Thankfully Russian has only three tenses and two moods, so the simple verbs themselves aren't that bad, but their system of describing motion is ridiculously complex and difficult for non-native speakers: directional/aspectual prefix plus directional/frequentive verb stem plus whatever conjugation you need.  So: prefix у- means to depart, stem идти normally means to go in a single direction on foot with an imperfective aspect, but the combination turns it into a perfective stem, and it's irregular anyway, giving us уйд- for "[someone] will depart on foot".  There is no verb for just plain "go".  Of all the things I've taught, no single topic creates more headaches for students, though it turns out to be hugely efficient (look how short that stem is compared to the translation!)

                Comparatives are weird, because they have far more than the English three (adj, comparative, superlative).  For the adjective "beautiful" (красив-):

                enhanced: прекрасив-
                comparative: более красив-, красивее
                superlative: самый красив-, красивейш-, наикрасивейш-
                comparative superlative: более наикрасивейш-
                superlative superlative: самый наикрасивейш-

                Saint, n. A dead sinner revised and edited. - Ambrose Bierce

                by pico on Wed Sep 26, 2012 at 02:58:23 AM PDT

                [ Parent ]

                •  Haha, that's great (0+ / 0-)

                  This is why I love learning languages; there's so many different ways to express things and different peoples have come up with all sorts of neat approaches.  Neat about the comparatives and directions (the latter I assume only applies to verbs of motion).

                  One of the things that Icelandic has that I understand is rare is the "middle voice".  That is to say, there's active voice (subject does something to object) and passive voice (object does something to subject), but with the middle voice, both subject and object are mutually doing the action.  So, for example, a common way to say goodbye is "sjáumst" or "sjáumst seinna".  "Sjá" is "to see", and "sjáumst" is the middle-voice third person plural indicative.  It's literally "[We] [will] see each other (later)".

                  I still don't get the concept of forming adjectives from nouns.  Seems to me to get an adjective from a noun you also need an implicit passive verb.  That is, what is the adjective form of "spoon"?  Well, you could have something that "looks" like a spoon, "tastes" like a spoon, "scoops" like a spoon, "shines" like a spoon, etc.  But how would you do that generically?  Making adjectives from verbs is easy, because the implicit noun needed is simply "one".  "One who jumps" -> "jumping" -> "jumping spider".  "One who burns" -> "burning" -> "burning tree".  Etc.  The same sort of thing like the English -ing construction is done in Icelandic, except of course there's 120 forms of the adjective and you have to know the past participle.

                  But adjectives from nouns without a verb?  I don't get that concept.  What's the implicit verb?

                  Also there's the issue of getting nouns from verbs, but that's a whole different story!  :)

                  Also, would you mind explaining how that stem morph in Russian works and how it's not an exception?  Like I said, I looked up a page that said how to form Russian plurals and they didn't mention stem morphs, so clearly it wasn't nearly comprehensive enough of a page.  :)  Maybe if it's not an exception  then it's a pattern or something, and they just didn't go into patterns?  You've got me curious. Icelandic, too, has some stem-morphing pluralization patterns beyond the universal stem-bending rules like the u-umlaut rule - Js showing up or disappearing, rotating stem vowel patterns like ö-ö-e/i-a:e/i-e/i-ö-a, and so forth.


                  •  There are at least three (that I know of) (0+ / 0-)

                    paradigms in Russian declension that involve stem morphs.  I wouldn't call them exceptions because they're not: it's one thing if a word or two follows the pattern, but we're talking categories of words.  The example I gave you holds true for (among others) ребенок, котенок, медвежонок, цыпленок, ягненок, гусенок, индюшонок, козленок, олененок, etc. etc. and the slightly irregular дитя, which otherwise follows the pattern.   At what point is this no longer an exception but a bona fide paradigm?  And it's a productive category (meaning the paradigmatic stem can be  used to create new words as well.)

                    Other stem-changing paradigms: words like гражданин- which have a variable infix -ин- that disappears in plural cases (so you get stem variants like северянин-/северян-).  This category deals with nouns referring to types of people, and involves a large list of words as well.   And another, albeit smaller paradigm for all the neuter nouns that end in -мя, where oblique cases involve a stem shift from -мя -> -мен-, words like имя (name), время (time), etc.  

                    I haven't included all the nouns whose stems end in consonant clusters, so the oblique cases sometimes add vowels for pronunciation purposes.   It's not a genuine stem shift in the way you're describing it (it has to do with the language's medieval roots), but it's plenty common to see things like окно -> окон, бабушка -> бабушек, etc.  

                    I don't understand the question about making adjectives from nouns.  Why do you need a verb in there?  In English we tend to use the nouns themselves as modifiers (i.e. "milk chocolate", milk is a modifier), but Russian requires an adjectival form here, so they just create one out of the noun: молоко -> молочный, etc.

                    Saint, n. A dead sinner revised and edited. - Ambrose Bierce

                    by pico on Wed Sep 26, 2012 at 11:53:24 PM PDT

                    [ Parent ]

                  •  Adjectives from nouns? Spoony! (0+ / 0-)

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site