Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: Pluralis på svensk och tysk (was: Re: Performative verbs (was: Re: here is some stuff i want all of ya'll to look at)

From:Mark Reed <markjreed@...>
Date:Wednesday, September 8, 2004, 14:14
On Wed, 8 Sep 2004 10:07:28 +0200, Andreas Johansson <andjo@...> wrote:
> I guess there's no hard line for when the probability for a certain formation is > high enough to say it's the regular one, but I intuitively feel that 67% is too > low, and that I, as a native speaker, do not feel that -er plurals are any more > unexpected than -ar ones on this kind of nouns.
The magnitude of the probability doesn't really have anything to do with regularity. There are (at least) two types of regularity: statistical and, for want of a better word, morphological. They answer two different questions: 1. Given a randomly chosen noun from language X, which of the possible plural forms is more likely than any of the others to be correct? This is statistical regularity. The absolute magnitude of the probability doesn't matter; all that matters is the *relative* probabilities of the various options. Whichever form has the plurality of cases would be the statistically regular one. 2. Given a novel noun being introduced to language X for the first time, which of the possible plural forms is most likely to be applied to it by a native speaker? This is morphological regularity, and concerns the default form rather than the most common. There are other ways of identifying the morphologically regular form - for instance, application of a rule takes slightly longer than looking up a memorized form, so you can time how long a speaker takes to answer when asked for the plural of word X, and over enough speakers and trials you will generally find a clear indicator that some type of plural consistently takes longer than others to come out. The two are not necessarily the same; as John said, German is a good example. The statistically regular plural is probably -en, but the morphologically regular plural is -s - even though it's statistically the *least* common form. To apply the test to Swedish, let's say you were discussing baseball in that language (not very likely, perhaps) and you wished to express the number of RBIs(*) a certain player has accumulated this year. Without expanding or translating "RBI", how would you make it plural? -Marcos I won't listen to any arguments from my fellow Americans that the plural of "RBI" should be "RBI" because that's what you get if you expand the acronym, make the head noun plural, and re-abbreviate it. It's lexicalized. Deal with it. :)


Andreas Johansson <andjo@...>Pluralis på svensk och tys k (was: Re: Performative verbs (was: Re: here is some stuff i want all of ya'll to look at)