A while ago, I posted a thought on what I termed conlanging by fiat. I did not provide any really clear example, since I am pretty bad at coming up with such things - especially ones that seem convincing and realistic. (I could easily come up with contrived ones, but there's a reason I don't use that, and it starts with con- and ends with -trived.)
As luck would have it, I ran into such an example in a recent discussion about polysynthetic languages. The discussion had other terminological confusions in it, and I am really happy it came up, as some confusion was resolved, and it seems people were willing to learn - a commendable trait, due to its infrequent appearance in human populations in general. I have asked for permission to use this example by its originator, and I really appreciate their cooperation on this.
So, the question was what to use instead of cases, if one wanted not to have them in a polysynthetic language. A suggestion was presented: use suffixes that are invariant. (In other terms, suffixes with only one allomorph per suffix.) It is superficially similar to cases, but it is not the same thing.
The reasoning behind this is based on a mistaken notion of what a case is, but the interesting thing is the approach, and I will exaggerate the approach taken in that discussion a bit - beyond what my source did at the very least, but not by far:
Have a set of suffixes that are applied to nouns, and mark their role in the sentence. Assert that these are not case forms. Say your language therefore manages to mark sentence roles efficiently, but without a case system.
Anyone see the problem? We have essentially created cases, and the only thing that makes our forms non-cases is our assertion that they indeed aren't cases.
No comments:
Post a Comment