Re: [saweli] Saweli Dictionary

 

2017-11-24 12:23 GMT+03:00 MorphemeAddict lytlesw@hidden.email [saweli] <saweli@yahoogroups.com>:

 

I share your attitude regarding Lojban's argument structures. They seem almost completely arbitrary after the x1 (subject) is picked. At least in Loglan it seems that there was some attempt to classify the arguments eventually by the role they play in a predicate/clause, but I don't think the arguments were ever assigned a default order, e.g., possibly because there were so many meanings for the various arguments, which Latejami/Saweli-Saxita resolve by having core arguments vs. oblique ones, which are always marked by prepositions (case tags). Lojban has BAI words for much the same reason, but they are used sporadically and don't add any clarity or order to the argument structures of the language, i.e., they don't make it any more systematic and hence predictable. 

https://mw.lojban.org/papri/Why_positional_case_system

 

On Fri, Nov 24, 2017 at 1:22 AM, Geoff Hacker geoff.hacker@hidden.email [saweli] <saweli@yahoogroups.com> wrote:

 

Not quite.

To begin with, I can certainly see how it's something that Latejami almost already does, just because you could use its absolute argument suffixes on absolutely any root, including attaching an argument structure suffix to a root that merely mirror's the root's default structure. Whilst this is redundant, it is also explicit. So the only difference that there would be to Latejami if you removed the default place structures would be that you put the place structure suffixes on everything, and this would at least make it easier to remember what argument structure a word had, because you could always determine it by the suffix on the root. It makes the language more verbose, but it was already pretty verbose anyway.

On the other hand, because Saweli-Saxita's suffixes change depending on the default place structure of the root, it is impossible to understand what they mean without knowing that place structure, so the language is already far more dependent on that default structure than Latejami. If you were going to get rid of the default argument structure here, then you could construct places morpheme by morpheme, but there would then be little difference between this approach and the approach taken by Latejami, just because Latejami already uses mnemonically compositional suffixes.

Anyway, I do think that both Latejami and Saweli-Saxita have a huge improvement over Lojban just because the place structures are so much more flexible. And at the end of the day, I doubt that any two people would be happy with quite the same conlang anyway. I'm just glad to see that such a well thought-out and comprhensive work like Latejami is still having an influence out there. 

Geoff

On Fri, 24 Nov 2017 at 15:02 MorphemeAddict lytlesw@hidden.email [saweli] <saweli@yahoogroups.com> wrote:

 

If I understand correctly what you're saying, it seems that that is what both Latejami and Saweli-Saxita already do. 

stevo

On Thu, Nov 23, 2017 at 1:38 PM, geoff.hacker@hidden.email [saweli] <saweli@yahoogroups.com> wrote:

 

For example, I often thought that the Lojban place strucures were fairly arbitrary. Why would one argument be an agent while another an abstraction? And for that matter, I think that Rick Morneau himself provides a good example of a root that he reuses extensively, "zog". 

The argument structure is just an entry point into most concepts, and when people impose a default structure onto a concept I often think that it's because they are trying to map it onto the root that their own language provides. But that is what the Lojbanists would call a "forced choice". You might find that some concepts lend themselves more easily to some argument structures than others, but I would argue that the argument structure is every bit as much a part of the concept as any "root" (which is to say that there is no true "root" distinct from its argument structure) and it would then be interest to see how much the meaning of the innermost morpheme can change depending on the different argument structures that you give it.