On Fri, 16 Oct 2015 14:04:54 +0300, Gleki Arxokuna <[log in to unmask]> wrote:

>2015-10-16 13:44 GMT+03:00 Alex Fink <[log in to unmask]>:
>> I dare say that I
>> find a metaphorical case system obviously more ergonomic than a numbered
>> one.  A numbered system demands that each predicate's complementation be
>> learned separately, and is fundamentally hostile to transferring knowledge
>> of one predicate that one knows well to another unfamiliar one.
>I certainly disagree with this one. Where do you learn this metaphorical
>case system from?  From your mother tongue, supposedly a SAE one, isn't it?
>Then certainly from auxlanish purposes such metaphorical system can be
>learnable faster ("donor" is marked with Dative-like case tag etc.)

SAE is not the only source of metaphoric case systems.  If I were designing a metaphoric case system for an ergonomic log-leaning language in fully serious mode, I would start instead from an analysis like that done in the paper of Hartmann, Haspelmath, Cysouw
recently mentioned here, and do something like run a clustering algorithm on the space of microroles.

If your argument for numbered cases is "using any case system in the well-populated region of natlang space will give speakers of some natlangs an unfair advantage, so we have to do something entirely different"... whatever sort of argument that is, it's not an argument from ergonomics.

Somewhat impertinent question: is there any feature of Lojban which you think is actually suboptimal?  Your attitude towards Lojban on this list is one I might characterise as boosterism (if clear of the taboo sort of advocacy).