Indeed... I've remarked on this before. If knowing what a word means
requires that the listener... well, be you, (since nobody else can
deduce the meaning from the word), then how does that enable
communication? And if it doesn't communicate, what linguistic purpose
does it serve?
That's really the issue with fu'ivla in general, and especially with
unmarked ones. Marking them with a rafsi (like bakrto) is at least a
start, but in a case like this where the whole issue is _which_ specific
bakni you are talking about, it doesn't do much.
I could maybe see an argument for default naming of animal/plant species
by fu'ivla derived semi-obviously from their standard Linnean binomials
(though even then it would be a pain for most people to look them up),
since there is at least a standard. But when faced with a word whose
source could literally be *any* language (that Pierre has heard of), and
possibly even modified idiosyncratically (.alzaitu.? Well, stuck the
al- article from Arabic to the Arabic word for olives (zayt) in order to
get the consonant cluster. And how was I supposed to guess that?)
If you mean to say things to people, they need to have SOME way to work
out what you said.
~mark