If the original purpose of Lojban is to test the Sapir-Whorf Hypothesis, which is that a language shapes the thoughts of its speakers, and most of the gismu are relations with English-worded definitions, isn't Lojban inevitably just as restrictive as English is as far as vocabulary is concerned?
No. The definitions include clarifiers that, for instance, salci "celebrate" includes funerals.
One can attempt to define, say, Japanese words in English, but just because you learn the English definition doesn't mean you truly grasp the nuances of the Japanese word. As, I think, xod also believes, we are in the process of discovering the nuances of the words. The definitions of the words in the gismu list aren't the real meanings of the words, merely that which can be encoded in English. --jay
Huh? Why? Not that many of the gismu correspond closely with any English word; they tend to have broader meanings. Dictionary "definitions", including the gismu definitions, are not mathematical definitions, but more like indications of the typical meanings. The real meanings live in that mushy stuff in people's heads.
This is an absolutist argument. If it held in the strong sense, then the only way to create a truly artificial conlang would be to base it on a theory which could not be described in a natural language at all! Otherwise you could say that, after all, it was only an encoding of its natural language description, and not "artificial" at all.
--But there is indeed a sense in which this is a valid
concern. If a person only learns gismu by the keyword
and not by the sumti, they will become an un-lobykai
user of Lojban, and their jbosku will tend to resemble
encoded English. All of us, in truth, start that way. But after
you have begun to learn to distinguish gismu with similar-
meaning English keywords, their true meanings will more and more
influence your usage. This is not, however, an especially Whorfish
effect--yet. That will only occur when you are perceiving in terms of those relationships...