For several weeks now I have been reading what strike me as stranger and stranger claims about Xorban. This leads me to wonder if they seem strange to me simply because I have misunderstood what Xorban is about. So I have decided to say what I have taken the project to be and what I understand that to entail. I hope this will lead someone to say where I have misled myself about what the project is up to or what it is taken to involve.
A loglan is a language readily usable by humans -- as opposed to by, for pertinent examples, machines and Vulcans -- which displays unequivocally the logical form of the utterances made in it. The display need not be transparent, the form immediately perceivable, but the discovery process must have a unique (and correct) product for each utterance, and must be
one that a normal person could carry out in a reasonable time. The logical form is defined by formulae in a developed logical language. JCB, who was not actually building a loglang for all that happened later to his project, chose FOPL, which, to be fair, was about all that was available to him. Since that time, many more systems have been developed (the current edition of the
Handbook of Philosophical Logic runs to over 6,000 pages and many items are mentioned only in bibliographic notes) to cover logically significant matters that FOPL does not. There are also many not logically significant utterances which are vital to language but which have not place in logical systems. JCB dealt with new items that came to his attention by either working them into the syntax of the underlying logical system or by slapping them onto his language ad hoc. In this he was on the right track,
since most of the later logical systems have developed their languages by expanding syntactic categories in FOPL or by just adding stuff on to it, with rules to explain where it fits in. So, now, starting a loglang, we can take a logical system that covers all we thing it should cover (which probably means about as broad a system -- or rather set of systems, since no one system covers all of language usage -- as possible). I suppose that this is an intensional logic with added pieces for questions and commands, for various other speech acts, and for epistemic and deontic structures (at least -- but these may be within the predicate structure), and presumably a little set theory (C and L). As noted, we could start from FOPL and build, but the results are likely to be chaotic (Loglan and Lojban, e.g.) in various areas.
I take the Xorban project to be to build a loglang better than Loglan or Lojban (or gua/spi, for all I
know).
As & has pointed out, making a loglang is almost trivial: you just need to assign words to all the dingbats of the formal language. Ta DA! Poof! Ah, but the humanly usable part. Logical systems produce formulae of extreme length and incredible redundancy (even with the friendliest notation) and are meant to be written down so that one can skim and does not have to remember details until checking is needed. So the real task of loglang construction is compressing the log into a lang, producing something that keeps the structure available but without the repetitions or the long separations between first and second mentions of the same things.
It seems to me that doing that optimally requires having a pretty good idea of what sorts of things you will have to deal with and in what sorts of contexts. It is mainly at this point then that I have trouble with the current work. We keep seeing thing like "We
will have no use for properties", which suggests that the intensional logic, which features properties as a major category, has not been looked at. Or "We can treat these all the same" said of things of radically different structures. Or "We don't need to mark intensionality" -- in an intensional logic!
I am surely missing something here, but I would like an explanation of what it is.