[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] rewrite rules

Good project. I suggest you put those in a public Google spreadsheet, so people can add alternative definitions (or suggestions of improvements) in additional columns. The expansions of JOI available on the BPFK pages are absent from your list.

Ultimately you can express all that Lojban can express with a very limited set of cmavo, but by using only those, utterances get extremely verbose.

For example, {ro mlatu cu mabru gi'e se tuple vo da} can be rephrased as:

{roldza fa lo ka zilkanxe fa lo du'u ce'u goi ko'a mabru ku lo du'u li vo kaidza lo ka ce'u tuple ko'a}

which can be further impoverished in cmavo diversity by removing those FA and expanding counting quantification (vo da / li vo kaidza…) using only existential quantification, but that'll make thing even much more verbose.

Do you know of the Tersmu project ( https://gitorious.org/tersmu/tersmu/ ), whose purpose is to translate between Lojban and a logic notation? It's not a complete project (many cmavo aren't yet handled properly) but it's already pretty good. If you have access to the Lojban rooms on IRC, Telegram, Slack or Discord, you can a Tersmu interface available there to test it without installing anything.

—Ilmen.


On 08/08/2017 16:33, vpbroman@gmail.com wrote:
Because of my minimalist preferences, I'm interested in seeing how small a subset of lojban would be capable of expressing the same things as the full language.
If a hard word can be completely replaced by a rephrasing with easy words, then that word is useful for conciseness, but is otherwise dispensible.
If you had rewrite rules to macro expand words in terms of other words, and if you ordered words from complex to simple and required that the words in the rewrite be simpler than the word being rewritten (to avoid self-referential loops) then you don't infinitely regress, you eventually expand everything to its simplest possible terms.

This is like the concept of Semantic Primes, which alludes to the prime factorization of numbers.
As an inspiring example, Schemers point to how all computations can be expressed in terms of: lambda, recursion, if, quote, cons, car, cdr, and maybe a thing or two more -- I forget.

So, content words in lojban get defined in a dictionary in terms of other words.
But I'm interested in how our function words are defined in terms of other words.
Lots of examples are in CLL and the BPFK docs, such as this from xorlo.

lo broda ku = zohe noi keha broda kuho

I've collected all the rewrites that I could find (plus a handful I made up) and listed them in this rough draft document.
cmavo_equiv.txt
https://app.box.com/s/hjis4vnshj8tvnu2q2vg7hkcfh65fcyc

Suggesting additions/corrections would be appreciated.

Many are inexact equivalents, for example BAI is defined in terms of fiho, even though BAI fits syntactically in a number of places where a fiho phrase is not allowed.
Most of UI can be roughly equivalent to a sei-sehu phrase, at least when the UI appears at the start of the sentence, but sei-sehu doesn't seem to comment on anything more localized than the entire bridi. E.g.

ui = sei mi lo nu nei ca gleki sehu

Except ui can attach to individual words and indicate pleasure in them more specifically, not just in the whole statement.
My rewrites for UI->sei are very rough and incomplete, but they are paired with lojban definitions from jbovlaste.

The prosumti and probridi seem harder to rewrite. I wonder whether the following would make sense.

mi = lo cusku be luhe nei
do = lo te cusku be luhe nei

mihe bremenli

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.