User:Gleki/xorlo as seen by La Gleki
Abbrevs
- UD
- Universe of Discourse
Scope generators
da series, prepositions create scope.
lo bridi can have a prenex that declares da.
All da that are not declared in the prenex of the bridi where they are used are inherited from lo bridi from higher levels: su'o da zo'u da broda lo nu da brode - here, the last da is inherited from the first da.
All da that are declared in the prenex of the bridi where they are used shadow all the da declared at higher levels: su'o da zo'u da broda lo nu su'o da da zo'u da brode - here, the last da is given a new scope, it's actually a new variable and in this inner bridi the first da is no longer valid.
Prepositions, pa da-like and pa broda-like nouns create scope to the right of each of them.
Pure da with its quantification declared in the prenex doesn't create scope.
lo-like nouns have the top priority scope within the entire UD to the left and to the right of them up to the boundaries of UD, they don't scope over each other.
lo
- This construct doesn't have a prenex thus every usage of it creates a new scope and a new variable.
- Additionally, all variables implicitly created by lo have top scope.
lo means the same as da poi except it's implicitly declared in an implicit UD-wide prenex and therefore has a different scope.
UD-wide prenex is always implicit, in other words lo broda has no prenex, it's always quantified only in the place where it is used. lo is impossible to redeclare in UD.
No matter how much context is given lo can always be rephrased using a da poi prenex wrapped into new scopes
lo plise cu kukte su'o da poi plise zo'u da kukte Apples are tasty. |
The difference is that lo doesn't require and is not capable of any explicit variable declarations.
More examples:
lo pa valsi cu se vamji ko'a rau da poi valsi zo'u da se vamji ko'a One word (any word) is worth ko'a. |
lo pa bangu noroi jai banzu no nu pa da bangu cu banzu One language (any language) is never enough. |
Pragmatic utility of lo
The only utility of lo compared to da poi is
- the ability to percolate through boundaries imposed by prepositions, da and pa broda-like nouns
As one of the consequences of that is that lo allows for free word order.
lo, da poi, pa broda similarity and differences
The basic and most important similarity is that
- nothing within UD is able to penetrate the scope of lo.
- declared da from upper scopes is not able to penetrate the scope of another da declared at levels inner to the first (upper) da. When needed goi-type assignments to intermediate pointers can be used.
Examples:
- su'o da zo'u da gasnu lo nu su'o da zo'u da brode
- da is not able to redeclare the da in da brode because the latter da is explicitly declared in the inner bridi.
- the scope of da brode is beyond the scope of the first da
- su'o da zo'u da gasnu lo brode
- da is not able to redeclare lo brode (because there are no explicit variables)
- the scope of lo brode is beyond the scope of da
- su'o da zo'u da gasnu pa brode
- da is not able to redeclare pa brode (because there are no explicit variables)
- the scope of pa brode is within the scope of da
pa broda
- This construct doesn't have a prenex thus every usage of it creates a new scope and a new variable.
Examples of scope boundaries
Some scopes are shown here with their boundaries marked with [ ... ].
- nu [mi broda] kei
- su'o da [broda de]
- mi na ku [bai ku [klama da]]
ci gerku and ci lo gerku
By a quirk of Lojban syntax, it is possible to omit the descriptor "lo", but never any other descriptor, from a description like that of Example 7.5; namely, one which has an explicit outer quantifier but no explicit inner quantifier.
— CLL 1.0
In other words, in CLL 1.0 ci gerku and ci lo gerku meant the same. After the xorlo reform the expansions are:
ci gerku = ci da poi ke'a gerku
ci lo gerku = ci me lo gerku = ci da poi ke'a me lo gerku
Hence, the difference is only in top-level declarations of dogs (by saying or implying lo gerku somewhere in UD) in ci lo gerku construct.
After the xorlo reform the difference is still very subtle to think of any useful difference between them.
le and particularity vs. any-ness
- le = non-veridical descriptor: the one(s) described as ...
- le zarci = one-or-more-specific-things-each-of-which-I-describe-as being-a-market
"le" is quite close in meaning to English "the"...
..."le" ... indicates that the speaker has one or more specific markets in mind (whether or not the listener knows which ones they are). Second, it also indicates that the speaker is merely describing the things he or she has in mind ..., without being committed to the truth of that description.— CLL 1.0
The description of the meaning of le is not very precise. Instead of
- le zarci = one-or-more-specific-things-each-of-which-I-describe-as being-a-market
I suggest this glossing:
- le zarci = one-or-more-specific-things-each-of-which-I-describe-in-this-UD-as being-a-market
Other explanations of le above are fine whether or not they are useful to turn Lojban into a colloquial language to be used to bake cookies. Namely, these explanations are fine if we assume that "the" is to be used anaphorically/cataphorically/exophorically:
"le" is quite close in meaning to English "the"
whether or not the listener knows which ones they are
In CLL 1.0 both da poi and lo can be translated as "some/any". le is opposed to both of them.
voi
le was poorly understood because people were relying on logical theories that for the most part ignored the notion of discourse (that is more related to linguistics). However, although le was more or less caught a similar particle voi was not popular.
The relation is
- noi/poi - lo
- voi - le
Here, voi starts a relative clause but unlike noi and poi it refers to things/events/etc. that the speaker has in their mind, that is that have already been created for the speaker's UD.
za'a melbi vanbi i lo vacri voi melbi vanbi cu vifne Such a nice neighbourhood. The air here is fresh. |
mu da prenu gi'e zvati le kumfa i pa pendo be mi ku voi zvati cu xanka There are five people present in the room. There is one friend of mine present there and he/she is nervous. |
Veridicality
- lo = at least one of all of those which really are
CLL deals with veridicality when showing lo/le distinction.
"veritas" implies knowledge, thus implies either jetnu2 or djuno1 and thus requires the presence of either epistemology or the subject who knows.
- for lo
at least one of all of those which really are
, the definition provided by CLL, is fine provided that veridicality is explained (even if implicitly or by context) in UD.
Hence, the whole idea of veridicality is correct and only needs pedagogical rephrasings leading to a colloquial "logical" Lojban becoming real.
outer quantifiers for predicate arguments in the discourse
Nouns to be referred to can also be quantified:
ci gerku cu batci lo remna i re le gerku cu citka lo rectu There are 3 dogs that bite humans. For two of them: each eats meat. |
Notice that when using this method "ci gerku"/"ci lo gerku" distinction is unreachable.
pre-xorlo expansions of lo, le, la
They were expanded as
- le:
- ro le su'o
- all of the at-least-one described as
- lo:
- su'o lo ro
- at least one of all of those which really are
- la:
- ro la su'o
- all of the at least one named
It's too late to take them into account due to a natural shift in Lojban usage leading to the adoption of BPFK expansions.
"Back to CLL 1.0" ideology (Neo-prexorloism)
With the explanations provided above the distinction between " lo/da and le ", which is described in CLL 1.0 (no matter how awkward or inutile the description is for colloquial usage), is nevertheless in general fine and useful for the language design and its applications (e.g. through translations into Laymen English).
In short it describes the difference between indefiniteness (lo and da) and definiteness, i.e. anaphora/cataphora/exophora.