[lojban] Re: New PA-proposal

[djandus <jandew@gmail.com>]

This seems really impressive and well-thought-out to me and deserves careful inspection.

I have to admit, from the start I was afraid that your demands for the grammar would make it far too complicated, but it actually seems rather intuitive. Just to make sure I understand things, I'm seeing that to make a string:

PA1 combine freely

PA2 exist individually

PA3 modify the string to the right

PA4 separate strings, modifying internally

PA5 are PA4 but with single usage

PA6 are specialized individual numbers

Just looking at that, the "pretty pretty simple simple" part of me wants to sort them primarily by basic grammatical usage. That is, combine PA2 and PA6 since they both have words that stand alone as number strings, and combine PA4 and PA5 since they both modify internally. The extra grammatical rules can be explained in subgroups. (e.g. refer to PA6 as PA2* or something) To clarify further, I feel that while it is certainly a valuable goal to make every grammatical PA construct be meaningful, it is also lofty and should be sacrificed when necessary for the sake of having a primary grammar that is clear and easy to understand. Additionally, the secondary rules are more particulars on interpretation or restrictions to provide a computer than a human. (As in, expressing that the words you call PA5 can only be used once is extremely simple to tack on as an extra subcondition, while it is more valuable for structure and grammar to group it with PA4)

To illustrate my point, I'll use notation entirely distinct from selma'o -- I'll use "#" for an unmodified number string, which can be composed of what you call PA1 (any combination), PA2, or PA6. I'll use "." for an internal modifier, and "-" for a left-hand modifier.

With this simple categorization, I'd like to ask for clarification on your more interesting grammar suggestions. You claim that right-grouping is necessary, but the example you provided seems (to me, correct me if I'm wrong) to illustrate left-grouping! That is, I'm getting the gist of your grammar suggestion to be that you can have any number of number strings automatically associated together. (Providing there is some way to tell them apart, implying in my mind that either another internal separator or number string terminator is desired.) That is:

# # # #

is parsed as having four number descriptions all describing the same amount. Additionally, I see you applying a primarily left-grouping methodology with PA3, saying that you can have any number of left-hand modifiers "prefixing" any string:

+++++# +++++#

would have two strings with five modifiers each, both of which are numerical descriptions of the same amount.

Additionally, I would describe the internal modifiers less as separating numerical strings, and more like joining strings together. (Though this is simply a matter of perspective.) Either way, that part says you can simply have

#.# #.#.#

Which would be two numerical descriptions, one composed of two substrings and the other composed of three.

This sort of grammar seems pretty simple, though implementing it certainly would require a few very critical editations to the current grammar. (I am certainly fond of fi'u for the golden ratio -- something that could easily be accommodated for in this grammar with the use of a special separator/marker/terminator of some kind.)

Before I close, I'd like to address your example of {li rau su'o pa} with how I see your grammar would be implemented. {rau}, being of your PA2, represents a full string (#). {su'o}, of your PA3, is a left-hand modifier (+). {pa} is, of course, PA1, and without any adjacent PA1 represents a full string (#). Thus, the structure is simply

# +#

with the two descriptions "enough" and "at least one" purported as equivalent descriptions of the same value. With this outlook, there is no issue of "left grouping" or "right grouping", as there are simply stacked left-hand modifiers that only modify the next string. One issue would be declaring that the internal modifiers are "close grouping", with {pi} and {pi'e} grouping before left-hand modifiers as expected. Otherwise, the grammar would be complete, intuitive, and often more so than the current, if your {li rau su'o pa} assessment is accurate. (I apparently don't know the current PA grammar well enough, as it simply seems silly to me to try to group {<rau su'o>}.)

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/RImbxLmyHV0J.
 
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/lojban?hl=en.