Periods and semicolons

Tinkerers Anonymous: Some people can't help making changes to "fix" Toki Pona. This is a playground for their ideas.
Tokiponidistoj: Iuj homoj nepre volas fari ŝanĝojn por "ripari" Tokiponon. Jen ludejo por iliaj ideoj.
janMato
Posts: 1545
Joined: Wed Dec 02, 2009 12:21 pm
Location: Takoma Park, MD
Contact:

Re: Periods and semicolons

Post by janMato »

I think I'm getting your point that "ni" conveys a definiteness that "ona" doesn't. So are we trying to say that deixis in context free speech (like on the internet) where you can't point, where really all there is is a stream of words, the "this" and "ni" and so on become a way of being more definite? Insightful, I hadn't thought of that and it kind of clears up the difference between anaphora and diexis in context free speech.

However, I still don't like "ona"....
jan Ote wrote:
janMato wrote:ona is such a lousy word-- it co-ordinates with *everything*, I would rather repeat.
Context rules, as usual. Every language I know has pronouns, to avoid repetition of noun. It seems to be an important part of language and I wouldn't hesitate to use it.
Yeah, but if it isn't bringing much to the table--imho--why not drop it all together? I have to choose to use ona/ni/wan/ijo/jan, guidance is mostly left up to calquing one's mother tongue. I'm using ni/ona from time to time, but for no particular reason other than in my own language I might do so at that point (save for definiteness, ni is definitely more definite-- but if I wanted an indefinite ni, I wouldn't choose ona). It would have been more true to a pigdin design to drop the subject or object rather than use ona/ni/wan/ijo/jan. Sadly, tp has obligatory subjects and seemingly obligatory objects for vi words. I don't think at all the solution would be to add more co-ordination strategies like lojban.
janMato wrote:Odds are low that I will ever have the opportunity to point at something and say "ni!"
You forgot Po... uhm, well... You use an English equivalent of "ni" almost every time you say more than one sentence.
My point isn't that "this/the/an/it" are useless in Polish or English or Hungarian. My point is that when you take away everything else, too, such as verb inflections indicating who did it, take away contrastive pronouns (he vs it vs they), take away things that contrast based on definativeness, leaving only "ona" and "ni" which are being used (rare but it exists I'll admit) in in-person discourse, then what good is it to have one or two? Better to have zero or not use them (again unless one is conveying definiteness).
While we often use "ni" this way:
ona li utala ala. ni li pona.
 He doesn't fight. That is good.
(NB: here we are not able to use finger while speaking, and point at "something", because there is no real object to point at! The fact he doesn't fight is good)
Then this is an non-deixis usage. This is some sort of pronoun that can stand in for an entire sentence, probably as a result of being from a language that cared about genders. What gender is "he doesn't fight"? It's either neuter or better yet, invent a construction that doesn't require saying "he". My mother tongue being English, I grasp gender (of the the inanimate) poorly. So "He doesn't fight. It is good." sounds just fine. In French, I suppose they couldn't easily use "it" because it would be masculine or feminine and would have to resort to a "this" construction. I think--I'm not really good at French.
Still "ona" doesn't indicate whether the replaced noun is an animated one or not? No problem. Hungarian doesn't have grammatical gender or a grammatical distinction between animate and inanimate.
Like I said before, Hungarian hasn't been eviscerated of it's grammatical machinery and vocabulary. It has tons of pronouns that co-ordinate with anaphora on 3+ diminsions. Moreover, it the verb has gobs of personal suffixes to step in to co-ordinate things again. Toki pona has one, (three if we count ni and wan). ona co-ordinates with everything. That which refers to everything, refers to nothing. A language doesn't need to have anaphora that co-ordinates on 2 or 20 diminisions, but if it co-ordinates not at all, when why bother using that word?
janMato wrote:Pointing in toki pona is hypothetical-- it's never happened (that two toki pona speakers were in the same room)
You disappeared me and jan lili mi. We certainly have used "ni" many times. ...
[/quote]
Nothing personal, just trying to make the point that toki pona is for the most part, an online, not an in person phenomenon. This contrasts heavily with pidgin and creoles which are mostly in person only phenomenon.

There exists a language in the world that as a grammatical requirement requires you to inflect the verb to indicate if the action is up-river or down-river. If an internet language incorporated it, the distinction would be meaningless, as it is a rule that works best in person, standing by a river. Similarly, a deixic pronoun that implies pointing with a finger is going on makes no sense on the internet. If we see a deixic pronoun being used on the internet, it probably doesn't mean "The one I'm pointing at"
"ike a! ona li moku e ona!". No ambiguity within the context.
Here ona isn't really referring to anything because it isn't more descriptive than
* ike a! li moku!

The ona just serves as a place holder. ijo has been time to time suggested to be that place holder as well
ike a! ona li moku e ona! (eating is going on!)
ika a! ijo li moku e ijo! (eating is going on!)
janKipo
Posts: 3064
Joined: Fri Oct 09, 2009 2:20 pm

Re: Periods and semicolons

Post by janKipo »

Context! Context! Context! In the given context, the first 'ona' is the spider, the second the fly. Without context it is unintelligible and, indeed, ungrammatical, because there are no preceding words to replace (you can't repeat what ain't happened yet). 'ona' is useful precisely because it can replace any word ("the empty is useful because everything fits it" more or less). But, in a given context, it can only be replacing a noun from the previous couple of sentences. This still allows for ambiguity of course, though the last case cited isn't one: try "Flash turned to Ming. He hit him." (or is it Buck or Mongo? it's been a while). Since tp has no categories for nouns (like gender or even number) there is no way to have several anaphorics that are so differentiated. The only real possibilities are by initial letter (but we don't have any letter words either) or by grammatical slot (subject/DO/ obj of prep/modifier on each). Lojban has both and a number of others but still can't do a good job on many cases (the absolutely precise ones are inevitably to complex to manage on the fly). And neither is very tpish. "ni" is similar -- in texting versions, it can only apply to some previous sentence or make definite some previous reference. In the real world, of course, it can also be used to refer to stuff outside language (and is).
janMato
Posts: 1545
Joined: Wed Dec 02, 2009 12:21 pm
Location: Takoma Park, MD
Contact:

Re: Periods and semicolons

Post by janMato »

ona has 2 syllables and so do most head nouns. One could save breath by avoiding ona and repeating in some cases.

Context in a context impoverished environment like online? Maybe my point is better aimed at internet conlangs (and natural languages used online). The context is the initial set of messages in a series, and maybe I guess some inference about a collocutors unseen environment. (We probably are both on the same planet, read the same news, etc) That sort of context can be referred to demonstratively. But in the dialog itself, not so much.

Anyhow, when ever I see (what for me is) a pain point in tp, I suspect it is a likely place for future language development, were there to be a hypothetical large number of users and 100 years to develop.

jan would be come he/she/they
ijo would be come it
ni would become "the person", "the thing"
wan would become "a person", "a thing"
ona would wither away and die or take on head nouns as modiers, or just be left as a the placeholder, like "it" is in "It is raining"

re: Letters as noun class (or markers on pronouns for noun classes)? How unstable! A few hundred years of pronunciation shifts and the system would be wrecked. I figure that if a easylang is to have word classes, the categorization should be based on pragmatics and semantics, not phonetics, or worse, arbitrary groupings.
User avatar
jan Josan
Posts: 326
Joined: Sun Oct 18, 2009 12:41 pm
Location: ma tomo Nujoka
Contact:

Re: Periods and semicolons

Post by jan Josan »

Just as an experiment, Let's look at the first paragraphs of jan Ote's jan Kikamesi, and try to get rid of the ona:
ma tomo Uluku li lon. sinpin pi ma tomo Uluku li sewi li suli. sinpin ni li kiwen. tomo Enana li tomo pi jan sewi meli Inana. sinpin pi tomo Enana li kiwen. sinpin ni li walo sama suno. o tawa tomo pi jan sewi Inana! o lukin! jan ante li pali ala e tomo sama. jan lawa ante li pali ala e tomo sama. jan seme li kama pali e ijo ali ni? jan Kikamesi li jan lawa pi ma tomo Uluku. ona[1] li kama pali e tomo sewi ni e sinpin ni. ona[2] li suli. jan Kikamesi li utala sama soweli wawa. ona[3] li pakala e sinpin kiwen. ona[4] li lawa e kulupu pi jan utala ona[5]. ona[6] li jan lawa. jan seme li ken sama e ona[7]? jan seme ken toki e ni: "mi jan lawa"?

meli sewi Ninsun en jan lawa Lukapanta li mama ona[8]. jan Kikamesi li suli li pona lukin. ona[9] li wawa li sama jan sewi. jan ala li wile utala e ona[10].
[1][3][7][9][10] could be replaced with 'jan ni' but it's more verbose and doesn't gain much.
ona[1] refers to either the jan or the tomo, and in context the tomo makes no sense.
ona[3] could I suppose refer to jan K or soweli wawa, but since they are being compared as the same it doesn't really matter.
ona[9] and ona[10] could refer to (jan Kikamesi), or (jan K + meli sewi N + jan lawa L), but 'jan ni' doesn't help us here if that's how someone decided to read it. In fact, to my ears, using 'jan ni' here would make me look twice at everyone mentioned in this paragraph, where the immediacy of 'ona' makes me feel more assured that it is just jan K. That might just be me though.

[8] here you would have to use 'pi jan ni' and we are up to three words to replace one ona with no added benefit.

[2] 'jan ni' here, or 'sinpin ni' could help us know for a fact if it is jan K or the walls that are suli. Context though, makes me pretty sure these are tall walls.
You could use 'ni' here, but then ni li suli could also mean something more abstract like "this [building of the walls] was important."

[4][5]and[6] might be the best argument for the importance of ona. Look at how much worse this is:
jan ni li lawa e kulupu pi jan utala pi jan ni. jan ni li jan lawa. Maybe you have another approach to this one I'm not seeing.
jan would be come he/she/they
ijo would be come it
ni would become "the person", "the thing"
wan would become "a person", "a thing"
ona would wither away and die or take on head nouns as modiers, or just be left as a the placeholder, like "it" is in "It is raining"
I'm trying to imagine how you mean this but I couldn't come up with a good way to approach the above paragraph with these substitutions.
janKipo
Posts: 3064
Joined: Fri Oct 09, 2009 2:20 pm

Re: Periods and semicolons

Post by janKipo »

Well, yes, in a hundred years being used by mainly English speakers, a lot of those things might happen -- and, of course, are legitimate even now, but not common style. So? I personally have enough trouble figuring out how to say things in tp now, without spending my time thinking about what in may be come in my great grandchildren's day (it should live so long!). Are you suggesting right now that what you are talking about is a better system? I don't see it myself. As a minimalist, the one anaphore seems about right, especially since it is open to all kinds of supplementation as needed. Ditto the one deictor. I can, if the need arises, reproduce the effect of most anaphoric and deictic system I know (and several I only suspect) with these. It may be that some of those supplemented cases come to be more common, to join the ranks of compound compounds like 'jan pona', but I see no reason to think they will supplant the basic form. And I certainly don't see any reason to encourage the change. Of course, we get back to the basic point that I don't see 'ona' and 'ni' as weak points in tp (nothing compared to 'ala/ale/ali' say or the absence of numeral names) and so ain't driven to try to "correct" them. Since you seem to be able to use them correctly and understand them in contexts, where it counts, I don't see your problem either.
Kuti
Posts: 358
Joined: Sun Feb 28, 2010 3:48 pm

Re: Periods and semicolons

Post by Kuti »

MESSAGE DELETED
Last edited by Kuti on Thu Aug 27, 2015 10:52 am, edited 1 time in total.
janKipo
Posts: 3064
Joined: Fri Oct 09, 2009 2:20 pm

Re: Periods and semicolons

Post by janKipo »

This is, of course, pure deixis, pointing to the event/sentence just before. It is not anaphora, since there is nothing -- except, perhaps, the whole sentence -- to repeat.
Post Reply