Archive:Terminator: Difference between revisions
(more examples in terminator listing; move cy to skippables) |
m (Uakci moved page Gamma:Terminator to Archive:Terminator without leaving a redirect) |
||
(11 intermediate revisions by 3 users not shown) | |||
Line 1: | Line 1: | ||
{{Hatnote|This applied to previous versions of Toaq but does not apply to the current version. See [[#Toaq Delta]].}} | |||
:''Not to be confused with [[wikipedia:Terminator_(franchise)|Terminator]].'' | :''Not to be confused with [[wikipedia:Terminator_(franchise)|Terminator]].'' | ||
A '''terminator''' is a [[particle]] that carries no meaning whose purpose is to show where a [[phrase]] of a certain kind ends. Some [[particle]]s, like {{t|po}}, {{t|mı}}, {{tone|5}}, open phrase – for example, {{t|po}} expects a [[noun phrase]] to immediately follow, and {{tone|5}} opens a [[clause|full “nested sentence”]]. And sometimes it is necessary to specify where that phrase ends, for unambiguity’s sake. | A '''terminator''' is a [[particle]] (that carries no meaning) whose purpose is to show where a [[phrase]] of a certain kind ends. Some [[particle]]s, like {{t|po}}, {{t|mı}}, {{tone|5}}, open a new phrase – for example, {{t|po}} expects a [[noun phrase]] to immediately follow, and {{tone|5}} opens a [[clause|full “nested sentence”]]. And sometimes it is necessary to specify where that phrase ends, for unambiguity’s sake. | ||
== A common example: {{class|po|object-incorporating verb}} phrases == | == A common example: {{class|po|object-incorporating verb}} phrases == | ||
Line 58: | Line 59: | ||
|- | |- | ||
| colspan=3 | | | colspan=3 | | ||
| ╱<br>CopV<sub>F</sub>P | |||
| | | | ||
|   <!-- strut --> | |   <!-- strut --> | ||
Line 64: | Line 65: | ||
|- | |- | ||
| | | | ||
| ╱<br>CopVP | |||
| colspan=2 | | | colspan=2 | | ||
| ╲<br>CopV<sub>F</sub> | | ╲<br>CopV<sub>F</sub> | ||
| | | | ||
| △<br>{{t|tỏa}} | |||
|- | |- | ||
| ╱<br>CopV | | ╱<br>CopV | ||
| | | | ||
| ╲<br>DP | |||
| | | | ||
| | | │<br>{{t|ga}} | ||
|- | |||
| │<br>{{t|pỏ}} | |||
| | |||
| △<br>{{t|jí}} | |||
|} | |} | ||
(Note: some internal CopVP structure was omitted for brevity.) | (Note: some internal CopVP structure was omitted for brevity.) X<sub>F</sub>P is an arbitrary convention for a phrase that terminates XP; see [https://github.com/toaq/zugai/issues/10 this] issue on the [[zugaı]] issue tracker for normative diagrams. | ||
In fact it would not be far-fetched to remove terminators from a future version of Toaq in their entirety. This quote due to [[And Rosta]]: | |||
: I agree both that spoken parentheses (as they are described and conceptualized by the loglangs that employ them) are unnatural and that they can be reconceptualized as something more natural. What's less redeemable, I think, is the unnaturalness of the conditions on the so-called elidability of the terminators. As a loglanger I should hardly carp at unnaturalness, but I would at least want the conditions to be explicitly specified and demonstrated to be ergonomic. My introspective sense is that it's easy enough to insert/unelide terminators at a kind of editing and proofreading stage, a stage at which you tweak the text to avoid misparses by the reader, but not at the first stage of composition, nor easy to formulate the rules of elidability. | |||
== Avoiding terminators altogether == | == Avoiding terminators altogether == | ||
It | It usually suffices to place heavy constituents on the far right of a sentence, corroborating Toaq’s right-branching syntax, and if necessary move any interlopers to the [[topic]] (which is terminated by {{t|bı}} – but this terminator usage is fine because a topic phrase is not ''introduced'' by any particle). Here’s an example of these repair strategies in use: | ||
<blockquote><poem> | <blockquote><poem> | ||
Line 124: | Line 132: | ||
|- | |- | ||
| {{class|lu|free relative clause (as in “the one who …”)}} || {{t|ky}} | | {{class|lu|free relative clause (as in “the one who …”)}} || {{t|ky}} | ||
| {{t| | | {{t|Mảı jí sa <u>lủ mảı hóa fúy</u> da}} ''I love somebody who loves me'' || {{t|<u>Lủ mảı hóa jí ky</u> nháo da}} ''They are such that they love me'' | ||
|- | |- | ||
| [[clause]] ({{tone|3}} or {{tone|5}}) || {{t|cy}}<ref group=note name=type-2-termination /> | | [[clause]] ({{tone|3}} or {{tone|5}}) || {{t|cy}}<ref group=note name=type-2-termination /> | ||
| {{t|Dảı <u>rûaq jí róu</u>}} ''It’s possible that I state it'' || {{t|Dảı <u>rûaq jí cy</u> róu}} ''Given this, it’s possible that I state'' | | {{t|Dảı <u>rûaq jí róu</u> da}} ''It’s possible that I state it'' || {{t|Dảı <u>rûaq jí cy</u> róu da}} ''Given this, it’s possible that I state'' | ||
|} | |} | ||
Line 155: | Line 163: | ||
| {{t|Gảı jí sa rủa da, ju kủao máq da}} ''I saw a flower, which was green.'' | | {{t|Gảı jí sa rủa da, ju kủao máq da}} ''I saw a flower, which was green.'' | ||
|} | |} | ||
==Toaq Delta== | |||
Toaq Delta no longer uses the terminators described as skippable above. It also does not use {{class|da|illocution}} to terminate {{T|ju}}. | |||
For {{class|po|object-incorporating verbs}} and {{class|mı|name marker}}, they now come after what they describe, as with all other adjectives. | |||
{| class="wikitable" | |||
!Toaq Delta | |||
!Toaq Gamma | |||
!English | |||
|- | |||
|{{t|nıaı po súq}} | |||
|{{t|<u>pỏ súq ga</u> nỉaı}} | |||
|your animal | |||
|- | |||
|{{t|káto mı mıao}} | |||
|{{t|<u>mỉ mỉao ga</u> kảto}} | |||
|“Moon” the cat | |||
|} | |||
For subordinate clauses and relative clauses, [[self-termination]] is applied, meaning if an argument is added after a subclause or relative clause has already been “filled,” the argument will go to the level above. An unfortunate side effect of this is that now, one cannot attach an adverbial to a main clause that ends with a subordinate clause because it would attach to that subclause instead. The same is true for relative clauses. Instead, one should move the adverbial to come after the verb but before the first argument. | |||
{| class="wikitable" | |||
!Toaq Delta | |||
!Toaq Gamma | |||
!English | |||
|- | |||
|{{t|<u>Ꝡe maı hóa jí</u> nháo da}}||{{t|<u>Lủ mảı hóa jí ky</u> nháo da}} | |||
|They are such that they love me. | |||
|- | |||
|{{t|Paı káto <u>ꝡë zudeq hóa Tóaqzu</u> tú da}}||{{t|Pảı káto <u>züdeq hóa Tóaqzu cy</u> tu da}} | |||
|The cat who speaks Toaq is a friend to all. | |||
|- | |||
|{{t|Ca <u>é tı súq ní</u> <u>ꝡä jaı jí</u> da.}}||{{t|Cả <u>tî súq ní cy</u> <u>jâı jí (cy)</u> da.}} | |||
|You being here makes me happy. | |||
|} | |||
{{t|Mo}}, {{t|kıo}}, and topics and still terminated by {{t|teo}}, {{t|kı}}, and {{t|bï}} (previously {{t|bı}}), respectively. |
Latest revision as of 20:21, 4 November 2023
- Not to be confused with Terminator.
A terminator is a particle (that carries no meaning) whose purpose is to show where a phrase of a certain kind ends. Some particles, like po, mı, , open a new phrase – for example, po expects a noun phrase to immediately follow, and opens a full “nested sentence”. And sometimes it is necessary to specify where that phrase ends, for unambiguity’s sake.
A common example: po phrases
Say we wanted to translate “a cat’s claw” to Toaq. “A cat’s” would be po baq kảto. “Claw” is choıceaq. So our first idea would be to put one right next to the other like an adjective, forming a serial:
*po baq kảto chỏıceaq
a cat-claw’s [something]
But this is incorrect because the noun phrase that po expects to fit in its bellows never ends! So baq kảto chỏıceaq ends up forming one noun phrase, meaning “a cat-claw” (something that is a cat and a claw at the same time 🤔). In this case we need to signal that kảto needs to go inside po, whereas chỏıbeaq needs to stay outside of it. This is where terminators come into play: ga, the terminator for po, serves as a fence guarding the content (the noun phrase) inside from the content (rest of serial verb) outside:
po baq kảto ga chỏıceaq
a cat’s claw
(Green marks the extent of the po-phrase – so in a sense, ga cuts it short.)
Skippable terminators
Some terminators are skippable (also known as elidable/elidible or omissible in loglanging circles) – they don’t need to appear every time the corresponding particle appears. po’s ga is an example of such a terminator. For example, while it is perfectly fine to say
Pỏ jí ga ní rủa da.
This flower is mine.
we can safely remove the ga without any repercussions:
Pỏ jí ní rủa da.
This is because after we see jí, a complete noun phrase, we encounter ní rủa – yet another noun phrase – which can’t possibly fit inside the po-phrase because po only ever asked for one noun phrase. This is different from the last example because chỏıbeaq could expand the baq kảto noun phrase by forming the serial verb kảto chỏıbeaq. It follows that ní rủa has to stay outside of the po-phrase, which ends prematurely.
Other terminators are always expected to be there, for example kıo’s kı. The two are used for parenthetical statements (kıo like this one! kı) and it wouldn’t make sense to allow the kı to be skipped (elided).
Some phrases don’t need terminators at all – for example, ju (which opens a parenthetical statement) ends as soon as the statement (a full sentence) ends (with an illocution like da). In this case we could also consider da as terminators of ju (that depends on whether you require terminators not to carry any meaning but be purely structural).
How to learn these?
There are two schools of thought on how to teach terminators, and both are equally valid (but it’s hard to say which one is “better”):
- Pretend every particle that has a terminator, skippable or not, intrinsically implies that terminator being there – for example, we would say that the grammatical structure for possessives is not po ___ but po ___ ga. Then specify that some of these terminators may be skipped under some circumstances.
- Don’t include terminators in the definitions of grammatical structures. Warn whenever one phrase bleeds into another (and introduce terminators as a stopgap mechanism for such occasions).
With (1) learners end up terminating too much; with (2), too little. (1) does not cause ungrammaticalities, but introduces yet another delay for the learner to produce sentences in the language as it is actually spoken (and spoken Toaq is not replete in terminators, which are in general avoided; see below section). (2) does engender ungrammatical productions, and also causes nuisance down the line as terminators are not easy to explain, but gets the learner much closer to actual Toaq 90% of the time.
Attitudes towards terminators
The practice of having constructs come in starter–terminator pairs like po–ga to aid in unambiguity has a precedent in Lojban. However, Lojban’s terminators are much more essential and commonplace since it has a much more nested and Byzantine structure, whereas Toaq is relatively flat.
Hoemaı considers terminators a necessary evil as they are not easy to codify in formal syntax terms, or at least come with a lot of useless syntax: a terminated phrase like po ___ ga ___ requires two heads, one for the po (which is semantically bearing), one for the ga (which is semantically vacuous):
VP | ||||||
╱ CopVFP |
╲ V | |||||
╱ CopVP |
╲ CopVF |
△ tỏa | ||||
╱ CopV |
╲ DP |
│ ga | ||||
│ pỏ |
△ jí |
(Note: some internal CopVP structure was omitted for brevity.) XFP is an arbitrary convention for a phrase that terminates XP; see this issue on the zugaı issue tracker for normative diagrams.
In fact it would not be far-fetched to remove terminators from a future version of Toaq in their entirety. This quote due to And Rosta:
- I agree both that spoken parentheses (as they are described and conceptualized by the loglangs that employ them) are unnatural and that they can be reconceptualized as something more natural. What's less redeemable, I think, is the unnaturalness of the conditions on the so-called elidability of the terminators. As a loglanger I should hardly carp at unnaturalness, but I would at least want the conditions to be explicitly specified and demonstrated to be ergonomic. My introspective sense is that it's easy enough to insert/unelide terminators at a kind of editing and proofreading stage, a stage at which you tweak the text to avoid misparses by the reader, but not at the first stage of composition, nor easy to formulate the rules of elidability.
Avoiding terminators altogether
It usually suffices to place heavy constituents on the far right of a sentence, corroborating Toaq’s right-branching syntax, and if necessary move any interlopers to the topic (which is terminated by bı – but this terminator usage is fine because a topic phrase is not introduced by any particle). Here’s an example of these repair strategies in use:
Cả rôq nháo cy rôq mao jí da.
That they cried made me cry too. (Ugly: there is a heavy constituent rôq nháo (cy) on the left. We need to make do with a terminator, plus the sentence is much less nice to read.)
↓
Rôq nháo bı, cả róu rôq mao jí da.
As for them crying, it caused me to cry too. (Much nicer, if longer.)
Rỏq nháo cà rôq mao jí da.
They cried, causing me to cry too. (Short and sweet, but we’ve had to pay the cost of changing the meaning of the sentence.)
Mảı pó pó jíbo pỉa ga pảı ga sẻo sa mẻaheo da.
My sibling’s friend’s spouse loves somebody else. (😖 Terminators galore!)
↓
Pó jíbo pỉa ga pảı bı, mảı zébo sẻo sa mẻaheo da.
As for my sibling’s friend, their spouse loves somebody else. (Better, and an actual repair strategy used in natural languages when the nesting becomes insufferable, but we’ve still not gotten rid of ga.)
Séo pö páı pö jíbo pỉa bı, mảı séo sa mẻaheo da.
As for the spouse of the friend of my sibling, they love somebody else. (We still do copious nesting, but at least bı is there to take it all down. If this sort of noun phrase appeared in the middle of the sentence, like Mảı séo pö páı pö jíbo pỉa cy cy sa mẻaheo da, we’d be back at cascading terminators like the cy cy here. So best to keep this sort of nesting in the topic, where it’s easiest and most natural to annul.)
Jíbo pỉa hóbo pảı hóbo sẻo bı, mảı hó sa mẻaheo da.
As for my friend, as for their friend, as for their sibling, they love somebody else. (Long but very easy to follow.)
The takeaway: prefer short and simple sentences to long and convoluted ones. Sometimes it’s best to distribute the information over multiple sentences (like in Modern Greek) rather than try to pack them into one large one. And if you need to right-branch early, bı is your friend.
List of terminators
Skippable
Structure | Terminator | Example (underline shows how far the phrase extends) | ||
---|---|---|---|---|
Without terminator | With terminator | |||
po | ga | pỏ súq nỉaı you, an animal’s | pỏ súq ga nỉaı your animal | |
mı | ga (unofficially, ceı) | mỉ mỉao kảto “Moon-Cat” | mỉ mỉao ga kảto “Moon” the cat | |
lu | ky | Mảı jí sa lủ mảı hóa fúy da I love somebody who loves me | Lủ mảı hóa jí ky nháo da They are such that they love me | |
clause ( or ) | cy[note 1] | Dảı rûaq jí róu da It’s possible that I state it | Dảı rûaq jí cy róu da Given this, it’s possible that I state |
- ↑ While not officially confirmed or denied, clauses seem to be terminable with adverbials. For example, in Cả rôq nháo põı jảq mẻo rôq mao jí jẽo nãı da That they cried in great sadness now caused me to cry as well, the noun phrase rôq… can’t possibly be included in the first rôq… clause as it appears after põı jảq mẻo, and adverbs go at the end of a clause – and nãı is similarly ousted to the outer clause because it’s preceded by jẽo, which is type-2 (and type-2 always follows type-1). There is a proposal for supplanting cy with nã, where na would be a vacuous type-2 adverbial (jeo?).
Apart from all these, hu, which are used for vocatives, are sometimes understood as terminated with ga. However this would seem unnecessary as hu are commonly used between sentences, in which case you may use lả, e.g., hu káto lả chủq súq ba O cat, you should eat!.
Non-skippable
Structure | Starts with | Ends with | An example just for you |
---|---|---|---|
full-text quote | mo | teo | Kủq nháo mó «Mảı jí súq da» teo da. They said, “I love you”. |
parenthetical statement | kıo | kı | Mả tỉjuı súq moq? (kıo ỉq jí da kı) Are you close by? (I am) |
topic | (start of clause) | bı | lâ seku bủ bı, chỉ bủ jí …that as for most false things, I deem them false |
incidental statement | ju | da | Gảı jí sa rủa da, ju kủao máq da I saw a flower, which was green. |
Toaq Delta
Toaq Delta no longer uses the terminators described as skippable above. It also does not use da to terminate ju.
For po and mı, they now come after what they describe, as with all other adjectives.
Toaq Delta | Toaq Gamma | English |
---|---|---|
nıaı po súq | pỏ súq ga nỉaı | your animal |
káto mı mıao | mỉ mỉao ga kảto | “Moon” the cat |
For subordinate clauses and relative clauses, self-termination is applied, meaning if an argument is added after a subclause or relative clause has already been “filled,” the argument will go to the level above. An unfortunate side effect of this is that now, one cannot attach an adverbial to a main clause that ends with a subordinate clause because it would attach to that subclause instead. The same is true for relative clauses. Instead, one should move the adverbial to come after the verb but before the first argument.
Toaq Delta | Toaq Gamma | English |
---|---|---|
Ꝡe maı hóa jí nháo da | Lủ mảı hóa jí ky nháo da | They are such that they love me. |
Paı káto ꝡë zudeq hóa Tóaqzu tú da | Pảı káto züdeq hóa Tóaqzu cy tu da | The cat who speaks Toaq is a friend to all. |
Ca é tı súq ní ꝡä jaı jí da. | Cả tî súq ní cy jâı jí (cy) da. | You being here makes me happy. |
Mo, kıo, and topics and still terminated by teo, kı, and bï (previously bı), respectively.