Open problems in translation: Difference between revisions

From The Toaq Wiki
(moved unicorn example to a different section)
(add {{t|dujı}} solution cuz im cute)
 
(18 intermediate revisions by 2 users not shown)
Line 1: Line 1:
The following is a list of English sentences for which there is no widely accepted Toaq translation. The list is broken into three sections: sentences for which no translation is known at all, sentences for which translations are known but are impractically length or unacceptable for some other reason, and sentences for which acceptable translations have since been found.
This page gives a list of English sentences and phrases for which there is no widely accepted Toaq translation. The list is broken into three sections: sentences for which no accurate translation is known at all, sentences for which accurate translations are known but are impractically long or unacceptable for some other reason, and sentences which used to appear in one of these two categories but for which acceptable translations have since been found. Each unsolved sentence is accompanied by some discussion of the difficulty, possibly including some incomplete or not widely accepted solutions.


===Sentences that have not been translated===
The existence or nonexistence of a solution is determined relative to official grammatical features of Toaq. Unofficial predicates are allowed, but sufficiently technical utility predicates should be considered grammatical features in spirit.


* "If I owned a unicorn, I would feed it."
==Sentences that have not been translated==
** {{T|She hâqdoa jí tu shıtuaq bö jí hó, keo bo jí sıa hao da.}} has to be phrased much differently than the original (enough so that it might not even mean the right thing (?)), and is long.
** {{T|Baq guo sü kîaı jí sa shıtuaq bı, mu paq gúo hâqdoa jí shítuaq päqtao gúo da.}} demonstrates the seed of different kind of phrasing, but is very long and awkward as-is. The idea is to have a single mechanism that covers both "if a unicorn existed, it would have a horn" and "if an event of me having a unicorn existed, it would include me feeding it" with some mechanism for saying "X's have property Y" even when X's don't really exist, and then to make X "events of me having a unicorn" and Y the property "the haver of X gives food to the havee of X".


===Sentences that have not been translated well===
==="It's as though..."===


* "I know the names of many things."
"You keep insulting me, as though you think it will make me change my mind."
** There are several attempts at this
*** {{T|Sa puı tu shı mëa púı bı, dua jí chûa hı shí da.}} works, but is quite long.
*** {{T|Sa puı bı, dua jí mû chua tu shı mea púı hı da.}} works if you allow {{T|tu}} to scope over {{T|hı}}, but it's also long.
*** {{T|Hıdua chua jí sa puı da.}} almost works if you make {{T|chua}} nondistributive, except that even the modified version of the predicate can't say which names go to which things.
*** {{T|Puy shı bı, dua chua sa shí da.}} works if you allow {{T|puy}} to be {{T|puı}} in SA.
*** {{T|Chuadua jí sa puı da.}} works but it isn't clear how this would generalize to other cases.
*** {{T|Shıjeo sa puı dûa chua jí sa ja da.}} works with {{T|shıjeo}} meaning "every individual among X satisfies Y".


===Solved sentences===
The hard part here is the "as though". What predicates are involved in an "as though X" construction?
 
==="If I owned a unicorn, I would feed it"===
 
Attempts:
 
* {{T|She hâqdoa jí tu shıtuaq bö jí hó, keo bo jí sıa hao da.}} has to be phrased much differently than the original (enough so that it might not even mean the right thing (?)), and is long.
* {{T|Baq guo sü kîaı jí sa shıtuaq bı, mu paq gúo hâqdoa jí shítuaq päqtao gúo da.}} demonstrates the seed of different kind of phrasing, but is very long and awkward as-is. The idea is to have a single mechanism that covers both "if a unicorn existed, it would have a horn" and "if an event of me having a unicorn existed, it would include me feeding it" with some mechanism for saying "X's have property Y" even when X's don't really exist, and then to make X "events of me having a unicorn" and Y the property "the haver of X gives food to the havee of X".
 
==="Foxes are small to medium-sized creatures"===
 
Attempts:
 
* {{T|Nuı rVV saotuao nıaı baq hupı da.}} requires a new conjunction {{T|rVV}} which produces the appropriate meaning when used in serials. It's unclear (to me, Hoaqgıo, as I write this) how the conjunction would work, because it doesn't seem like there's a meaning for it to have when joining two noun phrases, unless they both happen to be kinds.
* {{T|Rıe nuı roı saotuao nıaı baq hupı da.}} requires {{T|roı}} to be allowed in serials and for some mechanism to be put in place that gives it reasonable behavior, including correct behavior in this case.
 
==Sentences that have not been translated well==
 
==="a very large thorn"===
 
This translation exercise is a phrase rather than a sentence. The phrase is "very large thorn", and although it seems like a perfect use of Toaq's serial predicates, it is not possible to achieve the desired meaning through official serial behavior alone. Here are some attempts, beginning with two attempts that use official serial behavior (and fail to achieve the correct meaning):
 
* {{T|jaq sao choıbeq}} This is the obvious attempt, but it fails because serials are right-grouping. The serial groups as {{T|jaq (sao choıbeq)}}, and so refers to something that is "very much a large thorn". This is different than the intended meaning in that we are not modifying the largeness but rather modifying the "large thorn"-ness. Something which is undeniably large but only sort of a thorn would be a very large thorn, but would not be "very much a large thorn". And arguably, a *gigantic* thorn would not be "very much a large thorn" either, since it's too large to be such a great match for the description "large thorn".
* {{T|choıbeq jaq sao}} This solution groups as {{T|choıbeq (jaq sao)}}, which has {{T|jaq sao}} grouped together as it should be. However, the modifier ends up on the right side of the serialization with {{T|choıbeq}}, where it cannot take on attributive adjective semantics as per official Toaq serial rules.
* {{T|choıbeq jäq sao pè baq choıbeq cy}} This attempt translates literally as "thorn that is very large for a thorn". This can't really fail to have the right meaning, but it's way too long.
* {{T|choıbeq jäq sao cy}} This attempt omits the explicit standard from the adjective "very large".
* {{T|jaq sao pı choıbeq}} This attempt uses an experimental particle {{T|pı}} whose behavior is to stratify the serialization mechanism into two layers: first the things between the {{T|pı}}'s are serialized, and then the results of those serializations are serialized. {{T|a b c pı d e f g pı h}} groups as {{T|(a b c) (d e f g) (h)}}. The correct semantics are achieved here (via experimental grammar).
 
==="I know the names of many things."===
 
This sentence is difficult to translate because the desired semantics are "There are many single X's for which I know the name of X", but "many single X's" is difficult to express in Toaq. There are basically two families of proposed solution: those which attempt to express these semantics precisely, and those which only approximate them and leave the details to context.
 
Precise translations:
 
* {{T|Sa puı tu shı mëa púı bı, dua jí chûa hı shí da.}} This is a direct translation of the above semantics into official Toaq. It is much longer than the English and requires a fair bit of forethought.
 
* {{T|Sa puı bı, dua jí mû chua tu shı mea púı hı da.}} This is basically the same as the previous sentence, except that it allows {{T|tu}} to scope over {{T|hı}} in order to move the phrase {{T|tu shı mea púı}} inside the subordinate clause.
 
* {{T|Puy shı bı, dua jí chûa hı shí da.}} This is also a direct translation of the above semantics, but it relies on an unofficial quantifier {{T|puy}} meaning "many X's" to make the sentence much more succinct.
 
* {{T|Dua mu chua jí puy shı sa da.}} This is basically the same as the previous sentence except it uses a slightly different expression of the indirect question part of the sentence to make things even simpler.
 
* {{T|Shıjeo sa puı dûa chua jí sa ja da.}} This example uses a utility predicate {{T|shıjeo}} meaning "every individual among X satisfies Y" to avoid having to depart too far from "natural" plural logic usage.
 
* {{T|Chuadua jí sa puı da.}} This example uses the predicate {{T|chuadua}} to route around the distributivity issue. It is not clear how this would generalize to other similar examples (such as "I know the ages of many people").
 
* {{T|Puı tuyq düa jí chûa hóa hı da.}} Uses an unofficial quantifier {{T|tuyq}} which takes the union of all things satisfying a certain property. (Note that the union itself may not satisfy the property that its constituents do.)
 
* {{T|Dua jí tôıtuq chua hı sa puı da.}} should work, but requires an unnaturalistic turn of phrase.
 
* {{T|Dua jí chûabo dujı sa puı hı da.}} uses {{t|dujı|focus marker: for every aggregate amongst X, unboxed, …|#cLo4JaQT5}}.
 
Imprecise translations:
 
* {{T|Hıdua chua jí sa puı da.}} This solution assumes a nondistributive {{T|chua}} whose meaning would be something like "___ are collectively the names of ___". This solutions does not directly claim that I know which name is the name of which thing; that detail would be left for the audience to infer from context.
 
==Solved sentences==
 
No open problems have been solved since the creation of this page.

Latest revision as of 19:26, 2 November 2022

This page gives a list of English sentences and phrases for which there is no widely accepted Toaq translation. The list is broken into three sections: sentences for which no accurate translation is known at all, sentences for which accurate translations are known but are impractically long or unacceptable for some other reason, and sentences which used to appear in one of these two categories but for which acceptable translations have since been found. Each unsolved sentence is accompanied by some discussion of the difficulty, possibly including some incomplete or not widely accepted solutions.

The existence or nonexistence of a solution is determined relative to official grammatical features of Toaq. Unofficial predicates are allowed, but sufficiently technical utility predicates should be considered grammatical features in spirit.

Sentences that have not been translated

"It's as though..."

"You keep insulting me, as though you think it will make me change my mind."

The hard part here is the "as though". What predicates are involved in an "as though X" construction?

"If I owned a unicorn, I would feed it"

Attempts:

  • She hâqdoa jí tu shıtuaq bö jí hó, keo bo jí sıa hao da. has to be phrased much differently than the original (enough so that it might not even mean the right thing (?)), and is long.
  • Baq guo sü kîaı jí sa shıtuaq bı, mu paq gúo hâqdoa jí shítuaq päqtao gúo da. demonstrates the seed of different kind of phrasing, but is very long and awkward as-is. The idea is to have a single mechanism that covers both "if a unicorn existed, it would have a horn" and "if an event of me having a unicorn existed, it would include me feeding it" with some mechanism for saying "X's have property Y" even when X's don't really exist, and then to make X "events of me having a unicorn" and Y the property "the haver of X gives food to the havee of X".

"Foxes are small to medium-sized creatures"

Attempts:

  • Nuı rVV saotuao nıaı baq hupı da. requires a new conjunction rVV which produces the appropriate meaning when used in serials. It's unclear (to me, Hoaqgıo, as I write this) how the conjunction would work, because it doesn't seem like there's a meaning for it to have when joining two noun phrases, unless they both happen to be kinds.
  • Rıe nuı roı saotuao nıaı baq hupı da. requires roı to be allowed in serials and for some mechanism to be put in place that gives it reasonable behavior, including correct behavior in this case.

Sentences that have not been translated well

"a very large thorn"

This translation exercise is a phrase rather than a sentence. The phrase is "very large thorn", and although it seems like a perfect use of Toaq's serial predicates, it is not possible to achieve the desired meaning through official serial behavior alone. Here are some attempts, beginning with two attempts that use official serial behavior (and fail to achieve the correct meaning):

  • jaq sao choıbeq This is the obvious attempt, but it fails because serials are right-grouping. The serial groups as jaq (sao choıbeq), and so refers to something that is "very much a large thorn". This is different than the intended meaning in that we are not modifying the largeness but rather modifying the "large thorn"-ness. Something which is undeniably large but only sort of a thorn would be a very large thorn, but would not be "very much a large thorn". And arguably, a *gigantic* thorn would not be "very much a large thorn" either, since it's too large to be such a great match for the description "large thorn".
  • choıbeq jaq sao This solution groups as choıbeq (jaq sao), which has jaq sao grouped together as it should be. However, the modifier ends up on the right side of the serialization with choıbeq, where it cannot take on attributive adjective semantics as per official Toaq serial rules.
  • choıbeq jäq sao pè baq choıbeq cy This attempt translates literally as "thorn that is very large for a thorn". This can't really fail to have the right meaning, but it's way too long.
  • choıbeq jäq sao cy This attempt omits the explicit standard from the adjective "very large".
  • jaq sao pı choıbeq This attempt uses an experimental particle whose behavior is to stratify the serialization mechanism into two layers: first the things between the 's are serialized, and then the results of those serializations are serialized. a b c pı d e f g pı h groups as (a b c) (d e f g) (h). The correct semantics are achieved here (via experimental grammar).

"I know the names of many things."

This sentence is difficult to translate because the desired semantics are "There are many single X's for which I know the name of X", but "many single X's" is difficult to express in Toaq. There are basically two families of proposed solution: those which attempt to express these semantics precisely, and those which only approximate them and leave the details to context.

Precise translations:

  • Sa puı tu shı mëa púı bı, dua jí chûa hı shí da. This is a direct translation of the above semantics into official Toaq. It is much longer than the English and requires a fair bit of forethought.
  • Sa puı bı, dua jí mû chua tu shı mea púı hı da. This is basically the same as the previous sentence, except that it allows tu to scope over in order to move the phrase tu shı mea púı inside the subordinate clause.
  • Puy shı bı, dua jí chûa hı shí da. This is also a direct translation of the above semantics, but it relies on an unofficial quantifier puy meaning "many X's" to make the sentence much more succinct.
  • Dua mu chua jí puy shı sa da. This is basically the same as the previous sentence except it uses a slightly different expression of the indirect question part of the sentence to make things even simpler.
  • Shıjeo sa puı dûa chua jí sa ja da. This example uses a utility predicate shıjeo meaning "every individual among X satisfies Y" to avoid having to depart too far from "natural" plural logic usage.
  • Chuadua jí sa puı da. This example uses the predicate chuadua to route around the distributivity issue. It is not clear how this would generalize to other similar examples (such as "I know the ages of many people").
  • Puı tuyq düa jí chûa hóa hı da. Uses an unofficial quantifier tuyq which takes the union of all things satisfying a certain property. (Note that the union itself may not satisfy the property that its constituents do.)
  • Dua jí tôıtuq chua hı sa puı da. should work, but requires an unnaturalistic turn of phrase.
  • Dua jí chûabo dujı sa puı hı da. uses dujı.

Imprecise translations:

  • Hıdua chua jí sa puı da. This solution assumes a nondistributive chua whose meaning would be something like "___ are collectively the names of ___". This solutions does not directly claim that I know which name is the name of which thing; that detail would be left for the audience to infer from context.

Solved sentences

No open problems have been solved since the creation of this page.