Prefix Reform: Difference between revisions

From The Toaq Wiki
(Category)
m (Fix typo: [q] → [ŋ])
 
(7 intermediate revisions by 2 users not shown)
Line 1: Line 1:
'''Prefix Reform''' is a proposal by [[uakci]] to change how [[prefix]]es interact with the stems of words.
'''Prefix Reform''' is a {{proposal}} by [[uakci]] to change how [[prefix]]es interact with the stems of words.


Note: the two parts of the proposal, [[#Spelling changes|the spelling changes]] and [[#Pronunciation changes|the pronunciation changes]], could be considered independently of each other.
Note: the two parts of the proposal, [[#Spelling changes|the spelling changes]] and [[#Pronunciation changes|the pronunciation changes]], could be considered independently of each other.


==Reasons==
==Reasons==
 
Prefixes as done officially are subtle and hard to teach, especially in terms of pronunciation:
#Prefixes currently act weird.
* They interact with vowel length in non-obvious ways: {{t|bộtao}} [bŏʔŏ] vs. {{t|bô'otao}} [boːʔo].
##They are unable to interact with {{Done|3}}. This proposal would allow prefixes to go on words with {{done|3}}.
* They’re not computer-friendly to spell. Some combinations, especially ones involving {{t|ı̣}} (that’s a dotless i with an underdot diacritic!), take a lot of effort to input (and Unicode normalizes that character to ị, which is not the same) and display poorly on many people’s systems.
##It is unclear how exactly they affect vocalic length. Plus, those interactions are often quite subtle and easily missed – consider {{t|bộtao}} vs. {{t|bô'otao}}. This makes prefixes hard to teach and possibly discouraging to a new learner trying to produce them in speech.
* There's a gotcha involving vowel-initial stems: {{t|e-}} + {{t|ano}} is not {{t|ẹano}} [ˈʔĕʔĕano] but {{t|ẹ'ano}} [ˈʔɛ̆ʔɛ̆ʔaːno]. The latter decomposes as {{t|ea-}} + {{t|no}}! Looking at the IPA transcriptions, we may conclude that the language is actually sensitive to three vowel lengths, [ɛ̆~ĕ ɛ eː].
#The current spelling of prefixes is not computer-friendly. They are difficult (if not impossible) to type for many and some characters, like {{T|ı̣}} often do not display correctly in many fonts. This proposal would allow for easier input and guarantee compatibility with most good fonts.
#*There's a gotcha where {{t|e-}} (any prefix) + {{t|ano}} (any vowel-initial stem) is not {{t|ẹano}} [ˈʔĕʔĕano] but {{t|ẹ'ano}} [ˈʔɛ̆ʔɛ̆ʔaːno]. The latter is actually {{t|ea-}} + {{t|no}}! – but this subtlety is not explained or taught to people, and it's likewise unintuitive. Worse yet, the glottal interposition (which turns /ɛ/ and /ɛː/ into [ɛ̆ʔɛ̆]) effectively splinters vowel length into three instead of two!


==Spelling changes==
==Spelling changes==
Line 15: Line 13:


{| class="wikitable"
{| class="wikitable"
!Official
!  !! Official !! Proposal
|{{t|mụfoaq}}
|{{t|lạ́maı}}
|{{t|tọ̈ꝡa}} <small>{{red|''(invalid word)''}}</small>
|{{t|bộtao}}
|-
|-
!Proposal
| {{Done|1}} || {{t|mụfoaq}} || {{t|mufòaq}}
|{{t|mufòaq}}
|-
|{{t|lamáı}}
| {{Done|2}} || {{t|lạ́maı}} || {{t|lamáı}}
|{{t|toꝡä}}
|-
|{{t|botâo}}
| {{Done|3}} || {{t|tọ̈ꝡa}} <small>{{red|''(invalid word)''}}</small> || {{t|toꝡä}}
|-
| {{Done|4}} || {{t|bộtao}} || {{t|botâo}}
|}
|}
:<small>The {{t|ẹano}} gotcha does not disappear per se. However, with the prefix reform, it is much easier to see that prefixes are not “special” raku that can be tacked on the front of a word willy-nilly: {{t|e<u>'àno</u>}} is spelled exactly the same as {{t|<u>e'ano</u>}}, just with the stress shifted away from the first syllable. Colloquially, the glottal stop could be omitted – {{t|e<u>àno</u>}} and {{t|buq<u>ùgı</u>}} can only be understood in exactly one way – however, watch out for codal {{t|m}}: {{t|ra<u>màno</u>}} ≠ {{t|ram<u>'àno</u>}}.</small>


==Pronunciation changes==
==Pronunciation changes==
:<b>TL;DR:</b> You always stress the first syllable of the stem. Prefixes should be quieter, shorter, and use mid tone.
[[File:PrefixReform.svg|300px|thumb|right|A chart showing the spelling and pronunciation changes. The pink segments are pronounced with unstressed, short, closed vowels (here [kʰʊ̌˧]).]]


*Stress the first syllable of the stem rather than the prefix. For instance, in {{t|puchumgòıchuq}} ‘was taking medicine’, stress the {{t|-goı-}}, no different than {{t|bugòıchuq}} ‘doesn’t/isn’t taking medicine’ or even just {{t|goıchuq}} ‘take medicine’.
* Stress the first syllable of the stem rather than the prefix. For instance, in {{t|puchumgòıchuq}} ‘was taking medicine’, stress the {{t|-goı-}}, no different than {{t|bugòıchuq}} ‘doesn’t/isn’t taking medicine’ or even just {{t|goıchuq}} ‘take medicine’.
*This stressed syllable should be louder and/or longer and/or more extreme in terms of the tone contour.
* This stressed syllable should be louder and/or longer and/or more extreme in terms of the tone contour.
*The unstressed prefix syllables should be shortened with regards to regular vowel length: {{t|bunúq}} ‘the non-snake’ could be [bŭnʊq]. (◌̆, the breve, is used to signal “extra-short” vowel length in the IPA.) At any rate, they should sound shorter than the stressed syllable of the stem (so [bunuː] for {{t|bunú}} is fine).
* The unstressed prefix syllables should use the weak forms of their core vowels: /u/ goes to [ʊ], /i/ goes to [ɪ], /o/ goes to [ɔ]. These contextual allophones are already used elsewhere in the language – namely, [ɔ] appears in {{t|oı}} /ɔj/, and the three are also triggered before {{t|q}}, e.g., {{t|bıq}} [bɪŋ] and not [biŋ]. Now imagine this is yet another context for them to be weakened. This leads to spicy phonoaesthetics not seen anywhere else in the language: {{t|jıa-}} [dʑɪa]!
*Optionally, the unstressed prefix syllables may use the weak forms of their core vowel: /u/ goes to [ʊ], /i/ goes to [ɪ], /o/ goes to [ɔ]. These contextual allophones are already used elsewhere in the language – namely, [ɔ] appears in {{t|oı}} /ɔj/, and the three are also triggered before {{t|q}}, e.g., {{t|bıq}} [bɪŋ] and not [biŋ]. Now imagine this is yet another context for them to be weakened. This leads to spicy phonoaesthetics not seen anywhere else in the language: {{t|jıa-}} [dʑɪa]!!!
* The unstressed prefix syllables should be shortened with regards to regular vowel length: {{t|bunúq}} ‘the non-snake’ could be [bʊ̌nʊŋ]. (◌̆, the breve, is used to signal “extra-short” vowel length in the IPA.) At any rate, they should sound shorter than the stressed syllable of the stem (so [bʊnuː] for {{t|bunú}} is fine).
*In terms of tone, there are two options, the hard option and the easy option:
* In terms of tone: always pronounce all prefixes with the mid level tone, [˧]. Extra care should be taken not to allow the tone to slide upwards or downwards as it’s pronounced – in other words, avoid *[˧˦] or *[˧˨].
*;The hard option: Make the prefix or prefixes’ contour “anticipate” the stem’s contour. For example, if the stem starts high, the prefixes should be high. If the stem starts lower, the prefixes should be there as well. This option has turned out hard in practice because it’s not as easy to control the tone contour of a weak unstressed syllable.
*;The easy option: Always pronounce all prefixes with the mid level tone, [˧]. Extra care should be taken not to allow the tone to slide upwards or downwards as it’s pronounced – in other words, avoid *[˧˦] or *[˧˨].


Illustrative examples (using the easy option):
Illustrative examples:


{| class="wikitable"
{| class="wikitable"
Line 57: Line 53:
The value of this proposal, apart from more flexibility and less ambiguity, is that stems no longer alternate between stressed and unstressed depending on whether they have prefixes attached to them (think {{t|<u>juı</u>taq}} vs. {{t|bụ<u>juı</u>}}). As another pleasant side effect, poly[[raku]]<nowiki/>ic words in {{done|3}} are now possible (like {{t|äımu}}, which before was ambiguous with the mid-falling allotonal forms of {{t|ạımu}} and {{t|ạ́ımu}}).
The value of this proposal, apart from more flexibility and less ambiguity, is that stems no longer alternate between stressed and unstressed depending on whether they have prefixes attached to them (think {{t|<u>juı</u>taq}} vs. {{t|bụ<u>juı</u>}}). As another pleasant side effect, poly[[raku]]<nowiki/>ic words in {{done|3}} are now possible (like {{t|äımu}}, which before was ambiguous with the mid-falling allotonal forms of {{t|ạımu}} and {{t|ạ́ımu}}).


==Thinking clitically – a fifth tone==
==Hoemaı's comments==
Let us think about clitics – little segments of words that attach to other words rather than forming full units. For instance, in English, the possessive marker <i>’s</i> can attach to words like <i>California’s reputation</i> or entire phrases like <i>my most beloved partner’s necklace</i> or <i>that one guy we saw in the street’s student card</i>. But then there’s nothing stopping us from spelling it as a separate word: <i>California ’s reputation</i>, for instance. More on clitics [[wikipedia:Clitic|over on Wikipedia]] if you’re interested.
[[Hoemaı]] commented on the proposal on [https://discord.com/channels/311223912044167168/311223912044167168/1206242007438073876 February 11, 2024 in #general]. {{Transcript|<blockquote>
First of all, I agree with the motivation behind the proposal; the morphology isn't as simple as it could be, but there are reasons for this.
 
The two downsides I see with the prefix proposal
 
1) In Toaq beta, there were constructions like the following
 
{{t|Gı mu hâo súq}}
 
{{t|Bủ dủa jí hı râi ní.}}
 
where the "prefixes" mu and hı would appear outside of the {{Tone|5}} clause that they belong to.


Similarly in Toaq, but let’s for a moment assume that the clitical “prefix toneme” is spelled like {{tone|1}} (as a macron). Its pronunciation is outlined above, but we could summarize it as a mid-flat tone that also nullifies vowel length: {{t|jıa}} /dʑiːa˥˩/ but {{t|jīa}} /dʑia˧/. Putting this together into a new table:
When I created the prefix morphology of Toaq Delta, this was something I wanted to avoid. The prefix proposal looks like it would lead to something similar, e.g. {{t|buháo}}. This is why the morphology is as complicated as it is; I wanted both the stress and the word-tone to be on the first syllable.  


{| class="wikitable"
2) I see a great danger in the unstressed mid tone turning into a low tone in casual speech, making it very difficult to distinguish {{t|haobu háo}} and {{t|hao buháo}}. This is something that universal initial stress avoids. (I see that the most recent version of the wiki page has a rule about a change in vowel quality, which would help with this a bit. I'd have to mull it over)
!Official
|{{t|puchụmtao}}
|{{t|jı̣achia}}
|{{t|lạ́maı bẹ́ıroı}}
| style="background:#ccc" |N/A
|{{t|jụ̂aqjuaı}}
|-
!“Halfway”<ref group=note>[[Lynn]]’s idea: https://discord.com/channels/311223912044167168/334810940392341514/1174698020579856434</ref>
|{{t|puchụmtao}}
|{{t|jı̣achıa}}
|{{t|lạmáı bẹıróı}}
|{{t|tọꝡä}}
|{{t|jụaqjûaı}}
|-
!Main proposal
|{{t|puchumtào}}
|{{t|jıachìa}}
|{{t|lamáı beıróı}}
|{{t|toꝡä}}
|{{t|juaqjûaı}}
|-
!This idea
!{{t|pū chūm tao}}
!{{t|jīa chıa}}
!{{t|lā máı bēı róı}}
!{{t|tō ꝡä}}
!{{t|jūaq jûaı}}
|-
!A respelling<ref group=note>As a ludicrous flight of fancy, imagining the prefix tone is spelled the “lightest”, without any tone marks, {{done|1}} taking a new obligate tone mark, ◌̀, in its place.</ref>
|{{t|pu chum tào}}
|{{t|jıa chìa}}
|{{t|la máı beı róı}}
|{{t|to ꝡä}}
|{{t|juaq jûaı}}
|}
<small><references group=note /></small>


With this in mind, we may look at the neat symmetry in which [[focus marker]]s behave when alternated between their inherent {{done|2}} and this new {{tone|1}}:
Bottom line:


{|
I wish I could just be 100% in favor of the proposal, because it's a good one. However, I want to be transparent about why I'm not and why the morphology is the way it is: it's not that I simply missed an easier alternative, but that I started with certain goals in mind (see above) and had to make the rules fit the goals. Are these goals worth the complexity? Depends on one's priorities.
| {{t|luı}} || {{t|do}} || {{t|jí}} || {{t|súq}} || {{green|{{t|kú}}}} || {{green|{{t|tú}}}} || {{green|{{t|maoja}}}}
</blockquote>}}
|-
| colspan=7 | ‘I have given you {{green|''all the apples''}}.’
|-
| {{t|luı}} || {{t|do}} || {{t|jí}} || {{t|súq}} || {{orange|{{t|kū}}}} || {{orange|{{t|tú}}}} || {{t|maoja}}
|-
| colspan=7 | ‘I have given you {{orange|''all''}} the apples.’
|}


==See also==
==See also==


* [https://discord.com/channels/311223912044167168/889589074011230230/1113157808934883469 uakci’s original message]
* [https://discord.com/channels/311223912044167168/889589074011230230/1113157808934883469 uakci’s original message]
[[Category:Proposals]]
* A “logical conclusion” of the proposal posits that prefixes could be treated as separate words carrying a special [[prefix toneme]].

Latest revision as of 02:31, 10 September 2024

Prefix Reform is a proposal by uakci to change how prefixes interact with the stems of words.

Note: the two parts of the proposal, the spelling changes and the pronunciation changes, could be considered independently of each other.

Reasons

Prefixes as done officially are subtle and hard to teach, especially in terms of pronunciation:

  • They interact with vowel length in non-obvious ways: bộtao [bŏʔŏ] vs. bô'otao [boːʔo].
  • They’re not computer-friendly to spell. Some combinations, especially ones involving ı̣ (that’s a dotless i with an underdot diacritic!), take a lot of effort to input (and Unicode normalizes that character to ị, which is not the same) and display poorly on many people’s systems.
  • There's a gotcha involving vowel-initial stems: e- + ano is not ẹano [ˈʔĕʔĕano] but ẹ'ano [ˈʔɛ̆ʔɛ̆ʔaːno]. The latter decomposes as ea- + no! Looking at the IPA transcriptions, we may conclude that the language is actually sensitive to three vowel lengths, [ɛ̆~ĕ ɛ eː].

Spelling changes

Instead of putting the tone mark on the first syllable, put it on the first syllable of the stem. In the case of a word with falling tone, use mid-falling tone as the diacritic. Examples:

Official Proposal
falling tone mụfoaq mufòaq
rising tone lạ́maı lamáı
glottal tone tọ̈ꝡa (invalid word) toꝡä
hiatus tone bộtao botâo
The ẹano gotcha does not disappear per se. However, with the prefix reform, it is much easier to see that prefixes are not “special” raku that can be tacked on the front of a word willy-nilly: e'àno is spelled exactly the same as e'ano, just with the stress shifted away from the first syllable. Colloquially, the glottal stop could be omitted – eàno and buqùgı can only be understood in exactly one way – however, watch out for codal m: ramàno ≠ ram'àno.

Pronunciation changes

A chart showing the spelling and pronunciation changes. The pink segments are pronounced with unstressed, short, closed vowels (here [kʰʊ̌˧]).
  • Stress the first syllable of the stem rather than the prefix. For instance, in puchumgòıchuq ‘was taking medicine’, stress the -goı-, no different than bugòıchuq ‘doesn’t/isn’t taking medicine’ or even just goıchuq ‘take medicine’.
  • This stressed syllable should be louder and/or longer and/or more extreme in terms of the tone contour.
  • The unstressed prefix syllables should use the weak forms of their core vowels: /u/ goes to [ʊ], /i/ goes to [ɪ], /o/ goes to [ɔ]. These contextual allophones are already used elsewhere in the language – namely, [ɔ] appears in oı /ɔj/, and the three are also triggered before q, e.g., bıq [bɪŋ] and not [biŋ]. Now imagine this is yet another context for them to be weakened. This leads to spicy phonoaesthetics not seen anywhere else in the language: jıa- [dʑɪa]!
  • The unstressed prefix syllables should be shortened with regards to regular vowel length: bunúq ‘the non-snake’ could be [bʊ̌nʊŋ]. (◌̆, the breve, is used to signal “extra-short” vowel length in the IPA.) At any rate, they should sound shorter than the stressed syllable of the stem (so [bʊnuː] for bunú is fine).
  • In terms of tone: always pronounce all prefixes with the mid level tone, [˧]. Extra care should be taken not to allow the tone to slide upwards or downwards as it’s pronounced – in other words, avoid *[˧˦] or *[˧˨].

Illustrative examples:

Official Proposal
puchụmtao [ˈpuː˥˨t͡ɕŭʔŭm˨˩tʰaw˩] puchumtào [pʊ˧t͡ɕʊm˧ˈtaːw˥˩]
jı̣achıa [ˈd͡ʑĭʔĭa˥˨t͡ɕia˨˩] jıachìa [d͡ʑɪa˧ˈt͡ɕiːa˥˩]
lạ́maı bẹ́ıroı [ˈlăʔă˨˦maj˦˥ ˈbɛ̆ʔɛ̆j˨˩rɔj˩] lamáı beıróı [la˧ˈmaj˧˥ bɛj˧ˈrɔj˧˩]
*tọ̈ꝡa (invalid word) toꝡä [tʰɔ˧ˈja˧˨ʔa˨˩]
jụ̂aqjuaı [ˈd͡ʑŭʔŭaŋ˧˥˨d͡ʑu˨˩aj˩] juaqjûaı [d͡ʑʊaŋ˧ˈd͡ʑuː˧˥˨aj˨˩]

The value of this proposal, apart from more flexibility and less ambiguity, is that stems no longer alternate between stressed and unstressed depending on whether they have prefixes attached to them (think juıtaq vs. bụjuı). As another pleasant side effect, polyrakuic words in glottal tone are now possible (like äımu, which before was ambiguous with the mid-falling allotonal forms of ạımu and ạ́ımu).

Hoemaı's comments

Hoemaı commented on the proposal on February 11, 2024 in #general.

First of all, I agree with the motivation behind the proposal; the morphology isn't as simple as it could be, but there are reasons for this.

The two downsides I see with the prefix proposal

1) In Toaq beta, there were constructions like the following

Gı mu hâo súq

Bủ dủa jí hı râı ní.

where the "prefixes" mu and hı would appear outside of the rising-falling tone clause that they belong to.

When I created the prefix morphology of Toaq Delta, this was something I wanted to avoid. The prefix proposal looks like it would lead to something similar, e.g. buháo. This is why the morphology is as complicated as it is; I wanted both the stress and the word-tone to be on the first syllable.

2) I see a great danger in the unstressed mid tone turning into a low tone in casual speech, making it very difficult to distinguish haobu háo and hao buháo. This is something that universal initial stress avoids. (I see that the most recent version of the wiki page has a rule about a change in vowel quality, which would help with this a bit. I'd have to mull it over)

Bottom line:

I wish I could just be 100% in favor of the proposal, because it's a good one. However, I want to be transparent about why I'm not and why the morphology is the way it is: it's not that I simply missed an easier alternative, but that I started with certain goals in mind (see above) and had to make the rules fit the goals. Are these goals worth the complexity? Depends on one's priorities.

See also

Retrieved from "https://toaq.me/index.php?title=Prefix_Reform&oldid=2109"