"Make these work together" game.

A forum for all topics related to constructed languages
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Fri 14 Jun 2013, 18:42

I wrote:OK, if no-one proposes a solution (or, better, an overview of possible solutions) in, like, a couple days, I'll post some hints.
So, perhaps I should post something indeed.

To remind what the challenge was about:
I wrote:I am thinking of a natural-looking language with the following features:

1. It's monosyllabic, has at least three tones which are never subject to positional neutralizations, and has very few potential clitics (or, better, none); that is, even if there are "grammatical" words, they are also used independently or at least can carry full phrasal stress.
2. It is not isolating. Its nouns have at least three cases which, among other things, disambiguate between agent-like and patient-like arguments of transitive verbs, as well as at least two numbers; it may also have some morphology in verbs; bonus points for having three or more genders, preferably with at least partly non-semantic-based gender assignment.
3. There are no restrictions on the phonetic shape of the least marked form of noun, i. e. any phonotactically permitted syllable can be in principle the dictionary form of some noun.

(Sure, I am asking not for a detailed description, only for an outline of main traits, morphological techniques in the first place.)
From previous discussion it appears that a clarification is needed on "3": it does not mean that all possible syllables have meanings assigned to them; it does mean that every well-formed syllable can be *in principle* treated as a regular declinable noun (e. g. if it's a loan). In fact, it would be cool to apply the same principle to open classes of words other than nouns.

Now, hints.

A. None of my conditions says or implies that there are no syllable-internal alternations.

B. Reduplication, including such that alters the shape of one or both syllables in some ways, can be viewed as not violating "1" if treated properly (e. g. if the phonetic features of both syllables are needed to recover the full information of the original monosyllabic word and its grammatical form, i. e. there is a functional factor disfavoring any phonetic erosion that might lead to coalescence into one phonetic word.)

(This is, BTW, the main idea of my condition "1": the language must be as close as possible to your ideal prototypical monosyllabic language, with its structures offering little incentive to reinterpret any sequences of syllables as polysyllabic words.)

Also, a couple meta considerations about the game.

It seems to me that the idea of the game provokes touching upon exotic and/or theoretically difficult features. Therefore, it is very easy to misinterpret the challenge's author's vision of the challenge and possible solutions.

On the other hand, due to the same characteristic of the game, it may be very interesting to see solutions to related problems, slightly differing from the author's original vision.

So I propose that two things to be encouraged:

(1) Conscious modifications to challenges posted by others.

(2) Restating the challenge one is responding to, to make it clear how it was understood in designing a solution. (As an example: I did that instinctively in replying to Keenir's challenge above.)

Also, it would be cool to see multiple responses to each challenge. So, one more thing:

(3) Using more spoiler tags.
Keenir
runic
runic
Posts: 2457
Joined: Tue 22 May 2012, 02:05

Re: "Make these work together" game.

Post by Keenir » Fri 14 Jun 2013, 22:16

basilius wrote:So I propose that two things to be encouraged:

(1) Conscious modifications to challenges posted by others.
this may make people less inclined to take on the challenges. (and it also appears to run afoul of the next proposal)
(2) Restating the challenge one is responding to, to make it clear how it was understood in designing a solution.
admittedly I did that more in my first challenge reply, than in my second.
Also, it would be cool to see multiple responses to each challenge.
ditto.
So, one more thing:
have fun.
At work on Apaan: viewtopic.php?f=6&t=4799
User avatar
eldin raigmore
fire
fire
Posts: 5684
Joined: Sat 14 Aug 2010, 18:38
Location: SouthEast Michigan

Re: "Make these work together" game.

Post by eldin raigmore » Fri 14 Jun 2013, 22:31

basilius wrote:It seems to me that the idea of the game provokes touching upon exotic and/or theoretically difficult features. Therefore, it is very easy to misinterpret the challenge's author's vision of the challenge and possible solutions.
IMO one of the most or first interesting things about receiving responses to challenges I post, is seeing what the responder thought I meant, especially if it's a bit different than what I actually meant.
Certainly, seeing a solution along lines I never thought of, is entertaining and inspiring.
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Mon 17 Jun 2013, 16:33

So, no-one is going to respond to the clarified version of my challenge?

No-one at all?
User avatar
Click
darkness
darkness
Posts: 3328
Joined: Sat 21 Jan 2012, 12:17

Re: "Make these work together" game.

Post by Click » Mon 17 Jun 2013, 17:00

I'd like to, but I can't find it even though it's probably in a super-obvious place.
User avatar
DesEsseintes
cleardarkness
cleardarkness
Posts: 4475
Joined: Sun 31 Mar 2013, 12:16

Re: "Make these work together" game.

Post by DesEsseintes » Mon 17 Jun 2013, 17:19

I'd like to, too, but I haven't figured out a convincing solution. [:S]
Keenir
runic
runic
Posts: 2457
Joined: Tue 22 May 2012, 02:05

Re: "Make these work together" game.

Post by Keenir » Mon 17 Jun 2013, 22:32

basilius wrote:So, no-one is going to respond to the clarified version of my challenge?
can you re-post it, please?
At work on Apaan: viewtopic.php?f=6&t=4799
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Mon 17 Jun 2013, 22:39

Sure.
basilius wrote:To remind what the challenge was about:
I wrote:I am thinking of a natural-looking language with the following features:

1. It's monosyllabic, has at least three tones which are never subject to positional neutralizations, and has very few potential clitics (or, better, none); that is, even if there are "grammatical" words, they are also used independently or at least can carry full phrasal stress.
2. It is not isolating. Its nouns have at least three cases which, among other things, disambiguate between agent-like and patient-like arguments of transitive verbs, as well as at least two numbers; it may also have some morphology in verbs; bonus points for having three or more genders, preferably with at least partly non-semantic-based gender assignment.
3. There are no restrictions on the phonetic shape of the least marked form of noun, i. e. any phonotactically permitted syllable can be in principle the dictionary form of some noun.

(Sure, I am asking not for a detailed description, only for an outline of main traits, morphological techniques in the first place.)
From previous discussion it appears that a clarification is needed on "3": it does not mean that all possible syllables have meanings assigned to them; it does mean that every well-formed syllable can be *in principle* treated as a regular declinable noun (e. g. if it's a loan). In fact, it would be cool to apply the same principle to open classes of words other than nouns.

Now, hints.

A. None of my conditions says or implies that there are no syllable-internal alternations.

B. Reduplication, including such that alters the shape of one or both syllables in some ways, can be viewed as not violating "1" if treated properly (e. g. if the phonetic features of both syllables are needed to recover the full information of the original monosyllabic word and its grammatical form, i. e. there is a functional factor disfavoring any phonetic erosion that might lead to coalescence into one phonetic word.)

(This is, BTW, the main idea of my condition "1": the language must be as close as possible to your ideal prototypical monosyllabic language, with its structures offering little incentive to reinterpret any sequences of syllables as polysyllabic words.)
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Tue 18 Jun 2013, 00:32

OK, as I said above, there are some solutions that seemed more-less obvious but engelangish to myself. For example:
Spoiler:
There are three cases (A, P, D), three declensions (1st, 2nd, 3rd), and three tones (à, a, á).

In singular, cases are marked by tone:

Code: Select all

       Declensions →    1st  2nd   3rd

Cases ↓

P                       à     a     á

A                       a     á     à                       

D                       á     à     a
Plural can be formed by e. g. reduplication.

Similar patterns of circular alternations can be applied to vowels, onsets and/or codas.
Also, like I said, if you see a nice solution which would work had my challenge been slightly modified, it would be very interesting to see the challenge restated and solved this way, too.
Keenir
runic
runic
Posts: 2457
Joined: Tue 22 May 2012, 02:05

Re: "Make these work together" game.

Post by Keenir » Tue 18 Jun 2013, 01:31

Keenir wrote:
basilius wrote:So, no-one is going to respond to the clarified version of my challenge?
can you re-post it, please?
basilius wrote:OK, as I said above, there are some solutions that seemed more-less obvious but engelangish to myself.
Also, like I said, if you see a nice solution which would work had my challenge been slightly modified, it would be very interesting to see the challenge restated and solved this way, too.
oh. not sure about the others, but I thought you had posted a new challenge after I had posted (then explained) my answer to your prior challenge, and it was the new challenge which you were asking for respondings to.
At work on Apaan: viewtopic.php?f=6&t=4799
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Wed 19 Jun 2013, 00:50

Well, no, I didn't post a new challenge... although it seems that I've already got something new to post.

But I am kinda disappointed about no-one wanting to solve the old one.

Also, it would be cool to see a challenge from someone. Seriously. Other people's challenges are always more interesting :)
Keenir
runic
runic
Posts: 2457
Joined: Tue 22 May 2012, 02:05

Re: "Make these work together" game.

Post by Keenir » Wed 19 Jun 2013, 02:42

basilius wrote:Well, no, I didn't post a new challenge... although it seems that I've already got something new to post.
please do.
But I am kinda disappointed about no-one wanting to solve the old one.
well, if the other games are any indication, while there is no problem with two or more people tackling the same prompt and posting them within a minute or few - nobody tackles the prompt after a response has already been posted.
Also, it would be cool to see a challenge from someone. Seriously. Other people's challenges are always more interesting :)
perhaps a new thread should be started...reserving this one for discussion of aspects of the creations of the two threads.
At work on Apaan: viewtopic.php?f=6&t=4799
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Wed 19 Jun 2013, 17:13

Keenir wrote:<...> nobody tackles the prompt after a response has already been posted.
I see. Obviously, I was born into a wrong world :(

For it's much more interesting to see a variety of solutions to a good problem, and some discussion (like, which of its interpretations make sense and which don't, and why certain classes of solutions will or won't work, and the like.)

OK, I don't consider my previous challenge solved (the solution in my spoiler isn't naturalistic, and yours discards an important condition). Perhaps, later I'll post an example of what would look like a solution for myself.
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Wed 19 Jun 2013, 17:20

Keenir wrote:
basilius wrote:Well, no, I didn't post a new challenge... although it seems that I've already got something new to post.
please do.
OK, sort of the classic one. Unlike my previous challenge, this one seems to be more directly connected to some natlang precedents.

(1) Word order in a transitive clause like The cat saw the mouse should be free.

This means that ideally, 'cat', 'mouse' and 'see' must allow for all the six combinatorically available orderings. However, simple and elegant solutions partly restricting the placement of the verb are also welcome.

(2) In the abovesaid type of clauses, word order alone must make it clear who saw whom.

This means that cases, adpositions, voices, all types of agreement, lexical animacy hierarchies, syntactically conditioned sandhi and the like should be irrelevant. Also, the dependency tree should remain the same at least for 'cat' and 'mouse' (not to mention that the clause must remain one clause; resorting to clefts is considered cheating).

With this challenge it would be especially cool to see multiple and diverse solutions.
User avatar
Click
darkness
darkness
Posts: 3328
Joined: Sat 21 Jan 2012, 12:17

Re: "Make these work together" game.

Post by Click » Mon 24 Jun 2013, 21:25

basilius wrote:This means that cases, adpositions, voices, all types of agreement, lexical animacy hierarchies, syntactically conditioned sandhi and the like should be irrelevant.
Is prosody relevant?
User avatar
Shemtov
mayan
mayan
Posts: 2160
Joined: Mon 29 Apr 2013, 03:06

Re: "Make these work together" game.

Post by Shemtov » Tue 25 Jun 2013, 07:09

basilius wrote:
Keenir wrote:
basilius wrote:Well, no, I didn't post a new challenge... although it seems that I've already got something new to post.
please do.
OK, sort of the classic one. Unlike my previous challenge, this one seems to be more directly connected to some natlang precedents.

(1) Word order in a transitive clause like The cat saw the mouse should be free.

This means that ideally, 'cat', 'mouse' and 'see' must allow for all the six combinatorically available orderings. However, simple and elegant solutions partly restricting the placement of the verb are also welcome.

(2) In the abovesaid type of clauses, word order alone must make it clear who saw whom.

This means that cases, adpositions, voices, all types of agreement, lexical animacy hierarchies, syntactically conditioned sandhi and the like should be irrelevant. Also, the dependency tree should remain the same at least for 'cat' and 'mouse' (not to mention that the clause must remain one clause; resorting to clefts is considered cheating).

With this challenge it would be especially cool to see multiple and diverse solutions.
Marking on the verb determines word-order:
SVO:
Zazhe oman-spo ll'atein
Cat saw-SVO mouse

VSO:
Oman-pos zazhe ll'atein
Saw-VSO Cat Mouse

OSV
Ll'atein zazhe oman-osp
Mouse Cat saw-OSV
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien
basilius
MVP
MVP
Posts: 244
Joined: Thu 27 Jan 2011, 18:22

Re: "Make these work together" game.

Post by basilius » Tue 25 Jun 2013, 14:41

Click wrote:Is prosody relevant?
Yeah, this may be an interesting type of solutions. Please share what possibilities you see!

Although I suspect technically they will be about "syntactically conditioned sandhi" ;)

But anyway...
Shemtov wrote:Marking on the verb determines word-order:
SVO:
Zazhe oman-spo ll'atein
Cat saw-SVO mouse

VSO:
Oman-pos zazhe ll'atein
Saw-VSO Cat Mouse

OSV
Ll'atein zazhe oman-osp
Mouse Cat saw-OSV
An interesting option, too.

But it is hardly about "word order alone" ;)

Also, I don't know enough about natlang precedents for morphology which is sensitive to the ordering of dependents but cannot be claimed to represent essentially a type of voice marking or somesuch. They must exist, though.

Anyway, I insist that there are at least two types of solutions that take the "word order alone" condition quite literally.

And some others, which may be at risk with some forcibly broadened interpretations of some of my conditions ;)
Keenir
runic
runic
Posts: 2457
Joined: Tue 22 May 2012, 02:05

Re: "Make these work together" game.

Post by Keenir » Tue 25 Jun 2013, 14:47

basilius wrote:Anyway, I insist that there are at least two types of solutions that take the "word order alone" condition quite literally.
then why don't you just show us, and stop insisting we read your mind? the point of a game is fun, is to enjoy ourselves and enjoy stretching our minds.
At work on Apaan: viewtopic.php?f=6&t=4799
User avatar
Click
darkness
darkness
Posts: 3328
Joined: Sat 21 Jan 2012, 12:17

Re: "Make these work together" game.

Post by Click » Tue 25 Jun 2013, 15:05

basilius wrote:
Click wrote:Is prosody relevant?
Yeah, this may be an interesting type of solutions. Please share what possibilities you see!

Although I suspect technically they will be about "syntactically conditioned sandhi" ;)
That's precisely why I asked. [;)]
User avatar
DesEsseintes
cleardarkness
cleardarkness
Posts: 4475
Joined: Sun 31 Mar 2013, 12:16

Re: "Make these work together" game.

Post by DesEsseintes » Tue 25 Jun 2013, 16:50

basilius wrote:
(1) Word order in a transitive clause like The cat saw the mouse should be free.

This means that ideally, 'cat', 'mouse' and 'see' must allow for all the six combinatorically available orderings. However, simple and elegant solutions partly restricting the placement of the verb are also welcome.

(2) In the abovesaid type of clauses, word order alone must make it clear who saw whom.

This means that cases, adpositions, voices, all types of agreement, lexical animacy hierarchies, syntactically conditioned sandhi and the like should be irrelevant. Also, the dependency tree should remain the same at least for 'cat' and 'mouse' (not to mention that the clause must remain one clause; resorting to clefts is considered cheating).

With this challenge it would be especially cool to see multiple and diverse solutions.
This is a brief outline for a solution to Basilius's latest challenge. Unfortunately, I don't have time to present it nicely, so you must excuse my hastily-scribbled ramblings. Also, I make no claim of this being a fully worked-out solution, but I would like to see whether people think this is going in the right direction.

I have just used English for my examples.
Spoiler:
Let's posit a language which has an underlying topic-comment structure.

Definiteness is only indirectly indicated by word order. Components towards the front of the clause are more definite (and therefore "topicalised") than arguments later on in the clause.

In intransitive clauses, generally, new subjects come after the verb, while known subjects come before the verb.

Came-in mouse.
= A mouse came in.

Mouse came-in.
= The mouse came in.

If I'm talking about a cat, the natural word order would be:

1. (Cat woke up.) Cat saw mouse.

The mouse could be definite or indefinite, according to context.

2. (Came in mouse.) Cat mouse saw. (But cat mouse didn't eat.)

This word order would be preferred if one were stressing the fact that the cat saw the mouse and didn't, say, eat it.

The object would be shifted to the front of the clause to mark it as topic:

3. (Mouse and rat left hole.) Mouse cat saw. Rat dog saw.

Meaning something like "as for the mouse, it was (the/a) cat that saw it."

4. (Mouse was walking along, when) mouse saw cat.

Would instead introduce a cat as the "unexpected" argument. It is probably inevitable that this word order would convey a somewhat passive "colouring" despite not being overtly marked on the verb.
Obviously, this word order would be the most problematic, so in this scenario, the speakers of the language would probably tend to use a qualifier for "cat" here, such as "a passing cat", "a cat that happened to be there". They might also stress the word "cat". Such qualifiers would however not be necessary if context were clear, so as not to violate Basilius's condition.

5. Saw cat mouse(, and so our story begins.)

Would be used when the arguments cat and mouse were both being introduced into discourse for the first time.

6. (Was not dog.) Saw mouse cat.

It was (the/a) cat that saw mouse.

Basically, all analyses of who sees whom (or what sees what) are based entirely on context.

Since Click asked about prosody, I concede that for the above to work, some variation in stress patterns would almost certainly arise, but this kind of topic-comment structure is the only solution I can see at present.

I actually don't think such a language would be out of the question.
Post Reply