null

If you're new to these arts, this is the place to ask "stupid" questions and get directions!
Post Reply
wyl118
rupestrian
rupestrian
Posts: 6
Joined: 12 Jan 2020 14:15

null

Post by wyl118 »

As a novice in the field of phonology, I've been confused about the analysis by OT theory for a long time. Could you guys help me with these two questions?





Actually, the main part I cannot deal with is the markedness contraints and faithfulness constraints. And do you guys have any tips on the steps of working out this kind of questions? Thank you in advance!
Last edited by wyl118 on 19 Jun 2020 07:57, edited 4 times in total.
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: How to analyze the data with Optimality Theory?

Post by sangi39 »

An assumption, but might this be for an assignment? Or is it just for something to do?


So, as I understand it, which, to be fair, is not a lot, Optimality Theory basically states "here is some input A, here are some constraints B and their rankings C in language D, thus these are the mostly likely outcomes E given B and C".

So, for example, let's say you have an "input" /plank/. You can throw that at any language you want, and, according to OT, there'll be some likely outcome based on whatever constraints the language has in its phonology, and how it "ranks" those constraints.

The constraints, as they're described, are universal, i.e. they apply to all languages, but they are ranked differently. So, for example, you could have the following constraints:

1) "coda nasals are not deleted" (which I think is a faithfulness constraint since it's retaining something of the original input)
2) "clusters of nasals and plosives reduce to the plosive" (which is a markedness constraint since it actually makes a change to the input)
3) "clusters not between two vowels are broken up by vowels" (another markedness constraint, since it's now adding to the word).
4) "clusters after vowels are retained" (a faithfulness constraint)

These may be ranked in 3 languages as follows, for example:

Language A: 1 > 4 > 2 > 3
Language B: 2 > 1 > 3 > 4
Language C: 1 > 3 > 4 > 2

Which would yield the following outputs:

Language A: [palank] (the coda nasal is retained, constraint 2 doesn't apply because it violates 1, clusters after the vowel are retained, so 3 only applies to the initial cluster)
Language B: [palak] (the coda nasal is dropped, so 1 doesn't apply, the initial cluster is broken up (3 would apply to the final cluster, but this was reduced according to 2), and rule 4 can't apply because it violates 3)
Language C: [palanak] (the coda nasal is retained, but then both clusters are broken up according to 4, which 3 violates to it doesn't apply, and since the cluster is broken up anyway, 2 no long applies).

The output, given any input, is the one that then conforms most closely to the constraints the language places on itself in the order that they are ranked. So, for example, you couldn't have [palak] in Language B, because there's a constraint against nasal+plosive clusters, and while you could have [palak] in Language A, it violates its own constraints on deleting certain parts of the input.



Someone else will likely have a much better understanding of this than I do, and honestly that person is likely to be your teacher/lecturer/professor/tutor/whatever, so I'd suspect they're actually the best person to ask. There's no harm in asking for help, especially from the person whose career is built around helping people to learn [:)] It shows that you understand the gaps in your knowledge, and that you're willing to actively fill those gaps, and getting some more time with an educator isn't always a bad thing. It might help build a stronger relationship, which could aid you in the long run.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Khemehekis
mongolian
mongolian
Posts: 3884
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How to analyze the data with Optimality Theory?

Post by Khemehekis »

wyl118 wrote: 09 Mar 2020 10:11 a. Syncope permitted
sep[a]rate choc[o]late pers[o]nal

b. Syncope blocked
monog[a]my quarr[e]ling ult[i]mate
Syncope blocked in "quarreling", really? For me, "quarrel" is one syllable. Rhymes with "whorl". Just as "squirrel" in my signature rhymes with "girl".
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
wyl118
rupestrian
rupestrian
Posts: 6
Joined: 12 Jan 2020 14:15

Re: How to analyze the data with Optimality Theory?

Post by wyl118 »

Actually it's the basic questions for the final examination of this semester and I was expecting to prepare for it in advance. I've been study by myself but OT is the biggest problem. [:'(] Anyway, thank you so much for your help! [:)]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: How to analyze the data with Optimality Theory?

Post by sangi39 »

wyl118 wrote: 10 Mar 2020 14:16 Actually it's the basic questions for the final examination of this semester and I was expecting to prepare for it in advance. I've been study by myself but OT is the biggest problem. [:'(] Anyway, thank you so much for your help! [:)]
Was OT discussed much in your classes, or was it more "oh, and here's this other theoretical way of doing this", and you're being tested to see if you've grasped the basics?

I remember we had that in our lectures (archaeology), where the lecturer would predominantly discuss the side of things they agreed with (or directly worked on, which was normally the case, e.g. Kevin Kuykendall and his work on early human evolution, Mike Parker-Pearson and his work on Stonehenge, Dawn Hadley and Iron Age Scandinavia, John Bennet and his work on the Bronze Age Aegean and writing, and so on), but then they'd go on to briefly discuss different or opposing views (Marek Zvelebil, for example, would reel off several different hypotheses regarding the spread of early Indo-European languages in the space of half and hour if you gave him the chance). We were still expected to look into them as best as available resources would allow us to, of course, and you always needed to defend your work, but the basic understanding of a given topic usual came with a hint of bias in the lectures.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: How to analyze the data with Optimality Theory?

Post by Salmoneus »

Khemehekis wrote: 10 Mar 2020 05:26
Syncope blocked in "quarreling", really? For me, "quarrel" is one syllable. Rhymes with "whorl". Just as "squirrel" in my signature rhymes with "girl".
Yeah, but that's a weird American thing - deletion of schwa after /r/, followed by (in the case of 'squirrel') some weird vowel shift.

Giving the famous Bushisms "merkin" (for 'American') and "tourism" (for "terrorism").
Khemehekis
mongolian
mongolian
Posts: 3884
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How to analyze the data with Optimality Theory?

Post by Khemehekis »

Salmoneus wrote: 12 Mar 2020 03:24 Yeah, but that's a weird American thing - deletion of schwa after /r/, followed by (in the case of 'squirrel') some weird vowel shift.
Huh.

https://en.wiktionary.org/wiki/squirrel

I just checked the Wiktionary entry for "squirrel", and apparently outside of North America people pronounce it to rhyme with the way we Americans say "virile"! Never knew that.
Giving the famous Bushisms "merkin" (for 'American') and "tourism" (for "terrorism").
I've seen people write "terrism" when mocking Bush. I don't think I've ever heard the "tourism" pronunciation.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: How to analyze the data with Optimality Theory?

Post by sangi39 »

Wait, what is the reason why there's no syncope in 2b? I had thought maybe it was down to "not a valid cluster elsewhere", but then there's "stigma", which counters that idea for "monogamy". That switched my thinking to "morphemes", since "monogamy" contains a still valid independent-ish morpheme, but then "personal", I suppose, is still seen as two morphemes. Am I missing something?
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: How to analyze the data with Optimality Theory?

Post by Salmoneus »

sangi39 wrote: 15 Mar 2020 01:54 Wait, what is the reason why there's no syncope in 2b? I had thought maybe it was down to "not a valid cluster elsewhere", but then there's "stigma", which counters that idea for "monogamy". That switched my thinking to "morphemes", since "monogamy" contains a still valid independent-ish morpheme, but then "personal", I suppose, is still seen as two morphemes. Am I missing something?
I think your first answer is right: /gm/, /rrl/ and /tm/ aren't standard valid clusters.

It's true that there are words with /gm/. But they're very few, and they're learned loanwords, and they're limited (there's no words with /gmi/, are there? It's one thing to have a cluster before schwa, another when the following consonant is actually partially stressed...).

The same is also true, of course of /tm/ - there are words like 'tmesis', after all. And of course 'atmosphere'. And 'mahatma'. And 'hetman'. But they're very rare.

After all, /fT/ is technically a cluster in English (phthisis, naphtha, etc), but it doesn't really count...
Khemehekis
mongolian
mongolian
Posts: 3884
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How to analyze the data with Optimality Theory?

Post by Khemehekis »

Salmoneus wrote: 15 Mar 2020 02:38=
It's true that there are words with /gm/. But they're very few, and they're learned loanwords, and they're limited (there's no words with /gmi/, are there? It's one thing to have a cluster before schwa, another when the following consonant is actually partially stressed...).
Most of them come from the Greek ending -gma: stigma, enigma, sigma, magma, zeugma, dogma, smegma, just off the top of my head. There are also pigment, figment, fragment, segment, but that's over the course of two morphemes, so they don't really count.

I can think of only one /gmi/ word: pygmy.
After all, /fT/ is technically a cluster in English (phthisis, naphtha, etc), but it doesn't really count...
Huh. My old family doctor pronounced "phthisiology" with a /t/, as if it were "tisiology".

EDIT: It just occurred to me we have it in "fifth" and "twelfth". Of course, you might argue that those don't count, since the f and th come from two different morphemes.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
Post Reply