syntax-prosody-ot / main

A webapp for the syntax-prosody analyst working in Optimality Theory, with automated Gen, Con and Eval. Download build files from syntax-prosody-ot/build
https://spot.sites.ucsc.edu/
GNU General Public License v2.0
12 stars 2 forks source link

Swedish accent #514

Open jbellik opened 3 years ago

jbellik commented 3 years ago

A function that takes the output of GEN and for each tree in the candidate set, returns several trees...

addMinimalPhiHeads(ptree): returns {all trees q such that q is like ptree except that every minimal phi has a head labeled either at its left edge or its right edge}

Nick Kalivoda's specifications: every minimal φ is marked by Gen as either being left-headed or right-headed. So for three words, just looking at the minimal φ's,:

(a b c) (a b c) (a) (b c) (a) (b c) (a b) (c) (*a b) (c*) (a) (b) (c*)

NOT generated: (a b* c)

nkalivoda commented 3 years ago

Let's name this something like "minimal φ head side".

The thing that ended up italics in the above list is supposed to be (a b) (c) with an asterisk after "a" and one after "c".

Switching to acute accents for this comment to avoid further italics

Something that the three-word list didn't make clear is that we want left- and right-headed minimal φs to be able to co-occur. So for (ω ω) (ω ω), we want:

LL: (ώ ω) (ώ ω) LR: (ώ ω) (ω ώ) RL: (ω ώ) (ώ ω) RR: (ω ώ) (ω ώ)

And just to reiterate, only words in a minimal φ can be accented. Not allowed: {ώ...}, (...φ...ώ...), etc.

Theoretially, the thought behind this is that children in level-skipping configurations, and/or nodes lower than one of their siblings, are "adjoined", so they can't be the head of their parent. Equivalently (I think), if we interpret "head" as "accented", then running AccentAsHead on any of these structures should give 0 violations.

nkalivoda commented 3 years ago

The function seems to be incorrect, unless I'm missing something. In the second tableau in minimalPhiHead.html, exactly one daughter of ι is head-marked, and there is no head-marking within minimal φs. For instance, {(a b c)′}, {a b c′}, {(a b) c′}, which I thought would be impossible. For three words and [–Hd, –Exh, –NR, +Branching], I would have expected the following list:

{(a b c′)} {(a′ b c)} {((a b′) c)} {((a′ b) c)} {(a (b c′))} {(a (b′ c))} {a b c} {(a b′) c} {(a′ b) c} {a (b c′)} {a (b′ c)}

jbellik commented 3 years ago

Hi Nick, Despite its name, the function is actually written to be generalizable across categories, to create edge-aligned heads of minimal nodes of any specified category. The second tableau specifies that we want heads of minimal iotas to be labeled. That's my bad for not labeling the tableaux at all! Sorry for the confusion. I just labeled the tableaux in the test file. I think the function is working correctly but do let me know if I've missed something.

On Sun, Jan 31, 2021 at 11:41 PM Nick Kalivoda notifications@github.com wrote:

The function seems to be incorrect, unless I'm missing something. In the second tableau in minimalPhiHead.html, exactly one daughter of ι is head-marked, and there is no head-marking within minimal φs. For instance, {(a b c)′}, {a b c′}, {(a b) c′}, which I thought would be impossible. For three words and [–Hd, –Exh, –NR, +Branching], I would have expected the following list:

{(a b c′)} {(a′ b c)} {((a b′) c)} {((a′ b) c)} {(a (b c′))} {(a (b′ c))} {a b c} {(a b′) c} {(a′ b) c} {a (b c′)} {a (b′ c)}

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/syntax-prosody-ot/main/issues/514#issuecomment-770642954, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC7ROKSE26EMMZ7X3UPVXHLS4ZLLVANCNFSM4UJLTZ4A .

--

-- Jennifer Bellik, PhD Post-doctoral researcher & lecturer UC Santa Cruz https://people.ucsc.edu/~jbellik/

nkalivoda commented 3 years ago

Excellent, glad it was intentional!