Paradigm Uniformity is not part of Phonology Noam

  • Slides: 52
Download presentation
 Paradigm Uniformity is not part of “Phonology” Noam Faust Université Paris 8

Paradigm Uniformity is not part of “Phonology” Noam Faust Université Paris 8

Architecture Word formation (UR formation) /UR/ phonology has a say inside here Phonology has

Architecture Word formation (UR formation) /UR/ phonology has a say inside here Phonology has a say in this process Phonetic form Is it the same Phonology ? Possibly not!

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of a paradigm if it is triggered or blocked only in one » (my definition)

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of a paradigm if it is triggered or blocked only in one » Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity « a phonological process is triggered or blocked The occurence of throughout

Paradigm uniformity « a phonological process is triggered or blocked The occurence of throughout the word-forms of a paradigm if it is these schwas is triggered or blocked only in one » determined by PU Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of a paradigm if it is phonologically triggered or blocked only in one » unnecessary here: /tʁɔknə/ is fine. Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity phonologically « a phonological process is triggered or blocked necessary here: throughout

Paradigm uniformity phonologically « a phonological process is triggered or blocked necessary here: throughout the word-forms of a paradigm if it is */tʁɔknən/ is phonologically triggered or blocked only in one » impossible. unnecessary here: /tʁɔknə/ is fine. Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity phonologically « a phonological process is triggered or blocked necessary here: throughout

Paradigm uniformity phonologically « a phonological process is triggered or blocked necessary here: throughout the word-forms of a paradigm if it is */tʁɔknən/ is and therefore Phonologically (? ? ) triggered or blocked only in one » impossible. unnecessary here Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of a paradigm if it is triggered or blocked only in one » Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity phonologically « a phonological process is triggered or blocked « SHELL »

Paradigm uniformity phonologically « a phonological process is triggered or blocked « SHELL » necessary here: throughout the word-forms of a paradigm if it is a sequence CLiəLi is out if C<L */tʁɔknən/ is triggered or blocked only in one » and L=L in sonority impossible. (L=sonorant) Raffelsiefen (1995): Schwa economy in German /tʁɔkən/ ‘dry’ /tʁɔkənə/ ‘dry (fem. acc)’ /tʁɔkənən/ ‘dry (ms. acc)’

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of

Paradigm uniformity « a phonological process is triggered or blocked throughout the word-forms of a paradigm if it is triggered or blocked only in one » when SHELL is not violated in the paradigm, [ə] is not inserted. Raffelsiefen (1995): Schwa economy in German /dʊnkəl/ ‘dark’ /dʊnklə/ ‘dark (fem. acc)’ */dʊŋkələ/ /dʊnklən/ ‘dark (ms. acc)’ */dʊŋkələn/

(Raffelsiefen 1995 in fact claims, as I will claim here, that paradigm uniformity operates

(Raffelsiefen 1995 in fact claims, as I will claim here, that paradigm uniformity operates at the phonemic level, where no C is syllabic in German, and yields): /tʁɔkən, tʁɔkənən. . / /dʊnkəl, dʊnk_lən…/ For this reason, Raffelsiefen must say that Unsuffixed base must not be part of the paradigm…

(Raffelsiefen 1995 in fact claims, as I will claim here, that paradigm uniformity operates

(Raffelsiefen 1995 in fact claims, as I will claim here, that paradigm uniformity operates at the phonemic level, where no C is syllabic in German, and yields): /tʁɔkən, tʁɔkənən. . / /dʊnkəl, dʊnk_lən…/ For this reason, Raffelsiefen must say that Unsuffixed base must not be part of the paradigm…

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə]

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə] ‘open. pl’ [ofnə] unattested adj. but

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə]

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə] ‘open. pl’ [ofnə] unattested adj. but a) the suffix /-n/ surfaces as [m] after [n] [ofənəm] (Cnəm not a SHELL violation!)

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə]

Paradigm uniformity Meanwhile, in Standard Yiddish Same impossibility as German [ofn ] ‘open’ [ofənə] ‘open. pl’ [ofnə] unattested adj. but a) the suffix /-n/ surfaces as [m] after [n] [ofənəm] (Cnəm not a SHELL violation!) b) [Cnəm] is phonologically OK if [ə] is in base: [modnə] ‘strange’ [modnəm] ‘ms. dat’, [modənəm] unattested

Paradigm uniformity Meanwhile, in Standard Yiddish The base must be important Same impossibility as

Paradigm uniformity Meanwhile, in Standard Yiddish The base must be important Same impossibility as German [ofn ] ‘open’ [ofənə] ‘open. pl’ [ofnə] unattested adj. but a) the suffix /-n/ surfaces as [m] after [n] [ofənəm] (Cnəm not a SHELL violation!) b) [Cnəm] is phonologically OK if [ə] is in base: [modnə] ‘strange’ [modnəm] ‘ms. dat’, [modənəm] unattested

Proposal Given the phonetic form of the bases [ofn ] ‘open’ [modnə] ‘strange’ Language

Proposal Given the phonetic form of the bases [ofn ] ‘open’ [modnə] ‘strange’ Language seeks to stabilize one underlying form for the entire paradigm. For [ofn ] there are two candidates: /ofn/ => [ofn ] /ofən/ => [ofn ] (CəL# is generally ruled out in Yiddish)

Proposal For [ofn ] there are two candidates: /ofn/ => [ofn ] /ofən/ =>

Proposal For [ofn ] there are two candidates: /ofn/ => [ofn ] /ofən/ => [ofn ] (CəL# is generally ruled out in Yiddish) These two candidates are confronted with the set of suffixes /ə/ and /-N, -ʁ/ (without /ə/): /ofn, ofn-ə, ofn-N, ofn-ʁ/ vs. /ofən, ofən-ə, ofən-N, ofən-ʁ/

Proposal For [ofn ] there are two candidates: /ofn/ => [ofn ] /ofən/ =>

Proposal For [ofn ] there are two candidates: /ofn/ => [ofn ] /ofən/ => [ofn ] (CəL# is generally ruled out in Yiddish) These two candidates are confronted with the set of suffixes /ə/ and /-N, -ʁ/ (without /ə/): /ofn, ofn-ə, ofn-N, ofn-ʁ/ vs. /ofən, ofən-ə, ofən-N, ofən-ʁ/ A bad UR, n cannot be the nucleus of N#

Proposal The selected base being /ofən/ The lexicon therefore delivers /ofən, ofən-ə, ofən-N, ofən-ʁ/

Proposal The selected base being /ofən/ The lexicon therefore delivers /ofən, ofən-ə, ofən-N, ofən-ʁ/ The phonology gives [ofn , ofən-əm, ofən-əʁ]

Proposal For [modnə] there is only one candidate /modnə/ (As we already saw /modənə/

Proposal For [modnə] there is only one candidate /modnə/ (As we already saw /modənə/ will give unattested [modənə]) Thes lexicon therefore delivers /modnə, modnə-N, modnə-ʁ/ And phonology gives [modnə, modnəm, modnəʁ]

Crucially Language seeks to stabilize one underlying form for the entire paradigm = Lexicon-optimization,

Crucially Language seeks to stabilize one underlying form for the entire paradigm = Lexicon-optimization, rather than phonology = Paradigm Uniformity as an anti-allomorphy force (Raffelsiefen 2016)

Architecture PU effects Word formation (UR formation) belong here /UR/ phonology has a say

Architecture PU effects Word formation (UR formation) belong here /UR/ phonology has a say inside here Phonology has a say in this process Phonetic form Is it the same Phonology ? Apparently not!

More about Proposal Joins the view that PU happens in learning: Speakers learn that

More about Proposal Joins the view that PU happens in learning: Speakers learn that stems ending in /Cn / are illicit adjectival stems. In other words, in order to enter the verbal system of Yiddish, there are phonological requirements (see appendix for Modern Hebrew analogue). Prediction: these requirements will not necessarily hold for /Cl / (appendix abour [ejdl ] ‘genteel’)

Thank you!

Thank you!

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] [ejdəln

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] [ejdəln ] [ejdl t] [ejdələʁ] [ejdl st] [ejdlən] COMPARE!!

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no -ʁ in the [ejdəln ] [ejdl t] verbal [ejdələʁ] [ejdl st] paradigm! [ejdlən] COMPARE!!

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no -ʁ in the [ejdəln ] [ejdl t] verbal [ejdələʁ] [ejdl st] paradigm! [ejdlən] adj: /ejdl/ or /ejdəl/ given /ejdl-n, ejdl-ə, ejdl-ʁ/

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no -ʁ in the [ejdəln ] [ejdl t] verbal [ejdələʁ] [ejdl st] paradigm! [ejdlən] adj: /ejdl/ or /ejdəl/ given /ejdl-n, ejdl-ə, *ejdl-ʁ/ A bad UR, l cannot be the nucleus of ʁ#

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no -ʁ in the [ejdəln ] [ejdl t] verbal [ejdələʁ] [ejdl st] paradigm! [ejdlən] verb: /ejdl/ or /ejdəl/ given /ejdl-t, ejdl-st, ejdl-n/

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ [ejdələ] [ejdl ] no -ʁ in the [ejdəln ] [ejdl t] verbal [ejdələʁ] [ejdl st] paradigm! [ejdlən] verb: /ejdl/ or /ejdəl/ given /ejdl-t, ejdl-st, ejdl-n/ no problem, no need to insert schwa for now

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ /ejdəl/ /ejdəl+ə/ => [ejdələ]

Yiddish [ejdl ] ‘genteel’ Adj verb (+[zix]) ‘play it genteel’ /ejdəl/ /ejdəl+ə/ => [ejdələ] /ejdl+ø/ => [ejdl ] /ejdəl+n/ => [ejdəln ] /ejdl+t/ => [ejdl t] /ejdəl+ʁ/ =>[ejdələʁ] /ejdl+t/ => [ejdl st] /ejdl+n/ =>[ejdlən]

Two underlying forms for the same morpheme? ? ? ? verb adj /ejdəl/ /ejdəl+ə/

Two underlying forms for the same morpheme? ? ? ? verb adj /ejdəl/ /ejdəl+ə/ => [ejdələ] /ejdəl+n/ => [ejdəln ] /ejdəl+ʁ/ =>[ejdələʁ] /ejdl/ /ejdl+ø/ => [ejdl ] /ejdl+t/ => [ejdl t] /ejdl+t/ => [ejdl st] /ejdl+n/ =>[ejdlən] Same item, two different URs?

Two underlying forms for the same morpheme? √ejdl verb adj /ejdəl/ /ejdəl+ə/ => [ejdələ]

Two underlying forms for the same morpheme? √ejdl verb adj /ejdəl/ /ejdəl+ə/ => [ejdələ] /ejdəl+n/ => [ejdəln ] /ejdəl+ʁ/ =>[ejdələʁ] /ejdl/ /ejdl+ø/ => [ejdl ] /ejdl+t/ => [ejdl t] /ejdl+t/ => [ejdl st] /ejdl+n/ =>[ejdlən] Phonological Index ≠ UR

Noun => Verb in Modern Hebrew Noun Verb Katalog + i e

Noun => Verb in Modern Hebrew Noun Verb Katalog + i e

Noun => Verb in Modern Hebrew Noun Verb Katalog + i e => [kitleg]

Noun => Verb in Modern Hebrew Noun Verb Katalog + i e => [kitleg]

Noun => Verb in Modern Hebrew Noun Verb Katalog i e

Noun => Verb in Modern Hebrew Noun Verb Katalog i e

Noun => Verb in Modern Hebrew Noun Verb K ta l g i e

Noun => Verb in Modern Hebrew Noun Verb K ta l g i e

Noun => Verb in Modern Hebrew Noun Verb K t l g i e

Noun => Verb in Modern Hebrew Noun Verb K t l g i e

Noun => Verb in Modern Hebrew Noun Verb xantariʃ + => [xintreʃ] i e

Noun => Verb in Modern Hebrew Noun Verb xantariʃ + => [xintreʃ] i e ‘charlatan’ ‘charlatanize’

Noun => Verb in Modern Hebrew Noun Verb e xantariʃ i

Noun => Verb in Modern Hebrew Noun Verb e xantariʃ i

Noun => Verb in Modern Hebrew Noun Verb e x ntar ʃ i

Noun => Verb in Modern Hebrew Noun Verb e x ntar ʃ i

Noun => Verb in Modern Hebrew Noun Verb e x nt r ʃ i

Noun => Verb in Modern Hebrew Noun Verb e x nt r ʃ i

Noun => Verb in Modern Hebrew Noun Verb e x nt r ʃ i

Noun => Verb in Modern Hebrew Noun Verb e x nt r ʃ i The derivation may create a novel branching onset

Noun => Verb in Modern Hebrew Noun Verb ramatkal + => ? ? ?

Noun => Verb in Modern Hebrew Noun Verb ramatkal + => ? ? ? i e ‘chief of army’ ‘act as Co. A”

Noun => Verb in Modern Hebrew Noun Verb ramatkal + => *** i e

Noun => Verb in Modern Hebrew Noun Verb ramatkal + => *** i e (*rimtkel) Not everything is possible: phonology blocks some stem formations.

Architecture Stem formation Word formation (UR formation) /UR/ phonology has a say Phonology has

Architecture Stem formation Word formation (UR formation) /UR/ phonology has a say Phonology has a say in this process Phonetic form Is it the same Phonology ? Apparently not!

Inflection cannot create branching onset sg. kiteʁ pl. kitʁu kinter kinteʁu, *kintʁu ‘complain incessantly’

Inflection cannot create branching onset sg. kiteʁ pl. kitʁu kinter kinteʁu, *kintʁu ‘complain incessantly’ ‘taunt’

Inflection cannot create branching onset sg. kiteʁ kinter pl. kitʁu ‘complain incessantly’ syncope kinteʁu,

Inflection cannot create branching onset sg. kiteʁ kinter pl. kitʁu ‘complain incessantly’ syncope kinteʁu, *kintʁu ‘taunt’ Phonological rules cannot create branching onsets.

Inflection cannot create branching onset sg. kiteʁ kinter pl. kitʁu ‘complain incessantly’ syncope kinteʁu,

Inflection cannot create branching onset sg. kiteʁ kinter pl. kitʁu ‘complain incessantly’ syncope kinteʁu, *kintʁu ‘taunt’ n o i t a Phonological rules cannot create m r fo branching onsets. R U e k i l Un

Architecture PU effects Word formation (UR formation) belong here /UR/ phonology has a say

Architecture PU effects Word formation (UR formation) belong here /UR/ phonology has a say inside here Phonology has a say in this process Phonetic form Is it the same Phonology ? Apparently not! This phonology cannot create branching onsets