ÿWPC! ÆxýÈ•éý#x¶fõxŽS ¤_‘€¢Vº‚o=Q,Gïˆ .*µ˜N¹kCIu¾r¹¼Où&B<ÃJ䣡(vòü‰ùŒ†Ò/¬©ö/ÔùtkJ´%ÅÂ…¥ü}¥S••É¢ct2yå´PÁi?MR^o¡ý¡!ð¥òKê^ZD)¹Õ²cТžâ›ž œÀ³ÀŒƒPT£ß!£™ùßÜA¡‘L³(SØ—ûn6“|ê~ì]ÎH‹Ãp~¨wï7™4â©û¥2¨5ų„6¸0Í*`E›4+›t×áar-ouŸµ¤kXûi¡g¥ŸB^é¼oÞ<ú@o‡ƒHx@€ªå&!¶"Äqê‡ôî¹B®&Ÿà.€Àë)P.Jí"žC¶Ì½sÆ—6˜KØ“§¨áh³‹‡â.x¶Ü'Ƈ ºü¿Š&Öu’D…Ò‡zC;mÄfèWaDAÀX×$ {±¿Óê̹8Òª‚ÃaM OÐ$tvÁŠt{ÿ?õ —w)¥íÍdV"z^•º>2 ,³RißTq_hú1`¨Ë3ÖQR¡«ÿ`—R¸at­£›ªß/ŠXŒa·nºÌ`%çÀyA',~ä.ØkÇ©j”äû'4U-BØ %¥  0ìÅ×± 0ìˆ 0CtI· 09 ÙM B&‰hUBñ3n®C 0ÑñÂ"ÂUBäH&UBnnnnnnnnnn B°°°°°°°°°°°°°°°°°°w@Í4 ô\  `*Times New RomanTTX ›Ý ƒv 'ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  ÝòòArmin€Mester€&€Jaye€PadgettóóÐ X Ð(Ä2v $§§Ý ƒR;!ÝÑ7€÷nXXdìdì7ÑÑX°ÑÑX°ÑÓ  ÓÑ  ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  Ý ÍÝ ƒv (ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  ÝÓ  ÓòòDirectional€Syllabification€in€Generalized€AlignmentóóÐ X ÐÓ  ÓÌ(Ä2v $¨¨Ý ƒR;!ÝÑ7€÷nXXdìdì7ÑÑX°ÑÑX°ÑÓ  ÓÑ  ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  Ý(ÿÿ$òòÚ  ÚÚ  Úóó ?Ô€ôêòôXXÔ*This€note€was€first€posted,€in€essentially€its€present€form,€as€an€electronic€text€in€the€Rutgers€Optimality€Archive€(FallÐ X Ð1993).€We€gratefully€acknowledge€significant€help€from€discussions€and€correspondence€with€Junko€Itð=ð,€EllenÏBroselow,€John€McCarthy,€Alan€Prince,Cheryl€Zoll,€and€PASC€editors€Jason€Merchant€and€Rachel€Walker.€It€was€Zoll'sÏwork€on€Yawelmani€Yokuts€(Zoll€1993)€that€first€inspired€us€to€confront€the€issues€dealt€with€in€this€note.€As€usual,€allÏremaining€shortcomings€are€our€responsibility€alone. Ý ƒ ÏÂ$ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÓÓÔ€ôêòôXX÷nÔòòÚ  Ú1Ú  ÚóóÝ  Ý€€Though€not€with€level€ordering.€See€McCarthy€and€Prince€1993a€for€some€argumentation.€(ø2ÏÂ$¤¤Ý ƒR;!ÝÑ7€÷nXXdìdì7ÑÑX°ÑÑX°ÑÓ  ÓÑ  ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  ÝÓÓÔ€ôzÉôXX÷nÔòòÚ  Ú0Ú  Úóó ÏÝ ƒR;!ÝÑ7À÷nXXdìdì7ÑÑ@X°ÑÑ@X°ÑÓ  ÓÑ @ ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  ÝÔ€ôzÉôXX÷nÔòòÚ  Ú2Ú  Úóó€€John€McCarthy€and€Alan€Prince€independently€pointed€out€the€feasibility€of€such€a€'same-edge'€analysis.€Our€originalÐ X Ðproposal€was€stated€in€opposite-edge€terms.€See€below€for€discussion. 8Ý ƒ ÏÂ$ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÓÓÔ€ôzÉôXX÷nÔòòÚ  Ú3Ú  ÚóóÝ  Ý€€The€format€of€our€constraint€tableau,€lifted€directly€from€a€previous€ASCII€instantiation,€should€be€understood€as€follows:Ð X Ð'm'€denotes€a€mora;€(!)...(!)€indicates€indeterminacy€of€fatality€of€violation,€given€the€indeterminacy€of€ranking€between€FillÏand€Syll„Align.€The€optimal€form€is€marked€on€the€right€of€the€tableau€with€'$',€in€the€notation€of€Grimshaw€(1993). Ý ƒ ÏÂ$ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÓÓÔ€ôzÉôXX÷nÔòòÚ  Ú4Ú  ÚóóÝ  Ý€€As€comparison€of€(5a&b)€should€make€clear,€we€cannot€achieve€the€directionality€effect€by€counting€syllables€themselves.Ð X ÐOn€the€possibility€of€counting€segments,€see€later€discussion.=þ6X9`(*Courier 10pt12cpi3|P:«ô\  `*Times New RomanTTXXPô\  PŽ6QXPô\  `*Times New RomanTT¼¼^ô\  PŽ6Q¼Pô\  `*Times New RomanTTôôCô\  PŽ6QôP=þ6X9`(*Courier 10pt12cpiììdþ6X@É?BQì@(©R;$¡¡Ñ7€÷nXXdìdì7ÑÑX°ÑÑX°ÑÓ  ÓÑ  ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔ Ý ƒ a$ÝÓ  ÓÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÓÓÔ€ôzÉôXX÷nÔÒ°ÒòòÚ  Ú1Ú  ÚóóÝ  Ý€€Though€not€with€level€ordering.€See€McCarthy€and€Prince€1993a€for€some€argumentation.€ô\  `*Times New RomanTT >Ô€ôzÉôXXÔThis€note€was€first€posted,€in€essentially€its€present€form,€as€an€electronic€text€in€the€Rutgers€Optimality€Archive€(FallÐ X Ð1993).€We€gratefully€acknowledge€significant€help€from€discussions€and€correspondence€with€Junko€Itð=ð,€EllenÏBroselow,€John€McCarthy,€Alan€Prince,Cheryl€Zoll,€and€PASC€editors€Jason€Merchant€and€Rachel€Walker.€It€was€Zoll'sÏwork€on€Yawelmani€Yokuts€(Zoll€1993)€that€first€inspired€us€to€confront€the€issues€dealt€with€in€this€note.€As€usual,€allÏremaining€shortcomings€are€our€responsibility€alone.ô\  `*Times New RomanTT d /Q0Q1Q23t4t6t7t8:ËŠ~? Ý ƒR;!ÝÑ7€÷nXXdìdÈ7ÑÑX°ÑÑX°ÑÓ  ÓÑ  ÑÔ€X÷nXXXÔÔ€X÷nXXX÷nÔÝ  ÝÖ€ÿÿÖÖ€ÿÿÖÑ  ÑØOØà€  @(#àPhonology€at€Santa€Cruz,€Vol.€3,€1994,€pp.€79„85ˆÐ X ÐÌÌÌÓ  Óò òÔ€¼|¼XX÷nÔDirectional€Syllabification€in€Generalized€Alignmentó óÔ€X÷nX¼¼|Ôñ7ññ3ñÙ*ƒR;!Ùñ3ññ7ññ7ñÙ*ƒR;!Ùñ7ññ8ñ׃×ñ8ññ8ñ׃×ñ8ñÝ ƒÿÿÝòòÚ  Ú*Ú  ÚóóÝ  Ý×  ×ñ6ññ4ñÙ*R;!Ùñ4ññ6ññ6ñÙ*R;!Ùñ6ñØØÐ ° ÐÓ  ÓÌÓ  ÓArmin€Mester€&€Jaye€PadgettUniversity€of€California,€Santa€Cruz¹Ó  ÓÌÌÌOne€of€the€central€analytical€strategies€of€generative€phonology€in€the€past€has€been€the€attempt€toÏexplain€all€ò òdirectionality€effectsó ó€as€the€result€of€ò òdirectional€structure€building€algorithmsó ó€whichÐ ¤ L  Ðmove€in€a€step„by„step€iterative€fashion€from€one€end€of€an€input€form€to€the€other€end.€As€anÏexample,€consider€the€familiar€case€where€disyllabic€trochaic€feet€are€built€from€left€to€right€(as€inÏPintupi,€according€to€Hayes€1991).€Given€an€input€with€an€odd€number€of€syllables€like€ò ò/SSSSSSS/ó óÐ l  Ð(S=syllable),€the€relevant€part€of€its€derivational€history€looks€as€follows:€ò ò/SSSSSSS/€ð8ð€[SS]SSSSSÐ \ Ðð8ð€[SS][SS]SSS€ð8ð€[SS][SS][SS]Só ó,€terminating€in€a€representation€where€the€foot€parse€is€flush€withÐ Lô Ðthe€left€edge,€but€not€with€the€right€edge€(i.e.,€with€a€left„edge€directional€bias).€For€the€same€kind€ofÏinput,€right„to„left€foot€building€yields€the€opposite€result:€ò òS[SS][SS][SS]ó ó.€Ð (Ð Ðà  àIn€other€words,€standard€derivational€models€of€phonology€seek€the€explanation€for€such€kindsÏof€directional€bias€in€the€way€in€which€the€output€structures€have€been€built€up€during€the€derivation,Ïpiece€by€piece.€Directionality€effects€are€not€limited€to€prosodic€phonology€and€have€been€found€andÏdiagnosed€in€essentially€all€areas€of€phonology€and€prosodic€morphology,€including€tonal€association,Ïsegmental€processes,€harmony€processes,€feature„sized€morpheme€associations,€etc.€Ìà  àAs€is€well€known,€Optimality€Theory€(Prince€and€Smolensky€1993),€an€approach€€basedÏexclusively€on€wellformedness€constraints€instead€of€operations,€proposes€to€do€away€with€aÏsequential€derivation.ñ2ñ׃ d×ñ2ññ2ñ׃d×ñ2ñÝ ƒÿÿÝòòÚ  Ú1Ú  ÚóóÝ  Ý×  ×€In€its€place,€the€theory€posits€parallel€evaluation€of€an€indefinitely€large€andÐ Œ4 Ðinclusive€pool€of€candidate€outputs.€This€evaluation€results€in€the€selection€of€the€candidate(s)€thatÏoptimally€satisfy€a€given€set€of€strictly€ranked€constraints.€These€constraints,€the€totality€of€whichÏmake€up€the€grammar€(here,€phonology)€of€a€language,€are€in€principle€all€violable;€optimality€consistsÏin€minimal€violation€(see€Prince€and€Smolensky€1993,€McCarthy€and€Prince€1993a,b,c,€Itð=ð,€Mester,Ïand€Padgett€(1993,€to€appear)€and€other€related€work€for€further€discussion).€Ìà  àIn€its€crucial€reliance€on€the€existence€of€a€sequential€derivation,€the€derivational€theory€ofÏdirectionality€effects€is€sometimes€perceived€as€a€serious€obstacle€to€modern€constraint„basedÏtheories;€the€issue€arises€in€a€number€of€constraint„based€frameworks€other€than€Optimality€Theory,Ïall€of€which€shun€sequential€derivations€e.g.,€Declarative€ÌPhonology€(Scobbie€1991,€1993,€Broe€1993),€Harmonic€Phonology€(Goldsmith€1990,€1991),€andÏothers.€Ìà  àIn€their€paper€introducing€Generalized€Alignment,€McCarthy€and€Prince€(1993b)€(henceforth,ÏGA)€have€developed€a€technique€for€capturing€directional€foot€parsing€effects€in€Optimality€Theory.ÏBuilding€on€this€work,€the€purpose€of€the€present€note€is€a€very€modest€one:€we€would€like€extendÐ t'%(W Ðtheir€result€to€directional€syllabification€effects,€which€were€systematically€explored€and€analyzed,€inÏthe€context€of€an€overall€prosodic€theory€of€syllable€structure,€in€Itð=ð's€work€(Itð=ð€1989,€241-254;€1986,Ï163-219).€Ìà  àSomewhat€comparable€to€the€reanalyses€of€'phonotactics€plus€repair'€theories€found€in€chapterÏ10€of€Prince€and€Smolensky€1993,€the€status€of€our€work€is€very€much€that€of€a€demonstration€ofÏexistence:€we€show,€for€a€representative€case€with€directional€syllabification€effects,€how€anÏoptimality„theoretic€equivalent€to€the€standard€operational€analysis€can€be€constructed.€We€are€notÏclaiming€that€the€analysis€outlined€is€necessarily€the€best€way€of€capturing€the€effects.€In€particular,Ïour€reconstruction€is€strictly€based€on€Itð=ð's€(1986,€1989)€work€and€does€not€take€into€account€theÏresults€of€work€done€since€then.€Ìà  àOur€proposal€makes€use€of€a€central€analytical€strategy€in€GA,€where€directionality€effects€inÏfoot€parsing€are€derived€by€means€of€a€constraint€requiring€the€alignment€of€every€foot€edgeÏ(left/right)€with€a€prosodic€word-edge€(left/right).€As€demonstrated€in€GA,€the€doctrine€of€minimalÏviolation€ensures€the€selection€of€the€candidate€displaying€the€desired€directional€footing€pattern€(aÏsuggestion€originally€due€to€R.€Kirchner).€This€is€so€in€spite€of€massive€violations€of€the€alignmentÏconstraint€incurred€by€all€feet€not€directly€located€at€the€relevant€word€edge.Ìà  àTaking€up€the€alignment€strategy€of€GA,€we€show€that€directional€syllabification€patterns€areÏthe€result€of€alignment€constraints€requiring€that€every€syllable€edge€coincide€with€a€prosodicÏword-edge.€The€general€schema€is€given€in€(1),€where€E(dge)€is€a€variable€ranging€over€{L(eft),ÏR(ight)}€and€'Syll'€abbreviates€'syllable':׃ ×Ý ƒÿÿÝòòÚ  Ú2Ú  ÚóóÝ  Ý×  ×Ð |„ ÐÌ(1)€€Syll-Align€(Syll,Edge,PrWd,Edge)ÌÌà  àFor€clarification,€we€note€that€this€statement€is€to€be€interpreted€as€quantifying€ò òuniversallyó óÐ ,4 Ðover€tokens€of€Syll€and€ò òexistentiallyó ó€over€tokens€of€PrWd€(see€GA€for€a€more€formal€developmentÐ $ Ðand€for€further€exemplification):€ò òeveryó ó€syllable€must€be€Edge-aligned€with€ò òsomeó ó€prosodic€word.€InÐ   Ðmost€situations,€(1)€is€massively€violated€by€syllables€not€located€at€Edge.€For€a€given€input,€theÏoptimal€candidate€is€the€one€with€the€fewest€violations€(modulo€the€overall€ranking€of€constraints,€asÏusual€in€OT).€Ìà  àRight„Left€(R„L)€syllabification€and€epenthesis€(as€in€Iraqi€Arabic;€see€Itð=ð€1989,€who€buildsÏon€analyses€of€Broselow€1980,€1982€and€Selkirk€1981)€shows€the€pattern€in€(2)€(the€'Coda'-€pattern:Ïthe€stray€òòcóó€is€syllabified€as€a€coda€to€the€epenthetic€vowel,€whose€site€is€marked€by€capital€'V'):Ð ˜!  ÐÌ(2)€€[...]cvcccv[...]€ðð€[...].cv.cVc.cv.[...]Ð p#x! ÐÌThus€/gil-t-l-a/€appears€as€òò.gi.lIt.la.óó€'I€said€to€him',€etc.€This€pattern€follows€from€the€L-Edge€versionÐ H%P # Ðof€the€alignment€constraint€in€(3),€under€the€overall€constraint€ranking€in€(4)€(see€Prince€andÏSmolensky€1993€and€McCarthy€and€Prince€1993b€on€constraint€formulations).ÌÌ(3)€€Syll-ALIGN(L):€Align€(Syll,L,PrWd,L)Ìâ âÐ ä)ì$( Ðñ:ñÔ& ° Ôâ âñ:ñ(4)à  à|€ONSET€à ¸ à|€à  à>>€à h à|Syll-ALIGN(L)à p à|à È à€>>à  à€NO-CODAÐ ä Ðâ  €€€à  à|€PARSE€à ¸ à|€€€€€€à  àà h à|€FILL€à  àà p à|Ð Ðì Ѐ€€€€à  à|*COMPLEXà ¸ à|Ð ¼Ø Ðñ:ñÔ'°ä4Ôñ:ñÌà  àIn€order€to€see€how€this€works,€the€reader€is€invited€to€inspect€the€tableau€in€(5),€whichÏdemonstrates€that€a€gradiently€violable€understanding€of€Syll„Align€achieves€the€needed€directionalityÏeffect.׃ ×Ý ƒÿÿÝòòÚ  Ú3Ú  ÚóóÝ  Ý×  ×€We€leave€out€the€undominated€constraints€for€the€purposes€of€the€present€discussion.€TakingÐ l ˆ Ðup€a€suggestion€by€John€McCarthy,€we€assume€that€violations€of€alignment€constraints€are€reckonedÏin€terms€of€the€prosodic€unit€immediately€adjacent€(one€level€down)€in€the€prosodic€hierarchy€fromÏthe€unit€referred€to€in€the€first€part€of€the€constraint.€(3)€quantifies€universally€over€syllables€(loosely,Ï"is€a€constraint€on€syllables"),€hence€it€makes€sense€for€violations€to€be€computed€in€terms€of€moraÏcount.׃ Ü×Ý ƒÿÿÝòòÚ  Ú4Ú  ÚóóÝ  Ý×  ×€Let€us€focus€on€the€violations€of€Syll„Align(L)€here,€given€syllable„by„syllable€forÐ $  Ðconvenience.€The€constraint,€recall,€requires€every€Syll€to€be€left„aligned€with€PrWd.€The€first€syllableÏof€any€form€satisfies€the€constraint€completely,€naturally,€and€so€no€violations€are€counted€in€the€firstÏcolumn.€Some€differentiation€occurs,€however,€already€by€the€second€syllable,€which€in€candidate€(5b)Ïfails€Syll„Align(L)€by€two€moras.€We€do€not€indicate€a€fatal€violation€in€this€column;€as€McCarthy€andÏPrince€(1993b)€note,€the€desired€result€can€be€had€by€taking€together€at€once€violations€incurred€byÐ ¤À Ðò òalló ó€syllables.€Thus,€the€optimal€candidate€(5a)€violates€Syll„Align(L)€by€four€moras€total;€(5b,c)€areÐ ¬ Ðremoved€from€consideration€by€incurring€five€violations€or€more.ÌÌ(5)€€Align(L)€syllabification€tableau€(corresponds€to€R„L€syllabification)ÌÌÔ€ìzÅ ìXX÷nÔÔ€ìcy ì ììzÅÔ€€€€€€€€€€€€€€€||€FILL€€:€€€Syll-ALIGN(L)€€€€€€€€€€€€€€€€|€€NoCodaÐ 0L Ð===============||=======:================================|==========Ì€€€/cvcccv/€€€€||€€€€€€€:€€€S1€€€S2€€€€€S3€€€€€€S4€€€€€€€|Ì---------„„„---||-------:--------------------------------|----------Ìa.€cv.cVc.cv€€€||€€*€€€€:€€€€€€€€m€€€€€€mmm€€€€€€€€€€€€€€|€€€€*€€€€ò ò$ó óÐ ÀÜ Ð------„„„------||-------:--------------------------------|----------Ìb.€cvc.cV.cv€€€||€€*€€€€:€€€€€€€€mm€€€€€mmm!€€€€€€€€€€€€€|€€€€*Ì---„„„---------||-------:--------------------------------|----------Ìc.€cv.cV.cV.cv€||€€**(!):€€€€€€€€m€€€€€€mm€€€€€€mm(!)m€€€|Ì„„„------------||-------:--------------------------------|----------Ô€Xk– X ììcyÔÔ€X÷nX XXk–ÔÐ ü ÐÌà  àL„R€syllabification€and€epenthesis€(as€in€Cairene€Arabic;€see€Itð=ð€1989€and€references€therein)Ïshows€the€pattern€in€(6)€('Onset'-pattern:€the€stray€òòcóó€is€taken€as€the€onset€to€the€epenthetic€vowel):€Ð x"”" ÐÌ(6)€€[...]cvcccv[...]€ðð€[...].cvc.cV.cv.[...]Ð P$l$ ÐÌThus€/?ul-t-l-u/€appears€as€òò.?ul.tI.lu.óó€'I€said€to€him',€etc.€This€pattern€follows€from€the€R-Edge€versionÐ (&D & Ðof€the€alignment€constraint€(7),€under€the€overall€constraint€ranking€in€(8)€(which€is€identical€to€thatÏof€(4)€in€all€other€respects).Ð ("( Ї(7)€€Syll-ALIGN(R):€Align€(Syll,R,PrWd,R)ÌÌ(8)à  à|€ONSET€à ¸ à|€à  à>>€à h à|Syll-ALIGN(R)à p à|à È à€>>à  à€NO-CODAÐ ÐØ Ѐ€€€€à  à|€PARSE€à ¸ à|€€€€€€à  àà h à|€FILL€à  àà p à|Ð ¼Ä Ѐ€€€€à  à|*COMPLEXà ¸ à|Ð ¨° ÐÌTableau€(9)€illustrates€this€point.€Though€'S1'€still€denotes€the€initial€syllable,€Syll„Align(R)€nowÏrequires€that€it€be€aligned€with€PrWd€at€the€right€edge.€Candidate€(9b)€is€the€winner.ÌÌÔ& ì Ô(9)€€Align(R)€syllabification€tableau€(corresponds€to€L„R€syllabification)ÌÌÔ€ìzÅ ìXX÷nÔÔ€ìcy ì ììzÅÔ€€€€€€€€€€€€€€€||€FILL€€:€€€Syll-ALIGN(R)€€€€€€€€€€€€€€€€|€€NoCodaÐ $  Ð===============||=======:================================|==========Ì€€€/cvcccv/€€€€||€€€€€€€:€€€S1€€€S2€€€€€S3€€€€€€S4€€€€€€€|Ì---„„„---------||-------:--------------------------------|----------Ìa.€cv.cVc.cv€€€||€€*€€€€:€€€mmm€€m!€€€€€€€€€€€€€€€€€€€€€€|€€€€*€€€Ì---„„„---------||-------:--------------------------------|----------Ìb.€cvc.cV.cv€€€||€€*€€€€:€€€mm€€€m€€€€€€€€€€€€€€€€€€€€€€€|€€€€*€€€€ò ò$ó óÐ ôü  Ð---„„„---------||-------:--------------------------------|----------Ìc.€cv.cV.cV.cv€||€€**(!):€€€mmm€€m(!)m€€m€€€€€€€€€€€€€€€€|Ì---„„„---------||-------:--------------------------------|----------ÌÔ'ìD ½-ÔÔ€Xk– X ììcyÔÔ€X÷nX XXk–ÔÌà  àThe€major€result€of€Itð=ð's€(1989)€analysis€is€its€abilityð!ð€without€invoking€any€additionalÏconstraintsð!ð€to€account€for€the€convergence€of€the€two€opposing€patterns€(R„L:€Iraqi€Arabic;€L„R:ÏCairene€Arabic)€in€the€case€of€òòccccóó€clusters:€both€languages€syllabify€this€case€as€in€(10).Ð PX ÐÌ(10)€[...]cvccccv[...]€ðð€[...].cvc.cVc.cv.[...]Ð (0 ÐÌà  àThus€R„L€Iraqi€Arabic€syllabifies€/gil-t-l-ha/€as€òò.gil.tIl.ha.€óó'I€said€to€her',€and€L„R€CaireneÐ  ÐArabic€answers€by€syllabifying€/?ul-t-l-ha/€as€òò.?ul.tIl.ha.óó€'I€said€to€her',€etc.€Our€alignment€analysisÐ ìô Ðreplicates€Itð=ð's€result,€as€demonstrated€without€commentary€in€(11)€and€(12).ÌÌÔ& 4 Ô(11)€Align(L)€syllabification€of€/...cccc.../ÌÌÔ€ìzÅ ìXX÷nÔÔ€ìcy ì ììzÅÔ€€€€€€€€€€€€||€FILL€€:€€€Syll-ALIGN(L)€€€€€€€€€€€€€€€€|€€NoCodaÐ ˆ!" Ð============||=======:================================|==========Ì/cvccccv/€€€||€€€€€€€:€€€S1€€€S2€€€€€S3€€€€€€S4€€€€€€€|Ì------------||-------:--------------------------------|----------Ìcvc.cVc.cv€€||€€*€€€€:€€€€€€€€€mm€€€€mmmm€€€€€€€€€€€€€|€€€€**€€€ò ò$ó óÐ $ & Ð------------||-------:--------------------------------|----------Ìcvc.cV.cV.cv||€€**(!):€€€€€€€€€mm€€€€mmm€€€€€mm(!)mm€€|€€€€*Ì------------||-------:--------------------------------|----------Ìcv.cV.cVc.cv||€€**(!):€€€€€€€€€m€€€€€mm€€€€€€mmmm(!)€€|€€€€*Ì------------||-------:--------------------------------|----------Ìâ âÔ€Xk– X ììcyÔÔ€X÷nX XXk–ÔÔ'4 °Y4ÔÐ ø'#, ÐÔ& ì Ôâ â(12)€Align(R)€syllabification€of€/...cccc.../ÌÌÔ€ìzÅ ìXX÷nÔÔ€ìcy ì ììzÅÔ€€€€€€€€€€€€||€FILL€€:€€€Syll-ALIGN(R)€€€€€€€€€€€€€€€€|€€NoCodaÐ ¼Ø Ð============||=======:================================|==========Ì/cvccccv/€€€||€€€€€€€:€€€S1€€€S2€€€€€S3€€€€€€S4€€€€€€€|Ì------------||-------:--------------------------------|----------Ìcvc.cVc.cv€€||€€*€€€€:€€€mmm€€m€€€€€€€€€€€€€€€€€€€€€€€|€€€€**€€€ò ò$ó óÐ L h Ð------------||-------:--------------------------------|----------Ìcvc.cV.cV.cv||€€**(!):€€€mmm€€mm(!)€€m€€€€€€€€€€€€€€€€|€€€€*Ì------------||-------:--------------------------------|----------Ìcv.cV.cVc.cv||€€**(!):€€€mmmm€m(!)mm€m€€€€€€€€€€€€€€€€|€€€€*Ì------------||-------:--------------------------------|----------ÌÔ'ìäî7ÔÔ€Xk– X ììcyÔÔ€X÷nX XXk–ÔÌà  àThis€approach€derives€the€Maximality€effect€(Itð=ð€1989,€reanalyzing€Selkirk's€1981€principleÏof€syllable€number€minimization)€as€a€theorem:€the€more€syllables,€the€more€Fill€violations,€and€theÏmore€alignment€violations.€Hence€Itð=ð's€directionality-driven€convergence€effect€between€L„R€and€R„LÏsystems€in€òòccccóó€clusters€follows€directly,€without€additional€legislation.€The€redundant€overlapÐ Üø  Ðbetween€Fill€and€Align€should€be€noted;€on€the€positive€side,€it€should€also€be€noted€that€MaximalityÏeffects€can€be€obtained,€in€this€approach,€ò òindependentó ó€of€the€domination€status€of€Fill€(i.e.,Ð ´Ð  Ðindependent€of€epenthesis),€opening€up€new€possibilities€of€analysis€which€are€perhaps€neededÏelsewhere.€Ìà  àPerhaps€a€more€disturbing€finding€is€the€fact€that,€in€the€current€version€of€GeneralizedÏAlignment€Theory,€the€L„R€and€R„L€patterns€can€also€be€derived€under€an€ò òOpposite€Edgeó ó€settingÐ h„ Ðof€the€alignment€constraints,€i.e.€as€Align(Syll,Edge,PrWd,Edge'),€as€the€reader€may€verify€forÏher/him-self€by€recasting€the€tableaux€given€above€in€the€appropriate€way.€And€they€can€be€derivedÏby€counting€segments€instead€of€counting€moras.€They€can€even€be€derived€with€a€strictly€segmentalÏsetting,€requiring€each€ò òsegment€ó óto€be€as€closely€aligned€to€one€edge€as€possibleð!ð€this€is€sufficient€toÐ 8 Ðpush€the€epenthetic€vowel€as€far€in€the€opposite€direction€as€possible,€with€multiple€epenthesisÏdiscouraged€as€before.€Aside€from€the€question€whether€such€possibilities€are€required€or€desirable€Ì(e.g.€could€segmental€count€be€necessary€to€derive€directionality€in€a€language€without€weightÏdistinctions,€where€there€is€no€coda€mora€to€count?),€we€can€conclude€that€the€theory€of€GA€needsÏto€be€tightened€up.€The€task€is€to€isolate,€among€the€set€of€statements€of€the€formÏ'Align(x,Edge(i),y,Edge(j))'€made€available€by€the€general€Alignment€scheme,€those€that€have€trueÏlinguistic€significance.€Ìà  àIn€conclusion,€we€return€to€the€beginning:€what€we€have€tried€to€demonstrate€here€is€theÏexistence€of€a€method€for€capturing€directional€syllabification€effects€in€Optimality€ÌTheory.€It€seems€quite€clear€that€our€proposal€is€only€an€initial€step,€and€we€are€convinced€that€it€canÏbe€improved€on€in€many€ways.€Our€demonstration€and€argumentation€has€remained€abstract€andÏschematic€throughout,€using€the€contrasting€Arabic€cases€merely€as€convenient€demonstration€objectsÏfor€our€method:€our€goal€was€to€replicate€directionality€effects€as€they€were€argued€to€exist€in€earlierÏliterature€(Itð=ð€1986,€1989).€Some€recent€research€has€indicated€that,€taken€seriously€as€purportedÏinstances€of€directional€syllabification€patterns,€Arabic€dialects€present€a€host€of€further€issues€thatÏwould€have€to€be€dealt€with.€Both€John€McCarthy€and€Ellen€Broselow€have€pointed€out€that€all€theÏproposals€presented€above€fail€to€adequately€deal€with€the€way€in€which€ò òinitialó ó€òò#ccóó€clusters€areÐ Ì*è$+ Ðresolved€in€the€various€dialects€of€Arabic€(see€Broselow€1992).€Since€we€are€not€familiar€with€theÏempirical€details,€we€refrain€from€fabricating€hypotheses,€and€leave€the€issue€unresolved.€In€fact,€itÏis€not€unthinkable€that€the€ultimate€analysis€of€Cairene€and€Iraqi€Arabic€cases€does€not€even€involveÏdirectionality,€but€rather€other€factors€(as€argued€in€Broselow€1992).€Ð €.œ(/ Ðà  àWe€note€that€ò òsomeó ó€cases€of€special€behavior€in€PrWd-initial€position,€in€non-Arabic€languages,Ð ø Ðhave€a€straightforward€account€in€our€scenario.€Thus€Itð=ð€(1989,€252-254)€analyzes€Temiar€epenthesisÏas€basically€R„L€(Coda-preference;€in€our€terms,€ALIGN-L),€in€spite€of€the€fact€that€ò òinitialó ó€òò#ccvóóÐ ÔÜ Ðconfigurations€are€resolved€as€òò#.cV.cv...óó€and€not€aòòs€#.Vc.cóóv...€This€follows€directly€in€our€approach:Ð ÄÌ Ðin€òò#.cV.cv...óó,€the€second€syllable€is€separated€from€the€initial€word€edge€by€one€mora;€in€òò#.Vc.cv...óó,Ð °¸ Ðthe€mora€count€is€two.€Again€we€find€some€explanatory€redundancy,€however:€Onset€is€a€dominatingÏconstraint€in€Temiar,€as€pointed€out€by€Itð=ð€(1989),€whose€gradiently€fulfillable€principle€'AvoidÏOnsetless€Syllables'€is€among€the€precursors€of€current€Optimality-Theoretic€reasoning.ÌÌò òReferencesó óÐ L T  ÐÌñ1ñà0  àà ° àñ1ñBroe,€Michael.€1993.€Specification€Theory.€The€Treatment€of€Redundancy€in€Generative€Phonology.ÏDoctoral€Dissertation,€University€of€Edinburgh,€United€Kingdom.Ð (#(# Ðñ0ñà0  àà ° àñ0ñBroselow,€Ellen.€1980.€Syllable€Structure€in€Two€Arabic€Dialects.€òòStudies€in€the€Linguistic€SciencesóóÐ ü  Ð10:€13„24.Ð (#(# Ðñ/ñà0  àà ° àñ/ñBroselow,€Ellen.€1982.€On€Predicting€the€Interaction€of€Stress€and€Epenthesis.€òòGlossaóó.€Simon€FraserÐ ÔÜ  ÐUniversity,€Burnaby,€B.C.:€115„132.Ð (#(# Ðà0  àà ° àBroselow,€Ellen.€1992.€Parametric€Variation€in€Arabic€Dialect€Phonology.€In€Ellen€Broselow,ÏMushira€Eid,€and€John€McCarthy,€eds.,€òòPerspectives€on€Arabic€Linguistics€IV€(Papers€fromÐ ˜  Ðthe€Fourth€Annual€Symposium€on€Arabic€Linguistics)óó.€John€Benjamins:€Amsterdam.€Pp.€7„Ð „Œ Ð45.Ð (#(# ÐGoldsmith,€John.€1990.€òòAutosegmental€and€Metrical€Phonologyóó.€Blackwell:€Cambridge,€MA.Ð \d Ðà0  àà ° àGoldsmith,€John.€1991.€Harmonic€Phonology.€In€John€Goldsmith,€ed.,€òòThe€Last€Phonological€Ruleóó.Ð HP ÐUniversity€of€Chicago€Press:€Chicago.€Pp.€21„60.Ð (#(# Ðà0  àà ° àGrimshaw,€Jane.€1993.€Minimal€Projection,€Heads€and€Inversion.€Ms.,€Rutgers€University,€NewÏBrunswick.Ð (#(# ÐHayes,€Bruce.€1991.€òòMetrical€Stress€Theory:€Principles€and€Case€Studiesóó.€Ms.,€UCLA.Ð ø Ðà0  àà ° àItð=ð,€Junko.€1986.€òòSyllable€Theory€in€Prosodic€Phonologyóó.€Ph.D.€Dissertation,€University€ofÐ äì ÐMassachusetts,€Amherst.€[Published€by€Garland,€New€York,€1988].Ð (#(# ÐItð=ð,€Junko.€1989.€A€Prosodic€Theory€of€Epenthesis.€òòNLLTóó€7:€217„259.Ð ¼Ä Ðà0  àà ° àItð=ð,€Junko,€Mester,€Armin€and€Jaye€Padgett.€1993.€Licensing€and€Redundancy:€UnderspecificationÏin€Optimality€Theory.€Report€no.€LRC„93„07,€Linguistics€Research€Center,€University€ofÏCalifornia,€Santa€Cruz.Ð (#(# Ðà0  àà ° àItð=ð,€Junko,€Mester,€Armin€and€Jaye€Padgett.€To€appear.€Licensing€and€Redundancy.€òòLinguisticÐ l#t! ÐInquiryóó.ÐX$`"(#(# Ðà0  àà ° àMcCarthy,€John€J.€&€Alan€Prince.€1993a.€òòProsodic€Morphology€I:€Constraint€Interaction€andÐ D%L # ÐSatisfactionóó.€Ms.,€University€of€Massachusetts,€Amherst€and€Rutgers€University,€NewÐ 0&8!$ ÐBrunswick.€[Forthcoming€MIT€Press].Ð (#(# Ðà0  àà ° àMcCarthy,€John€J.€and€Alan€Prince.€1993b.€Generalized€Alignment.€Ms.,€University€ofÏMassachusetts,€Amherst,€and€Rutgers€University,€New€Brunswick.Ð (#(# Ðà0  àà ° àMcCarthy,€John€J.€and€Alan€Prince.€1993c.€The€Emergence€of€the€Unmarked.€Optimality€in€ProsodicÏMorphology.€Ms.,€University€of€Massachusetts,€Amherst,€and€Rutgers€University,€NewÏBrunswick.€[To€appear€in€proceedings€of€NELS€1993].Ð (#(# Ðà0  àà ° àPrince,€Alan€and€Paul€Smolensky.€1993.€òòOptimality€Theory.€Constraint€Interaction€in€GenerativeÐ ¤,¬'+ ÐGrammaróó.€Ms.,€Rutgers€University,€New€Brunswick,€and€the€University€of€Colorado,Ð -˜(, ÐBoulder.€[Forthcoming€MIT€Press].Ð|.„)-(#(# Ðà0  àà ° àScobbie,€James.€1991.€òòAttribute€Value€Phonologyóó.€Ph.D.€Dissertation,€University€of€Edinburgh,Ð ä ÐUnited€Kingdom.Ð (#(# Ðà0  àà ° àScobbie,€James.€1993.€Constraint€Violation€and€Conflict€from€the€Perspective€of€DeclarativeÏPhonology.€òòCanadian€Journal€of€Linguisticsóó€32.2:€155-167.ШÄ(#(# Ðà0  àà ° àSelkirk,€Elisabeth€O.€1981.€Epenthesis€and€Degenerate€Syllables€in€Cairene€Arabic.€In€Hagit€BorerÏand€Yosef€Aoun,€eds.,€òòTheoretical€Issues€in€the€Grammar€of€Semitic€Languagesóó.€MITÐ € œ ÐWorking€Papers€in€Linguistics,€Vol.3:€209„232.Ð (#(# Ðà0  àà ° àZoll,€Cheryl.€1993.€Directionless€Syllabification€and€Ghosts€in€Yawelmani.€Ms.,€University€ofÏCalifornia,€Berkeley.€[Transcript€of€talk€presented€at€the€first€Rutgers€Optimality€Workshop,ÏOctober€23,€1993,€Rutgers€University].