EGGSchool,BanjaLukaJuly/August2018

An Introduction to Historical Phonology 1

PatrickHoneybone

UniversityofEdinburghpatrick.honeybone@ed.ac.uk

Thecontentsofthissession

1.Historicalphonology–what’sitallabout...?

2.Howdoweknowphonologyhaschanged?

3.Whatisphonology,anyway?

4.Aretheredifferent‘types’and‘parts’ofphonologicalchanges?

5.Historicalphonologyandphonologicalhistory

Historicalphonology–what’sitallabout...?

Really,‘historicalphonology’involvesanythingthatcombines‘phonology’and‘thepast’

• thismayseemobvious,butthetwocanbecombinedinanumberofways...o itinvolvesbothsynchronicanddiachronicstudy(NB!)o itinvolvesbothgeneralhistoricalphonologyandlanguage-specificphonologicalhistory

• wewanttounderstandthedetailsofspecificchangeso andwewanttoreconstructpaststagesoflanguages’phonologies

• wewanttounderstandhowandwhyphonologycanchangeinprincipleo andwewanttoknowhowwecanreconstructpastsynchronicstagesoflanguages

[NB:reconstructingpaststagesoflanguageswasthestartofscientificlinguistics...]

Thesearebigquestions;theyrequireanumberofthings:

• evidenceofwhatthepastwaslike• knowledgeofwhatphonologyislike• anunderstandingofhownewthingscanbeinnovatedandintegratedintoaphonology

“Wewillfirstlyconsidersomethingofthebroadcontextthathistoricalphonologyexists

in–allthekindsofthingthatwewouldneedtounderstandinordertofigureoutboth

whatthephonologyofparticularlanguageswasatvariousstagesinthepast,andwhat

kindsofchangeshaveoccurredbetweensuchstages.”

We’llseeifwecananswersomeofthesequestions:

• whatdoesitmeantosaythatphonologyhaschanged?• howcanweknowthatphonologyhaschanged?• whatkindsofchangeshaveoccurredinlanguages?

• aretheredifferent‘types’ofphonologicalchange?• aretheredifferent‘parts’ofaphonologicalchange?• aretheredifferentmotivationsforphonologicalchange?

• aretherecharacteristicsthatphonologicalchanges(orparticulartypesofchanges)alwaysshow?

• canwedistinguishbetweenpossibleandimpossiblephonologicalchanges?

• whatcanphonologicaltheorysayabouthowchangesareintegratedintoorlostfromagrammar?

It’snotcompletelystraightforwardtodefinepreciselywhatwemeanby‘phonologicalchange’

• aretheseexamples?• phonologicalchangesaretypicallyshownusingthediachronic‘shaftlessarrow’>o thisisNOTthesameasthesynchronic‘shaftedarrow’®

[muːsi] > [maɪs] English ‘mice’

[pund] > [pfund] HighGerman ‘pound’

[ɡʷénh2-] > [ben] Irish ‘woman’

[keture] > [tʃwetiri] B/C/M/S~SerBo-Croat ‘four’

[koren] > [t ʃoːsen] English ‘chosen’

[hilpθ] > [hɛlps] English ‘helps’

NB:allofthesediachronicequationsaretrue,butmostofthemaremissingthepoint

• becausethechangesinvolveddidn’tjustaffectindividualwords

Here’soneattempttodefinephonologicalchange:

• Φ=aphonologicalentity• x¹y• P=person,population,place,phonology• T=time• i=thesame• >=indiachroniccorrespondence

NB:Φdoesnotjustrefertotranscriptionsofwords([muːsi],[maɪs])

• itcanalsobearealizationofasegmentand/oraphonological‘rule’o aphonologicalgeneralisation

NB:thediachronicarrow‘>’isambiguous:

• doesitrelatethedirectlypre-changeandpost-changestates?o orwereintermediatestages–weremorethanonequantainvolved?

• doesachangeneedtooccurwithinalanguageordialect?orcanitcreatenewones• ordoeschangeoccurwithinaspeaker?oragrammar?• somearguethat,strictlyspeaking,there’snosuchthingasphonologicalchangeWeneedtotalkaboutplaceaswellastime?Oneattempttodefine‘phonologicalchange’isasfollows Φx > Φy Pi,T1 Pi,T>1• Φ=aphonologicalformorgeneralisation• y¹x• P=place,person,phonology?population• T=time• i=thesame

changesaretypicallyshownusingthediachronic‘shaftlessarrow’‘>’• thisisNOTthesynchronicshaftedu®y

Joseph&Janda(2003)proposethatweshoulduse‘>...>’forcorrespondencesthat

involveseveralquantaandreserve‘>’todescribesingle-stepinnovations

• withtheseconventions,wecouldsaythatEnglishchangeinvolves:

o muːsi>...>maɪs

Thisisbecauseweasurethataseriesofchangesareinvolvedinthiscorrespondence:

muːsi > myːs > miːs > maɪs

NB:thediachronicarrow>hasafurtherproblem:

• itdoesnotbydifferentiatebetweensegmentsofdifferentphonologicalstatuses

o whataboutthedistinctionbetweenunderlying(‘phonemic’,contrastive)phonologyandsurface(‘allophonic’,predictable)phonology?

o canchangeoccuratbothlevels?

Weneedknowledgeofwhatphonologyislike

Thisisanintrocourse,butI’massumingyouknowsomethingaboutsymbols...

• ð,β,ŋ,ʃ,ʌ,ʊ,ə,æ,ø

...andthatyouknowsomethingaboutfeatures...

10

Elements in English vowels Harris (1994, 115) uses the GP set of elements to come up with the following set of representations for English lax vowels; he uses a quite ‘American’ set of symbols for these, and makes certain analytical assumptions that differ from those of Giegerich (1992), for example...

He also gives representations for tense vowels, accounting for a wide range of varieties, and assuming a very dialectal model of phonology (unlike Giegerich 1992’s partially panlectal approach) • read Harris (1994) discussion of segmental phonology for the full picture...

Elements in Consonants Harris (1994) also sets out a full set of representations for consonants. These in fact require 3 other elements which can straightforwardly occur in vowels (H, L, N), and 3 other elements (R, ?, h), but, importantly, do make some considerable use of the ‘vocalic’ elements in consonantal representations • more recent developments have sought to remove (R, ?, h, N) from the set of elements by reusing

the elements that can occur in vowels to represent all consonantal properties

The full set of elements in Harris (1994) - that is, in ‘classical’ Government Phonology (following Kaye, Lowenstamm & Vergnaud) is...

Element Gloss Independent manifestation A lowness a I palatality i/j U round/labiality u/w @ velarity ´ H stiff vocal cords high tone = ‘voiceless’ L slack vocal cords low tone = ‘voiced’ R coronality R ? occlusion / h noise h N nasality ???

These are used to give such representations as the following for English (where headedness is represented by underlining only where necessary)...

p b t k g s T n m R h /

x x x x x x x x x x x x | | | | | | | | | | | | U U R @ @ R R R U R h ?

| | | | | | | | | ? ? ? ? ?

h h ? ?

| | | | | | |

h h h h h N N | | | H H H

...andthatyouknowsomethingaboutphonologicalprocesses...

• itisstandardinphonologytodistinguishatleasttwomainlevelsofrepresentationinsynchronicphonologicalknowledge

So...formanyaccentsofEnglish(includingRP)thisdataisrepresentative:

[l] [ɫ]

light pillblame kiltpillar pulpitfeeling feel NB!

Inaccentslikethis

• [l]occursinanonset� NB:[l]and[ɫ]donotcontrast

• [ɫ]occursinarhyme� Arulecandescribeallthis:

l-velarisation=

13

4.TakingphonologicaltheoryseriouslyinunderstandingphonologicalchangeTheaboveassumesasimplistic(concrete)modelofphonology...• ifwetakephonologyseriouslywhenweconsiderhistoricalphonologyarangeofother

optionsopenup,extendingthewaysinwhichwecanconsidertheinteractionofchangesandofthephonologicalobjectsinvolvedinchangesbeyondthosethatwere/areconsideredintraditionalhistoricalphonology

Itisstandardinphonologytodistinguishbetweentwomainlevelsofrepresentationinthedescriptionofsynchronicphonologicalknowledge;sometheoriesargueformorethantwolevels,too:

‘phonemic’=underlying=lexicalrepresentation=UR

‘phonetic‘=surface=derivedrepresentation=SR

OnetraditionalwaytomaptheURontotheSRingenerativephonologywas(andformanyphonologistsstillis)usingphonologicalrules,andtheunitsthatexistatphonologicallevelsarethoughttobemadeupofcomplexrepresentations(usingfeatures,syllabicconstituentsetc)• themappingfromtheURtotheSRisknownasaderivation• therehasbeenconsiderabledisagreementastohowdifferenttheURandSRcanbedueto

aderivation:howabstractisphonology?• currentworkinOptimalityTheoryworkswithadifferentwayofmappingURtoSR,but

thetwolevelsremain,ifinasomewhatreinterpretedway

Oneproblemwiththewaythatthe‘>’conventionisoftenusedinhistoricalphonologyisthatitdoesnotbyitselfdifferentiatebetweensegmentsofdifferentphonologicalstatuses• ifthere’sadistinctionbetweenanunderlying(‘phonemic’)andsurface(‘allophonic’)level,

weneedtoconsiderwhetherchangecanoccurateitherlevel;anditcan...

Underlyingchangeandsurfacechange• thedevelopmentofi-umlautinthehistoryofEnglishisrelevanthere:o originally,therewerefrontroundedvowels–ifwethinkaboutitinthelightofphonological

theory,we’dsaythattheFRVsarederivedbyrule,o eg,u®y/__(C)i,j(thelengthofthevowelsinvolvedisirrelevant)

Wheni-umlautwasfirstinnovated,thedistributionof[u]and[y]waspredictable• [y]occurredwhenan/i/or/j/followed;[u]occurredelsewhere• thephonesarephonetically/featurallysimilaro thereiseveryreasontoassumethati-umlautwasoriginallyinnovatedasaphonologicalrule

(acaseof‘allophony’)• however,oncethe/i,j/werelost(duetoseparatechanges),therulesceasedtobe

synchronicallyactive,duetoareanalysiswhichcreatedofnewunderlyingsegments,like/y/o therewerethusnewcontrastso eg,/u:y/and/uː:yː/ ‘mouse’ ‘mice’(i) PGmc /muːs/®[muːs] /muːs+iz/®[muːsiz](ii) introductionofumlaut(+lossof-z) /muːs/®[muːs] /muːs+i/®[myːsi](iii) lossof-i+reanalysis=contrast! /muːs/®[muːs] /myːs/®[myːs]

• stage(ii)=theintroductionoftheumlautrule(=surfacechange)• stage(iii)issometimesknownasphonologisationo itisbetter,though,tocallitphonemicisationbecausephonologisationimpliesthatsomething

comeunderphonologicalcontrol,andthatdoesnotneedtoinvolvethe‘phonemic’orunderlyinglevel–isn’tstage(ii)phonological,too?

• inthestressedvowelinwordslikemice,thereisunderlyingchangewithoutsurfacechangeo thiscouldbedescribedas/uː/>/yː/o thisintroducesa‘phonemicsplit’intothelanguage

R /l/ ® [ɫ] __

Butthere’smore...

Thefollowingdataisrepresentativeofmyaccent,atleast:

feet [fiːt] feel [fiəɫ] feeling [fiːlɪŋ]

deep [diːp] deal [fiəɫ] helix [hiːlɪks]

seem [siːm] sealed [siəɫd] sealant [siːlənt]

Inaccentslikethis

• [iː]occursbefore[l]

• [iə]occursbefore[ɫ]

Thisissometimesdescribedas‘HighVowelBreaking’or‘schwainsertion’

• HVBcanbeformalisedasaphonologicalrule:

iː®iə/__ɫ

Thereisacrucialinteractionbetweenthetworules:

feet feel feeling UR /fiːt/ /fiːl/ /fiːl+ɪŋ/

syllabification .fiːt. .fiːl. .fiː.lɪŋ.

l-velarisation — fiːɫ —

HVB — fɪəɫ —

SR [fiːt] [fɪəɫ] [fiːlɪŋ]

InRule-BasedPhonology,onerulecanapplytotheoutputofanotherrule

• thisshowsruleordering:itisacaseoffeedingordero infeeding,rule1createsanenvironmentwhichallowsrule2tooccur:LVfeedsHVB

ThemappingfromtheURtotheSRisknownasaderivation• therehasbeenconsiderabledisagreementastohowdifferenttheURandSRcanbeduetoaderivation:howabstractisphonology?

• currentworkinOptimalityTheoryworkswithadifferentwayofmappingURtoSR,butthetwolevelsremain,ifinasomewhatreinterpretedway

• currentworkinrepresentationalmodels,suchasGovernmentPhonology,typicallyassumestwolevels,whetherthisismadeexplicitornot

Ruleorderingcangetmorefun...

Whataboutthesetranscriptions,whichrepresentsomevarietiesofAmEng:

set sent [sɛt] [sɛt]

cat can’t

[kʰat] [kʰat]

• arethereunderlyinglynasalvowelsinEnglish?• wedonotneedtosaythisifweallowforruleordering(orsomeanalogousmechanism)

Wesimplyneedtoassumetwoorderedrules:

set sent UR /sɛt/ /sɛnt/

V®VÑ /__[nasal]. — sɛnt

n®Ø/__C. — sɛt

SR [sɛt] [sɛt]

Thisinvolvesopacity

• ‘counter-bleeding’=alaterruleremovesthecontextthatallowsanearlierruletoapply

Whyisallthisrelevanthere...?

Thiswasanattempttodefinephonologicalchange:

NB:Φdoesnotjustrefertotranscriptionsofwords([muːsi],[maɪs])

• itcanalsobearealizationofasegmentand/oraphonological‘rule’o aphonologicalgeneralisation

So...

• Φcanbearule?• Φcanbearuleordering?• ΦcanbeanUR-SRmapping?

• doesachangeneedtooccurwithinalanguageordialect?orcanitcreatenewones• ordoeschangeoccurwithinaspeaker?oragrammar?• somearguethat,strictlyspeaking,there’snosuchthingasphonologicalchangeWeneedtotalkaboutplaceaswellastime?Oneattempttodefine‘phonologicalchange’isasfollows Φx > Φy Pi,T1 Pi,T>1• Φ=aphonologicalformorgeneralisation• y¹x• P=place,person,phonology?population• T=time• i=thesame

changesaretypicallyshownusingthediachronic‘shaftlessarrow’‘>’• thisisNOTthesynchronicshaftedu®y

Let’sgetdowntobusiness–let’sconsidersomedatarelevanttohistoricalphonology...

thereisaregularsegmentalcorrespondencebetween(i)northernaccentsinEngland

(suchasPresent-DaytraditionaldialectYorkshire)and(ii)southernaccents(suchas

Present-DayRP,whichhassouthernroots)intermsoftheirlaxvowelphonology:

o everyoccurrenceof[ʌ]inPDRPcorrespondsto[ʊ]inPDYorkshire,asinthefollowingsetofwords:

PDYorkshire PDRPlung [lʊŋ] [lʌŋ]

blush [blʊʃ] [blʌʃ]

cup [khʊp] [khʌp]

gulf [ɡʊɫf] [ɡʌɫf]

love [lʊv] [lʌv]

• thisisn’talwaysthecase,however:someoccurrencesofYorkshire[ʊ]correspondtoRP[ʊ],asinthefollowingsecondsetofwords:

PDYorkshire PDRPbush [bʊʃ] [bʊʃ]

put [phʊt] [phʊt]

full [fʊɫ] [fʊɫ]

wolf [wʊɫf] [wʊɫf]

pull [phʊɫ] [phʊɫ]

Whyistherethissituation?Weknow

thattheselinguisticsystemsareclosely

relatedbecauseeverythingelseaboutall

thesewordsisthesameinbothvarieties.

Tosaythatthesetwolectsarecloselyrelatedmeansthattheydivergednotallthat

longago–theywerethesamerelativelyrecently

• astheyarenowdifferentinvowelphonologysomekindofchangemusthaveoccurred• elsewhywouldtherebethesamecorrespondenceinthefirstwholesetofwords?

Thedatainthetwosetsofwordsaboveisthusevidenceforhistoricalphonology

• someofthemostcrucialevidenceforpaststatesoflanguages–someofthebestevidenceforphonologicalchange–comesfromthecomparisonofcontemporarysystemswhicharerelated

o thisiscomparativeevidence

[Weneedevidenceofwhatthepastwaslike]

Whatkindofchangegaverisetothedata?

• fromthisdata,thereseemtobetwooptions

ʊ>ʌin(thehistoryofvarietieslike)RPorʌ>ʊin(thehistoryofvarietieslike)Yorkshire

Thischangeismoreinterestingthanthat,though:

• itinvolvesdifferentnumbersofcontrastso inPDRP,/ʊ:ʌ/–thisisawayofshowingthat[ʊ]contrastswith[ʌ](asinputvsputt)o inPDYorkshire,thereisnosuchcontrast:/ʌ/doesnotexist–everywordwitheither[ʌ]or[ʊ]inRPhas[ʊ]inYorkshire,soputandputtsoundexactlythesame

• thismeansthattherewaseitherasplitinonesetofvarieties(includingRP)...

ʊ>ʊ:ʌ =acontrastdeveloped

• ...oramergerintheothersetofvarieties(includingYorkshire)

ʊ:ʌ>ʊ =acontrastwaslost

2

Tosaythatthesetwolectsarecloselyrelatedmeansthattheydivergednotallthatlongago–theywerethesamerelativelyrecently• astheyarenowdifferentintheirvowelphonologyachangemusthaveoccurred• elsewhywouldtherebethesamecorrespondenceinthefirstwholesetofwords?

Thedatainthetwosetsofwordsaboveisthusevidenceforhistoricalphonology• someofthemostcrucialevidenceforpaststatesoflanguages–someofthebestevidencefor

phonologicalchange–comesfromthecomparisonofcontemporarysystemswhicharerelatedo thisiscomparativeevidence

Whatkindofchangegaverisetothedata• fromthisdata,thereseemtobetwooptionso changesaretypicallyshownusingthediachronic‘shaftlessarrow’

ʊ>ʌin(thehistoryof)RP(andrelateddialects)orʌ>ʊin(thehistoryof)Yorkshire(andrelateddialects)

• thischangeismoreinterestingthanthat,though:o itinvolvesdifferentnumbersofcontrastso inPDRP,/ʊ:ʌ/–thisisawayofshowingthat[ʊ]contrastswith[ʌ](asinputvsputt)o inPDYorkshire,thereisnosuchcontrast:/ʌ/doesnotexist–thismeansthateveryword

witheither[ʌ]or[ʊ]inRPhas[ʊ]inYorkshire,soputandputtsoundexactlythesame

• thismeansthattherewaseitherasplitinonesetofvarieties(includingRP)...

ʊ ʊ>ʊ:ʌ =acontrastdevelopedʊ ʌ

• ...oramergerintheothersetofvarieties(includingYorkshire)

ʊ ʊ:ʌ>ʊ =acontrastwaslost ʊ ʌ

WehaveawiderangeofevidenceforEnglishwhichputsbeyonddoubtwhichchangeactuallyoccurred• wecancomparethedatawithlinguisticsystemsthataremoredistantlyrelatedo eg,lunghasthefollowingcognateinPDGerman:Lunge/lʊŋə/o eg,bushhasthefollowingcognateinPDGerman:Busch/bʊʃ/o wordsinbothsetshave/ʊ/inmoredistantlyrelatedsystemssoitseemslikelythatthe

olderEnglishstate(fromwhichbothPDRPandPDYorkshirearederived)hadonly/ʊ/

Theothermainsourceofevidenceforpastphonologicalstatesinmanylanguages(includingEnglish)comesfromwriting• inearlystagesofmoredistantlyrelatedlanguages,bothsetsofwordsarespeltwith‹u›o OldFrisianlungeno OldNorsebuskr• thisindicatesthattheylikelyhadabackhighroundedvowel(like[ʊ],butnotlike[ʌ])

Englishitselfhaswrittenrecordswhichgobackforaroundamillenniumandahalf• inearlierstagesofEnglish,bothsetsofwordsarealsospeltwith‹u›o OldEnglishlungen‘lung’o MiddleEnglishputhe‘put’

• indeed,thefactthatwenowspellmanywordsinbothsetswith‹u›inPDEnglishspellingisevidenceforourpurposes,too,becauseinthiscasethespellingsystemreflectsearlierstagesofthephonologyofalanguage(becausethephonologyhaschangedbutthespellinghasnot)

Awiderangeofevidenceshowsthatthischangewasasplitin(thehistoryof)varietieslikeRPo itisoftencalledthe‘FOOT/STRUTsplit’inworkonEnglish

2

Tosaythatthesetwolectsarecloselyrelatedmeansthattheydivergednotallthatlongago–theywerethesamerelativelyrecently• astheyarenowdifferentintheirvowelphonologyachangemusthaveoccurred• elsewhywouldtherebethesamecorrespondenceinthefirstwholesetofwords?

Thedatainthetwosetsofwordsaboveisthusevidenceforhistoricalphonology• someofthemostcrucialevidenceforpaststatesoflanguages–someofthebestevidencefor

phonologicalchange–comesfromthecomparisonofcontemporarysystemswhicharerelatedo thisiscomparativeevidence

Whatkindofchangegaverisetothedata• fromthisdata,thereseemtobetwooptionso changesaretypicallyshownusingthediachronic‘shaftlessarrow’

ʊ>ʌin(thehistoryof)RP(andrelateddialects)orʌ>ʊin(thehistoryof)Yorkshire(andrelateddialects)

• thischangeismoreinterestingthanthat,though:o itinvolvesdifferentnumbersofcontrastso inPDRP,/ʊ:ʌ/–thisisawayofshowingthat[ʊ]contrastswith[ʌ](asinputvsputt)o inPDYorkshire,thereisnosuchcontrast:/ʌ/doesnotexist–thismeansthateveryword

witheither[ʌ]or[ʊ]inRPhas[ʊ]inYorkshire,soputandputtsoundexactlythesame

• thismeansthattherewaseitherasplitinonesetofvarieties(includingRP)...

ʊ ʊ>ʊ:ʌ =acontrastdevelopedʊ ʌ

• ...oramergerintheothersetofvarieties(includingYorkshire)

ʊ ʊ:ʌ>ʊ =acontrastwaslost ʊ ʌ

WehaveawiderangeofevidenceforEnglishwhichputsbeyonddoubtwhichchangeactuallyoccurred• wecancomparethedatawithlinguisticsystemsthataremoredistantlyrelatedo eg,lunghasthefollowingcognateinPDGerman:Lunge/lʊŋə/o eg,bushhasthefollowingcognateinPDGerman:Busch/bʊʃ/o wordsinbothsetshave/ʊ/inmoredistantlyrelatedsystemssoitseemslikelythatthe

olderEnglishstate(fromwhichbothPDRPandPDYorkshirearederived)hadonly/ʊ/

Theothermainsourceofevidenceforpastphonologicalstatesinmanylanguages(includingEnglish)comesfromwriting• inearlystagesofmoredistantlyrelatedlanguages,bothsetsofwordsarespeltwith‹u›o OldFrisianlungeno OldNorsebuskr• thisindicatesthattheylikelyhadabackhighroundedvowel(like[ʊ],butnotlike[ʌ])

Englishitselfhaswrittenrecordswhichgobackforaroundamillenniumandahalf• inearlierstagesofEnglish,bothsetsofwordsarealsospeltwith‹u›o OldEnglishlungen‘lung’o MiddleEnglishputhe‘put’

• indeed,thefactthatwenowspellmanywordsinbothsetswith‹u›inPDEnglishspellingisevidenceforourpurposes,too,becauseinthiscasethespellingsystemreflectsearlierstagesofthephonologyofalanguage(becausethephonologyhaschangedbutthespellinghasnot)

Awiderangeofevidenceshowsthatthischangewasasplitin(thehistoryof)varietieslikeRPo itisoftencalledthe‘FOOT/STRUTsplit’inworkonEnglish

Wehavearangeofevidencewhichputsbeyonddoubtwhichchangeactuallyoccurred

• wecancomparethedatawithlinguisticsystemsthataremoredistantlyrelated

o eg,lung(/ʊ:ʌ/)hasthefollowingcognateinPDGerman: /lʊŋə/Lunge

o eg,bush(/ʊ/)hasthefollowingcognateinPDGerman: /bʊʃ/Busch

o wordsinbothsetshave/ʊ/inmoredistantlyrelatedsystemssoitseemslikelythattheolderEnglishstate(fromwhichbothPDRPandPDYorkshirederive)hadonly/ʊ/

Anotherkeysourceofevidenceforpastphonologicalstatesinmanylanguagescomes

fromwriting,especiallyisalanguageiswrittenusinganalphabet

• inearlystagesofmoredistantlyrelatedlanguages,bothsetsofwordsarespeltwith‹u›

o lung(/ʊ:ʌ/) OldFrisian lungen

o bush(/ʊ/) OldNorse buskr

• thisindicatesthattheylikelyhadabackhighroundedvowel(like[ʊ],butnotlike[ʌ])

• iftheyhaddifferentvowels,wewouldexpectthistoberepresentedinearlierwritingusingdifferentletters

o thebasicalphabeticprincipleisprettymuchthephonemicprinciple

o whenalanguageisfirstwritteninanalphabet,atypeofphonemicanalysisisinvolved

Englishitselfhaswrittenrecordswhichgobackforaroundamillenniumandahalf

• inearlierstagesofEnglish,bothsetsofwordsarealsospeltwith‹u›

o lung(/ʊ:ʌ/) OldEnglish lungen

o bush(/ʊ/) MiddleEnglish busche

• indeed,thefactthatmanywordsinbothsetsarespeltwith‹u›inPDEnglishspellingisevidence,too,becauseinthiscasethespellingsystemreflectsanearlierstageof

thephonologyofalanguage(thephonologyhaschangedbutspellinghasnot)

Awiderangeofevidenceshowsthatthiswasasegmentalsplitin(thehistoryof)

varietieslikeRP

• itisoftencalledthe‘FOOT/STRUTsplit’inworkonEnglish

o theFOOTvowel =thevowelthatadialecthasinthewordfootandothers

o theSTRUTvowel =thevowelthatadialecthasinthewordstrutandothers

Thisstillleavesalotofquestionsaboutthischange,forexample:whendidithappen?

AlthoughEnglishspellingdoesnotreflectthischange,writtenrecordscanstillhelp

withourdatingofit–here’soneway:

• WilliamShakespearelived1564-1616;hewrotetowardstheendofthe16thandstartofthe17thcentury,includinglineslikethefollowing;hewaswritingforaLondon

audience,soitislikelythatthisrepresentsthepronunciationofthatplaceandtime:

I have been closely shrouded in this bush, And marked you both, and for you both did blush.

Love’s Labour’s Lost iv, 3, 137-8

Scale of dragon, tooth of wolf Witches’ mummy, maw and gulf...

Macbeth iv, 2, 22-3

o theserhymesnowonlyworkinNorthernEnglishaccents;buttheyalsoworkedattheturnofthe17thcenturyinSouthernEnglishaccents(whichformedthebasisofPDRP)

o theserecordshelpustodatethechange–itmusthavehappenedafterShakespeare• or,toputitmoreprecisely:thisevidenceshowsthatpronunciationswith[ʊ]werestillnormal/unexceptionalinLondonspeechinbothsetsofwordsintheearly1600s

o thisisindirectwrittenevidence:inindirectwrittenevidence,wecaninterpretawrittensourceinsomeunintendedwaytogiveevidenceofthephonologythatitrepresented

Therecanalsobedirectwrittenevidencefromthepast,whichinvolvesexplicit

commentsonpronunciationfromearlyphonetician-phonologists,spellingreformers,

writersofdictionariesandlanguage-learningguidesetc

• thereisquitearangeofdirectevidenceforlanguageslikeEnglishfromsomeperiods

• Dobson(1968)hasinterpretedawiderangeofit,andsaysthat“Hodges,Wilkins,Coles,andCooperfailtodistinguishMEŭinfreepositionfromstressedandunstressed[ə]...thisisonlypossibleifMEŭis[ʌ]”(1968,586)

o theauthorsthatDobsonmentionsexplicitlydiscussedthesoundsinsetsofwords,andstatethatthevowelinwordslikelungandblushiscomparableto[ə],whichweknowisarticulatorilyverysimilarto[ʌ]

o HodgesistheearliestoftheauthorsthatDobsonmentions;hepublishedanumberofworksinthe1640sinwhichhe“...suggest[s]waysinwhichthetraditional

orthographymaybeimprovedandmademoreconsistent,and...invent[s]asystem

ofdiacriticsbywhichreadingandpronunciationmaybetaughtfromtheordinary

orthography”(Dobson1968,165)

Allthisevidencegivesusaremarkablypreciseindicationofthedatingofthechange:

• itmusthavebecome‘normal’inLondonspeechinthemid-decadesofthe17thcentury

Wehavethusbeguntoconsiderwhatkindsofevidencewecanusetodiscoverthe

subjectmatterofhistoricalphonology:

• comparativeevidencefromrelatedbutdistinctforms• evidencefromwrittenrecords

Wehavealsodonesomereconstructionofapastsynchronicphonologicalstate:

• 15thcenturyEnglishdidnothave/ʌ/(ie,itdidnothavethe/ʊ:ʌ/contrast)

Thisstillleavesuswithalotofquestionsthatwecouldaskaboutthechange,including

thefollowing:

• wheredidithappen?onlyinLondon...?• howdidithappen?howdoesasplitenterthephonologyofalanguage?• whatconditionedthesplit?whydiditoccurinsomewords(eg,blush)butnotothers(eg,bush)?

• isthisthekindofthingthatnormallyhappensinphonologicalchange?• whydidithappen?whatkindoffactorsleadtophonologicalchange?o toanswerthesekindsofquestions,wewillneedtoconsidersomefurtherfundamentalissuesinhistoricalphonology...

Innovationvspropagationofchange(the‘parts’ofachange)Indiscussingthechangementionedabove,wehaveaskedquestionsabouttwo

distinctaspectsofthechange

• theinnovationinvolved=thealterationinphonologicalsegmentsorstructureso =thestructuralaspectofthechangeo eg:isthisthekindofthingthatnormallyhappensinphonologicalchange?

• thepropagationinvolved=thewayinwhichaninnovationistakenupbyspeakerso =thesocialaspectofthechangeo eg:wheredidithappen?onlyinLondon...?

Innovationandpropagationareconceptuallyseparable,buttherearenoeasily

identifiablediachroniceventsunlessbothoccur

• aparticularinnovationmustbepossibleinaparticularphonologicalsystemandmustoccurinthespeechofoneorseveralspeakers,andmustthen‘catchon’and

bepropagatedthroughaspeechcommunitybyitsspeakers

• somechangespropagatethroughallspeakersofalanguage• somechangespropagatethroughthespeakersinacommunitywhichtakesuponlyonepartoftheareawherealanguageisspoken

o thismayinvolvethechangespreadingthroughonlyoneorseveraldialectsofalanguageo orsometimessucheventsarearguedtoleadtothecreationofnewdialects(orlanguages)

Anychangeinvolvesadiachronicdifferenceinsomeaspectoflinguisticstructure,and

forthatdifferencetobecomepartofthephonologicalhistoryofalanguage(andsoto

becomesubjectmatterforhistoricalphonology),‘social’propagationisessential

• the‘social’aspectofachange(broadlyconstrued)involvesboththesociologicalcategoriesthatspeakersfitintoandthegeographyofwheretheyexist

• wewillneedtoconsidersomeaspectsofthepropagationofchangeinthiscourse,inordertounderstandthefulldetailsofindividualchangesthathaveoccurredinlanguages

o however,wedoneedtofocus:theemphasisinthiscourseisonthestructuralaspectofchange–oninnovations

Tosaysomethingaboutpropagation,onewell-knownobservationisthat:whenachangedoesnotspreadthroughallareasofaspeechcommunity,anisoglossresults

• inchangesthatoccurredinthemediumordistantpast,wecantrackthegeographicaspectoftheirpropagationthroughthemethodsofdialectology

• theʊ>ʌchange(theFOOT/STRUTsplit)wasinnovatedinSouthernpartsofthe‘Englishspeechcommunity’,butdidnotpropagatetoNorthernparts;anisogloss

indicatestheextentofthepropagationofthechange

Thisisamapwithinformationfromthe

SurveyofEnglishDialects• lowerunbrokenline=FOOT/STRUTo (upperdashedline=BATH)

EachdotwasasamplinglocalityintheSED• thoseabovethelowerisoglosshadonly/ʊ/• thosebelowtheisoglosshadboth/ʊ/and/ʌ/

Thisflagsuptheinteractionbetweenhistorical

phonologyanddialectology

• toanextenttheyarestudyingthesamethings• dialectologyfocusesonpropagation

Ifanumberofisoglossescoincide(thiscanbe

duetopoliticalreasons),differentdialectsor

languagescanemerge.

Vowels in British Isles accents 7

Figure 3 Median and interquartile range for the 11 monophthongs produced by six male speakers of sse.

A complete description of each and every vowel for all 13 accents would indeed beextremely tedious. Instead, we will focus on a restricted set – varying from one dialect to thenext – of the most notable phenomena.

3 Results

3.1 Standard Southern English (sse)The accent sse serves as a reference against which the remaining accents are compared. Ithas been described in detail in many publications (Nolan 1998, Jones 2003, Upton 2004,Hawkins & Midgley 2005, Wells 2008, etc.). As figure 3 shows, the vowels of who’d andhood (which are supposed to instantiate the GOOSE and FOOT sets, respectively) are morefront than recent pronunciation dictionaries suggest (Jones 2003, Wells 2008). This finding ishowever in accordance with up-to-date acoustic–phonetic descriptions (Hawkins & Midgley2005, McDougall & Nolan 2007).

As far as closing diphthongs are concerned, figure 4 shows that the vowel of PRICE has arather back starting element while that of MOUTH is rather front, the difference being audiblyperceptible. In earlier descriptions, O’Connor (1973) and Gimson (1980) used the symbols/a/ and /ɑ/ for the starting element of PRICE and MOUTH, respectively. Nowadays, althoughJones (2003) and Wells (2008) use a single symbol for the first vowel of PRICE and MOUTH,the vowel plots provided in these dictionaries (Jones 2003: viii; Wells 2008: xxiii) explicitlyshow that the authors keep considering that the first element in MOUTH is slightly moreback than that of PRICE. In contrast to the latter, other recent publications have symbols forPRICE and MOUTH that are much more in line with our findings. Upton (2004) records twovariants for PRICE: [aI] in traditional RP and [ØI] in contemporary RP. He, however, notesjust one possibility for MOUTH: [aυ]. Olausson & Sangster (2006) also use [ØI] (PRICE) and[aυ] (MOUTH). In a review of vowel symbols in dictionaries, Windsor Lewis (2003: 147)comments on the use of [ØI] instead of [aI]: according to the author, ‘[t]he apparent Uptonsuggestion that /aI/ and /aυ/ have now reversed their relative starting positions in mainstreamusage is not supported by my observations and I know of no-one else of such an opinion’.Windsor Lewis goes so far as to claim that the symbols [aI] and [ɑυ] (for PRICE and MOUTH)

Inchangesthatarehappeningnow–thatis,inchangesinprogress–wecantrackthe

narrowlysocialaspectoftheirpropagationbyinvestigatingthevariationthatexists

(intermsoftherealisationofthephonologicalcategoriesinvolvedinthechange)in

differentsocialgroups,followingLabov’sideas(eg,Labov2001)

• theFOOT-STRUTsplitisn’tinprogress,solet’sconsideranotherchangewhichis:GOOSE-fronting(thatis,uː>ʉː),whichhasrecentlybeenreportedinanumberof

areas/dialectsofEnglish(eg,theSEofEngland,Birmingham,Nottingham,Manchester)

o thiscanbeseenintheplaceofwho’donthisF1-F2plotforRP-typespeakersfromLondon(takenfromFerragne&Pellegrino2010)

Inareaswherethischangeiscurrentlypropagating,suchasCarlisle,thereis

variationbetweensocialgroupsofspeakersintermsofwhethertheGOOSEvowelis

frontornot,asshowninthesetwoF1-F2plotsmadefromrecordingscollectedin

2007/2008(takenfromJansen2017)

• the60-year-oldmalehasabackrealisationoftheGOOSEvowel(around[u]),sodoesnothaveGOOSE-fronting–heisnottakingpartinthepropagationofthechange

• the23-year-oldfemalehasafront/centralrealisationoftheGOOSEvowel(around[ʉ]),sodoeshaveGOOSE-fronting–sheisnottakingpartinthepropagationofthechange

o [bothspeakersarejudgedtobemiddleclass]

Onthisevidence,thechangehaspropagatedtoCarlisle,buthasnot(yet?)

propagatedthroughoutthewholespeechcommunityinCarlisle

• itiscommoninsuchsituationstoconsiderthetypesofvariationinvolvedbetweengroupsasshowingevidenceforchangeinapparenttime

• ifolderspeakershaveonevariantandyoungerspeakersadifferentvariant,thiscanbeevidencethatchangeisinprogress,withthevariantusedbyyoungerspeakers

replacingthatusedbytheolderspeakers

o Jansen(2017)showsthisclearlytobethecaseinCarlisleGOOSE-fronting

0.5

1.0

1.5

2.0

Old Middle−aged YoungAge Group

Norm

alise

d F2

Wewillnotfocusinthiscourseonthesocialmechanismsthroughwhichachange

progressesthroughaspeechcommunityoronthegeographyofchanges(although

bothofthesearepartoftheoverallfieldofhistoricalphonology)

• muchofwhatwewilldiscusswillfocusoninnovation,askingquestionslike:

o whatkindsofinnovationarepossible?

o (aspartofthiswewillneedtoconsiderwhatkindsofinnovationhavehappened–thusfarweknowthatʊ>ʌ,uː>ʉː,andiː>aɪarepossible...)

o howcaninnovationsbeintegratedintoaphonologicalsystem?

Innovationscanbeendogenousorexogenous(themotivationsofchange)

Anotherfundamentalquestionis:wheredoinnovationscomefrom?

Wecanrecogniseafundamentaldistinctionbetweensourcesofinnovation;therecanbe:

• exogenouschange(‘internal’change)o theseareduetoinnovationsthatareimposedonaphonologicalsystemfromwithout

• endogenouschange(‘external’change)o theseareduetoinnovationsthatarearisewithinaphonologicalsystem

Exogenyinvolvessomesortoflinguisticcontact;however,notallcontactleadsto

change–languagescanborrowindividualwordswithoutanyrealchangeif:

• thewordsfitwiththephonologyoftheborrowinglanguageo eg,tax/taks/(fromtaxer)andsoufflé/sufle/fromFrench• orifawordisadaptedtotheborrowinglanguage’sphonologyo eg,muesli/mysli/>/mjuzli/(German),gherkin/xʏrkɪn/>/ɡɜ(r)kɪn/(Dutchgurkkijn)

However,languagescangainspecificfeaturesorevenwholesegmentsthrough

borrowing‘loanwords’–orothertypesofcontact

• eg,theEnglishvowel/ɔɪ/wasintroducedthroughborrowings(eg,choice,employ,join,buoy)

• eg,theexistenceoftoneinsomelanguagesisthoughttobeduetotheexistenceoftoneinneighbouring,butunrelatedlanguages

Endogenouschangedoesnotinvolvecontact;itcanbeduetosuchthingsas:

• system-internalpressures(eg,tomaintaincontrasts)• therealisationofpathwaysallowedbyconstraintsonphonologicalrepresentations• thephonologisationofphoneticbiases

Toreturntothemainchangeconsideredsofar,thereisnoevidencethatcontactwas

involvedintheʊ>ʌchange,soitmustbeakindofchangethatisendogenously

possibleinhistoricalphonology

• oneimportantquestionforhistoricalphonologyis:whatkindsofchangearepossible?

Muchofwhatwewilldiscussonthiscoursewillfocusonendogeny,becausethisis

wherephonologicalstructureislikelytoshowitsimpact

• it’simportanttobearexogenyinmind,however,inordertobesurethatachangeisendogenous

Adistinctionofapproach:historicalphonologyandphonologicalhistoryInordertofullyunderstandaspecificphonologicalchange,wewouldneedtoanswerawholerangeofquestions:

1.whendidithappen?

2.wheredidithappen?howfardiditspread?

3.whatwasitssocialpatterningasitpropagated?

4.howwasitinnovated:endogenouslyorexogenously?–whichfactorsgaveriseto

theinnovation?

5.whatwasitsprecisephonologicalpatterning?wasitconditionedphonologically?

6.howdiditaffectthephonologicalsystem?diditaffectthesetofcontrasts?

phonotactics?stress?diditinvolvechanginganalreadyexistingrule/processor

wasitentirelynew?

7.diditaffectonlyanumberofwords,ordiditaffecteverythingintherelevant

phonologicalenvironment?

8.inwhatwaydiditgetintothephonologicalsystem?howwasitphonologised?

9.isitacommontypeofchange?orisitsurprising?

10.couldithavepatterneddifferently?dochangesofthattypenormallypatterninthatway?

11.whichgroupofspeaker-listenersinnovatedit?children?adults?speakers?listeners?

12.howdoweknowthatthechangehappened?what’stheevidenceforthechange?

13.whydidithappenatpreciselythattimeandthatplace?

Therearedifferentkindsofquestionsinthelistabove,implyingdifferentkindsoffocus–itcanbehelpfulinthisconnectiontodistinguishbetweentwosubfieldsof

‘historicalphonology’

• phonologicalhistory=partofthestudyofspecificlanguages(sometimescalled‘philology’)o =questions1–7,withaninterestin8,andinteractingwith12

• generalhistoricalphonology=partofthestudyofgenerallinguisticso =questions9-11,withaninterestingeneralisingover3-8

• differentkindsofevidenceandargumentationarerelevanttodifferentquestions

Thesedifferencesoffocusare,however,intimatelylinked:

• inordertounderstandanyindividualchangeinthephonologicalhistoryofalanguage,weneedtounderstandhowlanguagescanchangeingeneral

• inordertounderstandhowlanguagescanchangeingeneral,weneedtoknowaboutlotsofindividualchangesinlotsoflanguages

• buttheapproachesareoftenpursueddistinctly:

o someworkfocusesonthephonologicalhistoryofx(x=English,French,Russian,Swahili,etc)

o otherworkisongeneralhistoricalphonology,aimingtoworkoutprinciplesofchange

9.isitacommontypeofchange?orisitsurprising?

• thisinvitesfurtherquestions:o whatare‘commonchanges’inthehistoryoflanguages?o whatkindsofthingsdoweexpecttofindinphonologicalchange?o arethere‘possible’changesand‘impossible’changes?–sometimesseenasthecrucialquestiono isthereasetofpossiblechangeswhichlanguagescaninnovate?• thesearetypologicalquestionso thestudyofdiachronictypologyisoneaspectofhistoricalphonologyo synchronicphonologicaltypologycanalsoactasacontrolonwhatweshouldreconstruct

13.whydidithappenatpreciselythattimeandthatplace?• thisis,infact,likelyunanswerable–canweeverhopetopredictwhenchangeswilloccur?o Weinreich,Labov&Herzog(1968)callthistheactuationproblem:o “Whatfactorscanaccountfortheactuationofchanges?Whydochangesinastructuralfeaturetakeplaceinaparticularlanguageatagiventime,butnotin

otherlanguageswiththesamefeature,orinthesamelanguageatothertimes?This

actuationproblemcanberegardedastheveryheartofthematter.”

o itisoftenarguedthatwecan’thopetobeabletopredictexactlywhenparticularchangeswilloccur

o rather,however,wecanhopetounderstandtheprinciplesthatgoverntheintroductionofchangeswhentheyoccur:whatcouldhappenandwhatcouldnot