Varro: Delingua Latina
Introduction,Text,Translation,andCommentary
VolumeI
Introduction,Text,andTranslation
EditedandTranslatedby WOLFGANGDAVIDCIRILODEMELO
3
GreatClarendonStreet,Oxford,OX26DP, UnitedKingdom
OxfordUniversityPressisadepartmentoftheUniversityofOxford. ItfurtherstheUniversity’sobjectiveofexcellenceinresearch,scholarship, andeducationbypublishingworldwide.Oxfordisaregisteredtrademarkof OxfordUniversityPressintheUKandincertainothercountries
©WolfgangdeMelo2019
Themoralrightsoftheauthorhavebeenasserted
FirstEditionpublishedin2019
Impression:1
Allrightsreserved.Nopartofthispublicationmaybereproduced,storedin aretrievalsystem,ortransmitted,inanyformorbyanymeans,withoutthe priorpermissioninwritingofOxfordUniversityPress,orasexpresslypermitted bylaw,bylicenceorundertermsagreedwiththeappropriatereprographics rightsorganization.Enquiriesconcerningreproductionoutsidethescopeofthe aboveshouldbesenttotheRightsDepartment,OxfordUniversityPress,atthe addressabove
Youmustnotcirculatethisworkinanyotherform andyoumustimposethissameconditiononanyacquirer
PublishedintheUnitedStatesofAmericabyOxfordUniversityPress 198MadisonAvenue,NewYork,NY10016,UnitedStatesofAmerica
BritishLibraryCataloguinginPublicationData Dataavailable
LibraryofCongressControlNumber:2018946419
ISBN978–0–19–882923–2(vol1)
ISBN978–0–19–882924–9(vol2) ISBN978–0–19–965973–9(set)
Printedandboundby CPIGroup(UK)Ltd,Croydon,CR04YY
LinkstothirdpartywebsitesareprovidedbyOxfordingoodfaithand forinformationonly.Oxforddisclaimsanyresponsibilityforthematerials containedinanythirdpartywebsitereferencedinthiswork.
Tomywifeandmydaughter
Preface
Onacool,foggyautumnmorninginEichstätttwenty-oneyearsago,Iwas sittinginalectureonLucretius.IcannotfullyrememberwhatProfessor TschiedelwassayingaboutLucretiusthatday,butassooften,therewasa learnedexcursusonotherauthors.Varrowasmentionedinpassing,andour professorremarkedthatwestudentsshouldbegratefulthatsomuchofhis workhadbeenlostbecauseourdegreesweretakinglongenoughasitwas.Iwas mildlyoutragedatthetime,butinretrospectIcanunderstandthesentiment. Varrowroteaboutseventy-fourworksin620books,andalthoughanancient bookcorrespondstoamodernbookchapterinlength,thisisstillanenormous output,especiallyforsomeonewhowasalsopoliticallyactiveandranafarm.
OfVarro’slinguisticoutput,the DelinguaLatina wasundoubtedlythemost importantwork.Thereweretwenty-fivevolumesoriginally,withaclearand straightforwardstructure:anintroductoryvolumewasfollowedbysixbooks onetymology,sixonmorphology,andtwelveonsyntax.Whatwenowhave isBooks5–10indirecttransmission,aswellasarelativelysmallnumberof fragmentsfromtheotherbooks.Inotherwords,wehavehalfoftheetymology, halfofthemorphology,andverylittleofthesyntax.The DelinguaLatina isnot onlyfragmentarybutalsohighlytechnical,andsoIamundernoillusionthat myworkwillenduponanundergraduatesyllabusanytimesoon.However, Ihavetriedtopresentthisworkinsuchawaythatitisnotjustgearedtowards thehandfulofVarrospecialistsworldwide,butsothatanyonewithaninterest inLatinorinlinguisticscanreadit.LikeVarro,Ihavedivideditintothree majorparts:acriticaleditionaccompaniedbyafairlyliteraltranslation;an introductionthatdealswithVarro’slifeandworks,hisapproachtolinguistics andourmodernlinguisticmethods,andVarro’ssometimesidiosyncraticstyle; and,finally,achapter-by-chaptercommentarythatiskeptasbriefaspossible. Noteverybodyisgoingtohavethetimeorinclinationtoreadmywork fromcovertocover.Forthoseforwhomthisistheirfirstintroductionto Romangrammarians,Irecommendstartingwiththeintroductionratherthan theeditionandcommentary.ThosewithmoreampleexperienceofRoman grammariansandtheirpeculiaritiesmayprefertodoittheotherwayround.
WedonotknowwhatpiquedVarro’sinterestingrammar.However,what iscleartoanyonewhohasploughedthroughtheendlesslyrepetitivepagesof Keil’s GrammaticiLatini isthatVarroisdifferent.VarrodoesnottreatLatin grammarasatoolthatcanhelpthebuddingoratororpoet;forVarro,Latin grammarisanendinitself,somethingthatisworthstudyingregardlessofits otherpotentialapplications.ThereisagreatdealofGreeklearninginVarro, butthereisalsoaverynoticeableprideinRomanculture.Theformeris
particularlyobviousinthemorphologicalpart,whilethelatterisespecially visibleintheetymologicalpart;here,weseeaveryRomanobsessionwithpork productsandsausages,withweaponsandtheRomanarmy,andwithRoman publicoffice.Varronianlinguisticsmayoftenbemorenaïveandlessadvanced thantheapproachesofhissuccessors,butwecanseeinhisworkadegreeofindependentthoughtandoriginalitythatissorelylackinginlatergrammarians.
IshouldatthispointstatethatIhaveapproachedVarro’sworkasalinguist, notasahistorianofculture.Spaceislimited,andmyworkisalreadylengthyas itis.IdoexplainRomangeographyandfeasts,butundoubtedlynotinasmuch detailassomereadersmighthope.Nocommentarycansatisfyeveryone,and Ihopethatsuchreaderswillalsoconsultoldercommentaries,whichareoften moreexplicitonthesepointswhileprovidinglesslinguisticmaterial.Bythe sametoken,Idonotconsideranampleorcompletebibliographyagoalin itself.MyreferencescontainonlythoseworksthatIhavecited.Whatisnot citedwasnotaccessibletome,orwasnotrelevantformypurposes.
IhavebeenworkingonVarroforalittlemorethanfiveyearsnow,and justasatthebeginning,Varrofillsmewithadmirationandexasperationin equalmeasure.ThereisindeedmuchtoadmireaboutVarro,whoattimes readslikeamodernlinguist—notthatthatisnecessarilyagoodthing.In themorphologicalpart,Varrotalksaboutthelearnabilityoflanguage;heisthe firsttomakethedistinctionbetweenderivationandinflection;andheisthe onlyRomangrammariantorealizethatthefutureperfectisnotasubjunctive. Theetymologyislessadvancedbyourstandards,buthere,too,therearemany valuableinsights,andalsomanyquotationsofearlytextsthatwouldbelostto usotherwise.Ontheotherhand,exasperationisalsoanormalreaction;the De linguaLatina wasclearlycomposedinhaste,andthisshowsnotonlyinVarro’s style,butalsointhecontents,wheremanyslipscouldhavebeenavoided.What ismore,Varrocanbequiterepetitive.However,readingVarroshouldatleast onsomelevelbeanedifyingundertaking,andIhopethatmyadmirationwill bemorevisiblethanmyoccasionalexasperation.
Myworkwouldhavetakenagreatdealmoretimehaditnotbeenfor agenerousgrantfromtheLeverhulmeFoundation,whichfreedmefrom undergraduateteachingandmanyofmyadministrativedutiesduringthe wholeof2017.ThisgavemeawonderfullyproductiveyearinwhichIwrote nearlyasmuchasintheprecedingfouryears.Thegrantalsoenabledmetogo toFlorenceandexaminethemostimportantVarromanuscriptinperson.
Fourscholarshaveinfluencedmegreatlyoverthelasttwodecades.In chronologicalorder,theyareFritzHeberlein,JohnPenney,AnnaMorpurgo Davies,andJimAdams.Jiminparticularhasgenerouslysupportedthiswork withadviceandhelpfromitsinceptiontoitscompletion.
IhavebenefitedgreatlyfromolderbibliographiessuchasCollart(1964) andCardauns(1982),butthebiggestbibliographicalhelpcamefromPhilipp Brandenburgintheformofhiswebsite(http://www.varro-grammaticus.de/).
Preface
Atanearlystage,JohnTrappes-LomaxreadtheentireLatintextandtranslationandprovidedmanycorrectionsandinsights.SarahMahmoodreadthe entiremorphologicaldiscussioninthestudytwiceandgaveclearandextensive feedbackthathashelpedmealot.Overthelasttwoyears,manypeoplegave expertfeedbackonthesmaller,individualetymologicalsections:JohnPenney helpedmewithTocharianandCeltic;MaryMcRobertandJanFellerer,with Slavonic;RobinMeyer,withArmenian;DonRinge,withGermanic;Kazuhiko Yoshida,withAnatolian;andElizabethTucker,withIndo-Iranian.Panagiotis FilosansweredmanyGreekqueries.
IowemuchtodiscussionswithJimBensonandLouiseMycock.Articlesthat werehardtocomebyweresenttomebyMicheleBianconi,RolandoFerri, AnnaZago,GregorBitto,RobertMaltby,andGiorgioPiras.Thestaffatthe SacklerLibraryandattheBibliotecaMediceaLaurenzianawerealwayshelpful andkind.
MyErasmuspartnershaveprovidedmewithexcellentdiscussions wheneverIvisitedEichstättandGhent.Inparticular,Iwouldliketothank FritzHeberlein,GregorBitto,BardoGauly,GernotMüller,MarkJanse,and GiovanbattistaGaldi.
Thisbookwouldnotbewhatitiswithoutthesurehandofmycopyeditor,MirandaBethell,whoseexpertguidancemadethelastmonthsbefore publicationeasyandstraightforward.
Myfamily,inparticularmywifeanddaughter,haveprobablyhadtohear moreaboutVarrothantheybargainedfor.Iamgratefulfortheirpatienceand supportandlove.
Andnow,Varro,myfriendandfoeoverhalfadecade,farewell!
Oxford, 31December2017
Abbreviations xv
1.Varro’sLifeandWorks1 1.1AFullLife1 1.2Varro’sWorks2 1.2.1Thegrammaticalworks4 1.2.2Thedateofthe DelinguaLatina 4 2.TheTransmissionofthe DelinguaLatina 5
2.1F,OurMostImportantWitness6
2.2FasaSourceforVarronianSpelling?7
2.2.1Latinpronunciationinthefirstcenturybc7
2.2.2Varro’sspelling10
3.AnOverviewofGreekandRomanGrammaticalStudies25 3.1TheStoics27
3.2TheAlexandrians28
3.3The χνηγραμματικ andApolloniusDyscolus29
3.4Varro’sImmediateEnvironment30
3.5LatinGrammarintheFirstCenturyad31
3.6Donatus,Priscian,andtheGenresofLateAntiqueGrammar32
3.7Excursus:TheDutiesoftheGrammarianandtheQuestion of ηνισμ /Latınitas 34
4.Etymology35
4.1Varro’sApproachtoEtymology36
4.1.1Varro’setymologicaltheoryinBooks2–436
4.1.2Varro’setymologicalpracticeinBooks5–740
4.1.3Varro’streatmentofloans43
4.1.4WhatcanmodernetymologylearnfromVarro?45
4.2TheStructureofBooks5–745
4.2.1Book545
4.2.2Excursus:ThegeographyofRome47 4.2.3Book651
4.2.4ExcursusontheRomancalendar52
4.2.5Book754
4.2.6ExcursusonLatinmetre55
4.3ModernEtymology60
4.3.1TheNeogrammarianhypothesis60
4.3.2Actuationandspreadofsoundchanges61
4.3.3Soundchangeandanalogywithinparadigms63
4.3.4Thecomparativemethodandinternalreconstruction63
4.3.5SubgroupingwithinIndo-European66
4.3.6Wheredoesthisleaveus?67
4.4SoundChangesinIndividualBranchesofIndo-European68
4.4.1Indo-Europeanphonology68
4.4.2Ablaut71
4.4.3Indo-Europeanphonologyandsoundchanges72
4.4.4Italicsoundchanges73
4.4.5Greeksoundchanges80
4.4.6Indo-Iraniansoundchanges87
4.4.7Balto-Slavicsoundchanges92
4.4.8Germanicsoundchanges99
4.4.9Celticsoundchanges103
4.4.10Anatoliansoundchanges108
4.4.11Tochariansoundchanges111
4.4.12Armeniansoundchanges115
4.4.13Albaniansoundchanges119
4.4.14Excursus:Alistofcommonderivationalsuffixes123
5.Morphology126
5.1GeneralRemarks126
5.1.1Varro’smorphologicalbooksandtheiraims126
5.1.2Basicterminology128
5.2TheIndividualMorphologicalCategories129
5.2.1Partsofspeech129
5.2.2PartsofspeechinVarro131
5.2.3Derivationandinflection133
5.2.4DerivationandinflectioninVarro136
5.2.5Case137
5.2.6CaseinVarro145
5.2.7Numberonnounsandpronouns148
5.2.8NumberonnounsandpronounsinVarro150
5.2.9Gender152
5.2.10GenderinVarro156
5.2.11Diminutives157
5.2.12DiminutivesinVarro160
5.2.13Gradationofadjectives161
5.2.14GradationofadjectivesinVarro164
5.2.15Definitenessandspecificity167
5.2.16DefinitenessandspecificityinVarro170
5.2.17Inflectionclasses171
5.2.18InflectionclassesinVarro174
5.2.19TheLatinthirddeclension175
5.2.20TheLatinthirddeclensioninVarro177
5.2.21Tenseandaspect178
5.2.22TenseandaspectinVarro184
5.2.23Mood186
5.2.24MoodinVarro191
5.2.25Voice192
5.2.26VoiceinVarro195
5.2.27Personandnumberonverbs197
5.2.28PersonandnumberonverbsinVarro200
5.2.29Morphologicalintegrationofloanwords200
5.2.30MorphologicalintegrationofloanwordsinVarro204
5.3TheStructureofBooks8–10205
5.3.1Book8206
5.3.2Book9207
5.3.3Book10209
5.4Varro’sGrammaticalTerminology211
6.Varro’sLanguageandStyle236
6.1DeviantNominalMorphology238
6.2 Quibus and quıs 240
6.3Adverbsin -tim 241
6.4Non-AgreeingPronouns242
6.5ImpersonalConstructions246
6.6MoodinIndirectQuestions247
6.7WordOrder:PositionofSubordinators250
DELINGUALATINA
Abbreviations
Abbreviationsofancientauthorsarebyandlargeasinthe ThesaurusLinguae
Latinae.Boldisusedforwordsinthenativealphabets;italicisusedforLatin script.Formoredetailonthemanuscripts,seepp.6–7and18–19.
Sigla
FCodexLaurentianusLI.10,fos2–34
FvvariantreadingsofFrecordedbyPetrusVictoriusandIacobusDiacetius intheircopyoftheeditioprinceps (Fv)FinagreementwiththeeditioprincepsaccordingtoPetrusVictoriusand IacobusDiacetius
fCodexLaurentianusLI.5
B15th-cvariantreadingsrecordedbyPetrusVictoriusincopyofthe Editio Gryphiana;thisorsimilarusedbyAugustinus
GCodexGothanus
HCodexHavniensis
MCodexGuelferbytanus896
VCodexVindobonensisLXIII
aCodexParisinus7489
bCodexParisinus6142
cCodexParisinus7535
pCodexBasiliensisF.iv.13
α twodifferentMSSbuthere,CodexVaticanus1556
Frag.Cass.asinglefoliooftheCodexCassinensis361
<>addition
[]deletion
add.addition
del.deletion
Dictionaries
DVVaan,M.A.C.de(2008), EtymologicalDictionaryofLatinandtheOther ItalicLanguages (LeidenandBoston)
FriskFrisk,H.(1960), GriechischesetymologischesWörterbuch,2vols(Heidelberg)
MaltbyMaltby,R.(1991), ALexiconofAncientLatinEtymologies (Melksham) OLDOxfordLatinDictionary,2ndedn(2012)
W-HWalde,A.andHofmann,J.B.(1965), LateinischesetymologischesWörterbuch,2vols,4thedn(Heidelberg)
Editions
Barwick(1925)Charisius
Skutsch(1986)Ennius’ Annals
Jocelyn(1967)Ennius’tragedies
Bakkum(2009)Faliscan
Lindsay(1913)FestusandPaultheDeacon
Keil(1857–80)GrammariansotherthanCharisius
Marx(1904–5)Lucilius
Ribbeck(1897)NaeviusandAccius
Schierl(2006)Pacuvius
deMelo(2011–13)Plautus
Rix(2002)Sabellic
Introduction
1.VARRO’SLIFEANDWORKS 1.1.AFullLife
NomorethanthebriefestsketchofVarro’slifecanbegivenhere.Readers interestedinproperbiographiesshouldconsultDahlmann(1935)orDella Corte(1970;despitethecriticismbyDahlmann(1955)levelledagainstan earlieredition).Butterfield(2015b)isapleasantlywritten,intelligentoutline ofVarro’slifeandworksthatissuccinctandlearned.
MarcusTerentiusVarrowasbornin116bcinReate,modern-dayRieti, andthusonSabineterritory,hencehissecondcognomen Reatınus,givento himpartlyinordertodistinguishhimfromPubliusTerentiusVarroAtacinus, ayoungercontemporaryandpoet.Varrowasbornintothesenatorialclass; DellaCorte(1978:228)stresseshisfinancialindependence,whichallowedhim topursuehisstudiesforthesakeofstudyingratherthanasameansofearning aliving,afactwhichsethimapartfrommanycontemporarygrammarians andscholars.
ReateisnotfarfromRome,andsoitcomesasnosurprisethatVarrodid mostofhisstudiesatRome,undertheguidanceofthegreatestRomanscholar atthetime,LuciusAeliusStilo,whoalsotaughtCicero.And,likeyoungCicero andanyotherRomanteenagerofgoodfamily,hewassentofftoAthens,where hecontinuedhisstudiesunderAntiochusofAscalon,headoftheAcademic schoolofphilosophy,whoreformedthePlatonists’teachingsbyeclectically integratingStoicandPeripateticelementsintothem.
Varrowasdestinedtoapoliticalcareer,acareerwhichwassmoothonly inhisearlieryears.In90bc,Varrobecamea triumuircapitalis,amemberof aboardoverseeingpunishments,includingcapitalpunishment.In86/85,he becameaquaestor,andshortlythereafter,tribuneoftheplebs.In68,probably, hebecameapraetorandthusoneofthehighest-rankingofficials.
Thispoliticalcareercannotbedivorcedfromhismilitarycareer.Inmilitary matters,VarrowasclosetoPompeyandservedashisproquaestorinthewar againstQuintusSertorius,afellowSabinewhohadtakenituponhimselfto leadtheLusitaniansintheiranti-RomanrevoltinwhatisnowPortugaland westernSpain.Thiswasinthemid70s.Notmuchlater,in67,hewasactivein
thewaragainstthepirates,wherehewonthemuchcoveted coronanaualis,a decorationawardedtoaleaderofsupremebravery.
AlthoughVarrowasclosetoPompey,hewasnotkeenonthefirsttriumvirate,theinformalalliancefrom59to53betweenPompey,Caesar,and Crassus.DespitehisdistrustofCaesar,hejoinedtheboardoftwentythat sawthroughCaesar’scontroversialagrarianlaws(foraperhapsoverlyharsh viewonVarro’sbehaviourtowardsPompey,seeLaughton1956:37).However, whencivilwarbrokeoutbetweenCaesarandtheoptimatefactionofthe senatorsledbyPompey,VarrorejoinedPompeyandbecamehislegateinSpain, whereheeventuallyhadtosurrendertoCaesar.HemanagedtojoinPompey againatDyrrhachium,inwhatisnowAlbania,butwasultimatelydefeatedat Pharsalusin48.
Inhisdesiretoappeargenerous,Caesarpardonedhim,whichledVarroto dedicatehalfofhis Antiquities,the‘divine’part,toCaesar,aclevermovethat ensuredthathewascommissionedaslibrarianfortheenormouspubliclibrary thatCaesarhadplanned.This,however,nevercametofruition,asCaesar wasassassinatedin44.Varroretiredfrompubliclifetodedicatetherestof histimetoscholarship,whichdidnotsparehimfromproscriptionbyMark Antony,whopersecutedsupposedenemiesofCaesarinordertogetholdof theirmoney.Varro’svillaatCasinumwasdestroyed,buthislifewassavedby hisfriendQuintusFufiusCalenus(foranegativeviewonthisfriendship,see againLaughton1956:37).VarrospenttherestofhislifeinhisvillasinCumae andTusculum,studyingandwriting.Hediedanoldmanin27bc.
Collart(1954a:365)sumsupthemaintraitsofVarro’scharacterandscholarshipasindependence,eclecticism,encyclopaedicknowledge,andmethod; buteventhoughwemaysympathizewithhisloveofantiquityandhis‘purely Romanempiricism’(Ferrante1962:171),wemustalsoacknowledgehisopportunism.WhileIagreewiththisassessment,Ialsobelievethatpeopleshould bejudgedfortheirfinestachievementsratherthanfortheirbehaviourintheir weakesthours;andVarrohadagreatmanyfineachievements.
1.2.Varro’sWorks
Varro’sliteraryoutputwasextremelylargeandhadalreadygainedhimaname inantiquity:Cicero(Att. 13.18)describeshimas πο υγρα τατο ‘writing enormousamounts’;Quintilian(inst. 10.1.95)callshimthe uirRomanorum erudıtissimus ‘mostlearnedmanoftheRomans’;andAugustine(civ. 6.2)states that‘hereadsomuchthatwemustbesurprisedthathehadtimelefttowrite; andhewrotesomuchthatwecanhardlybelieveanyonecouldhavereaditall’. However,apartfromVarro’sagriculturalworkandthe DelinguaLatina,we onlyhavefragmentsoftheremainder,andthesearedifficulttocollectandto assess.WhenErnout(1972:162)statesthataneditionofallVarro’sworkina
singleplacewouldbedesirable,ithastobesaidthatthisisasimpossiblenow asitwouldhavebeenwhenallVarro’sworkswerestillavailable.
ThedistinguishedreputationVarrohadinantiquityhastosomeextent beenshatteredinthemorerecentpast:Boissier(1875:28–9)saysthatitis notsurprisingthatVarrowrotesomuchbecausemostofhisworkswerelittle morethancompilationsandsummariesofwhathehadread.AndLaughton (1965:64)isevenharsher:‘Onemightevenventuretheoutrageoussuggestion thatVarrowouldbelesshighlythoughtofifhisworkhadbeenbetterpreserved.’Assooften,thetruthliessomewhereinthemiddle,andwehavetobe ascautiousnottobecarriedawayinourpraiseofVarro’sachievements,aswe havetobecarefulnottoseehimasamerecopyist.
Dahlmann(1973)providesagoodoverviewofVarro’sworks,butCardauns (2001)ismoreuptodateandfocusesonwhatwehaveratherthanreconstructionofwhatwenolongerhave.ItisnowcommonlyassumedthatVarrowrote aboutseventy-fourworksin620books(seeRitschl1848);evenifweconsider thattheaveragebookwasnolongerthanachapterinamodernbook,thisis anastonishingoutputforanyone,letalonesomeonewhowasnotafull-time writer.OnlyasnapshotofVarro’sworkwillbeprovidedherebeforewemove ontohisgrammaticaltreatises.
TheonlycompleteworkwehaveofVarroishis Dererustica,aworkon agricultureinthreebooks,writtenin37bcandthusinVarro’soldage.Forthe text,seeRodgers(2015).
OfthemanylinguisticworkscomposedbyVarro,the DelinguaLatina was withoutdoubtthemostimportantone.Wewilllookbrieflyattheothersinthe nextsection.
The DelinguaLatina hasstrongantiquarianelementsintheetymological sections.Varrowasverystronglydrawntoantiquarianismandhismost importantworkoverallwasthe Antiquitatesrerumhumanarumetdiuinarum, inforty-onevolumes.The Antiquities werebeguninabout55bcandfinished in47.Thefirsttwenty-fivevolumesdealtwithRome,theRomanpeople, andRomancustoms,whilethefinalsixteenvolumeswereaboutreligionas institutedforthegodsbymen.Manyfragmentsoftheworksurvive,butmostly fromthefinalsixteenbooks,quotedbyChristianauthors.Howreliablethese ChristianauthorsareasasourceforVarroisnotclear,sincesomeofthem veryobviouslydistorthimforpolemicalandideologicalreasons.Lactantius andArnobiusseemtoremainobjective,byandlarge,whileAugustine providesanaccountthatisattimeshighlymisleading(Holford-Strevens 2015:153–7).
Despitethefragmentsthatwehave,roughly600andmostlytransmitted byNonius,theprosimetricMenippeansatiresremainsomewhatelusive.They wereoriginallyin150booksandnamedaftertheCynicphilosopherMenippus ofGadara.Butterfield(2015b:14)describesthemas‘anewtypeofethical orsatiricaldiscourse,mixingmoral,paraeneticpreachingwithstreaksof subversivehumourandverbalplayfulness’.
Finally,the Logistorici,inseventy-sixbooks,weredialoguesinprose.Each onetookthenameofafamoushistoricalpersonandthespecificvirtuehe embodied,andconstitutedanantiquarianandmoraldiscussionofthisvirtue. ManymoreofVarro’sworkswoulddeservediscussion,butthesearethe mostsignificantoneseitherbecauseoftheirlengthorbecauseoftheirimpact. Wemoveontohislinguisticpublications.
1.2.1.Thegrammaticalworks
Apartfromthe DelinguaLatina,Varrowroteseveralothergrammatical treatises(testimoniaandfragmentsinFunaioli1907:179–371).WithGötz (1909:89),wecandistinguishthreeperiods.Thegrammaticalworksofthe earlyperioddealtwithspecializedtopics.Thus,his Deantiquitatelitterarum wasahistoryofthealphabet;DellaCorte(1981:154)showsthatVarro consideredtheaspiration H nottobealetter,andarguedfortheeliminationof K and Q asbeingsuperfluous.The Deutilitatesermonis discussedanalogyand anomaly.Andofhis Desimilitudineuerborum,wehaveonlyonefragment, sonothingcertainisknown.Towardstheendofthisearlyperiod,Varro composedhis DeoriginelinguaeLatinae,aworkdedicatedtotherelationship betweenLatinandGreek.
Inthemiddleperiod,wefindthe DelinguaLatina,butalsothe Desermone Latino,infivevolumes,aworkwhichtreatedorthography,accent,quantity, metre,andstyle.Götzbelievesthatthe ερ χαρακτ ρων alsofellintothemiddleperiod;theworkwasaboutwordformationandcontainedthreeormore volumes.
Theonlyrelevantworkinthelateperiodwasthe Disciplinaeliberales, meantasaneducationalhandbookinninevolumes.Onlythefirstofthese volumeswasaboutgrammar.Barwick(1922:18)believesthatthisfirstbook alsoincludedashortsketchofmorphology.
1.2.2.Thedateofthe DelinguaLatina
Wecannowlookatthecompositiondateofthe DelinguaLatina.Forthis, wehaveexternalsourcesinthecorrespondencebetweenCiceroandAtticus, buttherearealsointernalcriteriathathelpusnarrowdownthetimeframein whichourworkwaswritten.Themostimportantdiscussionsofthistopicare Barwick(1957)andRösch-Binde(2001).
LetusbeginwiththecorrespondencebetweenCiceroandAtticusandthe questionofthededication.As5.1shows,Books2–4discussingthetheoryof etymologywereoriginallydedicatedtoSeptumius.BasedonVarro’sstatement inthesamechapterthathewantstowritethreebooksforCicero,Barwick (1957:301)concludesthatBooks5–7werealsooriginallyanindependent work.However,ultimatelytheentireworkwasdedicatedtoCicero,andlater
grammariansquotingevenBooks2–4citethemas adCiceronem (Barwick 1957:298,testimoniainRösch-Binde2001:228–30n.13).
ThecorrespondencebetweenCiceroandAtticusisquitestraightforward. Cicerohaddedicatedhisfourbooksof Academica toVarro,hopingthat Varrowoulddedicatesomethingtohiminreturn.FromCic. Att. 13.12.3, aletterdatedto23or24June45bc,welearnthatVarrohadpromised todedicatesomethingbigtoCicero,butthatnothinghadhappenedintwo years.Cic. fam. 9.8.1,datedto11or12July45,isapolitereminderby CicerotoVarrothathewouldpleaseliketohavethisdedicationnow.Barwick (1957:298)concludesthatVarrobeganhisworkin47andfinishedbefore7 December43,whenCicerowasmurdered,asotherwisethededicationwould notmakesense.
Basedoninternalevidence,Rösch-BindeexpandsandmodifiesBarwick’s findings.Books2–4wereoriginallydedicatedtoSeptumius,Varro’squaestor, andRösch-Binde(2001:242)assumesthathewasquaestorduringtheCivil War,whichlastedfrom49to45bc.Wecanconcludethatthistriadwas dedicatedtoPostumiusshortlyafterwards.
ForBooks5–7wehavemoreevidence,butnotallofthesamequality. In5.100,therecentexhibitionofagiraffeinRomeismentioned.Plin. nat. 8.69tellsusthatthiseventtookplaceattheendof46bc.Lesssignificant inmyopinionisthefactthatin5.169–74,onlybronzeandsilvercoinsare mentioned,butnotthegoldcoinsintroducedbyCaesarin46/45bc(RöschBinde2001:238–9);Varrowithhisantiquarianinterestswouldnotbother aboutthemostrecentdevelopments.Similarly,unlikeRösch-Binde(2001:239) Idonotfinditveryimportantthatin6.34,Varrofailstomentionthechange ofthename Quıntılis ‘July’to Iulius,achangewhichhappenedin44bc;this wouldnotmattersomuchfortheantiquarian,althoughamentionwould nothavebeenoutofplace.Rösch-BindeconcludesthatVarroprobablyhad completedadraftofthistriadatanearlystagebutfinisheditafter46,yet before45/44.
Finally,in5.165,wearetoldofalltheclosuresofthegateofJanus,butthe closurebyOctavianin29bcisnotmentioned.However,allthistellsusisthat theworkwasnotmodifiedforasecondeditionlater,afterCicero’sdeath.
2.THETRANSMISSIONOFTHE DELINGUALATINA
Wecannowturntoissuesoftextualtransmission.Oftheoriginaltwenty-five books,onlyBooks5–10surviveindirecttransmissionandthusmoreorless intact.Oftheremainingbooks,wehaveasmallnumberoffragmentsfrom grammariansandsimilarsources.Theyposetheirownsetofproblemsand willbedealtwithseparatelyattheendofthissection.Beforethat,weneedto
lookatthedirecttransmission;here,manuscriptFdeservesspecialdiscussion. Thiswillbefollowedbyabriefoutlineofthemostimportantlatermanuscripts, andbyalistofthemostsignificanteditions.
2.1.F,OurMostImportantWitness
AmoderatelylargenumberofmanuscriptstransmitBooks5–10.However, despiteAntonibon(1899a:22–3),theyallgoback,directlyorindirectly,to amanuscriptstillextantinFlorence,theCodexLaurentianusLI.10,folios 2–34,abbreviatedtoF.FwaswritteninBeneventanminusculeintheeleventh century;therearefortylinesperpage.TheproblemsofFaresummarizedby Taylor(1996b:33–34):
Itserrorsbetrayascribewhowascarelessbynatureandperhapsalsobyvolition; whosemanifestlypooreyesightwasundoubtedlyworsenedbytheconditions inwhichhehadtowork;whosemispronunciationofLatin,nervousness,and independentlymindedfingersexacerbatedhisafflictedvision;andwhoseother physicalandmentalinfirmities,whatevertheymayhavebeen,cannotfullyexplain theplethoraoferrorswefindoneverypageinF.
Itisperhapsunfairtobequitesohardonourcopyist;Ihavespentmany hoursworkingwithF,andthebiggestproblemIencounteredisfadedink, notscribalincompetence.Iwouldgosofarastosaythatwheretheinkis notfaded,readingFisnotparticularlydifficult.Besides,mistakesaccumulate overthecenturies,andnotallgobacktothelastpersoncopyingthetext. Infact,somemistakesclearlygobackalongway.Forinstance,in10.62we havetoread initium withGroth,butFhas inillum;original initium ismisread moreeasilyas inillum inearliercapitalsthaninlaterminuscule.Othererrors, suchasthefrequentconfusionof a and t,musthavearisenfromanexemplar inBeneventanscript,butthisneednothavebeenthedirectancestorofF. OnemightalsoaddthatVarro’stextwasmangledbyVarrohimself(Serbat 1985:271),whocomposedinhasteandcertainlydidnotcreateatidytext.We mustresistthetemptationtoemendandrewriteVarroinanicestyle(Flobert 1978b).Vetter(1958)containsasensibletypologyoferrorsinF;evenifIoften disagreewithhisproposals,theprinciplesaresoundanddeservetobetaken seriously.Forexample,Vettershowsthatmany‘difficult’wordsweretrivialized inthetransmissionprocess;thatsmallwordslike sıc ‘thus’and sit ‘itmaybe’ weremixedup;thatoftenlineswereleftoutbyaccident;andthatinturneditors haveoverly‘emended’byexpandingVarro’scompressedstyle.
Foftencontainscorrections.AsalreadyrecognizedbyDahlmann(1957), thesecorrectionsarebyadifferentscribe,andtheyarenotbasedonmanuscript readings,butarehisownideas.
Theremainingmanuscripts,beingdirectorindirectcopiesofF,canoffer interestingconjectures,butnomorethanthat.Theexceptiontothisconcerns
5.119–6.61,alongpieceoftextthatgotlostfromFwhenaquaternionwent missing.Itisnotclearwhenexactlythislosstookplace,butithappenedafter 1521,theyearinwhichPetrusVictoriusandIacobusDiacetiuscollatedFwith the editioprinceps.WhereFislost,theircollationmusttaketheplaceofa manuscript.However,itshouldbepointedoutthatinthosepartswhereF isstillavailable,thecollationcanbeshowntocontainerrors,sointhelost portionsofFitmustcontainerrorsaswell.WhereFdisagreeswiththe editio princeps,thereadingsofFasrecordedbythetwoscholarsarereferredtoasFv; wherethereisagreementbetweenthetwo,itismarked(Fv).
2.2.FasaSourceforVarronianSpelling?
Animportantquestionthathasbeenanswereddifferentlybydifferentscholars iswhetherFreflectsVarronianspelling.However,beforewecanturnto ancientandmedievalspellingconventions,weneedtolookatVarronian pronunciation,asspeechisprimaryandwritingsecondary.
2.2.1.Latinpronunciationinthefirstcenturybc
WehaveaverygoodideaofwhatLatinsoundedlikeinthefirstcenturybc.All phonologicalcontrastscanbereconstructedfairlyaccurately,andinmostcases wealsoknowagreatdealofthephoneticrealizationsofindividualphonemes. ThisisbecausewehavemanydifferentsourcesthatallowustoexamineLatin pronunciation(detailsinAllen1978).
Inthefirstplace,thereareexplicitstatementsbygrammariansandlaymen withgrammaticalinterests.Noneofthemwastrainedinphonetics,ofcourse, andsotheirstatementsaresomewhatimpressionistic.Buttheydoremaina valuableresource,evenif,togivejustoneexample,Romangrammariansnever foundouthowvoicedandvoicelessconsonantsdifferfromeachother.Onthe otherhand,theElderPliny(inPrisc. gramm. ii.29)andothergrammarians dodescribewellhowthetwoallophonesof/l/differfromeachother.Latin hadavariantthatwas tenue ‘thin’andonethatwas pingue ‘fat’;todaywe speakof‘clear’and‘dark’orvelarized l.Thesomewhatartificial‘Received Pronunciation’ofEnglish,aswellasmanyofthedialects,likeMancunian,have thesamealternation,withclear l foundatthebeginningofasyllable(light ) anddark l foundatitsend(dull).Romangrammarianseventellusthatthe distributionofthetwoallophonesdifferedaccordingtodialect;AfricanLatin wasrenownedforhavingonlythedarkvariety,justassomeAmericanvarieties ofEnglishonlyhavethedarktype.
Latinmetreisanothergreatsourceforlearningaboutpronunciation.Much ofwhatweknowaboutvowelquantitiescomesdirectlyfromRomanpoetry. Romanpoetrywithitsrulesofelisionalsoteachesusthatfinal -m didnothave
fullconsonantalvalue,asitdoesnotpreventelision;thesamegoesforwordinitial h-.
Latinspellingalsotellsusagreatdeal,especiallythespellingofthoseat thelowerendoftheeducationalspectrum.Latinspellingwasneverfully standardized,soeventhemostformalinscriptionscancontributetoour knowledge,butthereweresomeconventions,andwhereverthesewerebroken consistently,itissignificant.Forinstance,thefrequentconfusionbetween theletters i and e inlatertextsisindicativeofthemergerof ˘ ı and e in pronunciation,amergerreflectedinItalianandSpanish; ı and ˘ e remained qualitativelydistinct,butalsolostthelengthdistinction,resultinginthreefront vowelsdifferinginheight.
Absenceofspellingmistakesdoesnotnecessarilyindicatethatsound changeshavenottakenplace;thewritersinquestionmaysimplyhavehad moreeducation.Onlyoccasionallyistheabsenceofpoorspellingsignificant. Forinstance,NorthAfricaninscriptionsdonotconfusetheletters i and e, despiteaboundinginallsortsofotherspellingerrors.Thatisanindication thattheabsenceofconfusionisnotsimplytheresultofsuperioreducation.St AugustinetellsusthattheAfricanearcannotdistinguishbetweenlongand shortvowels(doctr.christ. 4.10.24;Adams2007:260–5),andthisstatementin conjunctionwiththeabsenceofconfusionbetween i and e indicatesthatthe NorthAfricanvowelsystemwasdifferent;NorthAfricanLatinlostphonemic vowellength,likeothervarietiesofLatin,butmerged ı and ˘ ı into i,and e and ˘ e into e,withoutcreatingthreedistinctfrontvowels.
NeighbouringlanguagesthatarewellunderstoodcanprovidesupportingevidencethroughtransliterationsfromLatinintothemor viceversa,or throughloanwords.Thus,theLatinname Caesar istransliteratedinGreekas α σαρ,withunaspirated κ ratherthanaspirated χ.Thisdemonstrateshandily thatLatinvoicelessstopswereunaspirated,astheystillareinFrenchorItalian. FurtherevidencecomesfromPunic,wherethename Caesar hadinitial q rather than k.Thisisinterestingbecause q wasauvularstop,liketheArabic qaf,rather thanavelarstoplikeLatin c.Punic k,ontheotherhand,wasvelarbut,likeits biblicalHebrewcounterpart,aspirated.SpeakersofPunicmusthaveregarded theabsenceofaspirationintheLatinstopasmoreimportantthanitsplaceof articulation.
However,therearelimitstotheusefulnessofsuchtransliterations.Atsome pointtheycanbecomeconventional,withouttakingfurthersoundchanges intoaccount.WeknowthatinRome,thediphthong -ae- monophthongized to -e- inthefirstcenturyad.TheGreekspelling α σαρ mayreflectstable transliterationconventionsorparallelsoundchange;theGreekdiphthong -αιunderwentasimilarmonophthongization.Similarly,Latin -n- waslostbefore -f- or -s-,leadingtonasalizationandlengtheningoftheprecedingvowel.The spelling,asin Constantınus,remainedconservative,andisreflectedinGreek ωνσταντ νο ,wherethenasalispronouncedbecauseofthespelling.Butthe
hypocoristicform στα isbasedontheearlier,spokenformandshowsthe absenceofthenasal.
Finally,thefactthatsoundchangeisregular,byandlarge,helpsusto reconstructLatinpronunciation.EvidencefromRomanceaswellasIndoEuropeancanbeusedprofitably.Metreisnotparticularlyusefulforvowel quantitiesinclosedsyllables,whichcountasheavyregardlessofwhetherthe vowelinquestionislongorshort.ButweknowthatItalian i goesbacktoLatin ı ratherthan ˘ ı,soItalian villa mustgobacktoLatin uılla ‘farmhouse’witha longvowel.Similarly,weknowthatinnon-initialsyllablesonlyshortvowels wereweakenedintheprehistoryofLatin.Thus,nextto agere ‘todo,act’,weget actus asaparticiple,butthelongvowel,theresultofLachmann’slaw,canonly beproventoexistwhenwelookattheprefixedvariant redactus;thecontrast with facere ‘todo,make’, f ˘ actus,andweakened ref˘ectus isinstructive.
Latinhasthefollowingconsonantalphonemes:
BilabialLabio-dentalDental/alveolarPalatalVelarLabio-velarGlottal
Stops pbtdkgkw g w
Nasals mn [ŋ]
Fricatives fsh
Liquids lr
Glides u i
Allconsonantalphonemesotherthan h andthevoicedlabiovelarcanbe geminates.Thevoicedlabiovelarstoponlyoccursafternasal.Thevelarnasal isanallophoneoftheothernasalsbeforeavelarstop.Theglottalfricativeisa marginalphoneme,pronetoearlyloss,atleastinsomeregisters;noRomance languagehasanytracesofit.Theglidesareheretreatedasseparatephonemes, buttheycouldalsobeinterpretedasnon-syllabicvariantsoftheshorthigh vowels.ItshouldbementionedthatinGreekloanswealsofindaspirated versionsof p, t,and c,butinpre-classicalLatinsuchaspiratesweremostly renderedasunaspiratedstops.Intheclassicalperiod,ahandfulofnativewords alsoacquiredaspiration,suchas sepulchrum ‘tomb’,probablybecausespeakers wronglybelievedthemtobeofGreekorigin.
Latinhasfiveshortvowels, ˘ a, ˘ e, ˘ o, ˘ ı,and ˘ u,andfivelongones, a, e, o, ı, and u.Thelowvowels, ˘ a and a,onlydifferedinlength,but,amongtheother pairs,theshortvowelswereslightlymoreopen,asiscommonforshortvowels (Gussenhoven2007).InGreekloans,wealsofind ˘ y and y,roundedhighfront vowels,butespeciallyintheearlyperiodthesewererenderedas ˘ u and u.In someenvironments,syllable-finalnasalscanbelost,andherethevowelsare lengthenedbycompensationandnasalized.
Onelastquestionthatdeservesbriefdiscussionishowloanwordsarehandledinphonology.Tosomeextent,theanswermustbesociolinguistic.Loans fromlanguagesthataregenerallypoorlyknownand/ordonotenjoymuch
prestigetendtobenativizedphonologically.Languagesthatarewidelytaught and/orareconsideredprestigioushaveabetterchanceofbeingpronounced accordingtotherulesofthelanguageoforigin.WecanseethisinEnglish.In Britain,untilveryrecentlytheforeignlanguagethatwastaughtmostwidely andenjoyedthegreatestprestigewasFrench.Accordingly,Britonsoftenmake anefforttopronouncesuchwordsinaFrenchway.Lesseffortisgenerally madewithGermanorItalianwords,andwordsfromlanguagesfromoutside EuropeareregularlypronouncedasifEnglish.
ThesituationinLatinisnotverydifferent.Latinhasloansfromawiderange oflanguages,butmostofthemgetnativizedinphonologyandmorphology. ThesituationisdifferentforGreek.Intheearlyperiod,Etruscanwasthe languageofprestige,andLivyinformsusthatRomansusedtosendtheir childrentoEtruriaforeducation(9.36.2–4).Greekstillhadconnotationsof servilityinPlautus,andinscriptionsofthetimedidnotuse y or z ortheletter combinations ph, th,and ch torepresentGreeksoundsabsentfromLatin.In pronunciation,loanswereprobablyfullynativized.Weseethisinmanyloans whichsubsequentlyunderwentLatinsoundchanges: machina isfromDoric Greek μαχαν ‘device’,andithasundergonevowelweakening;similarly,Italian colpo isfromfullynativized colapus ‘blow’,aloanfrom κ α ο pronounced withoutaspiration.
Allthischangedinthesecondcenturybc.Greekswerestillviewedwith suspicion,buttheRomanscouldnothelpbutrecognizetheirculturaland scientificachievements.Philhellenismbecamemorewidespread,andmany peopletriedtheirbesttopronounceGreekinGreekfashion.
2.2.2.Varro’sspelling
NowthatwehavelookedatthephonologyoftheLatinspokeninVarro’sday, wecanturntospelling.Latinspellingisoftenconsideredtobenear-phonemic, andwhileitisgenerallyconsideredgoodtohaveanear-phonemicspelling system,afullyphonemiconecanalsohavedownsides(overviewanddetails inCoulmas1989).Letuslookatsomeofthembeforeturningtospelling conventionsinVarro’stime.
Italianspellingmaynotfollowthephonemicprincipleinthatthesame soundcansometimesbewrittenintwoways;forinstance,/k/iswritten <c> beforebackvowels(cosa ‘thing’),but <ch> beforefrontvowels(chi ‘who’). However,theforeignlearnerofItalianonlyneedstomasterahandful ofspellingconventionsbeforebeingabletoknowhowmostwordsare pronounced.Thatsaid,theorthographyisunderspecifiedinafewareas. Forinstance,instressedsyllables,standardItalian,whichisbasedonthe dialectofTuscany,hassevenvowelphonemes,butonlyfiveletters.Theletter <e> canstandforaclosedandanopenfrontvowel,andtheletter <o> cansimilarlystandforaclosedandanopenbackvowel.Thisisgenerally
