Вы находитесь на странице: 1из 22

American Philological Association

Homeric Discourse and Enjambement: A Cognitive Approach Author(s): Egbert J. Bakker Source: Transactions of the American Philological Association (1974-), Vol. 120 (1990), pp. 1-21 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/283975 . Accessed: 28/07/2013 00:03
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Philological Association and The Johns Hopkins University Press are collaborating with JSTOR to digitize, preserve and extend access to Transactions of the American Philological Association (1974-).

http://www.jstor.org

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

120 (1990) 1-21 Association Philological American ofthe Transactions

HOMERIC DISCOURSE AND ENJAMBEMENT: A COGNITIVE APPROACH


EGBERT J.BAKKER University ofLeiden 1. Introduction in Homerhas sinceMilmanParry's(1929)1 The discussionof enjambement The as 'oralpoetry.' ofHomer article topicinthestudy beena recurrent seminal in Homerforsentence end to coincidewithverseend was madeby tendency theunequivofrom intoa basis on whichHomercouldbe differentiated Parry The difference of Rhodesand Virgil. hexameter poetsApollonius callyliterate in by Parry Homerand thelaterpoetswas explained between in versification of whichtheHomericstyleis the technique of theoral compositional terms reflex. orat leasta strong product agree thatin Homertheend of theversecoincidesmore Most authors or unit('sentence')thanin Apollonius withtheend of a linguistic frequently handtheyhave to concedethat poet.2Yet on theother Virgilor any literate by the'sense' is notcompleted many cases where Homerdoes indeedcontain a case of what accordingly, we have to admit, theend of theverseand where maneuattractive and,indeed, A popular enjambement.' called 'necessary Parry to the sophis(by oral standards) subtleties such stylistic ver is to attribute But in doingso, in Homeric anyway. poetry we have to recognize that tication as to literate poets,which applyto Homerthesame standards we implicitly pointof view), an aesthetic factor (from reduces'orality'to a merelimiting we imply that so well. Moreover, whichHomer'sgeniuscould freeitself from 'idea,' and especthought,' worked ('complete theconcepts withwhichParry in oral poetry studies, and uncontroversial ially 'sentence') are self-evident arenot.3 which they be done,in myopinion, that I intend should In thisarticle to do something or stylistic we attribute to Homer'sliterary genius.This is the anything before is an oral style.Of course,there of Homer's styleas primarily appreciation
ed. [1971]). (A. Parry, the collectedwritings 11 will cite from 2 See Lord (1960: 54; 284), M. W. Edwards(1966), Kirk (1966), G. P. Edwards units(on which (1971: 93-99). The strongcoincidenceof metricaland linguistic by Peabody (1975: 4; 141-43) in see also Visser [1987: 31-321) was formulated test, test for orality.' Accordingto the enjambement termsof the 'enjambment independence of the linguistic expression of verse end should not, or very seldom,occur in an oral style. 3 One piece of criticism [19771), (Clayman& Van Nortwick of Parry'sstatistics conceptionof 'sentence' motivatedby a different for example, was apparently refuted (Barnes [19791), but The critiquehas been sufficiently fromthatof Parry. the confusionremains.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

J.Bakker Egbert

in callingHomericstyle'oral.' controversial, even something trivial, nothing of 'oral Homerin terms whohavestudied scholars many But stillI believethat or not whether in thatthey keptapplying, poetry'have notgone farenough, In thisarticle to Homer. language of written standards thelinguistic implicitly, consequencesof what it linguistic and cognitive I will accept the ultimate to modemlinguistic I willresort To do this, meansto speakaboutoralpoetry. orevenoralpoetry. philology ofHomeric thesphere outside on orality research in of 'enjambement' is thereexamination ofthearticle The secondpurpose use. By adopting oforallanguage properties of thelinguistic Homerin thelight to show point of view,I shalltry and discourse-oriented a functional linguistic the usedhitherto, byall standards that evenincases where is enjambement there so thatit is units, betweenlinguistic end of theline is stilla clear boundary Thissecondpurpose altogether. 'enjambement' theterm to suppress preferable basis toEdwards's(1966: 122andcognitive a linguistic amounts to providing cases a 'breakin sense' evenin the'harshest' attempts to look for 37) justified ofenjambemenLt4 oforalnarrative 2. Theproduction ornotin style, whether on Homeric whohaveworked Manyofthescholars that of that style havenoticed thekindoffeature with connection enjambement, in terms 'cumulation' of 'adding'(Parry), (Kirk)or somesuch can be dealtwith nature of theHomericstyleis intimately The cumulative characterization.5 diswhich ofenjambement Parry of thetwotypes theless harsh connected with ofHalicarnassus):6 hecalled 'unperiodic' Dionysius andwhich (after tinguished whichis in a clauseand a phrase fallsbetween verseend in Homerfrequently or some a participial or a prepositional to it, either some way in apposition for example: other phrase, expanding (1)
&W'a pL?v &q>' 't'ILwv IH, xac HI?isaavSpov

Sovp'i i3ak&v ip?S; asiOo;. (A 143-44).

Xapai?,

of thepreceding extension clause,so phraseis a non-essential The participial it createsis considerednot to be very strong.But thatthe enjambement secof thecases ofParry's to many as I shallargue, also applies, 'cumulation' in whichverseend falls ond kindof enjambement ('necessary'enjambement), for that seemtobelong example: constituents closelytogether, between
(2) A'ta; 6? KkXovXookv'OiXta6B6; opotoa; Ccbv ?x, PXa<POvxa caxza cXovov.(11 330-31).

4 Edwards (1966: 122-23): "actual overlappingof sense seldom occurs; one whole componentunit of the sentence and the sense has normallybeen exto the remaining pressed, and (thanksto the case-endings)even its relationship the in comprehending fairly clear. Thus thereis littledifficulty partsis generally even if an idea of its place in the sentence, and in forming sense of a word-group, verse and theverb and occupy the first (forinstance)the subjectand its modifiers its object fall into the second; or vice versa." [1953: 12ff.]). 'construction appositionelle'(Chantraine S See also Chantraine's 6 Kirk (1966) uses the term'progressive,' whichis more in line withthe linear term. stylistic of the discoursethanParry'sprimarily additiveorganization

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

andEnjambement Discourse Homeric

is between break thelinguistic In thisexample, EXeandfkapOtevta tCo6v the verse endin(1),while endin(2) with verse which coincides similar tothat + object parts ofoneandthesameclause(subject essential seems toseparate from oneanother. andverb) instylistic inHomer mainly terms. This hasbeendiscussed 'Cumulation' is seenas the Homeric text inwhich literary approach, dominantly reflects the wehaveto however, When wetake orality seriously, a finished product. static, inorallanguage as a dynamic, ongoing process. Recent research lookatHomer Homeric newperspective itallowsus to deal with possible; use makes this linguistic expresin a cognitive in which we approach framework, discourse ofthe narrator, not from processes ofthe cognitive the point ofview sions from the bya reader. standards applied ofview ofthe point 1985,1987, (1980,1982, Wallace Chafe In a series the linguist ofarticles, that speakers normally do notproduce their thepoint 1989) has elaborated spoken laninanuninterrupted Instead, andsentences. flow ofwords utterances 1987:1). These ina series (Chafe ofvocalization' of'spurts guage is produced properhavethe following callsthem, units), as Chafe (orintonation ideaunits often with contour, coherent intonation havea single ties(1985:106):(1) they preceded (2) they areoften a sense ofclosure; effecting a falling atthe end pitch independent may becomplete, (3) they andfollowed phenomena; byhesitation butthey mayalso be and (a) nounphrase(s), of a verb clauses, consisting units; (4) they onother dependent aresyntactically that chunks ofinformation less. arerelatively but often words, about seven short, that what he callsthe nature ofinformal Chafe fragmented hypothesizes thecognitive in theconcatenation reflects of idea units, speech, appearing amount ofinorthe consciousness, organism. Active ofthe human limitations as limited, is severely can focus on at anyonetime, that a person formation As a result, [1956]). suggests (Miller memory dataon short-time experimental by consists canbe processed a story thetotal ofwhich ofinformation amount insmall would seemtobe andthis chunks ata time, an oralstory-teller only inideaunits. discourse ofthe linear organization inthe fragmented reflected research as wellas ofrelated ofChafe's Oneofthe findings, consequences losesmuch of its in conversation of 'sentence' is that theconcept analysis, When what production. ongoing language with when wearedealing importance atthe endofany canoccur falling intonation wecallthe typical sentence-fmal to has not been brought unit,even whenthe syntax idea (intonation) haveany is. Ifsentences what a sentence so clear itis not anymore completion, thatis it is a function speechand narrative, function at all in ongoing in inthe linear organization that result constraints ofthe cognitive independent ina sentences the fact that from 46-47]).Thisappears (Chafe [1987: ideaunits to in length from in length andmayalso vary speaker uneven text arevery sentences oftexts, tothe may related aspect productive Instead ofbeing speaker. ofa as to thepresentation decisions be seenas theresult of a speaker's than ofthe rather cognitive orstyle, a matter ofrhetoric, they arethus narrative; inthe mind. activation speaker's ofideaunits sentences tend articulation oftexts, As a major ofthe rhetorical component to speaking in writing in speech. (and than Contrary to be more prominent

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Egbert J.Bakker

bycognitive that arelessdetermined areactivities (andreading) hearing), writing an oral position than leisurely often in a more A writer is very constraints. quality. integrated hasa more text heproduces the narrator; consequently, edited, way a wholes, andthe informal as integral texts arepresented written Sentences the howheenvisages linked with indissolubly a sentence is often writer begins be opofspeechcan the fragmentation Thus ofwhat hehasbegun. completion integration ofwriting.7 posed tothe tothe as opposed nature ofwriting Anexample integrated oftheedited, inthat shewent faster like "And then is a sentence nature ofspeech fragmented was this sentence (1989:4) data, inthe class."InChafe's beginning classthan ideaunits: ofthree as a succession orally produced
(3) a. ...and thenshe...wentfaster.
b. ...in that class,

class. c. ...you know than in the...beginning

with the appropriate clause, first ideaunit as a complete realized the Thespeaker inthetranscription). The stop (indicated bya full falling pitch sentence-final tothe its first, as anafterthought after a full pause, wasadded, second ideaunit finally, thethird unit, was to follow; that more pitch signalling non-falling one by the to thesecond connected theutterance to itsend,being brought know." device "you ideaunit linking English typically inideaunits has speech ofongoing organization Thetypically fragmented written educated differs considerably from formal which its own syntax, than on theproon thetext produced is more where theemphasis language, for arerelevant my ofongoing speech ofthe syntax Twoaspects duction itself. and(2) theloosely ofideaunits usedfor thelinking purpose: (1) thedevices tothe clause with andadverbial respect status ofnoun phrases connected phrases I mean what 5 belowI shallexplain In section bythe to which they belong. first dealwith the one. wewillnow second aspect; and[a] noun a verb phrase[s]) with properties (i.e. with clausal Idea units and ofwhich markers, oflinkage bya number toeachother may be connected (Schiffrin [1987:150]; in English frequent neutral andthemost is themost and Ina grammar sentences, 'perfect' describing formal, Chafe [1989:10-11]). that aresituated clauses linking conjunction as a coordinating may be described Inthe sentences.8 gramon the complex level, thereby yielding samesyntactic as a linking device hand, and serves on theother language, marof spoken
7 Notice thatthe distinction is relatedto and fragmentation betweenintegration tendsto and speech. Writing but by no means identicalwiththatbetweenwriting than speech, but it has its less formalvarieties. Likewise, be more integrated to but it has manygenres,rangingfrominformal fragmented, speech is typically formal (see now Martin [1989: 11]). Especially oral poetry may be highly (Bakker:in prep.A) in preparation and 'rehearsed.'In workcurrently premeditated I address the tensionbetweenconscious design ('rhetoric')on the one hand and orality('cognition') on the otherin Homericpoetry.In the presentarticleI am exclusivelyconcernedwith 'cognition.' 8 On the level of (logical) semantics,and signals that when the one of the clauses is 'true,' the otheris true too. This is what in propositional component logic appears as the logical relation'conjunction'(&), see Allwood et al. (1977: 32), McCawley (1981: 16).

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Homeric Discourse andEnjambement

between twoclausal idea units, in function ranging from an empty space-filler between twovocalizations9 toa continuative (progressive) connective signalling thefollowing that idea unitis morethan just a rewording of theprevious one (Chafe [1989: 11]). In narrative therelation between twoidea unitslinked by and is most oftena temporal one, but this is not due to and itself:as a maximally neutral element, theparticle just signalsthat there is a relation of somekind, without it. further specifying The application to theHomeric style of theconcept ofthefragmentation of oral narrative intoidea unitshas,I think, an immediate appeal.And theabandonment of thesyntactic conceptof 'sentence'in favorof theidea unithas important consequences for thediscussion ofenjambement. The factthat many citations from Homerists can be givenwhosewording closelyresembles that of Chafe10 maysuggest that thetheory of oralnarrative production is a statement in theoretical terms of whatclassicistshave alwaysfeltintuitively aboutthe addingand cumulative styleof theHomeric poems.In theremainder of this paperI shall try In section3 below I addressthe to workout thisstatement. phenomenon of clause linkage in Homerin terms of idea units. 4 and Sections 5 deal withenjambement in Homerwithin thecognitive framework of oral narrative production. This discussion ignores to someextent thetechnicalities of theproduction in section of theverse;these willbe dealtwith 6. cursorily 3. Clause linkage in Homer andidea units One ofthefeatures of theHomeric that modstyle most clearly agreeswith ern data on oral narrative is the linkageof shortsyntactic cola by 5E. The occurrence of connective particleslike 5e is a constitutive partof Parry's of 'sentence,"theconceptwhichunderlies definition all the discussionsof in Homer:what&e introduces enjambement a new sentence.11 It is, forParry, seems preferable, however,to describe5e (in Homer,thatis) as a linkage marker between clausalidea units, signalling that there is some,as yetunspeciWhat F connects fied,relation betweentwo linkedunits.12 are notso much in therhetorical unitsthatare a factor of texts, as thesuccessive articulation ideason which thenarrator linked focusses whileunfolding thestory. The units andoften more than one ofthem by &e arequiteshort go intoone line: (4)
6R 6d xv.Cov' E
61 aixcipov na xij 61 O{paGE (E 416)

9 In Beaman (1984: 57), and in spoken discourse is denied the status of coordinator on these grounds. 10 For example, Parry's (1971: 253) definition of 'unperiodic' enjambement: "...the sentence,at the verse end, already gives a completethought, althoughit goes on in the nextsentence, addingfree ideas by new word groups" (emphasis mine), or Kirk (1976: 152): "...any simple and paratactic narrative is cumulative; each new piece of information, as the storyproceeds,can be envisaged as being heaped upon its predecessor." 11 Parry(1971: 253): "I definethe sentenceas any independent clause or group of clauses introduced by a co-ordinate conjunction or by asyndeton." 12 It should be noted thatthe function of SC in later Greek (e.g. in cultivated Attic prose) is quite different: in breakingup the text into meaningful units,it does serve a rhetorical, text-structuring purpose (which I discuss in Bakker [in prep. B]).

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Egbert J.Bakker

as is shownin areon thewholeofequal length, they thefactthat Furthermore, they are connected suggests that withtheflow passage,strongly thefollowing presentation. In other words, with conscious rather than of ideasin thenarrative or stylistically: notrhetorically determined, is cognitively their length
(5) 8' Et 6Xov aZiv tkamEV &X0oXagagE, 'Ex-Twp 'dkwXv5' '4'a Soipa xcact otpat6v `Xvyon'vtn, 8 i quXointv 6op{vcvvgaXiGaaOat, EYEtp& aivijv. Oi8' iEXiXOiaav ica' iaVriot LEYtav'AXat@v, 'Apytot6' iupcoOev iicaptvivavtospaXayya;. msav 8' avtiot EV 8' 'Ayag4c'vwv apruvO Si gaiX%, M cEhXv8i ioX0i RpogacOat &naivxov. npaao; O6pouo', (A 211-17).

whicha cola, between narrative theunitslinkedby U? are short Most often, need notbe a temporal exists.But therelation (sequential)relation temporal linking device,5e in Homersimplymarksthe neutral one. As a maximally and ina (nonEnglish also in cases where ofone idea to another,13 progression is impossible: oral!) translation (6) jl pa' vv
goi
tt

dt. niOoto, xaasiyvio; SC' otoi

(H 48)14

fortheuse of 68 seems to be thatthesecondof two condition The minimal in thefirst: of whatis stated justa rephrasing has to be morethan linked units wordgroups Whenthesecondoftwocoordinated marker.15 &? is a progression image, one and thesame mental and bothexpress of thefirst, is a restatement of &?: Kca orx? have to be used,instead
(7) v xcat &64vro ipo6v pjiap (A 84)16 o6ppa gEv 'i otfl 8' yvt4 iptu'v xa' ipcioa?o xutptnaxctn A 355).

(8)

oi

(l &VatO6 tr' jEXto; oato'wr6O'-i naoat &yvtai


'nei o'v O yVVtpv

O'gTIy&p'e;

N'' iy'voVro

(A 37)

388)17

of one and thesame idea. statements of twodifferent These examplesconsist which as one idea unit, maybe introduced I wouldanalyzethem Consequently,
in the normalway by 8k.&I
13 See also the treatment of SC in Ruijgh (1971: 128), who uses the term 'transitive' to characterizethe transitionfromone 'fact' (as he calls it) to of Se as a svSzas,o; another.Compare Apollonius Dyscolus' characterization

underthe by the ancientgrammarians 14 Cases like this one have been treated heading "6 8e avr't toivyap," because a causal relationbetweenthe two clauses actually seems to obtain. This may be true, but the relation is not therefore exsressedby SE'. See also Chafe (1989: 11). t' 16 Comparethe equivalentcase E6; O icE &9pIc('xr)at / &q vTja; eiVooeu0 icva; vca et (A 194, 209). 'tepov E leXto; 17 When the second memberof the linkedpair containsnew information, 8% is used instead of te: 6{aETo 6' hE'Xto;, tvrikeato 8? Epyov 'AXatC6v(H 465). Notice, however,thatthe vulg. has Xr?. clausal con18 See also Ruijgh (1971: 131), who analyzes xai as coordinating stituentswithin the scope of units linked by 8*". Ruijgh's terminologyand

J?exacaxticot;.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Homeric Discourse and Enjambement 4. Homeric enjambement andthe ideaunit

Units linked by &? in Homeric (likethose byand in oral Greek linked English narrative) haveclausal properties, that is,they consist ofminimally a verb, very often accompanied bysubject and/or object constituents. Suchbasic clauses'9 havebytheir very a sense ofcompleteness. Thisis why Parry nature doesnot ofenjambement speak when verse endfalls between two linked clauses by &.20 Instead, he uses 'enjambement' forthose situations where a verse opens with a non-clausal phrase which is insome waysyntactically related to whatprecedes in theversebefore. The enjambement is 'unperiodic' or 'necessary' according tothe degree ofgrammatical closure that isreached atthe ofthe end preceding verse.21 in terms Now in an analysis of idea unitsthisconception is quite In this interms drastically altered. analysis we arenot thinking anymore ofa complex sentence thatoccupiesmoremetrical space than just one line. Approaching the problem from the cognitive, text-productive side, we haveto think interms ofa mental Thenarrator on picture anditsverbalization. focusses The the various aspects ofthis picture, thereby producing a series ofideaunits. ofthe grammatical andsemantic relatedness ofthese units reflects the coherence is the mental picture. Thewritten, edited correlate ofsuch a series ofideaunits text with sentence, butthefact that Homer is materially a written sentence should not inpunctuation inourmodern articulation (appearing printed text) induce us tothink that itis conceived in thewaywritten texts are.Thenonoforal integrated, fragmented quality ofthe Homeric style is highly indicative initssoftened, narrative, andthe term 'enjambement,' even 'unperiodic' sense, shown to ofthe verse canbereasonably should be suppressed as long as the end I have inmind, the To get consider fall between twoideaunits. anideaofwhat following passage:
(9)

c a c Aapxi6 no .vi (a69v, Xo T"o8ei8i voapiveooo, 11 .n xCexlC^dvo; av8polxX ,.p'n' ''i)


"Iko-o Aap8avib8uo,naalov)

ar&ap 'AXi4av8po;, 'EXEv1; not; hiiioiioto,


(A 369-72). 811poyepowco;.

frommine, but the two accounts,as far as are different conceptual framework the same. HomericGreekis concemed,are essentially we ("by 'nuclearpredication' 19 Cf. Dik's (1978: 15) term'nuclearpredication' number of termsfunctioning of a predicateto an appropriate mean the application of thatpredicate"). as arguments by whentheclause introduced 20 However,he speaks of necessary enjambement clause in the preceding the main clause of the subordinate BE is, syntactically, verse(as in A 57-58: oi 6' inel ozv "yep(kV o'tIYPycp; '' ytevoVto / toiot 6' 'AxtXXti).This practice,however,amounts 06&x c%i o&vtcytavo; p?ipr (and hence of apodotic 68) in Homer: the of subordination to a misjudgment 'main clause' ( aojot 6' avto'pcvo; ick.) thatallegedly completesthe thought begun by the precedingsubclause, is in realitya freshclausal unit which is to whatprecedesin thenormalway by 6E` apended ' Of course,the distinction is to some extentsubjective.See also Kirk (1976:
150).

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Egbert J.Bakker

analyzed as a succession of 8 idea units, 2 per This passagecan be effectively sentence image,rather thanas an extended line,unfolding a coherent mental 4 lines.The first occupying unitis a merename(acUxca-p 'AX,itav8po;).The as whatmaybe called a topic-shifting name,standing on its own,functions ofthenarrator andthehearer from one warrior devicewhich directs theattention theparticipant in thenarrative aboutwhom in thebattle to another: itmentions thepoet is going to say something. Owing to the forceof accuip, the unit In functional linguistic terms, we maycall effects a boundary in thediscourse.22 fragmented style willbe said in ita theme.23 Moreon 'themes'in theHomeric section 5 below. as a The second unit,'EXEvvil; soo6t; 'iK6fioto, is added to the first in apposition. of the Fromthepoint ofviewof theproduction qualifying phrase extendthe name verse,however,the function of the unitis to optionally of instance we encounter 'AX'tav6po; to theend of theline.This is thefirst theinteraction of theproduction with theproduction of theverse. of idea units In this ofthelatter. case theformer is in service The first unit of thenextline(Tiu65i6ii''t6c ctut-raiveo) is theclausal of thewhole.By all standards thislineis a case of neceselement usedhitherto, since thealleged subjectof theclause ('AXE'av6po;), is saryenjambement, ina linein which in myopinion However, expressed theverbdoes notoccur.24 constituent at all: the thematic constituent 'AX,E'av6po; is nota necessary it stands outsidetheclause proper, constituent, beinga syntactically detached it. Thus,line without and Tx6e?8qit 6x is perfectly complete 6ta tattcvr-ro aimedhisbow" butas "Alexandros, 369-70 shouldnotbe readas "Alexandros is no enjambement here.In secthere (...), he aimedhis bow." Consequently, tion 5 I shall say more about the special status of noun phrases like 'AXEtav8po; in Homericdiscourseand its consequencesforthe studyof enjambement This Xacov. Line 370 is rounded phrasentotgvt offwiththeappositional the with which thespacebetween tothestereotyped phrases expression belongs of these bucolicdiaeresisand theend of theline maybe filled.The function andinan easyway all circumstances phrases is toenablethepoettoreachunder a dactylic and closure theendof thelinewhentheendofa unit has,metrically, An other fallsat thebucolicdiaeresis. 6oup) exampleis 8otup(pa tvCp (6tEX with verbs of killing or wounding(see Visser 1987: 80-82; Bakker &
22 Notice, incidentally, of thatin laterGreek this is preciselya major function fromone clause to anotheris connectedwith a shift U: whenevera transition fromone topic in a narrativeto another,be, and not cai, is the connective Bakker(in prep.B). particle.See further 23 Theme constituents are noun phrases or adverbial elements(subclauses or detachedand placed beforetheirmain clauses. participles)thatare syntactically into the discourse,a topic a new referent in introducing Nominal 'themes' effect, shift; adverbial themes effect, in marking an incision between two action Bakker(forthc.). break.See further a paragraph sequences in a story, o iapiltova, o te pow pow pynv, 24 Kirk (1976: 150) addressesa similarcase: Mevot-rti&aoBpaijvat. Like the piXkatovav8p&v, / goip' not Hlarpo6ickoto case under investigationhere, the first verse ends with an apposed phrase (9ptkcazov &vSpxv) betweenwhichand the followingline a pause (in sense as well as in sound) is conceivable,while the syntaxof the sentencegoes on.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Homeric Discourse andEnjambement

theassociativechain,28 theperson to whomthetomb continue by mentioning such as EXFvils; a verse-filling apposition belongsand by providing soot; in 1.369,respectively. iPSoK6`loto

units("Ikoi Accp&ovixao, itaXatoi 6r1t,oyFpovtor) leans. The third and fourth

versewith these standardized, load connected Fabbricotti: The cognitive fc.).25 comparing them, notto Chafe's we might consider closingunits is so low that between phenomena butto thevocalization/hesitation idea unitsthemselves, of thisidea is that whenthepoetis The point them (likeyouknowin English). what he is already planning or somesuchexpression, producing Xa6ov inotpL1vt comesafter. From unperiodic enjambement. ofParry's The nexttwolinesare instances Homeric the pointof view of sentential syntax, theycontainthe typically in theform ofthebasicclause: 'circumstantial ofparticipmodifiers' extensions or thesentence couldstopbefore phrases. Grammatically, ial and prepositional is 'unperiodic' (or afterany of them,and this is why the enjambement however, thetwolinesare In theanalysis in terms of idea units, 'progressive'). alreadybegun,as additionsto the of a sentence not so muchan extension the Parisis seeking coverbehind whileshooting at Diomedes, mental picture: is gravestone on thetomb ofTroy'sfounder Ilos. The completion of thepicture focusses successively. on which thenarrator vocalizedinfour idea units The first of theseunits(atinX, EKXtp1Vo;) iS a paticipial phraseof a in thefirst is very frequent halfof theverse(see also ex. (1) above);26 typethat expressclausalunit, as an afterthought tothepreceding itmaybe characterized reported took place.The following under which theaction ingthecircumstances formuinxt i'4tfpc) is a prepositional of a common unit(&v6poqnitxw phrase Paris stands on which theplace where thegravestone laic type27 that specifies

25 Sometimesthe stereotyped closing unitis as completeclause: niirc

5?

Xa6;,

6 The well knownplacing of middle choriambic in 'runover'posiparticiples relevance, of the verse, a usage with obvious 'versifying' tion at the beginning is a special case of this type.It is one of Russo's (1963: 242) more convincing formula.' examplesof the 'structural 27 This type consists of. the following structure:adjective (or noun) + preposition+ noun (or adjective). The expressionshave in common that the before the prepositionstands in dative singular ending of the adjective/noun
hiatus (see Bakker 1988). Examples are:
8t9pcpII, g6n iv'i

display a prepositionalexpressions in the firsthalf of the verse consistently structure:noun + preposition + adjective, whereby the (dative or different (e.g. nI?rpT ?x' U vn, vip iv correption 4enitive)endingof the noun suffers type has the same Note thatthis verse-initial yaqn ?V awptpwrcp, aXXo5yajj). as vica; Ent yaqpupa'; and similarexpressionsin which corpattern rhythmical reptiondoes not occur. An example of both typesin one line is Hes. WD 599, consciousnessof the narrator, 2- Note thatthe tombof Ilus is in the peripheral since it is mentionedsome time before (1. 166: nap' "Ikou ijj)la nakaXtoi Aap8avi&ao).
Catci ?

icu6tavdipnII

iv'i

Vt ?v'i vriI, ?Ut' V UooSikgp oiiCCp// etc. It is interesting to note that

;cat

EuipoiXak

iv a&xo.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

10

Bakker EgbertJ.

the passage sentence,29 Thus insteadof being one complexintegrated vocalizethe which together idea units, successive discussedfallsapartin eight picture: ofone coherent mental aspects various (10) a. ButAlexandros, hair, of Helenof thefair b. thehusband hisbowhe aimed, c. toTudeides of thesoldiers. d. theherdsman on thegravestone, e. leaning tomb, f. on theman-made ofIlus theson of Dardanus, g. (that) h. theelderof thepeople.

role;inrequiring a verse-technical clearly theunits b. andd. playmost Of these, to the above, be compared as I suggested theymight, energy, little cognitive inoralspeech idea units significant between phenomena hesitation/vocalization the a pause) fillthespacein which simply (often Thesephenomena or narrative. plans ahead for the subsequentdiscourse. To speak of speaker/narrator a is notdesigning in thiscase is uselessin so faras thenarrator 'enjambement' a mental picture. butdeploying sentence right-dislocation 5. Left-and In Parry's terms,the 'enjambement'in the passage discussed in the in question in thesense thatthesentence previoussectionis 'unperiodic'30 I discuss In this section verse. the enjambing an end before have come to could ofone andthesame'sentence' constituents essential twoseemingly cases where called 'necessary' verses.These are cases of whatParry fall in twodifferent enjambement as the to havesimpleand smallconstituents oraldiscourse tends Informal these object[s])ofa verbin a clause.Veryoften, direct complements (subject, pronouns. or demonstrative personal are as simpleas unstressed complements complexnounphrasesmay (i) highly is twofold: The reasonof thistendency an effective and henceimpair of utterances, theease of processing diminish in that it is contendsto be context-bound (ii) oral discourse communication; is there deal ofsituational knowledge: a great between tracted peoplewhoshare understands as theaddressee often no needto use fullnamesor nounphrases, a pronoun is used. evenwhen whoorwhatis meant bythespeaker, anyway is notsufficient reference (or is pronominal simple however, Equallyoften, has to be used in A fullnounphrase by thespeaker). sufficient notconsidered this is that in thisrespect of oraldiscourse thosecases, butthecrucialfeature of theclausein question. Instead, of thestructure is notmadea part constituent
29 For example, the one in Rieu's translation (p. 207): "But now Paris, the husbandof Helen of the lovely hair,drew a bow on Tydeides the greatcaptain, leaningforcover againstthe columnson the moundwhichmen of a bygoneage Ilus son of Dardanus." Notice especially the intehad made for theirchieftain, of the last two verses ("leaning for cover..."): gratedquality of the translation unitsis greatest. withHomer's fragmented here thedifference 30 Except forthe 'necessary'enjambement between11.369 and 370.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

andEnjambement Discourse Homeric

11

and'rightis called'left-dislocation' it;this orafter before either itis placed in the element is 'represented' TheL/R-dislocated respectively. dislocation' This (clitic) pronoun. or evenbound (as in English) clausebyan unstressed as process speech nature ofthe inlinewith thefragmented is entirely strategy inplanned discourse, written writing: ofedited nature integrated tothe opposed ofthe complement direct function as the may phrases noun themost complex theL/Rspeech, clauses.In unplanned integrated complex, forming verb, the orafter before uttered ideaunit, as a separate beseen may element dislocated one. clausal concomitant before ideaunits, as separate that areuttered, phrases noun Left-dislocated Anexample themes.3' belong, arecalled the they properly towhich clausal unit Thebasic theme was &u-r&p ofa L-dislocated 'AXiav6po; inex. (9) above. within of discourse) (universe is to specify the'domain' function of themes (see say(s) something clause(s) the subsequent which) entity about (orthe which a may be called what effects a theme narration, Dik [1978:132]).Incontinuous a inconversation while section, previous as we haveseeninthe switch, topic theleading topicof the to establish a theme constituent mayutter speaker on,to before going the theme, after s/he pauses Often, conversation. subsequent with familiarity hisorher ordeny toexpress opportunity the addressee givethe Anexample the is topic.32
(11) A. That student of yoursthatcame to my officeyesterday. B. Hm. production. A. She's going to do a paper on oral narrative

wants to reason orother, some for when a speaker, occurs Right-dislocation ofthe ofoneofthe referents preceding the identity add,bywayofafterthought, of oralfragmented feature Like left-dislocation, it is a typical clausalunit. Rreferent-establishing, than In being rather explanatory, ('loose') style.33 do i.e.they L-dislocated than ones, arelessdiscontinuous dislocated constituents examples, inthe Some a break discourse.34 not effect
(12) He's a nice fellow,yourbrother. I wantto buy it, thathouse. He gave it to him,the book.

of inGreek, inspite occurs L- andR-dislocation too, is that Nowmy claim Ancient Greek does from ourtexts. that this thefact doesnotovertly appear clitic unstressed pronominal and/or Western possess languages, likemodern not, constituent. for the However, which function as 'substitutes' displaced elements
31 Linguisticusage is highlyinconsistent here. Very oftenthe term'topic' is used forL-dislocation. 32 The identity betweena thatis negotiated is veryoftensomething of referents speaker and an addressee. See the study of L-dislocation in these terms in Geluykens(1987: ch. 5). 33 Linguistic terminology conis still more diffusethan for left-dislocated topic,' 'antitopic' stituents.I have found 'R-dislocated topic,' 'afterthought (Chafe [1985: 115]), and 'tail' (Dik [1978: 19, 153-56]). 34 For discussionof 'themes' and otherlinguistic of (discourse) devices in terms see Giv6n (1983). continuity,

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

12

EgbertJ. Bakker

Greek can express subject agreement, bytheinflected morphology oftheverb.35 This meansthatan expression like 1coicp6T1; E-I7 is principally ambiguous between thereadings "Socratessaid" and "Socrates, he said." Likewise,?ME would allow of thereading"He said, Socrates."36 In thecase of ,coiCp6-r7J; Greekdoes notpossessan overt meanstoexpress unstressed objectsin objects, a clause,that are understood in thecontext of utterance; instead, it uses 'zerorealization,' as in
(13)

("He gave it [sc. the armor] to his squireEreuthalion to wear'),

6U)x

6'

'Ep?uOaXiowvt

PiX(

Opaxov.Tt

pop vat

(H 149).37

But thismeansthat whentheobjectis overtly expressed bya fullnounphrase, Now Homeric Greeksometimes allows us to 'prove' thedislocated status In thecase of L-dislocation of a nominal constituent. thismaybe done on the basis oftheplace ofan enclitic andin thecase of R-dislocation on the particle, intheclause. basisoftheoccurrence ofa 'redundant' pronoun In a recentpaper,Ruijgh (1988) addressestheplace of encliticsin the ina reexamination Homeric phrase what ofWackemagel's Law. He arguesthat are seeming to theLaw (viz. enclitics that do notoccupythesecond exceptions in theclause) are in factno exceptions at all, because the buta laterposition preceding words areleft-dislocated. Ruijgh cites, among other examples: (14)
itot 'o y' C'; rt-irviccar'a&p' ?4Xro. (A 68),
"he gave it to E., the armor").

it may be read as a case of R-dislocation (*&Moc 6' 'Ep?xOcxXiOVt TE'r%ECC,

of theplace ofap', has to be readas ''Buthe,having which, on account spoken thus,he sat down." In otherwords,o y' is a themeconstituent (to whicha withicoxr'. Another participial phraseis attached);theclause properstarts exampleis:
(15)
otiOEM aL&ap

Klat oiU; jEv Xixev acUOt &va4 &v6pxv 'Ayap?4ivvw

bre't xepl ?xlt@va; 7rawpaivovta;, 3' rIGov TE pi] ic?a "Avlttpov i4Fvapit,v

101).31 (A 99_

35 Ruijgh (in personal communication) suggestedthatit is in fact possible to analyze the personal ending on the verb as pronominalsubject. The athematic to jioi, iot',tic. ending-gt would thenbe the missingnominative 36 See also Ruijgh (1979: 71). 37 See also A 109: "Avwttpov ?S i2aa? ak 1p& ti?Pt, iic 6' "Pax' Ynno)v, where the object (Antiphus) is unexpressed in the second unit. An extreme in 101), the is B 102-8. The aiClrpov (introduced example of the phenomenon leading topic of this passage, is object in this passage 6 times, but never ex3pressed. thatthispassage is a good exampleof the factthatHomericGreekoften exrNote as 6? in later Greek: in this passage, it is uses awtap with the same function with g?v in 1. 99. correlated

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

andEnjambement Homeric Discourse

13

case in It is notveryclearwhatp' does in 1. 101,39 buteven in theminimal in the it cannotbe placed randomly to avoid hiatus, whichit servesmerely So in thisexample of enclitics. clause: it has to abide thelaw of theplacement But and does notbelongto theclauseproper. o is L-dislocated it indicates that or of p', semantic whatever is thefunction I think: we can go a stepfurther, thanto other is used fora purpose it is clearthat theparticle merely prosodic, we Consequently, constituent. status of thepreceding indicate theL-dislocated the thematic, L-dislocatedstatusof a noun have to take into consideration deviant in its seemingly apac(or any otherenclitic)40 phrase even when 41 position is absenL maybe discourse of R-dislocation in Homeric The 'proof'of theexistence madeon thebasisofcases likethis: ava4 &BvpFv'Ayapi4ivov (B 402). considered to be 'pleonastic'(e.g. Visser pronoun is often The demonstrative [1987: 118]), on the groundsthatthe subjectof the sentenceis expressed in thesame verse.I believethat and that the thisanalysis is incorrect anyway which is looselyattached constituent is in reality a R-dislocated allegedsubject to the clause. And I want to extendthisanalysiseven to cases wherethe is absent, for example, pronoun 'pleonastic' (16)
ala&p

6 PoiV ivpVxa?V

(17)

ia't 6vi ponorXov,ca't ya(pt' ovvap?}c AtO6; iatv (A 64). 6; C' 1ixot 0 11 tO6aOV ?cXw0ato Doio; 'Aio6kkXv

Achilles element: maybe readas a R-dislocated The phrase 4olpof 'AnokkXOv that Apollois thegod who in context, himthere is no doubt and for is speaking to himas 'he,' but adds his name to avoid all senttheplague; so he refers is not visiblein our In cases like this,theR-dislocation misunderstanding. itmust havebeenmadeapparent butin oralnarration bymeansof intonatexts, tion. This is of course a different way of sayingthata R- (just as a L- ) uniL42 is a separate idea (intonation) dislocated constituent is for the studyof enjambement Now the relevanceof these insights from therest is separated a seemingly constituent obvious:whenever necessary calls a case of whatParry of itsclause by a verseboundary (thusconstituting thatit is a withthepossibility we have to reckon 'necessary enjambement'), of all its 'enjambement' whichwouldrobtheterm case of L- or R-dislocation, the status of a constituent theL-dislocated Sometimes just before application. in thesame way as in (14)115) above. verseend is provedby an enclitic, thefollowing Consider example (citedbyRuijgh1988):
39 Notice thatthereare variant l? or Bilptao6v here:eitherp' is omitted readings the is read (which presupposesitzva' ptl insteadof tE,?vapiE,ov).Apparently alreadyin Antiquity positionof p' caused trouble 40 For example oDv: oi 6' inzl oO'v Ij?YpV, 6prjyp?? X' 'yvovro (A 57), whereoi 6' is thetheme. 41 Compare, for example, with (15): cota'p 6 Pi caiv Bovpi gvr' &vriOFov HokX6opov (Y 407). more ado to all nominatives 42 Notice thatR-dislocation does not apply without in the case of rov ihbicojo; C1?Cx placed at the end of a clausal unit For instance, concept. does not seem to be an appropriate (A 36) R-dislocation Ai-jtw

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

14 (18)
v

J.Bakker Egbert
b' 'n IW)g' 'v qryz (H 220-21). 8' ?5 iXtv, p' tF?vMvpp8Lov, 'AZXXES6 XOL npoO00v noxqucv. P3i

att&p

by theposition ofapa: proved


(19)

ofa is a goodexample a&'&p 'AxtUXX6; constituent theme Theleft-dislocated from switches of thenarrator effected byamk6p: theattention topicswitch This toAchilles. for battle preparations andtheir Myrmidons andthe Patroclus a separate warrants which inthediscourse, creates a major boundary switch, with whom thesubsethetopic which introduces ideaunit nominal (person) will beconcerned. discourse quent phrase. noun a simple than is more constituent L-dislocated Sometimes the in both cases theL-dislocation is, again, examples; thefollowing Consider
viv?S i,vOa &' &vi p intv avspa Xc8aaOEi;n 'ui60 iyt?OV(cV. np&tO; &iE McVoItiov a&XXlgo; (H 306-8). a&rilx' ipa a?pzpOtvto; 'ApifiX6icou P&4 gnrjp6v

(20)

0; Ol p.?v xxaiovw?; Utpxato, voiat 8. Kipini ' tE' E?f3axzv Xapno6vX? cpaVcfi faAXavo6v nap' paAVXov

(x 241-42).

(a subdiviof 'integral' (1976:168)as an instance byKirk Ex. (19) is treated is Thispassage ina discussion ofHI306-50. sionof'necessary') enjambement of viewof the thepoint from complex to be extremely considered by Kirk usedtoseeing when onegets However, and"sentence."43 ofverse interaction L- andR-dislocations, concomitant andfull ofthe as fragmented style Homer's In 1.307-8inany case(= ex. [19]), is notso exceptional anymore. the passage the verse-end, before a theme constituent is actually enjambement thealleged link disthe preceding with the (ip no;) provides extension whose predicative is nota constituent (1988).Herethetheme Ex. (20) is cited byRuijgh course. the which the within domain rather, switch; noun indicating topic phrase simple betweentwo clause is meantto make sense is a relation following in clause.Thusthe'enjambement' which playa rolein this persons/entities ideaunits: two tofall between happens endwhich toa verse down (20) boils
(21) a. Andto them Circe, b. acorns she threw.

limited verse is not endofthe the before ofL-dislocation Thephenomenon or verseis an enclitic thesecondwordof thefollowing to cases where caseof(14) and(15),the We may justas inthe saythat particle. postpositive to markthe L-dislocation. a has not been added merely particlea&p
43 'This passage, in which internalstops and integralenjambements and the of verse-endare rife, in which the simplicityof the verse as the overrunning (however one happens to unit is suppressedor transcended rhythmical primary passages in which of heavilycumulative look at it), lies at the oppositeextreme one verse leads to the next either with a new sentence or with progressive (1976: 168). enjambement."

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

andEnjambement Homeric Discourse

15

Here is absent. too whentheparticle there maybe L-dislocation Consequently, from Kirk'spassagerl 306-50 (cf.also ex. (2) above):44 aresomeexamples (22)
(DivWlSi;

E90p1 6pW6pcVO;

pvtv &vOpcnov n'Xvrai. (HI 313-15, cf. 321-22).


(23) jptne & irpona5poi0. Mapt; '
a'roaoXE&

8' "AuplkxXoviqpoppn&vrna Solcciaa; npVFvOV Oa?cio;, 'EvOa idiXtaro;

'AvcrtX6lX 'OVpovac xaoslyvyrjolo xXoOcki. (H 319-20).


V{V' Thi~ow

Soupi

(24)

tlcli; ioao Mnlptovr;8' 'Axac6Lavra a43ra6peVOV ia

KaXaXi4oItO 8c410v dgov. (H

342-43).

a to Kirk)is in reality ('integral' according In all three cases theenjambement at theendoftheline.In (23) thecomplexity constituent ending complextheme ("AndMaris,nearby descriptive detailis addedto thetheme lies in thefactthat L-dislocated with hisspear").In theother twocases,(22) and(24), thecomplex by meansof a partwowarriors at once and specifies, constituent introduces ticipialphrase,the relationbetweenthem.This createsa suitablecontext and nominal reference clause can do without thesubsequent ('frame'),so that a pairofidea Thisyields can leavetheobjectunexpressed (see ex. (13) above).45 in eachcase: units
(25) (26)

b. he hithimfirst highin his thigh.46

of Amphiclus dodgingthe onrush, a. And Phyleides,

Acamas withswiftfeet, a. And Meriones,overtaking shoulder as he mounted his chariot. b. he struck him in the right

of thelineseemsto belong at thebeginning Whena nameor nounphrase we haveto consider analogously R-dislocation, to theclause in thelinebefore, at theend of a line. in terms of nounphrases of L-dislocation to thetreatment in the formof the Often,thereis an overt index of this phenomenon, contains consideration o see (16) above).Thepassageunder 'redundant' pronoun twoinstances:
44 In all, thereare 12 cases of L-dislocation'enjambing'into the next line in Kirk's passage (11. 307, 313, 319, 321, 323, 326, 330, 333, 337, 342, 345, in the passage 349). This means that Kirk's figureof 'integral' enjambement the total. from are subtracted 22 to 10 whenthe L-dislocation-cases dropsfrom to note thatthisrelationbetweenthe complexthemeand the It is interesting subsequent clause (each theme specifyingthe relation between a new pair of whichBeye (1964) discovthe discoursestructure resembles in the flght) warriors ered to be the common property of both epic catalogue entries and of fact(the A-part, &v8poKxacsiat,the so-called 'ABC-scheme': a bare statement deis followedby a further e.g. a killing,i.e. a relationbetweentwo warriors) (C); the two partsare oftenseparatedby an anecdote(B) about one of scription in A. In the examplesunderstudythe theme,expressing the persons mentioned in the list, resembles the A-part and the clausal element the 'killing-entry' whichprovidesmore detail,the C-part. prp,per, 4 Notice that the second 'enjambement'in (22), equally 'integral' in Kirk's the "whereit is thickest, can be analyzed as case of R-dislocation: classification, muscleof a man" (xtnetat in 1. 315 is redundant).

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

16
(27)

EgbertJ. Bakker
NEaropi&at 8' 6 )1?v ol'Wraa''AM'vtov 64ei 8oupi ? X6cEov iEyXo;. (H 317-8). 'Av?ikoXo;, Xwanplr; 8i? t8u XaaE

(28)

H1v?Xci;. (H 339-40).

6 8' sn' oZvato;

a{uxva

Ocive

constituent Neoopt6at 6' ("Andas lineopenswiththetheme In (27) thefirst lines;47 ofthefollowing thetwovictors which forthesonsofNestor") specifies theexploitof one of thetwo ("theone, he by a unitnaming thisis followed to which is added his name, as an explicatory wounded Atymnius"),48 in on thesceneas a participant present In (28), Peneleosis already afterthought on 6 6' marks participant to theother a switch first, so that Lyconis described is no morethanan position thescene,whose namein 1. 340 in R-dislocated highly fortheversification, its function thatis, apartfrom optionalreminder oforaldiscourse.49 characteristic in (27) and (28) to Kirk'stable(1976: 177) the'enjambement' According no doubton resp.),50 ('periodic'and 'progressive' types belongsto themilder or abthepresence However, of o in theclause before. of thepresence account forthe'degree'ofenjambeelement does notmakeanydifference senceof this of thenextline maybe Ras thename at thebeginning mentin question, in Kirk'stypology which examples, Consider thefollowing anyway. dislocated type): (viz. the'integral' cases ofenjambement wouldbe 'harsher'
(29) oiUvuca x6v Xpu0sa-v ilT4Lao?v

the action (cf. 1. 335

rvexe

6i A1aK0V 'r

muv?6pagov).

The action of

(A 11-12). Axpetn8TI.
"EKxTop H1ptagii8t1;,
o5nr o'

apTItrlpa

(30)

i'vOa tiva npCorov, tiva 8' iusratov


ZiU;

icUio;

i4?Vapt4?V E8(oicev; (A 299-300).

the and thiscan be easily deduced from whomApollo's angeris directed, thepoetadds 'Arp,6t8, by wayof contextButtoavoidall misunderstanding, itis perfectly clearwho The sameappliesto (30). In thecontext R-dislocation. at thispointis solelyconcerned is thesubjectof 4eva6pt4v, as thenarrative butthepoetadds Hector'sname,in of Hector; of theglory withthebeginning of this On account R-dislocation. redundant ofa typically oral,slightly theform of of theverseis an elucidation in whichthenameat thebeginning analysis, not to speak of the clause before,not a part of it, it is highlypreferable
47 Notice thatin traditional would be describedas the thisconstituent grammar, sub u ect, followedby two partialappositions. Greek,viz. to it has also in post-Homeric Note thatp?v has here the function signal thatmore is to come: closure (and hence coherence)of the chunkof dis&? is reached. only whena following course in questionis attained 49 Other, WunrOpovat I a&r,ap 6opL;Vt VTp&OI napijcvo; examples: equivalent nHxfo; viA;, r68a; @Uic; 'AxtX4I;'q (A 488-89), 6 6' aija npO6tepo; stoyevh; Ka uipcov / ijpw; HpxcrXXo; a'piluo; (B 707-8). p types in any 50 I cannot see why (27) and (28) should belong to different typology.

In (29), in uttering illigtocaev, the poet has Agamemnon in mind, the king to

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

andEnjambement Homeric Discourse

17

coincideswiththeend of an as theverseend neatly anymore, 'enjambement' idea uniLt. ofconstituents andtheorder 6. Versification of discussion of a class of instances I concludethisarticlewitha short of whichhas notso muchto be discussedin terms enjambement' 'necessary the of versification, as in terms oral discourse and fragmented L/R-dislocation hexameter. spaceofthedactylic inthemetrical andphrases distribution ofwords of fragmentation Homer'sstylein terms Up to thispointI havediscussed was thatthe of oral discourse.The leading thought as the crucial feature does notyetknowhowhe is ofdiscourse, a stretch poet,in beginning Homeric prodthan a fixed process rather is a dynamic he produces to endit,as thetext ofHomeric as one ofthecharacteristics 'cumulation' uct Thisis whatunderlies is 'integrative': style theHomeric Butthere is also a realsensein which poetry. is often determined an utterance thewaythepoetbegins in an important respect, the force hereis meter: by whatwill have to be theend of it.The dominating narrative upon the preferred exigenciesof the verse oftenput constraints in that in a unithas to be reversed just to theorder of constituents expression, ofcharacteristic a number Thisyields intothehexameter. fit a givenexpression example: Consider thefollowing cases of 'necessary enjambement'.
(31) " tov p' uio; TcXapcovo;usn'otaro; y-i axpiCp & ii (N 177-78). C?Ona0ev iyxo;. VA4%,

hereis similarto ex. (24) above in that At first sight, theenjambement there is in thesecondline (vi''). However, bothhave an enjambing predicate In (24), vi4' openswhatmaybe called a clause that difference. an important hand,vt,6' In (31), on theother of itstheme. thecontext makessense within stands quitealone,becausea new clausalunit(tic 6' 'r"acnEv EY7o;) immedibe calleda thematic theforegoing cannot expression atelyfollows.Moreover, elements: it of L-dislocated because it lacks thecrucialproperty constituent, in thediscourse.In fact,withits does not introduce (a) new participant(s) as object, which refers backto1. 171,itmaybe treated pronoun (IrOv) anaphoric as the C-partof Beye's (1964) ABC-scheme(see also note 45), in which the ofopening is expressed abouta killing.51 Thus,instead information specifiic of thepreceding in (31) is an essential part a theme, clausalunit following v-Ut' clause,viz. itsverb. to notice thatif vUt' had be placed beforethe Now it is important it (as in *6ov p' i6o; modifiers{ns' oukato; iSyXFit J.acpwinsteadof after aC?i iKpC), it would have been easy to TrXa.FWovo;v t?v, {nr' odaro; idea units: ofthree as a succession conceive of theexpression
(32) a. Him the son of Telamon stabbed, b. undertheear, c. withhis long spear.

51 See also Visser (1987: 48-49 forthisdiscoursestructure. The ABC-schemeis part of Visser's typologyof 'killing scenes,' the discourse type to which he limitshis valuable discussionof Homericversification.

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

18

J.Bakker Egbert

And ifviu'' had stoodat theendof theverseand inc' oauaao; EYEt gaKpy) at an wouldhavebeen,in sentential of thenextverse,there terms, thebeginning However, thisis notthecase and enjambemenL case of unperiodic unequivocal vit' is separated from itsclause bythetwomodifiers. The reasonforthisis to butof theverse. idea units, notof thecognitive be foundin therealization, to the referring halfof theversewiththeterms After havingfilledthefirst with proceedin linearfashion thepoetmayin principle victim and thevictor, notto do so, becausehe has a ready thepredicate of theclause,buthe prefers areplaced meaning "with thespear" adonicsequences waytofillthesecond part: frequently after thebucolic andtheremaining spacebetween metrical diaeresis,52 filled thetrochaic caesura andthebucolicdiaeresis (- - -) canbe conveniently of As a consequence was wounded. wherethevictim by a phrasespecifying Its positionat the these localizations,the verb vwu' has to be postponed. intended by specifically anyemphasis of theversedoes notindicate beginning examples The following reasons. for metrical purely thepoet:vig' is dislocated aresimilar.
(33) ?ov & IOi; v{4', (34) o'6?
pL?,paFa

ausp6vrn U6pv o

pLvrnxtov 64E; Bovp\ oXOc , Xakxo5ppta.

(A 95-6). ; . (A 424-25).

8oupi cxa?a np0xplltv Am' &Goit80; 6gp9aXOoE - O68' ?v iicovi(ot i?owV ?X? yaiav ayou v{ aicolt

(35)

? ip?poie auiv telotv OpOnX0EvTZ; j Mbow' -a5ap?aoo; 8'? Poi yt?vEr' &Ot ipo. (A 49-50).

in (31): adonic to that The secondhalfoftheversein A 95 (ex. [33]) is similar theplace of specifying bya modifier for"withthespear"preceded expression in thefirst halfof the In (34), thespearhas already beenmentioned wounding. specifying phrase bya prepositional thesecondhalfcan be filled verse,so that In (35), finally, theparticipial phraseoiv TEciXEtv theplace of wounding. thepost-caesural filling phrase a ready-made 0pi1x0Ev-rF; is to suchan extent line.In all ahead to thefollowing itpushesthepredicate partof theversethat there is no room of versification is suchthat pattem three cases the(traditional) half oftheverse. inthesecond for thepredicate andconclusion 7. Summary basisforthediscussion I havetried to laya linguistic In theabove sections in so faras manytreatThis is necessary of orality. of Homeric stylein terms do notdiffer that subjectin terms ments of Homericstylestilldeal withtheir even when theydo of written fromthe treatment literature, significantly features that has many Homer Of course, approach. subscribe to theoralpoetry to the call fora 'literate'approach;but thisshouldnot make us insensitive
52 Beside iEycxipxaxcpp and thereare Sovpi pa?v(, 6 '?i Soxp\, vik?i XakKX see Visser xaAX4. For the role of these expressionsin the versification (fc.). (1987: 81-2) and especiallyBakker & Fabbricotti

64ti

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

and Homeric Discourse Enjambement

19

or perhaps even orality, textreflecting evidence in theHomeric massive pointing toanoralorigin. the concept the lines ofChafe, along inabandoning, basis consists Theoral in favor in terms of 'idea units' reflecting the of an approach of 'sentence' ofa series of doesnot consist Anoralstory narrator. processes ofthe cognitive best as tothe judgments bythe narrator's is determined sentences whose length of thenarrator's consists an oral story Rather, of thestory. presentation consists (oneat the story information ofwhich ofthe onsmall pieces focussing Of ofideaunits. ina series ofconsciousness' these 'foci a time) andvocalizing because ofthe complicated, inHomer dealmore process is a great this course, applies. principle basically I believe that the verse, but posed bythe exigencies for the discussionof This approachyieldsa suitableframework in sentential has tobe called syntax instances ofwhat Many 'enjambement'. atall, tobe noenjambement approach cognitive turn out inthe 'enjambement' inideaunits: whenever oforal discourse organization fragmented because ofthe it two such units, between shown tofall theendofthe verse canbe reasonably in however, that tosay, Thisis not ofenjambement. not tospeak is preferable ceasestobe enjambeevery caseof'enjambement' the oral-cognitive approach constituents between endfalling remain cases of verse There simply ment. The cites such cases). which a strong cohesion obtains [1966] (Edwards between cases makethose I haveadvocated cannot which oral linguistic approach intheappropriate in which light, butitdoesallowus tosee them disappear, than did.Thisis another wayofsaying remarkable they still more they appear before we andoriginality, Homer's we cannot genius that appreciate properly hadtostart which this genius from the oralbasis have appreciated

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

20

Egbert J.Bakker

References Allwood, J., Anderson, L.-G., Dahl, 0. 1977. Logic in Linguistics, C.U.P. Cambridge: and Hiatus in Early Greek Epic: Bakker,E. J. 1988. "Long Diphthongs Mnem.40, 1-25. and theRole ofEpic Diction," Phonology and IndirectDiscourse. Temporal . Forthcoming. "NarrativeStructure in Journalof ShortStory,"forthcoming Subclauses in a Herodotean Pragmatics. Linguistic . In prep. A. "HomericPoetry:fromCognitionto Rhetoric. Iliad." inthe for Evidence Orality in AncientGreek.Ak as a Discourse. In prep. B. "Clause Combining Particle." organizational "Peripheraland Nuclear F. Forthcoming. Bakker, E. J. & Fabbricotti, Semanticsin Homeric Diction: the Case of Dative Expressionsfor inMnemosyne. 'Spear'," forthcoming TAPA 109, 1-10. and Oral Composition," Barnes,H. R. 1979. "Enjambement Revisited: Syntactic and Subordination Beaman, K. 1984. "Coordination in: D. Tannen Discourse," Narrative in Spokenand Written Complexity N.J:Ablex, Norwood, Discourse, in Spoken and Written (ed.), Coherence pp. 45-80. HSCP 68, 345andCatalogues," Narrative Battle Beye,C. R. 1964."Homeric 73. in theProduction of a of Consciousness Chafe,W. L. 1980. "The Deployment Cultural in: W. L. Chafe(ed.), The Pear Stories.Cognitive, Narrative," N.J:Ablex, Norwood, Production, AspectsofNarrative and Linguistic pp.9-50. in Speaking,Writing, and Oral . 1982. "Integration and Involvement Language.Exploring and Written (ed.), Spoken in: D. Tannen Literature," NJ:Ablex,pp. 35-53. andLiteracy, Norwood, Orality between Speaking . 1985. "Linguistic by Differences Produced Differences in: D. R. Olson, N. Torrance& A. Hildyard(eds.), and Writing," Language and Learning.The Nature and Consequencesof Literacy, C.U.P., pp. 105-23. Cambridge: Readingand Writing, . 1987. "CognitiveConstraints Flow," in: R. S. Tomlin on Information Amsterdam: inDiscourse, Benjamins. and Grounding (ed.),Coherence in: J.Haimanand S. . 1989. "Linking Unitsin SpokenEnglish," Intonation and Discourse, in Grammar A. Thompson(eds.), Clause Combining 21-57. Benjamins, Amsterdam: Pais: Klincksieck. ii Syntaxe, P. 1953. Grammaire homerique, Chantraine, in GreekHexameter T. 1977."Enjambement D. L. & Van Nortwick, Clayman, TAPA 107,85-92. Poetry," Foris. Dordrecht Grammar, Dik, S. C. 1978.Functional Edwards,G. P. 1971. The Language of Hesiod in its TraditionalContext, Blackwell. Oxford:

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Homeric Discourse andEnjambement

21

of Homeric Craftmanship," TAPA97, Edwards,M. W. 1966. "Some Features 115-79. in Englishdiscourse, Ph.D. Dissertation, Geluykens, R. 1987.Left-dislocation Univ.ofAntwerp. in Discourse: An Introduction," in: T. Givon, T. 1983. "Topic Continuity in Discourse. A QuantitativeCrossGivon, ed., Topic Continuity Benjamins, pp. 1-41. Lanouage Study, Amsterdam: Kirk,G. S. 1966. "Studiesin Some TechnicalAspectsof HomericStyle ii. in Homer,"YCS 20, 105-51 Verse Structure and SentenceStructure Cambridge 1976,pp. 146(reprinted in: Homerand theOral Tradition, aretothis 82; references text). Mass: Harvard Univ.Press. Lord,A. B. 1960.TheSinger ofTales,Cambridge, thatLinguistsHave Always Wantedto McCawley,J. D. 1981. Everything to Ask,Chicago:Univ. of Chigaco Know aboutLogic, butwereAfraid Press. Miller, G. A. 1956. "The Magical NumberSeven, Plus or Minus Two," Psychological Review63, 81-97. in Homeric Character of Enjambement Parry,M. 1929. "The Distinctive in A. Parry[ed.],The Makingof Verse,"TAPA 60, 200-220 (reprinted 1971]pp. PapersofMilman Parry [Oxford Homeric Verse.TheCollected 251-65). Albany: StateUniv.ofNew YorkPress. Peabody, B. 1975. The Winged Word, Ruijgh,C. J. 1971. Autourde rE e'pique.Etudes sur la syntaxegrecque, Amsterdam: Hakkert. . 1979.Reviewof Ch. H. Kahn,The Verb'Be' in Ancient Greek, Lingua48, 43-83. . 1988."La place des enclitiques d' apres dansl'ordre des mots chez Homere la loi de Wackernagel," paperread at theErnstRisch Colloquiumin Basle. TAPA94, 235-47. Formulas," Russo,J.A. 1963. "A CloserLook at Homeric C.U.P. D. 1987.DiscourseMarkers, Cambridge: Schiffrin, Versucheiner RekonVisser,E. 1987. HomerischeVersifikationstechnik. Peter Frankfurt: Lang. struktion,

This content downloaded from 41.226.66.100 on Sun, 28 Jul 2013 00:03:58 AM All use subject to JSTOR Terms and Conditions

Вам также может понравиться