Recognition Scenes

Formal Analysis of Recognition Scenes in the "Odyssey" Author(s): Peter Gainsford Source: The Journal of Hellenic Studies, Vol. 123 (2003), pp. 41-59 Published by: The Society for the Promotion of Hellenic Studies Stable URL: http://www.jstor.org/stable/3246259 . Accessed: 03/10/2013 08:59
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 [email protected].

.

The Society for the Promotion of Hellenic Studies is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Hellenic Studies.

http://www.jstor.org

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

Journal of Hellenic Studies 123 (2003) 41-59

SCENES IN THE ODYSSEY FORMALANALYSIS OF RECOGNITION
Abstract: Type-sceneshave been studied and analysed for over seventy years. This paperpresents a more detailed with the aim of moving towards analysis of one type-scene,the 'recognitionscene', thanhas previouslybeen attempted, of of of the recognitionscene is disthe The structure a better-structured 'syntax' type-scenes generally. understanding sected into motifs and 'moves', all of which are tabulated;this is the core of the analysis. The ensuing points of clarification elaborateon the definitions and assumptionsbuilt into the analysis. Following this is an assessment of the 'syntax' and quantifiableelements of the recognition scene. The discussion closes with a general assessment of the more literaryface of recognitionscenes, discussing them in the context of the plan of the second half of the Odyssey.

THEsecond half of the Odyssey features a set of type-scenes that I term recognition scenes, which are representednowhere else in Archaicpoetry. There are fifteen such scenes throughout Od. 13-24. They are defined by a common set of motifs in a fairly stable sequence, and therefore fall into the category of 'typical scenes', describedoriginally by Arendtand explored more recently by Lord, Edwards,Reece and others.' This article presents the most detailed formal analysis to date. The sheer numberof examples of this type-scene, and the complexity of its formalstructure, make this an ideal case for examiningthe 'syntax' of type-scenes. By syntax I mean regularities in the use of motifs in a type-scene formatand in how they relate to one another. For in these fifteen examples there emerge certaintrendsthat would not be obvious to the casual observer: trends in, for example, how the sequence of motifs may be shuffled;how the recognitionscene relates to other kinds of type-scenes; and how motifs, strings of motifs, or even entire scenes, may be reduplicated. After a while the accumulationof these regularitiesbegins to look somerules, thoughfar less strictthanthatof an everydaylanguage:thus thing like a set of grammatical the term 'syntax'.2 for furtherdevelopThe specific findings presentedhere may also be seen as programmatic ment of a set of narratological tools and vocabularyfor studyingthe mechanicsof Homericnarrativeand 'composition-by-theme'. ('Theme' referssimply to any regularcombinationof motifs in oral poetry, following Lord's usage; by extension, it has come to be used more generally of poetry, such as type-scenes in Homeric epic.3) This approach, any patternsin oral-traditional formal analysis, complementsan approachsuch as thatof Minchin,who focuses on emphasizing the type-scene as an aspect of performanceand of the performingpoet's skill.4 The semantics of the recognitionscene are one importantaspect of its 'grammar'. A formal structurethat has specific culturaland literarysignificance is of obvious importanceboth for 'oral theory' (in the sense of the study of 'thematic'composition) and for the semiotics of the Odyssey. However, given the technical detail of this analysis, the literarysignificance of the type-scene in each of its instantiationscannotbe adequatelydiscussed here; this is not the place to deliver a detailed commentaryon half a Homeric epic.
I Arendt(1933); Lord(1960); Fenik (1968); Krischer (1971); Edwards(1980), (1992); Reece (1993); Minchin (2001) 32-72. (I omit Fenik (1974), as he says less there about type-scenes.) 2 This use of the word 'syntax' may be comparedto the notion of 'story grammars',an idea derived from the hypothesis in some anthropologicalschools of thought that generalized tale-types underlie a particular story. Zumthor(1983) 125-44 also speaksof 'une grammairede la po~sie orale' and makes an equationof oral style to formulaic style. However, the idea of 'story grammars'is perhaps most notable for the criticisms it has attracted: see e.g. Garnham(1983), Black and Wilensky (1984), Rosenberg (1991) 99-100, 256-67, Smith (1978) 193-4 (cited by Rosenberg). 3 Lord (1960). Note that for many scholars 'typescenes' or 'typical scenes' are understood as repeated sequences of everyday actions, which is clearly not the case here: Odysseus' reunion with his family is by no means routine. In my usage 'type-scene' refers to any repeated set of motifs, especially where framing and sequence are distinctive features. 4 Minchin (2001) 32-72.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

42

PETERGAINSFORD

Instead,this paperwill first delineatethe formal structureof this type-scene, proceed to disconstraintsandpossibilities revealedin the variouspermutations cuss in detail the 'grammatical' of the structure and motifs of the type-scene, and concludeby assessing in generaltermsthe thematic use and significance of this type-scene in the plan of the second half of the Odyssey. Immediatelybelow in part 1, Formal structure,I first present a list of the componentsof the recognitionscene; this is followed by a complete tabulationof the Odyssey'sfifteen recognition scenes, giving line-numbersfor each motif. Next comes part 2, Points of clarification, which explains issues in the previoussection and the assumptionsunderlyingthe dissection of the typescene into motifs. Part3, Analysis, examines in detail the morphologicalconstraintsand possiof the recognitionscene - its 'grammar'. Finally in bilities shown in the various permutations in the part 4, Recognitionscenes plan of Odyssey 13-24, I assess the use of this type-scene in more literaryterms. 1. FORMALSTRUCTURE I label these type-scenes 'recognition' scenes not because all of them feature a recognition of Odysseus by a memberof his family, but because whereverthere is such a recognition,it is narratedin one of these scenes.5 Recognition, in other words, is the most prototypicalfunction of this type-scene, but not a sine qua non. This peculiarityof terminologyis indicativeof the fact that I am sidelining recognitionas such - as a literaryand culturalphenomenonin the Odyssey - in favourof a more technical discussion.6 Withineach recognitionscene thereare fourpossible moves,andwithin each move a sequence of motifs. Analysis of type-scenes into motifs follows the traditionof, for example, Bernard Fenik's analysis of battle scenes in the Iliad. To pick out a higher level of structurewithin the type-scene (the 'move') is closer to Steve Reece's analysisof hospitalityscenes, which has some hierarchization of motifs; and indeed the term 'move' is taken from VladimirPropp, in whose a tale can consist of multiplemoves, which may interrupt one anothmodel of folktalenarratives er or follow in series; but withinthe move a regularsequenceof motifs is normallyfollowed.7 The four possible moves in the recognitionscene are: (1) TESTING: Odysseus tests the addressee's loyalty.8 Functionally this is not simply the addressee'sloyalty to Odysseus personally;it is the addressee'scommitmentto the integrity of the oikos. (2) DECEPTION: Odysseus deceives the addressee. He disguises his identityand tells a false travels.9 about his story (3) FORETELLING: Odysseus, disguised, foretells the returnof the real Odysseus. RECOGNITION: (4) Odysseus reveals his true identity and a full reunion takes place. The moves are a multiformof one another,and thereforeno scene and RECOGNITION FORETELLING
and/or DECEPTION, features both of them: after a TESTING Odysseus may either foretell his

returnor reveal himself. Consequently,in accordancewith a variationon the principle of economy, the two moves sharemany of their motifs.
5 The analysis of Emlyn-Jones(1998a) 131-4 is a particularlyimportantprecursorto this analysis, and prefigoutlinedhere. ures many elements of the formalstructure 6 On recognition, reunion, and recognition scenes in the broader sense, see for example Kakridis (1971); Stewart (1976); Richardson(1983); Murnaghan(1987); Most (1989); Goldhill (1991) ch.1; Walter (1992); Henderson(1997); Gainsford(2001). See also EmlynJones (1998b) for furtherbibliographyand review. 7 Fenik (1968); Reece (1993); Propp(1968). Reece's summary of the hospitality scene pattern is given 6-7; discussion of motifs 12-39; analysis of all examples of the scene 207-31. 8 Cf stage 3 in Emlyn-Jones' analysis, 'Odysseus tests the other's loyalty; the test is passed (or, in the case of the suitorsand disloyal servants,failed)' (Emlyn-Jones (1998a) 131). 9 2 in Emlyn-Jones'analysis, '... Odysseus Cf. stage is pressedfor his identity,in reply to which he tells a false story ...' (Emlyn-Jones(1998a) 131).

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

43

In the summaryof recognition scene motifs below, T refers to a motif in the TESTING move, F is FORETELLING and R is RECOGNITION. Moves tend to overlap, but motifs D is DECEPTION, within moves tend to follow a regularsequence (anotherparallelwith Propp). The 'protagonist' is typically Odysseus himself, still disguised at the beginning of the scene (the 'Stranger'). Thereare exceptions, however:in scenes 12-13 it is Eurykleiawho breaksthe news of Odysseus' returnto Penelope, and in the 'two-way' recognition scenes (1, 7 and 14; see Points of clarification (g), below) both charactersact simultaneouslyas protagonistand addressee. TI: Unknown to the addressee,the protagonist- and the audience- observe one or more of the following: T1a: the addresseedisplays evidence of his/her loyalty (often by excellent performance in a hospitalityscene; cf motif II, below); Tlb: the addressee displays the detrimental effects of the Enemy and of Odysseus' absence; the addresseedisplays disbelief that Odysseus is still alive (cf F2, R3 below).10 Tic: T2: The protagonistdecides to test the addressee(cf Di).11 T3: The protagonistquestions the addresseewith a view to testing him/her. T4: The relationshipis shown to be intact, or the loyalty of the addresseeis revealed. DI: D2: D3: D4: The protagonistdecides to deceive the addressee(cf T2). The protagonistgives a false identity. He tells a false story of his travels and how he gained informationon them. He recalls meeting Odysseus.12

Fl: The protagonistforetells Odysseus' return. F2 (=R3): The addresseeexpresses disbelief.13 F3 (=R4): The addresseewishes it were true. F4 (=R5): The addresseeasserts that Odysseus is dead. F5 (=R6): The protagonistis willing to swear an oath that Odysseus will return. F6: The addresseerefuses, rejectingthe Strangerand/orreiteratingdisbelief. Ri: The protagonist'sappearanceis enhancedby Athene, thus adding impact to his revelation; this often involves a bath. R2: The protagonistreveals him/herself. R3 (=F2): The addresseeexpresses disbelief. R4 (=F3): The addresseewishes it were true. R5 (=F4): The addresseeasserts that Odysseus is dead. R6 (=F5): The protagonistis willing to swear an oath thatOdysseus has returned. R7: The addresseerequestsevidence. R8: The protagonistgives evidence.14 R9: Joy and weeping at recognition.15
10

Cf stage 2 in Emlyn-Jones' analysis, '... The other

speaker refers frequently in conversation to Odysseus, usually introducingthe topic shortly after meeting him' (Emlyn-Jones(1998a) 131). 11Cf Emlyn-Jones(1998a) 132, 'a desire to provoke or upset is clearly in Odysseus' mindjust before his main conversationwith Penelope at 19.45-6'.
12 Cf stage 2 in Emlyn-Jones' analysis, '... a false

13 Cf stage 5 in Emlyn-Jones' analysis, 'The other refuses to believe' (Emlyn-Jones(1998a) 131). 14 Cf stage 6 in Emlyn-Jones' analysis, 'Odysseus gives a sign (ofi'ga) as a proof of identity'(Emlyn-Jones (1998a) 131). 15Cf stage 7 in Emlyn-Jones'analysis, 'Final recognition, accompanied by great emotion on both sides' (Emlyn-Jones(1998a) 132).

story in which he claims to have seen Odysseus on his
travels ...' (Emlyn-Jones (1998a) 131).

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

44

PETERGAINSFORD

In additionto the above, certainmotifs may be found in recognitionscenes which do not necessarily belong to a specific move: I: Boundarymarkingstartof scene. II: Hospitalityscene motifs; the motifs of a meal (IXa, b, c in Reece's analysis) and questioning (XIa) are often co-extensive with Ti (see underTl a, above).16 III: Boundarymarkingend of scene. When following motif R9, this often takes the form of a character'sweeping being interrupted (or potentiallyinterrupted).17 Table 1 is a tabulationof scenes and the motifs they contain,with line-references. Note that: (i) All the main motifs in the type-scene are listed in the order shown above, except that motifs I and II are at the top of the table and motif III at the bottom. (ii) Each scene is numberedand also designated by the name of the addressee: thus '6. Telemachos'. Certaincharactersappearin more than one scene (in most cases, this is when the disguised Odysseus foretells his return several times); these scenes are designated 'Eumaios#1', 'Eumaios#2', etc. (iii) The term 'inv.' after a name refers to the 'inverse' reading of a two-way recognition scene: for an explanationof this term see Points of clarification (g), following the tables. Briefly, a two-way scene is one in which the two charactersinvolved act simultaneouslyas protagonistand as addressee;thus in 'la. Athene', Athene is the addressee,but in 'lb. Athene inv.' she is the protagonist. 2. POINTS OF CLARIFICATION The following points are to clarify the above lists and the assumptionsunderlyingthe analysis. (a) Motifs I and III:framing move in the recogniAs noted above, motifs I and III are not an integralpart of any particular tion scene. I and III are concerned with the framing of the type-scene as a whole, discrete as a cognitive cue both for the performingpoet andthe audiepisode. Framingis often important ence, assisting each in creatinga conceptualframeworkfor their respective activities of narratthe narrative. For a discussion of framing techniques used, see ing and hearing/interpreting Analysis (b) below. Normally the four moves overlap to some extent; they are not generally framed as discrete from each other,though there is usually a sort of caesurabefore the FORETELLING/RECOGNITION move (see Analysis (d), below). However, within each move the orderof the motifs is generalsee Analysis (e)): this strongly suggests that moves have a ly regular(with some irregularities; as sequentialpatternsthat govern how a scene develops. their status of own, quasi-autonomous As noted above, this is parallelto Propp'suse of the term 'move'. (b) Motifll: hospitality Oftenanother scene', analysedin detailby Steve Reece (1993), interlocks type-scene,the 'hospitality with or overlapswith a recognitionscene. Such situationsare designatedhere very generallyas 'motif II'. In several such situationsthe hospitalityscene actuallyinteractswith the recognition as a host in a hospitalityscene scene and serves the functionofmotifTla: excellent performance shows a kind of integrity in the addressee's relationshipto the oikos and to its ethical responsibilities. In particular,the motif of giving the guest a meal (hospitality motif IX) can be
All references to hospitality scene motifs follow Reece's numeration.
16
17 8 in Emlyn-Jones'analysis, "'On to busiCf. stage ness"' (Emlyn-Jones(1998a) 132).

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

45

employed as a means of showing thatthe addresseeis fit to be a full memberof Odysseus' oikos (recognitionmotif T1a); however, that then permitsthe meal to be used in isolation from a hospitality scene, and it becomes a common featureof the recognitionscene as well, as in scenes 4 to the recognition scene from the hospitality and 5, motif II. It is, in other words, appropriated scene. The same thing happensfrequentlywith two otherhospitalityscene motifs, namely identificationof the guest (XI; cf recognitionmotif D2) and exchange of information(XII; cf recognition motifs D3-D4). It is clear that there is a great deal of overlap in the semantics of hospitality and recognitionscenes; but whereas hospitalityscenes appeareven outside Homer,recognition scenes are confined to the second half of the Odyssey.18 (c) Motif TI: observing the addressee Motif TI is concernedwith displayingthe addresseeas someone who still participatesin a firm, if dormant,relationshipwith Odysseus. The three sub-motifsT1a, T1b and Tic fulfil this funcof relationshipsand the reintion in differentways. They all signify the potentialfor restoration in other words. to create a character's be of the oikos: reintegrated, they eligibility tegration this these there are three distinctive of three subrepresentedby ways expressing Although motifs nevertheless,because they all serve essentially the same function,there is no set order in which they appearand they can be repeatedalmost without limit, as in scene 2. I have therefore designatedthem all as hierarchizedunderone motif, Ti. (See Analysis (c), below, for further discussion.) I have labelled some passages as motif T1a and T1b where an addresseeexpresses the motif on behalf of anothermemberof Odysseus' household:where, for example, a characterdescribes the detrimental effects of Odysseus' absence on anothercharacter.So scene 6 (Telemachos)contains exemplars of motif T1b on behalf of Penelope (16.126-7) and Laertes (137-53). In the of motifs T1a and T1b are shown with line numbersin tables above these vicarious appearances squarebrackets. (d) Moves and their independence and RECOGNITION are multiformsof one other,a single recognition Because FORETELLING scene can containno more thanthreeof the four moves. Indeed,only six scenes containthe maximum of three. Additionally,only a small proportionof scenes thatI have identifiedas examples of the type-scene do containa RECOGNITION.In other cases it is a misnomerto call this a 'recognition' scene, but I have continued to do so, partly out of respect for traditionand partly for convenience, but primarilybecause the idea of recognitionis always the teleological focus of the scene. The criterionfor distinguishingthe moves from each other is independence,which is to say, no move can requireanotheras a prerequisitefor its presence. A DECEPTION, for example, is usua 3 has scene a without a TESTING,and this DECEPTION TESTING; however, ally accompaniedby demonstratestheir independence. The FORETELLING and RECOGNITION moves, as noted above, are mutuallyexclusive, but this is only an apparentexception, as they are variantsof one another. A single move may even appearon its own: so scene 5 contains only a TESTING,and 12-13 contain only a RECOGNITION. A result of this independenceis that each move has a quasi-autonomous force (as suggested above, in (a)). Each is a miniaturetype-scene in its own right, with its own teleology, sequence, etc.; in practice, of course, they tend to agglomerate. In other words, the moves are collected togetherat a relatively low level in a cognitive hierarchyof scene patterns:the largerpatternof the 'recognitionscene' subsumes, or conversely emerges from, an agglomerationof patternsof testing, deception, etc.
18If more of the Cyclic epics survived,of course, this picturemight be very different.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

46
lb. la. Athene, Athene 13.253- inv., 13. 86 187-96, 221-371 Change of scene, 187b 2. Eumaios #1, 14. 1-173 Change of scene, 1-4

PETERGAINSFORD
3. 4. 5. 6. 7a. Eumaios Eumaios Eumaios Telemachos, Penelope 16. 1-220 #1, 19.44 #2, 14. #3, 14. #4, 15. 185-408 453-533 301-39 (or 96)251 Explicit change of topic, 185 Bridgescene (end of dinner), 453-6 453-6 Change of scene, 301-2 Bridgescene (day-break), 1-3 41-5, 49-55, 57-9, etc. 41-89, 126-7, 137-53 41-5, 57-9, 78-84 96-105 124-61 127-8, 137-56, [160-1] 7b. Penelope #1 inv.,19. 213-51 Formulaic, 213

Motif I

II T1 Tla

Tlb

Tic

Book 14, passim 3-108, 133-8a 3-14, 20-5, 33-58a, 80, 96-104 16b-19, 26-8, 40-3, 58b-67, 81-108, 137b-8a 44, 68-71, 89-90, 133-7a 189-96 256-86 287-99a 189-96 221b-5 237-49 115-17 138b-47

185-90

301-2

69-77, 85-9, 124-36, 154, 157-9 [126-7], [137-53]

T2 T3 T4 D1 D2 D3 D4 Fl F2 F3 F4 F5 F6 R1 R2 R3 R4 R5 R6 R7 R8 R9 III

253-5 256-86 256-86

459-61 503-6 508-19 459-61 199-234 235-359 321-30 331-3 363-71 366b-70 391-400 386-7, 401-6 468-82 468-502

303-6 307-24 326-39

44-6 91-8 112-17,147-9 62-4 65-6 (100b-1) 203-12 44-6 172-84 168b-70 185-202

215-17 218-19 221-48

120b 118-20a 149-64 166-71a 171b-3 151-60

299b-310 324-7

154-85 187-9 192-200

328 344-52 353-60 Bridge- Explicit scene, change of 361-71 topic, 185 Bridge- Bridge- Changeof scene scene speaker, 340 (dinner- (nightfall) 520-33 time), 407-8

202-12 213-19

215-19 221-48 249-50

Formulaic, Formulaic, Formulaic, 220 251 251

Table 1.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY
8. 9. 10. 12. 13. 11. Penelope Eurykleia, Philoitios, Philoitios Penelope Penelope #2, 19. 19. 317 (or 20.185- & Eumaios, #3, 23. #4, 23. 252-316 343)-507 239 21.188-229 1-38 39-84 Motif I Formulaic, 251-2 II TI Tla 253-60 253-7a 317-56 343-68 343-8 New Change character of scene, enters 185-8 188-90 190-4 190-225 190-200a, 209-10, 217-20a, 224-5 200b-23 207-10 193-4 476-90 492-502 195-8 200-4 85b-7, 107b-16 174-81a 181b-204 Change of scene, 1-4 14a. Penelope #5, 23. 85-116, 153-246 Change of scene, 85 14b. Penelope #5 inv., 23. 85-116, 153-246 Change of scene, 85 93-5

47
15. Laertes, 24.205360 Change of scene, 205-6 297-301 205-34, 281-96 205-12, 281-6

Tlb Tic T2 T3 T4 DI D2 D3 D4 F1 287-93 271-99 268-70, 300-2a, 305-7 312-13 309-11 314-15 302b-4 316 257b-60

361-8

226-34 289-96 235-41 242-314 315-17 235-41 303-6a 306b-8 265-79, 309-13a

232-4

F2 F3 F4 F5 F6 R1 R2 R3 R4 R5 R6 R7 R8 R9 III

236-9 227-31 386-8a 388b-94, 473-5 205-16 5-9 11-24 40-57 58-67 60-1 67-8 78-9 105-10, 174-81a 395-466 467-72 217-21 222-5 Change of Formulaic focus, change of 240 topic, 226-9 26-31 32-8 Bridgespeech, 80-4 70-9 113-16, 181b-204 205-40 209-30 231-40 328-9 331-44 345-8 153-63 96-103, 166-72 205-8 (365-82) 320-6

Change of topic 317

Bridgescene, 503-7

Formulaic, Formulaic, Bridge-, 241-6 241-6 speeches, 349-60

Table 1. continued

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

48

PETERGAINSFORD

To say thatscene patternsare linked and agglomerateis to say thatthey are connotative;there are strong semantic links, as well as formal links, between them. DECEPTION connotes TESTING - why does Odysseus deceive Eumaios,Penelope, Laertes? Thereare particular reasonsin each his for them none whereas reason is always the but applies universally; testing particularcase, there will tend to be a TESTING same, to test theirloyalty -, and so where there is a DECEPTION as well. The link works simultaneouslyat the semantic level (deception is an effective way of testing someone) and at the formallevel (both events will tend to be framedtogetherin a single scene, thus agglomeratingas a single type-scene). Conversely, it is because of the tendency of moves to agglomeratethat they connote each other.19This agglomerativetendency seems also to have broughtaboutthe stronglink between the recognitionscene and the hospitalityscene, though this link appearsto be primarilysemantic ratherthan formal. (e) Formal slippage due to semantic connotations This tendency for moves to connote one anothercreatesslippages between them: a certainmudor a motif dling of motifs. A motif can appeareven when its usual significance is inappropriate; can be omitted even when it is needed for the scene to make sense fully. An extreme is reached in scene 7, where hospitality scene motifs take the place of recognition motif Tla but do not actuallydemonstratePenelope's fidelity or her adherenceto norms of hospitality. She provides the Stranger(i.e. Odysseus disguised) with a seat; the incidentis elaborated(19.96-102, hospitalitymotif VIII);and then she questionshim abouthis identity(19.1045, hospitalitymotif XIa) using typical formulaethatprovide a strongformulaiclink with formal hospitalityscenes.20 Odysseus later gives a (false) identity and origins (XIb, 19.165-84 = recognitionmotifs D24), and otherhospitalitymotifs follow later(bath,offer of a bed, etc.). The above sequence does not in any obvious sense exemplify any moral rectitude through hospitality or fidelity to the oikos; ratherit implies, or signifies continuedmembershipof the oikos (motif Tla).21 Laertes' questioningof the Stranger(24.297-301),22 to which Odysseus again gives a lying answer(again motifs D2-4), serves the same function in the same way. Good performanceas a host in a hospitality scene is a formalizedsign of moral rectitude,just as bad performanceas a host also has wider ramifications(Polyphemos is the prime example); elements of the hospitalitytype-scene disseminateinto the recognitionscene, and vice versa. In a case such as the recognitionscene with Laertes(scene 15), the cruelty of the TESTING of Laertesmakes it seem thatevidence of the addressee'sloyalty is being presentednot for the sake of Odysseus but for the sake of the audience of the epic; or, perhapsone might say, for the sake of the formalized significative effect of the recognition scene.23 The addresseenever fails the
19Lord (1991) 54, similarly discusses deception and recognition as conjoint events. From comparativeevidence he makes the furtherobservationthat deception is atypically, athematically absent from the Nausikaa episode in Odyssey 6; and so, we should conclude, its absence is significant. Where deception is absent,so too are acknowledgementand reintegration:Odysseus will never set up oikos with Nausikaa. However, Homeric narrativediffers from the Slavic variantsthat Lord cites: there the Strangertypically reportsthe hero's death and burial,while in the Odysseythe hero's imminentreturnis consistently foretold. 20 19.105 = 1.170, 10.325, 14.187, 24.298, h. Cer. 113; and 19.105a = 7.238a. 19.105 also = 15.264, not in a formalhospitalityscene, but even thatpassage displays other hospitalitymotifs. 21The constructionof this scene is discussed in detail in Gainsford(2001). 22Note again the formula,24.298 = 19.105. 23 Book 24 is included in this analysis in spite of reservations among Homeric scholars concerning the 'authenticity'of that book. This is because the recognition scene with Laertesis, formallyspeaking,flawless. It with the forms thereforereflects a farbetteracquaintance of Homeric narrativestyle than that displayed by any poet or scholar since the beginnings of Homeric scholarthen, it is as 'authentic',in the ship. By formalstandards, sense of 'traditional',as any piece of Homeric narrative can ever be.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

49

within a formalrecognitionscene; therefore,in cases where there is no 'genuine' doubt TESTING of the addressee'sloyalty, such as that of Laertesor when an addresseeundergoesmultipletestis more symbolic than anythingelse: it acts as a confirmaings, the importanceof the TESTING tion of familial integrity, ratherthan a rebuilding of that integrity. This point is particularly move is overlaid by a formal hospitality scene, which may potenimportantwhen the TESTING even take the of the move altogether,as in the case of 19.96-105 discussed above. place tially A similar muddling might be expected to arise between two motifs within the recognition scene, Tlc and F4/R5, since both have the same semantic content: each represents the addressee'sexpression of a belief that Odysseus is dead. But as it turnsout, in all exemplarsof these motifs (scenes 2, 8 and 10 for Tic, and 3, 8 and 13 for F4/R5), it is trivial to determine from the context whether an expression of disbelief fits into the TESTING sequence or to the It FORETELLING/RECOGNITION would that the force of sequence. appear sequence - or, from the a of be a better word in the thematic structure of perspective performingpoet, 'cueing' might the narrativeis, in this instance, enough to pin an expressionof disbelief either to the one move or to the other,without ambiguity.

(f Isolated motifs appearing outside recognitionscenes I have omitted from the analysis the motif of a portentwhich is correctlyinterpreted as an omen of Odysseus' imminent returnor of the Suitors' deaths. This could be arguedto belong to the foretelling patternon the grounds that it often appearsin conjunctionwith motifs F2 and F4. and DECEPTION. More However, it always appearsin isolation from the othermoves of TESTING here and there the whole and occasionalimportantly, portentsappear throughout Odyssey only ly involve Odysseus himself, whereas the scenes that I have identified as examples of a formal 'recognition scene' occur only in the second half of the epic. I do not present a separateanalysis of scenes featuringthis motif, as it has no obvious connectionwith the action of reintegration of the family.24
(g) Two-wayrecognitionscenes The recognitionscene is bilateral. Each character needs to recognize, or ratheracknowledge,the other; and this acknowledgementmust be both a semantic or ideological statementand also a formal sign within the narrative language of Homeric epic. This bilaterality is especially betrayedby the existence of two-way recognition scenes: scenes 1, 7 and 14 below. In these scenes each characteris testing, deceiving and revealing him/herselfto the other, acting simultaneously as protagonistand addressee, in two recognition scenes superimposedon top of one another. The effect is reminiscentof a stretto in a fugue, where a theme overlapswith itself. In such a situationit is not simply a case of motifs from the two simultaneousscenes interlocking with each other,alternatingfrom one scene to the other;rather,passages serve as nucleic motifs in both scenes at once. A single event (e.g. the lying story of Odysseus in scene 1, 13.256-86) may serve two functions: as a deception motif in one sequence (DI, D2, D3: Odysseus deceiving Athene) and as a testing motif in the other (T3: Athene makes sure of Odysseus' continuingmetis by his attemptto deceive her).

24 Examplesare 2.146-208, 15.160-83, 15.525-38 and 17.150-65. Odysseus is involved with examples at 19.535-69 and 20.102-21. A slim connectionwith recognition scenes propercould be arguedin the formerof these as the passage is juxtaposed with scene 9 (Eurykleia, motifs 19.343-507). Other examples of FORETELLING

occurring in isolation include 1.415-16 (F4, F2), where Telemachosexpresses a lack of faith in divination;2.2814 (F1), whereAthene mentionsthe impendingdeathof the Suitors to Telemachos in passing; 2.361-6 (F4), where Eurykleiaexpressesa belief that Odysseusis dead.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

50

PETERGAINSFORD

The 'other' characterin these two-way scenes is female in the examples we have: Athene once, Penelope twice. Threeexemplarsare insufficientto tell whetheror not this is coincidence. In both cases the bilateralcharacterof the scene impartsa kind of ambiguityto the narrative:in Penelope's case this is certainlyrelatedto the ambiguityof genderroles and the indeterminacies her narrativegenerally.25In Athene's relationshipto Odysseus there is a different surrounding kind of ambiguity:whereas Odysseus is the keystone to his family's reintegration, Athene, as divine patron,is partof the key to his own recovery of his heroic identity. 3. ANALYSIS (a) Scene length Scenes range in length from 34 lines (scene la, Athene) to 224 lines (scene 3, Eumaios#2). The shortestscenes with full framingat the beginning and end (i.e. motifs I and III) are the 39-line scenes 5 (Eumaios#4) and 7b (Penelope #1 inv.); the shortestscene to featuremultiplemoves is the 42-line scene 11 (Philoitios and Eumaios), which is exceptionally tightly constructedand framed. The mean length of recognition scenes using the line-numbersin the tables above is 112.4 deviationof 69.1. These figureswill vary dependingon whetherinterludes lines, with a standard in for the fake partyand Odysseus' bath in scenes 14a/b;the scar-narrative the preparations (e.g. scene 9) and framingdevices (motifs I and III) are included. This is much shorterthanthe length of hospitalityscenes as analysedby Reece, where a scene may occupy a whole book or even multiple books;26but longer than, say, the typical arming scene analysedby Arendt,which varies between six and 35 lines.27 On the other hand,the armof an extendedaristeia, which is far longer.28These ing scene is generallybuilt into the structure variations,and the large standarddeviation even just in the case of recognition scenes, do not suggest that there is any particular 'standard' length for a type-scene; rather the reverse. However,it may be noted thatthe longest exemplarof a given type-scene never seems to exceed the shortestin length by a factor of more than ten. (b) Theframing of recognitionscenes Certaincharacteristic way of framingthe ways of framinga scene exist. The most characteristic a of is scene of a change setting (scenes ib, 2, 11, 12, 14a/b, 15). (motif I) recognition beginning of a new characteris also a typical occasion to begin a recogniThe appearanceor introduction tion scene (scenes 2, 6, 10, 15). Twice with Eumaiosa recognitionscene begins with a new con~ i pov Kcct8 versationaftera meal, with the formulaari&p inlEn oi O Io6rlto; (14.454 = wvro a with also formula,i18'r9ine oiv 15.303, scenes 4 and 5);29 twice scenes with Penelope begin = On one occasion it is sim7b and scenes 19.251, 8).30 (19.213 y6oto toh•o•aKpiotow tX&p(prl ply a change of topic in conversation(scene 3); on another,the very powerful conceptualbreak of a night provides the narrativebreak(scene 6).
25 See, for example, Foley (1978), Katz (1991), Felson-Rubin(1994). 26 The hospitality scene where Eumaios welcomes Odysseus occupies most of Od. 14; the hospitalityscene where Alkinoos and the Phaiakianswelcome him has its motifs spreadout over Books 6, 7, 8, 11 and 13. 27 In the Iliad Idomeneus' and Teukros' arming scenes are six lines each (13.240-5 and 15.478-83 respectively); Achilles' is 35 lines (19.364-98). The shortest fully developed armingscene is Paris'(11 lines, II. 3.32838). In the Odyssey there are brief quasi-armingscenes at 23.366-70 and 24.496-501, both involving several charactersarming themselves ratherthan a single individual. 28 Agamemnon's, Idomeneus', Patroklos' and Achilles' armingscenes all act as preludesto theirrespective aristeiai (II. 11.15-46, 13.240-5, 16.130-54, 19.36498); Teukros'appearsat the end of his aristeia. The arming scene in Od. 24 also has strong semantic links with the idea of aristeia. 29 'When they had put aside their desire for drinking and food ....' 30 'And when she had taken her fill of much-tearful lamenting....'

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

51

Scene-endingsalso follow typical patterns. The end of a conversationends a scene six times (scenes ib, 3, 4, 9, 10, 13); six times emotionalweeping ensuing from the conversationproduces the break (scenes 6, 7a/7b, 11, 14a/14b), with associated formulae:scenes 6 and 11 close with (flepX 6) 9og v K)'068)potSVototv oto, ISi (16.220-1 = 21.226-7);31 cf scenes 14a/bKai 8V ,L KC•i K' p Ei I vI) 068)potivoto p6Ovl o1)oKoo kog; 'H'6;g, , lg (23.241-2)32and scenes 7a/b, i1~8 8' Ri o0v rtcp~Orl 1toX81va)Kp'oto y6oto (19.251 = motif I in scenes 7a and 8, above).33 On other occasions it is a change of topic, usually with a change of speakeras well (scenes 2, 5, 15; the formula 16poEt, pil tot ppE~i oatt concludes scenes lb and 15, IEX6tVrO)V34 ttaxa eLr& 13.362 = 24.357); and once it is the powerful breakof nightfall (scene 4; notice also the parallels in the formula 16.220-1 = 21.226-7 e 23.241-2, noted above). (c) Reduplication: 'couplets'of scenes and multiplecycles of a move where two scenes appearpairedtogetherwith the same charScenes may appearin 'couplets',35 acters as protagonistand addressee. Examples of this are scenes 2-3 (Eumaios #2 - Eumaios #3), 7-8 (Penelope #1 - Penelope #2, including both 7a and 7b) and 12-13 (Penelope #3 Penelope #4, with Eurykleiaas protagonist). 'Couplets' of a scene have a kind of antistrophic effect, where one scene reinforcesthe other. The more idiosyncratic,even atypical, aspects of one will tend to be repeated in the second, as in scenes 2-3, where Eumaios repeats the same move. strikingclaim that beggars are liars in the context of the FORETELLING in Note especially the thirdcouplet, scenes 12-13, thatthe couplet as a whole is framedat the start (motif I, 23.1-4) and at the end (motif III, 23.80-4), but that there is no distinct framing between the two scenes. Conversely,in the couplet of scenes 2-3, both scenes are framedindividually,both at the beginning and end, and in between as well (14.1-4, 14.185, 14.407-8). The couplet of scenes 7-8 is not framedat the start,but is framedin the middle (19.251 doubles up as motif III in scenes 7a/b and as motif I in scene 8) and at the end of scene 8 (19.317), which incidentallyleads directly into the recognitionscene with Eurykleia,scene 9. Thereare reduplicationsnot only of scenes but also of moves within a single scene. To avoid confusion, I refer to such repetitions of a move as multiple cycles of that move: it would be wrong to assume a priori that the tendency to have pairs of moves is the same tendency as the reduplicationof whole recognitionscenes, though that idea is obviously very attractive. move in scene 6; two cycles of RECOGNITION in Examplesof thisare:two cycles of the TESTING and DECEPTION in scene 15. I tabulatethese in Table2, scene 14a;and two cycles of both TESTING below. The double cycle in scene 14a is explainedby the interludein the middle of the scene, where are made for the fake wedding which deceives the Suitors'families: there the first preparations move is more properlya foreshadowingof the actual RECOGNITION. cycle of the RECOGNITION Note also that in view of the existence of these cycles, it is unclearwhetherscenes 12 and 13 should properly be regardedas a couplet, or instead as two cycles of the RECOGNITION move within one framedscene. I have tabulatedthem as a couplet of two scenes, but the alternativeis perfectly viable. The unclaritytends to validate the idea that the tendencies to have couplets of scenes, and repeatedcycles of moves within scenes, are in fact the same tendency.

31 'And now the light of the sun would have set on them as they wept, but ....' 32 'And now the rosy-fingered Dawn would have appearedon them as they wept, but .... 33See n.30 above. 34 'Take heart; let these things not cause trouble to your thoughts.'

35This term is to be distinguishedfrom 'doublets'or 'doubling', wordswhich in Fenik's usage referto all multiple occurrences of a motif, and not solely to pairs. Fenik uses the terms to refer to reduplicationof characters, names, locations, etc. Examples in the Odyssey are the doublets of the maidservants, Eurykleia and Eurynome; or of the two farmsteads of Eumaios and Laertes;or of the two herdsmen,Eumaiosand Philoitios.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

52 Motif T1
T3

PETERGAINSFORD 6. Telemachos 2ndcycle 1stcycle 16.41-89 [126-7],
[137-53] 91-8

Motif
R1

14a. Penelope #5 2ndcycle 1stcycle
153-63

R2
R7 R8 R9

23.96-103
105-10 113-16 -

166-72
174-81a 181b-204 205-40

T4

112-17

147-9

Motif Tla Tlb Tic

15. Laertes 1st cycle 2nd cycle 281-6 24.205-12 226-34 289-96

Motif Dl D2 D3

15. Laertes 1st cycle 2nd cycle 24.235-41 303-6a 306b-8

Table 2.

D4

265-79

309-13a

Trying to pinpoint these cycles also raises issues over how to interpretthe very protracted and DECEPTION example of motif Ti in scene 2 (Eumaios #1). The sequence of the TESTING moves appearsas in Table 3: Table3

Table 3.

Tla: 3-14, 20-5, 33-58a, 96-104 Tlb: 16b-19, 26-8, 40-3, 58b-7, 81-108 Tic: 44, 68-71, 89-90

T3: 115-17

D4-D3: 118-20

Tlb: 137b-8a Tic: 133-7a T4: 138b-47

Unlike the situationin scenes 6 and 15, no definite cycles may be observed here. Even in scenes 6, 7a and 10 thereare multiple appearancesof the sub-motifsTla, b and c in no particular order. It thereforeappearsthat these sub-motifs can be repeatedalmost without limit. The need to validate the charreason that motif Ti is so protractedhere is surely the extraordinary acter of Eumaios, so as to give him a place in the narrative. He is being observed not only by Odysseusbut by the audiencetoo, who, being unfamiliarwith a figure so farremovedfrom heroic myth, need an especially elaborateintroductionto him. Philoitios is similarly subjectedto a observation:motif Ti occupies two thirdsof his recognitionscene (scene 10).36 fairly protracted (d) Overlappingand embeddingof motifs and moves Motifs can be co-extensive. In scene la, motifs D2 and D3 are co-extensive; in scenes ib, 4 and the appearance of hospitalityscene motifs 15, motifs T2 and D 1 coincide. Motif II, representing in the recognitionscene, is particularly proneto this: in scene 5 it coincides with motif I, in scene 10 with partof motif T1, and in scene 15 with partof motif T3. Hospitalityscenes also fill large proportionsof Books 14, 16 and 19 of the Odyssey,thus overlappingwith scenes 2, 3, 4, 6, 7a and 9 as well: the overlapis usually with motif I, the opening frameof the scene (scenes 2, 3 and 4), or TI, in which the addressee demonstratescontinued loyalty to Odysseus' household by showing good hospitalityto the protagonist(scenes 2, 6 and 9). move. There is a tendency for motifs to be embeddedin other motifs within the DECEPTION In scene 3, motif D4 is embeddedin D3; and in scenes 4 and 8, motif D3 is embeddedin D4. move (tabThis can occur in othermoves as well: in scene 6, in the second cycle of the TESTING ulated above), the motif T4 is embedded in TI; and in scene 15 the whole second cycle of the T1 motif is embeddedwithin motif T3.
36Comparealso the extraordinary introduction given to anothermarginalfigure, Theoklymenos,in 15.222-58.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

53

Moves can also be embeddedin other moves or overlap with them, but this is perhapsto be move (the exception is expected. It is notable thatthe containingmove is normallythe TESTING that is doing the containing). In scenes ib, 2, 4 and 7a, the scene 8, where it is the FORETELLING move is embedded in the TESTING DECEPTION move is move; in scene 6, a brief DECEPTION embedded within motif TI; in scene 9 a substantialRECOGNITION move is embedded in the and in scene 15, both cycles of the DECEPTION move are embeddedin the TESTING. In TESTING; scene 8, motif FI is used to framethe DECEPTION which is followed a FORETELLING move, by full move (includinga reiteration of F1). Similarlyin scene 7a, motif T2/D 1 is followed aftera break move (D2-D3-D4), and finally the close by a full exposition of T 1, then the rest of the DECEPTION of the TESTING move (T4). More strikingis when two moves are embeddedor co-extensive, but there is a sharpconceptual breakbetween them and the thirdmove. This occurs in scenes l b, 2 and 7a, where a case move embeddedin the TESTING of the DECEPTION move is followed by a clean breakbefore the FORETELLING/RECOGNITION move, in scene lb (13.299, a change of topic in mid-speech), scene 7 (14.148, a change of speaker),and scene 7a (19.212-15, a new conversationbeginningwith the formulail 8' rE•i oiVvtipq0'pI... = motif I in scenes 7b and 8, motif III in scenes 7a/b). The FORETELLING/RECOGNITION is normallymarkedoff by a breakof some kind fromwhathas that is not separatedfrom the previgone before. The only case of a FORETELLING/RECOGNITION ous moves is in scene 9, where, as noted above, the RECOGNITION is embeddedin the TESTING. This is the scene with Eurykleia,and this is not the only respect in which it is unusual.37 (e) Motifs out of sequence The tabulationof motifs shows that there are occasions where the sequence of motifs within a move is broken. This happenswith greaterfrequencythan one might predicton the basis of an analogy with Propp'smodel of 'wondertalefunctions': Proppinsists that the sequence of functions within a move, in his model, is virtuallyunbreakable of a move. In except by reduplication scenes there is rathermore variation, and FORETELLING recognition especially in the DECEPTION moves. Note that the variationof sequence in the RECOGNITION move is much smaller,which is that the FORETELLING and moves RECOGNITION are multiformsof one another. surprising,given In the TESTING move thereare no anomaliesthatcannotbe accountedfor by eitherthe embedof motifs others or the presence of two cycles of the TESTING within move. Unusual ding in the order of b c and within motif TI are not sequences actually anomalous,as pointed T1a, out earlier. These sub-motifs are denoted as sub-motifsprecisely because there seems to be no fixed sequence to them, and because it appearsthey can be replicatedalmost indefinitely:scene 2 is the main example of this seemingly limitless replication(tabulatedabove, separately). There is variationin the sequence of DECEPTION motifs in scenes 2 and 7a: in scene 2 the in is and scene 7a the is D4-D3, sequence sequence D1-D3-D2-D4. This variabilityseems to cormove respondto the relativelyfrequentoccurrenceof embeddingof motifs within the DECEPTION
move (scene 15). (see above on scenes 3, 4 and 8) and one case of a double cycle of the DECEPTION

In FORETELLING moves the sequence is less regular,following the sequence as tabulatedonly in scene 3, but not in scenes 2, 8 and 10. Motif F5 moves aroundwithin the sequence flexibly: in scene 2 it is embeddedwithin F1; in scene 8 it is preceded and followed by multiple occurrences of F 1; and in scene 10 it appearsbefore F1. It appearsthat F5 tends naturallyto appear eitherbefore or afterthe clusterof F2-F3-F4, as these motifs are spokenby the addressee,whereas F1 and F5 are spoken by the protagonist.38(In my tabulation,F5 is given its late placing because of the case of scene 3, and the parallelplacing of R6 in scene 13.)
37 E.g. the 71-line scar-narrative, which acts as an extraordinarily lengthymotif R8, Odysseus'evidenceof his identity.The samestorycomesto only five lines in scene 11.
38 I am indebtedto an anonymousreaderat JHS for pointing this out.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

54

PETERGAINSFORD

Within RECOGNITION moves it is difficult to tell if there is a similar tendency with the same = motif (motif R6 F5), as the motif appearsonly once, in scene 13; there the motif is embedded within motif R8. The sequence of motifs is anomalous elsewhere only in scene 9, where the orderis R1-R2-R8-R9-R2;that is, the only anomaly is the reduplicationof motif R2 at the end, before the resumptionof the TESTING move. 4. RECOGNITION SCENES IN THE PLAN OF ODYSSEY 13-24 (a) Odysseus'reunionwith hisfamily The recognition scene is all about reunionand is an integralpart of the nostos-narrative of the at it takes levels: is a nostos not about epic. Odysseus' place multiple just geographicalreturn, first to Ithaca,then to his house, then to his bedroom;nor is it confined to the political restoration that is enactedby the slaughterof the Suitorsand by Odysseus' resumptionof his status as 'king' of Ithaca. Even the final appeasementof Poseidon, of which Odysseus reminds us in 23.264-84, is not the centralobjective of the nostos-narrative. Recognitionrepresentsa furtheraspect of nostos, what we might call a 'familial'restoration: of Odysseus to his appropriate functionor role in the system of relationships that is, a restoration that make up his family. Not only does he have to be restoredas head of his oikos: an individual relationshipwith each member of his household must be restored,one by one, recognition scene by recognitionscene. The importanceof this 'functional'aspect of nostos is kept in view even in the first half of the epic. It is Kalypso's desire to usurpthe position of wife to Odysseus that keeps him from his nostos (1.15, 5.209-18). Nausikaawants to assimilate Odysseus into a newly constitutedoikos of theirown, with its own succession and genealogy,and with guest-gifts from the Phaiakianscoming dangerouslyclose to doubling up as a dowry. While Odysseus is absent from home he is nameless not only to the Phaiakians,but to his family too, as is suggested when he introduceshimself to the Cyclops: 'My motherand fatherand all my friends name me Nobody' (9.366-7). The consequenceof listeningto the Sirens,we are told, is to be deprived of one's nostos and one's family (12.41-5). In these situationsthe loss of family stands for the loss of nostos. And for Odysseus the regainingof family, and consequentachievementof nostos, is representedthrougha sequence of formalrecognitionscenes. All the various kinds of restorationcombined- geographical,political and familial - togeththat er constitutethe nostos-narrative and,to stretcha point, the Odysseyitself. The reintegration Odysseus seeks is reintegrationwith the household as a whole; he acts as a keystone to the integrity and continuity of the oikos. We thereforehave a bipolar relationship:on one side, Odysseus, and on the other,the entire family, as a corporateentity which depends on him as a source of genealogy, and as a symbol thatgives it its identity(it source of safety, as a patriarchal is not just any family, it is the family of Odysseus). Just as Odysseus' presence equates to a state of reintegration(for the oikos, for Odysseus' own heroic identity,and for the status of Telemachos,Penelope and others- that is to say, both for the family as a corporateentity and for individualfamily-memberstoo), so also his absence equatesto disintegration,disjunctionand disorder,again for all alike, and with all the concomiof genealogy, and so tant risks of invasion by outsiders,Suitors,the threatof the discontinuation on. While Odysseus is absent,the family might as well be any family: it is without a name,just as Odysseus is a hero withouta name. In this bipolarrelationshipof hero and family, Odysseus' absence contains within it the disjunctionof the family as a whole. Not just their disjunction from him, but their disjunctionper se: from each other and from their normativeroles. As long as Odysseus is absent,or not fully reunitedwith the oikos, Penelope will always be on the verge ofinfidelity; Telemachos'intermediatestatus between youth and adult hero will never be fully will not receive from his mastera home, a plot of land, resolved; Eumaios,more pragmatically,

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

55

or a wife (14.62-7, 21.213-15); and Laertes'grief and separationfrom the oikos (both geographical and political) will forever keep him marginalizedand liminal, 'on the thresholdof old age' (eiti yjpaog oi8&t, 15.348). '[F]or each of these figures the process of (mis)recognitionof Odysseus is different;and for each something differentdepends on Odysseus' return'(Goldhill (1991) 7). So as well as the one-to-manyreintegration(Odysseus and family) there is a cluster of oneto-one idiosyncratic, particularized relationships (Odysseus and Penelope, Odysseus and Telemachos,etc.). It is from the clusteringthatthe largernarrativeemerges. Each of these oneto-one relationships is expressed in recognition. Household-memberscannot make just any acknowledgement of the semantic, symbolic and ideological fact of Odysseus' return: it is encoded in recognition type-scenes, a formalized system, expressed through the medium of repeatedsequences of motifs. In this respect it is the same process in each figure. Everyonehas somethingdifferentat stake, yes, but simultaneouslyeveryone has something the same at stake. It is within the formal system that the differencesappear;and it is in the very type-nessof typescenes that the meaningfuldimension of reintegrationlies. So the formal system, the sameness of the type-scenes, exists alongside the individualcharof each individual's situation;it is not a case of individualcharacters acters, the particularities being fitted into a niche that stood vacant during Odysseus' absence, re-inserted into an autonomous,abstractsystem of relationships. The existence of two-way recognitionscenes, in tells us that it is a two-way, personal,relationshipthat is being restored. particular, Although the problems faced by each individual household-member, noted above (Telemachos'liminality and lack of assuranceof succession; Penelope's ambiguity;and so on), vary, the solutions to these problems are strikingly similar. Perhapssurprisingly,the solution implicatesthe statusof being a hero in one way or another. This is most explicit for Telemachos and Laertes,who standby Odysseus at the end of the epic, where Laertesrejoices to see his son and grandsoncompeting in arete (24.514-15); and it is Laerteswho makes the only kill in the battle of Book 24 and, uniquely in the Odyssey,his victory is honouredwith the formula8o6ntrqt (24.525 'he fell with a crash,andhis armsclattered oev &87tnea(ov, &p3d Ei•6a EEy' Fit' &w• For these two takes the form of coming as near as possicharacters,reintegration upon him'). ble to, or in Laertes'case returningto, a state of full hero-hood. It is a demarginalization: reinfrom the towards to the centre of means towards the norcoming away margins things, tegration mative position of being a hero. They have been marginalizedin a global sense by Odysseus' absence, and in a particular,individualized sense by, in Telemachos' case, this liminal state between youth and adulthood,and in Laertes'case by his age and his physical separationfrom the oikos. These two have somethingthe same at stake:one is a not-yet-hero,the other one is a vivid for them. Telemachosconstantlyneeds to fill once-was-hero,so the process is particularly his father's shoes, while Laertesis (cruelly?)provokedout of his retirement. For other figures the same holds true;if less graphically,it is because they do not need quite so much provocation,because there is less of a contest between them and Odysseus, or because they do not have the marginalizinggrief of old age. Eumaios and Philoitios are also in on the mnesterophonia,and have a place in one of the (heroic) armingscenes that lead to the battle of Book 24 (23.366-70). Penelope cannot fight in battles;but she, like Helen alone, has kleos and arete as a woman, and an army of Suitors. She is Odysseus' counterpart in every way, a parafor a non-masculine heroic ethos. She is the to digm neipicpov ('prudent') Odysseus' rohinrlitg ('cunning'), the memberof his household who tricks him in the recognitionscene of Book 23.39

39 Cf also the confusion of gender roles in similes: see Foley (1978). This is particularlyapplicable to the

simile in 23.233-40 which compares Penelope, rather than Odysseus, to a sailor escaping the sea.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

56

PETERGAINSFORD

(b) Eligibility to takepart in a recognitionscene Since the means of representingeach individualreintegration is the recognitiontype-scene, eliin take a scene an to is issue. Since the recognitionscene has gibility part recognition important to do with reunion, there will be none between, say, Odysseus and any of the Suitors:when Odysseus reveals himself to them, a very different set of symbols is needed. Members of his own family are the prime candidates,and by extension, membersof the oikos in a broadersense: the two herdsmen,Eurykleia,and even Athene. For this purposeAthene counts as a 'householdmember'inasmuchas she acts thematicallyas patronto the oikos, as well as the heroic patronof Odysseus, in the sense thatApollo is a patronfor Hektor:she plays the partof a domestic, rather thanheroic, deity.40 These four characters, along with otherswho have abbreviated recognitions in (the faithfulmaidservants 22.497-501; Dolios' family in 24.391-411), are not membersof the family per se but are the tendrils of a socio-culturaloikos that extends beyond the sanguinary relationshipsof the family. These charactersare both inside and outside the oikos. On the one hand, they are beyond genealogical succession, and the problemsposed for them by Odysseus' absence are not oikosthreateningones, but threaten them alone. Preservationof these charactersis incidental to Odysseus'nostos; they are expendable. (Notice how Melanthiosand the unfaithfulmaidservants are killed with no hint that their continued existence is of any importanceto the existence or stability of the oikos.) On the otherhand,they are assimilatedinto the sphereof family and succession, so thattheir own concerns are displaced by those of Odysseus' oikos. These thematic tendrilshave a tendency to fill gaps in the staff of Odysseus' household as needed, drawingin materialfrom outside and recontextualizingit; Eumaiosand Eurykleia,formernobles themselves who have been removed from their original familial context, partakemost strikingly in this characteristicof 'drawn-in-ness'. Eumaios, at 14.140-4, explicitly affirms that his membershipin Odysseus' oikos takes precedenceover any sense of belonging to the one he was born into. This drawingin is something that needs justification and ratification:it provides the occasion for Eumaios' telling his story. The dividing-linebetween the genealogicalfamily and those beyond that limen is particularly emphasizedwhen Odysseus makes the two herdsmen,in their conjunctrecognitionscene (scene 11), this promise (21.215-16):
K]Ca iot Etelrta

K rqpo reiaayvaotlCFo -e ~0oeov. TrieXeji6o... and then, as far as I am concerned, be the companionsand brothersof Telemachos. will you
'rdpo E KicaotY• Oe

are astonishingly strong words. With this promise Odysseus actually

makes them candidatesfor succession: he assigns them a place inside the sanguinarysystem of relationships. to the point conAside from assigning them a place in succession, these lines are particularly cerning the issue of eligibility for the recognitionscene. Eumaios and Philoitios, placed in the
40 Hoekstra (1989), commenting on Od. 13.221-5, argues, 'Athena's intervention ... is a reminiscence of Mycenaean times when ... she was the patron of the princes and their household goddess.' He draws the parallel of 7.80-1, whereAthene comes to Athens and enters the house of Erechtheus. Hainsworth (1988) 325 observes that the vocabulary there ('EpEXqifOg1uoctvbv 066'ov)suggests an association with the royal household of cult', thoughAthene's ratherthan with 'the apparatus is certainly relationship with Erechtheus/Erichthonios in cult too. paralleled

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

57

same position relativeto Odysseus as is Telemachos,are therebyassimilatedto the actantrole of 'addressee'in the recognition scene. These lines, in effect, are their authorityfor taking part in a recognition scene at all. Yet because their newly acquiredstatus is specific to the formalized context of the recognition scene, room is left for any wider implicationsof the lines - such as any genuine heredity- to be ignored at the narrative'sconvenience; they can be 'companions but this can be conveniently forgottenlateron. and brothers'for the purposeof reintegration, (c) Recognitionscenes and semiotics. type-sceneas signifier Odysseus' dog Argos (17.291-327) is a differentmatter. The scene with Argos involves recognition, but it is not a formal recognition type-scene; other considerationstake priority. For example, this narrativeof someone long-absentreturninghome to find his aged dog still waiting for him sounds suspiciously folkloric: if other narrativepatterns already overdeterminethis scene, then a formal recognition scene would be out of place.41 But it is still recognition, of a sort, if not a formal sort; the absence of the type-scene itself tells us something about recognition scenes. Goldhill (1991) 12-13 examines this scene in particular depth,thoughhe introduces it by emphasizing its functionalsimilarity to other recognitions. He identifies four contexts in which the scene has significance: as 'anotherarrival ... at an animal-guarded threshold',as an unmediatedrecognition 'withoutthe vagariesof speech', as a model of the 'faithfulphilos', and as a way of focusing on the moment of entranceinto the house.42 The second of these is most pertinenthere. All the otherrelationshipsOdysseus has with his oikos are mediatedby recognition scenes (and still more pervasively,by speech, accordingto Goldhill), but here no tokens are needed for there is no mediation;there is no disjunctionto be repaired. There has never been not language, anythingcoming between him and Odysseus, nothingmediatingtheirrelationship: nor any obligations. There is no formal recognition scene for Argos because there is nothing needing to be acknowledged. Tokens, signs for the benefit of the addressee that Odysseus has truly returned,are not the only signs that are importantin recognition scenes: a recognition scene, as a type-scene, and as a formal, conventional structure,is itself a sign. The scene itself denotes reintegration, both to an audience of the epic and also as partof the system of signs that make up the narrativestructure of the Odyssey. But conversely, if the question is asked, what is this thing 'reintegration' that the recognition scene signifies, the answer must be that reintegrationlooks like a series of recognitionscenes. The type-scene is a sign representingsomething happeningin the narrative, and simultaneouslyis one of the building-blocksof the narrative. 5. CLOSINGREMARKS In terms of the 'grammatical'properties of Homeric narrative,the details discussed in the Analysis above are the most importantformal results of this analysis. The discussions there of effect are of general importance (a) scene-length and (c) 'couplets' and their quasi-antistrophic to understanding how an oral poet makes use of formal structuresin organizinghis narrative.
41 Slavic parallels enrich still furtherthe traditional overtones of this scene by suggesting the possibility that Eumaios, also present in this scene, might recognize Odysseus as a result of Argos' recognition. Lord's thematic outline of Slavic Return Songs ((1960) 252-5, Theme Six, 'Arrival Home and Recognitions') shows parallels for a hero being recognized by his horse, and in one case hounds,and consequentlyby a groom or trusted servant:see especially stories E and e. Cf Lord (1991) 55: 'again an almost-recognitionby Eumaeus is interruptedby the partof the storyconcernedwith Odysseus's son [i.e. 17.328-9]'. As anotheroverdeterminingfactor thereis the fact thatthe 'dog at the door' is a typical motif in hospitalityscenes, motif IV in Reece's analysis ((1993) 15, 169-70). One of the anonymousreadersat JHS has suggested a furtherparallel in the CentralAsian epic of Alpamysh, where Argos' role is given to an aged camel: a translationis now accessible in Reichl (2001). 42See also Goldhill (1988), Rose (1979).

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

58

PETERGAINSFORD

given to (b) framingof scenes and (e) sequenceof motifs is well suited Secondly,the consideration to presentinterestin frametheory,which Bakkerrecently has used for his linguistic analyses of but also has much in common with the older style of framingtechniquesin Homericnarrative,43 narrative as divided into discrete episodes (represented that Homeric sees by, for examreading and Notopoulos44). ple, van Groningen Finally,the notes on (d) overlappingand embeddingof moves and motifs, taken in conjunction with the closing discussion of the functionof recognitionscenes in Odyssey 13-24, provide an avenue for approachingthe problemof how formal structurescan adopt a semantic function work such as the Odyssey:when themes overlap, it as literarythemes in a literary-mythological is importantto realize the layers of meaning that are being added. Points of clarification (e), above, shows that this can be importantwhen there is an overlap between a recognition scene and a hospitalityscene. this approachcan give due As well as throwinglight on the mechanics of Homericnarrative, skill for the and of to the involved in the manipuladegree quality recognition performingpoet tion of type-scene motifs. It is to be hoped, for example, thatthe narrativecomplexities and levels of meaning revealed by this analysis will suggest that some comparativelyunpopularpas- deserve rehabilitation as complex pieces of litsages of the Odyssey- Book 14, in particular eraturethat are worth looking at closely. Odyssey 14 receives a less favourableassessmentfrom Kirk(1962) 360-8; but the analysis here shows that, in terms of formalstructure alone, the book is composed of three recognition scenes and a hospitality scene. Much thought and care have gone into Odyssey 14. If no attentionis paid to the literaryvalues representedby and contained in a formal structureof this kind, then the analysis of type-scenes is little more than an esoteric hobby; but with that recognition,type-scene analysis can be a powerful interpretivetool.
PETER GAINSFORD

VictoriaUniversityof Wellington

BIBLIOGRAPHY
Arendt,W. (1933) Die typischenScenen bei Homer (Berlin) Bakker,E. (1997) Poetry in Speech: Oralityand Homeric Discourse (Ithaca,NY)

of storygrammars', Science3, 213-29 R. (1984)'An evaluation Black,J.B.andWilensky, Cognitive in 1-28 HSCPh and 'Convention M.W. Iliad 1', 84, Edwards, individuality (1980)
- (1992) 'Homer and oral tradition:the type-scene', Oral Tradition7.2, 284-330

andWalcot in McAuslan andOdysseus', of Penelope C. (1998a)'Thereunion (1998) 126Emlyn-Jones, in G&R31 (1984) 1-18) 43 (originally - (1998b)'Addendum', andWalcot in McAuslan (1998) 153-4
Felson-Rubin,N. (1994) RegardingPenelope: From Characterto Poetics (Princeton) Fenik, B. (1968) TypicalBattle-Scenesin the Iliad (Hermes Einzelschrift21, Wiesbaden) - (1974) Studies in the Odyssey(Hermes Einzelschrift30, Wiesbaden) Foley, H. (1978) "'Reversesimiles" and sex roles in the Odyssey',Arethusa 11, 7-26

43See especially Bakker (1997).

44 See, e.g., van Groningen (1935),

(1937);

(1949)(esp.7-9), (1951)(esp.83-7),(1964) Notopoulos
58-60.

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions

FORMALANALYSIS OF RECOGNITIONSCENES IN THE ODYSSEY

59

Gainsford,P. (2001) 'Cognition and type-scenes: the aoidos at work', in F. Budelmannand P. Michelakis (eds), Homer, Tragedy,and Beyond: Essays in Honour of RE. Easterling (London) 3-21 Garnham,A. (1983) 'What's wrong with story grammars',Cognition 15, 145-54 Goldhill, S. (1988) 'Reading differences:juxtapositionand the Odyssey',Ramus 17.1, 1-31 - (199 1) ThePoet 's Voice:Essays on Poetics and GreekLiterature(Cambridge) on Homer's Hainsworth,J.B. (1988) in A. Heubeck, S. West and J.B. Hainsworth(eds), A Commentary Odyssey 1 (Oxford) Henderson,J. (1997) 'The name of the tree: recountingOdysseyXXIV 340-2', JHS 117, 87-116 on Homer'sOdyssey2 (Oxford) Hoekstra,A. (1989) in A. HeubeckandA. Hoekstra(eds), A Commentary Kakridis,J. (1971) 'The recognition of Odysseus', in J. Kakridis,Homer Revisited (Publications of the New Society of Letters at Lund 64, Lund) 151-63 Katz, M.A. (1991) Penelope 'sRenown:Meaning and Indeterminacyin the Odyssey (Princeton) Kirk, G.S. (1962) TheSongs of Homer (Cambridge) Krischer,T. (1971) Formale Konventionender homerischenEpik (Munich) Lord,A.B. (1960) TheSinger of Tales(Cambridge,Mass.) - (1991) 'Homeric echoes in Bihac', in A.B. Lord, Epic Singers and Oral Tradition(Ithaca,NY) 49-56 (originally in Zbornikza narodniiivot i obiqaje40 (1962) 313-20) McAuslan, I. and Walcot,P. (eds) (1998) Homer (Oxford) Minchin, E. (2001) Homer and the Resources of Memory.Some Applicationsof Cognitive Theoryto the Iliad and the Odyssey (Oxford) self-disclosureand self-sufficiency in Greekculture',JHS 109, Most, G. (1989) 'The stranger'sstratagem: 114-33 Murnaghan,S. (1987) Disguise and Recognition in the Odyssey (Princeton) 80, Notopoulos, J.A. (1949) 'Parataxisin Homer:a new approachto Homeric literarycriticism', TAPhA 1-23 - (1951) 'Continuityand interconnexionin Homeric oral composition', TAPhA 82, 81-101 in Greek 1-77 'Studies HSCPh 68, -(1964) early poetry', Propp,V.I. (1968) Morphologyof the Folktale, tr. L. Scott and L.A. Wagner(2nd edn, Austin) (originally Morfol6gijaskazki (Leningrad1928)) Reece, S. (1993) The Stranger's Welcome:Oral Theoryand the Aesthetics of the Homeric Hospitality Scene (Ann Arbor) Reichl, K. (2001) Das usbekischeHeldeneposAlpomish:Einflihrung,Text,Obersetzung(Turcologica48, Wiesbaden) Richardson,N.J. (1983) 'Recognition scenes in the Odyssey', Papers of the Liverpool Latin Seminar 4, 223-5 Rose, G. (1979) 'Odysseus' barkingheart', TAPhA109, 216-20 Rosenberg,B.A. (1991) Folklore and Literature:Rival Siblings (Knoxville) Smith, B.H. (1978) On the Marginsof Discourse (Chicago) Stewart,D.J. (1976) The Disguised Guest:Rank,Role and Identityin the Odyssey (Lewisburg) van Groningen, B.A. (1935) 'Elements inorganiquesdans la composition de l'Iliade et de l'Odyss6e', Revue des Etudes Homeriques5, 3-24 - (1937) Paratactische Compositiein de Oudste GriekcheLiteratuur(Amsterdam) in Homers Odyssee. UOberlegungen zur Wirkungsweise Walter,H. (1992) 'Die Wiedererkennungsszenen des Anagnorismos',RPL 15, 61-9 Zumthor,P. (1983) Introductiona la podsie orale (Paris) (also published in English as Oral Poetry: An Introduction,tr. K. Murphy-Jody (Minneapolis 1990))

This content downloaded from 212.128.129.100 on Thu, 3 Oct 2013 08:59:50 AM All use subject to JSTOR Terms and Conditions