Tracking the South Slavic Epic Register

The Milman Parry Collection of Oral Literature, housed in Harvard University’s Widener Library, offers a unique window into the past, and specifically into the world of Yugoslav oral epic as it existed in the 1930s among a non-literate population. Thanks to Milman Parry’s foresight, scholarly rigor, and perseverance, the materials allow scholars working nearly eight decades later to experience this no-longer-living oral tradition almost at first hand. Central to the collection are continuous audio recordings (on aluminum discs), and notebooks with transcriptions made by a team member who was present at (and took part in) the recording process. Furthermore, the recordings include not only lengthy song texts (sung to the gusle as well as recited) but also conversations with the singers. We can thus analyze the language of the songs in detail, and also compare the language of song with the language used by the same individuals in a more neutral, non-singing situation.
Indeed, one of Parry’s primary goals in his Yugoslav field research was to find an analogue to the Homeric Kunstsprache within a living tradition, and to develop an empirical understanding of its inner workings. Could a single individual create and remember songs of the length and complexity of the Iliad and the Odyssey? If so, by what methods? Parry’s great achievement was to demonstrate that singers are able to compose in performance because they have internalized a specific set of rules which allows (and requires) elements of speech to be combined in just the right ways to achieve the desired communication. Albert Lord, who took on Parry’s mantle after the latter’s untimely death in late 1935, called this set of rules “a specialized poetic grammar within the grammar of the language.” [1]
Lord’s description of this “poetic grammar,” in terms of theme and formula, falls roughly within the framework of the Chomskyan linguistic theory prevalent at the time, according to which the role of grammar was to describe the competence of a speaker of a particular language. Chomsky’s focus, and that of linguists who have followed in this vein, was on the idealized grammar which represented this “competence”; they referred to instances of actual speech production as “performance.” [2] For them, the opposition between competence and performance was more or less that between a consistent overall system vs. the often imperfect actual realizations of this system; their primary focus was (and still is) on achieving a full description of the idealized underlying system.
Other linguists felt that the idea of “linguistic competence” should not be restricted to the confines of an idealized grammar, but that the context of speech must also be taken into account. To right the balance, the anthropological linguist Dell Hymes introduced the term “communicative competence,” [3] and the notion that competence in a language meant, not only the ability to recognize and produce grammatically correct utterances, but also the ability to recognize and produce utterances appropriate to the particular speech situation. The study of these different “ways of speaking,” and the fact that they were frequently associated with particular groups or communities, evolved into the discipline known as the ethnography of speaking.
Linguists outside the Chomskyan sphere also began to focus more attention on the context of speech. The discipline of sociolinguistics, which studies varieties of language that are correlated with (and affected by) socially-determined factors of language use, came to the forefront in the 1960s and has continued to grow. Within sociolinguistics, the concept that correlates most closely with communicative competence is that of register. The term itself was first coined in 1956 by T. B. W. Reid, and the concept was popularized by M. A. K. Halliday, who defined it as “the configuration of semantic resources that a member of a culture typically associates with a situation type.” [4] As in the case of “ways of speaking,” registers do not exist in isolation. Rather, as Jean Ure notes, “the register range of a language comprises the range of social situations recognized and controlled by its speakers—situations for which appropriate patterns are available.” [5] Each register is tied to a particular recurring social situation, and functions both to signal the speakers’ recognition of that situation (as opposed to some other situation) and to allow the communication appropriate to that particular situation to take place more effectively. Nearly every community recognizes something akin to a “formal” register and a “casual” register, but other registers are more specialized.
The fact of performance itself, in the sense of poetic performance, also came to be the focus of separate sub-disciplines. Anthropologists began to focus on performance in the sense of verbal artistry, and “an understanding of performance as socially constitutive and efficacious,” [6] while folklorists began to view their objects of study not just as cultural items but also as communicative acts. [7] Within the ethnography of speaking, anthropologists studied “ethnopoetics,” or the different ways in which oral performance could be structured, and the necessity for textual transcripts to take this into account. [8]
How do these new theoretical advances relate to the work of Parry and Lord? Which of the above tools is most relevant to our study of the materials of the Parry collection? To address this question, let us first return to Chomsky’s opposition of competence and performance (idealized grammar vs. individual and often imperfect implementations in actual speech), an opposition frequently compared to the Saussurian dichotomy of langue and parole. One, fairly obvious, interpretation of Lord’s “specialized poetic grammar” that exists “within the grammar of the language” would be that it corresponds to the “competence” of Chomsky’s idealized “speaker-hearer”—in which the latter concept would correspond to Parry and Lord’s idealized “singer of tales.”
Yet even in 1960 Lord recorded his hesitation that the dichotomy of langue and parole would apply to the language of oral performance. “It might be worth suggesting,” he wrote, “that we have in the case of oral epic performance something that is neither langue nor parole, but some third form,” or even “something that is both langue and parole at the same time under different aspects, thus making a third form of communication, or of relationship, peculiar to oral verbal art.” [9] This observation would indicate that Lord already realized (as did Parry before him) that an analysis of the language of epic song must make reference not only to “competence” in both the senses outlined above, but to “performance” as well. In other words, an appropriate framework would be one that would include all of the perspectives mentioned above.
Such a framework has in fact been proposed by John Foley. Making reference not only to Halliday’s definition of register but also to performance theory, ethnopoetics, and the ethnography of speaking broadly defined, he has spoken in some detail of the “epic register” with reference to verbal art of different sorts in different cultures. [10] This is a very productive move, in that it allows us to refer to all the different elements of this very complex phenomenon under the umbrella of a single term. Indeed, by the inclusion of the word “epic,” this conceptual framework introduces yet another element, that of genre. If one views genre in the more traditional sense, that of established literary canons, then the term “epic register”—part conventionalized and static and part situational and dynamic—might seem something of an oxymoron.
More recent views of genre, however, see it in much more fluid terms. Sociolinguists now view register, genre, and style as three distinct but overlapping concepts. Biber and Conrad, for instance, note that both genre and register make reference to the intended purpose of a particular instance of a “text variety” (which can be either spoken or written), as well as to its situational context, and that they differ only with reference to the actual linguistic features, which are “functional” in the case of register and “conventional” in the case of genre. [11] Linguistic anthropologists go even further in pointing out not only the “porousness” of each of the two categories but also the fact that each is continually renegotiated in practice. [12] In the case of the epic singers studied by Parry and Lord, therefore, where each song is composed anew in performance, which in essence meant that each instance recreates the genre, such a multi-faced approach holds a great deal of promise.
Of course, as with any powerful conceptual approach, there will be things that still need working out “on the ground.” In the hopes that this method will become practicable, I should like to make two cautionary notes, each having to do with the obvious fact that registers do not exist in isolation. The first is related to the fact that any one register must be defined with reference to a specific community, and the second to the fact that it must be defined in relation to other registers available to that particular community. In what follows I speak only of the Parry-Lord material; however the cautions are general ones which may be applicable in other instances as well.
As is well known, Parry and Lord carried out their extensive fieldwork in the Yugoslavia of the 1930s, and Lord did additional fieldwork in the Yugoslavia of the 1950s. The series in which portions of the materials have been published bears the title Serbo-Croatian Heroic Songs, where the adjective identifies the language in which the songs were sung. Clearly, therefore, the speech community to which the songs “belonged” was seen as unified by language. Of course, even then the language was pluricentric—a single language with recognizably different norms serving populations in different regions—and anyone familiar with the material knew that the singing tradition they described was native only within part of the language area covered under the rubric “Serbo-Croatian.” Still, because the language was generally recognized as a single entity, this term was sufficient as a rough identification of the type of tradition being discussed.
When Yugoslavia broke up in 1991, however, the language formerly known as Serbo-Croatian split into Bosnian, Croatian, and Serbian, a division formally recognized by the Dayton agreement of 1995 (the independent state of Montenegro, established in 2006, has proclaimed Montenegrin as its official language, and although codification of that language is underway, it does not yet have the generational time-depth of recognition either among its speakers, or within the world at large, as the other three). This of course places scholars of the epic tradition in a quandary: how is one to refer to the unified tradition studied extensively by Parry and Lord? The choice made by Foley is to call it “South Slavic epic tradition,” and it is clear, when he states that he uses this term “to designate genres of verbal art that … can occur, for example, among either ethnic Serbs or ethnic Croats,” [13] that he intends it as a substitute for the term “Serbo-Croatian.”
But this is misleading: the term South Slavic clearly includes not only the former Serbo-Croatian, but also Slovenian, Bulgarian, and Macedonian. It is true that Parry himself consistently used the adjective “Southslavic,” but always with the more general meaning in mind, even mentioning once in this context the hope that “the Macedonian and Bulgarian regions may furnish more definite evidence.” [14] Now that the evidence is in, however, the use of this adjective to modify the noun “epic” is seen to be highly misleading, since Slovenian does not have an epic tradition at all, whereas the Bulgarian epic tradition, now much better known outside Bulgaria, is clearly so different from that studied by Parry and Lord that no one who knows them both would ever group them together. Thus, whereas until quite recently one could reasonably speak of a modern-day group unified by the fact of speaking “Serbo-Croatian,” one can in no way speak of a modern-day group unified by the fact of speaking “South Slavic.” Thus, the phrase “a particular version of the South Slavic language devoted solely to the performance of epic poetry” simply has no meaning. [15]
The problem becomes even more acute when one uses the phrase “South Slavic epic” to modify the noun “register.” The name of a language used as a cover term can reasonably apply to all speakers of a language or, with express qualification, to a subset of them; and the name of a register used as a cover term (such as “baby talk” or “legalese”) can also apply to the traits common to such registers in different languages. But, as nearly all definitions of “register” make clear, any one register is defined by the fact of functioning within (and at the service of) a particular well-defined social unit. The fact that there is no such unit definable as “South Slavic” deprives this term of practical functional meaning.
The breakup of what used to be Serbo-Croatian has created many a dilemma for which there is no easy solution, and this instance is no exception. One possible choice would be to speak of the “Bosniak epic register,” where the adjective refers to what is now a clearly defined social unit—Muslims who speak what used to be called Serbo-Croatian. Another would be to speak directly of a particular set of singers and texts, and speak of the “Herzegovinian Stolac epic tradition,” a choice which might seem apt in this instance, since Foley has clearly identified all his examples as deriving from his close and thorough study of singers from this specific area. A third would be to use an ad hoc term such as “P-L epic tradition,” referring to the specific singing register that functioned in the 1930s in the regions where Parry and Lord worked. But if one is to reap the full benefits of the concept “register,” there must be a clearer, and more concretely grounded, identification of the social group within which it functions.
The second caution has to do with the fact that a register is by definition relative. One can only identify “formal” speech in relation to speech that is less formal, “baby talk” in relation to the kind of talk which occurs outside the presence of babies, “academic jargon” in relation to more neutral prose, and the like. Indeed, the attractiveness, as well as the explanatory power, of the idea “epic register” is that it points not just to a type of language but also to a particular situation, one in which that language is not only appropriate and expected, but one which has also contributed to the formation of the register itself (for still living traditions, one may add “… and which continues to contribute to its ongoing definition”). To arrive at a thorough and nuanced understanding of this complex phenomenon, therefore, it is clearly necessary to define the distinctive traits of the register; and this in turn can only be done by assuming that those who command the register (either actively or passively) have the competence which allows them to differentiate it from the other registers which they also command.
Which other registers should these be? Should one contrast the register appropriate to the specific performance arena of epic song with that of other specific performance arenas and speak of the folktale, or the proverb, or the charm register? Or should one think more broadly, and contrast the overall performance arena (and the register with which it is most commonly associated, the epic) with that of speech outside the performance arena? The second choice seems to make the most sense, and this is what Foley has done. Thus he has stated his intention to contrast the language of epic song of the singers of Stolac with that of “the unmarked conversational standard” of these same singers, [16] specifying further that this should be the singers’ “own conversational registers.” [17] Not only is this the correct choice, but it is also a particularly apposite one, in light of the fact that the Parry Collection includes audio recordings not only of extended stretches of epic singing, but also of extended stretches of conversation between Parry, his field assistant Nikola Vujnović, and the same individuals who did the singing. Thus, although the tradition is no longer living in the form in which Parry and Lord encountered it, one has the next best thing in the ability to contrast extensive recorded samples of these two speech styles.
But the problem that arises here is that the “conversational register” represented by the recordings of these particular conversations is far from neutral, everyday speech. One might counter that there is no such thing as absolutely neutral speech, and this is of course true. Still, most people have a rough sense of the neutral and unmarked, and the fact that they do is what allows us to identify conventional genres as marked in some specific way, or common registers as defined in some specific way. In the case of the unlettered population of a small provincial town such as Stolac in the 1930s, the best approximation of the “unmarked conversational standard” would be that of the local dialect as it is described by ethnographers and dialectologists: here, the description made by Asim Peco on the basis of work with elderly and highly traditional speakers gives the best approximation. [18]
In Parry’s case, although the speech he recorded was clearly conversational, the social context was far from neutral. Neither the physical environment in which the conversations took place, nor the social situation of which each of the participants was highly aware, could be regarded as ordinary or unmarked. One must therefore take these factors of “unnaturalness” into account if one uses this speech style as the standard against which to define the epic register. Of course, Parry was not a sociolinguist or a dialectologist, and he was much less concerned with providing a natural speech environment than he was with making maximal amounts of durable quality audio recordings of continuous singing or speech performance. He accomplished this latter feat by creating an ad hoc recording studio out of two adjoining hotel rooms, placing himself, Nikola, and the singer before a microphone in one room and the recording apparatus (managed by his student assistant, Albert Lord) in the other room. Using the apparatus specially designed by Parry, Lord was able to manipulate the different aluminum disks in a way that allowed recording to proceed without interruption.
From the singer’s point of view, this was a highly marked situation. First, the conversation took place not in the comfortable and familiar environment where relaxed chatting would normally take place, such as the village square or the yard of someone’s home, but in the hotel room of a foreign visitor. Not only that, the singer was placed directly in front of a large mechanical device and expected to speak straight into it, in the knowledge that what he said would be recorded onto disks that were being manipulated in a neighboring room. Second, the social differences between the interlocutors could not go unnoticed. In most instances (and in all instances of Stolac singers), the singer was a Muslim who was being questioned by a countryman of a different religion (Nikola was Catholic) who in turn was clearly taking directions from a foreigner of yet a different religion (Parry was Protestant). Furthermore, the singer, an individual of some stature and consequence in his own community (both in terms of prestige and chronological age), was answering questions put to him by a Catholic compatriot in his late 20s, himself in the employ of an American who also had stature in his own faraway community but was nevertheless only in his early 30s. Finally, the singer was speaking in the knowledge that he would be judged by what he said and sang, and paid accordingly, by a young man who was himself receiving pay from the ultimate instigator of the entire situation, the youthful American professor representing a distant university.
No interview situation is completely neutral, of course, and one must always contend with the observer’s paradox. Still, the characteristics of this particular situation are much more marked than might be the case were these recordings to have been made by linguists, ethnographers, or folklorists working today. Furthermore, in interpreting the material of these conversations, we must take into account not only this highly marked contextual background but also the fact that in most cases our source is limited to a written transcription, and that the majority of these transcriptions (including all those relevant for the current analysis), were made several years after the fact. It is also relevant that the transcriber was Nikola Vujnović himself, working at that point in the U.S., where he had been brought specifically for this purpose. Nikola thus had the advantage (as well as the possible disadvantage) of his own recollections of the recording situation.
To demonstrate the possible action of all these factors, I will quote from the transcript of a section of the conversation with the Stolac singer Ibro Bašić, and then analyze the conversation not only on the basis of the written transcript but also the audio disks from which the transcription was made. [19] In the excerpt below, I have transcribed the text exactly as Nikola wrote it. Portions spoken by Nikola are given in italics; portions spoken by Ibro are in roman type. For purposes of analysis I have broken the text up into lines and sub-sections, added line numbers, and provided an English translation.

A. “Word for word” fidelity 1–13

1 Pjevaš li to Ibro, jednu pjesmu uvjek isto svaki put? Recimo riječ za riječ?
                    1 Ibro, when you sing a song, is it always the same each time? Word for word, so to speak?

2 Ma, ja, ja, pjevam brte ono kako, pa je zgodnije ono znaš
                    2 Well, I, I, when I sing, you know, it’s like, well it goes better, you know,
3 kad čojek uz gusle pjeva ono ode, sve redom oni pjeva onako,
                    3 when you sing it to the gusle. Then everything goes fine the wayit should when you sing.
4 kad kaziva riječ po riječ muka je to, neku zaboravi, neku prebaci, neku tako.
                    4 But when you recite it word for word that’s hard. You forget something, you omit something, stuff like that.
5 A onako Kad pjeva uz gusle ono ide sve ono jedan za drugijem tako.
                    5 But when you sing to the gusle everything goes in order just right.

6 A znaš li ti šta je to kazat riječ za riječ kad pjevaš?
                    6 So do you know what it means to sing something “word for word”?

7 To je rećemo besjeda, ko benzir to je reć taman ko ja i ti sad se razgovaramo,
                    7 That’s like talking, like, well, to say, just like you and I are talking now.
8 ono pa govorimo eto tako i pjesma ono govorimo biva.
                    8 Well, we talk, and that’s what a song is, it’s like talking.

9 A nepreskočiš nikada ništa kad pjevaš jedan put i dva put jednu pjesmu?
                    9 And you never skip anything when you sing a song once, and then the same song again?

10 Pa jok.
                    10 Oh, no.

11 Nikad ništa?
                    11 Never ever?

12 Jok nesmije se preskocit. Ono što zna i što čuje ono mora onako pjevat
                    12 No, you musn’t skip [anything]. You have to sing it just the way you know it and hear it,
13 i ono mora gonit onako.
                    13 and you have to push it through like that.

B. Learning songs as a youth 14–31

14 Ibro.
                    14 Ibro.

15 Ja.
                    15 Yes?

16 Kad si bijo mlad koliko ti je puta trebalo da čuješ jednu pjesmu pa da je naučiš?
                    16 When you were young, how many times did you need to hear a song in order to learn it?

17 Baš kad sam mlad bijo mogo sam je odjednom primit,
                    17 Well when I was young, I could pick it up right away [= in one hearing],
18 a Bogami sad Nikola nemogu. Nemogu je primit Bogami
                    18 but by God Nikola, I can’t do that now. I can’t pick it up
19 i što primi zaboravi, ono iz glave odma izađi.
                    19 and what I do pick up I forget. It just goes right out of my head.

20 I veliš ti samo jedan put da je čuješ?
                    20 So you say you only had to hear it once?

21 Ja kad sam bijo mlad, nek bi je čuo od pjevača jednom
                    21 When I was young, all I had to do was hear it once from a singer,
22 odma bi je primijo. A sad Bogami nemogu.
                    22 and I picked it up immediately. But now by God I can’t.

23 Ma kako to Ibro?
                    23 Well, how is that, Ibro?

24 Ama eto tako.
                    24 Well, it just is.

25 Kako se može tako dugačka pjesma primit.
                    25 How can someone pick up such a long song?

26 More se primit ono sluša čojek ono i drago mu [4406] kako oni pjevač pjeva
                    26 It’s possible to do. It’s like you listen, and you like the way the singer sings it,
27 pa sluša i drago mu. Jer to kako iđe, taman ko škola i učenje.
                    27 and you listen and you like it. That’s how it goes, just like school and learning.
28 Samo mu kaži rećemo jednom onu zadaću i ono uvrti u glavu.
                    28 You just say once what the task is, and it goes straight into one’s head.
29 Jergo sad neiđe nikakav u školu nemore primit više
                    29 And if you never went to school, you can’t get it any more;
30 vet u đetinjstvu primi, i mora primiti tako pa ono mu ište srce znaš, pa da primi.
                    30 you [have to] learn it in childhood. And when your heart wants it, you know,
31 Pa ondak teža pretegni, teža pretegni, i ono zaboravi.
                    31 then you get it. But then it gets harder and harder, and you forget.

C. In search of variants 32–50

32 A vidiš Ibro. Ti si pjevo jednu pjesmu o Đerđelez Aliji
                    32 Look now, Ibro. You sang a song about Đerđelez Alija,
33 i jednu pjesmu o begu Ljuboviću.
                    33 and one about Beg Ljubović.

34 Ja.
                    34 Yeah.

35 I jednu pjesmu o Alagić Aliji.
                    35 And one about Alagić Alija.

36 Jes.
                    36 Yes.

37 A bili ti znao koju drugu pjesmu da ima naprimjer još jedna o Alagiću
                    37 So would you know of any other songs, for instance another one about Alagić
38 ili jedna još o Đerđelezu, ili još jedna ili dvije o Ljuboviću, bili znao ti?
                    38 or another one about Đerđelez, or one or two more about Ljubović? Would you know [some of these]?

39 Nebi, nebi Bogami Nikola.
                    39 No I wouldn’t, by God, Nikola. I wouldn’t.

40 Nebi znao?
                    40 You wouldn’t know [any more]?

41 Neznam brte Bogami, nema fajde jer ono čojeku što treba,
                    41 I don’t know any, by God. There’s no point. If what somebody wants is
42 i ono tačno, ja bi pjevo, rećemo i pjevo, a koju neznam Bogami
                    42 something I know well, then I’d sing it. But by God, if I didn’t know it
43 brate nebi, jeli? Sad ono kad je ono u redu znaš,
                    43 I wouldn’t sing it, would I? If you know it, then it all goes in order, but [if not]
44 pa prebaci onu otud, pa onu otud, pa nekud neiđe.
                    44 you mix this thing in here, and that thing in there, and it ends up not working.

45 A jesili ti čuo kada druge pjesme o njima da pjevaju?
                    45 Well, have you ever heard other [singers] sing some different songs about them?

46 Ko? Ja?
                    46 Who, me?

47 E.
                    47 You.

48 Jesam slušavo ove pjesme a ja drugije slušavo sam i druge pjesme
                    48 Yes, I’ve heard these songs, and I’ve heard others, other songs,
49 pa Bogami zaboravijo, ama slušavo ima nazad rećemo dvajest godin
                    49 but by God I’ve forgotten them. It was some twenty years ago that I heard them,
50 pa ono čojek zaboravi odma. Zaboravi.
                    50 and that’s the sort of thing you forget right off. You forget.

D. “Word for word” reprise 51–69

51 Što ono znači riječ za riječ pjevati u pjesmi?
                    51 What does it mean to sing a song “word for word”?

52 Kako što znači?
                    52 What do you mean, what does it mean?

53 Kažeš, ja sam pjevo Bogami sve riječ za riječ,
                    53 When you say, “By God, I sang that song completely word for word,”
54 što je ono riječ za riječ?
                    54 what does that “word for word” mean?

55 Ono ti je ko ono rećemo što se zove besjede, rijeć,
                    55 That’s, you know, well, it’s like what they call speech, a word.
56 ono besjede rećemo, ono jeli taman ko što ja tebi sad kazivam,
                    56 It’s like speech, you know? It’s just like how I’m talking to you right now.
57 to je kaže to je riječ to je naš jezik, sve ono.
                    57 That’s, you know, that’s a word. It’s our language. A word, you know, it’s speech, all that.

58 Ali što je riječ? Što je riječ kažimi?
                    58 But what’s a word? Tell me what a word is.

59 Besjeda.
                    59 Speech.

60 Besjeda.
                    60 Speech?

61 Ja besjeda, to je reć to je reć riječ taman ko što ja tebi kažem sada,
                    61 Yes, speech. It’s like, like, a word is just like I [would] say to you now,
62 jeli to knjiga Nikola? Jeli to ibrik Nikola. E tako je to je reć.
                    62 “Is that a book, Nikola? Is that an ibrik [coffeepot], Nikola?” There, that’s how to say—how to say a word.

63 Što je recimo jedna riječ u pjesmi. Kažimi jednu riječ iz pjesme?
                    63 Well, then, what’s a word—a word in a song? Tell me a word, tell me, from a song.

64 Ovo ti je ovako rećemo to je reć riječ,
                    64 Here, then, let’s say, here’s [how you] say a word:
65 “Podranijo od Kladuše Mujo, na vrh tanke načinjene kule.”
                    65 “Mujo of Kladuša arose early atop his fine well-made tower.”

66 A to su stihovi.
                    66 But those are verses.

67 E Pa ja to je kod nas, to jes kod vas, a kod nas reče se to.
                    67 Ah—well, that’s how it is with us—I mean with you. Us, that’s how we say it.

68 Ahaha!
                    68 Aha!

69 Ja tako je to.
                    69 Yes, that’s how it is.

E. Pinning down “a word” 70–87

70 A na primjer kada kaže ono podranijo što je ono?
                    70 So for instance, when you say this “arose early,” what is that?

71 To je reć uranijo ili usto hoćeš, pa ono podranijo.
                    71 Well, that is to say “got up early,” or if you want, also “arose early.”

72 A što je ono?
                    72 And what is that?
73 Ja neznam što je ono vet ovo ti je ovako rećemo, šta je ono.
                    73 I don’t know what that is. It’s, you know, just the way we say it, that’s all.
74 Podranijo od Kladuše Mujo ili rećemo, to je reć podranijo,
                    74 “Mujo of Kladuša arose early,” or you could say maybe “arose early,
75 prije zore, ogranka sunca. To je reć podranijo, uranijo, podranijo, tako.
                    75 before dawn, at break of day.” You could say “arose early,” “got up early,” “arose early,” things like that.

76 A na primjer kada ti kažeš Ja sam Bašić, što je to?
                    76 And for instance, when you say “I am Bašić,” what is that?

77 Neznam Bogami.
                    77 God knows! I don’t.

78 Neznaš ti to?
                    78 You don’t know this?

79 Ja jok Bogami. Jer ja njesam školovan tako, pa ja sam Bašić, ja to,
                    79 No way, by God. I’m not schooled in such, I’m just Bašić, I, well—
80 jer treba tu čojek školovan pa da sve to istomači,
                    80 you need somebody with schooling to explain all that to you.
81 a ja neumijem to istomačit ono materinijem jezikom.
                    81 Me, I don’t know how to explain that, not in my mother tongue.

82 A jeli to riječ na primjer kazati sam?
                    82 So is it a word, for instance, to say “am”/“oneself” ?

83 Jes.
                    83 Yes.

84 Jes, to je riječ?
                    84 It is? That’s a word?

85 Ja, sam čojek, jes.
                    85 Yes, I—I, myself, a person. Yes.

86 Ali kako onda može biti ono riječ, Podranijo od Kladuše Mujo?
                    86 But then how can this “Mujo of Kladuša arose early” be a word [too]?

87 Ama ja neznam, tako je to reć, tako je to.
                    87 Well I don’t know, that’s just how you say it, that’s just how it is.

F. Return to song texts (88–)

88 U redu moj Iro [sic].
                    88 O.K., my Ibro.

89 Ja.
                    89 Yeah.

90 A znaš li ti još štogod o Gazi Hrustanbegu?
                    90 So do you know some other piece about Gazi Hrustanbeg?


This selection is of especial interest since it contains a particularly eloquent discussion of the concept of riječ (the singer’s speech act), [20] one which has been frequently quoted. Parry’s interest in the concept of “word for word” fidelity, a topic which comes up frequently in his conversations with the singers, is the central concern of section A. The conversation then moves to the process of transmission and learning (section B) and the singer’s repertoire (section C). Sections D and E return to the idea of “word for word.” Although section D is the one most frequently quoted, both are critical in this regard; indeed the near breakdown of communication in section E contributes significantly to our understanding of the concept in the mind of the singer. In section F Nikola moves to re-establish a communicative equilibrium with the singer; due to space limitations, only the beginning of this section has been included.
Comparison of the transcription with the audio records, which contain non-verbal information such as pauses, laughter, and intonational cues, is highly instructive. For one thing, although Parry’s words are only infrequently marked in the transcript, they are much more frequently heard in the background. Furthermore, it is clear from the authoritative tone of his voice, even if his words are not always intelligible, that he is in control of the conversation. Nikola and Ibro appear to accept this fact of control, though there are certain instances of discomfiture, heard both in non-verbal cues and in one purposeful word choice made by Nikola. Even the transcription itself provides evidence of this discomfiture. Most of it is accurate, but there are certain errors. Two rather significant ones concern omissions that occur relatively early in the text: in line 42, Ibro says not just ono tačno but ono tačno koju znam ‘exactly the one I know’, and in line 44 he says not just pa prebaci but pa ono nije pa prebaci ‘but if it isn’t then you mix …”.
Nearly all the errors, however, occur during a segment of the conversation when Nikola appeared to be caught between his need to obey his employer and his knowledge (as a singer himself) that Parry’s question will make no sense to Ibro. Because it is known that Nikola made the transcription of the conversation at least three (and perhaps four) years after it took place, [21] it would appear that his unease about this culture clash was still present at some level of his consciousness. I have included the correction for each of the errors in the translation, and will summarize the most significant of them below.
The sections of relevance are D and E. After line 50 (the end of section C), there is a long pause. Parry’s words in the background are not audible, but he is clearly directing Nikola to return to the issue of “word for word.” Both Nikola and Ibro seem disoriented at this: they appear to feel that the question had been settled already (in section A). Nikola transcribes his own speech incorrectly in line 51, adding an extra instance of za riječ (as if he had said the equivalent of “word for word for word”, which he clearly did not say), and Ibro’s intonation in line 52 conveys a clear sense of mistrust. Further on, Nikola writes for Ibro’s speech in line 56 tebi sad instead of sad tebi (“[talking] to you now”): there is no essential change in meaning, but the lapse is unusual for Nikola, who usually makes a very faithful transcription of what is on the record.
In line 57, however, his error is more significant: he leaves out an entire phrase in one of Ibro’s most significant speeches. Thus, Ibro says not just to je kaže to je riječ to je naš jezik sve ono, but rather to je kaže to je riječ to je naš jezik riječ kaže to je besjede sve ono “That’s, you know, that’s a word, you know, that’s our language. A word, you know, like speech, all that”. This new element, the idea of speech (or utterance), becomes the center of the next few lines (59–62), especially line 62, where Ibro illustrates the idea of a self-contained utterance very clearly (jeli to knjiga Nikola? Jeli to ibrik Nikola “Is that a book, Nikola? Is than an ibrik, Nikola?”). Parry’s voice is heard in the background immediately after this, and it is right at this point that Nikola again leaves out an entire phrase of Ibro’s. Thus, the conclusion of line 62 is not E tako je to je reć, but E tako je to je reći reć riječ (There, that’s how to say—how to say, to say a word.) Then in the next line, when Parry’s words are heard in the background again, Nikola once more omits to transcribe several words from his own speech. This time there is no change in the meaning, but the smoothness of the transcription hides the fact that Nikola was at this point feeling some unease. Thus, instead of Što je recimo jedna riječ u pjesmi, Kažimi jednu riječ iz pjesme?, Nikola says Što je recimo jedna riječ jedna riječ u pjesmi, Kažimi jednu riječ kaži iz pjesme? (“Well, what’s a word, a word in a song? Tell me a word, tell me, from a song”). In short, Nikola’s transcription of this whole section conveys his own apparent discomfort as he recalls the situation. Once he and Ibro agree on the all-important distinction between “words” and “verses,” however, in lines 66–67, equilibrium appears to be restored. Indeed, the intonational patterns at this point convey that all three sides (singer, singer/investigator, and scholar) are happy with this distinction.
But then Parry decides to push the issue further. After line 69, we hear his words clearly in the background, as he says to Nikola the equivalent of “Ask him if ja [‘I’] is a word, ja sam [‘I am’]”. This puts Nikola in a quandary. As a singer himself, he knows this question will make no sense to Ibro, so he not only omits to transcribe altogether the clearly audible line of Parry’s speech, but he also—after a notable pause—substitutes a different word, podranio ‘arose early’ (line 70). But then Parry repeats the request, after line 75, and Nikola is forced to put the question directly to Ibro (line 76). Ibro’s response (line 77) is accompanied by derisive laughter, followed by his highly significant observation about the incongruence between the worlds of orality and literacy (lines 79–81), at which point the conversation seems almost in danger of breaking down altogether. But Parry is still not satisfied: we hear his directions again in the background, in response to which Nikola asks Ibro again (line 82) whether sam is a word. In doing so, however, Nikola pronounces the word not in the meaning “am” (quite naturally, since the rules of the language do not allow this word to be pronounced in isolation), but rather with a long vowel, in the meaning “oneself”. After this exchange, even Parry seems to acknowledge the dead end, and he directs Nikola to shift the conversation back to Ibro’s repertoire (section F).
The intent of the above excursus, comprising a brief description of the social situation of the conversations in general and an examination of a specific portion of one of these conversations, has been to demonstrate that the conversational speech of singers (as recorded by Parry) cannot be taken as exemplary of a neutral register, an “unmarked conversational standard.” What this speech is, however, is much more interesting than what it is not. First, it is an example of the type of discourse which sociolinguists analyze in terms of participant manipulation, in terms of the choices made by each participant in a conversation as he negotiates the constant trade-off between the achievement of his own individual goals and the maintenance of the conversational flow. Space does not permit a full conversational analysis of this excerpt, though a few elements have been mentioned above. Second, this situation gives us valuable insight into the actual means by which Parry both obtained his remarkable collection and derived the significant insights which Lord then brought to full expression. It is paramount to remember as well that Parry also obtained nearly all of the most valuable epic song texts in his collection in this very same recording situation, one in which the two-person audience, though clearly a knowledgeable one, was not representative of the social situation in which singers normally performed.
This, of course, reminds us that the performance situation represented by the Parry-Lord material is one which we must largely reconstruct, and that we have relatively little direct information about the neutral context of South Slavic epic performance at the time Parry did his recordings. We have what Murko and Parry have told us about their observations, and we have a few isolated narratives by the singers themselves (indeed, at another point in his conversation with Nikola, [22] Ibro tells a vivid tale, replete with epic hyperbole, of how he out-sang several other singers on a particular occasion). But without actual videotape evidence or multiple detailed first-hand reports, we simply do not have enough information about the performance situation to achieve a full-fledged description of the performance register. At the same time, we can and must use the insights of modern scholarship to deepen and enhance our understanding of what the epic register must have been like in those portions of the South Slavic world of which Milman Parry and Albert Lord have left us such a vivid record.


Agha, A. 2006. Language and Social Relations. Cambridge.
Bauman, R. 1977. Verbal Art as Performance. Prospect Heights, IL.
Bauman, R., and C. L. Briggs. 1990. “Poetics and Performance as Critical Perspectives on Language and Social Life.” Annual Review of Anthropology 19:59–88.
———. 1991. “Genre, Intertextuality, and Social Power.” Journal of Linguistic Anthropology 2:131-172.
Biber, D., and S. Conrad. 2009. Register, Genre, and Style. Cambridge.
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA.
Halliday, M. A. K. 1978. Language as Social Semiotic: The Social Interpretation of Language and Meaning. Baltimore.
Hymes, D. 1966. “Two Types of Linguistic Relativity.” Sociolinguistics, ed. W. Bright, 114–158. The Hague.
———. 1994. “Ethnopoetics, Oral-Formulaic Theory, and Editing Texts.” Oral Tradition 9:330–370.
Foley, J. M. 1992. “Word-power, Performance, and Tradition.” Journal of American Folklore 105:275–301.
———. 1995. The Singer of Tales in Performance. Bloomington.
———. 1996. “Guslar and Aoidos: Traditional Register in South Slavic and Homeric Epic.” Transactions of the American Philological Association 126:11–41.
———. 1999. Homer’s Traditional Art. University Park, PA.
Lord, A. B. 1960. The Singer of Tales. Cambridge, MA.
Murko, M. 1929. La poésie populaire épique en Yougoslavie au début du XXe siècle. Paris.
———. 1951. Tragom srpskohrvatske epike. Zagreb.
Parry, M. 1987. “Ćor Huso: A Study of Southslavic Song, Extracts.” The Making of Homeric Verse: The Collected Papers of Milman Parry, ed. A. Parry, 437–464. Oxford.
Peco, A. 1964. “Istočnohercegovački govor.” Srpski dijalektološki zbornik 14:1–200.
Reid, T. B. W. 1956. “Linguistics, Structuralism and Philology.” Archivum Linguisticum 8:28–37.
Ure, J. 1982. “Introduction: Approaches to the Study of Register Range.” International Journal of the Sociology of Language 35:5–24.


[ back ] 1. Lord 1960:36.
[ back ] 2. Chomsky 1965.
[ back ] 3. Hymes 1966.
[ back ] 4. Halliday 1978:111.
[ back ] 5. Ure 1982:5.
[ back ] 6. Bauman and Briggs 1991:79.
[ back ] 7. Bauman 1977.
[ back ] 8. Hymes 1994.
[ back ] 9. Lord 1960:279–280.
[ back ] 10. Foley 1992, 1995, 1996, 1999. The self-contained article printed in 1996, “Guslar and Aoidos: Traditional Register in South Slavic and Homeric Epic”, was reprinted with only minimal changes as chapter 3 (entitled “Homer and the South Slavic Guslar: Traditional Register”) of the 1999 book Homer’s Traditional Art.
[ back ] 11. Biber and Conrad 2009:2.
[ back ] 12. See Bauman and Briggs 1991 on genre, and Agha 2006 on register.
[ back ] 13. Foley 1995:xiin3.
[ back ] 14. Parry 1987:445.
[ back ] 15. Foley 1996:25; 1999:74, my emphasis in each instance.
[ back ] 16. Foley 1996:25, 1999:74.
[ back ] 17. Foley 1996:25n32; 1999:293.
[ back ] 18. Peco 1964.
[ back ] 19. Both the full transcript (PN 6598) and the audio files of one portion of it (disks 4405–4406) were made available to me courtesy of the Milman Parry Collection of Oral Literature housed in Harvard’s Widener Library. I am particularly grateful to David Elmer and Peter McMurray for facilitating my access to audio recordings.
[ back ] 20. For unknown reasons, John Foley consistently used the Serbian-marked form reč ‘word’ to designate this concept, despite the fact that all the singers Parry worked with used the form riječ in the meaning ‘word’, and that nearly all of them were from regions (Bosnia, Herzegovina, and Montenegro) where riječ was the accepted standard form.
[ back ] 21. The conversation took place in June, 1935, and Nikola made the transcription at some point during his 18-month stay at Harvard, between 1938 and 1940.
[ back ] 22. PN 6598, disks 4403–4404.