Combining Digital Resources to Investigate the Oral-formulaic Diction of Sanskrit Texts (Part 1)

In my previous posts on The Digital Orientalist, I introduced two types of linguistic resources dedicated to the study of lexicon and morphosyntax, called WordNets and Treebanks, respectively. In this series of posts, I will show how the information contained in these two resources can be used together to study the formulaic diction of Sanskrit texts. Before presenting the methodology, it will be necessary to introduce the notion of formularity in oral poetry—especially considering that this notion is usually associated with texts from Western classical literature, such as the Homeric poems—along with a new linguistic approach to the study of the latter. In my next post, to illustrate the methodology, I will demonstrate how, thanks to the Sanskrit WordNet and the Vedic Treebank, it is possible to extract the basic formula KILL THE DRAGON from the Rigveda (RV), a collection of religious hymns written in Vedic Sanskrit that constitutes the oldest texts in Indian literature.

These posts are partly based on the following article (in Italian): Biagetti, Erica (2023). Integrare Sanskrit WordNet e Vedic TreeBank: uno studio pilota sulla formularità del Rigveda tra semantica e sintassi. (En: Linking the Sanskrit WordNet with the Vedic Treebank: a pilot study on formulaicy in the R̥igveda between syntax and semantics). In Bossolino, Isabella and Chiara Zanchi (eds.), E pluribus unum. Prospettive sull’Antico. Pavia: Pavia University Press.

Formulas as constructions

Studies on formulaic language originated and primarily developed within Homeric philology and were later extended to other oral poetic traditions. Milman Parry and his student Albert B. Lord demonstrated that the Iliad and the Odyssey were almost entirely composed by oral singers through the use of a vast repertoire of formulas, i.e., “a group of words which is regularly employed under the same metrical conditions to express a given essential idea” (Parry [1930] 1971, p. 272), and formulaic systems, i.e., “a group of phrases which have the same metrical value and which are enough alike in thought and words to leave no doubt that the poet who used them knew them not only as a single formula, but also as formulas of a certain type” (Parry [1930] 1971, p. 275). Through field research on the flourishing oral epic tradition in Yugoslavia at the time, Parry and Lord validated this theory, demonstrating that the craft of composing epic songs was learned in a manner akin to language acquisition, requiring poets to master formulas, themes, and narrative patterns.

Kiparsky (1976) was the first to propose a syntactic definition of a formula. He suggested comparing formulas with idiomatic expressions found in everyday language, distinguishing between fixed idiomatic expressions (1)a and fixed formulas (1)b on the one hand, and flexible idiomatic expressions (2)a and flexible formulas (2)b on the other:

The English idiomatic expression in (1)a does not usually undergo lexical or morphosyntactic modifications and is therefore said to be a fixed idiom. The idiomatic expression in (2)a, on the contrary, is said to be a schematic idiom because the verb spill can be inflected for different persons, tenses and mood, and the noun beans can be modified by a definite article (the) or by other determiners (some, more, his). The following sentences taken from the English Web 2021 corpus show some possible realizations of this schematic idiom:

(3) She can’t spill the beans; I mean that Veronica spills the beans; But Dan Vado … spilled some beans; Xiaomi would spill more beans; With the man instantly spilling his beans…; etc. (enTenTen 2021)

Similarly, the formula in (1)b is always attested in this form in the Homeric poems and can therefore be categorized as a fixed formula, whereas the formula in (2)b is schematic, as the verb path– ‘suffer’ can also be inflected for different persons, tenses and mood:

(4) álgea páskhōn ‘suffering sorrows’ 9×, álge’ épaskhon ‘(they) were suffering sorrows’ 1×, álgea páskhei ‘(he) suffers sorrows’ 2×, álgea páskhein ‘to suffer sorrows’ 2×, etc.

While recognizing the true essence of the formula in the “abstract connection between álgos and path-” (1976: 86), Kiparsky defined formulas as syntactic constituents dominated by a single node, thereby introducing a syntactic level into the underlying structure of a formula. For instance, the NP and VP labels in the notation of the schematic expressions in example (2) signal that such expressions consist of a Verb Phrase (VP) governing a Noun Phrase (NP).

Recently, scholars of different oral traditions—namely Bozzone (2014, 2024) for the Homeric tradition, Frog (2024a,b,c) for the Skaldic one, and Pagán Cánovas and Antović (2016a,b) for the South Slavic tradition—have expanded upon Lord’s (1960, p. 36) idea that formulaic language is acquired and functions much like a natural language and have proposed studying oral compositional techniques through the lens of Construction Grammar. The foundational hypothesis of this theory is that constructions—i.e., recurring associations of a particular form with a particular function—are the basic units of language (Fillmore and Kay 1993; Goldberg 1995). Any linguistic expression with characteristics that are not fully compositional or that need to be stored in memory falls under the notion of construction: idiomatic expressions are therefore constructions, but so are syntactic structures, as they carry meaning, however abstract, even before being “filled” with words. Constructions constitute an organized inventory of a speaker’s understanding of their language’s conventions (Langacker 1987, pp. 63–76). This inventory is depicted as a taxonomic network of constructions, with taxonomic relations illustrating the degree of schematicity between two constructions (Croft and Cruse 2004, pp. 262–63). Consider example (5), where the specific idiom in a. is one realization of the more schematic idiom The X-er, the Y-er:

Bozzone, Frog, Pagán Cánovas and Antović provide a fresh perspective by recognizing that formulas, like idiomatic expressions, are recurrent pairings of specific forms with specific functions, thus categorizing as constructions. As Bozzone (2014, pp. 40–41) points out, this view of formulas as constructions reflects varying levels of flexibility, ranging from schematic (6)a to partially fixed (6)b, to fully fixed (6)c.

Rigvedic formularity

Studies on formularity in ancient Indian poetry agree in attributing a certain degree of flexibility to the compositional technique of the Rigveda: Kiparsky (1976: 99-103)–yes, the same Kiparsky we met in the previous section–introduced this text into the debate on oral poetry, emphasizing that its compositional technique makes limited use of long, metrically defined, and fixed formulas, as the Homeric ones we commented on above. Rather, the Rigveda features a web of schematic formulas, which belong to the poets’ repertoire but take on different instantiations in the text through lexical or grammatical substitutions, rearrangements, semantic inversions, or metrical variation (Jamison and Brereton 2014, p. 14; see also Jamison 1998). Lindqvist (2015, pp. 32-33) defines the formulaic diction of this text as “a set of formulaic networks or continua, which owe their cohesion to similarities, their extension to differences between words and phrases”. In other words, we could say that Rigvedic diction is marked by formulaic continua, but fixed formulas occur far less frequently compared to Homeric diction, on which oral-formulaic theory was developed. To substantiate this claim and to demonstrate the challenges of applying a Parryan definition of formula to Rigvedic diction, Lindqvist (2015, pp. 29-35) illustrates two systems that include seemingly fixed formulas such as píbā sómam ‘drink the soma!’ and áhann áhim ‘(he) killed the serpent’.

The verb phrase píbā.impv sómam.acc [1-4] is regularly employed (18×) under specific metrical conditions (syllables 1-4 of the pāda) to express an essential idea, namely to invite the god Indra to drink the sacrificial beverage. However, the expression also occurs in other metrical positions and with an inversion in the order of constituents: sómam píbā̆ [3-6]. Furthermore, the phrase is characterized by morphosyntactic and lexical flexibility, and it can feature a verb in the second person dual (sómam pibatam ‘drink the soma!’) or an object in the genitive (píbā somásya.GEN ‘drink of the soma!’), and can contain the noun mádhu- ‘sweet drink’ instead of sóma- (píbā mádhvas.GEN ‘drink of the sweet drink!’). Finally, the expression can be interrupted by other elements, as in example (7):

(7) píbendra sómam
píbā indra sómam
drink.IMPV.2SG Indra.VOC soma.ACC
‘Drink the soma, Indra!’ (RV 2.16.4d)

As shown by the previous examples, all variations of the expression píbā sómam share the same semantic meaning of drink soma, but they differ in one or more levels of analysis, making it difficult to draw a clear boundary between expressions that constitute a formula and those that do not. For this reason, Lindqvist (2015, p. 31) suggests defining a formula in relative terms, as a portion of the formulaic continuum to which it belongs. 

Scholars focusing on Rigvedic diction often resort to the notions of web, network, and continuum to describe the type of formularity observed in this text. The attentive reader will remember that we introduced the notion of a network earlier in this post, when describing how constructions are organized in the speaker’s mind. In fact, formulaic networks or continua identified by Rigvedic scholars can be interpreted as manifestations of the network of constructions that made up the poet’s repertoire. In the next post, I will briefly introduce the necessary resources to put this idea into practice, namely the Sanskrit WordNet and the Vedic Treebank, before demonstrating how to apply this new methodology to the second formulaic network discussed by Lindqvist and primarily known through the work of Calvert Watkins (1995).

References

Biagetti, Erica. 2023. Integrare Sanskrit WordNet e Vedic TreeBank: Uno studio pilota sulla formularità del Rigveda tra semantica e sintassi. In E pluribus unum. Prospettive sull’Antico Per i Decennalia dei Cantieri d’Autunno: I seminari dell’Università di Pavia dedicati al mondo antico. Edited by Isabella Bossolino and Chiara Zanchi. Pavia: PUP, pp. 45–62.

Bozzone, Chiara. 2014. Homeric Constructions. Ph.D. thesis, University of California, Los Angeles, CA, USA.

Bozzone, Chiara. 2024. Homer’s Living Language: Formularity, Dialect, and Creativity in Oral-Traditional Poetry. Cambridge: Cambridge University Press.

Croft, William, and D. Alan Cruse. 2004. Cognitive Linguistics. Cambridge: Cambridge University Press.

Fillmore, Charles J., Paul Kay, and Mary Catherine O’Connor. 1988. Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone. Language 64: 501–38.

Frog. 2014a. Mythological Names in dróttkvætt Formulae I. Studia Metrica et Poetica 1: 100–39. 

Frog. 2014b. Mythological Names in dróttkvætt Formulae II. Studia Metrica et Poetica 1: 39–70. 

Frog. 2014c. Oral Poetry as Language Practice: A Perspective on Old Norse dróttkvætt Composition. In Song and Emergent Poetics—Laulu ja runo—Πесня и видoизменяющaяся пoэтикa. Edited by Pekka Huttu-Hiltunen, Frog, Karina Lukin and Eila Stepanova. Kuhmo: Juminkeko, pp. 279–307.

Goldberg, Adele E. 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: Chicago University Press.

Jamison, Stephanie W. 1998, “Rigvedic viśvátaḥ sīm, or, Why syntax needs poetics.” In Mir curad: Studies in honor of Calvert Watkins, edited by Jay Jasanoff, H. Craig Melchert and Lisi Oliver, 291–299. Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft.

Jamison, Stephanie W., and Joel P. Brereton, eds. 2014. The Rigveda: the earliest religious poetry of India. Oxford: Oxford University Press.

Kiparsky, Paul. 1976. Oral Poetry: Some Linguistic and Typological Considerations. In Oral Literature and the Formula. Edited by Benjamin A. Stolz and Richard S. Shannon. Ann Arbor: Center for Coordination of Ancient and Modern Studies, pp. 73–106.

Langacker, Ronald W. 1987. Foundations of Cognitive Grammar, Vol. 1: Theoretical Prerequisites. Stanford, CA: Stanford University Press.

Lord, Albert B. 1960. The Singer of Tales. Cambridge: Harvard University Press.

Pagán Cánovas, Cristóbal, and Mihailo Antović. 2016. Construction grammar and oral formulaic theory. In Oral Poetics and Cognitive Science. Edited by Mihailo Antovi´c and Cristóbal Pagán Cánovas. Berlin: De Gruyter, pp. 79–98.

Parry, Milman. 1971. The Traditional Epithet in Homer. In The Making of Homeric Verse: The Collected Papers of Milman Parry. Edited by Adam Parry. Oxford: Oxford University Press, pp. 1–190. First published 1928.

Watkins, Calvert. 1995. How to Kill a Dragon: Aspects of Indo-European Poetics. New York and Oxford: Oxford University Press.