Harris on Antović and Cánovas and eds., 'Oral Poetics and Cognitive Science'

Mihailo Antović, Cristóbal Pagán Cánovas, eds.
Randy Allen Harris

Mihailo Antović, Cristóbal Pagán Cánovas, eds. Oral Poetics and Cognitive Science. Linguae and Litterae Series. Berlin: De Gruyter, 2016. 203 pp. $140.00 (cloth), ISBN 978-3-11-034838-5.

Reviewed by Randy Allen Harris (University of Waterloo) Published on H-Rhetor (February, 2019) Commissioned by Cristen Fitzpatrick (St. John's University)

Printable Version: http://www.h-net.org/reviews/showpdf.php?id=52589

This edited collection of essays grew out of a 2013 conference at the Freiburg Institute for Advanced Studies bringing together two fields that fit naturally with each other and that both have considerable importance for rhetoric. Unfortunately for the field it adumbrates, cognitive oral poetics, the essays are individually programmatic and collectively disjointed; unfortunately for rhetoricians, there is little or no awareness of our discipline. I will summarize each of the essays, which are not without their (sometimes considerable) merits, and conclude with some comments on their value for our discipline.

The editors, who also organized the Frieburg conference, are Mihailo Antović, a linguist in the Department of English at the University of Niš, and Cristóbal Pagán Cánovas, a linguist and a classicist with the Institute for Culture and Society at the University of Navarra. Their introduction is a brief but promising set of speculations on the mutual relevance of oral poetics (the study of oral performance epics) and cognitive linguistics (the study of how general mental processes and dispositions shape language use). Oral poetics stands on the impressive early twentieth-century foundation of work by Milman Parry and his student Albert Lord. Before their research, the great works of Homer and cognate epics like Beowulf were assumed to be individual compositions of particular poets, who wrote them down—that is, the standard literary model that we see with epics like the Aeneid and the Faerie Queene. Parry forcefully demonstrated that they were records of oral performances, narratives assembled on the fly, improvised from an inventory of stock patterns and a repertoire of linguistic strategies, by bards. He did so elegantly, by careful and compelling comparisons with the performances of illiterate Serbo-Croatian oral singers, recorded on field trips to Bosnia in the early 1930s.

The similarities were indisputable, especially the creative use and recombination of metrical-grammatical units he called “oral formulae.” Parry died young, and Lord, who helped gather and analyze the Serbo-Croatian data, brought the theory home in his classic Singer of Tales (1960). There were few resources for an accompanying psychological theory when Parry and Lord were writing (in particular, the dominant theory, behaviorism, had nothing to say about such performances), but scholars like Walter Ong and Marshall McLuhan later realized the cognitive dimensions of this research and took it in remarkable directions. Antović and Cánovas briefly sketch out this background, connect it to cognitive linguistics, and give précis of the contributions to the book.

Elizabeth Minchin, a distinguished classicist and emeritus professor at the Australian National University, starts the collection proper off with the tightly argued, data-rich “Repetition in Homeric Epic: Cognitive and Linguistic Perspectives.” Her repetitions are not the formulae so well known of oral epics; rather, they are units of narrative structure, what the Parry-Lord tradition (though not Parry or Lord) call “type-scenes,” such as donning-armor events, which always proceed from greaves to spear in a specified order, and contest events (“the prizes are set up; the challenge is announced; competitors come forward; preparations for the competition are made; the contest takes place; and prizes are collected” [p. 19]). Curiously, she uses the term “script” for these type-scenes, rather than the parallel term “frame,” which also comes out of artificial intelligence but is more common to the cognitive humanities. She argues that these units are cognitive phenomena that reduce mnemonic burdens on the bards and processing burdens on the audience and that have clear ordinary language homologs.

Raymond F. Person, a scholar of religion and a Semiticist at Ohio Northern University, gives us the fascinating “From Grammar in Everyday Conversation to Special Grammar in Oral Traditions: A Case Study of Ring Composition.” Ring composition may be the most famous patterning device of oral narratives (Minchin touches on it briefly as well), the chiastic arrangement of episodes radiating out from a centering event, often schematized in a letter formalism ABCDXDCBA. This Beowulf sequence is a locus classicus of the pattern:

A. Hrothgar’s speech on Æschere’s death and description of the monster’s pool (1321-82)

B. Beowulf wants to avenge Æschere’s death (1383-96)

C. Danes and Geats go to the pool (1399-1421)

D. arming of Beowulf in his helm and byrnie (1441b-54)

X. the battle and defeat of Grendel’s mother (1455-1628)

D. disarming of Beowulf in his helm and byrnie (1629-30a)

C. Danes and Geats return to Heorot (1632-50)

B. Beowulf reports his successful revenge (1651-76)

A. Hrothgar’s speech on pride (1700-84) (pp. 34-35). 

Ring composition, and chiastic patterns in general, also have irresistible cognitive accounts, with their obvious reliance on mental affinities like repetition, contrast, and symmetry. Person uses ring composition to take on the “Great Divide” thesis in oral poetics, which relies on a special grammar, one that is qualitatively different from the grammar of everyday speech. Ordinary language, we know, is littered with chiastic expressions (“A place for everything and everything in its place,” “Do unto others as you would have others do unto you,” “Starkist doesn’t want tuna with good taste. It wants tuna that tastes good”). But Person bypasses them to focus on larger discursive patterns. We are all familiar, for instance, with the conference questioner who wraps three or four questions into one turn, and the speaker who responds, “let’s take the last one first,” and works inversely through them (Person has an example from a journalist and Ronald Reagan). The phenomenon is also present in anecdotal storytelling and dynamic conversations (a nesting of adjacency pairs, in conversational analysis terms). This kind of radiating structure is, Person argues, a way to achieve efficiency in communicative economies, in ordinary language just as in the vast narratives of oral performance epics. He does not overturn the Great Divide thesis, however, so much as bridge it part way, maintaining that there is a special grammar that is “superimposed” on the grammar of everyday speech.

Two Wills, William Duffy and William Michael Short, both classicists at The University of Texas at San Antonio, team up for “Metaphor as Ideology: The Greek ‘Folk Model’ of the Epic Tradition,” perhaps the most readable essay in the book for anyone unfamiliar with linguistic theory. The essay that in many ways triggered the cognitive linguistics movement in the latter twentieth century is Michael Reddy’s “The Conduit Metaphor: A Case of Frame Conflict in Our Language about Language,” the first close linguistic analysis of a so-called conceptual metaphor, and virtually a playbook for George Lakoff and Mark Johnson’s hugely influential Metaphors We Live By (1980).[1] The two Wills do not cite Reddy (the essay is relatively unknown), but they chart out the vocabulary of the Greek bardic tradition in terms very similar to Reddy’s chart of talk-about-talk vocabulary. Reddy notes the prevalence of expressions like “Whenever you have a good idea practice capturing it in words,” “Try to pack more thoughts into fewer words,” and “Insert those ideas elsewhere in the paragraph” (the examples, and the italics, are Reddy’s).[2] That is, we frequently talk about talk as if language were a kind of package into which speakers and writers put their thoughts, which then get shipped to listeners and readers, who take those thoughts out and store them away. This kind of cross-domain analogic vocabulary is well known to the rhetorical tradition, of course. Quintilian called it “allegory,” with scholars like Giambatissta Vico and Henry Peacham cataloguing numerous examples. Reddy coins the term “metaphorisms” for these cross-domain usages, which seems appropriate, but he is assuredly wrong, as is the Lakovian-Johnsonian tradition after him, in calling the analogic frames that shape such usage “metaphors,” let alone using the redundant modifier “conceptual” (all metaphors are conceptual). The two Wills document the analogic vocabulary of the Greek bardic tradition, which uses terminologies of building and binding and weaving to frame the performance of oral epics.

The editors team up for “Construction Grammar and Oral Formulaic Theory,” which I will discuss in connection with Hans C. Boas’s “Frames and Constructions for the Study of Oral Poetics.” The two essays cover much of the same territory in somewhat complementary (but also somewhat redundant) ways. Each has strengths the other lacks and weaknesses the other redresses, though there are surprisingly no cross references at all. Boas is better positioned to handle the linguistic theory end. He is a named professor (really named, tri-named: Boas is The Raymond Dickson, Alton C. Allen, and Dillon Anderson Centennial Professor) in Germanic studies and linguistics at The University of Texas at Austin. He is also one of the leading construction grammarians in the world. Both essays summarize construction grammar, the most important post-Chomskyan syntactic theory, and Boas does the same for the related post-Chomskyan semantic theory, frame semantics (the two theories fall in the general domain of cognitive linguistics); and both essays center on the way constructions, syntactically understood, relate to oral formulae, metrically understood. The most distinctive characteristic of a construction (any typified expression) in construction grammar is its construal as a form-meaning dyad. Take the familiar English ditransitive (two-object) pattern:

1. Fred gave Wilma the necklace.

This pattern was heavily researched in the Chomskyan paradigm, but it was always regarded as an abstract arrangement that required interpretation through a semantic rule system based on the properties of the verb (to give requiring a giver, a receiver, and an object). But now consider 2-5:

1. Fred handed Wilma the necklace.

2. Fred tossed Wilma the necklace.

3. Fred dropkicked Wilma the necklace.

4. Fred blorked Wilma the necklace.

We might say, I suppose, that to hand requires the same arrangement of semantic roles, but to toss? To dropkick? To blork? Yet, in each case—and most tellingly with the nonsense word—we understand the sentence as a motion carried out by Fred that results in Wilma getting the necklace. That is, the meaning is yoked directly to the form, not to the verb or the product of some independent semantic-interpretation rule. Oral-poetic formulae, on the other hand, are conceived quite differently. They come out of Parry and Lord virtually as formal monads, metrical units that are almost accidentally meaningful.

Boas gives the better account of construction grammar but has little to offer in terms of its application to oral performance epics, merely asserting there are riches to be found. “By carefully studying the broad variability of epic poems,” he says, “construction grammarians can investigate and analyze the inventory of grammatical constructions that serve to regulate and produce oral performances” (p. 120). If this is the case, and I believe with Antović and Cánovas that it is, then it is a fundamental misunderstanding to see formulae as formal monads. Boas offers no real evidence. Antović and Cánovas, on the other hand, do. They provide the kind of payoff this work can produce in their close analysis of one formula in Serbo-Croatian performance epics, a recurrent morphosyntactic pattern in which a verb that literally means “to sit” means “to mount” when it participates in a grammatical complex requiring the aorist, specific verbal prefixes, and, when the meter allows for it, the temporal particle, pa. As in the ditransitive transference example above, the meaning is bound to the form. This is an important illustration, too, that the literal/figural continuum is syntactic as much as it is semantic.

Boas also gives a very helpful overview of frame semantics, including an introduction to the web-available ontology, FrameNet, so one can explore the theory directly (an activity I recommend). FrameNet is an approach to lexical semantics that overlaps with, but is not fully integrated with, construction grammar. It sees words as functioning primarily to activate patterns of thought about situations and actions. Knowing and using the word “revenge,” for instance, is not a consequence of accessing some dictionary denotation, like “one causing harm to some other who has previously harmed one.” Rather, it is “the infliction of punishment in return for a wrong suffered. An AVENGER performs a PUNISHMENT on an OFFENDER as a consequence of an earlier action by the OFFENDER, the INJURY. The AVENGER inflicting PUNISHMENT need not be the same as the INJURED_PARTY who suffered the INJURY, but the AVENGER does have to share the judgment that the OFFENDER’S action was wrong. The judgment that the OFFENDER had inflicted an INJURY is made without regard to the law.”[3] That is, lexical meaning is a dynamic function of the way words participate in generic activities. They call on other words, serving interrelated roles (revenge requires, for example, words filling such roles as AVENGER, PUNISHMENT, and OFFENDER). Their symbolic action relies on a web of understandings (additional frames) of temporal sequence, social obligation, emotional response, and so on. In a word, they activate and potentiate stories. Boas shows how such a semantics naturally supports bardic performances (though, regrettably, he uses a literary artefact to do so, a Sherlock Holmes story, not an oral epic).

Anna Bonifazi, classicist and linguist at Universität Heidelberg, gives us “Particles as Cues to Structuring in Serbo-Croatian and Early Greek Epic,” perhaps the most technical essay of the lot. Particles are often seen as highly expendable linguistic superfluities, semantically and even phonologically lightweight (well, oh, the to of infinitives, and so on). Bonifazi argues that they are “cues to multimodal structuring ... inherent components of the epic oral tradition,” “precious features indexing a semiotically complex event” (pp. 144, 125).

Mark de Kreij’s “The Priming Act in Homeric Epic” notes that the vast storyworlds of epics present particular challenges for oral performance—especially for the audience to track multiple characters, relationships, and narrative entanglements, which the spatial perceptions of prose reading manage more easily. Kreij, a classicist currently holding a postdoctoral fellowship with romance studies and classics at Stockholm University, argues that fixed linguistic constructions help to stabilize and buttress the storyworlds, supporting the claim with a close analysis of constructions containing the “priming acts” of his title: grammatically encoded intimations of an upcoming character or event, micro-prolepses of a sort.

Sonja Zeman, a linguist in Germanic studies at Universität Bamberg, wins the prize for longest title with the final entry of the volume, “Orality, Visualization, and the Historical Mind: The ‘Visual Present’ in (Semi-)oral Epic Poems and Its Implications for a Theory of Cognitive Oral Poetics,” which builds a sophisticated argument around the use of present tense in oral and semi-oral (or “transitional”) epics. Zeman sees the historical present not in the conventionally stylistic sense of creating immediacy or vividness but as a deeply cognitive way of weaving a shared experience, mediated especially through visualization, foregrounding “the relationship between the ‘teller’ and the ‘told’ / within a shared mental world” (pp. 189-190). Rhetoricians will recall the notion of enargia in connection with Zeman’s insights.

The final essay reveals as fully as any of the essays in this volume that the cognitive science of the title is too broad. The contributions, from the introduction all the way through to Zeman’s essay, rarely stray beyond the cluster of positions and methods associated with cognitive linguistics. Even the closely allied hermeneutic discipline of cognitive poetics gets little more than lip service in the volume, and that only in the introduction—despite the presence of its godfather, Mark Turner, at the conference that spawned the collection.

The book does not go very far toward realizing the hopes of establishing a field of oral cognitive poetics, charting no cogent program forward. Nor, despite the whiff of manifesto in Antović and Cánovas’s introduction, does it provide the incitement to do so. The essays ply several themes relevant to the field: the reliance on metacommunication (talking about talking, or singing about singing) as a technique of audience orientation; the continuity between ordinary language and poetic language; and creativity as a function of the formulaic reuse, repurposing, and recombination of stock forms and templates. They also lay out some of the tools and approaches that the field would need to engage, most prominently construction grammar and frame semantics. But the authors and editors fail to work together; to agree on a common terminology; to divide up problems; to agree on paradigmatic devices, passages, or analyses; or even to reference each other’s work. With three years from conference to publication, more coherence might be expected.

There is, however, as I hope I have conveyed, some strong work here, and much of value for rhetoricians. The Homeric epics are early sources of political oratory, arranged on bardic principles. Oral epics as a genre are chock-full of narrative strategies, argumentation tactics, and protocols of description, exposition, and instruction. They are important preliterate rhetorical repositories. They are as full as the wine-dark sea of rhetorical figures. And importantly, the circumstances of performance—long, long stretches of coherent, metrical, aesthetically pleasing, and memorable narrative poetry—elevate the productive and processing demands of language use in a way that is particularly revealing of its cognitive substrata, the way a vast cathedral reveals the structural and material requirements of all architecture. Understanding oral cognition also has an exponentially increasing relevance for the communications of our still-young century. Ong used the term “secondary orality” for the way radio and television resurrected communicative principles dormant under the regimes of print, but a digital age has exploded out of his now-quaint-looking “electronic age.” We may be in an age, with its incessant facetweetagrammchatting, so far beyond Ong’s vision that the cardinality of the phenomenon is incalculable. It is way beyond the scope of some phrase like “tertiary orality” in any case. The many media of orality, and their rampancy, eclipse wholly all other periods of human history.

The contributors to this volume would certainly have benefited from some awareness of rhetoric, which is almost completely absent from the volume. The most egregious omission (with the one inevitable exception of something called metaphor) is the utter lack of rhetorical figures, which, properly understood, not only comprise the superset that includes oral poetic formulae but are also constructions in precisely the sense of construction grammar (form/meaning dyads[4]) and are profoundly cognitive, owing their salience to resonances with mental affinities (similarity, opposition, correlation, repetition, derangement, symmetry, and so on). But rhetoricians have never let our neighbors’ indifference prevent us from enriching our discipline, and there is value for us in this volume, albeit scattered and largely accidental.


[1]. Michael Reddy, “The Conduit Metaphor: A Case of Frame Conflict in Our Language about Language,” in Metaphor and Thought, ed. Andrew Ortony (Cambridge: Cambridge University Press, 1979), 284-324.

[2]. Ibid., 287.

[3]. “Revenge,” FrameNet, The FrameNet Project, Colin Baker, project manager, October 6, 2002, framenet2.icsi.berkeley.edu/fnReports/data/frameIndex.xml?frame=Revenge.

[4]. See Mark Turner’s unfortunately neglected essay, “Figure,” for some arguments of this front. “Figure,” in Figurative Language and Thought, ed. Cristina Cacciari, Ray Gibbs Jr., and Albert Katz (Oxford: Oxford University Press, 1997), 44-87.

Citation: Randy Allen Harris. Review of Antović, Mihailo; Cánovas, Cristóbal Pagán; eds., Oral Poetics and Cognitive Science. H-Rhetor, H-Net Reviews. February, 2019. URL: http://www.h-net.org/reviews/showrev.php?id=52589

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.

Editorial correction.

The proper title of this review should read: “Harris on Antović and Cánovas, eds., 'Oral Poetics and Cognitive Science'”