A genealogy of poetry

1 What is poetry?

Recognizing a poem is rather easy but defining poetry in general can be more difficult. Over the years, many famous poets have attempted to define their craft. Emily Dickinson, in a letter to Thomas Wentworth Higginson in 1870, famously wrote, “If I feel physically as if the top of my head were taken off, I know that is poetry.” This notion, that poetry is inextricably tied to emotion and feelings, is part of a long Romantic tradition. William Wordsworth, in his preface to Lyrical Ballads, published in 1801, made a similar statement when he wrote, “Poetry is the spontaneous overflow of powerful feelings: it takes its origin from emotion recollected in tranquility.” Wordsworth’s contemporary Percy Bysshe Shelley also located poetry within the individual but fixated on the imagination over emotions. In his essay “A Defense of Poetry,” written in 1821 and posthumously published in 1840, Shelly wrote, “Poetry, in a general sense, may be defined to be ‘the expression of the imagination’: and poetry is connate with the origin of man.” Shelly recognized poetry as something central to being human, yet he also understood poetry to be a production, something that is expressed. One of the most quoted statements regarding what poetry expresses comes from Samuel Taylor Coleridge, who, in 1827, famously made the distinction between poetry and prose, remarking, “I wish our clever young poets would remember my homely definition of prose and poetry; that is, prose = words in their best order; – poetry = the best words in the best order.” Coleridge focuses less on the affects or even the location of poetry, but rather turns his attention to the very construction of poetry. His eye towards the structural constitution of poetry would later be echoed by Rita Dove in 1993, after being named Poet Laureate of the United States, who said, “Poetry is language at its most distilled and most powerful.” Taken in concert, what all of these definitions reveal is that poetry is and does many things, and that the notion of poetry itself has been subject to change over time and place. Poetry thus varies both diachronically (in its history and evolution) and synchronically (across and within cultures). Structural or formal definitions of poetry will thus always be insufficient because a poem’s form is always relationally constrained by the time and place of its production. While there is a long tradition of examining poetics from a formalist standpoint that dates back to the ancient Greeks, what we need is a definition of poetry that can account for historical and cultural differences but that can also highlight similarities across various poetic traditions. What we need is a more functional approach to poetry, one that reveals that the evolution of poetry is at once historically contingent and culturally universal. Such an understanding of poetry recognizes its multiplicity and its continual becoming.

This article is an attempt to create the foundation for a functionalist poetics. Such an approach hinges on the fact that poetry, as both a concept and a genre, is not stable. Despite this instability, anthropological studies into oral poetic traditions reveal that there are aspects of poetry that we may consider to be universal. One crucial aspect, I will argue, is that poetry is a form of ritual language. Another is that poems are highly referential linguistic objects that express beliefs and values. This, I believe, is in fact the central function of poetry: it is a tool for creating and sharing complex thoughts, meaning, and emotions.

2 A functionalist poetics and the fluid concept of poetry

Poetry does not have a history; it has many histories. Yet poetry is not reducible to either the singular or the multiple. Poetry has no clear beginning or end, it is decentered, rhizomatic in the Deluzian sense, though unlike the nonsignifying systems of A Thousand Plateaus, poetry is inherently a signifying system. While pursuing a biological analogy of poetry’s evolution that looks at poetic traditions through the lenses of natural selection, genetic mutations, genetic drift, and horizontal gene transfer might prove useful for understanding the many ways in which poetry changes, more relevant to the argument here is that such an analogy provides a timescale of change, one that unites the seemingly antithetical aspects of poetry as both innate (something belonging to all languages) and historical (something that includes a variety of traditions both oral and written). The issue with most classical approaches to defining poetry is that they center on similarity – poems are understood to share certain formal characteristics. Under such frameworks, different types of poetry can be difficult to reconcile. In order to do so, theories of poetics have traditionally over relied on categorization, dividing poetry into distinct forms such as epic, lyric, ballad, pastoral, etc. The problem that then arises is that these categories themselves are always subject to change and revision, being historically and culturally determined, and so even this approach becomes limiting. A poetics that focuses on function, on the other hand, is much more capable of dealing with a broader and more amorphous definition of poetry.

The concept of poetry, like all concepts, is not stable. The definition of poetry is always in flux, as new poems challenge old conventions and revive lost traditions. Władysław Tatarkiewicz charts the historical evolution of the concept of poetry in his book A History of Six Ideas. Tatarkiewicz (1980: 12–14) notes that in early Greece, poetry was not considered art because the arts were defined by rules and precepts, and since the belief was that poetry springs from inspiration, it was considered closer to philosophy, music, or prophesy than to art.

Tatarkiewicz (1980: 75) notes that the early Greeks did not have a specific term for poetry, not because they did not have poetry but because they simply did not need the concept. Their poetry was oral, sung poetry, and while specific genres of poetry were defined, there was no need for a larger, catchall term.^[1] The ancient Greeks understood poetry as an expression of the highest form of knowledge because of its “psychagogic capacity” (Tatarkiewicz 1980: 84) to influence minds and to converse with the spiritual world of the divine (which is why Plato condemned the poets in his Republic). They had two concepts of poetry: one, as bardic inspiration, and the other defined by its form as verse (1980: 86). The view of poetry as supernatural inspiration in verse form persisted through the Middle Ages, and when poetry was included as part of the fine arts, the result was that all art became imbued with spirituality (1980: 116). Tatarkiewicz (1980: 21–23) traces the inclusion of poetry as an art to the Middle Ages, and the separation of the fine arts from the sciences to the Renaissance, and cites a parallel transformation that took place between 1500 and 1750 when art as mimesis yielded to a new conceptual understanding of art as the production of beauty and the imagination. The view of art as beauty then lead to the definition of art as expression in the nineteenth century, and later to that which produces aesthetic experience at the beginning of the twentieth century (1980: 31).

Returning to the concept of poetry, Tatarkiewicz (1980: 51–52) outlines how the handcrafts and sciences were eliminated from the concept of art, and how music and poetry were included as their rules developed, though this process took longer for poetry than for music. This conceptual transformation was achieved in large part due to the division of the fine arts from other, lesser arts during the Enlightenment (1980: 60). Tatarkiewicz (1980: 248) writes, “By the eighteenth century, the concept of creativity was appearing more frequently in art theory. It was linked with the concept of imagination, which in that period was on all lips.” The rise of creativity as central to the production of art during the nineteenth century was drawn from the earlier recognition of poetry as inspired by a divine creator (1980: 249–250).

While verse form had long been a precondition for poetry, Tatarkiewicz suggests that this condition began to erode during the nineteenth century. He writes:

The concept of poetry was broadened from versed speech, but narrowed to the lyric. The philosopher Jouffroy wrote that “there is no poetry other than the lyrical.” This concept of poetry no longer referred to verbal art alone; in this sense, poetry manifests itself in musical compositions and in paintings. Also it had transcended the limits of the arts: there is a poetry, too, in natural scenes and in human situations. This concept has become familiar and natural to us: it is well understood, if difficult to define. The matter is best conveyed by Paul Valéry’s description: “Poetry is the attempt to represent … through the artistic resources of language those things that are vaguely expressed by ears, caresses, kisses, sighs, etc. … The thing cannot be described in any other way.” (Tatarkiewicz 1980: 118)

While recent scholarship and archival recovery have revealed that the equation of poetry with the lyric was not widely held until the early twentieth century, it is interesting to note that this transition was already beginning a century before.^[2] Jonathan Culler (2015: 49, 72, 90) examines the evolution of the concept of the lyric in his book Theory of the Lyric, from its earliest incarnation in ancient Greece as “ritualized, performative speech,” to its formation as an encompassing poetic category (with multiple subgenres) in Italy and Spain during the sixteenth century, to its modern manifestation as a poetic genre that encompasses “short, nonnarrative poetry.”

What Tatarkiewicz and Culler help underscore is that the terms “poetry” and “lyric” are genres, and that genres change over time. Writing against the notion of genres as set distinction, Ian Patterson (2013: 217) suggests, “the term ‘lyric’ itself may be better considered as an aspect of poems – a mode – rather than as a hard and fast generic distinction.” Culler makes a similar argument against the use of genres in general, but notes that they make for useful categories in analysis. While I agree with Patterson that we can think of “lyric” as a quality of poetry, the main point that should be derived from these discussions is that genres should be thought of as continuums rather than as fixed categories.

3 Conceptual categorization: poetry as genre

To acknowledge poetry as a genre is to acknowledge it as a semantic concept, not a formal one (Devitt 2004: 17). While formal concepts are rigid, semantic concepts are fluid because they are always situational. Amy Devitt (2004: 25), in Writing Genres, suggests that to understand genre as a semantic concept entails acknowledging genres as existing in rhetorical situations, defined by culture, and in a nexus with other genres. This places genres (and poetry) as relation or matrix “between an individual’s actions and a socially defined context” (2004: 31). Central to navigating this matrix are specific purposes, actors, and language use, collectively known as a ‘register’, specific to the genre (2004: 17). Carolyn Miller (1984: 162) notes that since genres occur because of our sense of recurring rhetorical situations, they will inevitably vary from culture to culture depending on social structures and expectations.

In this light, genres can be understood as specific kinds of concepts, and will thus share conceptual structures. While there is much debate regarding the structure of concepts, the dominant view is that concepts are mental representations (Laurence and Margolis 1999: 77), or what George Lakoff (1999: 392) refers to as “cognitive models.” Edouard Machery (2009) asserts in Doing Without Concepts that there are in fact three types of concepts: prototypes, exemplars, and theories, and argues that, as his title suggests, we should stop using the term “concept” as it obfuscates the different kinds of concepts that exist. In the prototype model, prototypical concepts are essentially encoded lists of properties that we use to decide if an item belongs to a specific class; however, while items that fit into a specific prototypical category tend to reflect the properties of the list, items must only meet a sufficient number of features (Laurence and Margolis 1999: 27). An apple tree might be a prototypical example of a tree, but a palm tree possesses enough of the properties we might list under our prototypical concept “tree” to still be considered a tree. In the exemplar model, concepts are sets of exemplars (rather than sets of abstract lists as in the prototype model); in the exemplar model, categorization is based on exemplars, examples from experience that we draw on from long-term memory that serve as a typical or ideal representation of the category (Machery 2009: 93). Exemplar theory is based on similarity, because it requires us to compare items to individual memories of category members (2009: 182). In this model, the first time we see a palm tree we might reflect on our experiences and compare this new item to our memories of apple trees and other trees in order to categorize the palm tree as belonging to the categorical concept “tree.” In the theory model, categorization is not based on similarity (as it is in the prototype and exemplar models); instead, in the theory model, sometimes called Theory-Theory, concepts are mental theories analogous to scientific theories. In addition, in the theory model, concepts are structured mental representations whose structure consists of relations to other concepts and embedded theories (Laurence and Margolis 1999: 44–45). In Theory-Theory, rather than comparing an item to a list of properties, people ask whether or not an item has the right hidden property or essence (Laurence and Margolis 1999: 46). According to the theory model, when we attempt to categorize a palm tree, we test it against our mental theory of what a tree is to see if it meets our standard of “tree-ness.” Machery endorses a multi-process theory of categorization, and claims that there are at least three categorization processes, which include prototypes, exemplars, and theories.

Lakoff makes a similar argument that there are multiple ways in which we categorize, though he stresses the role of metonymic associations in the formation of concepts, and claims that current research regarding the discovery of prototypes as a cognitive structure is flawed. He explains that what the discovery of prototypical feature-bundles as well as exemplars in current research trials actually indicate are effects of our cognitive models and thus do not actually reveal anything directly about the structural composition or nature of concept formation and categorization (Lakoff 1999: 391). A major source of prototype effects, Lakoff (1999: 406) argues, is metonymically based reasoning (in which one aspect of something is used to represent the thing as a whole), of which there are numerous models that all yield prototype effects.

Alternatively, Eric Margolis and Stephen Laurence (1999: 71, 75) champion a version of Dual Prototype Theory in which each prototypical concept contains two poles with atomistic cores, meaning that they are not singular or unitary, and are instead composed of similarities and resemblances, with one pole being very typical and the other not. Lakoff and Johnson (1980: 71) provide the example, “small flying singing birds, like sparrows, robins, etc. are prototypical birds. Chickens, ostriches, and penguins are birds but are not central members of the category – they are nonprototypical birds. But they are birds nonetheless, because they bear sufficient family resemblances to the prototype; that is, they share enough of the relevant properties to be classified by people as birds.” The reason it is important to recognize poetry is both a concept and a genre is because doing so reminds us that the definition of poetry is never stable, which is one of the reasons why it is so difficult to define. As a concept, poetry exists within the mental representations, or cognitive models, of individuals. As a genre, poetry exists outside of individuals as a nexus linking the dialogical matrix of a poem’s creation and the cultural context in which it is created. All this is to suggest that “poetry” means different things to different people in different times and places.

Regardless of the underlying cognitive structures that give rise to concepts, what is clear is that concepts grow and are extended through analogy as we compare new experiences to past experiences (Hofstadter and Sander 2013). This suggests that each time someone encounters a poem, the experience modifies their concept of poetry. The more diverse range of poems one encounters, the more flexible one’s concept of poetry is likely to be and, similarly, the wider range of concepts for poetic subgenres one is likely to develop. If one were to only read Shakespearian sonnets, one’s prototypical conception of a poem would be much more constrained than someone familiar with sonnets, ballads, nursery rhymes, and free verse. The concept of poetry as a genre is thus never stable and will inevitably vary individually and culturally. This understanding allows for great flexibility in what counts as a poem, as the only requirement is that a poetic object trigger a prototypical effect such that one recognizes it as being sufficiently similar to a prototypical poem.^[3]

4 Oral poetry and poetic universals

If the genre “poetry” is contingent on the various rhetorical situations of its production, and the concept “poetry” gives rise to prototype effects, then although poetry is marked by difference it is also marked by similarity. To discover the aspects central to poetry, we need to expand our conceptual recognition beyond the West and written poetic traditions, and explore an inclusive, anthropological view that can also account for oral poetic traditions. Ruth Finnegan (1992: 13) notes that although what we consider lyric poetry today is likely a universal aspect of all human cultures, not all cultures possess a word for poetry. Some cultures may only have terms for specific types of poetry, as we saw with the ancients Greeks, and in some cultures, there may be difficulty differentiating between what we might consider prose and verse (1992: 26). John Miles Foley (2002: 28) calls into question our Western mode of categorization as well, and contends that in some cultures there are oral traditions that may fall outside of our notion of lyric verse, such as “charms, recipes, proverbs, genealogies, funeral laments” and other so-called “minor-genres,” but that we are wrong to impose our Western perspective on these generic forms and exclude them from the realm of poetry. Before outlining a poetics that can encompass such a broad range of poetic traditions, however, it is important to first address these issues raised by Finnegan and Foley. How are we to reconcile the fact that not all cultures recognize poetry as we recognize it, and the inherent implication here that people are always constrained by their own cultural perspectives?

This is a central concern for anthropologists, who must grapple with how to understand and explain cultural customs and their significance from a vantage point that is decidedly outside the cultures they observe. Brazilian anthropologist Eduardo Viveiros de Castro (2014: 41) writes that the challenge ethnographers face is “to grasp the consequences of the idea that those societies and cultures that are the object of anthropological research influence or, to put it more accurately, coproduce the theories of society and culture that it formulates.” The solution to this challenge, Viveiros de Castro (2014: 43) proposes, is to develop what he calls “Anthropology as comparative ontography.” Comparative seems the operative word here. What Viveiros de Castro implies is that it is the responsibility of anthropologists to not only explain the cultural relevance of certain behaviors from the perspective of the cultures they are studying, but also to situate their own theoretical framework in a comparative context that recognizes its own modes of cultural production. This is important because the tools of analysis always influence what can be studied. With respect to poetry, this implies that the demarcation of boundaries, of where we draw the line for what is and is not poetry, will dictate what we can (and cannot) recognize as poetry. Edwin Hutchins notes,

Every theory implies a set of ontological commitments and every ontological commitment emphasizes some kinds of connections over others. What looks like low connectivity under one theory may look like a region of high connectivity to another theory. As we tend to look where our theories indicate areas of interest, by looking in particular ways, we actually make the world appear to be the kind of place that our theories can address. (Hutchins 2010: 706)

Hutchins reminds us that theories are not objective claims about the nature of reality but are instead cultural products themselves that in turn produce the objects of their analysis. Einstein’s theory of General Relativity did not create spacetime, of course, but the concept of spacetime enabled us to think about space and time as a singular field. When thinking about concepts to apply cross-culturally, it is important to keep in mind not only how we have defined these concepts but how other cultures define them (and how we understand their definitions in relation to our own as well).

The problem we are confronted with then is how to talk about poetry in a way that does not impose preconceived notions or received ideas about what poetry is and can be when thinking about poetry in cross-cultural contexts. The solution, it seems to me, is to recognize our own vantagepoint, and precede accordingly, knowing that we may be limited by our own poetic histories but try to recognize these limitations when they appear. With respect to Finnegan’s point that not all cultures have a word for poetry, then are we required to accept that these cultures do not possess poetry? I do not believe this is the case. What I think this reveals, however, is the distinction between what the linguist Kenneth Pike coined emic and etic knowledge. Borrowed from the linguistic terms “phonemics” and “phonetics,” Pike understood that like certain linguistic categories, cultural behaviors can be understood from either an insider or an outsider perspective. An emic view of culture is an insider’s view, the perspective held by people with an intrinsic cultural awareness of the meaning within their culture, it is the view in anthropology from the perspective of the subject; while the etic view of a culture is that held by an outsider looking in, the perspective of the anthropologist observing another’s culture. With these terms in mind, we can see how while it may be true that from an emic stance not all cultures might recognize that they possess a form of poetry, we can say that from our etic perspective we might judge differently and find that even a culture without a word for poetry nonetheless possesses a highly symbolic or ritualized language that we might recognize as poetry.

If we accept this position, that not all cultures might recognize the term “poetry,” but that all cultures possess a form of language we might call poetry, then we are ready to begin an anthropological exploration into what we mean by poetry. Such an endeavor requires us to move beyond the written poetry so familiar to the West and begin to explore oral poetic traditions. Foley (2002: 38) remarks that oral poetries fundamentally differ from textual poetries in their composition, performance, and reception. These relate to what Foley describes as the three key concepts to the theory of Immanent Art: register, performance arena, and communicative economy. He asserts that oral poetry operates in a register different from and more coded than everyday speech that connects oral poems to oral traditions, that oral poetry is performed for an audience and as such functions as an enactment of ritual that creates a performance arena, and that all of this takes place in a communicative setting that is highly efficient for expression because both audience and performer are fluent in conversing in the coded structures of their oral poetic traditions (2002: 114–117). Foley develops a number of proverbs that he invents to describe what he sees as core elements of oral poetry. These include “oral poetry works like language, only more so” (2002: 127), which refers to his observation that oral poetry differs from normal speech in its broader array of expressive strategies, including the use of what he dubs “bigger words,” in which a “word” in an oral poem is perceived as “a unit of utterance, an irreducible atom of performance, a speech act” (2002: 13).^[4]

When “poetry” is expanded to include oral poetry as well as written poetry, we can begin to recognize poetry as a human universal. Giorgio Banti and Francesco Giannattasio (2004: 290) write, “Ethnographic research has shown that ‘poetic’ forms and behaviors are almost universally widespread … But looking at the facts more carefully one comes to realize that one of the most defining characteristics of poetry – and significantly of music as well – is the impossibility of defining it in any simple way.” Like Foley, Banti and Giannattasio (2004: 306) recognize that while poetry is defined by each culture, all poetries share in the fact that they occur in a register that differs from ordinary, conventional speech, including special morphology, syntax, lexicon, and other stylistic features. They posit that in addition to occurring in an alternate speech register, non-conventional speech also utilizes alternating melodic contours (maintaining a single pitch, or adding emphasis to articulation like a musical scale) and the segmentation of rhythmic utterances (2004: 295). Banti and Giannattasio (2004: 295) argue that these aspects of non-conventional speech heighten its symbolic effect, and that every culture possesses some form of expression “ideally intermediate between language and music.” They go on to propose that it is “likely that poetry and music are separate developments of an originally common procedure of sounds and rhythmic formalization of speech, as suggested by Molino (2002)” (2004: 298). In addition to poetry’s unique register and association with music, Banti and Giannattasio (2004: 308) note that poetic speech often involves a “veiled style” that is more heavily imbued with cultural and emotional significance. As Foley indicated with his example of “bigger words,” Banti and Giannattasio also describe how poetic speech utilizes multi-word metaphors that are purposefully made for interpretation.

5 Poetry as ritual language

The performative nature of poetry sets it apart from ordinary speech or writing and creates a setting ripe for interpretation. These aspects of poetry align with what David Tavárez (2014: 530) argues are the three key features of ritual language: “it is an overt structuring of speech performance, it is exercised reflexively, and it remains open to various forms of indeterminacy.” Poetry is inherently performative, and an important aspect of this performativity is the self-reflexive nature with which poems point to themselves as poems – linguistic objects that are highly interpretable and indeterminate, that reflect the conditions of their own production, and that exist in a register outside of normative discourse. Like poetry, ritual language is also characterized as a highly stylized and heightened form of ordinary language (Bell 1997: 51, 70).

Tavárez (2014: 518) outlines “five domains that inform many known forms of ritual language in different societies: parallelism and repetition; representation and mimicry; enaction and personification; authority; and reflexivity and indeterminacy.” By analyzing these domains in relation to poetry, we can see that poetry is a form of ritual language. As has already been established, poetry is highly self-reflexive and indeterminate due to its performativity. This performativity is partly derived from the performance arenas where poems are received as well as from the poet who claims authorial authority as someone steeped in the cultural and creative traditions of the poem’s production. As an embodied artform, poetry is inherently enacted (both in its production and its reception), and while not all poems rely on personification, it is a common poetic technique. There is a long tradition of understanding poetry as a form of mimesis that not only represents or mimics the world but also coproduces our social world through its enaction. Lastly, and arguably most importantly, poetry also relies heavily on parallelism and repetition. This is a key domain because parallel, iterative structures invite reflexivity and self-awareness, which are fundamental to poetry and ritual language.

6 Poetry and parallelism

Roman Jakobson (1966: 399) long ago noted the importance of parallelism to poetry, writing, “on every level of language the essence of poetic artifice consists in recurrent returns.” Parallelism can function at every level of language – narrative, symbolic, metrical, lexical, phonic, grammatical – often operating at multiple levels that interact with each other. Yet parallelism, it should be noted, is more than mere repetition. James Fox (2014: 33) explains, “Jakobson has urged a decidedly more open exploration of the pairing of elements in corresponding sequences. The essential feature of this pairing – whether it is based on supposed synonymy, antonymy or ‘synthetic’ determination – is that it involves simultaneously identification and differentiation. In this sense, parallelism is distinguished from repetition, which involves identification alone.” If repetition creates rhythm, parallelism creates interesting rhythms, allowing repetition leeway to play with and against itself and our expectations.^[5]

Related to this discussion of parallelism, and the importance of repetition, similarity, and difference, is the concept of recursion, which gives rise to nested, hierarchical repetition. Steven Pinker (1994: 481) defines recursion as, “A procedure that invokes an instance of itself, and thus can be applied repeatedly to create or analyze entities of any size.” This process of self-similar embedding allows for increasingly complex and potentially infinite sentences in which phrases may be continuously embedded in other phrases. A distinction between the two terms is important: while the structure of the “recurrent returns” generated by parallelism typically appears parallel, or linear, which creates rhythm, the structure of grammatical recursion is nested and hierarchical, which creates syntactical complexity. In terms of the information conveyed, parallelism at the level of narrative can serve the same purpose as recursion at the grammatical level – There is a dog. The dog is big. The dog is black. The dog belongs to my neighbor. I don’t like my neighbor… conveys the same information as My neighbor, whom I don’t like, has a big, black dog… – yet the transmission of information will not be as efficient. However, the recursive nature of a narrative creates a rhythm, built from repeated words and phrases that can serve as mnemonic aids. The significant difference between grammatical recursion and grammatical parallelism is that recursion has a fractal-like structure and requires self-similar embedding, while grammatical parallelism does not require this type of hierarchical nesting of its recurrent parts. From this we can infer that recursion is in fact a specific kind of grammatical parallelism. Brian Kim Stefans (2017: 30) notes the similarity between recursion and Jakobson’s use of parallelism, and writes, “Recursion’s primary avenue is through grammatical parallelism.” Parallelism allows for recursion at the level of grammar when repeated patters become nested, or embedded within the same sentence. While grammatical recursion is not necessary for language, or poetry, parallelism is a requirement for both.^[6] When grammatical parallelism gets co-opted by syntax and subsumed within a grammatical structure so that self-similar recurrences become self-similar embedding then parallelism becomes recursive. This occurs because recursion is a very efficient way to code and share information.^[7] While parallelism allows for a potentially infinite narrative, recursion allows for a potentially infinite sentence.

Parallelism has been recognized to operate at all levels of language, while analyses of linguistic recursion have primarily focused on the grammatical level. Linguists have tended to only conceive of recursion as operating at the grammatical level, yet I want to suggest this approach is overly constrictive and limiting in its scope. If we think about how recursion might operate at the narrative level, we can see how a poem that operates as a gestalt will necessarily have many narrative aspects that are self-referentially embedded within the poem’s narrative (one might think of Dante’s circles of hell as an example). The elements of a poem do not stand rigid like bricks, laying the edifice for meaning. Not only do the parts of a poem create a compounding of meaning as we encounter them sequentially, but as we analyze these parts, we learn to recognize them as referential signals that indicate they are part of a larger structure and that point to themselves as meaning embedded within larger meaning. It is in this sense that recursion is not necessary at the level of grammar but is a requirement at the narrative or structural level of a poem in order for a poem to function as a gestalt. From the standpoint of the reader, when we read a poem, although we read each line or piece sequentially, in parallel, as we construct our interpretation we do not think of the segmented parts as if we were constructing a house, looking at the function of each piece individually and trying to map it onto a blueprint; we read a poem the way a flower blooms, each stage giving way to the next, the various elements always imbedded within the whole structure as it emerges.

7 Poetic ontologies: poems as metapragmatic symbols

Because poetry is imbedded within culture and tradition, it becomes a framework for expressing beliefs and values. This means that poetry functions in a manner that Michael Silverstein has called metapragmatic. To understand what Silverstein means by metapragmatic, it is necessary to first make a distinction between semantics and pragmatics. Greg Urban (2006: 90) explains that “‘semantics’ refers to the explicit meaning of words” whereas “pragmatics refers to the meaning conveyed by speech that must be inferred from context and paralinguistic features.” A linguistic understanding of the pragmatic aspect of poetry, then, refers to its meaningfulness within the context of its performative communication. Thus, if we consider ways in which pragmatic aspects of language might be referential, then we can see that, as Urban (2006: 90) writes, “‘Metapragmatic’ refers to linguistic signs that are about the pragmatic code, about how to interpret the extrasemantic meaning encoded in speech.” The metapragmatic function of language, Silverstein (1993: 17, 39) notes, can be either explicit (e.g., gossip or reported speech) or implicit (e.g., speakers or listeners who attempt to contextualize the pragmatic aspects of a poem in order to appropriately interpret them). To say that poetry functions metapragmatically is to suggest that to engage with a poem is inherently a linguistically reflexive (and reflective) activity in which one cannot help but be concerned with the appropriate use of language.^[8] Luke Fleming and Michael Lempert (2014: 490) explain, “Poetry here does more than unitize and make comparable chunks of discourse: it helps fashion discourse into a metapragmatic icon, an image of an act imbued with cultural value.” While Fleming and Lempert are correct in their assertion that poetry is “imbued with cultural value,” and that knowledge of this aspect of poetry makes poetry function metapragmatically, their claim that poetry “helps fashion discourse into a metapragmatic icon” misinterprets Silverstein’s definition of metapragmatic. Metapragmatics, according to Silverstein, is either a type of discourse or a functional relationship – it is not an icon or an image “of an act imbued with cultural value,” it is a reflection on or about such cultural value (Silverstein 1993: 38–39). If we were to consider a poem as a metapragmatic object, a linguistic object whose function is in part a metapragmatic reflection on its cultural value, it would necessarily be a metapragmatic symbol, a complex sign with no clear or direct connection to its referent. Fleming and Lempert’s main point is that poetry is a ritualized interaction, in which the whole poetic event (rather than its individual parts) functions as a culturally meaningful symbol. The key element here is that poetry is formally structured in a highly distributed and patterned manner; these patterns emerge by way of likeness and difference among segments of a poem, and this dispersed patterning produces meaning that is cumulative and provisional (Fleming and Lempert 2014: 494).

Metapragmatic language is essentially language that makes one consider the ways in which it is culturally useful or valuable. Poetry, as a form of ritual language, always points to itself as non-conventional language and as imbued with culturally determined value, and thus all poems function metapragmatically to some extent. We cannot extract poems from their cultural context. As a metapragmatic symbol, the meaning within a poem is compounded, emerging as a unified whole through the parallel and disparate parts of the poem.^[9] Yet, as an explicitly performative act, poetry is provisional, as each performance opens the poem to new interpretations after the fact.^[10] Furthermore, a poem never functions alone; as Foley has shown, poems operate within traditions (even poems that work against tradition necessarily acknowledge the tradition or traditions they are working against). This means that poetry always occurs within a context (hopefully the discussion of genre should have already made clear why this is), and as a result a poem cannot exist in a vacuum unconnected to other poems; poetic performativity, writes Fleming and Lempert (2014: 503), is “an artifact of a highly distributed and configurationally semiotic assemblage.” Poems not only speak to us, but they speak to each other, and to culture at large.

Poems speak to us by means of each reader’s (or listener’s) active agency in re-creating a poem’s meaning through the performative act of reading and interpretation, they speak to other poems by means of their genre relations as well as their influential effects on poets (and thus on subsequent poems), and they speak to culture by means of their metapragmatic functions and cultural significance. As Silverstein writes,

Any event of metapragmatic semiosis stands in a stipulative or regimenting relationship to some discursive interaction that is understood thereby to constitute an interactional text in some respect or manner. The discursive interaction is modeled as consisting of such-and-such event structure(s), events being instantiable in causally contingent/consequential occurrences of indexical signs. (Note also that insofar as metapragmatic semiosis is itself an event, it implies, another, superordinate stratum of metapragmatic–pragmatic relationship.) (Silverstein 1993: 48)

As events of metapragmatic semiosis, poems are inherently relational and interactional. Their relations and interactions are situated within the event of their performance, which is itself imbued with cultural meaning. In this light we might say that poems speak to and through readers or reciters via their performative arena and their event structure, and are thus interactional by means of their artistic and aesthetic influence as well as their openness to interpretation.

Tavárez (2014: 518) suggest that this dynamic, in which poems seem to extend beyond themselves as complex matrices of meaning, enables poetry to function as a bridge between collective ritual and private ritual, in which “ritual” can be understood as “the repetitive and highly creative performance of multilayered symbolic acts by individuals or groups in order to secure a number of pragmatic aims.” Citing Clifford Geertz’s “poetics of power,” Tavárez (2014: 521) goes on to explain how meaning operates within ritual language, and argues that ritual language creates “discursively enacted representations of the world [that] are not mere mimicries of the universe, but domains with their own ontology.”^[11] This has clear ramifications for poetry because it implies that poems, as a specific type of ritual language, may represent or imitate aspects of our world, regardless of whether or not poetry is viewed through a mimetic lens. Yet each poem contains its own individual ontology and thus possesses a unique metaphysical epistemology.

As a world-making process, poetry is inherently ontological. Some theorists, such as Amie Thomasson (1999) and Anna Ribeiro (2015), have suggested that each genre of art possesses its own unique ontology, such that all poetry is characterized by a unique ontological framework. Despite offering universal ontologies for specific genres of art, such theories are often heterogeneous, acknowledging that individual poetic texts will inevitably possess their own internal ontological structure (McHale 1987). John Crowe Ransom (2003: 43), for example, acknowledges, “A poetry may be distinguished from a poetry by virtue of subject-matter, and subject-matter may be differentiated with respect to its ontology, or the reality of its being.” This position, that poetry possess a unique and variable ontology, tacitly admits what Tavárez argues openly – that each poem contains its own ontology. Peer Bundgaard and Frederik Stjernfelt (2015) claim that this is not only true for poetry but for all works of art, which as aesthetic objects each possess their own ontology, phenomenology, and semiotics.

Not only do poems have their own unique ways of being in the world (structurally, relationally, etc.), they also each contain their own unique domain of reality, knowledge, and truth, their own simulated existence that blossoms with each performance. Because all poetry is performative, an emergent negotiation of meaning between the performer and the audience, a poem, when performed, converts speech into ritual, making poetry not only performative but transformative as well.

8 From oral poetry to written poetry: poetic universals

What hopefully has become apparent is that contemporary, Western, written poetry shares many aspects of oral poetry, and that the majority of poetry throughout the world is in fact oral and not written. In order to recognize how oral and written poetry are similar, as well as a key way in which they affect cognition differently, it is important to examine how written poetry developed from oral poetry. While writing systems were first developed in concert with the rise of agriculture and the need to keep track of goods, it was not long before their use was expanded for other purposes including poetry. David Olson (1994: 68) describes writing as a metalinguistic model of language that has granted people an awareness of the structure of language.^[12] Written language brings words as linguistic artifacts into awareness (1994: 75). Olson does not suggest that oral cultures do not possess any metalinguistic concepts, but that writing greatly expands such concepts.^[13] He cites ethnographical studies of oral cultures and the differences between “straight speech” and “crooked” or “warped” speech to show how oral cultures do possess special discourses that are more open to interpretation than normal speech, but suggests that these cultures do not make a distinction to the same degree as literate cultures (1994: 103). Olson (1994: 241) argues, “It is writing rather than speech which encourages consciousness of the distinction between what I say and what I mean to say.” This distinction is possible in literary cultures because the language of a written text can be replicated with a precision not supported by oral language, and can thus be viewed as an object of study itself (1994: 105). Expanding on Benjamin Lee Whorf’s linguistic relativity hypothesis, Olson (1994: 259) proposes, “writing provides a set of categories for thinking about language.” Writing provides a model for us to understand concepts such as lexicon, morphology, syntax, and grammar. “Writing,” Olson (1994: 242) contends, “gives rise to the idea of writing.” Literacy turns thought in on itself, making thought the object of its own investigation.

Literacy opens the structure of language, and of poetry, to analysis. Such structural models can be applied to both oral and literate cultures, and can help us recognize universal aspects of poetry’s function and construction as metapragmatic, symbolic discourse. Yet how we think about and understand these structures can greatly differ across cultures. One of the seminal concepts in the study of poetic structure is that of parallelism, which Gerard Manly Hopkins helped to develop and which he associated with recurrence of phonemes in sequences of meter, rhythm, and rhyme that Hopkins saw as central to the combination of metrical units (Jakobson 1985: 39).^[14] In his essay “Closing Statements: Linguistics and Poetics,” Roman Jakobson (1960: 386) famously quotes an 1865 paper by Hopkins, in which he writes, “The artificial part of poetry, perhaps we shall be right to say all artifice, reduces itself to the principle of parallelism. The structure of poetry is that of continuous parallelism.” Jakobson further developed Hopkins’ analysis of poetic parallelism, synthesizing work by Edward Sapir, highlighting the importance of both symmetry and dissonance in the construction of parallel structures in poetry. Jakobson (1985: 40) analyzed parallel structures in ancient Hebrew, Chinese, Turkic, and Russian texts, among others, and concludes, “Parallelistic systems of verbal art give us a direct insight into the speakers’ own conception of the grammatical equivalences.” An analysis of parallelism within a poem, Jakobson suggests, reveals the poet’s conception of and navigation between equivalences and discrepancies. These equivalences and discrepancies take place at various linguistic levels – the phonemic, syntactic, morphological, lexical, semantic, etc. – which highlights the interactions between their continuities. The notion that parallelism is central to poetry, and that equivalence is central to parallelism, reveals another universal aspect of poetry: the role of similarity, and thus of analogy and metaphor.

The centrality of metaphor to poetry should not be surprising. As George Lakoff and Mark Johnson (1980) have shown, metaphors are the foundation of our conceptual system. Metaphors are based in physical, embodied experience, specifically, in the recognition of one experience as similar to another (Lakoff and Johnson 1980: 5). Lakoff and Johnson (1980: 57) argue that our conceptual system is metaphorically structured, and that the metaphors that shape our lives are derived from direct physical experience in a cultured world. One of the main consequences of understanding metaphor as based in experience is the notion that “understanding takes place in terms of entire domains of experience and not in terms of isolated concepts” (1980: 117). Metaphors emerge from actively being in the world. Lakoff and Johnson (1980: 247) note, “metaphor is a natural phenomenon. Conceptual metaphor is a natural part of human thought, and linguistic metaphor is a natural part of human language.” We cannot help but think and speak metaphorically, and because our brains are embodied, and all humans are biologically similar, many primary metaphors are universal (1980: 457). Complex metaphors, on the other hand, often arise from primary metaphors, and are culturally determined and propagated by ritual (1980: 234, 255).

Lakoff and Johnson have shown that we organize our experience as a gestalt, an organized whole that is more than the sum of its parts, and this is what makes our experiences coherent. Lakoff and Johnson (1980: 81) write, “experiential gestalts are multidimensional structured wholes.” We segment our perceptions of the world into experiential gestalts, a complex of recurrent properties more basic to our experience than their separate occurrence; these gestalts are themselves comprised of categorical concepts that emerge from experience; and these concepts can be defined by prototypes and by relations to other prototypes. When we have multiple similar experiences, we bundle them together to construct categories, which are organized prototypically. This is one aspect of how genres are formed; poetic genres are prototypical categories of different kinds of poetry. Categories are thus emergent from our experience and are necessary for us to understand and interact with our environment. This makes perfect sense in relation to poetry when we consider that we experience a poem as an ontologically unique metapragmatic text. Each reading of a poem instantiates, momentarily, a mental world. When we experience a poem, we encounter its world as an experiential gestalt.

Perhaps the most common experiential gestalts we are familiar with are rituals, both religious as well as the everyday rituals we enact in our lives. Lakoff and Johnson (1980: 234) make the claim, “Everyday personal rituals are also experiential gestalts consisting of sequences of actions structured along the natural dimensions of experience – a part-whole structure, stages, causal relationships, and means of accomplishing goals.” Making coffee in the morning, going to church, holiday meals, working out at the gym – all of these activities are rituals in that they are repeated, structured practices that we experience as a coherent sequence of actions (1980: 233–234). Writing poetry can be a ritual, as can reading it. Poetry relies on ritual language to help enact its performativity. This is true of other kinds of ritual as well and is arguably a feature of myth. While there is much debate among anthropologists regarding the primacy of and the relation between ritual and myth (Claude Lévi-Strauss has opposed ritual to myth, arguing that they are two contrasting processes), the dominant view currently is that they share common paradigms.^[15] While the oral-formulaic approach to poetry has revealed several universal aspects regarding the purpose and function of poetry, an examination of myth will more fully expose the structure of poetry that we have begun to dissect.^[16] Claude Lévi-Strauss explains that myths are hierarchically structured, not temporally or linearly structured. Constructed like prototypical concepts, “a myth is made up of all of its variants” (Lévi-Strauss 1955: 435).^[17] Myths produce meaning from constituent units that are not isolated relations, “but bundles of such relations” and exist in a mythological time that is at once synchronic and diachronic (1955: 431). Brazilian anthropologist Eduardo Viveiros de Castro stresses the importance of understanding concepts as we do myths when trying to comprehend Amerindian perspectives of experience. Citing Lévi-Strauss, Viveiros de Castro (2014: 21) writes that the basic units of a myth are relational: “A mytheme ‘is always made up of all of its variants’, just as a concept is irreducible to the arguments or propositions expressing it.” Here we can see how, like prototypes, we create myths from multiple similar variants that get bundled into a gestalt. With respect to poetry, we might recognize that in oral traditions, where the same poem may be told slightly differently each time, poems are also structured like gestalts. This is true for written poetry too, once we acknowledge that even textual poetry is performative, and each subsequent reading can never be experienced as exactly equivalent. As Heraclitus might have said, you can never step into the same poem twice.

9 The function of poetry: a tool for thought and communication

While the concept of poetry is far from stable, some commonalities do exist. Though it was not always conceived of as such, poetry is an art; it is itself a genre and also an umbrella term for multiple subgenres; it operates in unique registers; it is performed and is part of a tradition, though every culture maintains its own unique poetic traditions; its core contains both music and metaphor, expressed though language; and each poem is a metapragmatic symbol that contains its own ontology, and is experienced as a gestalt. Many of these elements are structural, but I believe that together they also reveal the primary functions of poetry. Because a poem contains its own temporality and ontology, outside of conventional experience, and because the language of poetry is more heavily symbolic than normal speech, I want to suggest that at its core, poetry attempts to fully express the original, primary functions of language as a means for developing complex thoughts and for communicating those thoughts to others (Young 2023).

For the poet, and for readers too, poetry is a tool for thought because it gives shape to ideas and emotions. Language brings concepts more fully into awareness, and poetry organizes and reorganizes concepts in ways that generate ever more connections and new ideas. Yet poetry is also a tool to communicate these concepts and ideas to others, and to enable others to use the poem as a tool for their own thinking. Poems are performative, and as such they are also communicative. Jane Hirshfield reflects on the many ways that poetry begins with concentration. She explores how a poem can focus our attention outward as well as inward, and she outlines the ways that a poem focuses concentration and creates its own world through its music, rhetoric, images, emotion, story, and voice (Hirshfield 1997: 6–7). Such concentration is both internal, a tool for thought, and external, as the poem is shared and communicated and leads others to concentration. Hirshfield (1997: 32) writes, “Poetry leads us into the self, but also away from it.” Poetry leads us into our own mind, but also outside ourselves, into the mind of the poem. Robert Bringhurst (2006: 143) also recognizes the centrality of concentration to poetry when he writes, “When you think intensely and beautifully, something happens. That something is called poetry.” Bringhurst (2006: 224) sees in poetry the manifestation of patterned thought and meaning.

Bringhurst (2006: 210) sees poetry as something natural to the world, as that which provokes the very existence of language, and as that which all language aspires to rejoin. As we might recall, Tatarkiewicz explains that this understanding of poetry as not only a genre of art but as a quality that exists in nature is historically dependent and is in fact a rather contemporary conceptual development. Bringhurst’s view is not too far from that of Ralph Waldo Emerson (1884: 726), who wrote, “poetry was all written before time was, and whenever we are so finely organized that we can penetrate into that region where the air is music, we hear those primal warblings and attempt to write them down, but we lose ever and anon a word or a verse and substitute something of our own, and thus miswrite the poem.” While Tatarkiewicz has shown that this broad, Romantic view of poetry is a rather new invention, I am inclined to believe it is true, just as I know that while the invention of the concepts for nouns, grammar, and lyric poetry have not always existed, the things themselves have been around as long as there has been language. If thought is so central to both language and poetry, perhaps Bringhurst (2006: 146) is correct when he writes, “People have tried to tell me that language is the source and basis of poetry. I’m pretty sure that’s backwards. Language is what thought and poetry produce. And stories are the fruit that language bears. You and I are stories told in ribonucleic acid. The Iliad is a story told in Greek. Stories are pretty ingenious at getting themselves told.” Although we haven’t always had a word for poetry (and some cultures still do not), as long as we have had words, poetry has existed. And while the function of each poem differs depending on the context of its production and reception, the function of poetry has always been to aid in the construction and distribution of the complex ideas and emotions that make us human.