A database as a method of raising typological questions about poetic form

For many languages, there are ‘verbal arts’ (written, spoken or sung), which
involve additional regulation of language and other kinds of structure. Certain
kinds of regulation are cross‐linguistically common, including division of a text
into lines and sub‐ and super‐line structure (verse), counting and rhythmic
organization of syllables or similar units (metre), repetition‐with‐variation at the
word or higher syntactic level (parallelism), repetition of parts of syllables
(alliteration or rhyme), control over the placement of word boundaries (caesura
and bridge rules), etc. There are no general typologies relating to these poetic
forms (at most, there are partial surveys, usually in the form of histories).
What might a general typology for a very large number of poetic (including song) traditions from around the world tell linguists? We would understand how widespread certain language‐based poetic forms are, and whether related forms are generally historically‐culturally linked. We could ask in what ways forms can combine (e.g., kinds of rhyme with kinds of metre), what combinations are not found (Jakobson argued that metre and parallelism were almost never found together), and, when forms do co‐occur, whether they interact with one another or are best understood as entirely independent constraints on the text. With a sufficiently wide range of data we would be able to test hypotheses about how forms differ (e.g., the possibility that alliteration is always local, whereas rhyme can be non‐local). We can also correlate poetic forms with linguistic forms, particularly in the phonology, to test the ‘development hypothesis’ according to which linguistic accounts of poetry assume that the specific forms in a given poetry arise from phonological and other forms specific to its language (such that a poetry is matched to, or determined by the language). We discuss our construction of a (publicly available) prototype database for a number of poetic traditions in a sample set of languages: English, Old Hittite, Italian, Maltese, Khalkha Mongolian, Somali, and Bora. The database is structured as a set of questions, which can be addressed to any poetic tradition; we discuss how the questions are formulated, and how they help us make new discoveries. We look at how the structure of the database itself, and the questions it encourages us to ask, can lead us to new ways of thinking about poetic form, and its relation to language. We show this by looking at the specific case of Tswana and how the database‐procedure makes possible to discover a metrical form which was previously unsuspected for the languages of the sub‐Saharan area (cf. Greenberg 1960). Because the database includes any kind of language‐based regularity in
poetry, it enables a more holistic account of the connection between poetic form
and linguistic form, as well as of the relations between poetic forms, than has
previously been available. The construction of a database, as an empirical
methodology in the investigation of poetries in a wide variety of languages, thus
raises new theoretical and typological questions about the data.