Verse is text that is divided into lines (verse lines). One of the subtypes of verse is metrical verse. In metrical verse, the length of the lines is controlled by a set of rules (indirectly, all metrical rules count syllables). The lines in metrical verse are usually subject to other restrictions as well, most commonly restrictions on the rhythm of the line (based on stress, syllable weight, or lexical tone), and/or on a requirement that a syllable in a specific line-internal location be word-initial or word-final (a caesura rule). Some meters also include rules about rhyme or alliteration.
Although verse is probably a universal (see poetic form, universals of), found in all oral or literary traditions, there are poetic traditions without metrical verse, of which perhaps the best known is the Hebrew poetry of the Old Testament, which is based on syntactic parallelism rather than on counted syllables. Metrical verse is found in European literatures (Greek, English, the various Celtic, Germanic, Romance, and Slavic literatures, also Finnish), in Arabic and Islamic literatures (e.g., Persian, Urdu, Turkish, Hausa), and in literatures less clearly influenced by Arabic (such as Berber and Somali), in the literatures of South Asia (e.g., Sanskrit, Pali, Hindi, Malayalam, Tamil), of Southeast Asia (e.g., Thai, Burmese, Vietnamese), and of East Asia (e.g., Chinese, Korean, Japanese). Metrical verse is reported to be largely or completely absent in the poetry of ancient Semitic literatures and of Australia, non-Islamic Africa, the Americas, and New Guinea, but this may just be because researchers have not been looking for it (fieldworkers far too rarely ask questions about the poetics or poetic practice of a culture).
The variety of meters can be illustrated by some examples. English iambic pentameter requires a 10- or 11-syllable line, with even-numbered syllables tending to have stress. The French alexandrin requires a line of 12 or 13 syllables, with the sixth syllable both stressed and word-final. Swahili shairi requires a line of 16 syllables, with the eighth syllable word-final (and no control over rhythm). Greek iambic trimeter requires a line of 12 syllables, with even-numbered syllables heavy (containing a long vowel or ending in two consonants), and the third, seventh and eleventh syllables light (containing a short vowel ending in at most one consonant). Arabic kamil requires a line of between 12 and 15 syllables with a complex rhythmic control (in the shortest line the third, seventh and eleventh syllables are light and the others heavy). Sanskrit sardulavikridita requires a line of 19 syllables with the aperiodic rhythm “heavy heavy heavy light light heavy light heavy light light light heavy (word boundary) heavy heavy light heavy heavy light heavy.” Japanese meters require lines of five or seven light syllables (but permit a heavy syllable to substitute for two light syllables). A genre of Vietnamese pairs a six-syllable line with an eight-syllable line, in which the second, and sixth (and eighth) syllables belong to one tonal class and the fourth to another. Germanic alliterative meter requires between two and four stressed syllables, at least two of which must alliterate.
In literary studies, meter is usually discussed as an aid to interpretation, and less attention has been paid to the theory that underlies the meter than is desirable. The approach to meter most common in such studies is the foot combination and substitution approach. In this approach, a meter such as iambic pentameter is a template made by combining five iambic feet each of which is composed of a sequence of an unmarked syllable followed by a syllable that is marked. The resulting template is matched to a line whose syllables are unstressed or stressed, so that stressed syllables occupy marked positions in the template and unstressed syllables the unmarked positions. For lines that are not fully periodic (e.g., in an iambic pentameter line where the rhythm does not involve a uniform repetition of unstressed-stressed throughout), the template itself is changed by substituting a foot of a different kind (e.g., a spondee for an iamb) to match the stress pattern of the variant part of the line. This approach only describes the actual rhythm of the line, and though it offers a convenient vocabulary for the literary critic, it tells us nothing about the organization of the meter of the line or why some variations are possible in this meter and some not. Most recent theoretical accounts express strong reservations or total rejection of this approach.
Recent theoretical approaches to meter are based primarily on linguistic theory, particularly on the theory of phonology, following the foundational work of Morris Halle and Samuel Jay Keyser (1971). For metrical purposes, most such theories adopt mechanisms that are used in the theory of phonology, particularly the theory of word stress. Following Mark Liberman’s (1975) insight that stress is a matter of the relation between syllables, not a feature of syllables, different approaches explored the use of trees and grids as representations in accounts both of word stress and of metrical poetry (Kiparsky 1977; Hayes 1983). The phonological theory of optimality theory has also been adapted for use in metrical verse (e.g., Golston and Riad 2005). Nigel Fabb and Halle (2008) develop their account of poetic meter from a formalism proposed for word stress by William Idsardi (1992); it groups the syllables with the help of unpaired parentheses both in phonology (word stress) and in lines (metrical verse). While in most approaches the metrical representation is a template built by special rules and then matched to the line, for Fabb and Halle the metrical representation is generated from the line (much as in generative syntax the syntactic representation is generated from the terminal elements, such as words or morphemes).
In metrical verse, as noted, the length of the line is controlled. In most cases, the basic unit of measurement is the syllable. However, in many metrical traditions, some syllables are part of the line but uncounted. In a common convention, a syllable ending in a vowel precedes a syllable beginning in a vowel, but only one of the two syllables is counted for metrical purposes (though both are usually pronounced). The latter fact shows something important; it shows that the grouping and counting of syllables for metrical purposes is not directly dependent on the phonology of the lines. It also poses evident problems for other approaches, such as that of Kristin Hanson and Paul Kiparsky (1996), which attempt to account for variation in number of syllables by referring to the specific phonology of the language.
In Japanese, some Indian meters, and some other metrical traditions, morae are counted, a heavy syllable counting as two morae and a light syllable as one mora. It is often argued that the heavy syllable actually consists phonologically of two morae, but this is not necessary for an explanation of the meter, which can refer just to the syllable as projecting one or two metrical elements. In Sanskrit and later Indian meters, morae are thus counted, but some meters count morae while also controlling syllables: In the gana-counting meters, syllables form (typically) four-mora groups that are respected in composing the line. Some song meters (e.g., Tongan, Ugandan) also use mora counting as an organizing principle, but this may be a secondary effect in song traditions where heavy syllables match two beats, light syllables match one, and the number of beats is musically controlled. Here, a better understanding of the independent metrical status of text and tune is required.
In addition to controlling the length of the line, metrical rules also often control a pattern based on putting the syllables into two classes, one marked and the other unmarked. It is of particular interest that metrical rules differentiate two kinds of syllable but apparently never three or more kinds, even though this greater differentiation is phonetically possible in many languages. For example, in English metrical verse, the only strictly controlled syllables are those that carry main stress in a polysyllable; other syllables, whether stressed or not, are not strictly controlled, and this is why English metrical verse is rhythmically fairly variable (see the discussion in music, language and). This means that as regards the strict regulation of syllable types in English meters, the syllable carrying the main stress in a polysyllabic word is in one class, and all other syllables, whether stressed or unstressed, are in the other class. Yet English distinguishes several degrees of stress in longer words, such as autobiographical or onomatopoeic, and there is no question that in the perception of the rhythm of the line, we perceive more than two degrees of stress. In the quantitative meters of Greek, Sanskrit, or Arabic, syllable placement depends on whether a syllable is light or is heavy. In Vietnamese, there are phonologically six distinct kinds of tone, but the six types of syllable are grouped into just two tonal classes for metrical purposes. It is interesting in this connection to consider alliterative meters, such as the meter of Beowulf; in the normative line with four stressed syllables, the third must alliterate with the first and/or second but not with the fourth. Here, stressed syllables are partitioned into two types – alliterating and not alliterating.
A patterned distribution based on two metrical classes of syllable, such as the heavy and light syllables in Greek verse, is often thought of as the basis of the rhythm of the line. A major way in which theories of meter diverge is in their account of the relation between rhythm and meter. For example, in English iambic pentameter, there is a general tendency for odd-numbered syllables to be unstressed and even-numbered syllables to be stressed, but the actual pattern of stressed and unstressed syllables varies constantly from line to line, thus, lines in the same meter can vary in their rhythm. Some accounts of English meters attempt to explain the full range of rhythmic variation by building statistical tendencies into the metrical rules. In a different approach, Derek Attridge (1982) incorporates rhythm fully into his account of metrical verse, so that meter and rhythm are accounted for by a single theory. In his account, the metrical template also includes elements that match silences in the text (offbeats), thus building temporal notions into the metrical theory. This and similar accounts must cope with the fact that lines with the same metrical pattern can be realized with different rhythmic patterns, and vice versa. If rhythm is not explained by the metrical rules, several types of explanation are possible (and can be combined). For example, Fabb (2002) argues that the perception of rhythmic regularity involves pragmatic processes of pattern matching that are distinct from metrical rules (which govern those aspects of the line that are strictly controlled and, like other kinds of implicit linguistic rules, are not directly perceived). It is also possible that rhythmic patterns might be independently represented, perhaps by grids similar to those found in metrical verse. The link between the metrical form and the rhythmic form of the verse then may fall under a theory of text-to-tune matching.
Related Credo Articles
Metre (Greek, ‘measure’) is a human artistic development of the instinctive response all warm-blooded creatures have to the pulse....
The rhythmic pattern of a line of verse measured in terms of basic metrical units or feet. In accentual verse, as in English, a foot consists...
From the Greek word meaning `measure'. In poetry, metre is the measure of the rhythm of a line of verse, when the line is rhythmically...