Analysis of a corpus of poems A corpus of 1000 lines of poetry (ten samples of 100 lines by ten different authors) is analyzed by a computerized connectionist model of poetic meter. The analysis finds that poets use measurably distinct stress patterns and suggests that these patterns may be “fingerprinting” individual writers. Furthermore, the analysis shows that variations in metrical patterns are in accord with the prevailing verse aesthetics of the period in which the poets were writing. Introduction In English poetry, that genre's single most compelling discriminant—that which defines a poem as a poem—has traditionally been its meter. Meter defines the length of the line, and therefore the distinctive appearance of a poem on the page, and establishes, for the listener of a poem, the significant regularity of a rhythm. Whether this rhythm also carries with it the weight of part of a poem's meaning or is used only for a conventional aesthetic effect that invites the reader to take pleasure in its regularity or variations, meter is one of the central attributes of the genre. poetic. While the meter of a poem may or may not be strongly considered by the poem's audience, or by its critics, meter has always been a matter of substantial interest to poets (see Addison [1994]). At any point in a line of poetry a factor in the decision to favor one word or syntactic pattern over another has been the metrical impact of that choice. Furthermore, the limits of choice are not simply defined by a correctness rule like the following: all stressed positions must have stressed syllables, and no unstressed position can have a stressed syllable. Metric variations, which give rise to what Halle and Keyser (1971), and others, have termed "metric complexity" or "tension," are permitted and, in fact, produce much of the interest in the rhythm of a poem. Traugott (1989), for example, when speaking of Auden's poetry, notes that "one can identify a complex metrical design that integrates and enriches the multiple verbal icons functioning at other levels of language" (294). Indeed, poetic rhythm can only work when it destroys the very sense of design it invokes; the extreme position is taken by Shklovsky (1917), who states: "the problem is not to complicate the rhythm, but to disorder the rhythm" (p..