The software can be downloaded, but web-based versions of these programs work almost as well, which makes them easy for students to access from their personal computers. Using Gentle and Drift to Analyze Poetry: A Primerīasically, Gentle assesses word timing and tempo, and Drift is a highly sensitive pitch-tracker. Below, I’ll lay out a quick primer on Gentle and Drift before explaining how I use these tools to help students contextualize scansion and rethink their approach to gathering data. You can read more about MacArthur’s approach here and here. Gentle and Drift were developed in 2015 by Robert Ochshorn and Max Hawkins using Kaldi, an open-source speech recognition software toolkit that was developed at Johns Hopkins University. MacArthur uses Gentle and Drift, two open-source digital tools that align an audio recording with its transcription. I’ve focused on methods introduced by Marit MacArthur, who pioneered the use of forced-alignment speech software and pitch-tracking tools for identifying patterns in audio recordings of poems. Partnering with Tech’s Data Visualization Lab this spring, I’ve begun teaching first-year students how to use some of these tools to analyze poetry rhythms. Starting in the 1980s and 1990s, some began to explore ways of writing software to track quantifiable patterns in poems, especially syllabic rhythms (for more on the long history of poetry analysis software, check out Setsuko Yokoyama’s fascinating article, “Digital Technologies for Exploring Prosody: A Brief Historical Overview”). The tools of big data have proven useful to poetry scholars. Tech’s Data Viz Lab, featuring its large-format screen display on the right. Georgia Tech’s Data Visualization Lab, centered on its signature giant screen, reinforces this association. Automated software can process large amounts of data faster and more efficiently than ever before. In our age of “big data,” characterized by internet systems that gather way more information than we can easily process, we tend to associate data visualization with computers. Ultimately, teaching poetry analysis within a data visualization framework instills in students a widely applicable lesson: that how we gather and represent data is integrally related to how we interpret it. Pairing scansion with other, more recently developed methods of computerized poetry analysis, I guide students to think beyond data output (all those iambs and trochees, say) and consider how data emerges, turning their eagerness to critique scansion into a general critique of graphic representations of data. The fact that we don’t think of scansion as data visualization-but that students are so ready, even eager, to critique it-makes it an ideal teaching opportunity. In fact, it’s arguably the most famous data viz system in English literature today. Scansion is essentially a method for data visualization, just like a map or a pie chart it’s a graphical representation of the phonemic data that a poetry scholar locates within a poet’s words. For students taking my English courses at Georgia Tech, these flaws are a chance for thinking critically about the “V” for “visual” in Tech’s signature multimodal “WOVEN” approach. But it is the flaws themselves that concern me here. Why teach a flawed system? Because despite its flaws, it has helped poets write some absolutely stunning poems. I believe that poetry teachers have done students a great disservice by teaching this system without helping students explore these flaws. In fact, an iamb is not a stable, essential unit of English speech, as linguists have long observed. Or perhaps they noticed that this binary system, stressed or unstressed, is really quite rudimentary natural speech is never so simple. They had noticed that even so-called “standard” pronunciation varies wildly (for example, is “homage” pronounced “ho-MAGE” or “HO-mage”? In fact, both are pretty common, according to the pronunciation guide in the Oxford English Dictionary). They had noticed all those single-syllable words-like “hips”-which are never consistently stressed or unstressed. Most English speakers say suc-CESS and CIR-cuit SUC-cess and cir-CUIT are practically different words.īut while few of my first-year students start out knowing why this system works, almost everyone has some idea of its imperfections. They had not learned that in a stress-timed language, a listener’s comprehension of multisyllabic words depends on a speaker’s pronunciation stresses occur differently in mora- and syllable-timed languages. Few know that it works because English is a stress-timed language, and it doesn’t work the same way for, say, Japanese or French, which are mora-timed and syllable-timed, respectively. A dash and u-shaped marking are often used to denote stressed and unstressed syllables, respectively.īut few students enter college knowing why this system, “scansion,” was designed in the first place.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |