To quantify, to know for certain whether a piece of writing is good: this desire is one of my students. I tell them vague, mysterious qualifications. Does it help you transcend time and space and feel for a moment free of some constraint you didn’t know what there? Does it surprise you or lift a veil that you may see human experience more clearly? Does it have image, rhythm, and envoi? But are these quantifiable? Can we develop a system to give a poem a raw score and know whether it’s good?
The Poetry Assessor is designed to do just that.
After reviewing previous studies, Michael Coleman Dalvean was able to develop a set of criteria to weigh the poems. For instance, one study analyzed a sample of 85 known (“successful”) poems and 85 obscure (“unsuccessful”) poems. Successful poems had fewer syllables per word in the first line, tending toward the monosyllabic. They used more common words and had simpler syntax. Simple, direct poems were more likely to be reproduced in anthologies.
Another study reviewed 100 sample poems from professionals (from drawn from Contemporary American Poetry, Poulin and Waters, 2006), 100 from amateurs (at www.amateurwriting.com), and found that professionals used more concrete words. Amateur poets were more likely to use exact rhymes (rather than approximate, or “slant” rhymes), more alliteration, and less variety in word choice. This second study formed much of the research material of the computation study that has resulted in the score generator.
Using another set of variables—including Negative Emotion, Affect, Psychological Process—and the average age of acquisition of a word, the machine captures the content and the presumed difficulty of the words.
To score a poem, simply paste the body into the textbox and click the calculate button. A number will be generated. It would appear that poems scoring in the negative numbers are closer to amateur writing and those in the positive, more professional:
As an indication of how to interpret the placing of the poem on the scale, Sylvia Plath’s poem “Crossing the Water” scores 2.53 indicating that it shares more characteristics of a professional poem than an amateur poem. On the other hand, John Laws’ poem “There are so many Things” scores -2.05 indicating that it is closer to the amateur end of the spectrum.
Of course, I tried out some of my poems. My poems scored between -1.8 and +3.4, with most hitting 1.5, a median, well below the 2.53 Plath level. The poems on the lowest end of the spectrum are forthcoming in a couple magazines; the highest ranked are not. I’m prone to say that it doesn’t mean anything, but I am biased against a quantifiable world.
I suggest that you try some—either your own writing or some of your favorite poems—and let me know what you think. Is the Assessor accurate?