Sometimes I need a little help turning over the creative engine when starting a new poem. I have developed a tool that helps me to do just that, and am sharing it with the community in case it helps other poets to ignite their muse as well.
Poetic constraints--such as patterns of alliteration, metre, and rhyme--originally served as mnemonic devices in pre-literate societies. Patterned speech is inherently easier to remember, which is why recalling a nursery rhyme is still easier than memorising prose. Stylised forms of language remained in favour long after writing developed, but in the twenty-first century, the only requirement of a contemporary poet is that they somehow end up writing a poem.
The blank page can present the stress of infinite possibility. I find that adopting artificial constraints can often cause my inventive faculties to rise to the challenge, overcome that stress, and get going. One approach is to create prompts involving random words to incorporate into a draft poem. I have swapped many such prompts with colleagues, which adds an extra layer of motivation since we would later circulate what we wrote.
Over the years, I have found that these prompt-driven poems often turn into much more than merely a clever exercise. Sometimes these seemingly arbitrary footholds in language provide a way in to expressing deep feeling. The poem "Father-Son Conversation", for example, from my collection The Silence Teacher, began with a prompt.
It has been hard to find good random word generators for poetry prompts. Despite being an excellent dictionary, the OED on CD-ROM uses a pseudo-random algorithm that often gets stuck in a "rut", cycling through the same patterns of words. Others, like the Random Word Generator online, seemed like they could be improved upon to tailor to the needs of the kinds of poetry prompts I have found most useful. So I made my own.
I wrote software to analyse the nearly 3,000 poems that have been published in Poetry Magazine and made available online through The Poetry Foundation. I designed the system to filter out "stop words" (like "the" and "and") and then count how often each remaining word appears. As a result, I can select words in a range of most frequently to least frequently used in these poems, with very different results. I also incorporated the WordNet database, which lets me pick words by type (adjective, adverb, noun, and verb). Finally, I added in the Clark/Paivio extended word norms database, which is a comparatively small database of only 925 words, but has been enriched by quantitative research conducted on real people about how they perceive these words, which could open up some interesting possibilities in the future. I combined all of this with some other writing "challenges" that I brainstormed along the lines of the ones colleagues and I would issue to one another in prompts.
The result is a parameterised automated prompt generator that I plan to experiment with and tune to be optimally useful in generating different kinds of poetry-writing prompts. I am looking forward to playing with it, and seeing what kinds of poems it sparks. Perhaps you would like to give it a spin too.