Poet Tips Has Hatched

Share on FacebookTweet about this on Twitter

Nearly a decade ago, I asked the question, “Wouldn’t it be great if a service like Pandora [or Spotify] existed for poetry?” A variety of experiments in trying to “teach” computers how to analyse individual poems led me to conclude that this approach was a dead end.

Then one day in the shower, I had the idea that if I could instead “pool” the recommendations of real people about which poets (rather than poems) are similar to each other, I might be able to fulfil that same intent — to help people find their next favourite poet using technology — in a completely different way.

So, I put up a simple website and, one month and 20,000 lines of code later, the idea attracted nearly 100 beta testers to a private “rough draft” of the site. Their input has been critical to shaping the interface of the site, as well as clarifying the essential message.

Today, I am pleased to announce that the public beta — that is, an improved version of the site, but still very much a work in progress — is available for anyone to peruse.

Encouragingly, between the beta testers, and a few curious individuals who learned about it from the @poet_tips Twitter account and my Facebook feed, the site has grown in the past month by nearly two thousand recommendations or “tips”.

Line Graph of Tip Growth Votes by Country Pie Chart Votes by Country Map

That said, we still have a long way to go. This site will only be as useful as the tips and recommendations people are willing to contribute to it. So, if you are at all curious, please do visit and bookmark the site, and come back to both find new poets and to contribute your suggestions as often as you like. Like Wikipedia, this site will always be a work in progress, and the more we build it, tip by tip, the better it will get.

This kind of thing is obviously also a bit of a social experiment. There have already been some interesting moments.

For example, someone already figured out how to use the simple recommendation formula (“If you like ___, you might like ___”) for snark. A poet known for showmanship and media antics recently got the tip that if you like them, “you might also like PT Barnum”.

Clearly, the site will need some moderation and upkeep. Hopefully, enough people will contribute useful tips to drown out the inevitable bit of silliness and “noise”.

In response to the Tweets, two people also tried to report a “bug” with the site not recommending enough women. The site was “seeded” with canonical poets from history, which is male-skewed, but has long since outgrown that initial seed base with real tips contributed by real people.

So, at this point, if there is a “bug” in this crowd-sourced data set, it reflects a very real “bug” in our society. I’d like to think Poet Tips, presenting the opportunity for anyone to respond by contributing and voting on tips, could actually become a “bug fix” for inequality and underrepresentation in the poetry world.

With all of this, time will tell.

So far, I have already made some new discoveries, thanks to the influx of recommendations that have included poets I hadn’t heard of before (but should have), as well as votes on various associations that hadn’t occurred to me (but now make sense). In short, it’s starting to work. And all of us, click by click, can help to make it work better.

I leave you with a short screencast outlining the practicalities of using the site. If you have any thoughts, questions, or feedback — or if this has inspired you to want to help out with the building and maintaining of the site and its growing data set — please do get in touch.

<a href="https://youtu.be/FoWO91ltGjg" target="_blank"><img src="https://www.robertpeake.com/files/2016/03/pt-logo-splash.png" alt="Play Video" class="alignnone size-full wp-image-7600" style="margin: 0 auto; max-width: 640px; width: 100%" /></a>

Visit the Poet Tips website for more.

Share on FacebookTweet about this on Twitter

The Decline of Goodness in Poetry

Share on FacebookTweet about this on Twitter

“You do not have to be good.”

-Mary Oliver, “Wild Geese”

What kind of poetry will people be reading 100 years from now? It is impossible to predict for sure. Yet certain quantifiable trends in the poems published over the past hundred years give a definite indication of where poetry has been, and may give us some clues as to where it is going.

Methodology

As I have said before, aesthetic matters must be confronted on aesthetic terms. In 1968 a team of researchers asked people to rate different words in the English language on various numerical scales, such as the age the person first learned the meaning of this word and whether the word denotes something masculine or feminine. In 2004, another team extended this research, giving us the Clark and Paivio (2004) Norms — a set of 32 different scores for 925 special words (hereafter “Clark-Paivio words”).

Poetry magazine may be considered a bellwether of taste in American poetry, and conveniently has made nearly 3,000 poems stretching from its inception in 1912 to the present day all available online.

I trained computer software to analyse each one of these poems, counting how often a Clark-Paivio word appeared, which happened nearly 23,000 times in the available online corpus of poems. Armed with this large collection data, I then used a strategy similar to that of Michael Coleman Dalvean, creator of Poetry Assessor. I took the averages of the Clark-Paivio word scores across all 32 variables, rolling these up into an overall score for each poem. For example, the Clark-Paivio word with the lowest age of acquisition is “toy” at 1.5, whereas “bivouac” gets a score of 6.7. If a poem used both words, the poem itself would then get a score of (1.5 + 6.7) / 2 = 4.1 for the “age” variable.
Continue reading…

Share on FacebookTweet about this on Twitter

What Can Computers Teach Us About Poetry?

Share on FacebookTweet about this on Twitter

Colossus ComputerThe idea that analysing poetry with computers could teach us anything about the art is controversial. A recent survey I conducted of more than 300 tech-savvy poets confirmed that — while they generally agree that technology has been good for poetry in terms of fostering community, creating networking opportunities, and providing remote learning — they would rather computer scientists keep the ones and zeroes away from their iambs and spondees.

Intuitively, this makes sense — after all, we write poems for people, not machines. Poetry is one of the most intimately human of activities. Yet analytical methods, properly interpreted, can reveal new aspects of poetry that we readers and writers might miss. Blind spots can be corrected, what we sense intuitively can be confirmed scientifically, and computers may indeed help us to see old words with new eyes.
Continue reading…

Share on FacebookTweet about this on Twitter

No Such Thing as Bad Words

Share on FacebookTweet about this on Twitter

“The dose makes the poison.”

-Paracelsus

In response to my recent analysis of the frequency of words used in past issues of Poetry magazine, current editor Don Share issued me a good-humoured challenge:

So, I analysed 395 poems from 13 issues of Poetry edited by Don Share from October 2013 to November 2014.

I was at first surprised to discover that the nature of the results are not substantially different than those of the nearly 3,000 past issues.
Continue reading…

Share on FacebookTweet about this on Twitter

Top “Poetry Words”

Share on FacebookTweet about this on Twitter

Having counted the occurrence of words in nearly 3,000 poems published in Poetry Magazine to create a parameterised random word generator, I am making some other interesting discoveries about these words.

First, as one Twitter user pointed out, the words that come up at each “frequency of occurrence” setting on the generator have their own distinct feel, as if very different types of poets might gravitate toward different clusters of words:
Continue reading…

Share on FacebookTweet about this on Twitter