<M <Y

June Film Roundup: I guess the theme of June was mixing fact and fiction? I dunno why I feel the need to come up with a theme for all the random movies I watched in a month. This thing is long enough as it is. Here you go:

July is gonna be a huge month at the museum, as their theme for the month is "The American Epic". Movies that might show up in next month's review include Citizen Kane, The Grapes of Wrath, Do The Right Thing, Reds, The Right Stuff, Nashville, There Will Be Blood, and The Night of the Hunter. I'm tired already!

Mashteroids: As my birthday present to you, I present Mashteroids, Queneau assemblies of the IAU citations for minor planets. This showed up briefly on NYCB two years ago, but I've expanded the dataset, improved the sentence tokenization, and created a platform for future Queneaux.

A few samples:


Robert Shelton (b. 1948), nineteenth president of the University of Arizona, chaired the Keck Telescope Board from 1997 to 2000. The book promoted the Copernican system and became a best seller. Besides his scientific work, he is also the author of the well-known popularizations A Brief History of Time and Black holes and Baby Universes and Other Essays.


Named for the province of New Zealand on the eastern side of the South Island. He published his first story in Pilote magazine in 1972 and his first album in 1975. He has written several papers on the history of optics.


Junttura embodies the Finnish mentality to get things done, stubbornly and at all costs. He is also an authority on the poet and novelist Kenji Miyazawa and currently directs the museum at the Kenji Miyazawa Iihatobu Center. "Miminko" is Czech word that expresses the unique stage of innocence at the beginning of human life.

You also got an RSS feed.

[Comments] (4) Billy Collins, Stand-Up Comic (Bonus: How To Write Poetry): For reasons that need not concern us, I recently gave some advice on writing poetry. I don't know anything about poetry, but I was able to derive the most basic advice from first principles: "read a whole bunch of poetry before you try to write some." Adam Parrish knows more about poetry and offered some poetry-specific advice: "get over yourself".

I think a lot of incipient poets get caught in the idea that poetry is somehow about free self expression, and that the best poetry is that which most freely expresses the self—which, of course, isn't true. Poetry is a genre that you have to be literate in and a toolbox that you have to learn how to use.

If reading a bunch of poetry is too much work for you, you should at least take the time to reverse-engineer the findings of this paper by Michael Coleman (also via Adam), which uses machine learning to model the differences between poems written by members of the Academy of American Poets, and poems written by the general public. It gives some clues as to how the genre works and what's in the toolbox. e.g.:

The negative association with the PYMCP variable ‘Rhy’—a proxy for the extent to which words elicit other words that rhyme with the stimulus word—indicates that professional poets use words that are somewhat unusual but not necessarily complex. Professional poems have fewer words denoting affect but more words denoting number. Professional poems also refer less to the present and to time in general than amateur poems.

Run your stuff through Poetry Assessor until you start getting good scores. Now you're a poet! Well, sort of. The machine-learning algorithm can reliably tag crappy poems as crap, but it mainly looks at vocabulary and I don't think it knows about scansion at all. I ran the first paragraph of Bleak House—three ponderous Victorian sentences—through Poetry Assessor and it got a 1.8, making it a decent twentieth-century American poem. (And it's a very good paragraph, but you see the problem.)

I formulated my "read a lot of poetry" advice because that was also the techinque I used to figure out if I had any more specific advice to give. (I don't.) While reading a lot of poetry, I got really into the work of former US Poet Laureate Billy Collins. Collins has written a number of what I would call "NPR poems", poems that you could imagine him reading aloud on NPR, some of which he probably did read aloud on NPR. He's on NPR a lot. And at first glance the NPR poems have more in common with stand-up comedy than traditional or contemporary poetry.

I think it's best to think of the narrator of a Billy Collins poem as a fictional poet named "Billy Collins", a man whose bouts of incompetence and perpetual lack of inspiration are exploited by the real-world Billy Collins. Stand-up comics do the same thing. I became very interested in how Collins is able to use this persona to do serious poetic work through poems that aren't serious at all—again, something analogous to what a good stand-up comic does.

Some examples. I'm gonna start with Cheerios and Litany, two poems I don't really like. These poems are about as confrontational as Billy Collins gets, but it's not because of their subject matter: it's because they're poetry hacks.

"Cheerios" has a Poetry Assessor score of 0.8--barely professional quality. In "Cheerios" the incompetent poet "Billy Collins" keeps trying to launch a flight of poetic fancy using the overwrought abstract language associated with amateur poetry: "stooped and threadbare back", "more noble and enduring are the hills". But he can never get it off the ground because the engine keeps stalling on concrete imagery--the objective correlatives associated with professional poetry. The problem with that is the concrete imagery consists of nothing but different breakfast foods ("waited for my eggs and toast", "that dude's older than Cheerios", "illuminated my orange juice"). So it's deliberately bad amateur poetry interrupted by deliberately bad professional poetry. Just saying it's a bad poem isn't enough. It's bad in a very interesting, bathetic way.

On the other hand, "Litany" has the incredibly high Poetry Assessor score of 4.4. (The maximum score given in the Coleman paper corresponds to a PA score of 5.2.) What's his secret? Collins spends the entire poem blasting out objective correlatives at high speed. Some of them are taken directly from other poems ("the crystal goblet and the wine"), some of them are allusions ("the plums on the counter", "the burning wheel of the sun"), some are original ("the boat asleep in its boathouse"). But as he shoots those images out, he classifies them, like he's working on an assembly line, or brainstorming the poem he will eventually write. "Litany" is the opposite of "Cheerios". Collins is hacking the part of your brain that evaluates poetry, pushing all your buttons with free-floating imagery. It's a bad poem because you don't know enough about the people in the poem to understand what the imagery means.

Some other NPR poems, arranged roughly in ascending order of seriousness:

Pay special attention to "Another Reason Why I Don't Keep A Gun In The House" and "Nostalgia", two hilarious poems that are literally highbrow stand-up comedy. "Gun" is Seinfeld-esque, employing the tricks of modern poetry to take an exasperating everyday situation and blow it up into series of escalating fantastic images. (In case you were wondering, its Poetry Assessor score is 2.2, squarely on the "professional" side.) "Nostalgia" (1.3) is more of a Steve Martin kind of comedy, presenting logically flawed arguments and the dumb things people say when they're arguing on autopilot.

"Nostalgia" escalates not to a punchline--a funny kind of absurdity--but to a reductio ad absurdum, a logical absurdity. That makes it a good transition to two Collins poems that, although they deal with ephemeral topics, are more serious and less jokey. They both deal with words, the relationship between words and reality, and the fact that we're always putting words into boxes that themselves have no relationship with reality:

"First Reader" (2.9) is my favorite Collins poem. I feel like "American Sonnet" is the most professionally composed of his poems, and Poetry Assessor agrees, giving it the highest score (3.2) of any of the poems I tested. (Apart from "Litany", which is a poetry hack.) I tried writing down some analysis but these two are easy poems to appreciate, so I'll spare you. I want to close with two poems that I'm not crazy about as a whole, but which do a really interesting thing in the last stanza: they anthropomorphize individual words.

"Paperwork" shows fictional poet "Billy Collins" not being able to write a poem, dreaming in the end of gaining inspiration from an "ancient noun who lives alone in a forest." "Thesaurus" is all about anthropomorphizing words, but it's not until the end that the words leave "the warehouse of Roget" and take on independent lives, "wandering the world where they sometimes fall/in love with a completely different word."

Anthropomorphizing words is how Collins deals with the fact that poetry is a lonely business: writing things down all day, making sure to use exactly the right word all the time. Who else needs to be that careful about individual words? Stand-up comics, that's who. A punchline and a poem both rely on an unexpected word at exactly the right time. That word, when it comes along, is your best friend.

PS: Minor error in the Coleman paper which confused me when I was trying to convert between the paper's scores and Poetry Assessor scores.

For example, Robert Hass has two poems in the corpus, The Image and Our Lady of the Snows, which score in the high to very high range of .72 and .94, respectively.

Those numbers should be reversed. "The Image" has a score of .94 (PA: 5.2), and "Our Lady of the Snows" has .72 (PA: 1.1)

Reunion: I got a misdirected flyer in the mail inviting Leon Richardson to a high school reunion. Class of 1983. I was not yet in kindergarten in 1983, so I thought I might go and drop hints about the youth serum I'd invented.

On the other hand, the invitation is addressed to "Richardson Leon, or current resident". So I can go as myself. Anyone can show up to this high school reunion! They don't care!

In fact they're probably hoping a few current residents will show up to boost the numbers. The flyer seems acutely aware that high school reunions are increasingly an anachronism in this world of "Facebook, Twitter, and Smartphones", and is really desperate to prove the worth of in-person reunions.

It also informs me that "The bio-sheet deadline is Friday, August 30, 2013." Interestingly enough, that's also what a supervillain recently told the United Nations.

Apo11o ll: To celebrate the anniversary of the first moon landing, I packaged up a project I came up with a while back: Apo11o ll, a generative piece that performs Queneau assembly on the Apollo 11 transcripts (from The Apollo 11 Flight Journal and The Apollo 11 Surface Journal).

Duke: Rog. [Long pause.]

Armstrong: That's one small step for (a) man; one giant leap for mankind.

McCandless: Roger, 11. I have a T13 update for you. AOS Tananarive at 37:04, Simplex Alpha. Readback. If you want to go that way, crank it up, and then you can drive it around and look where you want. Over. 11, this is Houston. And we copy the VI.

Aldrin: Does it look to you like the [garble] the right way? Yes, they were working out - this elaborate scheme.

Collins: Unless you'd rather sleep up top, Buzz; I like - you guys ought to get a good night's sleep, going in that damn LM - How about - which would you prefer? I say the leak check is complete, and I'm proceeding with opening the hatch dump valve.

Aldrin: That enough?

McCandless: Apollo 11, this is Houston at 1 minute. Over.

First Mashteroids and now this? How am I doing all this Queneau space-magic? The answer is simple: Olipy, my library for artistic text generation (focusing on Queneau assembly, because it's the best). Check it out of Github and you'll have everything you need to create home versions of many of my works. It's like my own personal Boîte-en-valise! Want to create something new? Just grab some data and feed it to an Assembler class.

Loaded Dice 2013 Update: I fetched the BoardGameGeek data again, a yearly tradition, and put up another Loaded Dice update.

A few highlights:

If you go to the main page, you can download an amazing 17-megabyte JSON dump of BGG data I've compiled. It includes descriptions and genres for every game in the dataset, and three data samples that convey historical rating data over three years. At this point I feel like I'm adding enough on top of what the BGG API can give you (the historical rating data) that I can make the data dump available without apology.


Unless otherwise noted, all content licensed by Leonard Richardson
under a Creative Commons License.