Posts Tagged With: literature

Treaty of Greens: Generate this! (Reflections)

Treaty of Greens (generative poem)

Reflection 1

As a person who isn’t that into traditional poetry, I was less than enthused with an assignment that was labeled as “generative poetry.” In traditional poetry, I typically pick a specific subject or mood and run with that, so I did the same approach with this assignment. At the beginning of this assignment, I was in the process of writing a theorized letter to my CEO (at Walgreens), where I argued the lack of love for cashiers. I was feeling pretty passionate about my job, so that’s how I stumbled upon Walgreens as my subject matter. I planned on creating a poem that showed the virtues and triumphs of cashiers, until I had a terrible day at work. My goal evolved into showing the “dark side” of Walgreens.

The word choice part was pretty easy, at least in comparison to traditional poetry. I decided to have one group of words that portrayed more to my job title, and another category of words that portrayed to the customer. I decided to capitalize words that would show anger or aggression, or other words that could relate to such a subject. For example, I capitalized ASSHOLES because that’s typically something cashiers scream in their head at rude customers. In a different light, I capitalized PATIENCE because it’s something most customers seem to lack. I think by capitalizing certain words this emphasizes certain points in the poem, which seems to add a nice touch. Furthermore, this creates contrast. The verbs were a little bit more difficult to come up with, for some reason I cannot explain. I think part of it is because I was trying too hard to think of unique verbs. I felt that most of the verbs I managed to scrounge up were rather boring and didn’t paint a picture, but I wanted the verbs to relate to Walgreens. I did manage to get a few odd verbs in there as “engulf”, “defecates”, and “delegate”. These are still loosely tied to the job of a cashier, especially at my store, and I think it really puts a twist on the generative poem.

As mentioned previously, I wasn’t crazy about traditional poetry before this, so I wasn’t too excited for this assignment. While I didn’t love this assignment, I did enjoy that I could essentially have a computer create a poem for me, only each time it would magnificently different than the previous time. This poem definitely challenged me to rethink how poetry is composed in general. By creating poetry in this way, through code, it really changed what poetry can be. It helped me see that poetry, whether through code or traditional, follows some type of pattern with words. However, the generative poetry really expands poetry. Instead of having a traditional sentence, that most people would write, generative poetry can create these crazy, enlightening sentences that one would never think of creating. It’s this aspect that has challenged me to really rethink poetry; maybe I didn’t like traditional poetry because of all of the constraints and limitations. This generative poem has helped me see that anything can be poetry; I don’t need to conform to certain poetry idealism’s in order to create a great poem. Furthermore, this assignment has helped me to start to consider that code itself is poetry; it follows a certain pattern, adheres to certain rules, and creates meaning in something.

Reflection 2:

There are many people in society today who don’t believe that this very assignment on generative poetry is not a true form of literature; we could even argue that there is one of those nonbelievers among our graduate course (cough Jason cough). It’s understandable for most readers to first assume that generative poetry is unlike traditional poetry and literature in general, but after studying it and learning the essence behind codes, it can be argued that there really is no difference at all.

In Perspectives on Ergodic Literature Espen Aarseth (1997) argues that cyber text focuses on the “mechanical organization of the text, by posting the intricacies of the medium as an integral part of the literary exchange,” (p. 1). In simpler terms, the computer is not just the medium, it’s part of the text too.  Aarseth further argues that cyber text is no different from other texts because all literature is different for every reader, the reader has to make choices in order to make sense of the text, and a text can only be read in one sequence at a time (p. 2) All three of these standards apply to both the generative poem assignment, as well as traditional poetry or literature in general.

Generative poetry and electronic literature challenges traditional text, but that doesn’t mean that the newly invented literatures don’t qualify as literature. Aarseth writes that “text is something more than just marks upon a surface,” (p. 12), meaning that text is something that creates meaning and allows for the flow and exchange of ideas. In The Semantic Web Revisited, Nigel Shadbolt, Wendy Hall, and Tim Berners-Lee (2006) claim that the Web consists of “documents for humans to read to one that included data and information for computers to manipulate,” (p. 96).  Even if computers are manipulating the text, much like in the generative poem, meaning is still being made by the reader, or even, humans. And then, the same argument occurs: there is a difference between paper and computer texts. But what is the difference? Aarseth argues that “the real difference between paper texts and computer texts is not very clear,” (p. 10) and it is true; other than the medium, what is the difference?  There are obvious subtle differences, like computers run on electric and the words are coded to appear on a screen, but the argument is that this code is literature too. How? Code uses a certain language and follows a pattern in order to create something meaningful to the reader. Codes can change the color of a text or background, among millions of other things. In comparison, the human hand and mind can write poetry with a certain rhythm that displays different emotions. The medium is still literature.

Since we can consider generative poetry as a type of literature with the evidence presented, we must consider what this means for the composition and structure. Aarseth writes that cyber text “centers attention on the consumer, or the user, of the text,” (p. 1), which changes the way that we compose. Instead of composing a poem for a traditional reader, we must begin to consider other options. For example, readers can be users or even co-authors. We must write in such a way that can account for that; the text must be more interactive to allow for the co-authorship. However, this poses a bit of a threat for the “reader”. Aarseth argues that the cyber text reader “is not safe” which means we can argue that “they are not a reader,” (p. 3). Most books are predictable and allow for full control, but with these newly developed ways of writing, more risks are available for the reader. The reader can fail at understanding how to navigate through the text which leads to a lack of understanding.

Understanding then, is linked to interpretation. But not interpretation as we know it. In “What does it mean to ‘interpret’ code,” a blogger writes that interpretation is no longer what it used to be; it’s not that “search for what the author secretly meant,” but rather it is the exploration of “semiotic objects in order to explore culture and systems of meaning.” This definition changes how we view literature; it’s not about that problem or climax, it’s about the meaning behind the text, and the interaction the text has with the medium to create that meaning. Just as words work together on a page to create a narrative, or within a Haiku to show imagery and emotion, words work behind the screens of a screen with code and the computer to create meaning.


Aarseth, E. (1997). Introduction: Ergodic literature. In Cybertext: Perspectives on Ergodic            Literature. Baltimore, MD: Johns Hopkins University Press. Retrieved from  

Berners-Lee, T., Hall, W., Shadbolt, N., (2006). The semantic web revisited.

What does it mean to ‘interpret’ code? (n.d.) Critical Cod Studies.

Categories: Alphabetic Text Analysis, class activities, elit, ergodic literature, evidence, generative poem, images, information architecture, mapping, semantic web, technology | Tags: , , , , , , , , , | Leave a comment

Jason’s Generative Poem

1. My generative poem was about the themes and emotions of my work-in-progress novel, “Manifestation.” The novel is very emotional and involves a lot of tragedy, pain, and death. My goal was to see how those themes translated into this new format.


I chose names of some of the novel’s characters, cities, and other important places. I also chose descriptive words that I hoped would reflect the novel’s themes, including some words I found were among the most commonly used when I did a word search of the novel. By choosing from among the most commonly used words and phrases, I hope that the poem accurately reflects which themes and emotions are the most commonly seen throughout the story.


Some of the randomly generated lines reflect the dark nature of the theme:


Energy screams

believe in the broken

run from the bleeding

Belief whispers

mourn the bleeding

A dream kills.

Belief searches for help.


Others, however, tended to come out a bit more mundane. I think part of the issue there was the lack of control over the order the words went in. If I had a greater understanding of the code (beyond making the minor tweaks I added in), I’d likely change the generator to give me a wider variety of sentence structures. For example, the poem only seems to generate three types of lines:


“Subject verbs noun,” “Subject verbs,” and “verb the adjective adjective noun.”


By contrast, the opening of another poem I wrote in the past had a more varied structure:


“I saw my own shadow today
We sat under a tree
She didn’t have that much to say
Yet still she sat by me”


I (Subject ) saw (verbed) my own shadow (my own noun) today (at certain time)
We (Subject) sat under (verbed) a tree (a noun)
She (Subject) didn’t have(didn’t verb)  that much (adverb) to say (verb)
Yet (Adverb) still (adjective) she (subject) sat (verbed)  by me (by noun)


While some of these structures could be done in the current generative poem with a little tweaking, the code wouldn’t work so well for some of them. For example, the line “She didn’t have that much to say” wouldn’t fit well in the current code structure. If I added “She” in the “above” subject field, “didn’t have” in one of the verb fields, and “that much to say” in the “below” field, I could get the sentence shown above. But when considering my other words in the poem, I could also end up with sentences like “Gabby runs that much to say,” or “Tock didn’t have the school.” Those don’t make much sense to me and I don’t feel they would fit well. Similarly, I don’t think a phrase like “Yet still” would fit well in the code structure.


Without the skill needed to alter the coding itself, I ended up feeling stuck with only the ability to choose words and phrases, not the order they came in.


2. I don’t exactly feel like the generative poem can qualify as “literature,” even based on the ideas of “cybertext” and “ergodic literature” discussed by Espen Aarseth in Cybertext: Perspectives on Ergodic Literature. In the beginning chapter, “Introduction: Ergodic literature,” Aarseth defines “cybertext” as a concept which “focuses on the mechanical organization of the text” (p. 1). A generative poem clearly qualifies, since the code that runs the poem is directly responsible for the way the text is organized on the screen. The coding itself is at least as important as, if not more important than, the choice of words used in such a poem. Without the code, the poem would have no structure, and would simply be a meaningless jumble of random words piled together. The form the code creates, even though it is randomized, is what ensures that the poem can be read and understood.


However, I don’t feel that “understandable” randomly generated lines are necessarily literary in nature. Instead, I view them as more artistic. defines “art” as “the quality, production, expression, or realm, according to aesthetic principles, of what is beautiful, appealing, or of more than ordinary significance. “ A poem, even when randomly generated, can definitely have aesthetic appeal. When reading my randomly generated poem, I definitely see lines that generate emotions and reaction, and the unique combinations are definitely of “more than ordinary significance.” There is also a certain amount of skill involved in the aesthetic process of choosing words and phrases that will blend well together. A poorly chosen word bank will generate a poem without any appeal and which doesn’t provoke any emotions in the reader. Choosing words skillfully, on the other hand, can result in combinations that touch the reader and spark their curiosity and imagination.


Artistic and aesthetic appeal, however, aren’t necessarily literary in nature. It all depends on how the resulting work is interpreted. In the article “What does it mean to ‘interpret’ code?”, the author defines “interpretation” as “the systematic exploration of semiotic objects in order to explore culture and systems of meaning.” The article discusses examples such as the Coca Cola logo, based on signification. According to the article, a cultural studies scholar “does not ask what the Coca Cola Corporation intended by choosing red for the color of its cans and logo or what meaning it hi[d] in the signature-style of the words on the can. Instead, she performs a semiotic analysis on the possible meanings conveyed by those details (the color red, the cursive script) of the can in the contexts of the many surrounding sign systems and discusses what the can and, by extension the company, have come to signify” (p. 7). An example like this shows that an alphabetic text (such as a company logo) can be analyzed not merely by the importance of the words themselves, but instead based on stylistic details and metaphor.


Based on such an idea, I would examine a generative poem based on semiotic analysis of the code structure (such as what types of sentences it can generate), the on-screen display (font type and color, background, etc), and on the words in the word banks as individual objects. This is different from the way I would analyze a line in a traditional poem, where I would consider the meaning the author was trying to convey with the structure, word choice, rhythm and meter, and so on.


To show the difference in the way I would analyze a generative poem versus a traditional poem, I can examine a line from each type of poem that I have written (from the perspective of an outside reader). For example, when I wrote, “I saw my own shadow today,” I might analyze what the shadow signifies, such as whether it represents some dark secret of the writer’s past, and what “seeing” the shadow might mean with regards to the emotions evoked when being faced with those dark secrets.


When analyzing one of my lines from the generative poem, I would focus on different things. When looking at the line “believe in the broken,” I would read the line by understanding that the author separately chose the phrase “believe in” and the word “broken” without necessarily intending they would appear side by side. Therefore, I wouldn’t analyze what it means to believe in “the broken” specifically. I would instead consider the emotional impact of having words related to belief in a poem alongside negative words, such as “broken” and also “screams,” “bleeding,” and “kills.” This would give me a feeling that the author was trying to convey pain, sadness, and troubles, but perhaps with a feeling of hope remaining within.


The key difference here is between specific meaning (facing the emotions provoked by dark secrets of one’s past) and aesthetic emotional response (considering the concepts of pain and hope together). In my opinion, the former tells a story, while the latter does not. Both can generate emotion and provoke a response in the reader, as can any form of “art.” I feel that the lack of a “story” is what separates a literary poem from an artistic one.


It’s very possible that a more advanced work of coding could create a generative poem that I would feel told a story. For example, in Joyce’s (1987) “afternoon: a story” the individual pages remain static; it is only the order in which they are read which changes. This allows the author to express specific events, even though the order of events might change on subsequent readings. The difference is in the degree of control the author retains. More control is needed to tell a story of any coherence; less control shifts a work away from literary meaning and towards purely aesthetic, artistic appeal.

Categories: elit | Tags: , , , | 2 Comments

Create a free website or blog at