Blair Simmons: Computer-Generated Dialogue, In Progress

monkey with glasses sitting at a laptop

Charles O. Hartman, in his book Virtual Muse, insists of computer-generated poetry, “The question isn’t exactly whether a poet or a computer writes the poem but what kinds of collaboration might be interesting”

Hello again! I am incredibly excited to share with you another one of my findings. The last time I posted, I shared with you my favorite computer-generated prose by a program called Racter. For a refresher, my research involves archival research on and about computer-generated literature. Computer-generated literature is anything a human and computer collaborated on to write. This can take various forms and be executed with an endless number of techniques.

Since I am currently working on my own computer-generated literature, I can share some of my experiments with you all.

I am programming in the language Python and using the lexical database WordNet to navigate through nouns, finding their synonyms and hyponyms in order to create a dialogue that sounds akin to conversation. In WordNet, “Nouns, verbs, adjectives and adverbs are grouped into sets of cognitive synonyms (synsets), each expressing a distinct concept. “ In WordNet, words are not defined through definitions, but through relationships to other words. Dialogue that is created through the use of WordNet has a feeling of floating through and around meaning. The following is a short dialogue my program, in its current state of development, wrote based off of Pinter’s plays:

Davies These: I'm just a reed in the wind.

Lally: The wind is shouting.

Davies These: It's a whirlwind.

Lally: A whirlwind slams.

Davies These: A door slams upstairs.

Lally: A door is stumbling the beard that is yearning the accomplice.

Davies These: Ginger beard.

Lally: Ginger beard covers the necklace.

Davies These: Ginger beard.

Lally: Ah yes, ginger, dried ground gingerroot.

Davies These: I had a seafaring background.

Lally: Ah yes, background, the part of a scene (or picture) that lies behind objects in the foreground.

Davies These: I had a seafaring background.

Lally: Ah yes, background, extraneous signals that can be confused with the phenomenon to be observed or measured.

It can be seen in this excerpt that multiple meanings for one noun is often employed. For example, ‘background’ is defined twice, once as “the part of a scene” and again as “extraneous signals.”

I have a long way to go before my program is creating narrative structures, but I find this proof of concept pretty amusing and fun.

One thought on “Blair Simmons: Computer-Generated Dialogue, In Progress

Comments are closed.