Each section of text was generated by using the UNIX grep utility to parse the poetics log files for the entire year of 1995. Thus every word in this text was written by persons posting to Poetics, that is, one or more of the approximately 200 subscribers to Poetics. The input string was altered in some cases ("browning" for "brown," for example) to keep the output within the 3 to 100 line range which I chose as optimal for this composition.
Each section represents the unedited output for the grep returns on the indicated expression with two exceptions. (1) The "clear" section was truncated at 41 lines due to an excessive length. (2) Multiple repetitions of header lines were trimmed in the case of repeats due to hits based on threads. Images were retrieved using identical or similar source words through the Alta Vista search software. The first usable (i.e., non-proprietary) image encountered was incorporated into the installation.
The Input Phrases
Details about the input phrases are as follows:
|Title Word||Grep Word||Search Word|
Note: for images, first "search word" + "images" was tried (to force retrieval to contain image files. If this did not work, "search word" + "image" was tried. As a last resort, if nothing could otherwise be found, the search word on its own was tried, even though it meant looking through hundreds of hits. Interestingly, "fox" was the most difficult word to search because of numerous movie and sports hits containing proprietary material.
Date of Generation: 12 November 1996.