Monday, February 13, 2006

The history of a sentence...

I had my first experience with last night. It went well...I was turning in a paper on Aristotle's epistemology in On the Soul for PHIL 142. My paper is officially 93% original.

What surprised me about the originality count is this: TurnItIn appears to work on tracking sentences in a student's paper and how those sentences correspond with other sentences written in the history of the database...I initially expected TurnItIn to work citation-by-citation as well (for instance, the database tracked direct citations that I used from Aristotle's text and connected them to papers that other persons in my class wrote--they presumably used the same quote).

I was amazed to find, however, a segment of one of my sentences in the analysis section of my paper match up with a completely random article published in 2004 about residents in Florida preparing for Hurrican Charley. My point about Aristotelian knowledge was that when I know something, the expression of that knowledge is the same as the expression of the knowledge of when a "tree" becomes "wood." This point is significant for Aristotle because when I know something, I can change the "project" of things in the world, as they fit my own project (e.g., I know a tree on its own terms, but I can also know a tree on my terms--for instance, if I am a shipbuilder and I need material, I can use the wood of a tree to build my ship. Thus I know both "tree" and "wood," and I know the difference between the two).

To illustrate this point, I had a sentence that read as follows:

I know this: I know that when I cut down "tree" to accumulate "wood" to build a ship, I am changing the tree.

This sentence matched up with the following sentence from "Florida Prepares for Hurricane Charley; "American Quest".", The America's Intelligence Wire, August 1, 2004:

MALE: This is birch. This is also a birch. This is alder. It doesn't cause me any pain to cut down a tree, because I know that when I cut down that tree, I am going to plant a number of trees to take its place.

This similarity got me thinking: how is it that every sentence in my paper didn't get tracked to a particular, random article. I can understand how TurnItIn uses this sentence tracking technique to find significant blocks of text that have been stolen and not cited properly.

But, the element of humor in my example comes from the total unrelatedness of the actual words at the level of sentence (I do find it ironic, however, that the article that my sentence matched is actually proving an Aristotelian point, namely that a person can know a tree on its own terms, and also on terms that are helpful to their own personal project).

Was it a particular order of words? Or a particular percentage of words that matched between the two sentences? Oddly enough, only these words matched:


In total, that's eight words in my original sentence ("I" matched twice).

How is it that no other sentence, merely on a word-by-word basis, matched up with other articles or sources? Surely in the history of the database, every word I used culd be matched in a different article. Is it word order? I just happened to match the words "I know when I cut down tree"...I am amazed that no other sentences in my paper matched another sentence written within the database at some point...

This is an excellent lesson in language: sentences can correspond to a large percentage of exactitude while having no common meaning, and no significant correlation for the purpose of an argument...


Anonymous Jess said...

That your sentences did not match others in the database proves the lingustic truth that language is literally infinite. It seems strange that there weren't many matches, but the chances that you'd formulate an idea word for word as another did is mind-bogglingly tiny. That's exactly how this site works, too: the odds are so tiny that any match causes warranted suspicion. And then all the prof has to do is check that match to see if the student has plagiarized.

Question: does TurnItIn now own your paper? If I remember correctly, there are some very serious questions concerning intellectual rights with services like these.

6:35 AM  
Blogger Nick Z. said...

Good question, Jess.

I read through the entire license statement/agreement because I had heard things about intellectual property issues...I've found nothing (and that's only of the confusing clauses I could understand ;)

But, there is reason to think that TurnItIn does not own any student papers. One can click on quotes in your own paper that match other citations in the class, and when such notifications are clicked on, a message to the likes of, "since we do not own this paper, we cannot display its author" appears.

But, that's the best I have got...

11:48 PM  

Post a Comment

<< Home