Netting Plagiarists

Well, this plagiarism series has proven so fascinating and even popular that I may just blog on it forever.

Just kidding.

But I will blog on the topic today. Mark Goodacre, whose "editorial fatigue" principle I explained in my Friday entry, has posted an interesting response to my use of his principle.

Goodacre poses a number of questions concerning how to catch plagiarists who are not so naive, the truly clever ones:

(1) It's easier to catch the weak plagiarists who betray their sources all too quickly because they do not think to hide or know how to hide the distinctive phrases of their sources. But how easy is it to catch the better plagiarists, who can see the distinctive, Google-friendly phrases in their source material, and who change words but not sentence structure?

(2) On-line plagiarism is so easy to do that, as Jeffery points out, it is relatively easy to catch. But does this mean that some of our students are getting away with book-plagiarism all the more easily? If someone plagiarizes Sanders, Crossan or Wright, you can bet that we'll be able to find it. But if they plagiarize Howard Clark Kee, are we going to spot it?

(3) Does our satisfaction in discovering the on-line plagiarists mask us from seeing the students who have paid someone $75 to write their essay for them, or who have pulled their essay from a model stock essay for which they have paid $25 access? Google is not going to spot those, and however much we might suspect, are we going to be able to convict?

I have a few responses to these questions. Let me take them in reverse order.

Concerning number (3), how do we catch these cheaters who buy essays?

If a student pays somebody to write an essay, then Google wouldn't help, but if I suspect that a student didn't write an essay, I can retain the essay in my office and ask the student to come to pick it up. I can then ask the student questions about the research process or about the essay's arguments. I think that many such students would demonstrate their ignorance and confess under the pressure of even courteous questions.

If a student purchases an essay online, then in my experience, Google often finds these. Why? Partly because the stock essays recycle arguments in other stock essays, but partly because complete or partial copies of stock essays often appear online and accessible to Google.

Concerning number (2), how do we catch book-plagiarists?

I've found that although this poses a problem for tracking down some plagiarized passages, such as those passages that nobody has ever quoted online, it doesn't stop me from catching a plagiarist anyway because other plagiarized passages in the same student essay do show up online.

About the passages that don't show up online but that I figure are plagiarized, I can tell the student, "We've seen that you've plagiarized already. These other passages also look plagiarized. I suggest that you rewrite them and add the proper footnotes." This assumes that I'm giving the student a second chance, which I usually do since I'm teaching in Korea.

Concerning number (1), how do we catch the really clever online plagiarists, who know to change words while keeping the sentence structure in an attempt to elude Google?

My reply? Well, how much time do you want to spend? Catching such a plagiarist takes a bit more time, but I find them by constructing a search net. A copied passage modified in the way described still betrays its origin because it retains a pattern of expressions and structures from its source.

Take the following actual example of a passage containing plagiarism:

The cross of the vision in the beginning part of the poem is depicted as being lifted high in air presenting the images of light. This light imagery symbolizes the staggering spiritual brightness of God. The speaker becomes a receiver of God's light, an undefiled witness.

I wondered if it was plagiarized, so I began constructing a net to dredge it up from the abyss of the internet.

First, I opened Advanced Google. Then, in the category "with the exact phrase," I typed "The cross of the vision" from the suspect paragraph and clicked "Google Search." Up came only six items. That surprised me. For such a sentence, hundreds of items might have shown up. With only six, I might take a look first, but I also might not. If you do look at this early stage, then click on "cached" so that the text will have the words highlighted.

Anyway, I then backed up the browser to the category "with all of the words" and typed in the two-word expression "lifted high" in double quotation marks so that Google would look for the full expression rather than two separate words. Up came a single item. I looked using "cached" and found a passage with both expressions highlighted.

I then backed up the browser again and returned to the category "with all the words," where I gradually added the multiple-word expressions "in air," "light imagery," "spiritual brightness," and "of God's light" by a process of trial and error.

If an expression doesn't work, then I either shorten it or discard it. For example, I tried "staggering spiritual blindness" and found nothing, so I backed up and trimmed "staggering." The expression "spiritual blindness" did work.

If you play around with this process, you'll get the hang of it.

Finally, I had constructed my net. Look what I dragged in, a passage from Robert V. Graybill's "The Dream of the Rood: Apotheosis of Anglo-Saxon Paradox," Essays in Medieval Studies, Volume 1, 1984:

The cross of the vision, which is lifted high or "towers" in air, presents a double vision to the dreamer. It is "leohte bewundon," twined around with brightness as the arm of a braceleted Viking wench might be with golden jewelry, or "compassed" with shining gold. Here the light imagery crescendos for six lines to create a visual climax of stunning spiritual brightness. Then comes the antithesis. The light of God and the dirt of man are contrasted. The dreamer sees himself as dim and soiled with flesh, dark and loathsome as any coalpit stone, yet paradoxically and miraculously a recipient of God's light, a beholder of wondrous whiteness.

Compare to the plagiarist's attempt to hide it:

The cross of the vision in the beginning part of the poem is depicted as being lifted high in air presenting the images of light. This light imagery symbolizes the staggering spiritual brightness of God. The speaker becomes a receiver of God's light, an undefiled witness.

Not only are the highlighted words the same and in the same paragraph, but they are even in the same order. Note also that the student has altered "stunning" to "staggering" and "recipient" to "receiver," which thus provide more evidence of the plagiarist's source. Interestingly, the student also misunderstood "a beholder of wondrous whiteness" to refer to the state of the beholder rather than of the beheld and therefore described the "receiver of God's light" as "an undefiled witness" despite the original text having already described the recipient as "soiled."

Or has the plagiarist understood better than I?

At any rate, I've sufficiently made my point. Even a clever plagiarist can be caught by constructing a net cobbled together using multiple-word expressions.