I have a habit to peek at printouts that have been left at our common printer for more than a few hours. Yesterday, I found this.
At a first glance, this text reads like a badly written short essay on the applications of logic. But there was no title, no name, no indication of origin. What was it really?
What caught my attention was a combination of two factors. First, my group shares this printer with the Institute for Logic, Language and Computation (ILLC) of the University of Amsterdam, of which I learned from past printouts that they study automated semantic analysis of natural languages. Second, there was not one, but twelve copies of this text next to the printer, without any indication of origin or attribution.
It could have been an essay by a student to one of the ILLC staff, but then I would expect to see the name of the student and I wouldn’t expect twelve copies. The more likely explanation, given the context, is that this text constitutes an exercise, to be distributed to a class of students.
But then, what sense does it make for a teacher in logic (and semantics, I presume) to distribute a badly written text that only barely covers the basics of the topic?
Here I should mention that I did not study in this area, nor do I actually know the person who ordered the printout. So far I know, the twelve copies still lie around next to the printer. So I can only venture a guess, without knowledge of fact.
The guess I can make is inspired from my own experience as a teacher. In the past, I have designed assignments for students where the goal was to learn from the mistakes: by requesting them to recognize and explicitly identify what is wrong in a badly written computer program, students are taught in contrast to recognize what would make it correct. This text above reminded me of this assignment. A badly written essay about logic from a logic teacher would only have value, I thought, if the problems in the form of the essay were relevant to the topic.
So I re-read, and then bingo! If I am not mistaken, every single mistake in the text is a different sort of language error: lexical, syntactical/grammatical, logical, factual, with all kinds of variants thereof.
I was even quite appreciative when I realized the amount of effort made to ensure that all the errors were of different types, some quite obscure. The author of this text probably wanted to highlight that a human reader is readily able to skip over all these errors without even thinking much about them, while a computer would have a terrible time dealing with the situation.
I can see how this text would sustain a few hours of enlightening discussions about the fundamentals of that research area. This teacher is probably good!