I toss off at random either a serious or a flippant answer when people ask me why I quit programming back in 1988. My most frequent, only partly facetious retort is, "Because computers don't have a sense of humor." (Or "humour" when I'm talking to English-speakers outside of the U.S.). I may have to find another rejoinder in the future.
According to an August 2, 2007, article, "How to make a robot laugh," that appeared in my morning newspaper, The Globe and Mail, two researchers at the University of Cincinnati in Ohio, Julia Taylor and Lawrence Mazlack, have developed a prototype joke-detection program. At this point, their software recognizes simple puns, but it doesn't yet get more subtle plays on words or other forms of levity.
These are not the only experts trying to figure out how to give computers a sense of humor. Researchers at the University of North Texas in Denton have built a program that analyzes the frequencies of words found in jokes rather than trying to figure out why the jokes are funny.
I believe that the pursuit of knowledge and truth—any knowledge and truth—is noble. ("Knowledge and truth" is redundant, but it is a worthwhile redundancy, considering the gobbledygook that passes for knowledge among the superstitious masses.) Nevertheless, call me overly conservative—or even a stick in the mud if you must—but I'm skeptical about this being the best use of valuable research time and dollars. There are a number of medical researchers and physicists, among others, who, despite the low cost of computing today, are starved for computer cycles to do their computation-intensive research. What's more, every day, very bright people with PhDs who are trying to end hunger, cure diseases, or advance the cause of world peace, among other virtuous purposes, have their worthy grant applications turned down because of a lack of funds. So how does finding a way to tickle a computer's funny bone manage to pass through the research sieve? I'm just asking.
Intrigued, I put on my intrepid-journalist cap and tracked down Ms. Taylor's Web page on the University of Cincinnati's site. There, I found a link to The International Society for Humor Studies (ISHS), which, according to its Web site, is "a scholarly and professional organization dedicated to the advancement of humor research." Who knew that scholars were dedicating their lives to this academic pursuit? Spending your days figuring out how to get your jollies sounds like a heck of a career to me. Where do I sign up?
Following the story trail farther, I found a link to The International Journal of Humor Research, published by Walter de Gruyter, on the ISHS Web site. The journal costs €44 for a single issue, €175 annually for either the print version or the online version, or €185 annually for both. That's way over my budget for these columns, but, fortunately, the journal's Web site includes a free sample issue, namely Volume 18, Issue 1, which was published in April 2005.
The sample contains a paper titled "Analyzing conversational data in GTVH terms: A new approach to the issue of identity construction via humor" by Argiris Archakis and Villy Tsakona. Copyright laws do not allow me to reproduce the paper in full here, but the fair use doctrine permits me to quote a short passage. I'm sure you'll want to rush out and subscribe to the journal once I've whetted your appetite.
Here are the first two sentences of the paper's abstract: "The central aim of this paper is to apply the General Theory of Verbal Humor (henceforth GTVH; Attardo 2001) to conversational narratives and to integrate it with sociopragmatic approaches. We consider script opposition as a necessary prerequisite for humor and its perlocutionary effect (i.e. eliciting laughter) as a secondary criterion for the characterization of a narrative as humorous."
I kid you not. That's a direct quote from a paper on humor.
I tried reading a few scholarly paragraphs beyond the abstract, but I quickly gave up. Instead, I put on Groucho glasses and did a crouch walk around the neighborhood. I thought I'd learn a lot more about comedy that way.
Based on the available sample, the eggheads who write for this publication sound like a barrel of laughs, a regular riot. I'll bet they crack up their guests at intimate soirees. Can't you just hear these people telling a joke? "A Christian ecclesiastic predicant, a religious celebrant of the Judaic persuasion, and a muliebral pietistic sacristan ambulated into a bar. They suffered multiple diffuse axonal injuries and subdural hematomas. That notwithstanding, a prospective jurisprudent advocate passed the bar." Oh yeah, that's certain to have everyone rolling on the floor, laughing their you-know-whats off (ROTFLTYKWO) and involuntarily shooting soup out their noses (ISSOTN).
(Sorry, there are no prizes for guessing how much time I spent and how many different dictionaries and thesauri I had to consult to write the previous paragraph. Suffice it to say, they are very large numbers. And I'm proud to report that I stumped Word's spell checker on four words in that paragraph, which, excluding proper names, beats the humor journal's quote by two words. I only hope I came somewhat close to using the words properly.)
There have also been other less academic, but equally deadpan articles on humor. For example, the March/April 2006 issue of IEEE Intelligent Systems, a publication of the IEEE Computer Society, included an 11-page section titled "Computational Humor." It came complete with flow charts and footnotes. These fun-packed pages sported articles bearing the titles, "Frame-Shifting Humor in Simulation-Based Language Understanding," "Embodied Conversational Agents: 'A Little Humor Too'," "Automatic Production of Humorous Expressions for Catching the Attention and Remembering," and "The STANDUP Interactive Riddle-Builder." Computer people just can't bear to be without their acronyms. In the latter article's title, STANDUP is short for "System to Augment Nonspeakers' Dialogue Using Puns."
Getting back to the University of Cincinnati researchers, why did they want to develop a joke-detecting machine? According to the Globe and Mail article, the researchers said that "robots acting as human helpers will need a sense of cyber humour to make them more acceptable and less annoying."
If they really want to make computers more acceptable and less annoying, I have a suggestion for them. Instead of teaching a computer to laugh at my jokes, the researchers should use their technological talents to figure out how to compress to something less than half a day the time it takes for the operating system to load and let me get to work. Oh, and it would also be nice if, a few minutes after my half-day wait for the operating system to do its thing, I wasn't confronted with a message telling me that Windows has to restart to give effect to the software updates that it just pulled off the Web for the purpose of giving me a false sense of security by plugging a few more of the endless stream of grave security holes in Windows and the application software on my machine.
Another reason why I think that there are more valuable uses of the researchers' time is that, despite my tongue-in-cheek excuse for leaving programming behind, I no longer want my computer to have a sense of humor. I fear that, if it does, it might start heckling me as I write my allegedly humorous columns. I worry that when I key in what I think is one of my funniest lines, my computer speakers will shrilly deliver the words, "You call that a joke? People for the Ethical Treatment of Animals (PETA) and Greenpeace won't let you publish it because, if you do, laughing hyenas will go extinct by their own hands." Or "Don't send that one to MC Press. Cut out the middleman and send it directly to the morgue. It's going to die." Hey, I said that I was worried that my computer would start heckling me, not that it would be good at it.
Possibly worse, my sense-of-humor-enabled computer might remain silent as I create what I consider to be my most comical prose. Just in case that happens, I'll have to adopt and adapt some of the witticisms that stand-up comedians use to deal with such situations. For example, after experiencing silence when I expect my computer to issue a big guffaw, I might type, "I know you're out there; I can hear your disk drive spinning." Or "That's right; save up your laughter and output it to me big time at the end." Hey, I said that I'll need to adapt some of the comedians' quips, not that I'll be good at it.
Before I finish, I need to add a disclaimer. My research did not uncover anything to suggest that it is the case, but it's possible that the work of the University of Cincinnati researchers is a decade out of date because, as Will Rogers once said, "If the world comes to an end, I want to be in Cincinnati. Everything comes there ten years later." Now, if you'll excuse me, I have to stop writing so I can spend some time explaining that joke to my computer.
Joel Klebanoff is a consultant, a writer, and president of Klebanoff Associates, Inc., a Toronto, Canada-based marketing communications firm. He is also the author of BYTE-ing Satire, a compilation of a year's worth of his columns. Joel has 25 years experience working in IT, first as a programmer/analyst and then as a marketer. He holds a Bachelor of Science in computer science and an MBA, both from the University of Toronto. Contact Joel at firstname.lastname@example.org. He hopes that if his computer learns to chortle, it will laugh with him, not at him. He's sensitive and paranoid about things like that.