February 11, 2004

Blogging time

In the course of a post about the factual standards of linguistic journalism, Semantic Compositions remarks:

This morning, Semantic Compositions featured an article from the Boston Globe, which launched him off about the prescriptive/descriptive debate. Mark Liberman, however, sat down with a copy of the Oxford English Dictionary, and just demolished the prescriptive case being made. SC envies Prof. Liberman the time necessary to compile such a post, but feels obliged to answer his rhetorical question: "Can't anybody use a dictionary anymore?"

Though grateful for the plug, I'd like to correct two related misapprehensions: first, that an actual copy of the OED was involved, and second, that researching the post took a significant amount of time.

Here's how it was.

1. Over breakfast early last Monday morning, I read the new links on A.L.D., noted Powers' screed, snarfed the offending section, opened an html file, and pasted it in. Blogging time: about 20 seconds. (I'm not counting the time I spent reading his piece or other things, as I would have done that anyhow). The rest of the morning I spent answering email and participating in a weekly DARPA project conference call.

2. On most Mondays, the computational linguistics students at Penn hold a lunchtime talk series called Clunch. The Clunch Wallahs not only arrange a presentation but also provide a buffet lunch. Last Monday, the presentation was Julia Hockenmeier with a fascinating presentation about using CKY parsing techniques to predict protein folding from amino acid sequences, and the food was from Tandoor India. At noon, I sat down with my rice, saag paneer and naan, unpacked my laptop, logged into Penn's OED subscription via wifi, looked up Powers' three words, snarfed the relevant senses and historical citations, and pasted the results wholesale into the html page that I'd opened earlier. By the time Julia started her talk at 12:20 or so, I'd long since finished both my lunch and the OED interaction, checked my email, and spent several minutes talking with my neighbors. Blogging time: maybe 2 minutes.

3. Julia finished up her talk around 1:30. I had a couple of errands to run, and then I walked home, arriving about 2:30. I did the dishes, checked my email, and then sat down to write the post. I'd thought it through a bit while doing the dishes, and I already had all the quotes in place, so I just had to type it out and then load it into MovableType's somewhat clunky posting interface. I'm a reasonably fast typist, and I didn't do much editing beyond reading it through for obvious idiocies and fiddling with a couple of formatting issues. The post went in shortly after 3:00 p.m., and I turned my attention to an afternoon of actual work. Blogging time: about 20 minutes.

Total time to create the post -- about 23 minutes. Portion of that time devoted to research: about 2 minutes.

If I had needed to consult a paper dictionary and copy the citations by hand, the research would have taken a lot longer, probably half and hour or more. As a result, I probably wouldn't have done it. I can rarely afford to spend more than 10 or 20 minutes composing a post, generally using an odd little shard of time between one appointment or chore and another, when I can't hope to do anything that requires intellectual set-up time.

If there seems to be some real research in one of my posts, it's either something I did as part of one of my day jobs -- some stuff I compiled for a course lecture, or some examples for a paper, or the like -- or it's something that can be done in a couple of minutes, through the miracle of networked computing.

Using online dictionaries is now so easy that I'm reluctant to excuse people like Murphy and Powers for not bothering to check before drawing a line in the sand on word usage -- though I'm a tolerant person.

Posted by Mark Liberman at February 11, 2004 11:10 PM