More Grumpiness About Language Usage in Law Reviews

Mark Tushnet

1. The dangling "as such." Sometimes the phrase "as such" actually does something -- when it refers back to something that serves as an example of what's about to be said, or to some characteristic of the thing referred to that's about to be introduced. ("She was an Estonian. As such, she was especially interested in the travails of the national soccer team.") In law reviews, though, "as such" is almost always pointless throat clearing. I used to blame legal writing instructors for this, but I've located blog posts going back at least to 2006 by legal writing instructors criticizing the usage. So, where are law faculty and law review editors getting this?

2. The false parallelism "sometimes"/"other times." Law reviews now rather consistently use the parallelism "Sometimes .... Other times...." I couldn't locate on-line usage guides on this, but my search indicated that the parallelism is occasionally described as an "informal" usage. And, my search indicated that much more common is what I grew up with -- "Sometimes.... At other times ...." I wonder whether this is just an example of hypercorrection.

3. Possible innumeracy. I'm less sure about this (if it weren't the summer I'd consult with some colleagues on it), but I think that the following, from a recently published article in a top-ten law review, is wrong (describing the concept of linear and non-linear dose-response rates): "The relationship might be linear, in
which case if 1 gram of exposure causes a 1-in-100 risk of cancer, 0.01 grams of exposure should cause a 1-in-10,000 risk of cancer; or, the relationship might be non-linear, such that if 1 gram of exposure
causes a 1-in-100 risk of cancer, 0.01 grams of exposure causes a 1-in-1,000,000 risk of cancer." I think that this isn't right. With two points, you can't tell whether the relationship is linear or non-linear. The first might be a relationship in which the ratio is one-to-one, the second might be one in which the ratio is one-to-10,000 (I think -- I didn't say that I was fully numerate!), but both might be linear. In the second example, suppose you cut the exposure in half. A linear response, given the information, would be that the risk decreases to  1-500,000. A non-linear response might be that cutting the exposure in half decreases the risk to 1-100,000, but that cutting it by 99% (the example given) decreases it to 1-1,000,0o0. (I think.)
Posted 9:04 AM by Mark Tushnet [link]