I hereby define a Substantial Nit as a small matter of usage that has large consequences for a reader or listener's understanding of a significant point. In my judgment, most common nits, however worth picking they may be for other reasons, are not SNs. To qualify as an SN, a common mistake must routinely lead to significant misunderstandings; for example, I'm not interested in the stray case where the needless apostrophe in "The Simpson's" on the decorative boulder in front of my house might case a space shuttle to crash. By way of positive example, I begin the list with two charter members:
1. Percent vs. percentage points. For a recent example, I recently heard an NPR story that talked about an incidence of a disease rising by two percent. The story went on for a while, and I couldn't believe a two percent increase had created such a big story. At the end, the reporter finally mentioned the numbers--an increase of something like 4% to 6%, or two percentage points, but about fifty percent. No wonder it was a big deal! I find this to be a fairly common error that almost always makes a big difference in meaning.
2. Disinterested vs. uninterested. In a strict, old-fashioned sense, these are not synonyms. Uninterest is lack of interest in the sense of willingness to pay attention. If you stopped reading this entry before now, you were probably uninterested (as well as gravely misguided, of course). Disinterestedness means lack of personal or financial interest for the purposes of fair judging; Consumer Reports accepts no advertisements because it wants to make disinterested judgments. A lot of people now use "disinterested" to mean "uninterested" and "unbiased" to mean "disinterested." I think that doesn't work. For example, I generally think that buying new cars is a bad idea (in strictly financial terms), and I have good reasosn for thinking so. I'm biased against buying new cars, but my bias is disinterested because I derive no significant benefit from the sale of used cars. More importantly, it's important for politicians and journalistic analysts to cultivate disinterest, but it's fine for them to have well-founded biases that they can explain and defend. It's hard to encourage the disinterestedness that's essential to productive discourse if we lose the word for it.
3. "Proof." Contentions and fictions do not prove things. Brad DeLong makes this point today in his memorial post about J. K. Galbraith, zinging the New York Times obituary writers' use of "proofs" to describe the arguments of some of Galbraith's detractors: "Proofs? I know many people who find Becker's and Stigler's arguments powerful ones. I know nobody who would call them 'proofs.'" I come across similarly inexact usages of "proof" and its variants frequently, often in new college students' papers, which sometimes claim that a given text "proves" something about life. For example, Pride and Prejudice might "prove" that women in Jane Austen's time could find happines by defying social convention and holding out for true love in marriage. A novel can imagine something of that sort, and a novelist might be said to argue it, but a novel cannot prove a sociological or historical claim. To insist on more exact usage of the terms of proof is to encourage public discourse to distinguish among the kinds of proof that various situations allow or require.
4. "Beyond a shadow of a doubt" (in a legal context). This is related to and narrower than SN #3. Almost as often as not, I find, news coverage of court cases will slip at least once from the phrase "beyond a reasonable doubt" to "beyond a shadow of a doubt." Wouldn't it be lovely if reporters and analysts routinely used the correct phrase and helped their viewers or readers understand more precisely the legal standard of a "reasonable doubt" in a given case?