It’s not fresh news, but I still want to mention this year’s “Banished Words” as compiled by the good people at Lake Superior State University in Sault Ste. Marie, Michigan. They’ve been at this (admittedly silly) exercise since before most groups began naming a “Word of the Year” — the 2016 list was LSSU’s 41st swing at the curve ball that is our ever-changing vocabulary. As usual, they provided a few interesting things to think about.
This year I’ve done something different with LSSU’s list: I’ve run some Google Ngram analyses on the top half dozen of these exiles-to-be. My goal was not to find evidence that banishment is undeserved, but rather to quantify the irksomeness (to some) of these words compared to similar terms. For example, are these words really used as often as critics complain? Is their use a legitimately new phenomenon? Interesting things proceeded from this.
Let’s look at the first candidate for banishment: use of “So” at the start of a sentence (I’ll call it “prefix so“).
It’s an understandable choice. I’d noticed myself picking up this habit in less-formal writing a couple of years ago. You can probably find uses of it on this blog with little effort. It crept in at first, usually in correspondence, but then it gained momentum. When I noticed myself using it at least once in virtually every email, I realized there was a problem and began to prune it. I let it stay once in awhile, for the sake of variety (and sometimes to stress informality) but it’s very rare when it can’t be replaced with a better word or phrase or be removed entirely.
Within Ngrams, I tried comparing it to similar “word whiskers” that easily came to mind, specifically “Thus” and “Well.” To capture the prefix nature, I ran the search capitalized and with case-sensitivity on.
If you look at the graph this produces, you might notice something surprising: prefix “So” has actually seen heavy use for most of the last century and a half. Compared to the stuffy and academic “Thus” it has been more common since just before the Civil War, with the exception of a period of “Thus” dominance between the end of World War II and the late ’90s. Has the use of “So” picked up lately? Definitely: compared to its 1970s nadir, you’re 50% more likely to encounter it today. But that doesn’t strike me as a particular problem. It’s equally interesting that since the 1850s “Well” has followed a usage curve that’s almost a perfect mirror of “So“, but with lesser volume.
The rants against the use of “conversation” (specifically “join the conversation“) were more difficult to pin down with ngrams. In print, at least, “join the conversation” is practically a unicorn: good luck finding much use, let alone overuse. “Conversation” itself has been on an upward trend…but it’s still seeing only a shadow of the popularity it enjoyed in the first half of the 19th century. The uptick has recently brought it back to about half of its previous high.
One of the nominators of this ban lamented a list of synonyms which “conversation” has replaced. Since he’d done the work for us, I plugged the whole list into Ngrams, and received back what I feel were entirely predictable results: “conversation” rests in the middle of the pack, and it hasn’t seen the greatest increase in use (that distinction goes to “talk,” followed by “discourse“). “Join the conversation” might be an overused phrase in broadcast speech, which people can love to hate, but its overuse in other contexts is a matter of perception, not necessarily reality.
The next word, “problematic,” was one I thought we might see good results with — to match the calls for banishment. For a first run, I paired problematic with problem and — no great surprise — found that few uses of problematic are known before the middle of the 20th century. It began its steep rise in the ’60s. Other than that, this was no help.
Again using proposed synonyms, I fed it into Ngrams along with inconvenient, undesirable, and one of my own, troublesome. A truth appeared here: problematic has pushed those others aside in recent decades, and now sees more use than all of them combined. This isn’t a new development, though: it’s dominated since the late ’80s.
The popularity of “stakeholder” might be too recent for Ngrams to appreciate. The word exists in some form at least to the start of the 19th century, but it was rarely used before starting a steep climb around 1980. It has a strong air of corporate buzzspeak about it, which is probably why it made the list. Ngrams couldn’t tell me much else.
Many seem offended by the misuse of “price point.” This phrase used to have (and still does, for many) a specific meaning. It’s essentially shorthand for “the price we can get away charging without driving away so many customers that we lose money.” In other contexts it might mean “the best price to challenge a competing product.” It was tricky to find useful comparisons for Ngrams to work with, but an enlightening one is price point vs list price vs retail price. Note on this graph how the other two both begin to decline when price point starts its rise in the 1980s. But something odd might be going on here, as the combined use of all three phrases today doesn’t add up to their popularity 40 years ago. (It’s also quite possible that there are other factors involved and I’ve just failed to find good terms to compare.)
I’ll look at just one more here, and it’s a fun one: “secret sauce.” I chuckled when I saw it on the LSSU list, because I’ve recognized it as a cliché for some time. In fact, when a client used it in some material not too long ago I thought about deleting it and re-writing, but they were attached to it and it was a single use, so we let it stand.
Secret sauce can occasionally be found in food-related sources at least to the 1940s. Its older sister, special sauce, turns up at least to the first decade of the 20th century. But you can put a pin on the graph when McDonald’s began using the words “special sauce” in its jingle for the Big Mac: 1974. The Big Mac jingle was so ubiquitous at one time that it was adopted for other purposes: a teacher in my middle school had students learn it in Latin.
McDonald’s, by the way, never really claimed this to be a “secret sauce” and most people would argue that it’s not particularly special, since it’s simply a version of Thousand Island dressing, which has been relatively common in the US since at least the 1950s.
Saucy origins aside, there’s something to be said about the rise of “secret sauce.” It will be a few years before the Ngrams databank catches up, but indications are that it probably became more prominent than “special sauce” at some point within the past five years.
LSSU’s 2016 list contains a number or other terms. Look them over for yourself and decide which ones you’d personally banish. Try your own Ngrams views. While you’re on LSSU’s site, take a look at some of the previous lists, or the complete compilation of all 41 years of exiles. They provide an interesting inventory of perceived overuse and misuse, which now spans roughly two generations of American life. The whole thing is an exercise in pedantry and curmudgeonliness, but there are insights to be had (and plenty of entertainment). It’s interesting, for example, to see terms reappear and change over time (“so” was on the 1999 list, for a different reason), or to watch words migrate from anathema into the vernacular.
= = = = =
Some quick notes about Ngram data (and my use of it) in the above post:
- All my graphs were produced using the generic “English” corpus. You might see differences if you use American English, British English, or English Fiction. This is drawn from English-language printed material: mostly books, with some magazines and journals. It’s extremely useful…but represents only about 4 1/2 million volumes and around half a trillion individual data points. Google estimates this to be about 6% of all material ever published.
- Ngrams only reliably lets you search material through 2008. The submitters to LSSU’s lists might be well-attuned to the most recent trends, but Ngrams data can’t capture that. When I’ve written “today” in the above post, it’s not entirely true: it’s more like “today’s data set” which isn’t capturing most of the last decade. For some other criticism of Google’s ngrams (there’s plenty), start here.