Skip to content

Lex maniac

“Investigating changes in American English vocabulary over the last 50 years”

Tag Archives: analysis

February 2, 2021 outlier

outlier

(1980’s | mathematese? | “anomaly,” “exception (to the rule),” “outsider”)

Our definition of this term today owes much to statisticians, but it seems to have come into play earlier in geology and topography. An outlier is part of a formation that is physically separate from the rest of it, due to a fault line or possibly erosion. The OED cites several other meanings, most of which coalesce around the idea of an entity noticeably outside the norm. (One which doesn’t is my favorite: “person who sleeps or lives in the open air, or away from his or her place of business, duty, etc.,” now considered obsolete.) Available in statistical analysis for decades, the word showed up infrequently if at all before 1980 in the press. By now it is ordinary, although I suppose it retains a slightly technical flavor.

My image of an outlier: picture a number of data points sprinkled on a graph. Most of the points cluster together, and it’s easy to visualize a straight line running through the plot. But then there are a couple of points that are nowhere near the line, wrecking your experiment. That’s what “outlier” means to scientists — and to the rest of us too, by now. In 2020, we use it much more often to designate persons, companies, cities, nations than in 1990. But that sort of usage was hardly unheard of. As early as 1994, the word had a vogue in relation to Japan; American economists liked to point out that Japan needed to be treated differently (i.e., more punitively) because it was an “outlier” that didn’t do business like the other nations. In medical insurance jargon, the outlier is a special case of the general definition: a patient who uses an unusually large amount of medical resources (as in “outlier payment,” “outlier policy”), generally due to unpredictable complications.

The way you handle outliers says a lot about your approach to statistics. One way to look at it is that they suggest a failure of observation or a part of the experiment done incorrectly. According to this line of thinking, they are warning signs telling us that further checking is needed. On the other hand, since outliers throw off the data and complicate simple conclusions, there is always a temptation to explain them away or throw them out. In statistical analysis, that may be the right response, but in public discourse it can too easily turn into suppression of dissenting voices or alternative ideas. A lot of outliers don’t, in fact, have much useful to say, but dismissing too quickly anyone who disrupts the meeting will over time cause decision-making to become too insular. Which may be the idea.

It strikes me, though it may not be true all the time, that “outlier” when used of a person conveys a negative connotation. Maybe it’s because it sounds suspiciously like “out-and-out liar.” When you want to commend someone for staying outside the herd, you call them something else — principled, honorable, and so forth. Whether due to stubbornness or incompetence, the outlier’s predicament is self-inflicted.

Every now and then I unearth an “-er” noun with no corresponding verb, and “outlier” is one — it may be that scientists say that certain data points “outlie” the normal distribution, but if so, it hasn’t penetrated mainstream vocabulary. “Whistleblower” is a classic instance. Others: caregiver, doorbuster, headhunter, rainmaker, stakeholder, warfighter. (There are a number of two-word examples, too.) Then there are the past participle adjectives that lack present indicatives, as in “handwritten.” Outliers. Hmph. What did you expect?

Tags: , , , , , , , , , , , ,

August 14, 2014 unpack

unpack

(2000’s | academese | “analyze,” “get to the bottom of,” “delve into,” “explicate,” “untangle”)

This is one of those expressions, like “get with the program” or “reference,” that seems unnecessary, because we already had so many words that meant the same thing and covered the same ground. It is also an example of an old word put to new use, repurposed, as we might say, like “concerning” or “dress down.” One figurative sense of the word was loosely comparable to today’s usage: that of “unpacking” memories, where the mind is likened to a storage locker full of old stories that must be extracted.

“Unpack” remained firmly the property of articles on travel and tourism as late as 1980, but it grew a new meaning by the end of the decade. Suitcases were no longer the thing; now one might unpack a concept, a character, or a government policy. I’ve even seen “repack” used in conjunction with “unpack,” as in this jargon-laden quotation attributed to a U. of Oregon professor in 2009: “Multimedia projects offer an effective platform for students to unpack and repack technical problems such as grammar in a creative way.” I’m not sure what it means, but I think it’s two different methods of explaining grammar to younger students: first you demonstrate the elements and then they use that knowledge to recognize general rules and patterns. Well, maybe. It sounds less convincing when I spell it out.

It’s not easy to tell from LexisNexis where the new meaning came from. The earliest instances turned up in book reviews, articles about theater, that sort of thing. The first time I remember hearing the word used to mean “sort out” was in an English class with E.D. Hirsch at the University of Virginia in 1988. That memory leads me to suspect an academic origin, and the literary locus of most early sightings doesn’t contradict that, although, of course, arts writing is often a conduit for therapese. While there is no reason that “unpack” couldn’t have come out of therapese, I don’t know of any evidence that therapists use it much, even today, much less in 1990. The history of academic fads plays a role in my speculation. “Unpack” came along roughly at the end of deconstruction‘s heyday. The avant-garde had moved on by the late eighties, but they still needed a word that meant “unravel” or “take apart,” because that is what academics do. “Deconstruct” had done the job pretty well for a decade or so, but it had taken on baggage during that installment of the culture wars; in fact, it had started out with baggage, because it was popularized by a French philosopher. It’s not hard to imagine herds of English professors on the lookout for a word that would denote the same action without partaking of the controversy. E.D. Hirsch is no post-structuralist, but he still needed a way to say “let’s break this down.”

“Unpack” didn’t see widespread use until after 2000; by 2010 politicians had started to use it and it had spread well beyond the arts ghetto. Travel writers still need it, of course, and no substitute has presented itself for what you do when you get home from vacation. But the new meaning has firmly established itself. Unpacking is indicated whenever there is involution, when not every facet of a problem or concept or personality is visible at the same time. It is most often a matter of revealing hidden assumptions underlying an argument or a directive (we used to say “expose” or “lay bare,” which applied to the assumptions, not the framework). There is no essential element of mystery involved — in fact, when a professor announces she is about to unpack something, she already knows the point she intends to make — but complexity is necessary. Does “unpacking” necessarily suggest exposing deception, or at least revealing that which someone else would rather keep hidden? It is true that when one wants to conceal something, one tends to cover it over, whether one is dealing with concepts or contraband. So the term may be used that way, but certainly not always.

Thank you to my friend and occasional colleague Mark from the Bronx for proposing “unpack.” Keep those cards and letters coming!

Tags: , , , , , , , ,