Submitted by Lisa P. Slayton, Associate Director, Leadership Initiatives Pittsburgh Leadership Foundation
For roughly half a millennium, “robust” has been a slightly literary — but by no means technical — term for “sturdy,” “hardy,” “strong,” and “vigorous.” It has been used to refer to just about every kind of durability and strength, including moral rigor, physical power, architectural soundness, rich flavor, and even loud music.
Not long ago, there was a robust teenage garage band two doors away from me. As I recall, “robust” was not usually the term I or my neighbors used to describe their unique aesthetic.
There are a few fields in which “robust” is a jargon term in the strictest sense — that is, a word that is unlikely to be understood by people outside a given discipline. One of these apparently is wine-tasting, judging from the word’s frequent appearance in the pages of “Wine Spectator.” (It means, according to my personal dictionary, “wine so aggressive I probably won’t like it.”) But the most arcane and technical use of “robust” is in the field of statistics, where it describes mathematical tests that are so reliable they yield reasonably good results even when random events muck up some of the underlying assumptions.
(You wanna see real jargon in action? Here’s a quote from a 1979 issue of the journal “Nature” that the Oxford English Dictionary cites as an example of the word’s technical use: “The ANOVA assumes equality of variances, a condition not satisfied here; however the test is robust to small deviations in homoscedasticity.” I may be wrong, but I think deviations in homoscedasticity, robust or not, are illegal in several states.)
I’ve always admired the statisticians’ use of this word — even though they essentially hijacked it from centuries of use as an elegant but simple term for “strong.” What I love about the statisticians’ adoption of “robust” is that (a) it’s consistent with the earlier, non-jargon definition, in that it describes a statistic that is hardy and vigorous under trying circumstances; and (b) it provides a colorful word for an important concept that really does cry out for a special name, one that carries at least a whiff of significance even for people with no training in statistics. (You may not know how a robust statistic does its job, or even what that job is, but you can easily guess, from the word’s traditional meaning, that the statistic in question must stand up formidably to some tough circumstances.) The statisticians gave their vivid-but-technical idea a vivid name, entirely in keeping with the word’s prior use and Latin derivation (from robur: strength). I’m all for that.
The trouble with “robust” isn’t in the realms where it’s used as real jargon, but in the thoroughly un-technical way it is bandied about by people who use it merely as a kind of verbal shoulder-padding — something to make them look burly and tough, even though all they’re really doing is whispering sweet nothings. Most of the uses of “robust” that I see in the philanthropic and nonprofit world are just expressions of approval, dressed up in a strutting, tough-guy facade. When people refer to a “robust description” of some project, most of the time they mean nothing more than that it was a good description. When they say an evaluation yielded “robust” results, they’re usually not referring to the statisticians’ criteria (the results withstood lots of variation in the peripheral variables), they just mean the results made them happy. When you hear foundation officers tell you their new initiative is really “robust,” ask for a definition. Most of the time, I’ll wager, they simply mean “we’re spending a lot of money on it.” It’s easy to see why they wanted an intimidating, aggressive word to conceal such pedestrian notions.
“Robust” is a really nice word. It suffers not from being ugly jargon, but from being overused, forced into debilitating overtime labor, a muscular word applied to puny ideas. For those who are becoming sick of it, I ask only that you blame the users, not the word.
(Postscript: After completing this note, I came across yet another realm in which “robust” has a technical meaning: software engineering. In 1981, a computer scientist named Jon Postel coined an influential rule he called “the robustness principle.” I wouldn’t dare to attempt a description of what the rule means in the actual practice of software development, but metaphorically, as a general rule of life, it is pure poetry. It’s also an interesting application of the underlying idea of “robustness.” Postel’s rule is: “Be conservative in what you do; be liberal in what you accept from others.”)