Zipf's Law: Using The Same 135 Words

New York City

The Second Edition of The Oxford English Dictionary suggests the English language may possess a quarter of a million distinct words. Achieving an exact count is impossible, given the abundance of current use, obsolete use and derivative constructions. It is also likely that English contains more words than any other language. Sociolinguist John McWhorter tells us there have been 6,000 languages over time with another 10,000 dialects - most of which are dead or dying.

The linguist George Kingsley Zipf was credited with illustrating and proving what became known as Zipf's Law in which the frequency of any word is inversely proportional to its rank in frequency tables. That is to say, the more we use certain words, the more use they get. The Brown Corpus contains "only" 50,000 English-language words, but Zipf's mathematical formulations of the 1930s and 1940s demonstrated that nearly 50 percent of the corpus is comprised of the same 135 words.

Zipf's work is a classic case of Power Law. It's inverse logic and Pareto-like distribution underscore how "preferential attachment" works. As in the "the rich get richer" framework, use accumulates more use just as money accumulates more money. It makes sense that we should use such few words to be efficient and effective in daily conversation, conventions and commerce. How many time has somebody dropped a pretentious word in cocktail-party chatter only to silence the conversation? "Who was that creep, anyway?"

Nassim Taleb rightly tells us in The Black Swan that Zipf was not the first to understand this phenomenon, nor is Zipf's Law a "law" in any real sense of that word. Still, its implications for leaders and language are very clear. On the one hand, a leader must achieve concise, unambiguous meaning in language use. On the other hand, leaders seem to use the same words all the time, thus rendering those words relatively meaningless. Boring too! The key is to find the right word-use balance between what is direct and understood and what is interesting, compelling and different from everyone else's cliches and bromides.

With content analysis, CEOs and other leaders can actually determine how much they sound like everyone else in today's language risk-averse, "me-too" business environment. In doing so, they may decide to move fractionally out on Zipf's distribution and embrace a slightly more distinctive spoken and written vocabulary that will help them distinguish their organizations and themselves and extinguish the tired, lazy verbiage found too often in today's business lexicon.





p.s. Patti LuPone is her usual brilliant self, starring here in Gypsy. That girl has some pipes! I first saw her as Evita in 1979 here and, until Sunday, last saw her at Michael Feinstein's club. She is a delightful force of nature.