From the Philadelphia Inquirer, March 23, 1982.

Plain Prose: It’s Seldom Seen in Journals

Written by Dick Pothier

If you want to publish an article in some scientific or medical journal, here is some unusual advice from Scott Armstrong, a professor of marketing at the University of Pennsylvania’s Wharton School: Choose an unimportant topic. Agree with existing beliefs. Use convoluted methods. Withhold some of your data. And write the whole thing in stilted, obtuse prose.

Armstrong, who is the editor of a new research publication called the Journal of Forecasting, offered the advice in a serious, scholarly article last month in the journal’s first issue. He said yesterday that he had studied the publication process in research journals for years.

“Although these rules clearly run counter to the goal of contributing to scientific knowledge — the professed goal of academic journals — they do increase a paper’s chance of being published,” Armstrong said.

“Some readers may feel that the suggestions here ... are extreme,” he wrote in his article. “However, they provide a description of many papers published in the social sciences.... It is not by accident that intelligent and successful scientists produce such work.” Armstrong surveyed dozens of recent studies on how articles in such journals get published, and the result, he said, “was rather depressing, if our job is to get that research information out and have the readers benefit from it.”

In one study, Armstrong said, academics reading articles in scientific journals rated the authors’ competence higher when the writing was less intelligible than when it was clear.

In another study, Armstrong said, research papers were mailed to a sampling of dozens of researchers. Half the scientists received a paper that described an experiment confirming existing beliefs; the other half received a paper describing an identical experiment but with a different conclusion that challenged the consensus.

Although the methods used in the two sets of papers were identical, the scientists surveyed generally approved of the procedures used in the papers that confirmed existing beliefs and generally disapproved of the same methods when they were used to contradict what most scientists believed, Armstrong said.

“Papers with surprising results are especially important for adding significantly to what is known. Presumably, the editors of journals want to publish important papers,” Armstrong said. “On the other hand, they are concerned that the journal might look foolish — and so they reject many of the important papers.”

For young academics who wish to be published in such journals, Armstrong said, “the factors that would seem to be a deadly combination would be choosing an important problem and obtaining surprising results.”

Other studies, Armstrong said, indicate that obscure writing helps those who have little to say. And having little to say may also be an advantage, especially if the author withholds some significant data. “This will allow the researcher to continue publishing slightly different versions of the same research,” which Armstrong says is a common practice. Armstrong’s own specialty is the study of forecasting, and the new journal that published his how-to-publish article covers forecasting in many study areas, including economics, sociology, finance, psychology, and engineering.

While his advice was offered seriously, Armstrong concedes that his own how-to-publish article violates every how-to-publish rule he offered. “I tried to pick an important problem, write it clearly, and so forth.... In fact, a number of editors of other journals have seen it and most of them seemed interested. Two of them even asked the if they could publish it in their own journals.”