Explaining Winners and Losers Alike

The idea in plain words

The usual story of a technology contest goes like this. The winner had a better design, or a better strategy, or saw the future more clearly. The loser was too expensive, too complicated, too closed, too early. Each side is explained with a different kind of cause: the winner by its merits, the loser by its faults.

The symmetry principle says this is lazy. If you explain why one design succeeded by its technical superiority, you should be able to explain why the other failed by the same standard, and you will often find that the losing design was superior in some respects too. Usually the difference lies in something else: which users mattered, which problems counted, which alliances formed. A fair explanation uses the same kinds of causes for both.

Where it comes from

In 1931 the Cambridge historian Herbert Butterfield attacked what he called the Whig interpretation of history: the habit of writing the past as a story of progress leading to the present, sorting historical figures into those who helped bring about today’s world and those who stood in its way. The Whig historian, Butterfield argued, studies the past with one eye on the present, and so misunderstands both. He picks out the parts of the past that resemble the present and treats them as the important parts.

Technology history is especially prone to this, because technologies are routinely ranked by whether they “won”. The losers become cautionary tales and the winners become destiny.

In 1976 the sociologist David Bloor proposed a rule for the study of science that he called symmetry. Beliefs that turned out true and beliefs that turned out false should be explained by the same kinds of causes. It is not enough to say that scientists accepted a correct theory because it was correct and rejected an incorrect one because of prejudice. Both acceptance and rejection have causes, and those causes should be studied in the same way.

In 1984 Trevor Pinch and Wiebe Bijker applied the idea to machines. Their famous example was the bicycle. In the 1870s and 1880s the high-wheeled “ordinary” bicycle and the low “safety” bicycle coexisted. Young male athletes saw the high wheel as fast, daring and prestigious; to them, its danger was part of the point. Women, older riders and cautious buyers saw it as a machine that threw people over the handlebars. Pinch and Bijker called this interpretive flexibility: different social groups saw different artefacts in the same object, and the contest ended not when the safety bicycle was proven “better” in some neutral sense, but when one group’s definition of the problem won out. They called that ending closure.

What it does to the mind

The asymmetry is psychologically natural. Once we know the winner, its features look like reasons; the loser’s features look like excuses. This is hindsight working on explanation. It is also a matter of narrative economy: “the better product won” is a short story, while “the product that suited the groups that mattered most at a particular time won” is a long one.

There is a professional temptation as well. Engineers like to believe that technical merit wins, and marketers like to believe that strategy does. Each reads history to confirm their craft.

Where it shows up

The Betamax road is the classic test. Was Betamax technically better? Many said so. Did VHS win because of longer recording time, licensing strategy or rental shops? Each answer implies a different theory of what “better” meant to buyers in 1978. The programming languages essay argues that each language encoded a theory of who the programmer was; languages that “lost” often lost because the programmer they imagined was not the one who showed up in large numbers. The databases essay shows hierarchical and network databases that were efficient on the hardware of their day and lost when the definition of the problem shifted from machine time to human time.

What it asks of someone deciding

When you analyse a competitor that beat you, resist two stories at once: “they were lucky” and “we were worse”. Ask which customers decided the contest, and what problem those customers thought they were solving.

When you analyse your own success, apply the same rule. If you would explain a rival’s identical success by timing or distribution, those causes may explain yours too.

Remember the objection. Symmetry is a method for explaining outcomes fairly, not a claim that all designs are equal. Some things really do work better. The discipline is to show how “better” was defined, and by whom, before treating it as the explanation.