The Years Teach Much Which the Days Never Know

In March 1958 the scientist Charles David Keeling began measuring carbon dioxide at an observatory on Mauna Loa in Hawaii. The record he started shows a sawtooth line that rises and falls each year as northern forests take up and release the gas, and beneath the saw runs a slow climb. No single day or month could have shown the climb, since the seasonal swing is larger than the yearly increase. It took years of patient measurement to reveal a pattern that is now at the centre of world politics. The topic, a line from Ralph Waldo Emerson's essay Experience, says that the years teach much which the days never know. The essay asks what exactly the years teach and whether the lesson is reliable.

The simplest case is the difference between weather and climate. A cold week in July proves nothing about warming, and a hot week in January proves nothing in the other direction. Patterns that matter in economics, public health and ecology have the same structure. An unusually good monsoon does not end a drought, a single month of strong exports does not make an economy, and one year of low cases does not show that a disease is beaten. Those who make decisions from the news of the day are likely to be wrong in both directions.

Psychology has built its own long clocks. The Harvard Study of Adult Development began in 1938 with a group of Harvard students and a group of boys from poor Boston neighbourhoods, and it has followed them and their descendants for more than eight decades. Its directors have reported that the strongest predictor of health and happiness in old age was not wealth, fame or even cholesterol at fifty, but the quality of close relationships. The Dunedin study in New Zealand tracked a thousand children born in 1972 and 1973 and found that childhood self control predicted health, wealth and criminal records decades later. None of this would have been visible in a year, and much of it surprised the researchers.

Michael Apted's Up films offer a popular version of the same lesson. Beginning in 1964 with fourteen British children aged seven, the series returned every seven years to ask the same people the same kinds of questions. Viewers watched a boy who dreamed of one future and a girl who feared another become adults whom no one could have predicted from the first episode. The films showed that character is neither fixed nor free, and that the surprises of a life lie between the outlines of childhood and the choices of later years. Institutions keep similar records. India's census, held roughly every ten years since 1881, supplies changes in fertility, literacy and migration that daily life never reveals.

Ordinary people also learn by waiting. Parents notice only at the end of a decade what a child's early habits meant, and teachers learn which of their students mattered when the grown students come back to say so. Employees discover that the manager they resented was building their skills, and children understand their parents' rules when they set rules of their own. Gandhi's autobiography, written in instalments across the years, is a long reflection on events whose meaning he saw only gradually. The days give vivid detail, and the years give proportion.

The days give vivid detail, and the years give proportion.

Yet the years have their own way of distorting. Memory is not a recording but a reconstruction, and each retelling alters it. Daniel Kahneman and his colleagues showed that people judge a past episode largely by its most intense moment and by how it ended, and not by its length or its average, a pattern they called the peak end rule. In one experiment patients who had a longer colonoscopy that ended gently remembered it as less unpleasant than patients who had a shorter procedure that ended abruptly. Experiences of equal pain were remembered differently depending on how they closed. If memory behaves like that, then what the years teach may be a story that the mind has edited, and the lesson may be about the editor.

There is also a different objection, which is that the days teach things the years forget. A sick child's face, the smell of a classroom and the mood of a town on the night of an election are lost to retrospect, which reduces them to an outline. Historians know that contemporaries often saw what later accounts lost, such as the confusion of the moment and the possibility that things might have turned out differently. Hindsight makes the past look inevitable and so makes it seem easier to judge than it was.

Neither objection undoes the claim, but both sharpen it. The years teach best when they are recorded. Diaries, ledgers, photographs and data series keep the days intact long enough for the years to read them. Keeling's curve is trustworthy because he measured and wrote down every day, and a person who keeps a journal is able to check his memory of a decade against what he thought at the time. The teaching of the years depends on the discipline of the days, and the lesson improves when both are preserved.

Societies need the same discipline. Governments that measure rainfall, disease and learning over decades have a chance of seeing what no election cycle will show. A school system judged by its results this year will tend to neglect what only shows after twenty. The temptation of public life is to answer to the days, to the headline and the quarter, while the problems that matter are slow. Institutions that keep long records and listen to those who remember form a kind of collective patience.

One difficulty remains that the proverb does not admit. Over the years the person who learns is no longer the person who lived through the days. He has changed, and so has the standard by which he judges what he went through. A lesson drawn at seventy about events at twenty is a conversation between two strangers who share a name. It is unclear whether the years teach the truth about the days or only a version suited to the one who has survived them, and no one can step outside the years to check.