Error and Ignorance

The knowledge enterprise—Science, Inc., if you will—is often taken to produce something, i.e., knowledge. I think it is important to distinguish two kinds of learning, however, two kinds of coming-to-know. The first is probably the kind we most often think of when we think of science. Here, coming into the possession of knowledge means overcoming our ignorance; we come to know something that we were previously unknowing about. The idea of a scientific "discovery" points at this kind of knowledge production. But there is also another sense in which we gain knowledge, namely, by discarding false beliefs previously held. Here, learning is a matter of correcting our errors.

When Andrew Gelman talks about the "replication and criticism movement" he is drawing much-needed attention onto the part of science that corrects errors. Some scientists have in fact begun to work in the opposite direction, pushing back back against replication as a norm of scientific inquiry. Andrew and I think this demonstrates a measure of impatience. It is clearly done in the spirit of the ignorance-conquering conception of science, as if science should always be pushing its inquiry into the darkness of the unknown, coming back with new knowledge of how the world works. This eagerness to make entirely new discoveries forgets how much we are confused and confounded in our understanding of the world by the false light of mistaken notions. Some of these notions are age-old folk ideas, some are stubborn prejudices. But some are simply the often entirely honest mistakes of our scientific predecessors. It does them no dishonor to show that they were wrong.

I normally use a gardening metaphor to make this point. We can easily imagine people who want to focus on the "positive" side of the business: planting, watering, fertilizing the flowers. It would be absurd, however, to suggest that weeding fails to make a "contribution".