Tuesday, July 12, 2011

Genealogy and Science

From time to time, an interesting discussion erupts on genealogy mail lists about whether genealogy is a “science.” This is a tough question, and I think I am fully capable of oversimplifying it. It is also a good question, but, I think, the wrong question.

For starters, genealogy is indeed non-fiction. At least, its aims to be. Genealogists aim to ascertain chains of kinship, whether forward or backward in time. We do this by determining parent-child relationships. Hezekiah either was or wasn’t your ancestor.

That said, proof is often difficult. No matter how diligently we adhere to the Genealogical Proof Standard, we often can not state our conclusions with total certainty. DNA evidence can enhance our confidence, but we often must qualify our conclusions with words like “almost certainly,” “probably,” or “possibly.”

“Science” is commonly defined to involve observation, experimental investigation, testing of hypotheses and development of laws about the behavior of natural phenomena. When we think of science, we typically think of the physical sciences such as physics, chemistry and biology. But, broadly defined, the phrase “natural phenomena” can be extended to humans, and thus we sometimes talk of “social sciences” – sociology, psychology, political science, etc.

Most people don’t consider genealogy a science. Perhaps the most important reason is that genealogy is not experimental and is not concerned with establishing scientific laws – that is generalizations about relationships among phenomena that hold true under specified conditions (paraphrased from the American Heritage Dictionary, Second College Edition). History comes a bit closer to the extent that historians try to understand patterns in social, political or economic events, but the subject matter is so complex and the conditions so varied that historians find it hard to conclude that if A happens, B will follow.

So, genealogy is not a science. Who cares? For me, the more interesting question is: how do genealogical research methods compare to scientific method. Genealogists don’t run experiments in the same way scientists do, but they do observe data, develop hypotheses and draw conclusions. Many of the same rules of logic apply.

A famous philosopher of science (yes, there is such a thing), Karl Popper, said that a theory can not be scientific unless it is falsifiable. The basic idea is that, because it is so difficult to prove many things with certainty, we can approach truth by proving alternative theories false. Scientists posit a “null hypothesis,” and then set out to disprove it. If evidence suggests rejection of a null hypothesis, then the theory can be accepted with some confidence. For example, if a researcher hypothesized that all sheep are white, but null hypothesis would be that some sheep are not white. Because a researcher can never be certain that he has observed all sheep in the world, he can never prove that all sheep are white. That is, the all are white theory is not falsifiable. However, a null hypothesis that all sheep are not white can be proved with a single observation.

This principle applies to genealogy. Our research might lead us to hypothesize that the Hezekiah Hercules of was the father of Herky Hercules. We may never be able to prove the relationship, but we might be able to disprove it. If we can reject a series of hypothesized relationships, we may be left with only one plausible relationship standing.

Another principle of science is that of conceptual economy – that is, a preference for the simplest explanation among a set of competing theories with comparable explanatory power. Any particular theory may eventually be rejected, but this principle advises not making explanations more complicated than they need be. This problem arises often in genealogy. Many possible explanations, some of them Rube-Goldberg-complicated, may be available to explain why Great-Grandma Grace lived with her Aunt Alicia, but don’t dream up more complicated explanations than necessary. Unless you can disprove the simple explanation, it likely is true.

So, no, genealogy is not a science. But both scientists and genealogists make careful observations, develop theories about relationships, and draw defensible conclusions. This is what matters: skilled genealogists do their research with much the same care as scientists. I think we should worry less about whether we are a science, and more about how we do our research and write our proof arguments.