By Daniel Hubbard | April 14, 2013
I always like when I can connect my years as a scientist and my work as a genealogist. This week I got a bit of inspiration in that direction. I was reviewing some genealogy that was felt to be questionable. I have had to agree. There were a few key people who seemed wedged in where they did not belong. I could just stop here and discuss how the goal of genealogy is to find the right people, however few they might be, instead of gathering many people however wrong they might be, but I won’t stop here. I want to discuss chaos.
It is interesting to think about what happens in different situations when small changes are made. If a pitcher in a baseball game makes a mistake, it might mean that the ball ends up a few inches from where he intended. If pitching was part of a chaotic system, a small error in the way the ball left the pitcher’s hand might result in the ball doing a few crazy loops over the first baseman’s head before landing in a fan’s beer. How would genealogical research behave if a change is made to a birth date here or a surname there? Would it be a near miss or would a bystander get soaked in beer?
What is Chaos?
Chaos in this case is not total disorder and unpredictability, though that is a meaning for chaos. That is a pretty boring form of chaos. In mathematics and many sciences there are much more interesting versions of chaos called chaotic systems. They are not totally disordered or totally unpredictable but they are very different from how most of what we deal with behaves. If you want to drop a ball so that it bounces up to exactly three feet, you might try dropping it from four feet. It might bounce too high. Next time you might drop it from three and a half feet and it wouldn’t quite bounce high enough. Obviously next time you would want to try a height somewhere in between. You can do that because this is a situation where small changes to the starting point lead to small changes to the end point. The smaller the change you make to start with, the smaller the change you will make to the result. There is no sign of any behavior that anyone would call chaotic.
Some things behave very differently. It would probably surprise no one that the scientific concept of chaos was first discovered while working on weather prediction. A truly crude computer was running a very, very simple simulation of the weather. At some point it was rerun with numbers that it had produced part way through an earlier simulation. Had it been given exactly the same numbers it would have produced exactly the same result. The numbers weren’t the same. They were just very close. They were shortened a bit. By that I mean instead of putting in 0.654321, the number put in would have been 0.654. It was thought that this would be something like the difference between watching a ball bounce after rolling it off a three-foot table and rolling it off of a sheet of tissue paper on top of a three foot table. Yes, the starting point is different but not in any meaningful way.
What happened instead is that though at first the new results looked like the old ones, they rapidly became totally different. Howling north winds became light southern breezes. Chaos had been discovered.
Chaotic systems like weather, air turbulence, some parts of economics, etc. are very difficult when it comes to specific calculations. General predictions may be possible, climate models can work but exactly what the weather will be like more than a week from now is a very, very difficult problem. Predicting with precision when small changes do not balance out is not easy. A middle school student with a few history books can, in the space of an evening, get a rough idea of how a topic evolved over hundreds of years. There might be a near miss in understanding here and there but the gist will be right. Try a similar experiment with a genealogical problem involving poorly documented people and you won’t be able to say the same. Investigating is not easy when precision really matters and small errors can’t even out.
Chaos’s Big Three
There are three things that can be generally said about how chaotic systems change-
- The precise way you start makes a huge difference to where you end up.
- Different paths through the system can be very similar for a very long time and then suddenly do radically different things.
- Even starting points that are very far apart can, and eventually will, result in temporarily very similar situations.
Does anything in that list look familiar? In it’s own way, genealogical research is a chaotic system.
- There are often many different people who might fit while looking for a specific ancestor. As we research we intend to weed out the clearly wrong ones and study the near misses until we are sure that they do not fit. We want the correct starting point for the next step in our research. If, despite our best efforts, we chose incorrectly then we will take a path that isn’t “almost right” like that ball that fell from a fraction of an inch above three feet. We will wander off following a branch of the human family that is not actually ours.
- That wrong family might live in the same place as the right family for a generation or two or three. The people might do many of the same things. Their lives might be very similar. One more step back in time and one family’s immigration is discovered, as well as what their name was before it was Anglicized. The other family would have been found where they had been found before. Those families have suddenly diverged. Your path back through time no longer even resembles what it should have been.
- That research can converge on the same point from vastly different starting points is obvious in genealogy. To be human means to be related. It is not so odd that we discover tenth cousins twice removed that live half a world away. Sooner or later, every path will cross.
Genealogical research may show some of the signs of being a chaotic system but that simply means that you need to be sure of each step. With careful analysis you won’t take the wrong path or leave any spectators soaked in beer.Twitter It!