I am fascinated in the large murmurations of starlings. They get my attention whenever I see them outdoors and it is hard to get bored watching uploaded videos of them.
I was pretty young when I first saw them and that would have been in nature since Internet wasn’t around then and videos of them were not very common. It is one of those things that happens randomly and strikes you as awesome that you don’t really think about trying to figure it out or run and get a camera. Now that we all have cameras on our persons all the time and we are itching for any opportunity to film anything, this gets captured a lot more often. I’m grateful for that. I enjoy watching them.
But even though as a young person likening himself to be a beginner scientist and mathematician, it is not the kind of behavior that jumped out at me as something that can be explained as some kind of deterministic behavior or mathematical formula. What struck me was that I was watching a choreographed dance in the sky.
Starlings are not mindless bits of matter. If they feel threatened, they will flee and seek safety. They have a great deal of autonomy in how they choose to spend their time.
There are times that come when they conclude this is a good time for a murmuration. If they see one nearby they will fly to join it. If they are in one and see another in a distance, they will move toward joining the two into one. If a murmuration gets split up, they almost always rejoin but sometimes it is the smaller group rejoining the larger one, and sometimes it is the larger group rejoining the smaller one.
There is a lot of fascinating behaviors of the collective. It is fascinating enough that they never seem to collide with each other. But whole groups will instantly change direction and always the same direction. It is rare that the murmuration would just dissipate as you might expect if they all made independent choices to change directions. Sometimes different groups will converge at different angles and even in head on directions. Not only are there no collisions but there doesn’t even seem to be an sense that it could be dangerous. A couple surprising close calls would convince a few to drop out with an bird-brain equivalent reaction of “that’s enough, this is getting ridiculous”. One a murmuration starts, rarely will birds leave it individually until the entire thing dies down.
The other fascinating thing is that the murmuration keeps a well defined edge where one side has a certain density of flying birds and the other side has zero flying birds.
Clearly this is a behavior that begs for scientific study and there is no shortage of scientific interest in it. Mathematical and computer models have been built that show simple mathematical and mindless equations can reproduce the same kind of behavior. Given the obviously limited capacity of the bird’s brains, a simple mathematical behavior is probably close to what is really going on.
I have a couple objections. While perhaps an existing murmuration can be explained by simple formulas, it doesn’t explain why the birds voluntarily join it and at particular times and places where they seem to choose.
Look at the numbers of birds involved. This is not like disturbing a beehive and being swarmed by bees from the same hive. These birds cover large distances to join the murmuration to get as large as they do. They also seem to choose to stay over a flat surface. One popular video shows them over a body of water and when they turn they turn in the direction of more water. A mindless formula wouldn’t consider the terrain. Even a mindful one would recognize the water is not the most hospital place for these birds if they find themselves in need of landing. They choose to be there because it is a flat terrain, it gives them the opportunity bring the show close to the ground without having to stop.
It is not my nature to concede an animal behavior to a mechanistic mathematical explanation. (But that may also be an admission why I never became a scientist). It is my nature to acknowledge intelligence when something looks intelligent. Murmurations look very intelligent, not only at the individual level but at the collective level.
The best human analogy is a dance but humans don’t dance like that. The closest we come is choreographed dancing that requires a lot of practice and is still prone to irrecoverable mistakes. The starlings are joining their dance spontaneously but it is not like the spontaneous dancing of humans that show in videos where a lone individual starts dancing in crowded public space: everyone joins in dancing but mostly they are just dancing with their own bodies, there is no coordination even at a partner-dance level.
Perhaps a close analogy is that of line or square dancing where there is a caller who announcing a particular move and everyone knows what to do and when to do it. Something like that does seems to be happening with the starlings.
I make these observations because I wanted the opportunity to put in words some of the wonder I have when I see starling murmurations.
But I was also motivated by my recent theme of the hazards of dark data when dealing with large scale data (the countless birds in a murmuration is a lot like big data especially if you imagine recording each birds position and velocity every second). In my previous posts I used the term dark data as an analogy to cosmology’s dark energy to explain the expanding universe or dark matter to explain the motions of galaxies. The word dark means we propose some unknown thing to take the place of what must be missing to make our models to continue to work.
The hazard of dark data is that we end up convincing ourselves we solved something and that there is no reason to look further.
In the context of starling murmurations the dark data is what exactly signals the individual birds how to adjust their flight paths. We don’t know. We have no way of knowing. But we can suppose that they might be using their eyes and ears to judge the positions and motions of just their most immediate neighbors. With that limited data, we can show mathematically it is possible to generate the same behavior. This is a very simple explanation of only immediate tracking information on just a handful of neighbors.
I agree that maybe that is all that is happening. But I disagree that it necessarily is what is happening. I see no reason to dismiss that they have some method of relaying messages from large distances or to dismiss their capacity to visualize a larger number of their neighbors or even be alerted about an oncoming murmuration that the individually can not yet see.
I can’t prove that there is another explanation, but I don’t see proof that the simple explanation is what is really happening.
To me, murmurations appear to be evidence of intelligence and hints at an intelligence very unlike anything we experience. I am content to acknowledge I don’t understand what is going on. I don’t need to invent data so I can feed a human-generated equation.