One interesting thing that living matter and non-living matter don’t seem have to have in common:

  1. The interactions of non-living matter, like a stone, are most easily explained in terms of forward causation. Newton’s laws allow me to predict the future motions of a stone in a very straightforward manner. Our models go forwards in time.
  2. The interactions of living matter, like a starling, are often most easily explained in terms of backwards causation. If I want to understand why the starling stole my chips, the simplest explanation is something along the lines of “the starling wanted to obtain food because it was hungry” (or even just “it was hungry”). We’re modelling the starling’s behaviour (taking food) in terms of variables only defined in the future (a satiated starling). The model here goes backwards in time.

I came across this while reading a blog post about abiogenesis, the origin of life, but I can’t for the life of me remember what it was. The author had done his PhD on life, and thought that this criteria could be formalised as a measure of “how alive” some matter was. I’m not sure what I think of it right now, mainly because I know very little about life, but it’s a really interesting idea. And it’s easy to test, at least in principle - here’s a few examples:

Consider a stone. Most of the time, the only force that has any bearing on a stone’s life is gravity. Newton’s laws are a relatively simple way of predicting the movement of the stone in the future. As an explanation of why the stone moves the way it moves, Newton’s laws are extremely good!

Is it possible to explain a stone’s movement more simply in terms of reverse causation? I don’t think so, mainly because a stone doesn’t actively pursue any particular state of being. It doesn’t get hungry, it doesn’t want to reproduce, and it doesn’t sleep.

This is good news, because it means the measure works when you apply it to a stone. It predicts a stone is not alive, which is exactly what you’d want it to do.

The case of a starling stealing food is more interesting. First of all, is it possible to exlain why it stole my chips in terms of forward causation? In principle, yes - if you could measure exactly what the state of every single cell in the starling’s body was at some point in time, you could use the standard theories of physics to calculate how the starling would move. Actually measuring that state would be impossible in practice, but even supposing you could it would be an enormous amount of information. In other words, a poor explanation.

It’s much easier to understand the starling’s behaviour if you flip things around. A starling wants to be in a variety of desirable states, such as being well fed, reproducing, sleeping, being safe etc. Starlings are usually not in these desirable states, however, and act so as to improve that. The starling must have been hungry, so it took a chance and stole my chips to bring itself close to a desirable (satiated) state. This is a very different kind of explanation, but it is still predictive, and much simpler than the usual models of physics in this case.

Again, this is great news for the measure, because it works when you apply it to a starling. It seems to do quite well when you apply it to clear-cut cases. What happens when you try to measure a less clear case, such as a fire burning on a log?

In this case, it seems like there are pretty good forward and backward explanations, which is interesting. You can probably model the dynamics of fire using fluid mechanics, and that would give reasonable explanations of why the fire spreads the way it does. On the other hand, it’s not totally crazy to say that fire wants to eat as much wood as possible, in which case you can explain the spread of the fire in terms of its desire to eat all of the wood.

I’m not completely sure about my reasoning here, but if it holds up then it seems like the measure is agnostic as to whether fire is alive or not. I’m not sure what to make of that - at this stage it’s an interesting thought experiment, but I’d love to see it formalised as a bonafide mathematical theory.

It would be really useful for understanding how life could emerge (-genesis) from non-living matter (abio-).

I’m a big fan of the way anthropocentrism seems to be crashing.