In my last post I tried to argue that Humberto Maturana’s biology of language might have something to say about the biological nature of mathematics. This biology of language that Maturana proposes is understood in the context of autopoiesis (the continuous self-creation of any living system) that is his fundamental definition of life. I gave a far too quick account of a related theory (Karl Friston’s free energy principle) at the end of the post, citing it as a powerful extension of Maturana’s provocative ideas.
Building a convincing account of autopoiesis is a book-length enterprise. And the free energy principle is a mathematically complex idea. So I didn’t do justice to either of them in that post. But they are both important to a sense I’ve had, for a number of years now, that mathematics does have a biological nature. Today, I want to make the argument again, but a little differently. I’ll begin with some references to studies that have caught my attention because they concern mathematical behavior in insects. A very recent Science article reported on a study that suggests that bees are capable of responding to symbolic representations of addition and subtraction operations. It was a small study (just 14 bees), but the bees were trained to associate addition with the color blue and subtraction with the color yellow
Over the course of 100 appetitive-aversive (reward-punishment) reinforced choices, honeybees were trained to add or subtract one element based on the color of a sample stimulus.
The bees were placed at the entrance of a Y-shaped maze, where they were shown several shapes in either yellow or blue. If the shapes were blue, bees got a reward for choosing a set of objects at the end of the maze that was equal to the first number of shapes plus one. If the first set was yellow, they got a reward for choosing the set of shapes at the end of the maze that was equal to the first number shapes minus one. The alternative or incorrect choice could have more than one shape added, or it could have a smaller number of shapes than the initial set. The bees made the right choice 63% to 72% of the time, much better than random choices would allow.
The full study can be found here. In their introduction they explain why these bees are worth looking at.
Honeybees are a model for insect cognition and vision. Bees have demonstrated the ability to learn a number of rules and concepts to solve problems such as “left/right,“ above/below”, “same/different,” and “larger/smaller.” Honeybees have also shown some capacity for counting and number discrimination when trained using an appetitive (reward- only) differential conditioning framework. Recent advances in training protocols reveal that bees perform significantly better on perceptually difficult tasks when trained with an appetitive-aversive (reward-punishment) differential conditioning framework. This improved learning capacity is linked to attention in bees, and attention is a key aspect of advanced numerosity and spatial processing abilities in the human brain. Using this conditioning protocol, honeybees were recently shown to acquire the numerical rules of “greater than” and “less than” and subsequently apply these rules to demonstrate an understanding that an empty set, zero, lies at the lower end of the numerical continuum.
I understand being impressed with the fact that bees can acquire this kind of discriminating ability. But not so clear is why they are structurally capable of such things. I would argue that it’s because their living relies on the presence of structure that allows these kinds of responses. This is the level that interests me.
Here are a few clips from past posts::
Ants were seen to communicate some kind of numerical information about the location of food.
In the described experiments scouting ants actively manipulated with quantities, as they had to transfer to foragers in a laboratory nest the information about which branch of a ‘counting maze’ they had to go to in order to obtain syrup…
The likely explanation of the results concerning ants’ ability to find the ‘right’ branch is that they can evaluate the number of the branch in the sequence of branches in the maze and transmit this information to each other. Presumably, a scout could pass messages not about the number of the branch but about the distance to it or about the number of steps and so on. What is important is that even if ants operate with distance or with the number of steps, this shows that they are able to use quantitative values and pass on exact information about them.
Zhanna Reznikova and Boris Ryabko, 2011. Numerical competence in animals, with an insight from ants.Behaviour, Volume 148, Number 4, pp. 405-434, 2011
The abstract of a paper published in Nature in 2001 includes this:
…honeybees can interpolate visual information, exhibit associative recall, categorize visual information, and learn contextual information. Here we show that honeybees can form ‘sameness’ and ‘difference’ concepts. They learn to solve ‘delayed matching-to-sample’ tasks, in which they are required to respond to a matching stimulus, and ‘delayed non-matching-to-sample’ tasks, in which they are required to respond to a different stimulus; they can also transfer the learned rules to new stimuli of the same or a different sensory modality. Thus, not only can bees learn specific objects and their physical parameters, but they can also master abstract inter-relationships, such as sameness and difference.
Mathematical behavior without a brain?
But here’s something interesting about the slime mold – the abstract of a paper published in Nature in September of 2000 reads:
The plasmodium of the slime mould Physarum polycephalum is a large amoeba-like cell consisting of a dendritic network of tube-like structures (pseudopodia). It changes its shape as it crawls over a plain agar gel and, if food is placed at two different points, it will put out pseudopodia that connect the two food sources. Here we show that this simple organism has the ability to find the minimum-length solution between two points in a labyrinth. (emphasis added)
And here’s another strategy used by researchers that was reported by Tim Wogan in 2010 in Science.
They placed oat flakes (a slime mold favorite) on agar plates in a pattern that mimicked the locations of cities around Tokyo and impregnated the plates with P. polycephalum at the point representing Tokyo itself. They then watched the slime mold grow for 26 hours, creating tendrils that interconnected the food supplies.
Different plates exhibited a range of solutions, but the visual similarity to the Tokyo rail system was striking in many of them… Where the slime mold had chosen a different solution, its alternative was just as efficient.
Autopoietic systems are ones which, through their interactions and transformations, continuously produce or realize the network of processes that defines them. They continuously create themselves. Maturana and Varela proposed that every living system is autopoietic, from individual cells, to the nested autopoietic systems in organs, organisms, and even social organizations. In my last post I connected this interpretation of life to Karl Friston’s free energy principle. But it was pretty sketchy, so I would like to fill it in a little here.
I find it important that the free energy principle has the same circular model of living processes as autopoiesis. But for Friston, the key to a system’s continuously regenerating itself relies on how it manages to maximize expectations and minimize surprise. Minimizing surprise is essentially the same as maintaining a low entropy state, which is synonymous with maintaining ones structure. (The mathematics of entropy in information theory, is easily applied to entropy in thermodynamics.) And so minimizing surprise is the same as minimizing entropy. But the thing that holds it all together, the thing that can formalize the analysis, is the use of Baysian inferences or statistical models because this is a way to quantify uncertainty. With all of this, living systems maintain themselves by keeping themselves within a set of expectations (through sensory information, statistical inferencing, and their own action), If they stray too far from having those expectations met (like a fish out of water), they will no longer exist.
When I consider the different ways that mathematics is present in bees, ants, and slime molds from the perspective of autopoiesis or free energy, mathematics looks like its right in the middle of all the action – in the thick of the organism’s living. It will show up in the interactions and transformations that contribute to life itself (where life is the maintenance of the structural and functional integrity of oneself) because it is part of the regular flow of its living. According to the free energy principle, living systems live by either adjusting their expectations to match the flow of sensations, or adjusting the flow of sensations to match their expectations. It must be true that mathematics is as much a part of this as any biological process.
Recent Comments