The worm’s apparent simplicity makes it a favorite lab animal — for example, it was the first animal whose brain was mapped. Researchers have learned that, with fewer than 400 neurons, it can handle both associative and non-associative learning.
Cassell, author of Animal Algorithms (2021), offered a number of remarkable facts about the worm and I thought I’d summarize a few of them again here, along with some questions:
- The worm, 1 millimetre in length, consists of only 900 cells. Thus neurons comprise a large proportion of its total cell count.
The worm actually comes in two forms: males which have 385 neurons and hermaphrodites (both sexes) which have 302 neurons. In both cases, it seems that over one-third of all its cells are neurons. But if we compare the worm to the human, we see a considerable difference: The human body has roughly 30 trillion cells and the human brain only 86 billion. Of course, there are neurons distributed throughout the human body. Even so, the proportion of human neurons to other human cells seems much lower.
Perhaps the worm’s brain has roughly the minimum number of neurons any brain must have for simple bodily functions and learning — irrespective of the size of the rest of the body.
- Equipped with those 400 neurons, the worm can feed, fast, mate, lay eggs, swim in liquids and crawl on solids. In fact, the worms’ “social lives” can become, well, quite complex:
Cassell writes,
In addition to basic behaviors, C. elegans is also capable of learning, including associative and non-associative learning. A paper published in the Journal of Neurochemistry documented the learning behaviors, including attraction and aversion to salt, temperature, and other substances. What might be surprising to many is that this learning involves both short-term and long-term memory mechanisms, which include regulation of neurotransmitters.
“Even the tiniest brain,” March 17, 2025
- Cassell offers several observations that touch on this question:
… even though the brain is tiny, it does not have a simple structure. One might expect the smallest known brain to have a structure that is either relatively uniform or random. An example of a uniform structure is that found in crystals, which form a symmetrical lattice. A random structure would be expected if the positions of the neurons were not specified, but rather develop through a random process. Contrary to being either uniform or random, the brain does have a complex structure that is specified and repeatable.
“Tiniest brain”
Yes, that’s the problem of specified complexity: In a world where nature, left to itself, produces either uniform order or chaos, we find a level of information-rich order that requires an underlying intelligence. And in this case, that information-rich order is alive.
- And just when we think we might have finally got down to the truly simple, basic part:
A second observation is that the brain contains a large number (approximately 100) of different types of neurons, both in terms of design and function. They are not all identical. That also would not be expected for the smallest brain. A third observation is that small neural networks within the brain control various behaviors, such as the touch response network. It is possible that some of these neural networks are irreducibly complex.
“Tiniest brain”
Irreducibly complex means that the current structure cannot have arisen via a gradual buildup from simpler to more complex steps. It’s not that simpler versions could not do the job as efficiently; rather, none of them could do it at all.
- Cassell quotes a research paper that attempts to account for C. elegans’s unexpectedly busy little brain:
The mere existence of such structures may actually further underscore the directed evolution to form such clusters, which presumably carry fine functional roles along the neurites. Taken together, local compartmentalized activities, facilitated by the clustered synaptic organizations revealed herein, can enhance computational and memory capacities of a neural network. Such enhancement may be particularly relevant for animals with a compact neural network and with limited computational powers, thereby explaining the evolutionary forces for the emergence of these synaptic organizations.
Ruach, et al., “The synaptic organization in the Caenorhabditis elegans neural network suggests significant local compartmentalized computations,” PNAS, 2023, Vol. 120, No. 3.
But wait. Did the researchers say “directed evolution”? As Cassell notes, the term has never been generally accepted in the research literature. That’s probably because it implies underlying purpose or design. Here we have no quarrel with that idea. But it appears to abandon the idea of gradual, random assembly of even the C. elegans brain via natural selection acting on random mutation.
So Here We Are…
Looking at the very simplest brain known, we find both specified and irreducible complexity. It is all very far removed from the organic elements that are the building blocks of life. And yet we aren’t yet anywhere near the types of brains that think, in the sense that a dog thinks.
However evolution happens, it is beginning to sound far more complex than the sort of theory that made Richard Dawkins feel intellectually fulfilled as an atheist.
Posts
