Look at the image below. This rather bizarre thing is Tom Ardans’ depiction of an imagined creature, an Elder Thing , that figures in the Cthulhu Mythos , the product of the tormented imagination of Howard Phillips Lovecraft . The most famous appearance of an “Old One” happened in the short novel At the Mountains of Madness (which, so the legend goes, will be filmed by Guillermo Del Toro ), in which a scientific expedition to the icy realms of Antarctica happens to discover the remains of a bizarre alien civilization, whose members are thought to be extinct. Those members are the barrel-shaped “Elder Things” from the picture.

What, if anything, this has to do with evolutionary neuroscience? Well, read the brief review in the last paragraph again. It says there that the Elder Things formed a civilization. Just let the suspension of disbelief act a little, and let us forget that this is a work of fiction. Would this be possible? Would creatures with such an alien body plan be able to develop nervous systems which are complex enough to build a whole civilization? One important aspect of Lovecraft’s work, of course, relies on the assumption that the Elder Things (as all servitor races) are alien in origin – and, in the minds of fiction writers, are thus not amenable to the same evolutionary pressures that shaped Earthly organisms. Let us push that aside, as well. Here, on Earth, would it be possible to one such species – “monstrous barrel-shaped fossil of wholly unknown nature; probably vegetable unless overgrown specimen of unknown marine radiata” – to have a complex nervous system? The field of evolutionary neuroscience, in a sense, can help us to answer this question.

“Existing biology would have to be wholly revised, for this thing was no product of any cell growth science knows about”
To begin to answer the question of whether or not the Elder Things are plausible, let us return to the “dual nature” of evolutionary biology we mentioned before. Try to remember that evolutionary biology – as well as evolutionary neuroscience – has two main “themes”: variety and similarity. How can we begin to tackle the Old One problem using those two concepts?
One of Darwin’s most potent ideas was the realization that most living species shared some common features, as well as had their differences. Take the group Mammalia , for example. All mammals share common features: body covered in fur, the skull with two occipital condyles , seven cervical vertebrae… Other traits, however, are not shared by all mammals. For example, the ability to fly and to use echoes as a form of perception is unique to bats.
Darwin’s observations, however, were more than just “pattern recognition”. He proposed that the traits which are shared by species in the same taxonomic level are there because of common descent – that is, this trait was present in a species that was the common ancestor of the extant (living) species, and was transmitted to the descendents with at least some conservation. In other cases, the trait was modified so much in a given descendent that it can be said to be new.
In order to judge whether or not a given trait is primitive – that is, whether or not it was present in the common ancestor of two species – we must either have access to fossil records, or infer this from the state of extant species. By resorting to a cladistic approach – that is, by analyzing traits superimposed to a phylogenetic tree , which describes the evolutionary relations between species -, we can try to infer whether or not a trait is primitive. The lesson, here, is that, in order to understand how a given trait evolved, we must first locate it in a kind of map. This map is a little bit like this:

This is the representation of the relations between a few of the known species. “You are here”, in the figure, refers to the position of Homo sapiens in this phylogeny.
Why would we need such a map to understand whether Elder Things could possibly evolve complex nervous systems? Well, first of all we must understand whether they would be part of a branch of this tree which routinely evolved big, complex, highly interlinked brains. They need not be closely related to men. Another feature of the cladistic approach is that it allows us to understand when two evolutionary events are convergent – that is, when a given trait evolved independently in two species (or groups of species). Take electrolocation, for example. This is a remarkable sensory modality, in which the individual sense electric fields around its bodies and interpret distortions in it in order to navigate its environment, that evolved independently in sharks, two groups of fishes – Gymnotids and Mormyrids, distantly related with each other – and in platypuses. Given the fact that the common ancestor between those groups did not present electrolocation, it is more parsimonious (and, in evolutionary terms, more correct) to postulate that this trait evolved independetly in sharks, electric fishes, and platypuses.
The same thing, it appears, hapenned with brain sizes. Even though brain size may seem like a coarse variable to be analyzed – something that would fit a pre-Victorian neuroanatomist, but not scientists in the Genome Era -, it is in fact very informative. The evolutionary neuroscientist Georg Striedter dedicated most of his book, Principles of Brain Evolution , to the analysis of changes in brain size. One interesting fact about this variable is that brains increased and decreased in size many times in the evolutionary history, but reached the biggest sizes in primates, cetaceans (whales, porpoises and dolphins) and cephalopods (octopuses and squids). Those species are only distantly related, and a cladistic analysis reveals that the evolution of big, complex brains occurred independently in those species. Thus, in the central nervous systems, locating a species (or group of species) in a phylogenetic tree is very useful to begin the investigation of which evolutionary processes were responsible for the current state of the trait in question.
In what comes, then, we will begin our study of the evolution of Elder Things’ nervous systems by doing something that belongs to a whole different field of inquiry – taxonomy and systematics. Our first task will be to identify our species. Let us sacrifice some virgins to Azathoth and Shub-Niggurath, lest their ire against our attempt not destroy our whole world; and then, we will pursue the inglorious field of taxonomy.

Phylogenetic systematics for Elder Things
So far, we’ve been setting the field for our investigations on the nervous system of a fictional species, the Old One, an important part of Lovecraft’s Cthulhu Mythos. The first step here would be to determine the position of such a creature in the Tree of Life. Lovecraft’s description is interesting, because it tangentiates what was known at his time in terms of taxonomy. Let us come back to his description:
“Important discovery. Orrendorf and Watkins, working underground at 9:45 with light, found monstrous barrel-shaped fossil of wholly unknown nature; probably vegetable unless overgrown specimen of unknown marine radiata. Tissue evidently preserved by mineral salts. Tough as leather, but astonishing flexibility retained in places. Marks of broken-off parts at ends and around sides. Six feet end to end, 3.5 feet central diameter, tapering to 1 foot at each end. Like a barrel with five bulging ridges in place of staves. Lateral breakages, as of thinnish stalks, are at equator in middle of these ridges. In furrows between ridges are curious growths. Combs or wings that fold up and spread out like fans. All greatly damaged but one, which gives almost seven-foot wing spread. Arrangement reminds one of certain monsters of primal myth, especially fabled Elder Things in Necronomicon. These wings seem to be membraneous, stretched on a framework of glandular tubing. Apparent minute orifices in frame tubing at wing tips. Ends of body shrivelled, giving no clue to interior or to what has been broken off there.”
Of course, part of this description is intended to undermine any attempt of classification. The Elder Things have traits which are reminiscent of many Earthly creatures – from the most basal Eumetazoa , the radial animals from the phyla Cnidaria and Ctenophora , to plants and vertebrates. We will focus here on only one of those traits: their body plan.
Body plans are important features of organisms. Every organism has an inherited body plan that can be described in terms of inclusive categories, such as simmetry, presence or absence of body cavities, partition of bodily fluids, presence or absence of segmentation, degree of cephalization, and type of nervous system. The diversity of body plans of animals is restricted by its evolutionary history, the ecological niche it occupies, and its life history. One example: even though a worm that adopts a parasitic life style in the intestine of one vertebrate can look and behave in ways which are radically different from non-parasitic worms, both share the same body plan that is a trait of their phylum.
Let us take another close look at the depiction of an Elder Thing:

It is clear, from that image, that Elder Things are characterized mainly by their barrel-shaped body. This body can be divided in similar halves by more than two planes that cross the longitudinal axis of the organism. In taxonomy, this is called radial symmetry, and two phyla in the kingdom Animalia have it: Cnidaria and Ctenophora. This plan is particularly good for animals which are sessile or sedentary, or for animals which are free-swimming, because they can sense their environment from all sides equally. Notice, however, the pair of wings that Elder Things’ have. A paired structure such as this represents a variation in the “radial symmetry” theme, called biradial symmetry . The only phylum to present this type of organization is Ctenophora, composed of less than 100 species – all of them marine, occurring specially in warm oceanic waters. This is consistent with the hypothesis (made by Lovecraft, based on a few geological conjectures of his time) that the poles once were much warmer places. Ctenophora are named after the eight plate lines which are used for locomotion. With the exception of a few sessile and crawling species, Ctenophora are free-swimming. Even though they are weak swimmers and more common in superficial waters, some species are sometimes found in deep waters.

Elder Things present a few traits which are unique to them, in relation to other Ctenophora: they present wings, which the scientists in the original expedition narrated in “At the Mountains of Madness” postulated were used when specimen left the sea; and, most importantly, they have highly derived behavioral capacities, which are very very VERY different from those found in other members of their phylum. In the next section, we will examine the nervous system of Ctenophora with more care, in order to make those capacities more salient. Part of this information was passed to me by Yog-Sothoth, and let us hope that I do not get incredibly mad while I deal with the absurdity of it all.

The nervous systems of ctenophores
We concluded that the Elder Things from At the Mountains of Madness should be included in the phylum Ctenophora. This is because the Elder Things share many traits – most importantly, a basic body plan – with creatures from this phylum. Let’s now try to unravel the basic plan for the brains of Ctenophora.
Ctenophores present a nervous system which is very similar to that found in Cnidaria, the other phylum that is composed of animals with radial symmetry. This nervous system is organized mainly as a nervous network, in which each neuron is connected to many others in a seemingly random fashion. No central control is found here; the nervous systems of radial animals are the best example, in nature, of a diffuse nervous system. Two interconnected networks, one in the epidermis and one at the basis of the gastrodermis, are found. The axons from the neurons in this network terminate either in other neurons, or in junctions with sensorial or effector organs. As in other animals, signals are transmitted from one cell to the other by means of neurotransmitters; however (differently from other animals), the majority of sinapses have neurotransmitter vesicles in both sides, allowing transmission from any side of the synapse.
Since this network does not have any kind of central system, the conduction of a neural signal in it is diffuse – which means that, if we apply one stimulus in any given point in the animal’s body, a spike will be initiated which will be conducted equally in all directions and reach any other point in the network. Even if different lesions are made in axons from one point in the network, conduction will still be able to reach any other part of it because there are many alternative and redundant connections between two points. The fact that synapses are usually bi-directional also contributes to the diffuse mode of signal transmission in the network.
Myelin sheets are absent in the axons of Ctenophora. This fatty substance is responsible, in other animals, for the “saltatory” pattern of neural transmission. Myelin is electrically isolant; thus, action potentials are generated not in the regions of the axon which are covered in myelin, but in the regions which are not (nodes of Ranvier). This tends to reduce the leakage of ionic currents, which results in the maintainance of the electric potential conducted by the axon for longer distances, and increases the velocity of propagation. In Cnidaria and Ctenophora, however, myelin is not present, and their axons are fine and short. This is not very economic: short axons (which could be longer if myelin was present) demand a greater number of synapses in any given pathway, which retards the passage of the spike in the system. The general conclusion is that neural transmission in Ctenophora is slow.
Thus, if we apply one stimulus in one point of the body of one ctenophore, the ensuing spike will be conducted everywhere in the network; this process will be slow. Also, since there is no myelin to prevent the effects of current leakage, spikes will get smaller and smaller as they travel through the axon. Th behavioral consequences are simple: response strength, in any part of the animal, will be inversely proportional to the distance from this part to the origin of stimulation.
There is a secondary consequence to this: if we apply (say) an electric shock in one part of a ctenophore, there will usually be no response. However, if we apply a second shock (even after a brief interval), the response will be ellicited – in fact, it will be increased. Thus, repeated stimulation is usually necessary for the elliciation of a response. This phenomenon (called facilitation dependency ) can be understood as an adaptation to the constraints of the animal’s body plan and nervous system organization. First, facilitation dependency ensures the transmission of weak stimuli. Second, it ensures the economy of metabolic resources (action potentials consume energy in the form of ATP ): the animal does not respond to a single stimulation (which could be meaningless), but to a given number of stimuli. In a system that highly dependent on facilitation, responses will be more linked to the number of stimuli than to its intensity.

Ph’nglui mglw’nafh Cthulhu R’lyeh wgah’nagl fhtagn
Let us consider again what we know about the Ctenophora, the phylum that we identified as the one to which Elder Things pertain. They have biradial bodies, sessile or free-floating habits, come from aquatic environments, and have diffuse nervous systems. We know: 1) their position in the history of life; 2) their anatomical constraints (bi-radial symmetry); 3) their sensory environment (aquatic environments, from whence stimuli come from all sides at the same time); 4) their behavioral tendencies; and 5) the organization of their nervous systems. From and adaptive point of view, we can connect all those things.
Biradial bodies are probably adaptive in the aquatic environment (especially in benthic environments ), where sensory input can come from any direction. Of course, they are only adaptive if your foraging habits are not really complex: most ctenophores are not very active hunters, and some are even sessile. The physiological and anatomical properties of ctenophores’ diffuse neural network are useful when dealing with those limitations. Later in the history of evolution, other species diverged from this phylum, inventing new body plans and, consequently, new ways to deal with it. They did this to occupy new ecological niches.
When animals evolved a new body plan – one that did not involve radial symmetry, but bilateral symmetry – their nervous systems had to be reconfigured. In general, as animals evolved bilateral symmetry, they tended to divide their internal organs. For example, Platyhelminthes , Nemertea , and Gnathostomulida are the first phyla to present an bilateral organization; this was accompanied by two essential adaptations. First, while Ctenophorae and Cnidariae had two sheets in their embryos (ectoderm and endoderm), those latter animals evolved a third sheet, the mesoderm . Thus, while all internal and external structures from Ctenophora developed from one of those two sheets, the mesoderm developed new organs, tissues and systems in all latter animals. A second adaptation was the tendency for cephalization, which is a consequence of abandoning radial symmetry: as animals became bilaterally symmetrical, they also tended to be “directed” to one pole of their bodies. This was accompanied – and that is what is important here – with a greater organization of the central nervous system.
The psychologist Herbert Spencer thought that nervous systems had a tendency, in evolutionary history, towards greater complexity, centralization, and division . This is not always the case; for example, amphibians and lungfishes actually decreased the complexity of their brains when they first diverged from their ancestors as different species. However, a general tendency for greater “division of labor” is indeed found, at least in invertebrates (whether or not this tendency remains as strong in vertebrates as it is in invertebrates is another story). Take, for example, the case of Hydra, one member of the phylum Cnidaria. In this genus, most circuits are made by sensory-motor cells (or epithelial-muscular cells, which execute the functions of sensory reception as well as motor activity. If one of these cells is stimulated, it contracts, contracting the body and/or tentacles. In Ctenophora, however, most circuits are intermediated by one or more neurons; whenever a sensory cell is stimulated, it transmits a signal to another cell. In this case, then, there is a more widespread division of labor in the functions of both cells.
Besides this tendency to divide the functions of cells, there is also a tendency to segregate functions in systems. In a Cnidaria from the genus Cordylophora, we find only one diffuse network and no centralization. A derived state in other Cnidaria (as well as in Ctenophora), at least two networks are found: the principal network, located between the epidermis and muscles, and a secondary, less-developed network, associated with the gastrodermis. In vertebrates, this tendency is higly developed: we have complex central networks (“modules”) which segregate sensory modalities, and even within those networks there is considerable separation of labor – for example, the segregation of motion, color, and form perception in the mammalian visual system.
In conjunction to this tendency for division of labor, there is a tendency for centralization – that is, a tendency to form “clusters” of neurons and circuits which have equal functions. In the pedal discus of Hydra, the neurons of the diffuse network are more concentrated than in other regions, assuming a circular organization that “suggests” a nervous ring . In Hydra vulgaris, tere is an obvious nervous ring around tentacles and hipostome. In anemonae (animals from the phylum Cnidaria), there is an anatomical separation of slow systems (composed of neurons with slow conduction velocities) and fast direct systems (composed of parallel networks of bipolar neurons). Of course, no central nervous system can be found in Ctenophora and Cnidaria; but a tendency for modularity is already present.
Very well, we described the diffuse networks of some simple Cnidaria and Ctenophora; what about Elder Things? Let´s see a description:

“The nervous system was so complex and highly developed as to leave Lake aghast. Though excessively primitive and archaic in some respects, the thing had a set of ganglial centers and connectives arguing the very extremes of specialized development.
Its five-lobed brain was surprisingly advanced, and there were signs of a sensory equipment, served in part through the wiry cilia of the head, involving factors alien to any other terrestrial organism.”

As we thought, their nervous system is much more complex than that of ctenophores. They have a brain, organized (at least superficially) as a mammalian brain (“five-lobed brain”, I presume, is the closest Lovecraft could get to describing mammalian brains). This makes sense in light of our observations: First of all, they produced a civilization, which demands 1) consciousness, and, consequently, 2) a rather complex nervous system. In primates, brain size correlates rather well with group size , and, as brains increase in size (at least in vertebrates), they tend to increase in complexity – that is, they tend to be even more centralized and modular than the brains of small-brained animals from the same group. By analogy, a complex nervous system – centralized, composed of a brain that is divided in many modules – is needed to deal with social complexity. This is hardly the case for animals with diffuse nervous networks. However, the position of Elder Things in the Ctenophora radiation (inferred from their radial body plan) leads us to predict that their nervous system would be more akin to a diffuse network than to a more centralized nervous system that is typical of animals with bilateral symmetry.
Radial symmetry is an adaptation to the specific ecological challenges imposed by the specific niche occupied by radial animals. Radial animals have diffuse networks because their environment has not selected nervous systems which could not deal with it. A centralized nervous system can be selected only when there is a high degree of cephalization, which is a consequence of bilateral symmetry; and it is not capable of dealing with radial symmetry in body plans. Thus, the niche occupied by ctenophores demands radial symmetry, which, in its turn, demand diffuse networks. In order to maintain a radial body plan and evolve a centralized nervous system, too many evolutionary changes would be needed, in concert. This is highly improbable, though.
Of course, if we change our assumptions – especially the assumption that Elder Things evolved on our own planet, which is false in the Cthulhu Mythos – we can acommodate these problems. That is precisely the point Lovecraft made: the laws that apply to living beings on Earth are not necessarily appliable to uneartlhy creatures. Lovecraft is a superb horror writer precisely because of this: his main premise in devising the Cthulhu Mythos is that horror could emerge from the idea of an universe that is, in some sense, impervious to the general laws we uncover here.

In which we derive principles from what we speculated until now
Elder Things were used as a model from which we could explain some fundamental principles of brain evolution:
1) Brains evolve inside a phylogenetic context: We inferred that the nervous systems of Elder Things were organized as diffuse networks because that is how the brains of their relatives – other Ctenophora – are organized. This is the theme of “similarity”.
2) Brains evolve inside a ecological context: The nervous systems of Ctenophora are radial because this is adaptive in their ecological context. They are as well adapted to their niche as are other animals to their respective niches.
3) Brains are not free to evolve: The radial symmetry of ctenophores’ bodies constrains the evolution of their nervous systems; when new species, with bilateral organization, evolved, these species could evolve more centralized and cephalized nervous systems.
These three principles can be used to derived all other principles of brain evolution

Disclaimer
This work was made under the financing of Miskatonic University.