The brain: a concept in flux

Abstract

One of the most important aspects of the scientific endeavour is the definition of specific concepts as precisely as possible. However, it is also important not to lose sight of two facts: (i) we divide the study of nature into manageable parts in order to better understand it owing to our limited cognitive capacities and (ii) definitions are inherently arbitrary and heavily influenced by cultural norms, language, the current political climate, and even personal preferences, among many other factors. As a consequence of these facts, clear-cut definitions, despite their evident importance, are oftentimes quite difficult to formulate. One of the most illustrative examples about the difficulty of articulating precise scientific definitions is trying to define the concept of a brain. Even though the current thinking about the brain is beginning to take into account a variety of organisms, a vertebrocentric bias still tends to dominate the scientific discourse about this concept. Here I will briefly explore the evolution of our ‘thoughts about the brain’, highlighting the difficulty of constructing a universally (or even a generally) accepted formal definition of it and using planarians as one of the earliest examples of organisms proposed to possess a ‘traditional’, vertebrate-style brain. I also suggest that the time is right to attempt to expand our view of what a brain is, going beyond exclusively structural and taxa-specific criteria. Thus, I propose a classification that could represent a starting point in an effort to expand our current definitions of the brain, hopefully to help initiate conversations leading to changes of perspective on how we think about this concept.

This article is part of the theme issue ‘Liquid brains, solid brains: How distributed cognitive architectures process information’.

The difference in mind between man and the higher animals, great as it is, certainly is one of degree and not of kind.

—Charles Darwin

To extend our understanding of neural function to the most complex human physiological and psychological activities, it is essential that we first generate a clear and accurate view of the structure of the relevant centers, and of the human brain itself, so that the basic plan—the overview—can be grasped in the blink of an eye.

—Santiago Ramón y Cajal

1. Definitions

A built-in aspect of the scientific outlook is that we scientists are enthusiastic nitpickers, who oftentimes take pleasure in finding small inconsistencies and exceptions that will lead to reconsiderations and even in some cases complete reformulations of established scientific definitions. As we define, refine, reconsider and reformulate the ideas pertaining to a particular concept, we have a good chance of gaining profound insights that we would not be able to obtain otherwise. The fluidity of definitions is their main strength, because this flexibility is one of the most important factors that allow science to grow. An excellent example of this conceptual refinement is the process of trying to define the brain.

2. The idea of a brain

It stands to reason that ancient humans must have had an acute—albeit indirect—understanding of the importance of the brain, as it is undoubtable that such knowledge would give our ancestors a survival advantage. For example, they would have been acutely aware that a head injury would be the surest way to kill an animal or an enemy. It also stands to reason that they would have known of the specific effects of injuries on particular parts of their own heads. Alas, any practical knowledge on this matter that our ancestors might have accumulated over the millennia was lost until humans began to leave written records. It is widely believed that the first written repository of such knowledge came to us via the ancient Egyptians. Probably the earliest explicit reference to the human brain is found in an unnamed papyrus from about 6000 years ago [1], however, the best-known early instance of a written record about the human brain and what happens to it when disturbed is found in the Edwin Smith Papyrus. This is a 3500 year old document (likely a copy from an earlier source) usually recognized as the first formal text on trauma medicine [2]. This papyrus described in detail injuries caused by industrial accidents or as a consequence of battle, and established the first written example of treatment priorities (i.e. by classifying the injuries as ‘treatable’, ‘probably treatable’ or ‘untreatable’), articulating the basics of modern medical triage protocols. Most importantly for the purposes of this review, in this papyrus we also find what is probably the first written word explicitly meaning ‘brain’ (figure 1). This document also describes, among many other types of injuries, the undesirable effects of head injuries. Strangely, despite their evident knowledge of the importance of the integrity of the brain in terms of bodily function, the ancient Egyptians did not seem to consider the brain worthy of preservation during their burial practices, as opposed to the heart, which was carefully preserved in mummification procedures (for a more detailed exposition of this topic please see [3], chapter 4). Over time, our knowledge—as well as the interpretation of such knowledge—about the human brain evolved into Aristotle's dual brain concept, which defined the brain as consisting of the encephalon (or proper brain) and parencephalon (roughly, the cerebellum). An implicit assumption of Aristotle's view is that brains were part of the anatomy of vertebrate animals. Thus, historically, the way we thought of the brain and nervous system—not surprisingly owing to Aristotle's influence—was in the light of vertebrate organisms.

Figure 1.

Possibly the earliest known instance of the word brain in a written language: Ancient Egyptian hieroglyphics. As modified from Pagán [3]. Illustration by Alexis G. Pagán.

3. To define a brain

Thinking about the brain from the ‘vertebrocentric’ point of view [4] dictated the early discourse about its nature. This frame of mind was the basis of the formulation of several theories of brain architecture [5,6] which described the vertebrate brain in purely structural terms. The eventual sophistication of anatomical knowledge gave rise to the idea of segmental organization, which divided the brain into 10 structurally discrete parts: cerebral cortex, basal ganglia, thalamus, hypothalamus, tectum, tegmentum, pons, medulla, cerebellum and the spinal cord [5,6]. The segmental organization concept was refined by studying the ontogenetic (developmental) origin of the aforementioned parts alongside their phylogenetic (evolutionary) history. At this developmental/evolutionary point, this integration explicitly acknowledged the importance of ‘lower organisms’ in this story and consequently, neuroscientists began to cautiously wander away from a pure vertebrocentric view. Eventually, with the advent of the molecular biology revolution, genomics began to provide unique molecular insights into the above classifications. It is important to point out that since these theories were purely based on structural properties, they did not touch upon the function of such structures or upon any emergent properties (consciousness as a premier example) that might arise when these parts worked together. The study of such integration would come later as part of the study of the brain as a complex system.

At the point in time when scientists initiated the first efforts to systematically define a brain, functional considerations began to be taken into account in a more prominent way. In this sense, a brain was defined as an organ acting as the control hub of the nervous system in animals displaying cephalization. This definition implied that a brain must be located within the head of the animal. Further along, this definition was refined by stating that the brain is the organ tasked with the generation of processes like memory, motivation, thought and consciousness. As usually occurs in science, this refinement generated more questions; a case in point: consciousness. Defining consciousness is as difficult—likely even more difficult—as defining the term ‘brain’ itself. Alas, consciousness is a topic beyond the scope of this review.

4. An early effort to define the brain beyond the vertebrate paradigm

Approximately in the mid-twentieth century, neuroscientists began to think about the brain out of the proverbial ‘animal-centric’ box. More recently, as a result of certain interesting discoveries, some scholars are tentatively including microorganisms and plants in their thoughts about the brain, a trend that as we shall see, is currently under intense debate.

As we discussed, the original approach to answer the question ‘What is a brain?’ was built on three main aspects: (i) a purely structural approach, (ii) a vertebrocentric perspective, and (iii) a precise identification of what kind of vertebrate we would be referring to. Soon enough, scientists expanded their thoughts about the brain by including function in light of structural features and information about developmental and evolutionary origins. Probably one of the first explicit efforts that tried to come up with a wide enough definition of a ‘brain’ in order to be useful to a variety of scientists appeared in the paper: ‘What is a brain and who said so?’, a semi-satirical yet remarkably well-thought paper published in 1986 in the British Medical Journal by the late Dr Martin George Netsky [7]. In this paper Dr Netsky presented an insightful and quite entertaining examination of the multiple meanings that the word ‘brain’ can take, and I believe that its main message is still valid today, at least in principle. The main thesis of Dr Netsky's paper was to make the point that articulating a comprehensive, purely structural definition of a ‘brain’ is bound to be rather difficult or even impossible, at least in a way universally accepted by the general scientific community. In these lines, Dr Netsky proposed that a proper definition of a brain would include aspects of structure and function integrated with mechanistic explanations of how a brain would generate its multiple functions. Some of these functions include but are not limited to, the detection of environmental signals, the processing and interpretation of these signals in order to generate behaviours (ideally, from an evolutionary perspective, these behaviours would enhance the survival chances of an individual). Finally, other functional aspects of a brain must include learning and memory processes in order to make use of the interpretation of such signals to better cope with future environmental challenges. Additionally, Dr Netsky stated that any proposed definition of the term ‘brain’ should address questions like whether brain convolutions are essential to its function (in fact, they are not; many vertebrate species lack brain convolutions with no apparent disadvantages; case in point: birds).

As a self-professed lexicophile (a dictionary enthusiast), Dr Netsky explored a plethora of published definitions of a brain and found all of them lacking, since these definitions overwhelmingly focused on purely anatomical criteria. In his words, all of these structurally-oriented definitions merely described ‘… a dead brain’.

Dr Netsky's idea emphasized that a working knowledge of the differences between ganglia and nuclei in a neurobiological context is absolutely necessary in order to truly understand the brain. Nervous systems are assemblages composed of collections of neurons that work closely with each other to perform specific functions. A group of closely located/interacting nerve cells is arbitrarily defined as a ganglion (plural: ganglia) when it is located outside of a central nervous system and as a nucleus (plural: nuclei) when located within the central nervous system. The distinction between ganglia and nuclei is not absolute and is therefore a source of some confusion, since neuroanatomists came up with the ganglia and nuclei designations long after some parts of the human brain were named and described. For example, there are parts of the mammalian central nervous system like the aforementioned basal ganglia that are involved in important functions such as motor control and motor learning, executive functions like impulse control and mental flexibility, as well as the generation of emotions, among many other functions [8–10]. Since the basal ganglia are located within the confines of the central nervous system, technically, they should be called basal nuclei (and they are increasingly called that in recent works).

It is evident that one of the main influences on Dr Netsky's ideas about the brain was his interest on the evolution of nervous systems. In fact, Dr Netsky and a colleague, Dr Harvey B. Sarnat, wrote one of the seminal books on this topic [11]. According to Dr Netsky the ‘origin story’ of his ‘What is a brain and who said so?’ paper began when he and Dr Sarnat submitted a review paper exploring some ideas on the evolution of the human brain. They proposed that the aforementioned delineation of a clear distinction between a ganglion and a proper brain is essential to formalize a proper definition. Basically, they argued against the—at the time—standard designation of invertebrate central nervous tissue structures as ‘cephalic ganglia’. This designation is still sometimes used even though all available evidence indicates that such nerve conglomerates serve as the control centre of the organism; in other words, an actual, bona fide, brain. Sarnat and Netsky's point of contention was that a ganglion and a brain are fundamentally distinct entities, meaning that a ganglion, however large, can never developmentally transition into a brain. In these lines, they could not find any examples of this transition; in every example that they explored, a ganglion seems to remain a ganglion however big it would get. Their chosen example to illustrate the point was the well-known fact that certain dinosaur species displayed an enlarged lumbosacral ganglion, which actually was several times larger than the actual dinosaurian brain (this ganglion was colloquially called a second brain). They—I think correctly—argued that it was highly unlikely that this ‘hyperdeveloped’ dinosaurial ganglion endowed the organism with additional cognitive abilities. Upon submission, their paper was peer-reviewed by several invertebrate zoologists who oddly enough, took emphatic exception to the idea of an ‘invertebrate brain’, and in no uncertain terms advised Drs Netsky and Sarnat to abandon the concept of an invertebrate brain. Happily, they did not, and their paper was eventually published [12].

In the end, in his 1986 review paper Dr Netsky proposed the following definition, explicitly referring to vertebrate brains:

…that part of the central nervous system in the skull; connected to the spinal cord; the seat of sense, motion, thought, and of human speech; comprising two contiguous hemispheres connected by commissures; a cortex of neurons, the gray matter, surrounds both white matter and various subcortical neuronal clusters. [7, p. 1672].

Nonetheless, in a subsequent paper, Sarnat & Netsky [12] listed a series of structural and functional characteristics between ganglia and brains that established them as fundamentally distinct entities. In the end they concluded that: (i) no known animal possesses a cephalic ganglion, (ii) developmentally, the human brain does not start as a ganglion, and (iii) therefore, a ganglion never becomes an actual brain. Thus, a crucial point would be to estimate when the brain as an established structure appears within the context of the evolutionary history of animals [13].

5. The first brain

Planarians (figure 2) include several hundred non-parasitic flatworm species belonging to the phylum Platyhelminthes. These organisms are broadly classified according to their habitat; planarians occupy freshwater, marine, and terrestrial ecological niches [14]. The best-understood group of these worms is the freshwater type, particularly the ones belonging to the family Dugesiidae, which includes the genera Girardia, Dugesia and Schmidtea, comprising more than one hundred formally described species [15]. The most striking characteristic of many of these planarian species is their remarkable capacity for regeneration [16,17]. As an example of their regeneration abilities, one of the species most adept at this process, Girardia dorotocephala, only requires a fragment of about 0.08 mm³, the equivalent of approximately 10 000 cells, to achieve complete body regeneration [18–20]. As if this fact was not remarkable enough, the regeneration capabilities of some planarian species include the complete and correct reconstitution of their nervous tissue, including their brain [21–24]. It is difficult to overstate the potential relevance of these organisms' regenerative abilities for the possible development of medical treatments for neurological diseases and injures. Interestingly, there is also mounting evidence that decapitated planarians that regenerate their heads partially retain some learned habits or training acquired prior to the decapitation event, a fact that is particularly relevant to the exploration of where memories are stored in an organism [25–27].

Figure 2.

Representative planarian species, as indicated. The scale bar represents a length of 1 cm. (Online version in colour.)

In the process of developing their ideas about the brain, Sarnat and Netsky proposed as a working hypothesis that the planarian brain could be seen as a precursor of the human brain. In other words, they essentially proposed that the planarian brain was the first evolutionary example of a vertebrate-style brain. They were very careful to explicitly state that there was no evidence that planarians were in a direct line leading to vertebrates, and rightly so, since today we know that they actually are not. Nonetheless, the planarian and vertebrate brains at the very least represent a curious example of convergent evolution. Eventually, the planarian brain evolved into an important model in the neurosciences. Additionally, it has been recently acknowledged that the usefulness of planarians in biomedical research goes beyond developmental biology and regeneration. Planarians are being developed as an animal model in the pharmacological sciences with exciting results especially in light of neuropharmacology and the pharmacology of abused drugs [28–31].

Even though planarians are not in the direct line leading to chordates, their nervous system exhibits remarkable similarities to vertebrate nervous systems. These parallels range from their molecular characteristics all the way to morphological features. For example, virtually all described flatworms species, including planarians, make use of many of the same neurotransmitter systems as vertebrates do [28,31]. Planarian nerve cells also display a fair degree of synaptic complexity in terms of the number and concentration of different neurotransmitters within their synaptic vesicles, as well as neuronal morphology that is more reminiscent of vertebrates than invertebrates [12,13]. A particularly curious aspect of the planarian neuronal architecture is the presence of dendritic spines. These structures are widely distributed throughout the animal kingdom but are most commonly found in vertebrate organisms [32,33]. Dendritic spines are important regulators of synaptic function, especially neuronal plasticity and memory formation [34–36]. Interestingly, even though the most common type of invertebrate dendritic spines are the ones found in insects, the planarian dendritic spines are more similar to the ones found in vertebrate organisms [12,13,37].

The idea of the planarian brain as an anatomical forerunner of the vertebrate brain relies heavily on gross anatomical features. First, the typical planarian brain consists of two distinct sections, which is strongly reminiscent of the well-known dual vertebrate brain structure with its well-defined lobes. Moreover, these lobes display projections that seem analogous to vertebrate cranial nerves. The best characterized planarian species in this regard, Dugesia japonica [38], possesses nine such pairs of nerve extensions (figure 3). In contrast with the vertebrate nervous system in which a single spinal cord connects to the two brain hemispheres, planarians have two nerve cords, one from each lobe. Interestingly, the planarian nerve cords are connected by additional nerve fibres, giving them their characteristic ladder-like appearance (figure 4; for an extended discussion of the planarian nervous system please see [3], chapter 9).

Figure 3.

A simplified representation of the D. japonica brain, showing its nine pairs of nerve extensions. As modified from Pagán [3]. Illustration by Alexis G. Pagán.

Figure 4.

Side-by-side comparison of the human and planarian central nervous systems. The bar represents approximately 1 m for the drawing of a human and 1 cm for the planarian diagram. As modified from Pagán [3]. Illustrations by Alexis G. Pagán.

6. The—real—first brain (s)

The point of view stating that the planarian brain is a true representative of what ‘the first brain’ must have looked like is an interesting starting point to try to gain insights on vertebrate nervous systems. However, it must be pointed out that planarians are not the most basal animals [39,40], meaning that they cannot be the absolute first example of a brain, broadly defined. Current ideas about the phylogeny of animals are also in flux; this field is undergoing a minor controversy in specialized circles. In brief, traditionally, sponges were widely considered as the most primitive animals, defined in part by the lack of an unambiguously recognizable nervous system [41,42]. However, there is convincing evidence indicating that the phylogenetic lineages leading to sponges and other ‘primitive’ animals did possess an actual nervous system that was lost along the course of evolution [43]. All animals possess an unambiguously recognizable nervous system, with the possible exception of sponges. Sponges are filter feeders; however, there are some notable exceptions: at least four species of carnivorous sponges actively capture prey in a way curiously reminiscent of certain carnivorous plants [44]. Predation in these sponge species appeared to have evolved only once and it is not clear whether these sponges possess bona fide nerve cells. However, some lines of evidence indicate the presence of relatively sophisticated sensory cells in certain sponge species, as well as complex gene networks related to sensory functions in these organisms [42,45].

More recently, molecular data seems to indicate that ctenophores (jellyfish-like active predatory organisms) and sponges are ‘equally basal’, which as a first approximation is unlikely owing to the marked differences in their morphological and physiological complexity [46]. Recent reviews of the evidence for each position (i.e. sponges first, ctenophores first, or both equally basal) expresses the idea that even though both positions are plausible, there is no conclusive evidence to make a concrete ruling on this matter [47–49]. Clearly, additional data is needed to figure out the most likely candidate of an Urmetazoan (the last common ancestor of all metazoans)-like animal, and therefore the question about the first ‘true’ brain is still an open one. Regardless of the type of animal that ends being the best candidate for an Urmetazoan-like organism, the reality is that there are many organisms that do not possess an obvious ‘brain’ yet are nonetheless capable of performing many of the functions traditionally attributed to this interesting organ.

A fair working definition of a brain is ‘the control centre of a biological system’. Planarians (see above) represent some of the earliest organisms displaying cephalization and bilateral symmetry, a property which is directly related to a centralized nervous system and therefore, by definition, planarians have an obvious control centre [50]. However, let us not forget that there are many organisms that have a nervous system but no obvious centralization. These types of organisms not only survive but are also quite successful. At any rate, the fundamental unit of every nervous system known so far is the neuron. Interestingly, it seems that the evolutionary origin of neurons—an event that happened approximately 500 million years ago—coincided with the initiation of hostilities between animal species for nutritional purposes. In other words, neurons appeared shortly before animals began pursuing and eating each other [51]. Incidentally, this information will likely play a role in the sponges-ctenophores controversy (see above), as ctenophores are true predators, a characterization that tends to indicate a relatively sophisticated nervous system.

It seems that the simplest organisms that evolved ‘true’ nervous systems are the cnidarians (hydra, jellyfish, and related organisms; [52–54]). The lack of cephalization displayed in the architecture of the nervous systems of cnidarians seems to be a direct outcome of their radial symmetry [54]. An interesting particularity of the cnidarian nervous system is that ganglia are conspicuously absent [55]. This fact is consistent with the current consensus suggesting that the appearance of ganglia is a hallmark of nervous system centralization, as opposed to the distributed nervous systems in cnidarians and related organisms [55].

The most numerous type of animals on earth, the arthropods, display a ganglia-based nervous system, controlled by a brain (which is sometimes still denoted as a cephalic ganglion), with a series of ganglia alongside a ventral (as opposed as the vertebrate dorsal) nerve cord. The presence of ganglia connected by nerve tissue in arthropods is consistent with the segmented nature of their body architecture [56]. Then there is the remarkable nervous system of the cephalopods (octopuses, squid, etc.). These animals are undoubtedly the most intelligent of all invertebrates. They show a significant degree of functional convergent evolution with the nervous system of vertebrates [57]. However, structurally, the vertebrate and cephalopod brains look nothing alike, and in fact, it has been established that they do not share a monophyletic origin [52,53].

These are just a few examples of the best-known, non-vertebrate animal brains. In this issue, you will read about examples of organisms that show rather atypical systems capable of performing many, if not all of the activities that a ‘traditional’ brain can. In the particular example of organisms lacking an evident centralized nervous system, some scholars have proposed to speak of ‘aneural cognition’ or ‘headless thinking’ [58]. These entities do undoubtedly expand how we think about brains. Let us briefly explore some representative examples of such organisms.

7. Thinking plants

Plants are one of the most curious examples of organisms displaying brain-like functions. This is an admittedly controversial assertion, since plants do not possess any kind of centralized organs to speak of, let alone even the inkling of a nervous system, at least at the macroscopic level. The Greek philosopher Aristotle classified plants as ‘insensitive’ beings as a criterion to separate plants from animals. Due to Aristotle's fame and prestige, just as his views on the brain, his views on the nature of plants prevailed for centuries. He famously stated: ‘Life is found in animals and plants. But in animals it is patent and obvious, whereas in plants it is hidden and not clear.’ In essence, Aristotle was not even sure whether plants were alive.

It was not until the 1800s that the scientific consensus about plants began shifting in favour of plants as ‘sensitive’ beings (that is, unambiguously living entities). It is well-known that in the 1880s Charles Darwin and his son, Francis, produced groundbreaking studies on plant behavior [59,60]. However, it seems that this trend formally began in 1876 with Dr William Lauder Lindsay (1829–1880), a Scottish physician and amateur botanist, who published a paper titled ‘Mind in Plants’ in (of all places) a psychiatric journal [61]. In ‘Mind in Plants’, Lindsey told the story of his interest in whether the ‘attributes of mind’ were present in plants in addition to animals. After a thorough exploration of the idea, he surprised himself when he was unable to find an unambiguous demarcation between the ‘plant mind’ and the ‘animal mind’. He extrapolated his findings to the idea that he would neither be able to find such a clear distinction between ‘animal minds’ and ‘human minds’. One of the immediate consequences of his conclusions was that from then on, plants were considered to display true behaviour. Along the pursuit of this line of thought, a minor detail came up: in animals, behaviour is a direct result of the activity of a nervous system, a system that is conspicuously absent from plants. This ‘minor detail’ did not deter a cadre of innovative, truly out-of-the-box thinkers, who began talking about the ‘nervous life of plants’. This emerging trend led to the notion that scientists could begin talking about the neuroscience of plants or as it is better known, plant neurobiology.

To say that the concept of plant neurobiology proved controversial is an understatement. One of the main challenges that this field encountered was that plant neurobiology proponents chose to use certain terms that were the traditional purview of classical (animal-based) neurobiology as metaphors (and to be clear, plant neurobiology advocates explicitly stated that they were metaphors, no doubt about that). All appropriate disclaimers notwithstanding, fiercely -territorial as scientists are, some classical neurobiologists took exception to the use of ‘neurobiology’ as applied to plants, and in all fairness, it is not hard to understand why. To begin with, plants have no ‘nerves’ to speak of. Also, ‘neurobiology’ and ‘behaviour’ are words that are (until very recently) invariably associated with animals. The situation was further compounded when plant neurobiology enthusiasts began to use terms like ‘intelligence’, ‘mind’, ‘memory’, ‘cognition’ and ‘brain’, in the context of plant biology. The debates about the use of these words as applied to plants became as heated as scientific debates can be, and perhaps a little more than that. Human emotions, though ostensibly not actually running the scientific show, do strongly influence the debate for sure. The arguments that the phrase ‘plant neurobiology’ helped initiate are perfect examples of this fact. Just to give you a taste of how the discourse is taking place, on the advocate's side, there are published papers like ‘The mind of plants: thinking the unthinkable’ [62] and ‘Plants learn and remember: let's get used to it.’ [63]. On the detractor's side, there are titles like ‘Plant neurobiology: no brain, no gain?’ [64] as well as the (rather unkind) ‘Plant neurobiology: intelligent plants or stupid studies?’ [65]. Even in light of all the disagreements that plant neurobiology as a field of study triggers in its detractors, the undeniable reality is that plants do display behaviour. They also seem to be able to display phenomena very much like memory and learning, and without a doubt, they do react to their environment. These are uncontested facts that virtually everyone in the scientific community accepts; I know of no exceptions. A thorough discussion of how this field came to be is outside the scope of this review, but there are a few excellent articles that explore this story in detail [66–70]. One thing is for sure, these are exciting times for the brain sciences and I fully anticipate significant breakthroughs in the near future.

The final examples of atypical brain-like function that I will briefly mention here will illustrate the notion that to functionally redefine the brain as a single, static, and macroscopic structure is a near impossible proposition. Some of these examples are explored in much more detail in other articles in this issue.

8. Sociomicrobiology

Probably one of the most striking examples of atypical brains includes the activities of a variety of bacterial species. Many of these bacteria were known to display social behaviour before they were thought to have brain-like properties. Early biologists described complex macroscopic structures that they called ‘fruiting bodies’, which were recognized as being bacterial colonies since the early 1980s [71,72]. In the 1970s, researchers described behaviours of certain myxobacteria (colloquially called slime bacteria, not to be confused with slime moulds, of which we will talk further along) that were correctly interpreted as an early example of cooperative behaviour. In a nutshell, these bacteria coordinate predation. As an example, the myxobacterium Chondromyces crocatus has the ability of forming large colonies that travel together through soil in the form of an amoeboid body a couple of millimetres long. Upon encountering prey, the colony releases a coordinated ejection of digestive enzymes from individual bacterial cells, eventually digesting their target (reviewed in [73,74]). For a while after this discovery, the observation of apparent cooperative bacterial behaviour was merely considered a biological curiosity, no more than that. However, beginning in the 1990s it became apparent that bacterial cooperation is a much more widespread occurrence. The discovery of multiple examples of this phenomenon gave rise to a new field: sociomicrobiology (reviewed in [75]; a less technical exploration of sociomicrobiology is included in [76, pp. 156–161]). This discipline was created in great part in response to the recognition of the phenomenon of quorum sensing, which describes bacterial communication phenomena that seem to be related to the pathogenic nature of many bacterial species (reviewed in [77,78]). This fact has many public health implications. Quorum sensing disruption is currently investigated as a possible target to develop new classes of antibiotics [79,80]. Remarkably, recent developments indicate that bacterial communities rely on electrochemical processes for communication purposes, just as nerve cells do [81–83]. Moreover, one of the most exciting developments in this area is the discovery that bacteria can communicate with each other via ion channels, in a manner quite similar to how nerve cells communicate [83]. Surprisingly, this kind of communication can occur between different bacterial species [81,82].

A second example of sociomicrobiology is the case of certain eukaryotes including several species of social amoebas colloquially called slime moulds. Their colonies are made of individual amoeba-like organisms that were recognized as distinct species as late as the nineteenth century [84,85]. The two best-known genera of slime moulds are Dictyostelium and Physarum, which are represented by several extant species. In general, slime moulds live as unicellular, independent amoeboid organisms when there are enough quantities of nutrients available. During harsher environmental conditions, the individual cells organize themselves into a macroscopic ‘slug’ whose cells show division of labour, including the generation of structures formed by cells destined not to reproduce. Interestingly, there is a fundamental difference between Dictyostelium and Physarum in terms of their collective behaviour. Dictyostelium's cells retain their individual identity within the colony, just like a multicellular animal. In contrast, Physarum's individual amoebae fuse into a single mass, forming an acellular slug [86,87]. When in colony form, the slime moulds are able to cooperate to solve problems like navigating through a maze [88–90]. Also, they are able to solve more complex problems like evaluating a variety of food sources to achieve optimal nutrition [91]. Additionally, slime moulds display neuronal-like habituation [92], and are even able to transfer learned behaviours through cell fusion [93]. The notion of ‘proto-brains’ formed by individual bacterial or amoeboid cells is something that was unthinkable until very recently. Slime moulds are currently studied in terms of their ‘proto-cognitive’ abilities [94–96].

In summary, the fact is that certain types of social bacteria seem capable of performing some of the currently accepted brain functions. Examples of these are the perception of environmental signals and their analysis and interpretation (particularly when related to nutrients), as well as the coordination between individuals in order to acquire such nutrients. Also, eukaryotes like slime moulds are able to perform the aforementioned activities and they take it one step further, namely by engaging in learning and memory processes and on the generation of true division of labour, including the apparent sacrifice of certain individuals for the sake of the community and even the loss of individual cellular identity when in the macroscopic form. Moreover, we already saw that the signalling mechanisms between single-celled individuals in prokaryotic and eukaryotic colonies are closely reminiscent of signalling mechanisms traditionally associated with nerve cells, and that at least in the case of the slime moulds scientists are beginning to speak of ‘proto-cognitive’ behaviours. Taken together, all these facts at the very least argue for the characterization of the behaviour of social bacteria and slime moulds as brain-like. It seems that the study of these organisms illustrate an example of the research directions that could lead to a fundamental redefinition of what a brain is.

I would like to finish this section with some thoughts that I expressed a few years ago about slime moulds and social bacteria (from [3, pp. 152]):

Another rather interesting trait of slime molds and related organisms is that they are capable of rather impressive feats traditionally thought to be limited to ‘higher’ animals. These include behaviors like problem-solving skills and the ability to learn. Amazingly, they also display the ability to anticipate environmental changes based on prior experience. Still, just like bacteria, the amoeba-like cells in slime molds do not possess an actual animal-like nervous system. If we think about it, from our admittedly biased perspective, the absence of a nervous system makes the behavioral repertoire of bacteria and slime molds even more astonishing. Curious creature as I am, this makes me ask myself, wouldn't it be interesting to find out whether bacterial populations have some form of self-awareness? Would it be possible at some fundamental level that slime molds, well, wonder?

9. Superorganisms

A particularly intriguing phenomenon relevant to the exploration of the concept of a brain is the phenomenon of collective behaviour in superorganisms. Here we will find a series of eusocial organisms like the usual suspects, such as ants, termites and bees, but also organisms that we would not think of as eusocial, like certain mammals (the naked mole-rat; [97]) and even an eusocial type of worm (an as yet undescribed parasitic flatworm belonging to the Himasthla genus; [98,99]). Eusocial animals represent a particularly interesting example of atypical, ‘collective’ brains, because each individual possesses a brain of its own yet the interacting individuals generate emergent behaviours.

The comparison of superorganisms with brains was proposed as early as the 1990s ([100], reviewed in [101]; [102]; reviewed in nautil.us/issue/23/dominoes/ants-swarm-like-brains-think-rp). Also, a series of predictions about how the comparison between entities like ant nests and brains can throw light into the latter's fundamental processes [103]. Overall, these authors list a series of parallels between ant colonies and brains (this comparison can obviously be extrapolated to other types of superorganisms, i.e. bees, termites, etc.). Some of the most relevant parallels in terms of this review are:

• —The overall activities of ant colonies and brains are generated by the multiple interactions of numerous subunits displaying various degrees of independence (nerve cells, versus independent animals).
• —Each of the aforementioned subunits is unaware of the single, larger entity that is generated with their actions, and their collective behaviour can be likened to distributed and parallel processes.
• —Their collective behaviour generates phenomena that result from multiple individual actions at the level of the subunit (a neuron fires or does not fire; an ant moves away or towards the nest).
• —Collectively, the above actions result in emergent phenomena like memory and decision making among others.
• —Neither brains nor ant colonies possess an obvious control centre or any evident hierarchical chain of command.
• —Just as any two brains are distinct by virtue of the different ways in which their nerve cells interact, any two ant colonies will behave in their own, idiosyncratic way; in other words, they display individuality.

10. A proposal to expand the conceptual idea of a brain

In the spirit of the question raised by Dr Netsky [7], if we ask ourselves ‘What is a brain?’, we can say that a brain can be thought of in terms of ‘a wide variety of physical substrates capable of generating the functions of a brain’. Additionally, once we establish this definition about a brain, if we ask ourselves ‘Who says so?’, we could say that the answer is ‘the scientists working on the specific fields that study these various physical substrates’. There is a specific purpose for my choice of expressing the above statements in tautological form. Virtually any and all definitions that we can offer about the brain are by necessity, arbitrary (although it is clear that some definitions will be more precise than others). The undeniable reality is that at the end of the day, nature does not care about how we choose to characterize it. Nature is; it is that simple. Our definitions are for our exclusive and explicit benefit. As biology is a science of exceptions, we can periodically redefine a particular concept and there is a real possibility that we will discover a previously undescribed class of organisms that will be the proverbial exception that proves the rule. This being said, this constant delineation of definitions is largely responsible for the progress of science and is still the best strategy we have to date.

Another hard fact is that the human brain, with its approximately 86 billion nerve cells with thousands of connections between each other, and about as many glial cells as neurons [104] seems to possess a series of abilities in common with many evidently simpler systems. In a general article titled ‘Thinking about the brain’, the late Dr Francis Crick speculated on the best tools to study the workings of the brain—in its vertebrocentric meaning—including higher functions like perception, conception, imagination, volition and emotion (and here I would humbly add consciousness). He presciently stated that: ‘It would not be too surprising if the proper theoretical tool for approaching such problems turned out to be communication theory.’ [105, p. 219].

I believe that Dr Crick's assertion will be proven right sooner rather than later. Communication theory must be one of the fundamental disciplines that will help us understand what brains are and how they work. As a first approximation, a systematic, all- (or at least most-) encompassing definition of a brain could be that a brain seems to be an assemblage of independent yet closely related units. These independent units can be individuals in a superorganism, nerve cells organized into a ‘traditional brain’, amoebas in slime moulds, or bacteria in a biofilm, among other examples. These interacting units communicate with each other in various ways that nonetheless result in similar emergent phenomena that we collectively group as aspects of behaviour, memory, and other cognition-related processes. Although in many cases the units of a brain are within the same organism (i.e. neurons in a human brain), in many other cases these units are independent (again, think bacteria, slime moulds and individual ants), a fact consistent with the concept of ‘fluid-’ or ‘liquid-like’ entities.

And what about plants? What do we do about plants? Plants represent an especially difficult challenge in light of the fundamental differences between them and animal species that possess or generate nervous-system-like activities and serve as control centres of an individual. No plant displays any indication of obvious centralized control systems. It would be interesting to observe the development of a theoretical framework of plants using abilities traditionally within the purview of brain-like systems.

It seems clear that the multiple ways in which a variety of organisms have evolved the capacity to generate phenomena that we unambiguously recognize as ‘brain-like’ prevents the formulation of an all-encompassing definition. Based on this idea, I propose an organizational scheme (figure 5) in an effort to offer a systematic classification of the various possible versions of a brain. It is to be noted that I intend for this scheme to be a mere starting point, by necessity subject to modification as the field progresses. Also, the fact that I chose representative examples of selected organisms to illustrate the particular brain ‘type’ does not preclude the possibility of including other organisms. Finally, I only include living natural organisms in this scheme, although it is clear that artificial entities are and will continue to be discussed as possible brain candidates, especially in light of current technological trends.

Figure 5.

Proposed classification of brain types (see text). It is clear that some refinement is needed; for example, please note that cephalopods display undeniably higher cognitive capacities than many vertebrate species [52,53,57,106,107], and as we saw previously, sponges are multicellular but do not show clear evidence of neuronal processes [41,42,45]. A speculative, yet intriguing possibility is that this classification system might represent a starting point to develop neurobiological theories related to astrobiological research, as scientists are beginning to talk about possible nervous systems elsewhere in the universe [108].

11. Proposed classification of brain types

This classification system represents a first approximation in an effort to expand the conceptual idea of a brain. In this approximation, I do not include criteria like intelligence, thinking, reasoning, and other so-called cerebral higher functions, although I arbitrarily use evidence of cognitive processes to differentiate between multicellular vertebrate and invertebrate brains.

(a). Type 0: No brains

This classification represent organisms that live independently, namely single-celled organisms displaying evident behaviour but no evidence of collective or cooperative behaviour at the current state of knowledge. Organisms of this type would include organisms like paramecia and other protists, whose membrane are excitable in the neuronal sense and use this property to generate behaviour [109]. Also included in this group are solitary bacterial species, which also display true behaviour as well as some neuron-like characteristics [110].

(b). Type 1: Facultative brains

Single-celled organisms showing collective behaviour (i.e. ‘proto social’ traits). These types of organisms are capable of forming cell conglomerates that closely approximate the criteria for multicellularity; these conglomerates usually arise under challenging environmental conditions (i.e. scarcity of nutritional resources). When in their ‘pseudo-multicellular’ form, these organisms display a variety of functional characteristics closely reminiscent of traditional brain processes (memory, ability to solve problems, cooperative behaviour, etc.). In this classification we would include social bacteria, slime moulds (see above) and yeast; interestingly, these latter organisms are being proposed as models to study neuronal behaviour [111].

(c). Type 2: ‘True’ brains—invertebrates

This includes one of the traditional conceptual views of a brain. This classification includes the brains of multicellular organisms that at the present state of the field unmistakably display behaviour albeit with no concrete evidence of ‘cognitive’ processes. Generally belonging to ‘lower organisms’, these types of brains are neuron-based, with the exception of plants, which at the moment remain controversial members of this group (see above).

(d). Type 3: ‘True’ brains—vertebrates

This is the ‘classical’ conceptual view of a brain, namely, vertebrate brains. This type of brain is present in multicellular, ‘higher’ organisms with clear evidence of cognitive processes. These types of organisms may or may not display cooperative behaviour, but exclude eusocial organisms.

(e). Type 4: Brains within brains—superorganisms

This includes multicellular organisms displaying true eusociality. This classification includes organisms possessing ‘traditional’ nervous systems generating higher functions upon their interaction, thus, ‘brains within brains’. Eusocial organisms traditionally include the social insects (i.e. bees, ants, termites, etc.) among other types of organisms. Some authors argue that humans might belong in this category, but the general scientific consensus argues against this idea.

12. Concluding remarks

By studying the integration of the behaviour of individual components, from single cells to whole organisms, there is the potential of discovering general principles that might govern the emergence of what we currently call a ‘brain’. There are certain entities that stretch almost to the absolute limit what we would think of as a brain. For example, viruses are biologically evolving entities that to my knowledge have never been characterized as possessing brain-like behaviour. However, quite surprisingly, certain bacteriophages are able to detect bacterial biofilm-modulating signals [112] and use these signals to influence quorum sensing activities in their bacterial hosts. Moreover, certain viruses are capable of communicating with each other in order to actually coordinate strategies to infect bacterial hosts. This is a phenomenon that was hypothesized as early as the 1940s (reviewed in [113]), and formally reported in 2017 [114]. One of the main ideas discussed here is that communication between independent living entities is a characteristic of a brain. To state that a virus population displays brain-like characteristics will certainly prove controversial, and certainly more work on this idea is warranted. As an example of these efforts, viruses are currently thought of as complex adaptive systems [115]; this point of view will certainly initiate interesting conversations and speculations on whether the behaviour of communicating viruses qualify as an example of brain-like functions.

Perhaps a brain is best described not as an actual physical object, but as an emergent process whose individual structural components are subject to biological evolution and may possess physical independence in order to collectively generate aspects of behaviour. In other words, the brain can be seen as a rather fluid or liquid entity. There is little doubt that much more work is necessary to begin to understand and apply these new concepts. An exciting fact about this effort is that it is bound to generate stimulating new ways for us to understand the brain in its multiple incarnations; perhaps even ideas that our own brains have never pondered before.

Data accessibility

This article has no additional data.

Competing interests

I declare that I have no competing interests.

Funding

I received no funding for this study.