Log in

No account? Create an account

Bacterial Cognition

This post was written using notes taken during Pamela Lyon's talk at ASCS '07.


Cognitive science must have an answer for two key questions: what is cognition, and what kinds of creatures or systems are cognitive? Descartes argued that only humans were cognitive agents due to their rationality, reflectiveness and perhaps creativity. In the 20th century, behaviourism showed that much of animal behaviour--including human behaviour--could be explained without recourse to mentalizing of any kind, simply by examining stimulus and response. The cognitive revolution in the 1960s showed that even animals such as rats must use mental maps and representations of the external world to make decisions. At the same time computers were being programmed to undertake a multitude of tasks often considered cognitive (e.g. play chess, interpret sentences (e.g. Eliza). These two research paths suggested that perhaps cognition was not the unique faculty of humans at all?

Nevertheless, solving the puzzle of exactly what is cognition has remained. It is within this context that the study of bacterial cognition is interesting. Bacteria--and by extension unicellular eukarotes--have long been considered to simple, too reactive and too determined to be a member of the cognitive gang. However, Pamela Lyon argues that this exclusion is unwarranted. She suggests that bacteria are sensitive, communicative and decisive organisms and bacterial responses are more flexible, complex and adaptable than generally believed. In terms of re-defining cognition, Lyon argues that behaviour at the microbial level is precisely what must be understood in order to comprehend how more complex and specialized forms evolved and now function. Lyon claims that cognition is part of basic biological function, like respiration.[1] A good way to examine this is through a case study of myxococus xanthus (MX).

Case study - myxococcus xanthus

M.xanthus is a gram negative rod-shaped proteobacteria (picks up a stain on a test). M.xanthus dwells in soil and organic waste. M.xanthus is predatory, territorial and highly social. Almost no stage of its life cycle is solitary. The demands of cooperative living and navigating an epistemically polluted environment significantly shape its behaviour. M.xanthus dwells in a challenging, constantly fluctuating, informationally rich environment that is teeming with predators as well as prey; e.g. one gram of soil is estimated to habour up to 10 billion microbes of possibly thousands of species, which live alongside eukaryotic organisms (plant and animal). M.xanthus is well-equipped to cope with this environment.

There are three arguments against bacterial cognition, 1) Flexibility 2) Complexity and 3) Distilty


Lay-people generally assume that bacterial responses are inflexible and unmodifiable. That is, bacterial behaviour presumed to be rigidly determined by the organism’s genetic endowment and the nature of the stimulus. It is also thought that Bacterial behaviour does not require memory, is not sensitive to context or does not involve selection among alternative actions.

However, decision-making in bacteria has been known for over 30 years. Responses integrating multiple or conflicting stimuli are not always additive or linear. Bacteria have a huge repertoire of responses including: Swarming motility, ‘Wolf pack’ hunting and deceptive signaling, e.g. M.xanthus lures E. coli within killing range.

Bacteria sense features of their environment via two types of signal transduction system: 1-component and 2-component systems. 2CSTs have separate protein sensor and response regulators. The total number of ST systems has been likened to ‘bacterial IQ’ and the predominant type (1CST or 2CST) as reflecting ‘introversion’ or ‘extroversion’. To give an idea of how complicated bacteria are, E.coli shown to have 30 different sensors and 34 response regulators of differing activities and reaction speed—and that’s just its 2CSTs. The number of STs in M.xantus is unknown. However, due to complex lifestyle it is estimated to have 5 times as many as E. coli.

These findings suggest that bacterial behaviour is highly flexible and involves complicated decision-making.


Bacteria are thought to be very simple creatures, without a sufficiently complex nervous system to constitute a cognitive agent. They are not supposed to have a life-cycle or development. However, not only are bacteria sufficently complex as individuals, they also have communal living arrangements with incredible emergent complexity (e.g.colonies, swarms, multispecies biofilms). Not only this, but the information processing systems that generate bacterial behaviour share mechanistic similiarities (not simply conceptual similiarities) with those that generate human behaviour.


A biofilm is a highly structured living arrangement with 100s of species of bacteria. An analogy can be made between the biofilm and a city because within the biofilm there is division of labour and mutual living (i.e. bacteria live off each others waste-products and productions.) The biofilm allows access to resources that individuals cannot obtain or use. The biofilm provides protection against predators, optimized survival and reproduction.

Catheter biofilm

Bacterial development

Development in bacteria refers to changes to cell morphology or function. M.xanthus forms fruiting bodies similar to slime mold. This aggregation is like ‘great herd migrations’. In the formation of the fruiting bodies, 10-20% will form spores and survive and 80% commit suicide (autolysis) to provide nutrients for the group. A small percentage become ‘sentries’ on the perimeter. Incompetent cells are provided with proteins by conspecifics to make them competent. Cheater mutants never prosper in these ecosystems/set-up. Occasionally there is the arrival of very efficient cooperators who are very good at providing proteins (like a messiah bacteria)

Under starvation conditions, m.xanthus undergoes a magnificent developmental process in which roughly 100,000 individual cells aggregate to form a structure called the fruiting body over the course of several hours.

M.xanthus fruiting body

These examples show that bacteria are complex, social creatures.


'Distality' is a concept most easily understood by contrast to proximity. The term is used to investigate the distance between objects, events or properties that are significant for the perceiver and the behavioural response. Bacterial responses are generally thought to be proximal in three ways. a) they are within direct contact of stimulus physically b) they have a small amount of mechanistic complexity involved in processing the stimulus information (aka Dennett's unity position on intentionality) and c) temporally continuous (i.e. no memory), lack of decoupling mechanisms.

Although I don't have detailed notes for all aspects of the distality argument, an amazing example of a distal capacity of bacteria is their ability to communicate. Bacteria have a chemical messengering system called 'Quorum Sensing' (QS) that they use in the biofilm. There are intraspecies as well as interspecies quorum sensing (called bacterial ‘Esperanto’). There are at least five different quorum sensing systems known so far. Some bacterial strains are known to use three systems. Quorum sensing molecules are proximal indicators of distal states of affairs (e.g. the presence of (non) conspecifics).

Another distal example involves swarms of M.xanthus have been found to migrate non-randomly toward glass beads from a distance of 10 body lengths.[2]

It is also unclear exactly why distality is a necessary feature of cognition, rather than one of the properties that certain forms of cognition feature.


This post has a small fraction of information about bacterial cognition. Even though bacteria may not be aware, they certainly have complex behaviour and decision-making worth examining. Bacteria perceive, remember, problem-solve, learn and communicate. Understanding how they make group and individual decisions may contribute importantly to our understanding of cognition across many species including humans.

[1] This reminds me of Searle's argument that consciousness must be understood as a fundamentally biological output, just as lactose cannot come from equations mimicking function, thought cannot come from computation alone.

[2] An incredible article on bacterial directed movement towards objects can be found HERE.


Heh, I remember thinking that bacteria were the smartest things around, given how they often ruined my (not-so) carefully prepared genetics experiments! Those little bastards can live just about anywhere - it's very amusing to have to assay your antiseptic now and again to make sure it hasn't become infected!

Have you heard of the autopoiesis theory of Humberto Maturana and Francesco Varela? Their Santiago Theory of cognition approaches the same sort of thing from the point of view of biology and systems theory - almost the opposite to what this author does as I gather (approaching a biological system from a cognitive background). They come to similar conclusions: if cognition involves feedback between a given entity and its environment, the domain of what is cognitive is wider than our typical anthropological bias would lead us to believe.

I appreciate the merits of such thinking for general theory of cognition and challenging our preconceptions, but there are distinctions to be drawn - regulation of behaviour by chemical perception and organelles in bacteria may be structurally similar to human perception organs and nervous systems, but the differences are just as interesting as the similarities.
Earman has an article where he points out that function instantiation is a trivial matter. An ordinary table can instantiate the most complex functions provided that its various temporal slices are interpreted as having the right input-output relations. That to me is a strong argument against features like complexity being even in the ballpark of capturing cognition. I'm sure any actual cognitive system will be complex, but I think its sheer complexity is not philosophically important.

Rupert Sheldrake has suggested that the sun, or any star for that matter, may be conscious for largely the same reasons as you give here concerning bacteria.

The Planet Earth series has a wonderful piece on various fungi. The time-lapse photography reveals a plethora of teleological behavior. I especially like the insect fungi that force individual insects to engage in bizarre behaviors in order to improve the release range of the pernicious spores.

Have you ever read that story "There Made of Meat"?
"They're Made out of Meat"

I meant.
Totally!!! I read that as a ugrad but it is fun to read it again.
...function instantiation is a trivial matter. An ordinary table can instantiate the most complex functions provided that its various temporal slices are interpreted as having the right input-output relations. That to me is a strong argument against features like complexity being even in the ballpark of capturing cognition. I'm sure any actual cognitive system will be complex, but I think its sheer complexity is not philosophically important.

What is or is not philosophically important surely depends on the question being asked? I agree that complexity as you state it as instantiating functions doesn't 'capture' cognition. Of course not! Neither does an abacus 'count' just because at each moment of time it is 'representing' a number. The interest here is not a simplistic functional analysis of discrete states of the organism, but a project to extend the domain of cognitive science beyond creatures with developed nervous systems. Recognising how many fascinating puzzles exist in the behaviour of bacteria may affect a) the type of creature we investigate to help us research cognition and b) illuminate uniquely biological processes involved in typically 'cognitive' acts such as collective decision-making.

Exactly how this will inform the cognition debate is a bloody good question. Whatever cognition is, it probably needs a bunch of the properties that bacteria seem to have in abundance.

Rupert Sheldrake has suggested that the sun, or any star for that matter, may be conscious for largely the same reasons as you give here concerning bacteria.

First of all, consciousness is not the same as cognition. Secondly, Sheldrake's suggestion points out the problem with a purely functional analysis of the mind by contrasting states of the sun. The discussion above is absolutely about biological instantiation (see my footnote 1. re: Searle).

The bacteria participate in a remarkable range of activities which should be considered cognitive when compared to other animal behaviour that is also considered cognitive. Now, maybe the way we even approach cognition is flawed? That seems likely... it is going to be fun to try and figure it out.
Yeah complexity of function is obviously not sufficient for consciousness ie Ned Blocks China Brain thought experiment:


I thought it was another good counter argument against functionalism and representationalism.

Hmmm... My comments are meant to be interpreted as supporting your rebuttal to philosophy Jeff.

I did not think that the fact that bacteria could demonstrate a large set of states was particularly relavent, because their behavior is more useful evidence that they might be worthy of attributing cognition.

I can see cognition going beyond nervous systems. One animal smart animal that is on the edge of the (orthox) cognitive divide still registers as having a nervous system, but it is a small nervous system.


I agree that DNA is probably sufficient for some cognitions.
Do you know if the proceedings will be published? I'd like to read her paper and its references, since I'm kind of ignorant here. Despite that, i'll continue talking, and likely make this disclaimer redundant.

I, too, immediately thought of the Chinese Room. It feels like this points at an implementation of a simplified (generalized?) CR. It also points out what i consider to be the problem with the CR: it begs the question of immutable tables and therefore omits learning altogether.

It seems like one of the defining features of cognition would be a learning mechanism that alters the rule table, above and beyond environmental memory. In the case of the CR, this seems to add a requirement for "understanding" the tables in order to rewrite them. In the bacterial case, it could preclude the view that these systems are cognitive, as they don't learn. The learning mechanism is most likely selective pressure, so the learning mechanism stops applying to the individual (or individual system) and begins applying to the species as a whole. It even segues nicely into the smooth distribution between instinct and learned reaction, but there are probably a lot of little details to sort out there.
Have you read about Gerald Edelmans Theory of Neural Selection?

When I talked to the course coordinator of bioinformatics at University, and I was trying to explain that bacteria demonstrate altruism and other intelligent behavior he said `yes, but that is controversial it is safer to refer to it as adaptive behavior.'

Edelman probably argues that even higher order consciousness is the result of adaption, not just genetic adaption, or neural adaption (survival of the fittest neural repotoirs).

I think Edelman is argueing that each individual brain has information relative to itself - a stance not dissimilar to Searles `consciousness is a biological act like the secretion of bial acid in the intestine.'
Well, it again depends on what 'rudimentary cognitive ability' requires. Individual bacteria exhibit memory, decision-making etc... which are features of cognitive agents. They also act collectively.
This is a really good read for me, Must admit that you are one of the best bloggers I ever saw.Thanks for posting this informative article.
hermes handbags Coach handbags hermes birkin handbags hermes purses discount hermes handbags hermes kelly handbags hermes lindy handbags hermes wallets hermes birkin bags burberry handbags hermes handbag coach handbags

August 2016



Powered by LiveJournal.com