The evolutionary history of integrated perception, cognition, and action

[gmirey]

Wanted to draw a first plate about very early animals, and the evolution of early neurons and muscles, for integrating the first steps of P. Cisek’s model into it, but there are considerable ongoing debates on these questions…

Some lineages, whose molecular studies reveal as more ancient than seemingly “more primitive” others, show muscle and nerve cells… skipping a clade or two before muscle and nerve cells are found again… so it’s quite complicated to make sense of a sequence of events in this. Also, those splits happened long before the times for which we’re able to find fossils (and as a matter of fact involve soft-bodied animals only), exacerbating the fundamental issue with paleontology studies:

(one strip from The Far Side by Gary Larson)

The following main animal clades seem consensual, but the relationships between these seem to inspire as many theoretical models as there are combinations. Or close:

• Deuterostomia (leading to vertebrates => us)
• Protostomia (molluscs, arthropods, annelids)
• Xenacoelomorpha (quite simple multicellular things, basal bilaterians?)
• Cnidaria (sea anemones, jellyfishes)
• Placozoa (simplest of all)
• Porifera (sponges)
• Ctenophora (comb jellies)

There seems to be some consensus that Deuterostomia and Protostomia together, excluding all other above, makes a valid clade (we’ll call that Bilateria for now)… and that this clade in turn, together with Cnidaria, would forme a clade excluding sponges and comb jellies… but Xenacoelomorpha could be inserted as basal bilaterians, or even deuterostomes, and Placozoa + Cnidaria could turn out to be an all-exclusive clade, too.

So, although I believe that the following plate by Dr P. Cisek is a very good call for the evolution of early nervous systems, there are still some uncertainties and the possible emergence of simpler lineages at various points in-between the proposed steps. The neural arrangement of the ancestral bilaterian, here seen with simple and ventral “BNS”, may also be challenged.

In blue, BNS: Blastoporal Nervous System, responsible for locomotion and ingestion.
In tan, ANS: Apical Nervous System, with photosensitive and chemosensitive receptors, and controlling energy homeostasis.

ANS would in particular be able to know about “hunger” and availability of food in the immediate environment, signalling with dopamine whether the BNS should behave in “exploit” (feeding there) or “explore” (go somewhere else) mode… (cf. Levy walks).

Whether metabolic or behavioral, P. Cisek stresses that each function should not be studied as “input-output”, but as part of a “loop” whose purpose is always a return to an equilibrium. Eg. “exploration” : moving towards a place with more food, if successful, will ultimately allow to reduce hunger, lowering the need to continue exploring. Those loops can (and will) interact together (or support each other, as with the loop for “feeding” in the above example).

The fusion of ANS and BNS happening in the rostral part of the critter would form the basis for a “brain”, which will evolve afterwards, in the various bilaterian lineages.

[gmirey]

@gmirey

We’ve reached to P. Cisek and may be able to get some illustrations after all.

Some newer and better quality video (with an integrated slide player of some kind) for a presentation of the evolutionary view.

[rhyolight]

Researching Paul Cisek’s papers in preparation for upcoming journal club (Affordance Competition Hypothesis).

[gmirey]

Paul Cisek allowed us to get a first hand peek at his poster, to be presented at Society for Neuroscience meeting next week

(this image should be clickable to increased resolution if you wait a little while, but I did not manage to upload an even better one, to read the smallest characters)

[edit] Trying to host with better resolution, click for a high-res view…

pcisekposterfull.png

[gmirey]

I find there’s a nice link between Karl Friston speech below and thoughts about “control loop” and “affordances” by P. Cisek evoked here, and perhaps sensorimotor developments of HTM.

[Bitking]

I think of this as a sign that the field is starting to close in on consensus regarding the central themes after a paradigm shift. “New” ideas fit together nicely with the “new” general framework even though they are coming from radically different directions using very different methods.

I like to think about this using the shifts to the plate tectonics, evolution, or the periodic table models.

Before these emerged there was clearly a huge trove of facts and theories to try and fit observations in these related fields together. Some made more sense than others but they were not entirely successful in explaining what was being seen.

Once the “correct” model emerged all the facts started to fit together and predictive power started to emerge. I see this bit (predictive power) as one of the strongest confirmations that some field has rooted out the bedrock foundations that will stand the test of time. There may be some tuning around the edges but the basic model will stand this without falling apart, in fact, it is likely to get even stronger.

Failure to predict, such as planets in the “wrong” place or unexplained expansion of large scale celestial structures, drive further searches for a better model in a field.

The true root of science is “Hmmm, that’s odd?”

[weiglt]

Hi there,

i had a deep look into one of the more simple diagrams from Cisek and im not very sure if I understand it the right way.

I really have a problem with the interface between the environment and the ANS system. When there is a inhibitory or a exitatory feedback then it requires sensory neurons. So in this model there are two different types of sensory feedback.

Feedback 1:
At the one hand we get feedback from the environment trough maybe chemical receptors or similar stuff. This sensory feedback detects if there is food near the system or not. If there is no food than the sensory feedback inhibits the food neuron/state. If there is food near by the sensory feedback is exitatory and exite the “Food Neuron”.

Feedback 2:
There is also a feedback about the nutrient state/(inner system state). This means that if the “hunger neuron” is exited there are two different strategys to act on. But the selection of the action is dependent on the environmental feedback. If the system has hunger and there is no food then do action one and if there is food and the system has hunger than do action two. If there is no hunger need at all, the “hunger neuron” is inhibited through inner senory neurons (for example in the stomache) and there is no action selection needed in context of this feedback.

My try:
I tryed to make a simple neuron model of the above diagram but i have some issues because of the sensory feedback.

My model only works if there are always pairs of sensors which act in different directions. For example the two sensory neurons S1,S2 which collect feedback from the environment should act like this. If there is food near by the system sensor S1 fires action potentials to the Food Neuron.
And if there is no Food at all the sensor S2 inhibits the Food Neuron.

I have the same problem with the Neuron for the impetus (Hunger). I need two neurons. One which is firing when the systems has hunger and one neuron which inhibits the Neuron (Hunger) when the system has no hunger.

I never read anything of such neuron pairs so i have my doubts to this model. But its very interessting because it should work for action selection quite well which this configuration.

There is also another question which arose from this model. It is a kind of competition between the actions. Because when there is no Food near the system the main exitatory drive for action selection dont come from the “Food Neuron”. Instead it comes only in an exitatory way from the “Hunger Neuron” if this is exited. So from the “Hunger Neuron” there is a exitatory signal to both of the potential actions. And this is the point where competition makes the run between the two actions.

This competition between the actions would mainly be fought when there is no food near hte systems. I

Maybe one of you have some interesting thoughts on that because im not sure if i understand Cisek in the right manner. But it would be great to hear your feedback.

Best regardes

Tom

[gmirey]

@weiglt

Hi there,

i had a deep look into one of the more simple diagrams from Cisek and im not very sure if I understand it the right way.

Hi there, nice to see someone actually digging

I have several remarks to maybe help you get on track:

• I don’t know if the ANS and BNS “nets” are thought connected (with axons) in that primordial critter… in fact P. Cisek stresses that the signalling strategy between the two involves dopamine. I don’t know much about that neurotransmitter at this point, but I had the impression, reading that part, that it could use a much more diffuse way of propagation than direct synaptic contact.
• Anyway… diffuse or not, you’re quite allowed to represent its effect with either positive or negative arrows or conceptualize them as similar to axonal ‘wires’… I guess it’s okay in a schema.
• So… the question about sensor ‘pairs’. There are in fact lots of examples in the animal kingdom for the pairing of antagonistic sensors (or secondary relays) so that as much of a ‘low’ as a ‘high’ signal is finally carried towards an area. In humans, we have, eg, Cold vs Hot cells towards somatosensory regions, and On vs Off bipolar neurons in the retina, towards thalamus, then V1.
• But… I know about those from quite a complex creature (me ). Shall that ancestral critter also possess such pairs? I don’t feel they are necessary. You don’t need an opposite ‘inhibitory’ signal to have that ‘food’ cell in your schema stop firing when there’s no longer a ‘food’ sensor firing to drive it. It will stop naturally. And as a matter of fact, you don’t really need that food cell as a relay any more. Sensor either fires, or it does not… Food sensor is the food cell. And part of the ANS already. Same reasoning for the hunger detection part.
• Hunger and Food availability in the environment compete with each other (one excitatory, one inhibitory) for the control of another cell, which is still ‘inside’ ANS: that one will be active when the critter is in good shape (satiety or food-available), and release dopamine towards BNS… and releasing less dopamine towards BNS if it would be a good idea to travel farther (hunger and no food in the vicinity).
• BNS can then make a choice between two strategies (and motor implementation). ‘Exploit’ cell is directly activated when dopamine is high. ‘Explore’ cell gets inhibited by ‘Exploit’ cell, and can be driven to fire otherwise (there should exist “inherent” firing loops all over the place, easily able to drive this guy to activity, in the absence of inhibition). And… yes, choosing among those two strategies is a form of competition indeed.
• Note that my usage of “cell” above to refer to sensors or intermediate variables is not to be taken literally. The simplest of digital schemas could indeed represent each as one ‘cell’ and be enough (and efficient). But those may also have been be multi-somatic centers in the real critter… if only to ensure of less errors by redundancy… or to sample more of the environment… or to ensure enough levels of dopamine are able to be transmitted… or… whatever.

Hope that helps

[weiglt]

Hello gmirey,

wow this helps a lot.

• I don’t know if the ANS and BNS “nets” are thought connected (with axons) in that primordial critter… in fact P. Cisek stresses that the signalling strategy between the two involves dopamine. I don’t know much about that neurotransmitter at this point, but I had the impression, reading that part, that it could use a much more diffuse way of propagation than direct synaptic contact.

• Anyway… diffuse or not, you’re quite allowed to represent its effect with either positive or negative arrows or conceptualize them as similar to axonal ‘wires’… I guess it’s okay in a schema.

I think you are right. We probably can abstract the signaling between different connenctions this way.

• But… I know about those from quite a complex creature (me ). Shall that ancestral critter also possess such pairs? I don’t feel they are necessary. You don’t need an opposite ‘inhibitory’ signal to have that ‘food’ cell in your schema stop firing when there’s no longer a ‘food’ sensor firing to drive it. It will stop naturally. And as a matter of fact, you don’t really need that food cell as a relay any more. Sensor either fires, or it does not… Food sensor is the food cell. And part of the ANS already. Same reasoning for the hunger detection part.

I didnt thougt about only one sensor But this makes quite sense. I modeled the extra sensory neurons only for better understanding the whole thing. I know they are obsolete but for understanding quite helpfull.

• Note that my usage of “cell” above to refer to sensors or intermediate variables is not to be taken literally. The simplest of digital schemas could indeed represent each as one ‘cell’ and be enough (and efficient). But those may also have been be multi-somatic centers in the real critter… if only to ensure of less errors by redundancy… or to sample more of the environment… or to ensure enough levels of dopamine are able to be transmitted… or… whatever.

This was very helpfull for better understanding. I know that this is only a very abstract way of seeing this problem. The reality is like always much more complex. But for understanding how behaviour maybe emerges from sensory input in context of certain inner drives it is quite useful i think.

So this is the updated version without the inhibitory sensors which make sense. There is sensor S1 which firees if there is food near by and sensor S3 which fires when the system has hunger. Dont understand me wrong but i see this in a very abstract way.

Thank you a lot for your great help. I think if we understand those circuits a little bit more we can understand the more complex stuff a little bit easier in futur.

Best regardes

Tom

[gmirey]

@weiglt

Dont understand me wrong but i see this in a very abstract way.

No worries See, P. Cisek’s schema, as a starting point, is in fact even more “abstract”, I believe.

His nodes are not always intended to represent actual neurons (or bunch of them) but also… simply… concepts. I guess. You know, his original “required nutrient state” as driver of a “hunger” concept, and “nutrient state” as an inhibitor… need not to be actual cells, or sensors. I can conceptually imagine a cell able to detect current nutrient states, and comparing that to an evolution-tuned, internal “requirement level” to decide whether it ought to fire or not. Know what I mean? This whole ‘one plain arrow + one dashed towards same “hunger” node’ part of the schema could just be… one cell.
I understand this could have been quite confusing if thinking about them in terms of neurons.

I guess his choice of such an abstract schema make sense, still. To have all those factors up-playing or down-playing particular loops explicitly described, with emphasis on them, much more than, say, on cells themselves (and I believe we have little clue about the actual cells implementing those… I’m only quite confident that such an abstract schema is, as in your drawing, nicely conceptualizable with the basic building blocks we accept were available at that time).

Like, in your drawing, you’ve represented “Food” as directly inhibiting the 2nd strategy (exploring). This is true at the abstract level. If trying to represent actual cells (or groups of them), then, from Cisek’s description, I’d rather have a single excitatory arrow (dopamine) between ANS and BNS. And thus make that 1st strategy inhibiting the 2nd locally in BNS. (You may still keep that “neuropeptide Y” arrow from “hunger” in ANS to increase drive of both strategies in BNS, though… I simply did not develop that part in previous reply).

[Bitking]

Bats and Dolphins Evolved Echolocation in Same Way

The discovery that molecular convergence can be widespread in a genome is "bittersweet,” Castoe adds. Biologists building family trees are likely being misled into suggesting that some organisms are closely related because genes and proteins are similar due to convergence, and not because the organisms had a recent common ancestor. No family trees are entirely safe from these misleading effects, Castoe says. “And we currently have no way to deal with this.”

This paper casts doubt on the clear line of evolution of features, clouding the certainty that this or that feature is part of a regular progression.

[weiglt]

His nodes are not always intended to represent actual neurons (or bunch of them) but also… simply… concepts. I guess. You know, his original “required nutrient state” as driver of a “hunger” concept, and “nutrient state” as an inhibitor… need not to be actual cells, or sensors. I can conceptually imagine a cell able to detect current nutrient states, and comparing that to an evolution-tuned, internal “requirement level” to decide whether it ought to fire or not. Know what I mean? This whole ‘one plain arrow + one dashed towards same “hunger” node’ part of the schema could just be… one cell.
I understand this could have been quite confusing if thinking about them in terms of neurons.

yes its clear that he talks about concepts. I always try to understand it in terms of neurons but this could lead to problems sometimes as you said.

Thank you a lot for your response. You helped me with understanding this quite a lot:)

[cisekp]

Hello everyone. I just joined the forum and wanted to say how excited and flattered I am by all of the attention you’ve paid to my work. There is a lot to read through, here and in the other thread, so it will take me some time to absorb it and formulate some replies. But I will try to reply to the questions raised (at least to those where I think I know the answer) in the near future.

I should admit up-front that I’m not familiar with the HTM theory, so I can’t comment on the comparisons that some of you have made between it and my work. But I will read some of the papers and try to bring myself up to speed. It is always nice to find kindred spirits!

[Same post as on the “affordance competition” thread]

[cisekp]

Indeed it’s great that you’re digging into this!

Yes, my diagram was meant to be pretty abstract, so those excitatory and inhibitory arrows are not necessarily neurons (In fact, that whole scheme could be implemented by many unicellular animals). But for ancestral eumetazoans, with that ANS/BNS distinction, there is some specialization. The state control system roughly corresponds to the hypothalamus (or more precisely, its “terminal” or rostral segment), in which the signalling is primarily hormonal. That’s probably the case for the “tegmental neuropile” of lancelets, and of course our hypothalamus does a lot of hormonal control. In fact, chemical sensation might not even need “sensors” because chemicals directly influence metabolism and “downstream” systems as long as you have the right receptors in your membrane.

In any case, in modern animals neuropeptide Y (NPY) does appear to be a big part of a hunger signal, and the shift from locally exploiting to long-range exploring does appear to be governed by dopamine. Strictly speaking, originally it didn’t have to be an explicit competition, but rather a dopamine-dependent influence on how far you move before turning: As dopamine levels go up, you turn more, so you stay more local. Eventually, however, I believe the distinct needs of exploiting versus exploring motivated explicit specialization of different systems within the “peduncular” or caudal segment of the hypothalamus, which expanded into the telencephalon. Within it, the ventrolateral pallium (future piriform, insular, and neo-cortex) specialized for exploiting locally (learning local appetitive and aversive cues), whereas the medial pallium (future hippocampus) specialized for exploration (learning landmarks, odor gradients, etc). So at that point, some of those arrows correspond to neural circuits. And they make predictions - high tonic levels of dopamine should excite the ventrolateral circuits and inhibit the medial circuits.

[cisekp]

Indeed, it is sometimes challenging to determine whether shared features are due to common ancestry or whether they are due to convergent evolution. The key is to look at a lot of species whose phylogenetic relationships are known. In the past, phylogenetic relationships were determined by shared features, rendering the whole exercise dangerously circular. But now we have genetic methods that are not only independent of any morphological issues, but can be validated by gigantic data sets (especially by looking at “noncoding” regions of the genome). So we can derive phylogeny first, even with some estimates of timing, directly from the genome, and then only later do we start looking at the features. Again, as we look at more and more species, the question of homology versus convergence becomes easier to determine. That is not to say that pitfalls don’t remain - that’s why we need to be careful.

The case of echolocation is actually a pretty easy example. Since bats and dolphins are very far from each other and very few of the intervening cousins have anything like echolocation, it’s clearly a case of convergence. What is interesting is that the mechanisms that evolved in those independent branches are so similar. This is actually remarkably common, simply because we all live in the same world with the same physics.

[edpell]

Cisekp, I was surprised to learn from you that the neo-cortex is not new. Can you tell us a bit about how it has changed particularly in terms of the number of layers, now six, only five when marine mammals returned to the sea. Thanks your work is uniquely informative.

[gmirey]

@edpell

only five when marine mammals returned to the sea.

From that figure alone (with which I’m not familiar, but still…) I would tend to refrain from drawing the conclusion that 4, 5, 6 layers was a timely evolution which somehow “did not get to 6 at the time of cetacea divergence”.
Making more sense is neocortex at the mammalian (not placental, mind you… mammalian. That means kangoos and, wow, echidna) last-common-ancestor.
…
And from that quite well established common ground, some divergence for cetaceans (even if it somehow seems to “revert” to 5, this does not sound much of a “reversion” to me).

I think it was Jeff in some of the videos, pointing out that there seemed to be “two similar interleaved circuits” among layers of the neocortical sheet. Which, to me, would call for a one-time “oops, replicated that once too many, boss” kind of error.

[rhyolight]

@gmirey

Which, to me, would call for a one-time “oops, replicated that once too many, boss” kind of error.

An under appreciated mutation, indeed.

[gmirey]

Further info about isocortex specifically and our mammalian ancestry can be found in Aboitiz & Montiel papers mentionned by P. Cisek.

such as, this one, although I haven’t read it with full attention yet

Olfaction, navigation, and the origin of isocortex

showing