More city walls

Prompted firstly by the paper (reference 1) introducing a themed issue from the Royal Society called ‘Liquid brains, solid brains: how distributed cognitive architectures process information’, and secondly by a paper in that themed issue (reference 2) about homeostasis and termites. With the closing sentences of this latter paper being: ‘To put the matter provocatively: brains are cognitive systems that implement a telos [a term from the Ancient Greeks, roughly the purpose or drive which distinguishes life from the rest of creation] towards which the system trends. Homeostasis – that fundamental property of life – is the source of the telos. This is fundamentally ecological in character, shaped by the extended homeostasis and extended cognition of the many little lives the extended organism comprises. It is there, I assert, that a coherent theory of brains in all their forms may be built’.

The present purpose is rather more mundane, that is to say to speculate about the explanatory power of homeostasis in describing life forms, small and large. Including paying some attention to the sort of collective life forms of particular interest to the authors of references 1 and 2, for example the slime moulds, the collective animals (like corals) and the social insects.

I might add that homeostasis is much appealed to. It figures, for example, at some length in reference 3 and rather more briefly in reference 4.

Figure 1

I start with some homeostatic basics.

In the figure above we have a life form A, perhaps an amoeba from biology at school, perhaps some part or organ of some living organism like a liver, perhaps something else altogether like a seal, a city (with city walls, a notion first noticed at reference 6) or a termite mound, living in environment B, perhaps a fresh water lake or the salt sea, perhaps a Petri dish. There will be a clear boundary between A and B, perhaps some kind of semi-permeable membrane. B may or may not have a clear-cut external boundary. B may or may not be a life form in its own right.

We think in terms of both A and B being, at least in some large part, liquid, which liquidity facilitates transfers between the two. We make the simplifying but untrue assumption that these are well stirred liquids with simple physical properties like temperature and pressure and simple chemical properties like the amount of salt present. We suppose there to be just one set of properties, just one list of properties, with A being distinguished from B by there being at least some properties for which they take different values; differences which have to be maintained in the face of decay, decay maybe arising from the second law of thermodynamics which features in reference 6. Maintenance which is possible because A is a life form with intentionality (a technical term which I may have got wrong. Maybe it should be telos, as above. But see reference 7), even if B is not. We also allow A to have some properties which are not appropriate or relevant to B and vice versa. This is suggested by the two ‘NA’s in the figure, the not applicable of demographers.

The boundary between A and B has to be semi-permeable. Impermeable in order to maintain the differences between the two liquids, which would otherwise get mixed up, leading to the death of A. Permeable in order that A and B may interact, that A may draw it nutrients from B and expel its waste products to B. No interaction leads, at least eventually, to death. Generally speaking, the differences between the two liquids have to be small, with modest property gradients between A and B: this way A can manage the flows. Big differences and the flows would get beyond its control.

Generally speaking, A has reasonable control over its own properties, far less control over the properties of B. Although that said, A might be a B subcontractor, with the subcontract being about the maintenance of the rest of the interior of B.

To keep things simple, we suppose that all these properties take real values, positive, negative or zero. Some of them will be non-negative. Sometimes zero will have physical meaning, as in no salt; sometimes not, as in temperature, not usually measured from absolute zero, this last being far too low to be of much everyday interest.

Homeostasis is about the fact that most of these properties need to stay within fairly narrow limits if our entity is to survive and prosper. The rate of chemical processes, for example, is very sensitive to temperature, amongst other things. And brains, for example, do not work at all well if they are too cold or too hot. With the result that a good part of the energy, the activity of the life forms A, is devoted to homeostasis, maintaining their status quo, to maintenance and repair, rather than growth, recreation or reproduction.

This leads onto the next figure.

Figure 2

Here we still have life form A living in environment B. But we have added the entity X, some part of A, some process within A, a process which might be pervasive in A or which might occupy some particular place, some particular location in A. There might be just one such X (for example, the liver), there might be lots of them (for example, the billions of neurons in the brain). And despite the processes just mentioned, the usual idea is that all these entities are real things which you can see, if not with the naked eye, at least with an optical microscope, with no need for electrical trickery. In any event, the X will only prosper when the properties of A are kept within surprisingly narrow limits, this keeping being homeostasis. And A, being a life form, knows about this and works at it, within the limits set by the properties of B, and there may well be other entities Y, in A, whose business this is.

Scott Turner spends quite a lot of space in reference 11 on organisms like A which subcontract some of their homeostasis to entities Y which are in B rather than in A. Maybe close to A, but not actually an integral part of A. This might be called extended homeostasis.

And Damasio, in reference 3, talks of vague feelings of unwellness being generated when there is a serious failure of homeostasis. With part of the human problem being that there is no simple map from the vague feelings of unwellness to the root cause, even supposing there to be just one such, which is often not the case, even with physical rather than mental disorders.

It is easy enough for those who like modelling by diagram to get carried away here and we have at least three possibilities for complications: hierarchies, touching and gates.

Figure 3

One might have various types of entity, suggested here by the various shades of grey, and one might have rules about which entities are allowed to be in a subject-host relation.

Figure 4

One might allow entities to touch, either in a peer-to-peer way (left) or in a subject to grandfather way (right). Touching which would promote interaction between the entities so touching, interaction which would not have to make it across an intervening stretch of water (as it were).
Such touching might be more or less permanent, or it might be temporary, meeting some passing need.

Figure 5

One might want to model the transfers of material between two entities in more detail, going beyond a more or less homogenous, semi-permeable membrane. So here the circular entity has six gates, where the enlarged gate on the right might be open for traffic from left to right, one of many permutations available. A gate might, for example, only be good for a certain kind of chemical.

The gate, the dark blue square, is supposed to have some intelligence. It knows about the properties of the liquid to the left and about those of the liquid to the right and applies some algorithm to those properties to decide what to do about the gate. Open, shut, whatever. Which serves to point up an important property of homeostasis: homeostasis is a cognitive process with knowledge of state, present and desired. Perhaps also, the past. Knowledge which is distinct, physically and conceptually from the subject of that knowledge. Knowledge as signifier. Knowledge which is a basis for action. For discourse on which see reference 8. Something which non-life does not do, at least not to my knowledge, other than man-made things like computers. On the other hand, non-life can store plenty of history, possibly remote history rather than self-history, of which there is plenty in the stalagmites and stalactites of reference 9. One difference being that they do not do anything with that history; they might grow but they are otherwise inert.

Part of the point of all this is that the left hand world (of the gate at the right in the figure above) is going to be different from the right hand world. They are both some kind of liquid, possibly but not necessarily a watery liquid, both containing all kinds of chemicals, all milling about and mixing – but the mix on the left will be different from that on the right, a difference reflecting their differing functions and places in the world, a difference which takes work and energy to maintain. An arrangement which echoes the city wall of reference 6.

Using these three devices one can build elaborate models. But would such models help, and would they be much to do with the homeostasis of biologists?

Non life forms

At this point, I wondered whether one could find examples of this sort of thing, of A’s, B’s and boundaries, which did not involve life. And while I did come up with some big examples, like an earth sitting in its solar system or a continent sitting in its ocean, I did not come up with any small examples, of the same order of size as mammals. I did come up with the flint, beds of which grow in chalk, but on reflection flints do not qualify. Being solid, the speed of their interaction with their environment is very slow. Worse, they are very bound up with life, with chalk being dead life, with some flints almost amounting to fossils and with other flints growing in the tunnels left in the chalk by ancient burrowing animals. See references 13 and 14.

In the middle years of the last century, some people working on the origins of life enthused about crystals, which can exhibit very florid and plant-like growth. But in the end, this turned out to be nothing much to do with life, with the mechanisms for growth being entirely local and there is no action at a distance, no knowledge, no homeostasis. A crystal of copper sulphate floating in a solution of same is not the same as an amoeba floating in some nutritional medium in a Petri dish. A different order of being altogether.

The example of the blood supply

Figure 6

Blood is the well known red fluid delivered by the vascular system of hearts, arteries, veins and so forth, to all parts of the body – with our concern here being human bodies. Very roughly speaking, the blood is well mixed up and is pretty much the same all over.

The most important function of blood is the delivery of inputs and the fetching away of outputs. With glucose and oxygen being two such inputs and carbon dioxide being one such output. Note that glucose needs to be kept within its proper bounds, with low levels being a problem for glucose greedy organs like brains and with persistently high levels in the blood causing various kinds of serious damage – and for this reason the body includes complicated machinery for regulating the level of glucose in the blood, machinery which can go wrong.

Other functions include support for the immune system, water regulation and temperature regulation.
In addition to functions of this sort, the blood also carries a great deal of information, in particular about malfunctions of the various organs and systems which make up the body as a whole, information in the form of the presence or concentration of this or that chemical which modern testing machines can detect. The body makes use of some of this information.

One example here is the prostate-specific antigen (PSA), otherwise gamma-seminoprotein or kallikrein-3 (KLK3), mainly produced in the prostrate gland to help semen along. But it does leak into the blood, where elevated or rising levels may indicate prostrate cancer. The catch being that deciding at what level further testing is appropriate is not easy, not least because this level may vary with race and other factors. So PSA is useful in the semen but is not useful in the blood where it is an information carrying waste product.

Another example is C-reactive protein (CRP) produced by the liver in response to inflammation, with inflammation itself being a healthy response to unhealthy damage. The idea is that this protein binds to dead or dying cells in the inflamed areas so as to facilitate their disposal by other parts of the immune system. Rough speaking, high levels of this protein in the blood indicate that something somewhere is wrong. So CRP is rather more than an information carrying waste product, it is being transported by the blood for a reason, with a purpose.

A forced example

The hierarchy of Figure 3 above prompts the thought that maybe one could think of a computer in these terms. That we have a computer running a whole host of processes, arranged in a calling hierarchy.

Processes which need to be able to draw resources from their environment, perhaps memory, disc or processing time. Process which take information as input and give other information as output.
And some of these processes would be about maintaining the environment, rather in the way that the kidneys and the liver look after the blood. So one might have a process which makes memory space or disk space available, a process which is perhaps empowered to clean rubbish away. One might have another process charged with clearing away processes which have died or which are otherwise redundant.

All this can be dressed in vaguely homeostatic clothes, but such dressing does not seem to add much value.

A variant would be to imagine a more biological computer. A computer which might be a watery container for lots of processes. With the data being the make-up of that water, that environment, assumed to be well mixed. Water which might, in Visual Basic speak, be regarded as a collection of global variables. While a process might, for example, detect the presence or concentration of some chemical in the environment, and its contribution might be to signal that presence or concentration with some other chemical. One might construct logic gates along these lines.

One might easily have rules, one might need rules, yielding limits within which this environment needed to be maintained. But here again, the homeostatic clothes do not seem to be much more than that.

Organising principles

Newton came up with gravity, an organising principle for heavenly bodies which was enormously successful. He brought order out of something which while not chaos, was sometimes a bit chaotic. Gravity was enormously successful.

Other people have been searching for an organising principle for life, or perhaps for brains. One such is the homeostasis of reference 2. In an earlier book, reference 11, Scott Turner talked of energy, of the need for life forms to capture usable energy to keep the second law of thermodynamics at bay.
With the Davis of reference 6 putting some flesh on exactly how this keeping the second law of thermodynamics at bay was done at a molecular level, a trick that only life forms, with genes, can manage.

Others, for example Friston at reference 12, talk of free energy, by analogy with the concept of the same name of statistical mechanics, but here an information flavoured quantity to do with errors in prediction.

So is homeostasis our Holy Grail?

Summary so far

For the life form A of Figure 1 above to be in good working order, the inside properties need to be inside fixed ranges, possibly quite narrow ranges. For it to prosper, the outside properties also need to be inside fixed ranges, but with the ranges of the outside properties often being a lot less demanding than those of the inside properties.

In some sense, our life form knows about all these properties and is able to act on the inside properties, is able to change them. There will often be a knock-on effect on outside properties, although, to the extent that the outside is a lot bigger than the inside, this effect is apt to be small to the point of vanishing – although if numbers are large and timespans are long this may no longer be the case. There is knowledge involved here, a cognitive loop – a loop which, as is explained in references 1 and 2 is not necessarily implemented by our sort of central nervous system, nor, indeed, involve neurons at all. Our life form does have some control over its destiny.

Chemicals will be moving in both directions across the boundary, possibly but not necessarily moving down a density gradient. The life form is likely to include pumps, for pumping chemical uphill, as it were, as well as simple orifices or semi-permeable membranes.

There are chemical reactions going on the whole time, usually at great speed, at least in relation to our rather slow, conscious way of experiencing the outside world, in both the inside and the outside worlds.

As noted above, some of these movement and reactions can be described as cognitive loops.
If the boundary is breached in a serious way, the inside and outside worlds are going to get mixed up. The inside world, the chemical reactions keeping our life form going will collapse and it will die. All the stuff which had gone into it will be rapidly reabsorbed by the outside world and it will be, more or less, as if it had never been.

Otherwise, the life form is always subject to decay. In the ordinary course of things, these properties are going to drift out of their proper ranges. Some of this will result from the boundary being buffeted, pounded, as it were, by the seething chemical soup of the outside world. And, as noted just above, it is likely that there will, from time to time, be more serious external events, triggering step changes in some of these properties.

The life form may be growing and will be attempting to reproduce itself from time to time. All of this requires it to be in good working order, which includes homeostasis, and to have a good supply of fuel.

Conclusions

This is all well and good. Homeostasis provides a framework on which to hang a description of large parts of life forms; a good part of the business of life forms being the maintenance of homeostasis. But it is not the whole business.
 

Figure 7

For example, homeostasis does not help with the plumbing nicely diagrammed at reference 10 and reproduced above: this plumbing may be an important part of maintaining homeostasis, but the idea homeostasis in itself does not help particularly when discussing the plumbing; the issues are of a different kind, of a different order.

So maybe the lesson is that one does not want to push any one explanatory vehicle for life, or for brains, too far. It is unlikely to be up to the whole job. Which is a lot more complicated than Newton’s planets seen from a great distance – and the handsome differential equations which describe and encapsulate his law of gravity in the motions of these planets.

References

Reference 1: Liquid brains, solid brains - Ricard Solé , Melanie Moses  and Stephanie Forrest – 2019.

Reference 2: Homeostasis as a fundamental principle for a coherent theory of brains - J. Scott Turner – 2019.

Reference 3: Self comes to mind: constructing the conscious brain – Antonio Damasio – 2010. Contains an extended discussion of homeostasis and how failures of homeostasis might map onto feelings and emotions. Unfortunately a map which does not usually work in the other direction, in the sense that while we might feel unwell, we cannot say why.

Reference 4: Saving normal – Allen Frances – 2013. Frances includes a short discussion of homeostasis to prepare the ground for his critique of what he describes as the abuse of DSM IV – the fourth version of the diagnostic and statistical manual of mental disorders which underpins that side of health care in the US.

Reference 5: http://psmv2.blogspot.com/2013/06/dsm-5.html. Previous notice of reference 4.

Reference 6: https://psmv4.blogspot.com/2019/03/maxwells-demon-revisited.html. A rather different sort of story about a city wall, involving genes and amino acids.

Reference 7: https://plato.stanford.edu/entries/phenomenal-intentionality/. A version of intentionality which puts consciousness first, which I do not myself, on very limited acquaintance, hold with - but the issues raised in the article are real enough. In any event, only for the philosophically minded.

Reference 8: https://plato.stanford.edu/entries/peirce-semiotics/.

Reference 9: Precise timing of abrupt increase in dust activity in the Middle East coincident with 4.2kya social change - Stacy A. Carolin, Richard T. Walker, Christopher C. Day, Vasile Ersek, R. Alastair Sloan, Michael W. Dee, Morteza Talebian, and Gideon M. Henderson – 2018.

Reference 10: https://www.ncbi.nlm.nih.gov/books/NBK279250/.

Reference 11: The extended organism: the physiology of animal built structures - J. Scott Turner – 2000.

Reference 12: A free energy principle for the brain - Karl Friston, James Kilner, Lee Harrison – 2006.

Reference 13: https://www.flint-paramoudra.com/. A good source of information about flints, large and small. Also West Runton, a place, as it happened, where my scout troop had its summer camps, in its own summing camping ground.

Reference 14: https://en.wikipedia.org/wiki/Flint. An introduction to flints and their place in the world.