Edited transcript of the lecture, with additional notes (slides are available at doi.org/10.13140/RG.2.2.32704.69122/1)

It looks like we’re heading towards a new paradigm. The totality of what I have to say could make up a whole series of lectures so I can only give a very general description of most of it. The first point is that there are problems with the current paradigm in physics, it's getting a bit stuck you might say; people are almost saying it's a productive operation doing theory, that it doesn't matter too much that you can't fit the theory to experiment.

This slide lists various problems with the current paradigm. One is that while three of the fundamental forces, strong, weak and electromagnetic, can be unified by something called the Standard Model, it's a problem fitting gravity into that picture. People have tried something called supersymmetry but that predicts so-called superparticles — every particle has a corresponding superparticle; they hoped the Large Hadron Collider would show these, but it doesn’t. But things have turned up up that weren’t expected, so-called dark matter and dark energy. People have tried extensions of the Standard Model, but the simplest one turns out not to fit experiment, and going beyond that there are many possibilities and it is not clear which to choose. Another problem lies in interpreting quantum mechanics: because of this some people say forget about interpreting quantum mechanics, just shut up and calculate! To which others responded by saying ‘shut up and contemplate’.

So how are we to go beyond this? The answer seems to be that biologists know things that physicists don’t. And a paper from the 1980s, in collaboration with Michael Conrad and Dipankar Home (https://arxiv.org/abs/1610.09347), pointed to parallels between biology and quantum mechanics. The cause, it was suggested, lies in there being a deeper picture of reality, leading in some circumstances to quantum behaviour and in other cases biological behaviour.

Our diagnosis of the problem is that physics prefers a quantitative approach to nature based on the equations, which is nice because using equations you can test if your theory is right by whether it predicts the right numbers(1). Biology doesn't use equations very much; what it does is to create descriptions. This depends on finding certain patterns, and using appropriate terminology to characterise what you find. That works very well in biology. If nature is to a considerable degree biological, you would get problems with an approach based on equations. So perhaps we should start from biology and see if that can be adapted to physics.

This slide shows the general scheme of the talk. I have recently come across something called coordination dynamics, which I shall explain later in some detail. This is an approach to biology involving units called synergies, which is scientific, as opposed to philosophical, in that you can make observations to study such entities, as well as creating mathematical models that can be fitted to the observations.

Another thing, which probably few people know about unless they are in artificial intelligence, is a computer simulation of human language. It is extremely complicated, just as language is, but is able to understand complicated sentences, and is similar to coordination dynamics, suggesting that coordination dynamics is the right picture of nature.

Another piece of the picture is what I call ‘nature as designer’. Natural language is a process that seems to work better than anything human beings have come up with. Why does this happen? I will give an explanation of how this might happen.

Next, from this approach we get the idea that ideas may be able to influence things, which is quite controversial as it might justify intelligent design. Finally, there is the issue of connecting this approach with quantum mechanics: former particle physicist Karen Barad has written a book discussing their interrelationships.

That’s the general outline, and the first topic is coordination dynamics. The person who has been doing most of the development of this is Scott Kelso, and I can recommend a book chapter of his, which can be found at https://www.researchgate.net/publication/301949127_Coordination_Dynamics. There are things called synergies, which are similar to functions in computer programs, and the point about the organisation in both cases is that you can design a very complicated system bit by bit: design one function which does one thing, and then another that uses that function, and so on, building up complexity in that way. This structure of bits working together is something that lets you build up complex machinery bit by bit. Synergies involve subsystems that work together, and in coordination dynamics research you look for these pieces, which are degrees of freedom of the system, and the idea is that these systems work as well as they do as a result of feedback, just as when we learn to walk we learn to coordinate two aspects of activity such as stepping forward and balance. We learn coordination through feedback from trial and error, eventually acquiring stable behaviour. This is the process by which functions develop, if they can do.

There is a diagram in Kelso's account which compares this different approach with the usual approach in science. I’ve indicated the most important comparisons on this slide. An example is self-organisation, which happens in biology but not so much in physics. Again, structures in physics interact as a result of forces, whereas in synergies it is information exchange that mainly governs interactions, while the things that are interacting being movements rather than things. Thus while crystal is a structure where it is the positions that matter, in cases such as the flocking of birds it is not so much the positions that matter but the movements, these being the things that interact with each other to determine the organisation of the whole. Again, there is the fact that in physics you look for fixed laws, whereas in coordination dynamics the laws are evolving, and are context dependent. So it’s a different kind of beast: you bear in mind that this is what you are looking for and do the science that way. So we have this rather different kind of situation to what you have in physics, and it works differently.

Now I’m going to talk about Winograd’s language simulation. This is a complicated program for interpreting language. How does he go about it? It’s a bit similar to how one goes about synergetics: he starts off asking how language works, and this leads him to use a special kind of grammar, Halliday’s systemic grammar. This doesn’t look for static structures as approaches like Chomsky’s do, it looks at what is happening with language, asks what does the grammar do.

And if you are concerned with meaning, it is no good considering language in isolation, it has to be language in some world, so he simulates a ‘blocks world’ where you can ask questions about blocks, or ask it to do things with blocks in the simulation, and that can test whether the program is understanding what you were inputting to it.

Then, to write a program you hypothesise as to what is going on, you represent it as a flow chart, and then write code corresponding to the flow chart. To illustrate this, this slide shows the overall structure of the program, with the things on the left corresponding to the language part, things like grammar, and the things on the right are concerned with the blocks world, things like moving blocks. And here is a typical flow chart, showing how different actions are taken depending in whether a verb is transitive or intransitive, for example looking for a noun phrase in the case where the verb is transitive. There are very many of these.

And this slide shows the representation of meaning, illustrated by the fairly complicated sentence ‘Harry slept on the porch after he gave Alice the jewels, represented by the set of units (#SLEEP :HARRY :RELl) (#LOCATION :RELl :PORCH) (#GIVE :HARRY :ALICE :JEWELS :RELB) ( #AFTER :RELl :RELB). Here the first unit represents the fact that there is reference to Harry sleeping, while the second represents the fact that the location of that state of affairs is the porch, the two being connected by the fact that they share the term :RELI. The program has to turn this sentence into this kind of structure, looking at pieces of input in succession.

That gives a rough feel as to how the program works, and is nicely explained in Winograd’s thesis entitled ‘Understanding Natural Language’. One point I’d like to make is that the program does not state explicitly how that kind of structure is to be handled; what it is doing is finding pieces it can handle, thereby taking sentences part, and processing them in such a way as to make correct use of the collection at some future time possible. The success of this approach suggests that it provides a good picture of what is happening with real language use, and it is likely that the brain is working similarly, even though its hardware is very different.

Now I’m going on to this idea of Nature as Designer. Human language is amazing: it is something incredibly complex and Winograd’s program is only looking at a small part of it, that part needed for dealing with the simulated block world; human language is something vastly more complicated. How on earth does nature accomplish this trick? The speculative answer is that it evolves a bit at a time. There’s an idea that people in evolutionary biology have, Stuart Kauffman’s ‘adjacent possible’: you are at a certain level now, and there is something fairly similar to it which is a bit better. You may find that thing that is a bit better, and take that step forwards. An example that can be taken from English grammar is the Past Continuous tense, the use of which has been characterised as being ‘to say what we were in the middle of doing at a particular moment in the past’. Such a feature of grammar offers increased clarity and may have become adopted at some time in the past. You can see in principle — and clearly research would be needed to confirm it in detail — the way this amazing thing, human language, could have come into existence naturally, step by step.

So we can see everything fitting together: natural synergetic complexes are complicated structures but we can see in principle how they could have evolved. I’ll just mention the fact that the reason this works is that you have modules, so you can advance by adding a module or lightly modifying it. You are getting better and better matches between what you want to do and what you are able to do. It’s like a game of chance; every now and then you hit the jackpot, the jackpot being a new strategy for something like communication, ending up with more and more powerful mechanisms.

In connection with the last point I’ll just mention one thing that is rather interesting, again connected with human capacities. Terrance Deacon has written a book entitled ‘Man, the Symbolic Species, relating to the fact that there are three kinds of signs: iconic where the sign resembles what it relates to, indexical, used for building structures, both involving the current situation, and a third kind called the symbol, which deals with generalities that are not specifically related to the current situation, such as referring to something that might happen some time in the future. This seems to be a specifically human capability, so at some point this ability to use symbols, detaching from the current situation, gradually developed, leading among other things to abstract thought.

That is incidental to the present context, but here is something that is really important in connection with the picture I am developing, which I call the pairing phenomenon. When talking to you, I use some process in my brain to generate what I am saying, while you use a totally different though related process in your brain that hopefullyregenerates something resembling what I am talking about. That only works if the two processes are closely matched: if I were talking in French and you understood only English that outcome would not happen. Communication only works effectively if the two very different processes fit together, so the process requires coevolution of the two components.

This pairing will turn out to be very important, but let me give a familiar example, the double helix, where you get precise pairing of the two strands because of pairing constraints at the level of the bases from which the helical structures are constituted. Such pairing is important because it gives rise to precise replication.

Here are other examples where coordination involves pairing. When two people are doing things together, such as moving a heavy object, they initially wobble about a bit but learn to adjust their behaviours so this does not happen, something that involves a matching or pairing of the two systems. And a different kind of pairing involving language is the match that there is between the language of a community and the practices of that community, with language developing so as to encompass the practices of the community, both describing these practices and assisting their performance. And again there is a matching of specific ideas and behaviour motivated by the ideas, with ideas both deriving from behaviour and being applied to behaviour(2).

Back in 2006 I was contacted by one Ilexa Yardley, whose thinking sounded intriguing but did not entirely make sense to me. It is only now that I am beginning to understand what seems to be their profound nature. Here is an extract from the beginning of the book: “An entity is always part of a process, a process always part of a system, which is always part of an entity, process and system … this produces systems within systems, or systems of systems”. We can see a connection with what I have been talking about, if we take it that the systems relate to synergies. And a synergy as both a system aspect (the structure that defines its behaviour) and a process element, the behaviour that corresponds, a distinction that connects a spatial aspect (the interconnections) and a temporally aspect (what happens over time), and these are connected by her equivalent to pairing, which she terms oppositional dynamics.

The next slide shows a representation of this idea in pictorial terms. It shows two overlapping circles, portraying the fact that a system may sometimes be regarded either as two things or as one. But there is also a third element, since the pairing we have been talking about is always associated with a third entity looming in the background, namely the activity that depends on the two individual systems working closely together. This triadic aspect in fact originated in the work of Peirce in the 19th century, but Peirce focussed on the mind and symbolic aspect whereas Yardley sees this more as a physical process.

One of her ideas helps to see a mathematical side to this. The slide shows the two entities progressively separating. Yardley characterises the separation as an offset, and from the viewpoint of coordination dynamics we can see this as a control parameter. We can also see it as a process associated with pairing, as the intimate connection phase would be a source of correlation between the two systems. But the details must be for the future to decide.

There is a very interesting phenomenon, originally investigated by Michael Faraday and currently being investigated in more detail using the cymascope of John Stuart Reid. This supports some of these ideas by showing how correlations can develop between temporal behaviour and spatial structures even in the case of simple, non-biological systems. The lecture showed a video where music was played to water (the cymascope makes use of a loudspeaker attached to the container to achieve this), and the resulting patterns are observed by viewing the reflections of a light source off the surface of the water. As the music is played a variety of interesting patterns emerge, typically exhibiting symmetries of various kinds.

Getting back to Yardley’s ‘system of systems’ etc., this is something familiar in contexts such as astronomy, but there we don’t seem to see things like coordination dynamics, with information being important. The difference would appear to depend on the fact that in the cymatics case water is involved, which is a liquid. Nonlinearity is likely also to be relevant. There may also be connections with the work of Prigogine, who was interested in biological systems and in systems far from equilibrium.

Music is also relevant in connection with work carried out in collaboration with a musicologist in my college, Tethys Carpenter, who suggested that themes may have fundamental importance in view of the way the same theme may feature in a range of musical compositions. Our examination of the attributes of themes indicated the existence of a number of suggestive parallels between themes and genes (https://philpapers.org/archive/JOSWCM.pdf)

Another part of the picture is the idea of the biological signal, suggested by Benveniste on the basis of his experiments, since supported by the work of Montagnier, indicative of biological responses to specific signals. The hypothesis that specific temporal signals are as important in biology as is specific chemistry, but the idea has not been widely accepted.

Another proposal, going beyond mainstream science, is the idea that ideas play a significant role in contexts such as the origin of the universe and of life, and biological evolution. In the latter context, Yardley writes ‘There is a symbolic man in mind which is the idea of man which had to be present somewhere hidden (imaginary, an idea) before man could appear’. This is in contradiction to the dogma that the evolution of species is entirely explained by current science, but in this connection it must be borne in mind that the dogmatic position lacks quantitive support.

I’ll also briefly mention the work of Karen Barad. In her book ‘Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning’, she takes up the idea of parallels between the quantum domain and social behaviour, and develops the concepts of agential realism and ‘intra-action’, which are in essence equivalent to the concepts of coordination dynamics, and using these parallels is able to account for a number of mysterious features of the quantum domain. Ultimately we can expect to make precise connections between the quantum domain and coordination dynamics and then realise that quantum theory in its present form is a limited theory, to be superseded by a deeper perspective. We already have considerable understanding of what will be involved, and this should lead to a new era in science.

I’ll finish now with a video (https://youtu.be/-4XpIUDC0ok) showing what might be termed natural cymatics. The light source here is the sun, and the video shows it reflected off ripples on the surface of Lünersee in the Austrian Alps. The reflection is particularly spectacular near the edge of the lake, as can be seen towards the end of the clip.

(1) An additional factor is the treatment of the measurement process, the hypothesis that observation 'collapses the state vector' typically being replaced by the many-worlds view, according to which all possible outcomes exist in parallel universes. See my paper 'Biological Observer-Participation and Wheeler's 'Law without Law', at https://arxiv.org/abs/1108.4860, for a discussion of this issue.
(2) Relevant quotes from Yardley include 'Any idea is connected to a counter-idea (an opposite), or else the idea cannot exist', 'An idea is the hidden reality ... for whatever is showing', and 'From mind, ideas express. From ideas, symbols. From symbols, mind.'