Thursday 28 January 2021

Why "Physiology First" has become important to me

Often I find that the ideas which really change my life are not "new" ideas to me, but things which are revealed through reading and interacting with others which in fact I have always felt. When I encountered cybernetics through Oleg Liber in my early 30s, I encountered a way of thinking which I had alway felt - particularly through thinking about music. At the same time, when I encountered cybernetics, I already knew where the problems were. I just had a new name, and a lot of new intellectual friends, with whom I could explore those problems. But after a few years, I knew I couldn't ignore the thing that was missing in cybernetics (I tried to fill it with philosophy - particularly Critical Realism - but that just introduced a whole load of new problems!)

What I always felt was that all this logical stuff - mathematics, technology, programming - alongside all the artistic stuff - music, etc - was physiological. When I was about 18, I wrote a lot of tortuous prose about this (which I've still got) - although not understanding that it was physiological, I called it "emotional" - very much the kind of thing a hormonal teenager would write! The closest I could get to biology and physiology at this stage was Piaget, whose work I had been introduced to by a school friend (amazing how those peer conversations have such a profound impact - I remember him saying "that sounds like Piaget" and I said "who's that?")

Now, over 30 years later, I have a different vocabulary and a different theory for saying exactly the same thing. "We shall not cease from exploration, until at the end of all our exploring..." as Eliot said. This is thanks to the cellular biology of John Torday. Only now, perhaps, do I have a secure footing to say with more precision and evidence what I had been thinking as a teenager. 

The experience of rediscovering teenage thoughts is a process of unlocking a map which resituates lots of discarded elements that have been accumulated over the years. About 10 years ago in Bolton, a rather crazy academic visited us and talked endlessly about Husserl and his "Origins of Geometry". I tried to read it. But it didn't make any sense to me.

Now I read it, and it is a different story altogether. It's as if Husserl was Torday - or he saw almost exactly the same thing. What's he saying? Well, basically, he's saying that geometry (and by extension, logic, maths, etc) is not part of a mind-independent reality (so cyberneticians will be happy), but it does have a genealogy. Geometry has a beginning, and it continues to live in our consciousness. Where is this beginning? For that Torday makes more sense than Husserl: it's origins are tied up with the evolutionary origins of the cells which comprise us, and our consciousness in which triangles, lines and squares live. To think of the logical and apparently eternal geometric forms of triangles or squares is to dig into the deepest primordial origins of our physiology. 

Maybe this sounds too abstract or fanciful. But this is precisely the thought that I had as a teenager about music, computer programming and mathematics. In other words, it is an idea which resonates with my own physiological history. More to the point, Torday's ideas map out a fractal pattern of cellular development from simple single cell organisms, to you and me, and consciousness. If geometry does feel fundamental it is because it is embedded in this fractal structure in the deepest way - which, by what we understand of fractals - it must be (I think). This is basically what Husserl is saying - except that he didn't have the science of cells or fractals with which to say it.  

But if this is right, then the way we think about technology, which ontologises logic, mathematics, structure and reason, is a deep mistake. But, of course, we know it's a mistake, and have known for centuries. But we have not had a physiological science which could sufficiently challenge the rationalist impulse and the "unreasonable effectiveness" of mathematics as Wigner put it. But mathematics and logic are "unreasonably effective" not because they represent eternal truths about the universe, but because their origins lie at the origins of our physiology which is the ultimate measure of any effectiveness we can judge. 

Friday 22 January 2021

Ashby on Reconstructability

The following quotes are taken from the introduction by George Klir to Roger Conant's "Mechanisms and Intelligence: Ashby's writings on Cybernetics", which can be downloaded here:

The balance between the reduction of complexity to a model and nature itself was a key point for Ashby. Datafying things "throws away information", as he puts it, but on the one hand, this is essential for science, whilst on the other hand, good science cannot lose sight of the ways in which a model might be used to reconstruct the complexity from which it is derived. 

In most data analytic work today, there is much reduction. And then it stops - and assumes that the reduction can be used to shape the reality - that is the mode of Facebook, Cambridge Analytica, etc. But reconstruction is essential, otherwise, how are we to know that it is these variables and not those that we should be attending to?

This also means that a "system" is not a thing-in-the-world, but rather an idea. As Ashby puts it, it is a "set of variables":

"At this point we must be clear about how a 'system' is to be defined. Our first impulse is to point at the pendulum and to say 'the system is that thing there.' This method, however, has a fundamental disadvan­tage: every material object contains no less than an infinity of vari­ables and therefore of possible systems. The real pendulum, for in­stance has not only length and position; it has also mass, temperature, electric conductivity, crystalline structure, chemical impurities, some radio-activity, velocity, reflecting power, tensile strength, a surface of moisture, bacterial contamination, an optical absorption, elas­ticity, shape, specific gravity, and so on and on. Any suggestion that we should study 'all' the facts is unrealistic, and actually the attempt is never made. What is necessary is that we should pick out and study the facts that are relevant to some main interest that is already given. ...The system now means, not a thing, but a list of variables."

And then here is the problem of reconstructability: 

"systems models which have recently been developed in many different areas are almost invariably constructed from subsystems. While the subsystems, each associated with a subset of the set of variables of the overall system, are often well validated models of the phenomena involved, the question of the ability to reconstruct the overall system from the given subsystems is almost never raised. It seems that there has been a tendency among many systems modellers to take the reconstructability for granted. It is clear that without an analysis by which the reconstruction ability of systems model is determined, the model is likely to be fundamentally incorrect and might be vastly misleading."

But I think this is the remarkable thing: the whole enterprise is not about "complexity", but "simplicity". Reflecting that the variety of the scientists will always be overwhelmed by the variety of nature, he suggests that the whole point of science is to find effective approaches to simplification:

"...system theory (is) the attempt to develop scientific principles to aid us in our struggles with dynamic systems with highly interacting parts, possibly exceeding 10^100 who faces problems and processes that go vastly beyond this size. What is he to do? At this point, it seems to me, he must make up his mind whether to accept this limit or not. If he does not, let him attack it and attempt to find a way of defeating it. If he does accept it, let him accept it wholeheartedly and con­sistently. My own opinion is that this limit is much less likely to yield than, say, the law of conservation of energy. The energy law is essentially empirical, and may vanish overnight, as the law of conserva­tion of mass did, but the restriction that prevents a man with resources of 10^100 from carrying out a process that genuinely calls for more than this quantity rests on our basic ways of thinking about cause and ef­fect, and is entirely independent of the particular material on which it shows itself. If this view is right, systems theory must become based on methods of simplification, and will be founded, essentially, on the science of simplification. ...The systems theorist of the future, I suggest, must be an expert in how to simplify."

Monday 18 January 2021

Computing Conviviality

Conviviality is a state of being between members of group whose work within that group is felt by all to be meaningful, cohesive and enlivening. Classic descriptions of what it is like to engage in convivial work include the crop-mowing scene in Anna Karenina, where Tolstoy describes the aristocrat Levin's experience of sharing the manual labour with his workers. For Tolstoy, the aristocratic socialist that he was, the portrait of Levin was pretty much a self-portrait. But he knew conviviality was a thing relating work and technology.

Conviviality with modern technology is hard to conceive of. We may all be on Facebook, or are editing the odd shared spreadsheet, but each of us is pursuing personal goals and personal ambition: how can I look good by doing this? will this get me promoted? can I publish about this activity and become famous? This is particularly the tragedy of modern universities - even for academics who preach socialist principles, often the flip side is an ego which seeks to publish their excoriations of the system as a big-hitting journal paper. 

Conviviality entails the loss of ego as far as possible; it entails service to the group over the climb to the top. But most importantly, conviviality entails collective technological work. This is where things get very complicated for conviviality in the 21st century.

Our computer technology allows each of us to transform our environment in remarkable ways. For Illich, whose study of conviviality is one of the most penetrating, the antithesis of conviviality was represented by the mechanical digger: unlike the shovel which required the collective effort of many to dig a hole, the JCB could do the work of 100 people in one go. Computers are not shovels; they are JCBs. 

Heidegger had an interesting way of talking about the fruits of human technological labour: it served, he argued, to reorder nature into what he called the "standing reserve" - the world as revealed to us as "instrumental". Through such reordering, a field could become a coalmine, or a river bank could become a point of crossing with a bridge.  With computer technology, the standing reserve that is manipulated is often data: a disordered collection of data becomes categoried and compartmentalised; a stream of categorised data can coordinate new human action for a common purpose. In this way, technological work can be convivial perhaps - except that it rarely feels like it. 

Heidegger's metaphor can be useful if we are to reconceive conviviality in a the digital age. The problem is that computer tools are generally not very good tools for doing the jobs they are intended to do. This is partly because the design of tools is generally shaped by the work-as-imagined by the software developers, not the work-as-done by the actual users. In trying to engage in the work-as-done, users have to find ways of working around the work-as-imagined in order to solve their problems. Typically this involves many mouse clicks, the frustration that the interface doesn't make the "thing that you want it to do" easy, or that the system simply doesn't work as expected. These experiences apply particularly to educational technology. In my institution, where we've recently introduced Canvas and Teams, it is noticeable how much time is being spent by academics dealing with an interface which, whilst it is much better than Blackboard, still demonstrates the difference between work-as-imagined and work-as-done. 

This requires both what David Graeber calls "imaginative labour", and an awful lot of tedious clicking - so the high-level intellectual work of academics is replaced with low-level interface work.

A possible homologue of Heidegger's river bank or empty field is the software designer's interface constraints. Although such things are the product of technology, and in some sense, already "standing-reserve", it also presents humans with natural challenges in the form of stress, boredom, etc. The problem is that everybody is having to build their own bridge through the interface - the scope for collectively working with each other to build a bridge that everyone can use is constrained by barriers in the software, or institutional barriers which stop people getting access to aspects of the software which might be changed for the better for all. The second problem is that even if it was possible to build a bridge across the software's "river bank" for all, there would still be a gap between "work-as-imagined" and "work-as-done" for new users of the system.

Conviviality requires co-design. Enid Mumford's work on co-designing systems with nurses in the 1980s and 90s has become very important to me recently: she certainly deserves much greater recognition (and an alumnus of Liverpool University). For Mumford in the 80s, coding was sufficiently common to all forms of interface so that it enabled her to find ways of engaging nurses at quite a technical level in the functionality of the system, so that they could shape it to reflect work-as-done.

Today, the equivalent level of commonality is the Application Programming Interface. Good computer systems today are built of services, and good systems with good APIs make all the underlying services available to anyone who wants to access the data providing they have sufficient privileges. 

I've spent some time recently working with professional service staff and academics in working with APIs to achieve effective and efficient solutions to the problems of the interfaces. While the solutions we have developed together can be redistributed to other staff, it is much better if the means for creating those solutions is shared so that everybody understands how they can build bridges together. This is not to say that everybody should learn to program, but it is to say that engaging with a deeper technical understanding of systems is to open doors to deeper institutional discussions about the problems that academics and support staff face.

Conviviality occurs when we recognise the problems that each of us face, and relate them to our own problems, whilst recognising that working together can help us all. Coding creates a universal framework for articulating and sharing deep problems - it is not something that is done TO people: that merely creates a new set of constraints of "work as imagined". The magic only works if it is done WITH people. 

Friday 8 January 2021

Recursion and Function

I attended an interesting talk before Christmas on music and structure at Liverpool's PhD music discussion group (which is great fun!).  The talk was really about the "function" of musical chords, and how this function related to the structure of the music. What was being argued was that there was a kind of recursive relation between the deep structure and the function on the surface. This isn't a new idea, of course - Heinrich Schenker was the first to talk about this over 100 years ago - but it was interesting because it muddied the distinction between structure and function: function was something that is "done" by some component (a chord) which in aggregate articulates a structure. Or does it? Is it structure which determines the function of its surface components? The problem is that when "structure" is discussed, it is always comprised of some fundamental unit (in this case, a "function"), but one could not talk objectively about the structure if one were to consider a different function - a different way of carving up the structure. 

Our understanding of "function" is rather confused more generally - not just in music. It does appear to be the case that music has some kind of recursive structure - units of homologous processes repeat at different orders of scale. It is a fractal in this sense. But fractals in nature are not made from simply repeating some kind of pattern; they emerge through interference of multiple variables at different orders of scale. It is the difference between an L-system fractal (see L-system - Wikipedia), which can produce something that looks like a tree, and a real tree. Photographic holograms provide the real clue: these are images of 3D objects encoded into a 2D pattern by virtue of the interference between the beams of laser light which are used to encode the image. The fractal pattern that emerges is actually an encoding of space and time, since space and time are the fundamental values in the frequency of light.

We might look at a hologram produced in this way and say that it is comprised of a number of features which "repeat". We might even argue that there are particular functions which repeat. And yet these are some surface features which resut from a deeper process of interference which we cannot fully comprehend. Ascribing "function" to some element is our way of dealing with this uncertainty. But it can lead us astray.

What is the process of interference that leads to the fractal in which we identify "functional units"?

Ultimately, I cannot see how it is other than physical and physiological. In other words, it involves (quantum) mechanics and cells. Music is clearly within this realm: vibrations in the air and physiological responses. What is remarkable, given the variety of physiological possibilities, is that there can be any agreement about function at all. This suggests to me that the social function of coordination and agreement serves in some way to establish coherence and pattern in the fractal between social groups - and that indeed, this may be fundamental to inter-human (or indeed, inter-organism, or inter-cell) coordination.  

We are given to believing that there are fundamental surface units of functionality which produce coordination. Yet, we are continually reminded that this is not the case. From this we conclude:

  1. There is nothing fundamental in nature that unites us, or that can be harnessed to orchestrate our minds better: it's just "culture";
  2. There remains to be found some structural method which with sufficient force, vigour or control can mobilise collective action;
  3. It is the job of education to programme the young to reinforce the functions of existing ways of thinking, and to teach them that this is as good as it gets.

Point 1 will lead to destruction of our environment (as it is doing!). Point 2 leads to totalitarianism. Point 3 leads to the enslavement of education to those in power rather than an authentic inquiry into nature. 

Sunday 3 January 2021

Covid-19, Science and Education

One of the best things I watched on terristrial TV this new year was the Parliament channel's recording of a special meeting to discuss the scientific evidence on the new Covid-19 variant. It's not often we see scientists in almost a "native" mode, expressing the uncertainty around knowledge about a virus, and explaining this uncertainty as best they can to politicians whose job it is to formulate policy. You can watch it here: - Science and Technology Committee

The scientific rigour is impressive (and some of the questioning by the politicians is also very acute - did we inadvertantly cause this mutation? will the vaccine work against it? - no definitive answers to any of this). But it also struck me that this is a group of people who are totally focused on their scientific niche, when the virus presents a systemic problem and there is no coherent way to connect the rigour of virology to the broader social questions - particular those about children, viral transmission and schooling.

About schooling, Dawn Butler asked a pertinent question about whether schools should be closed. This was one of the more disappointing moments from the scientists. They fudged around the issue, saying things like "obviously not going to school damages children... so it's a balance".

Obviously? What do they mean by that exactly? Where is the comparable rigour behind that assumption that they show in their dealings with the virus? Calum Semple from the Sage committee was on the radio the other day even saying that the approach to education was a "whole system approach". Really? I don't know what his understanding of "whole system" is, but it doesn't look like something that systems theorists would understand as "whole system". "Whole system approach" has become a kind of sop to fend off awkward questions: interesting how the establishment appropriates systems thinking, and then does the opposite. 

We can't really have a whole system approach until we have some grasp of the relationship between natural systems and cultural systems. Our problem with education is that we conceive of it being entirely cultural, and that our culturally-defined parameters and metrics which determine the success of education are divorced from natural biological and psychological factors. Covid is screaming at us that these things are not seperable, and that we need better scientific theory in education and learning. 

In the time of Piaget, there was little doubt that learning was connected to nature. The challenge was to find the best way to connect the natural processes with the cultural, institutional organisation of society's education system. Now hardly anyone talks about Piaget in any depth. Constructivism (if anything pedagogical is discussed at all) takes its place, and rather like "whole systems", the word has become divorced from its scientific roots, and splashed around as a badge of honour to mask pedagogical and institutional approaches to learning which are the very opposite of what Piaget was talking about. Constructive alignment anyone?

Today the government let it be known that "online learning was a last resort" for education. A last resort from what exactly? From the cultural system that we call education which wants to maintain its convenient divorce from nature and science. As the world moves into a new era where a combination of working online alongside mass unemployment is going to bring intense stresses on existing social structures, are we really saying that face-to-face mass education which is little changed since 1900 is the way forwards? How does that prepare the kids for the world as its becoming? The answer is, it doesn't: it prepares the kids for their parents' and teachers' world that we are leaving behind. 

This is not to say that the online education we have today is good. It's obviously mostly terrible. But it's terrible because the computer is being used to reproduce the function of the old system. But a proper "whole system approach" would change the system. 

I don't think it's that difficult to imagine how things could be much better.  The effect of technology upholding ancient institutional practices and structures has been an increasing transactionalisation, increasingly mundane technological work for learners and teachers, and a fundamental loss of meaning in the activities of learners and teachers, from primary school onwards. Education risks meaninglessness and irrelevance. 

How do we put the meaning back into our teaching and learning activities? They must be reconceived around the manifest and fundamental uncertainties that face us all. Rather than drilling knowledge that everyone already knows into young minds, education could be a process of renewal where the questions that nobody knows the answer to are addressed by young and old together. And yes, you can teach maths like that! That was pretty much what was going on the Science and Technology Committee.