General Systems Theory Research Project
General, Systems, Theory, Research, Project
Peter caws1,2* 1 department of philosophy, the george washington university, washington, dc, usa 2 american association for the advancement of science, washington, dc, usa
This paper has three parts. First, i discuss what i take as the original stimulus and the pur- pose of general systems theory (gst) to be, why i think it is important, and how i came to be involved in it.
I reflect on von bertalanffy’s general system (sic) theory and the early debates on the topic, stressing the essential concept of isomorphism, with its rewards in following up parallel developments in different domains, and its risks and temptations in the projection of grand and all-inclusive systems.
Second, i discuss the direction my own work took after my term as president of the society for general systems research (1966–1967), and how it diverged from the early program, in particular in its emphasis on the difference between system and structure and on the essential role of individual subjectivity in the latter.
I stress the importance of the concept of ‘relation’ as underlying that of ‘system’, and in particular the difference between relations as embodied in physical systems and relations as components of intentional structures that may or may not corre- spond to physical systems. In the third and final part, i discuss the place of gst in the philosophy of science, especially in connection with the unity of science movement, and its potential for the organization of this domain.
I ask what light the concept of system can throw on our knowledge of the universe and its worlds (a distinction explained in the paper), and what the risks are of assuming tight isomorphisms between mathematical structures and physical systems, for example, in cosmology and quantum mechanics. Copyright © 2015 john wiley & sons, ltd.
Keywords general system theory (gst); philosophy of science; unity of science; isomorphism; history of the systems movement
It is an honor to have been invited to appear before you today; the more so because this lecture is named for ludwig von bertalanffy. It is also an unexpected honor – i did not know until recently that this meeting was even happening,
* correspondence to: peter caws, webb 101, mount vernon campus, the george washington university, washington, dc 20052, usa. E-mail: firstname.lastname@example.org † the ludwig von bertalanffy memorial lecture delivered at the annual conference of the international society for the systems sciences (isss), washington, dc, usa, on july 2014 was presented as twinned presenta- tions by peter caws and david rousseau, under the joint title ‘general sys- tems theory: past, present and potential’. This paper represents peter caws’ contribution reflecting on the past and potential of general system theory.
Copyright © 2015 john wiley & sons, ltd.
Systems research and behavioral science syst. Res. 32, 514–521 (2015) published online in wiley online library (wileyonlinelibrary.com) doi: 10.1002/sres.2353
And although my name has regularly appeared on the list of the board of distinguished advisors of this society i have, as far as i can recall, never given it any advice, distinguished or otherwise. It is a coincidence that the meeting should be taking place in washington, not only at my own university but also in my own neighborhood, within 10-min walk from my apartment – a case of the meeting coming to me rather than me having to go to the meeting.
It is also a coincidence that tom mandel (to whom i owe my thanks – i am sorry he cannot be with us) should have had the idea of bringing back, on this particular occasion, some of the early participants in the international society for the systems sciences (isss) or its precursor. If it had not been for all these things, i probably would not have been here at all. But i am very glad i am. Some of what i have to say will inevitably be
Autobiographical. My claim to attention is pre- sumably that i was once president of the society for general systems research, the precursor of your own isss. Bringing back a former president after almost 50 years has its risks. For one thing, unless he has been following things closely, which i have not, he was bound to be out of touch.
For another, anyone who has had the responsibility of addressing an annual meeting as its president, and who has taken that responsibility seriously, probably is, or at any rate was, pretty opinionated. I thought i had a necessary task back in 1966,
And i tried to carry it out; but it was not popular. At that time, i took my job to be deflationary. People were getting carried away by the idea of an overarching, all-embracing system, of which all the sciences were to be partial instantiations. I remember in particular a paper, which i had especially in mind in writing my address, that argued from a local distribution of small-mouth bass to a layered hierarchy of systems from the microscopic to the cosmic.
I thought this was extravagant, if not megalomaniacal, and would give systems theory a bad name, so i was at pains to point out its limitations. As i put it in the intro- duction to the reprinting of the address, in my book yorick’s world,
‘among some of my colleagues in the society i had detected a rampant tendency to suppose,
Somewhat after the manner of hegel, that ontology could be read off from logic – that if one could build a layered edifice of theoretical systems the world must contain somewhere their real counterparts. The argument of the address served as a gentle rebuke to these pansystematists’(caws,1993, 16).
Some of my listeners probably thought i was a killjoy – although i admit that i took some satis- faction in the fact that, after i had made my point in the presidential address, anatol rapoport thanked me for making it and said he wished he had done it himself.
All this was, of course, partly von bertalanffy’s fault, because he was something of an evangelist for what he originally called general system theory, in the singular, that is, the theory of a system that would embrace the diversity of the sciences and subsume the partic- ular systems that he was confident would be found repeating themselves at various levels of complexity.
To do him justice, he himself did not yield to the lofty pretensions i was gunning for. In his ‘response’ to the papers offered to him on his 70th birthday, compiled by ervin laszlo as the relevance of general systems theory, he says: ‘i did not find ultimate truth or “noth- ing-but” solutions, and never aspired toward …. A secular “extra ecclesiam nulla salus.” Rather, whatever i may have been able to con- tribute, leaves plenty for others to do better’ (laszlo, 1972, p.483),
Von bertalanffy started at a middle level, that of biological systems, where he introduced an essential and most fruitful distinction between closed and open systems, the latter providing as the former did not for metabolic exchanges across boundaries. Boundaries, as the theme of this conference suggests, are crucial.
However, it is worth pausing here. Open systems can be open in all sorts of ways – and they can be closed by the selective admission of adjacent elements. So the extent of the system becomes a matter of choice – what are its elements, in what relations, across what boundaries? This is consistent with the definition of ‘system’ it- self, deriving as it does all the way from its greek origin as nothing more specific than ‘a whole compounded of several parts or mem- bers’. But that is entirely indeterminate – what is the whole in question?
There seems to be no obvious answer to this a priori, as is evident from the astonishing variety of submissions to the present conference no prescribed field of study, no limitations on scope.
The scope is sometimes virtually all-embracing, as in the case of large-scale systems engineering, whose practitioners have what may seem the grandiose task of anticipating all possible boundary crossings at all degrees of scale or detail and in all interacting domains, whether natural or social, financial or logistical, physical or biological, ecological or meteorological, etc., not missing any contingencies but not overestimating any either, with huge conse- quences for budgets and human welfare hanging on every decision.
But i am getting ahead of myself here. I pro- mised some reflection on what the field was like when i got into it. I arrived in the united states, with a degree in physics under my belt but not otherwise committed, at an exciting time, catching the wave of what gregory bateson characterized as
‘the growing together of a number of ideas which had developed in different places dur- ing the second world war. We may call the aggregate of these’, he continued, ‘cybernet- ics, or communication theory, or information theory, or systems theory. …. All these sepa- rate developments in different intellectual centers dealt with communicational prob- lems, especially with the problem of what sort of thing is an organized system’ (bateson, 1972, p.483).
This is worth dwelling on too, given how cybernetics, and information, and communica- tion, and our own systems, have been rivals for dominance ever since. As david rousseau remarked to me yesterday, everyone wants to be the mother ship.
I was able to switch fields to philosophy, thanks to the generosity of yale and my mentor there, henry margenau, who used to work closely with c. West churchman, at that time, one of the edi- tors of the journal philosophy of science, in which
I published some of my early papers. In my dis- sertation work in 1956, i realized the importance of the concept of isomorphism as it applied to conceptual schemes and their mirroring (pace rorty) of physical structures.
I did not then know von bertalanffy’s work, or that he had spoken about ‘the structural isomorphy of laws in the different fields of science and reality’ (von bertalanffy, 1951), although i may have been indirectly influenced by it, because one of my professors was carl g. (‘peter’) hempel, who had commented on the paper in which von bertalanffy used the expression and may possibly have referred to it in class.
By an accident of academic fate, my first teach- ing job was not in philosophy but in ‘general science’, which meant that i had to read up on chemistry and genetics and geology, to add to the meteorology to which i had been introduced in school by an eager young physics teacher fresh out of the air force.
This constituted a pretty good basis for doing comparative work. I special- ized in the philosophy of science – and i have always believed that scholars who do that must have a first-hand acquaintance with as broad a range of the natural and social sciences as possible.
I gravitated naturally enough to the american association for the advancement of sciences (aaas) and gave my first paper to its annual meeting during that first year of teaching. All sorts of interesting developments were coming to light, particularly in studies on the brain and nervous system, and i remember being intro- duced to the work of mcculloch and pitts, and reading ross ashby’s (1960) design for a brain and of course his introduction to cybernetics (1956).
I do not remember how i first came across it, but one of the formative influences at the time was the work of an eccentric society called the artorga research group (for artificial organism), whose president was oliver d. Wells and whose committee consisted of gordon pask, heinz von foerster, ross ashby and stafford beer.
add in kenneth boulding, anatol rapoport, gregory bateson and margaret mead, and you get some idea of the firepower of these early pioneers. I did not know all of them personally, but some- how between artorga, the young society for general systems research, the aaas (of which i became vice-president for section l in 1967) and my first book on the philosophy of science (caws, 1965), i found myself delivering the address to which i have referred in 1966.
I want to pay special tribute to oliver wells, a neglected figure in this history.
He self-published a series of periodical pamphlets for artorga and one small book, how could you be so na- ïve? (wells, 1970) from his home in southern en- gland, and was responsible for bringing a lot of original work to the attention of his mailing list.
I am not sure whether to mention this, but i re- member being startled at the time, and maybe you will be too: one of the articles he republished in his book bore the title ‘science fiction – sex for ever: a new cybernetic project called interfuck’ (wells, 1970, 140-143), which proposed ‘the development of a group of sys- tems for the two-way transmission of sensory- sexual information’, based on the apparatus developed by masters and johnson for measur- ing the human sexual response.
It contained the laconic remark ‘the project is difficult to name in english,’ although oliver wells seems to have had no trouble with this. It seems to me a case of boundary crossing worth drawing to your attention. I also owe to wells a pithy formula, ‘the brain computes the world’, which summed up admirably a causal theory of per- ception that still holds water today. Artorga engaged in a collective effort to build
A self-reproducing machine, based on some ge- netic work by lionel penrose. In the penrose archive at university college london, i recently came across an interview with wells, in french, in the journal science et vie, in which the interviewer, gerald messadié, expressed his ad- miration for the systems work going on in the english-speaking world and concluded rather enviously:
‘there is today no creative mind which does not direct all its wishes to a profound re- newal of all the ideas with which we live. In- numerable original works are sleeping in the files of scientists and technologists. It is per- haps artorga that is preparing the synthesis and the reorganization that are necessary, a Veritable work of the encyclopedists [quite a compliment for a frenchman]. It only re- mains for france to join in’ (messadié, 1961).
In view of the plethora of systems literature to which gerald midgley referred the other day, it would seem that this work is as urgent as ever.
In my presidential address, which i entitled ‘science and system: on the unity and diversity of scientific theory’, i commented on the change from ‘theory’ to ‘research’ in the name of the soci- ety, which seemed to me to mark a becoming modesty.
a theory, as i pointed out, is really a way of looking at things – theoros in greek meant an official observer, who accompanied people to the consultation of oracles or to competition in the regional games, to ensure that things were done in proper order and reported correctly.
So a theory is not just any old way of looking, but one which carries some gravitas and will stand against challenge. A general theory would be a way of looking at many things, perhaps at all things, in a similar way.
The further transition from ‘theory’ to ‘science’ makes a stronger claim. What rendered all those conjectured isomorphisms suspect was that theo- retical possibilities do not always map onto phys- ical actualities. Natural systems come into being as they do, with the contingency of evolutionary accidents and pressures – many niches remain unfilled, so we cannot assume totality or even generality. It was a good move on the part of the international society for the systems sciences to drop the ‘general’ of the society for general systems research.
Already in my dissertation, i was stressing the need for the theoretician to accompany and ani- mate the theory, which could i suppose be taken as a version of the view that the observer has to be considered along with what is observed. That view, however, has to be handled with care.
That there might be a theory of theories, a science of science, seemed obvious to me, but that did not mean that there was anything wrong or naive in trying for a science free of observer bias, and in- deed that is a condition of success in most of the physical, as opposed to the social or human, sciences. The essential conditions for a science, it still seems to me, are three:
Syst. Res. Research paper
Copyright © 2015 john wiley & sons, ltd. Syst. Res. 32, 514–521 (2015) doi: 10.1002/sres.2353
General systems theory: its past and potential 517
(1) object constancy across knowing subjects, including agreed nomenclature (this means being sure we are talking about the same thing);
(2) replicable observations and predictions, sub- ject to common reporting standards (this means looking at the same elements of the world in the same way); and
(3) theoretical consistency, including to the extent possible, simplicity and plausibility (this means arguing openly and convincingly in the face of doubt or criticism).
These last conditions are sometimes definitive. One notable case for the test of simplicity is the switch from the ptolemaic to the copernican account of the solar system. As I pointed out in an earlier paper (caws, 1963), the advent of com- puters would have made predictions according to the ptolemaic view quite feasible, but the simpler copernican picture was easier to visua- lize and its predictions quicker to compute.
A contemporary challenge to the test of plausibility is presented by big bang theory and particularly by inflationary cosmology, which make extra- ordinary claims on belief in matters of time and causality.