If a system is defined by the collective actions of its parts, one may still ask why the parts should cooperate to create collective effects? And, especially if there are very many parts in the system, how is it possible for each part to keep track of what other parts are doing so each can coordinate their activities with the whole?
Collective action characteristic of the system evidently resides in the system, and not in any given part. So it is the system itself that provides to its parts the guidance they require to maintain their contribution to collective behavior. In other words, the system must create the conditions that allow and ensure that its parts act in a way that defines that system. This is true even for a designed system, i.e., a system constructed to act in a certain way. The system has an overall design, but physically it is not the design, but the system’s collective behavior, that ensures participation of its parts in the designed behavior.
The necessary effect of the collective actions of the system on each of its parts is captured by the “Rule of Provision”, namely, that a system must provide an environment for its parts that enables them to follow the rule of performance and at the same time ensures that they do so. It does this through the application to its parts of forces and constraints and, in the case where the parts are human, through provision of incentives.
Two examples help illustrate the rule of provision. One involves a university, which is a dynamical system that provisions its student parts. The other example concerns a student’s shirt, a dynamical system that provisions its button parts.
As discussed in an earlier post, the student follows the rule of performance with respect to the university by going to class, doing homework, playing on a sports team, paying fees, exchanging ideas with classmates, among many other activities that in their small way contribute to university functionality. The student behaves this way because there is something attractive about being a member of the university community, even when some of the actions required for attendance, such as paying fees or taking examinations, are usually not pleasant. The student wants to be a part of the university because of what it offers him, for example, the chance to become better educated and thus to improve his prospects for future well-being. He does those things that are required even when they themselves are not attractive—like paying fees and taking examinations—because the university requires (forces) him to do so to remain a student in good standing.
For sufficiently egregious violations of the rule of performance, for example not taking required examinations, a student part may be removed from the system, a not uncommon consequence for a part of any system that displays chronic disregard for the rule of performance. These are particular examples for a university of the twin strategies of incentive and force that every system uses to coopt or compel their human components to follow the rule of performance. Whether a clerk in a store, a soldier in the army, a member of a family, in each case, in order to help maintain a human part in “good standing”, the system incentivizes the part to follow the rule of performance (e.g., via a salary, the fraternity of a band of brothers, or bonds of familial affection and duty) and/or tries to compel performance by force or threat (e.g., via loss of employment, military discipline, or fear of social disapproval).
The second example of the rule of provision is for a technological system—the student’s shirt. The shirt is a dynamic system that contains parts such as threads, sleeves, collars, and buttons. We focus first on a given button. The button performs the task of helping hold the shirt together when the student is wearing it. The button may seem like an inert object that doesn’t really “perform” a task. But the button is a dynamic system that exerts forces on the shirt via elastic deformation in response to the pull of the fabric. It helps the shirt to function by aiding the critical task of following the motion of the student’s body–the energy source for the shirt.
In order to ensure that the button actually is able do the work necessary to hold the shirt together, the shirt offers a suitable environment, for example a buttonhole of proper dimensions—not so small that the button cannot fit through it, and not so large that it would slip out in response to normal tugging of the fabric.
The shirt also employs strong threading to bind the button in place against the pull of the fabric, and to ensure that the button does not pop off and become lost to the system. With these measures in place, as per the rule of provision, the shirt makes it likely that the button will follow the rule of performance.
However, if the button should by some chance fail the rule of performance, then, as with the nonperforming student, provision by the system can be expected to cease. If a button made of brittle material should crack and fall off the shirt, it would be unable to hold the shirt together. Its secure position and accustomed task provided by being a part of the shirt now replaced by the uncertain environment of the trashcan. Of course one might argue that the cracked button is now a part of a trashcan system—an argument that raises interesting questions about the role of waste (is waste a “part”?)—but for present purposes the lesson is, if a part doesn’t follow the rule of performance vis-à-vis a given system, then that system cannot be expected to follow the rule of provision vis-à-vis that part.
The general relation between (i) performance by parts and (ii) provision by a system is that parts perform “upward” in support of their host system whereas the system provides “downward” to enable its parts. Thus the student acts to support the university by engaging in approved activities, and the university provides rewards and incentives for the student that act to encourage or compel him to continue to do so. Likewise the button supports the function of the shirt by applying appropriate forces “upward” to help keep the shirt in place (on the student), and the shirt acts downward” on the button to provide a suitably matched local environment in which shirt-forces are able to hold the button in that favorable location.
The dyad of performance and provision thus takes the general form of a feedback loop, with the system helping the part perform in a way that supports the function of the system, and vice versa. This is a basic feature of the technosphere, whose dyadic relation with its human parts is one of the main topics of inquiry in this blog. In a later chapter (tentatively, 6.0 Being Purposeful in the Anthropocene) the performance-provision feedback relation, together with the transitive nature of the rule of performance, is used to develop a general picture of agency applicable to any system. This development will subsequently be essential to our discussion of the possibilities for human “intervention” in the behavior of the technosphere.
This is the end of chapter 4.0 Being a System
Further reading
Discussion of the rule of provision: “Humans and technology in the Anthropocene: Six rules”, Peter Haff (2014), in The Anthropocene Review, volume 1, pages 134-135.
Persistent citation for this post: P. K. Haff, 3.5 The rule of provision, in Being Human in the Anthropocene blog, 2018. https://perma.cc/P8JG-GS24
Next up: New chapter 4.0 Being a Part of a System, in which I look at the definition of a part, the role of hierarchy, and several new regulative rules that directly affect or limit the behavior of parts, including human parts of the technosphere.
When thinking further about this, I just remembered an entirely different approach: This is synergetics, as developed by the German physicist Hermann Haken decades ago. I come up with this because of your use of ‘collective behaviour’, and was wondering how this could be a physical concept. Haken’s theory is purely physical and mathematical, and it was deployed not only on physical phenomena (such as laser) but also on social and economic ones. I do not overview the literature, but my impression is that this theory somehow faded out in scholarly debates, probably because it was mainly a German, and hence ‘peripheral’ development? Anyway, synergetics is a powerful analytical tool to grasp your ‘rule of provision’.
Good point to bring up Haken, who did early work on a general picture of “slaving”, how slow degrees of freedom (dof) can under some conditions capture or direct the actions of faster dof in support of further growth in amplitude of the original slow dof. For some physical systems you can write down equations that capture this process, and apparently Haken did so for lasers (I havet read those papers). This is related to so-called effective field theories in physics. For example the motion of single electron in a multi-electron atom can be approximated by requiring that it be consistent with the average over the motions of all such atoms. The resulting electric potential governs the motion of the (fast) electrons, which in turn create the (slow) atomic potential. These examples can be mathematized, and represent a bona fide part of physics. This works among other reasons because the systems of interest are simple–you know (almost) exactly the properties of parts and the interactions between them. You could try to mathematize the technosphere, but its a tough game–we dont understand the properties of parts very well, nor the interactions among them. This is what Earth System Science tries to do–a worthwhile project–but with many uncertainties. Re collective behavior as physical, I only mean that we proceed something like was done for fluid mechanics, a science developed in the absence of any reference to its molecular components. Rather, we identify what seem like relevant collective phenomena (flows,waves), and try to generate a logical description of those phenomena that is consistent with observation and that doesnt violate anything we do know about physical law.
There is a lot of work by German scholars (perhaps also Japanese) on applying synergetics on social systems. A leader was one of Haken’s colleagues, Wolfgang Weidlich (theoretical physicist), who was also an active member in the German society for evolutionary economics, I learned a lot from him (he passed away in 2015). His works include: Concepts and Models of Quantitative Sociology, Springer Verlag 1983; Physics and Social Science – the approach of Synergetics, Physics Reports, 204, 1991, 1–163; Sociodynamics – a systematic approach to mathematical modeling in the social sciences, Gordon and Breach 2000, Taylor and Francis 2002, Reprint Dover 2006. The first book was co-authored with Günter Haag, who is with the STASA at Stuttgart, an institution for applied science, workong on systems analysis, with numerous contributions. I think one can build on that, but obviously, it has fallen into oblivion.
This is a central piece in the entire intellectual edifice that you are constructing. Against the background of our conversation about semiotics, allow me to add a few thoughts. Although the words pop up, they do not appear to be essential, namely ‘function’ and ‘design’. In the (bio)semiotics view, systemic interdependencies are always FUNCTIONS mediated by SIGNS, so that meaning and function are merged in one PHYSICAL phenomenon. The role of signs is obvious, I think, in the university example. I would like to stress that this goes beyond the working of incentives. There is a large literature in sociology and economics about this question. Economists only tend to see incentives, whereas sociologists have always also emphasized other factors, especially symbolic means of integrating collectives, reaching from Parson’s systemic view to Barry Barnes’ version of symbolic interactionism. The point is that symbolic integration (meaning, for example, adopting the identy of a student as part of one’s personal identity) creates pre-commitments and behavioural tendencies that reduce the need for direct incentives. I think that this is crucial for analyzing the technosphere in relation to humans: Our modern culture is one in which technologies increasingly become part and parcel of our identities (think of the ubiquituous smartphone), and so we actively support the further evolution of this technology. We live in a ‘culture of technology’, and that makes us parts of the technosphere.
The case of the button is more tricky. Here, the point is about DESIGN as a sign: The design of the shirt assigns a function to the button, and that design, in turn, defines the function of the shirt in a larger systemic context. Well, most people would immediately comment, this is human design, so semiotics comes into play via human intentionality. But that is only partly true, because firstly, this does not necessarily mean ‘individual intentionality’ but social and economic forces that evolve beyond human control (this is what most economists think about overwhelmingly complex markets), and secondly, if we extend the perspective to including the general discussion about ‘design’ in evolutionary theory, it is clear that all living systems involve design without intentionality. That means, I think that one should start out from such a universal theory of evolution, and then look at the single cases of interest, in order to ifentify the semiotic mechanisms. Thus, we have a button and its function in the shirt, a function of the shirt in human technologies of protecting the body and signalling to other humans, and these evolve in the larger context of human life, and so on. For analyzing these interdependencies, references to semiotic intermediation of physical causalities is indispensable, while remaining a physical phenomenon of its own. To come back to the button, I think that Petroski’s books about engineering everyday artefacts might present rich empirical material to substantiate your argument. And Petroski often also makes the point that certain artefacts evolve in solving problems created by artefacts, which amounts to an endogenously evolving functionality of the technosphere which guides our own intentionality.
Two points here. Not being an expert on C. S. Peirce’s philosophy nor his theory of signs (so kindly thank you for accepting my recent comments on such topics, as published in the recent post “Interlude on Semiotics”), I don’t want to dispute the “physical” nature of functions + signs, but, for me, such a claim needs a better grounding in concepts accepted and understood by physical scientists before proceeding with further analysis in those terms. So for example, I want to avoid claims like “Our modern culture is one in which technologies increasingly become part and parcel of our identities (think of the ubiquituous smartphone), and so we actively support the further evolution of this technology.” This may be true, but it is only the particular way in which the necessity of human support for the technosphere plays out; a more fundamental reason is that the function of the technosphere demands such support for purely ontological reasons. This topic will come up in a later set of blog posts about the role of agency in systems. One of my guiding principles in this blog is to stay away from complex explanations based on analysis of specific kinds of systems, what I call “constitutive” analyses, as they bring in the necessity for system-specific assumptions of the kind that the present enterprise might eventually want to challenge (such as that technological systems are constructed by humans).
Similar remarks apply to the comments on the button example. Not disputing your comments about design, but the button’s design is an event in the history of that button and ultimately in the development of the shirt-system. But its function in the shirt-system once it is there is, like for any part of any system, to help the shirt maintain its function (metabolism). Invoking the idea of human intentionality at the outset seems premature, and, in my view, is better subsumed at this stage into the analysis as a generic (regulative) and non-remarkable property of a human part, rather than with all the baggage it otherwise brings along of “what humans want”, a point of view that, if adopted, might, and in this case would, contradict an analysis built on a purely regulative picture of the technosphere.
I see now a little more clearly a reason for my resistance to jumping immediately to the use of Pierce’s triad of object-sign-interpretant, b/c there is a temptation to lean a little too heavily on ‘interpretant’ as summarizing what we think we know about how humans work. But if humans as interpretants is only one possibility, so that essentially anything can be an interpretant, as I think must be the case, then I would first like to clarify my thinking on how to put this in my own regulative language,rather than jumping to particulars such as “intentionality”.
Finally, Henry Petroski was my valued colleague for many years in civil engineering here at Duke, and I much appreciate his insight into the interaction of technology and society. He was often criticized (privately) for not working on “real” problems of engineering (meaning mathematical models), but he was, and is, a giant in the field of engineering.
I understand this analytical stance, basics first! Yet, let me emphasize that ‘interpretant’ should be approaches as a ontological category in the fundamentals. In my 2013 I tried to relade semiotics to information theory. I think that one problem in opening to semiotics results to be that this sounds like an arcane philosophical approach that spoils more precise scientific thinking. But if you switch to ‘information theory’, this trouble can be avoided, I think. This is simple: We have the Shannon-Weaver tradition of quantitative information theory, but we lack a physical theory of the semantics of information. This is where we can bring in Peirce’s triadic approach, because the interpretant is all about the semantics.
One scholar who discussed all that is Stanley Salthe, a biologist renowned as one of the intellectual protagonists of modern biosemiotics. I was lucky to co-author a paper with him, published in Biosystems. He even speaks of ‘physiosemiosis’. He contributed to hierarchy theory in biology and other important topics. Perhaps one needs to bring all these fringe approaches together, and then watch the puzzle pieces matching!
Great that you knew Petroski in person! I think that engineering is a somewhat neglected field in the philosophy of science and in the social sciences anyway, probably studying engineering as a systemic activity would give us great insight into how the technosphere works!