Systems
Systems theory
“in the past centuries, science tried to explain phenomena by reducing them to an interplay of elementary units which could be investigated independently of each other. In contemporary modern science, we find in all fields conceptions of what is rather vaguely termed ‘wholeness.’” (Bertalanffy, 1950)
From the perspective of multilayered structures, two concepts of systems will be introduced: the chaosmos of Deleuze and Guattari and the holistic theory of biologist Ludwig von Bertalanffy.
Bertalanffy's proposed theory aims at defining a unity of the sciences. General System Theory was outlined as a holistic theory, wherein the whole is equal to more than the sum of its parts and wherein interaction between heterogeneous elements plays a significant role (Bertallanffy 1950). With reference to Alfred North Whitehead, who defined every large organism as a unity of smaller organisms (Whitehead 1925:18,80,105,112), Bertallanffy's theory introduced the idea of strata as physical “wholes,” such as atoms, molecules, or crystals consisting of the union of other elements defining another strata. In so doing, Bertalanffy's holistic theory has features of hierarchical structure and, with that, a top-down approach (Bertalanffy 1950).
Although approached from a different perspective towards transdisciplinarity, the equivalent of Bertallanffy's “whole” for Deleuze and Guattari could be chaosmos, a complexity arising from chaotic structures through self-organizing processes. Chaosmos, similarly to Bertalanffy's “whole,” is an indivisible system38 of interacting elements. The difference is only in its structure, which is described as a non-hierarchical or, using Deleuze and Guattari's term, a non-arborescent system having no structural elements around which all “things” can be discussed. However, Deleuze and Guattari introduce a definition of strata that is different from Bertallanffy's strata. Within this alternative definition, all the earlier concepts of Deleuze's ontology – difference, connection, heterogeneity, cartography, virtuality, multiplicity, becoming (individuation), etc. – are included. Deleuze and Guattari's strata is divided into three main layers – physico-chemical, organic, and anthropomorphic – wherein, differently from Bertalanffy's strata, all the levels are equal and each of them could be introduced as a substratum of another at any order of magnitude (Deleuze & Guattari 1980:40,69). Although Deleuze and Guattari's strata is “layered,” as opposed to Bertalanffy's onion-like reductionist structure, the division between layers happens in a horizontal “grassroots” structure, where none of the layers are wrapped up with a new layer and where dependency between layers varies depending on perspective.
Therefore, Deleuze and Guattari's holistic approach should not be confused with the reductionistic definitions given by Whitehead or Bertalanffy. Instead of defining a system as a unity of smaller organisms, Deleuze and Guattari's system draws an abstract map of linked processes and elements interacting with each other at a horizontal level.
Networks
Physically, computer networks are connected to each other non-hierarchically and do not necessarily depend on intermediaries, e.g. servers routing digital information to the end computer.
Nevertheless, computers within the computer networks are usually connected over the network switch or wireless router, in such a way forming a decentralized (and at the same time semi-hierarchical) network. In order for the computers to interact with each other, the networks are based on certain rules set across the different protocols. The Internet Protocol Suite (TCP/IP) is one of the most widely used set of protocols forming a network of interconnected computers. One of its abstraction layers, the Internet Layer, facilitates the interconnection of networks, enabling digital data flow among computers. Its Internet Protocol (IP) defines the fundamental address spaces that are then controlled by Domain Name Servers (DNS). In other words, the Internet is defined by TCP/IP and DNS (Braden 1989).
Being the most prevailing, the Internet is also seen as a layer for other networks that can be constructed in order to permit the routing of digital data in predefined lower level networks. For example, the Local Area Network (LAN) is usually used for small-scale networks and is defined by physical space, like a room or house. A Wide Area Network (WAN) or Virtual Private Network (VPN) can overarch a much broader geographical area but are still used in a defined network by businesses or governments. Those networks are usually constructed upon hierarchical routing rules using intermediary servers, and self-organization within those networks could only be possible with a set of predefined exceptions, such as elements a, b, and c, which may become self-organized if all of them are connected to an element d.
In contrast to the network systems mentioned above, other network concepts are based on Peer-to-Peer (P2P) or computer-to-computer connections, which may have a decentralized or distributed character better suited to self-organized systems. However, as was mentioned earlier, they would still be dependent on the Internet layer organized upon a semi-hierarchical DNS system, which is controlled by the Internet Corporation for Assigned Names and Numbers (ICANN). This corporation is responsible for managing all IP addresses, and it therefore makes the Internet centralized from the perspective of IP addresses provided.
Autopoiesis
While speaking of autopoietic systems, Maturana and Varela put an emphasis on the cyclic process of interaction and production. On one hand, Maturana and Varela's machine is similar to Deleuze and Guattari's idea of a machine desiring to produce via repetitive processes. On the other hand, it is similar to Ashby's self-reproducing system, which operates within a network of processes in order to regenerate parts and to therefore stay continuous:
“An autopoietic machine is a machine organized (defined as a unity) as a network of processes of production (transformation and destruction) of components which: (i) through their interactions and transformations continuously regenerate and realize the network of processes (relations) that produced them and (ii) constitute it (the machine) as a concrete unity in space in which they (the components) exist by specifying the topological domain of its realization as such a network” (Maturana & Varela 1972:78).
Maturana and Varela see living systems as a self-referring, circular organization which is, first of all, defined by units of interactions that exist in an ambience (Maturana & Varela 1972:9- 11). This feature is comparable to the self-organization contexts introduced above. On the other hand, Maturana and Varela doubt self-organized systems and argue that organization of the thing would mean changing that thing, as in the organization of a chair, which, disassembled, would lose its identity as a chair (Maturana 1987:71). Therefore, they prefer using the term “autopoiesis” instead of “self-organization.” Among other features of a living system is the possibility of participating in interactions relevant to bigger systems that constitute higher order autopoietic unities. This feature is comparable to Whitehead's idea that a larger organism is defined as a unity of smaller organisms (Whitehead 1925:18, 80, 105, 112). This is also similar to Bertalanffy's holistic approach to the universe, where inter-relationships between elements, including atoms, all together form the whole (Bertalanffy 1950). The possibility of participating in interactions relevant to bigger systems is also clearly defined in Miller's hierarchical living systems, mentioned above (Miller 1965, 1978).
Although Maturana and Varela talk about machines as living organisms and organizations, they never discuss these machines as human-made creatures. Maturana and Varela give a car as an example of a non-autopoietic machine, a human-made machine that in itself is not a unity and whose components are produced via other processes. A crystal would not fit into the definition of a autopoietic machine either, because it is constituted of components that are specified by a lattice organization. The authors call that kind of organization static (Maturana & Varela 1972:79-81). Additionally, non-material elements, such as the coding or transmission of information, are not in the domain of the autopoietic machine (Maturana & Varela 1972:90). Consequently, the robot or software simulating living organisms are never conceived as living organisms. However, autopoiesis is a widely used metaphor for describing autonomous robotic systems within artistic contexts.
References and literature
- Bertalanffy, L. von (1950). “An Outline of General System Theory.” E:CO Issue Vol. 10 No. 2 2008 pp. 103-123. Available at: http://xa.yimg.com/kq/groups/18353846/275082385/name/34099391.pdf (Accessed: 28 April 2015).
- Bertalanffy, L. von (1968). General Systems Theory: Foundations, Development, Applications., General Systems Theory. Reprint, New York: George Braziller, revised edition, 1976.
- Deleuze, G., Guattari, F. (1980). A Thousand Plateaus. Reprint, Minneapolis: University of Minnesota Press, 2000.
- Maturana, H., Varela (1972). Autopoiesis: the Organization of the Living. Dordrecht: D. Reidel Publishing Company, 1980.
- Miller, J. G. (1965). “Living Systems: Basic Concepts,” Behavioral Science, 10(3), pp.193-237.
- Whitehead, A. N. (1925). Science and the Modern World. Reprint, New York: Pelican Mentor book, 1948.