What is the essence of “life” that makes a living system different from a non-living, inanimate, mechanistic system? Chemists approach this question by addressing the difference between “organic” and “inorganic” matter, and the emergence of carbon atoms that allowed for a tremendous variety and dexterity in molecular structures. Biologists address the question by defining the qualities or “symptoms” of life, such as the ability for an organism to reproduce and to move through space. However, these approaches do not address the heart of the matter of what makes a living system and what defines the “life” of that system. In this essay, I define “life” as a dynamic, self-organizing, autopoietic system and map out the manifestations of these autopoieitic processes in living beings at the cellular level, throughout the nervous system, and in the organism’s behavioral and perceptual experiences.

A living system operates through dynamic interactions with the environment rather than static “states” or mechanical transitions. This means that the state of the system is always changing through a dynamic flow governed by both internal self-regulation mechanisms and external constraints. Many cognitive scientists use the analogy of the organism as being like a computer, with measurable ‘input’ sensations and observable ‘output’ behaviors. The primary difference between computational models and dynamic models is the way that these models view change within a system. Computational models view change as the shift from one discrete, static state of a system to another static state, whereas dynamic models view change as something that happens in a system continuously through time. Thus, dynamic models measure the changes of a system over time rather than measuring and describing the explicit structure of the variables in a state space at any one particular moment in time.

The goal of computational models is to measure and describe the system in one moment in time and use this as a template to predict the plausible alignment of component variables in a future state space. Thus, change is measured as the difference between one static state space and the next static state space in a serial causality. Dynamic models are not concerned with the arrangement of variables in state space at any one moment in time (i.e., creating a “snapshot” of the state space) but instead explore the trajectories of variables within the state space over time. In dynamic systems, regularities in the interaction patterns of components within a dynamic system reveal an organizational unity.

What makes the dynamics of interaction within a living cell different from those of molecular transformation in natural, non-living processes? In cellular dynamics, the processes of the system (cellular metabolism, for instance) produce components that comprise the network of the self-organizing unity that produces them. Thus, the ‘product’ of the system and the ‘producer’ of the system are the same (Maturana & Varela, 1987). Further, some components of the system work to form a boundary for the system. This membrane boundary participates in the network of unified interactions within the cell while also simultaneously circumscribing the network as to limit the extent of its interactivity. Thus, an autopoietic system is a system that defines its own boundaries and has molecular networks that produce themselves.

Autopoeitic systems have their own internal mode of circular organization; organization of the system produces the components of the system and constructs a semi-permeable membrane that defines the system. The self-organizing mechanisms within the system constrain and delineate the domain of possible interactions between the system and the environment. Autopoietic systems can be compared to allopoietic systems, which are systems that rely on the interference of an external agent. Allopoietic systems rely on heteronomy (two systems, organisms and environment) whereas autopoietic systems are autonomous (one system, codependent origination, structural coupling, sensorimotor contingencies). Autopoietic systems have what is called “operational closure”, meaning that the internal processing functions of an organism are not singularly defined by the environment and that the products of interactions within the cell do not leave the cell. That is, the structure of the network of molecular components within a cell determines the way that environmental stimulations will be processed by the cell and hence by the organism.

Organisms exist within an environment that constrains the activity of the organism, but the organism also has a certain degree of autonomy. The organism is not just a simple input-output processor receiving ready-made information from the environment; instead the organism creates meaning from what is given in the environment. The nervous system, endocrine system, immune system, etc., influence what it is that the world becomes (perceptually and experientially) for the organism.

Living cells are a fascinating example of an autopoietic organization in that they are both self-producing and self-reproducing. This means that the process of cellular reproduction (mitosis) is triggered by a cellular fracture generated by internal mechanisms and not by external “splicing” from a separate entity or force. Humberto and Maturana (1987) hypothesize initially, in the history of the evolution of the living cell this cellular fracture was provoked by a “bumping” of one cellular unity with an external entity. Somewhere along the way cells evolved to produce this dividing mechanism internally and whatever processes and mechanisms that generated this splicing were preserved throughout the heredity of “daughter” cells.

Cellular reproduction creates two unities that are not identical to the original or to one another but have the same basic organization. In order for this structural similarity to be preserved, the mother system must be distributed and non-compartmentalized at the time of reproduction. Interestingly, during interphase (the non-reproducing timeframe in a cell’s lifespan) the cell’s components are compartmentalized and separated from one another and hence there can be no plane of fracture.

The process of reproduction produces some structural configurations that are preserved and others that vary with the ongoing generations of new cells. The structural configurations that reappear from generation to generation define the heredity characteristics of a class of cells. Some cellular components are resistant to variations and thus are conserved with only slight changes through reproduction. For instance, the synthesis of proteins with DNA remains largely unchanged through the historical lineage of cellular reproduction, and yet the type of proteins synthesized changes with a much greater range of variance with each successive generation.

Ontogeny is the history of structural changes within a unity without the loss of organization in that unity. The environment interacts with the unity from moment to moment to produce ongoing structural reconfigurations. The system also structurally changes from moment to moment based on its own internal dynamics. The cell unity constitutes the range of interactions that can be had with the environment based on its structure in any given instance.