Chapter 4: Functions are not Final

Chapter 4: Functions are not Final

Functions, Norms, Final Cause and Formal Cause  

Teleology is problematic if considered under substance metaphysics, however teleology can be reconceptualized under process metaphysics and enaction. The problem under substance ontology is expressed by David Hume’s aphorism that “an is does not imply an ought.” A thing’s existence under substance ontology does not imply its normativity, functionality nor teleology. The challenge remains for enaction and process ontology to show how normativity and functionality can come into being and under what conditions. 

An is does not imply an ought. The factual existence of a substantial thing does not imply its normativity nor functionality. Under substance ontology, things en soi do not possess teleological function by merit of their factual properties. However, if reconsidered under process ontology, an “is” is not a stable thing, but an abstraction from a stabilizing process. Further, a living processual form (an organism-environment system) is not a substantial being, but a processual becoming. A life form is not a Humean is (n) but an Kantian whole, an autonomous body of autopoiesis (v). A life form is a self-organizing process undergoing precarious operational closure and individuation. Thus, Hume’s aphorism does not necessarily apply to processual life forms.

An organism is not necessarily describable as a thing; an organism is not a static Humean is, but a Kantian whole. That is, organisms are not thing-like substances, but holistic processes. An organism’s structure-of-becoming operationalizes its being, and its precarious operational closure enables conditions for instrumental ought (per Stuart Kauffman), i.e. an organizational normativity (instrumental derives from the Greek root organon and refers here to the precarious organizational closure of an organism). 

Stuart Kauffman coins the term instrumental ought, which is rephrased as instrumental or organizational normativity in this chapter:

David Hume said one cannot deduce an ought from an is, the famous naturalistic fallacy. From the fact that mothers love their children, you cannot deduce that they ought to do so. But Hume is deeply wrong. Hume, in the tradition of British Empiricism, thought of a passive observing mind/brain in a vat and wondered how that observing mind could have reliable knowledge of the world. He rightly noted that from what is observed to be the case, one cannot deduce what ought to be the case. We live with the naturalistic fallacy. But Hume forgot that organisms act, even one as simple as an amoeba. Once doing enters the universe, doing it well or doing it poorly follows[....] In short, doing brings with it instrumental ought. We are agents, and doing something well or poorly matters! So we ought to do it well. This is instrumental ought, not moral ought. Instrumental ought, how to do something, is with us as soon as agency arrives. Thus, it is ancient. (2019, p. 94).

Instrumental norms are generated with organism processes of autopoiesis, autonomy and their precarious organizational closure. An organism is a “Kantian whole” according to Kauffman in that the parts exist for and by means of the whole. The organism is a circularly causal closure of constituent processes that organize a boundary of individuation (a self organization), which reciprocally form the constituent processes. 

Kauffman describes life as a Kantian whole in the following passage, and relates this to the organization of instrumental normativity:

Long before Darwin, Immanuel Kant understood this: ‘An organized being then, has the property that the parts exist for and by means of the whole.’ Call this a ‘Kantian whole.’ The heart exists for and by means of the whole organism of which it is a functioning part. [...] In short, functions are subsets of the causal consequences of parts of organisms. But how do we know which ones? [...] More generally, to be a function something must abet the survival of a Kantian whole. (2019, pp. 8-14, emphasis added).

Organisms undergo precarious operational closure, the whole closure actively constrains and differentiates its constituent processes, and these processes are interdependently integrated as the whole closure. In circular causality: the whole diversifies and develops the parts comprising and sustaining the whole. In the history of an organism, independent parts did not come together to aggregate a whole. The whole integrates and differentiates, and an observer abstracts interdependent parts and functions. How can an observer determine what counts as a normative function? The answer is due to the precariousness of an organism’s operational closure, i.e. functionality contributes to the ongoing metastable persistence of an organism-environment system against its equilibrium and collapse. 

For an organism, aggregate parts do not form the whole. Instead, a holistic concrescence integrates and differentiates, but parts are abstractions. Constituent processes are not independent parts aggregating together, but interdependent differentiations coevolving with each other and their environment. In this usage, identifying the cause of instrumental normativity is not a matter of reductionism, but a matter of irreducible operationality within a holistic concrescence. Function is instrumentally situated and actively attuning within the holistic organism-environment system’s precarious operational closure, but is neither pregiven nor final. Abstracted function is contingent to the precarious operationality of a dynamic processual form undergoing self-organization and individuation. Functions are enacted, not final. 

The word organism was previously defined from the Greek organon meaning “instrument” and ism meaning “doctrine.” Hence organ-ism can be operationally defined as an instrumental (organizarional) doctrine, i.e. an obligate precarious organizational closure (in this definition, doctrine implies obligate). Equivalently, an organism can be operationally defined as “a living complex adaptive system of organs that influence each other in such a way that they function in some way as a stable whole” (McGill university website). Thus, Kauffman’s term of instrumental ought is suitable as literally an organism’s organizationally mediated normativity. The instrumental normativity within an autonomy is an organic normativity. 

Instrumental normativity is operationally definable with respect to the concrete, interdependent interactions within a Kantian whole, but is irreducible to reified final-causality. An organism’s instrumental normativity is operationalized from the behaviors of “organs” interdependent with other “organs,” developmentally concrescent within (for-and-by-means-of) a Kantian whole. The parts exist for and by means of the whole, the organs exist for and by means of the organism. Organs comprise the organism, but not prior to the whole organism developmentally differentiating its organs; the process is a circular dynamic co-emergence.  

To Victoria N. Anderson, a researcher of teleology and complex systems, the word usage of instrumental is dubious: “ ‘Organ’ comes from the Greek organon, meaning ‘tool’ or ‘instrument,’ a somewhat unfortunate etymology for an organ is different: it helps create and is created by the individual in which it exists. Tools don’t do that.” (2019). Importantly, in this project’s word usage of instrumental normativity, this term is qualified as instrumental within an autonomy, not instrumental in the sense of a heteronomous tool. Organs are not instrumental in the sense of being heteronomous tools, but by being instrumental within the autonomy of an organism. Organs are considered instrumental for-and-by-means of the whole organizational closure, as differentiations that dynamically co-emerge with the integrated whole. Anderson’s concern regarding etymology can be reconciled if this consideration is made. The word usage of instrumental is only coherent if considered as autonomously instrumental. 

Parts as Abstractions and Avoiding Reification 

Abstraction of functions, parts, and norms can be useful for heuristic purposes, but should not be reified. Within an organ-ism, instrumental “parts” (organs) exist for and by means of their interdependent integration as a whole, even as the developing whole differentiates its “parts” in circular causality. For organisms, abstracted parts do not exist in themselves. Parts do not have functions definable outside the autonomy of an organism (heteronomy), and they do not independently come together to form a whole (heteropoiesis). As an analogy, a reductionist fallacy would be to shatter a vase, then conclude that vases are made by assembling types of shards. Applied to life, eggs cannot be put back together once broken; the precarious life cycle irreversibly collapses if organizational continuity breaks. Kantian wholes are not ontically reducible into aggregate parts, hence parts are abstractions. 

When considered under process ontology and enaction, function and normativity are contingent to the formal cause of life; a life form is a processual form. This is to conceptualize life as a meta-stable pattern of operational closures imposed on material-energetic flows. 

When considered under substance ontology and essentialism, function and norms imply the final causality of teleology as a property of life, life’s telos. Hence, the word usage of teleology under processual-enaction can be a useful abstraction, but is potentially misleading if teleology is misunderstood under implicit biases stemming from substance ontology. Victoria N. Anderson gives an example of how function, norms and purpose are abstractions from life as a processual form of autonomy: 

Human purpose is a specific type of a more general purpose in nature. Both can be defined abstractly and generally as forms of self-organized adaptation[....] [Y]our own purposes—which arise from your own self-organizing biological and semiotic processes—make you a purposeful being. (2019)

Teleology can be qualified under processual-enaction by reconceptualizing norms and functions as abstractions from the formal organizational dynamics of a biological system. If teleology is insisted as shorthand for function and purpose, then telos can be interpreted as actively generated instrumental (organizational) normativity contingent to the processual form and autonomy of life as a precarious self-organization of operational closure. 

Beyond its use as an abstraction, the concept of teleology can become a reification depending on underlying metaphysical assumptions. In other words, formal cause begets the appearance of final cause, but to claim an ontically real telos is a reification. To take a teleological stance is to use telos as shorthand as if the organism has final purposes, functions and norms. The shorthand is cashed-out by describing the active generation of normativity from the formal causes of an organism’s precarious operational closure. Even so, telos may be a misnomer under processual-enaction, as this cause is not final, but is actively generated by- and contingent upon life’s processual form.

Recap

To organize a life form is not to have a static form. Life actively individuates under process ontology, whereas life possesses an unchanging property of identity under substance ontology. Likewise, to self-organize instrumental (organizational) normativity is not to possess a static, pregiven telos. To reconceive functionality and normativity under enaction and process ontology is to describe the instrumental contingencies for the formal cause of a self-organizing precarious operational closure. However, to conceive functionality and normativity under substance ontology and essentialism is to ascribe final causes as reified properties of a living thing. 

Genesis of normativity is contingent on different time scales of precarious autopoiesis and autonomy, including but not limited to evolutionary timescales. Function and norms are not merely evolved (as in teleosemantic explanations), but have to be lived and enacted in situ. An organism is penetrated by multiple time scales of operational closures, e.g. scales of autopoiesis, learned experience, development, evolution and niche construction. Examples of closures are forms of life cycles, heredity, reinforcing feedback loops for experientially learned behaviors and ecological feedback loops. Thus, questions regarding functionality can be qualified by clarifying the time scales are under consideration, while still recognizing the interactions between slower and faster time scales of closure. The process of natural selection alone is not sufficient to escape the Humean naturalistic fallacy to account for functionality. An is does not imply an ought, even if the is has an evolutionary history. 

An “is” does not imply an ought, but the self organizing, precarious “becoming” of a precarious organizational closure does imply instrumental normativity. A life form is not a thing-like is, but a becoming; a metastable living pattern imposing itself on-and-with material-energetic flow in a relationship of needful freedom. In this way, function and normativity is contingent upon the instrumental (organizational) normativity of a self-organizing formal cause. Thus, function is not a final cause but an actively generated cause, specifically an enactively generated cause for organisms undergoing autopoiesis and autonomy. The formal cause that can generate instrumental norms and functions is described as the processual form of life. A life form is a self organizing organism-environment system undergoing precarious operational closure, with instrumental normativity pertinent to survival of its ongoing individuation. Life is a standing-streaming pattern imposed on material-energetic flow, not a static is per Hume. Life is a Kantian whole. 

The Unprestatable 

The previous section claimed that under process ontology, an enactive framework can ground normativity and functionality without appealing to telos as a strictly final cause. The challenge is to reconceive normativity and function not as coming from pre-given, prescriptive “final causes,” but as actively generated from the historical processes of living formal causes. This challenge is to replace final laws with self-organizing laws (autonomy). Functions are not final-causes of behavior. Functions are self-organizing formal causes instrumental to the autopoiesis of life as a pattern of precarious operational closure.

How predictable are organism behaviors and functions in various environments? Can a set of all possible behaviors, functions and norms be stated beforehand? Are functions and norms of behavior final in their predictability and pre-statability or are they un-prestatable? Is an account of functionality framed in evolution and natural selection (e.g. teleonomy) sufficient to explain a complete set of functions? 

These questions are answerable by revisiting the concept of sensorimotor contingencies as developed by Di Paolo et al (2017). Sensorimotor contingencies are divided into four classes. The sensorimotor environment relates a body to an environment, abstracted from the internal dynamics of an organism. The sensorimotor environment constrains, but does not fully determine, organism-environment system behavior. The sensorimotor habitat additionally takes into account the organism’s internal activity, sensorimotor coordination accounts for contextual tasks at play, and sensorimotor schemes account for normative frameworks. 

The structure of the sensorimotor environment [...] can constrain but does not fully determine the possible behavioral strategies for solving the task. It does not provide us with the full explanation of the behavior of the agent[....] The sensorimotor habitat describes the relationship between sensor activity and motor commands taking into account the internal dynamics of the agent. It is the set of all possible trajectories that the agent takes, given a range of boundary conditions and parameters. If we are dealing with a potentially complex, nonlinearly dynamical agent-environment system, providing a full analytical description of this set may be unfeasible. (p. 65, emphasis added.)

A philosophical framing problem prevents the sensorimotor environment from fully determining sensorimotor habits if isolated from an organism's autonomous activity. The sensorimotor environment enables and constrains (but neither causes nor fully determines) sensorimotor habits. Furthermore, the boundary conditions for a complete set of all possible sensorimotor habits cannot be framed if decontextualized from task conditions of sensorimotor coordination in situ and in vivo. Likewise, a normative framework is necessary to qualify the contextual situation. In short, the sensorimotor habits, functions and norms are not pre-stable in advance due to the framing problem. The internal activity or autonomy of a situated, living, self-organizing organism-environment system is what frames conditions necessary to set boundary conditions for potential habits, coordinations, and norms. In other words, autonomy frames function and norms, and teleonomy alone cannot predict nor frame the boundary conditions for a pre-stated set of functionalities. 

In the words of Stuart Kauffman: “No one knows or can know what shall become as the biosphere evolves and shapes its own future in ways we cannot state in advance. They are unprestatable.” (2019, p. 1).  The autonomy and autopoiesis of life generates the salient frames for its habits, and habits are not fully determinable by externalities; autonomous habits are neither heteronomous nor teleonomous. Further, “enablement, not cause, enters our explanatory vocabulary. Much of the becoming of the biosphere has to do with making possible” (Ibid., pp. 116-117). Sensorimotor environments enable and constrain, but do not fully determine, sensorimotor habits. Internal dynamics enable and constrain, but do not fully determine, contextual situations. Normative frameworks are enabled and constrained  (but are not fully determined) by sensorimotor environments, habits, and coordinations.

Pregiven, pre-ordained, or even pre-evolved conceptions of telos implies final causality. Complex systems such as life exhibit emergence and metastability, not finality. Complex systems fail to be completely predictable, therefore final causes and entailing final laws will fail to completely determine these systems. Prescripted telos and teleonomy fail in their roles to determine the final causation of life’s complex normativity. Processes of enaction and autonomy are better suited for this role. 

Telos, normative functionality and goals can be reconceived under enactive philosophy via autopoiesis, autonomy and operationally-contingent instrumental (organizational) normativity. In other words, teleology can be reconceptualized under the project of ontogenology (a typology of generative processes). 

Conclusion

Biological systems, by consequence of their autopoiesis, can yield un-prestatable self designating laws (autonomy) to replace prestated teleonomy and teleology. Functions can be operationally defined by their consequences within a system’s precarious organizational closure as a “Kantian whole.” From this precarious operational closure follows contingent “instrumental ought,” i.e. instrumental (organizational) normativity. Instrumental normativity is irreducibly operationalized within the concrescent processes of autopoietic beings as Kantian wholes. In this way a being is better described as a becoming and a self is better described as a self-ing. A Kantian whole can be defined as follows: within a self organizing being, such as an organism, the instrumental parts (organ, oργανον) exist for and by means of the whole, and the whole exists for and by means of the instrumental parts.

Processes of autopoiesis and enaction yield un-prestatable autonomous behaviors. The set of all possible behaviors is underdetermined and neither pre-statable nor fully entailed by teleology, teleonomy nor the environment. An organism’s behaviors are enabled (but not fully determined) by its environment. An organism’s autonomous internal dynamics frame its possible set of habits, but this set is still not fully determinable without situated context, history and normative framing. In a circularly causal way, life autonomously frames which behaviors are possible, salient and normative. Even still, evolutionarily and developmentally novel behaviors of exaptation (i.e. jury rigging, co-option) defy attempts at pre-stating a completely framed set of possible behaviors. The set of all possible norms, functions and behaviors of an organism-environment system is un-prestatable and open-ended. Function and norms are not final nor enumerable; they are autopoietic, autonomous and indefinitely open-ended. 

Pertaining to the concept of ontogenology from Chapter 1, Stuart Kauffman’s concept of “unprestatable becoming” can apply to a typology for various processes of becoming, i.e. self-organization, complexity, evolution, development and niche construction. Ontically precedent generative processes of becoming include enactive processes of autopoiesis, autonomy, individuation, normativity, agency, sensorimotor intentionality and sense making. The form of life is in a status nascendi, a state of becoming.