Mike Holliday

Section 3: An Alternative Metaphor

In the previous section, I used Hilary Putnam's criticisms of metaphysical realism to suggest that the metaphor of the 'mirror of nature' is deficient as a way of picturing the relation between our minds and the world, and hence is no longer a useful heuristic for understanding how our knowledge relates to reality. Putnam suggests that we picture ourselves, not as in some way separate from the world, observing it as if from a distance (a perspective that readily suggests the possibility of a God's-Eye-View and one determinate description of reality), but as involved in the world, and that such involvement concerns our abilities, including our perceptual abilities, as embodied creatures. Later on, I will be considering the views of a number of psychologists and theoretical biologists that can help to build a coherent version of this alternative 'metaphor' of how we relate to the world. First of all, however, I want to look at one attempt to develop a metaphysics that proceeds along similar lines.

1 The Metaphysics of Brian Cantwell Smith

Brian Cantwell Smith's book On the Origin of Objects (1996) is an extended meditation on the possibility of such a new metaphysics, his avowed aim being to find a path between naive realism and pure constructivism, whilst staying true to the underlying intuition behind each tradition: namely, the 'epistemic deference to the world' that underpins realism, and the 'human involvement in the world' that lies behind constructivism. Smith's background is computer sciences and his interest in metaphysics derives from problems he encountered in trying to develop a sound theory of computation. Some may be surprised by the connection between computing and metaphysics, but computer theorists require an ontology of objects, properties, and relations in order to formalize the steps in the computing programs that they are concerned with.1

The underlying problem that Smith found was that, even in the case of a program that tried to deal adequately with mathematics, he seemed to end up hamstrung by some sort of distinction or structural assumption that was made at the outset. It appeared that what was needed was the ability to 'make distinctions on the fly', rather than to have to formulate all necessary distinctions and concepts at the preliminary stage. He concluded (p. 41) that

[if even arithmetic, the paradigm of formal symbol manipulation] generates this much complexity, that lends strong support to the idea that in more general situations it will be even more inadequate to treat objects as having stable purpose-independent identities, without regard to the functions or regularities in which they participate.

Hence Smith's desire to develop a metaphysics that enables us to understand how objects (and therefore distinctions and structures) can be generated, rather than being taken for granted as an initial element of our analysis. He points out (pp. 17-21) the extent to which his own problems reflected those working in other disciplines, for example complexity theorists, quantum physicists, and social theorists who are dealing cultures other than their own, all of whom have similar difficulties in establishing (except by pre-theoretical fiat) the basic entities with which they are concerned.

Smith notes (p. 16) that in the analytic tradition it is normal to assume that a situation can be parsed or analyzed at the outset in terms of a set of objects, properties and relationships. However, in Smith's view, it is precisely at the so-called 'pre-theoretical' stage that one is liable to introduce presuppositions and biases that seriously limit the usefulness of the subsequent analysis. One particular difficulty is what Smith terms inscription errors, whereby ontological assumptions are inscribed to a system and those assumptions or their consequences are then read back off the system as constituting an empirical discovery or theoretical conclusion (p. 50). Of course, it is usually necessary to inscribe something; the error is in not recognizing the dependence of the result on the original assumption, especially when there are reasons of ideology, naiveté or prejudice for making that assumption in the first place. These sorts of concerns lead Smith to adopt a methodological criterion that he terms the principle of irreduction, whereby no theoretical assumption is taken as being a priori. This implies that for any assumption that we might make, we must be able to say (i) 'where we bought it', i.e. understand and appreciate the circumstances in which that assumption initially arose, (ii) 'how much we paid', for example by recognizing that we will somehow need to take into account those aspects of a situation that the assumption ignores or simplifies, and (iii) 'how we got it from here to there', i.e. we should appreciate that our analysis may be distorted because we are using assumptions and distinctions in circumstances that differ from those in which they first arose (pp. 78-79).

Smith's first basic intuition is that there is 'a world out there' or 'there is more to the world than us' and that any metaphysical account must retain this. This is a realist intuition. However, it is not an assumption that the world exists independently or externally from us, because (i) such an assumption would contradict the principle of irreduction, (ii) we are obviously participants in the world (so it can't be independent of us), and (iii) there is no sense in which the world as a whole can be external to anything. The latter is essentially Putnam's point that there can be no 'view from nowhere', or as Smith puts it:

... there is nothing in fact, or in logical possibility, or in metaphysical possibility, or accessible to the imagination, or in any other form, for the world to be dependent on, independent of, separated from, or external to. The world has no 'other' (p. 104).

Smith goes on to set out seven 'constraints' that this sort of realist intuition imposes on metaphysics (pp. 104-115). These are most definitely not assumptions made at the outset (which would, of course, contravene Smith's own principle of irreduction), but rather 'goals or desiderata' against which the developed metaphysics will subsequently be evaluated.

(i) symmetric realism: Any new metaphysics must treat subject and object equally; this implies that it must provide some sort of resource from which the distinction between subject and object can be derived.

(ii) a single cup: This is in effect a restatement of the realist intuition of 'one world'. Everything, including abstract entities, drinks from one cup.

(iii) Objectivity: The realist intuition 'spells defeat for Objectivity: the mythological idea that there is an attainable or coherent form of knowledge that is independent of any bias, interest, position, perspective, or other fact of subjective particularity ... [;] there is no position from which the whole is visible' (pp. 106-107).

(iv) objectivity: On the other hand, we need to retain some notion of objectivity (lower case), otherwise the result will be an extreme 'anything goes' relativism.

(v) pluralism: This is (roughly) the sort of concept envisaged by constructivism or post-modernism:

... one of the most difficult challenges a metaphysical theory must face is how to do justice to what is right about pluralist institutions, cultural sensitivity, and infinite variability, on the one hand, while at the same time retaining an appropriately rigorous notion of virtue or standards or worth, on the other. ... either without the other is by now too easy (p. 108).

(vi) directedness: Any future metaphysics will need to explain the 'world-directed commitments' of intentional agents.

(vii) objects: It must be recognized that an essential aspect of what it is to be an object (whether physical or otherwise) is its 'resistance', i.e. that we cannot fully control it.

The first step that Smith takes in developing a metaphysics that might meet these desiderata is to consider the characteristics of particularity and individuality, and the differences between them. Particularity refers to the fact that our usual notion of an object involves something that is located or happens, something that is specific. Individuality, on the other hand, involves individuation criteria and the idea that an object is conceived of as a form of unity that is separate from the rest of the world; as such, individuality rests on the notions of sameness and difference. For example, a keyboard is different from the other objects around it, but remains the same when I press down one of the keys. Smith thinks that it is possible for there to be particularity in the absence of individuality. His example is that, if you say 'It's raining', then you are referring to the feature 'raining' being exemplified but not to a specific object or individual. 'It is not features themselves that exhibit particularity without individuality. Rather, what is particular but not individual is the metaphysical patch or disturbance or stuff in the world that warrants the feature placement' (p. 125). It is therefore perfectly possible to identify a feature without at the same time separating out that part of the world as an individual. The lesson that Smith draws is that our metaphysics must not presume the category of 'object', and hence presume the distinction between subject and object. This would be a clear case of an inscription error; we would have already assumed the existence of reference and other forms of intentional participation in the world and so be unable to explain them.

Our next task is therefore to understand how the notion of an object can arise out of a base of sheer particularity. Smith introduces the concept of registration, which equates to 'making sense of', or 'parsing', or 'taking as being a certain way'. This formulation is designed to introduce some sort of relational activity whilst avoiding language that already suggests subject/object, individuation, or related concepts. As an example, the Count of Monte Cristo, when thrown into the sea in a sack, could be said to have registered, but not individuated, the sea. Registration has three key properties (pp. 197-198): (i) it is that activity which leads to an intentional attitude towards the world; (ii) it is neutral as to the location of the splits between subject and object and between subject and any supporting surround (e.g. culture, instruments); (iii) it does not ontologically single out 'objects'.

Another key element in understanding how objects are differentiated is what Smith terms 'intermediate flexibility'. The world is not like a set of gears, where a small movement in one location automatically transmits itself throughout the system; rather, it is normal for effects to dissipate. This flexibility means that one area of the world (which we might label a 'proto-subject' region) can track another area (which we might label a 'proto-object' region), because there is a degree of linkage between the two without them being completely coupled together (Smith uses as an example a 'super-sunflower' that continues to turn the necessary 1/4º per minute whilst the sun is obscured). Thus tracking can continue even if the region which is being tracked cannot for a while be discriminated (pp. 201-203). Without this intermediate flexibility, we would not be able to think of the world as containing an area which is proto-subject and an area which is proto-object; there would be either total flux or a complete gearing of all parts of the world to all others. Such flexibility also provides the proto-subject region with the ability to adjust its state without impacting too significantly on the surround; by adjusting its own state, it is able to maintain a form of coordination with the proto-object region.

The metaphysical picture painted by Smith is of 'a wholly particular world - not individuated at all ... a world filled with particular, deictic flux, riotous and differentiable at every rank ..., not committed to any a priori registration whatsoever' (p. 323). The term flux emphasizes that there are no boundaries and that change can be abrupt as well as gradual. The fact that registration arises from within 'the flux' and is prior to the subject/object distinction means that, for Smith, we cannot think of representation as a basic category: 'it is not just that we do not represent the tree; we are not even distinct from the tree, at least not at first' (p. 217). The coupled interaction between the two regions is the important area of stability in the flux, and this area crosses what will, at a later stage, become the boundary between subject and object. The proto-subject uses its separation from the proto-object to push the stability that exists in their relationship out into the world, i.e. into the object-region; this is the start of individuation and of abstraction away from irrelevant details. As the proto-object area changes, appropriate adjustments take place within the proto-subject area in order to extrude a stable 'object' (pp. 225-227, 241-242). Smith claims that the fact that the subject-region has to take responsibility for maintaining coordination with the object-region whenever effective coupling is broken is 'the origin of reasoning, representation, and syntax: the effective projection, onto the intentional agent, of the requisite arrangements for maintaining long-distance (semantic) coherence' (p. 220). Hence, Smith considers that intentionality is a way in which we utilize our freedom to act locally in order to coordinate with what is otherwise beyond our reach (p. 208).

Smith is careful to note that these arrangements for ensuring the continuation of coherence are 'projected' onto the intentional agent. In other words they are not created by the subject acting as an agent (indeed, this would give rise to a circularity of the very type that Smith is attempting to avoid, since it would presuppose the very category of 'a subject'), but arise out of the linkages between the two proto-regions of the underlying flux. It is this process of 'triangulation' between the two regions that enables ontology and reference to emerge:

It is not ... just that reference is preserved  - as if the objects to be referred to were independently supplied ... . Rather, reference is achieved. These processes of preservation, maintenance of invariance, and the like, are part of the very processes of stabilization that constitute that something as a something (pp. 263-264).

So far, Smith has emphasized the role of the subject in creating an ontology of objects. However, this process equally involves the object, in the sense that the proto-subject cannot stabilize arbitrary regions of the flux by coordinating with them in an appropriate manner. That there are some regions that can be stabilized and hence objectified in this manner is a brute fact of the world, and we have no a priori reason to believe that this must have been the case.

It is important to understand that objects are stabilized not simply through some sort of cognition by the subject but by means of a complex surrounding web of activities, a 'community of practice' or 'participatory surround':

The idea, rather, is that a full spectrum of behavior is ontologically implicated in a system's ability to individuate and register. We need to acknowledge the ... complex networks of actors and activities ... that make up the compensatory behavior necessary in order to triangulate onto an object. This is the realm of history, documents, instruments, social practices, organizations, beliefs ... , whose ontological importance sociology and social studies of science have been at such pains to point out (p. 288).

Smith argues that this metaphysical picture that he has provided is true to both the realist intuition (the world is not just us), and to a pluralist perspective. The fact that objects are 'extruded' from the whole, rather than being put together from predetermined pieces, means that there is an infinity of connections that are unregistered by us, whilst the possibility is left open that they are registered by others. The act of registration means that we are abstracting from what we take to be irrelevant detail, and this is a loss and in a sense a violence to a fuller picture. But we ourselves are part of the flux, and this struggle to individuate therefore also applies to our decisions and commitments; to think otherwise would be to fall back into an externalist viewpoint. There are political implications arising here, as Smith recognizes.2

In summarizing his metaphysics, Smith says that 'Ontology is the projection of registration onto the world. Representation is the projection of registration onto the subject or vehicle' (p. 349). But registration is the underlying notion and therefore we can never have a fully coherent theory of either ontology or representation, since each provides only one of the faces of registration. We can refer to the flux that underlies all reality, but we cannot describe it, since we can only describe anything subsequent to registration and to the extrusion of objects and subjects from the flux. The account given is therefore one of 'ontological pluralism sustained by metaphysical monism' (pp. 374-375). 

The metaphysics countenanced by On the Origin of Objects includes a number of important features that help us to develop an alternative view of our relationship with the world:

Firstly, the emphasis is on flux, rather than existence and objects.

Secondly, the key metaphysical notion of 'registration' is relational.

Thirdly, all concepts (including object, part, whole, causation, and implication) must be constructed in some way out of the flux.

Fourthly, there is an explicit aim to try and avoid either side of binary opposites, such as naive realism vs. idealism, and rational Objectivity vs. a free-for-all pluralism.

One word of warning is necessary here. We must be careful not to assume that 'processes', 'activity', 'relations', or other notions that are emphasized within Smith's metaphysical picture, are ontologically primary in some way. This would offend against Smith's principle of irreduction, which implies that our metaphysics must be 'grounded simpliciter', rather than 'grounded in x' (p. 83). In what follows, I will not be attempting to use the metaphysics set out in On the Origin of Objects in a foundational manner. Rather, I am suggesting that a rewarding way to attack apparently intractable conceptual problems is to assume that notions such as 'activity' and 'relation' are more elemental than the standard static conceptions of 'object' and 'property', and then see where our analysis takes us. Part of the justification for this approach is the belief that we carry around a number of unconscious, and possibly unwarranted, assumptions that arise from taking the more static notions as foundational.

First, however, I need to put some more flesh on the metaphysical bones. I will be looking at the views of a number of psychologists and theoretical biologists whose theories throw additional light on how we might view our relationship with the world. After all, if we reject the formalistic view of a rational mind reflecting a pre-given world, we are put back into the position of considering ourselves as the particular type of living creatures that we happen to be, continually involved in a world of which we are part.

2 The Santiago School

Whilst Brian Cantwell Smith's investigation is into the general nature of the origins of objects, the Chilean theoretical biologists Humberto Maturana and Francisco Varela are concerned with the question what sort of entity is a living being? (Maturana & Varela 1998). For Maturana and Varela, they are entities that 'produce themselves and specify their own limits' (p. 40). The first key element here is the notion of autopoiesis, or 'self-creation', whereby a system differentiates itself from its environment by its own dynamics. Maturana and Varela distinguish between an entity's organization, which consists of those relations that must be present for something to exist as a such-and-such, and the entity's structure, which consists of the actual relations that exemplify that entity's organization at a particular point in time. The idea is that a living being maintains its organization (until it dies), but changes its structure through time; in other words, the organism attempts to maintain the dynamic processes that constitute it, rather than any of its specific constituent parts. The second important element of a living organism is that of a boundary, a limit to the network of dynamic relations that comprise the living being. This boundary, or membrane, is itself part of the dynamic interactions that constitute the organism. Thirdly, an autopoietic system not only maintains its defining organization, but also regenerates the components out of which it is formed. Life is defined by Maturana and Varela as autopoiesis in physical systems, and the key living entity is therefore the cell.3

As an autopoietic system, an organism responds to perturbations from the environment by changing its structure (i.e. changing some of its elements and/or the relations between them) so as to maintain its organization. For Maturana and Varela, the environment does not determine the changes to the organism; instead, the outside world triggers some form of change in the organism, but what this change turns out to be is determined by the structural history of the organism itself. This gives rise to the concept of closure, which does not imply that an organism is isolated from its environment, but rather that the change of its state following some form of impact from the surrounding environment is propagated solely by means of the processes by which the organism is constituted. In complex organisms, the plasticity of the nervous system, including the brain, provides a vast domain of possible changes in structure in response to the impact of the outside world (pp. 163-166).

The domain of explanation that applies to living organisms is therefore different from that for physical phenomena. Because organisms are autopoietic entities, the nature of the phenomena that they bring about depends on the organism's organizational processes and not on the physical attributes of its components. The two domains of explanation are nevertheless consistent; this is because the organism's organization must be substantiated in a particular structure at any one time, and the physical components of that structure will be subject to the laws of physics (p. 51).

A living entity also acts on its environment (for example, an organism's alterations in response to the environment may include changes in its motor neurons, leading to physical action) and the two-way interactions between entity and environment will consist of reciprocal perturbations, producing what Maturana and Varela refer to as structural coupling. Therefore Maturana and Varela conceive of the environment as a source of perturbations to a living entity, rather than as a source of information (pp. 95-97, 169). Similarly, there is no place for 'representations' or 'ideas' in Maturana and Varela's ontology; of course, we as observers can describe an organism's behavior as if it were based on representations of the world outside it, but these descriptions do not reflect the causal mechanisms within the nervous system of the entity we are considering, and do not therefore, claim Maturana and Varela (pp. 131-132), provide a adequate scientific explanation of the organism's behavior.

For Maturana and Varela, thinking consists of the activity of a nervous system that interacts with certain of its own states as if they were entities that were independent of the organism. It is another type of process by which an autopoietic system adjusts to perturbations from the world outside, albeit an exceedingly complex and recursive process. Cognitive activity must therefore be an embodied activity, because it depends upon the changes in the structure of the organism; in effect, it represents flexible and adaptive interaction through the structural coupling between the organism and its environment. A basic cognitive activity is that of distinction, whereby an organism separates out an object from its background; for Maturana and Varela, this activity constitutes the object, as such. Categorization is therefore a key activity of a living system:

... the predictions implied in the organization of the living system are not predictions of particular events, but of classes of interactions. Every interaction is a particular interaction, but every prediction is a prediction of a class of interactions that is defined by those features of its elements that will allow the living system to retain its circular organization after the interaction, and thus, to interact again (Maturana 1970). 

Just as thinking is one way of adjusting to the impact of the outside world, so communication does not consist of the transfer of information but is instead a coordination of behavior through structural coupling between two or more organisms (Maturana & Varela 1998, p. 196). This accounts for the lack of similarity between linguistic behavior, for example using a word, and related actions, such as distinguishing what the word refers to (Maturana & Varela 1998, p. 208). There is an affinity here with Putnam's conclusion that one of the problems with metaphysical realism is the lack of similarity between the linguistic domain and the domain to which reference is made, leading to the idea that the two domains must be linked by some form of correspondence, a notion that turns out, claims Putnam, to be empty of substance.

Autopoietic theory deals with systems and how they are to be addressed from the perspective of the observer observing them. The concept of the system as a unity must therefore be understood as a unity with respect to the observer, rather than as a unity in a metaphysical sense: 'Strictly, the identity of a unit of interactions that otherwise changes continuously is maintained only with respect to the observer, for whom its character as a unit of interactions remains unchanged' (Maturana 1970). It is therefore important not to take 'autopoiesis' in a metaphysical manner, as this can lead one to think of it as some form of vitalist explanation of living systems (Whitaker 2003, entry for autopoiesis).

Because distinctions are made by living entities, objects cannot be said to exist in a domain that is prior to, or independent of, the observer:

... without observers nothing exists, because existence is specified in the operation of distinction of the observer. For epistemological reasons, we ask for a substratum that could provide an independent ultimate justification or validation of distinguishability, but, for ontological reasons, such a substratum remains beyond our reach as observers. All that we can say ontologically about the substratum that we need for epistemological reasons, is that its permits what it permits ... Once a domain of reality is brought forth, the observer can treat the objects or entities that constitute it both as if they were all that there is and as if they existed independently of the operations of distinction that bring them forth (Maturana 1988b).

What we see in Maturana and Varela are a number of themes that are similar to those that we have already come across in Brian Cantwell Smith, but now in a theoretical setting that is less general, a setting that stresses the embodiment and activity of living beings. In particular:

(i) Living entities (potential 'subjects') are areas of the world that differentiate themselves from the rest of the world (the 'environment') by means of their dynamical behavior, assisted by physical boundaries (or membranes) between themselves and their immediate environment.

(ii) The interactions between the living entity and the environment constitute a structural coupling, and the living entity alters its detailed structure so as to maintain its organizational dynamics in the face of perturbations from the environment.

(iii) A basic cognitive activity is distinction, whereby we are able to separate out an object from its environment - this separation is an activity of the observer and is not 'out there' in the world as such.

(iv) Hence objects do not exist per se, but only for an observer, and there are many different ways that a living entity can carve up the world, as it were, into different sets of objects.

(v) Similarly, communication, information, and representations are not items that we can say exist 'out in the world' but are characterizations that we make of the results of the structural coupling between ourselves as observers and our environments.

3 Howard Pattee

If Maturana and Varela are attempting to answer the question 'what is life?', the theoretical biologist and systems scientist Howard Pattee considers the related question 'how is it that matter ever came to be symbolic?' 

Pattee argues that physical theories and semantic theories provide complementary descriptions of reality. His basic intuition is that the distinction between these two different forms of description depends upon which aspects are taken to be invariant. Pattee explains that physical laws cannot describe the symbolic domain because such laws are framed so as to be as independent of observers as possible, thereby placing heavy emphasis on principles of universality, invariance, and symmetry. One implication of this is that the laws that we express in our theories of physics relate to those universal aspects of the world that have no significance for us as individuals. On the other hand, symbolic descriptions are specifically designed so as to be context-dependent, relevant to the local situation, and independent of physical processes (Pattee 1995a).

Pattee says that it may be difficult for us to see the connection between symbolic descriptions and physical descriptions because our symbolic systems have traveled so far from their beginnings in the first living beings. Nevertheless, says Pattee, the physical realm and the symbolic realm are complementary descriptions of the same reality:

I am a physical reductionist in the sense that I believe all symbolic behavior must have a material embodiment, following physical laws, that correlates with this behavior. ... However, I am not a reductionist in the sense of those who claim that symbols are 'nothing but' matter. 'Nothing but' implies that the only model that is required to understand symbols is a complete materialist or physical law model. ... My position is that no complete physical description of these material structures, although correct in all details, will tell us all we need to know about their symbolic function. Briefly, this is because symbol function, like all biological function, is not an intrinsic or law-based property of the material symbol vehicles but a selective survival property of the populations of individuals that use the symbols for material construction and control in a particular environment (Pattee 1995a). 

However, we cannot produce a single description that encompasses both the physical and symbolic aspects of a situation. For example, if we were to make a complete physical description of, say, a measuring device, we would lose sight of its functionality.

Pattee quotes von Neumann to the effect that we always have to make an epistemic cut, i.e. to divide the world between observed system and observer, even if this split is, to some extent, arbitrary. He believes that a number of the dualisms that have plagued Western Philosophy derive from the necessity of making some form of epistemic cut. The cut results in a distinction between physical descriptions and symbolic descriptions, which means that mind and matter (as well as symbols and their referents) tend to be seen as distinct domains, hence generating the problem of how one of the domains relates to the other:

It is the essential function of a symbol to 'stand for' something - its referent - that is, by definition, on the other side of the cut. This necessary distinction that appears to isolate symbol systems from the physical laws governing matter and energy allows us to imagine geometric and mathematical structures, as well as physical structures and even life itself, as abstract relations and Platonic forms. I believe this is the conceptual basis of Cartesian mind-matter dualism. This apparent isolation of symbolic expression from physics is born of an epistemic necessity, but ontologically it is still an illusion (Pattee 2001).

We can, I think, identify two rather different versions of the epistemic cut in Pattee's writings. Firstly, there is the cut that we make as scientists when we make the decision to describe a situation either in dynamic terms, e.g. when we do physics or chemistry, or in semantic terms, e.g. when we do biology. Secondly, there is an epistemic cut that we make as perceiving beings, when we say or think 'that is an x'. For example, Pattee explains that '[the epistemic cut] arises whenever a distinction must be made between a subject and an object, or in semiotic terms, when a distinction must be made between a symbol and its referent or between syntax and pragmatics' (Pattee 2001). This cut appears to be the source of knowledge, conceived of as knowledge of an external world.

Much of Pattee's work is concerned with how symbols and meaning first came about, and his discussion centres on the evolution of the genotype/phenotype distinction. Pattee views genetic material, such as DNA, as providing a physical substrate that is (relatively) stable and reliable - unlike proteins, which are embedded in a sea of continuous catalytic activity. This stability allows the genetic material to generate activities that we can describe as involving information storage and memory, thereby forming the starting point for the reliable construction of proteins and hence of living organisms. Pattee has termed this combination of symbolic and physical activity in one and the same system semantic closure

... self-reference that has open-ended evolutionary potential is an autonomous closure between the dynamics (physical laws) of the material aspects and the constraints (syntactic rules) of the symbolic aspects of a physical organization. I have called this self-referent relation semantic closure ... because only by virtue of the freely selected symbolic aspects of matter do the law-determined physical aspects of matter become functional (i.e. have survival value, goals, significance, meaning, self-awareness, etc.). Semantic closure requires complementary models of the material and symbolic aspects of the organism (Pattee 1995a)

The stable, genetic elements of the system lend themselves to a symbolic description, whereas other processes lend themselves to description in terms of dynamic processes. The overall system is co-determined by the two sorts of processes, and allows for the creation of domains within the system that show structural and functional autonomy. In effect, structure and meaning are created 'from within', rather than being obtained from outside the system (Cariani 2001).

Pattee also stresses the origin of the symbolic realm in the perceptual activity of the earliest living organisms. In order to symbolize something, we have first to classify the world, or distinguish between different elements: '[the epistemic cut] has a primitive origin and is found in all living organisms. It is simply an extreme case of the distinction made, even by the first cells, between stimuli that cannot be correlated and stimuli that can be correlated or that follow a recognizable pattern' (Pattee 2001). Ultimately, it is evolution that produces the symbolic realm; symbolic function is not a property that belongs to something because of some physical or formal law, but because of its contribution to the selective survival of the entities that use the symbols (Pattee 1995a; see also Rocha 1998, section 4.1).

But what exactly is the relationship between what appear to be two different explanations of the origin of symbolic existence, i.e. (i) the development of perception in early life forms, and (ii) the evolution of physically stable, 'information-bearing', genetic material? Pattee suggests that both the genotype and our perceptual distinctions revolve around the seeking out of invariance and compressibility of description; without compressibility life would not exist because, he claims (1995b), it is not possible to adapt behavior to an endless stream of random events. I read this as suggesting that in perception we have to identify what are, for us, invariants in the background of the environment (and ultimately this is achieved by natural selection favoring organisms which can 'identify' invariants that are relevant to the organism's survival). Similarly, the development of genetic material enabled an invariant structure (for the construction of proteins) to be passed on to descendent organisms. Now the first-mentioned development, as I have described it, could not have occurred without the second, since hereditable structure is needed in order for organisms with perceptual abilities to evolve. I believe that this is why Pattee usually describes the genotype/phenotype distinction as being the key step in the development of the semantic world.

From Pattee's account of the creation of the symbolic domain within living creatures, there is the clear implication that meaning must be embodied.

4 The Origin of Consciousness - Nicholas Humphrey

In his book A History of the Mind (1992), the psychologist Nicholas Humphrey discusses the origin of phenomenal experience in terms which bear a strong resemblance to Brian Cantwell Smith's account of an emergent metaphysics. Humphrey's starting point is also reminiscent of Maturana and Varela; he describes (p. 18) living creatures as 'self-integrating' and 'self-individuating', with boundaries that are 'self-imposed and actively maintained'. According to Humphrey, such boundaries are crucial for the emergence of consciousness, because early living creatures evolved so as to be sensitive to the exchange of matter, energy, and information across their boundaries. Initially, their surfaces may have simply reacted to what was impinging on them; at a later stage, the reactions spread to other parts of the organism leading, for example, to a movement away from, or towards, chemicals, light, or pressure that affected their surfaces. Still later, parts of the organism's boundary become 'sense organs', and it might, for example, move to deeper water in response to blue light and shallower water in response to red light. At this stage, sensitivity and responsiveness to the external environment were strongly linked. But eventually a more sophisticated approach developed whereby the responses of the animal were not automatically relayed so as to produce actions, but instead held 'in abeyance', as it were, as patterns of possible actions within the nervous system. Humphrey describes (p. 20) such patterns as action-based representations of the effects of the external environment on the organism's body, and claims that as such they are the first meaningful subjective phenomena.

But such information also represents the external world, so we have the possibility of both (i) a subjective 'how it affects me', and (ii) an objective 'how it is out there', from the same source. The first of these we can term 'sensation', and the second, 'perception'. The information processing to cope with (ii) has to be rather different than that for (i), and has to deal with the permanence of what are now taken to be 'objects', and the possibilities that such objects afford for the organism's future interaction with its environment. This way of dealing with the impact of the environment on the organism leads to intentional objects about which the organism can have cognition and knowledge. So, although the source of these two kinds of mental representation, of subjective feelings and of physical phenomena, are the same, they must be different types of representation (pp. 22-23). That they are still separate in humans can be deduced from some of the dissociations noted in the clinical literature, for example visual agnosia, where there is defective perception with normal sensation, and blindsight, characterized by perception with defective sensation. Humphrey suggests (pp. 62-73) that it is sensation, rather than perception, that produces the indexicality of our conscious experience, its 'here-ness', now-ness', and 'me-ness'.

Humphrey points out that, because perception involved the brain in 'constructing' the objects in the environment, it was important for there to be some form of error-correcting mechanism that enabled the organism to check that its perceptions did indeed generally conform to reality. One obvious way to detect errors would be to try and reconstruct the original sensory input from the perceptual result; the reconstruction could then be transmitted to the 'sensory center' to see if there was a match with what was actually being sensed. Mental images generated by the 'perceptual center' would be continually rejected by the error detection mechanism since they would conflict with the actual sensory stimuli, which is presumably why mental images are so fleeting. Humphrey's account (pp. 84-96) implies that the same area of the brain should be active both for an external stimulus and for a mental image, and there is indeed some evidence that this is the case. 

There are a number of characteristics that are common to both sensations and bodily actions, and this suggests to Humphrey that the former have their origin in the latter. These similarities are that both sorts of activities (i) belong to the subject, (ii) are tied to a particular site in the subject's body, (iii) have a specific quality (or modality, in the case of sensations), (iv) are present-tense, and (v) are phenomenally immediate or self-characterizing - in other words, someone who has a particular sensation knows immediately what its properties are (pp. 112-141). The fact that sensations are self-characterizing suggests to Humphrey that they are 'mine' in the way that my limbs are, and hence that sensations are close to being some kind of bodily activity in their own right.

Humphrey provides a hypothetical evolutionary account of how sensations might have originated in such a manner. Initially, says Humphrey, sensation was a method of doing something at the point on an organism's surface that was affected by the environment. The next step was for the information about the stimulation to be relayed to a more centrally placed neural location before a response was made. At a still later stage the response is not made to the bodily surface itself, but to the nerve between the bodily surface and the brain. Finally, the response is to an 'inner model of the body': 'I am still maintaining that to have a sensation involves the making of a "sensory response". But this response, which began its theoretical life as a real bodily activity, has now become some sort of brain activity' (p. 163). Humphrey suggests (pp. 185-187) that the sensory characteristics of specific sensations are 'skeuomorphic' features that derive from evolutionary more ancient patterns and functions that have been preserved by biological inertia.

Therefore, to be conscious is, for Humphrey, to have affect-laden mental representations of something happening to me here and now (pp. 110-111). The subject is therefore an embodied self; it would not exist in the absence of bodily sensations. Other types of mental activities, such as thoughts and beliefs, are conscious only to the extent that they are accompanied by some form of reminder of sensation, as is the case in mental imagery.

5 Discussion

We have been considering the views of Smith, Maturana and Varela, Pattee, and Humphrey in order to find an alternative way of conceiving our relationship to the world. The story that we can see developing might be characterized as follows: (i) The world is flux (Smith), out of which there emerge areas where the activity becomes complex and circular, and some of these areas develop boundaries which further enable the circularity and complexity of the processes that they enclose (Maturana). These areas can be characterized as living organisms. (ii) At an early stage, there developed basic genetic material, which enabled continuity of structure to be maintained (Pattee). (iii) Living entities react to the impact of the environment upon their boundaries, and to the flows across their boundaries, by means of internal changes (Maturana), and the evolution enabled by the existence of genetic material leads to the development of organisms that react in ways that contribute to their continued existence and reproduction. Reactions to the environment become more complex such that the organisms start to show behavior that distinguishes aspects or parts of the environment from the rest of the environment. Ontologically, this is the creation of objects (Smith, Humphrey, Maturana). It is also the way in which meaning is created within the system, rather than being imposed from the outside; ultimately, therefore it is evolution that allows for the creation of the semantic domain (Pattee). (iv) Organisms develop so that distinctions that originally resulted in simple physical behavior (e.g. a movement away from a chemical) now result in more complex internal behavior, with the development of sense organs and the nervous system (Humphrey). (v) But the universe has to be such that autopoietic beings can arise and that invariants can be extracted from the environment (Smith, Maturana). In other words, the universe has to allow for the compressibility of description (Pattee).

The metaphor that we might use to characterize the above account (which will be developed further in Section 4) is therefore one of emergence.4 This provides a way of looking at the world that does not suffer from the problems that Putnam attributes to the conception of an 'interface' between ourselves and the world, but without 'over-finessing' the problematic by endorsing a return to naive realism or a Wittgenstein-like rejection of such matters. The difficulty with metaphysical realism, as Putnam conceives it, is the separate nature of the two domains, world and language, which are in turn related to the two domains in the theory of perception, world and percepts. The separation of domains seems to require some form of interface, but this leads, suggests Putnam, to incoherence and a lack of real explanatory power when we come to consider how the domains relate to each other. The alternative metaphor of emergence takes as primary, not the separate domains, but the relationship between them. There is a world, but objects and subjects are not considered to exist as parts of the world but are taken to arise out of the relationships within the flux. As such, they are secondary to the relationships that constitute them and therefore (i) no interface is necessary, and (ii) meaning, truth, and reference arise out of the co-constitution of subject and object, rather than being mysterious relationships that have to be added to the two pre-existing domains. Such a metaphysics does not appear to be subject to the criticism that has been leveled at Putnam's internal realism, to the effect that it is unstable, always threatening to collapse into a cultural-relative anti-realism.5 Because the relationships that constitute object and subject are primary, there is always flexibility in the resulting ontology, but the subject-object constitution is of both at the same time, and not the constitution of a domain of objects by a preconceived subject (no matter whether we suppose such a subject to be defined as a transcendental subject, a cultural subject, or otherwise). Objects and subjects depend upon the relationships that arise from the world as flux; they do not arise de novo, by means of some form of metaphysical constitution.

This version started: July 2004

Last updated: July 2005

Here are links to:

To send me an email, click here.


[1] This can be seen in Tom Gruber's (1993) definition of ontology within computing science:

A body of formally represented knowledge is based on a conceptualization: [that is,] the objects, concepts, and other entities that are assumed to exist in some area of interest and the relationships that hold among them. ... An ontology is an explicit specification of a conceptualization.

[2] The political and social consequences of Smith's metaphysics will be touched upon in the section In Lieu of a Conclusion ... .

[3] It is worthwhile expanding somewhat on the significance of a 'boundary' for the emergence of cells, and hence of living organisms. In the view of Smith and Szathmary (1995), the important elements of living things are metabolism, i.e. a 'linked series of chemical reactions, driven by an extrinsic source of energy', and heredity, i.e. a form of 'template replication'. However, once we get past an early prebiotic stage, there is also some form of membrane to keep the molecules of the 'living thing' separate from the environment. Prior to this, it is possible to envisage a basic form of chemical metabolism and template replication on, say, the surface of a rock, but this is (according to Smith and Szathmary) simply a population of molecules interacting ecologically and not yet an individual in the sense that, say, a bacterium is. However, the formation of membranes enabled the existence of the first cellular organic structures. The compartments formed by the membranes can then support a division of labor between cells, with only some cells being replicators and others coding for enzymes that are useful for the system's general metabolism.

[4] Emergence is only a possibility because the universe happens to allow for the compressibility of description, as described by Pattee. Without this compressibility, there would be no invariants to be extracted from the environment, and therefore no entities to be objectified. Emergence therefore represents the possibility of such objectifications at multiple levels, such that one level is seen by us as 'arising out of' another level.

[5] See, for example, Norris 2002.