In this paper we present a new model for the mechanism underlying what is traditionally known in immunology as the “selfnonself” distinction. It turns out that in operational terms, the distinction effected by this model of the immune system is between a sufficiently numerous set of antigens present from the start of the ontogeny of the system on the one hand, and isolated antigens first introduced after the system has reached maturity on the other. The coincidence between this “founder versus late” distinction and the traditional “somatic self-foreign pathogen” one is essentially contingent, an example of the purely opportunistic tinkering characteristic of biological organization in general. We conclude that the so-called “self-nonself” distinction in immunology is a misleading misnomer. This raises the question as to what would genuinely count as a “self-nonself” distinction, a fundamental question for biology in general and Artificial Life in particular.

In the field of artificial life there is no agreement on what defines ‘autonomy’. This makes it difficult to measure progress made towards understanding as well as engineering autonomous systems. Here, we review the diversity of approaches and categorize them by introducing a conceptual distinction between behavioral and constitutive autonomy. Differences in the autonomy of artificial and biological agents tend to be marginalized for the former and treated as absolute for the latter. We argue that with this distinction the apparent opposition can be resolved.

Abstract. Biology is, better than anything else, about existence in time. Hence biological reality cannot be defined without reference to a temporally situated observer. The coupled or detached character of this observer (with respect to the own time variable of the system) provides a link between the observer and the observed. This connections delimits the kinds of scientific descriptions that can be given at all by an observer. In particular, two fundamentally different forms of description, corresponding to different epistemological attitudes and different philosophies of science, called endo- and exo-physics, can be distinguished. Two old puzzles, the Omniscience Problem (illustrated here on the example of Internal Chemistry) and the Chameleon Problem (originally an argument against philosophical functionalism) are reconsidered in the light of these distinctions. As application, the question, in what sense computer models of life can be suitable for studying life, is examined.

This paper aims at understanding modular organization in multivariate dynamical data. In contrast to information-theoretic approaches, the authors start from the complementary point of view of statistical modeling and prediction of dynamical systems. They arrive at the conclusion that modularity is not necessarily an objective property of a system’s organization but rather is inferred by cognitive systems as it can simplify learning and lead to gains in predictive power. This conclusion may prove useful for constructivist approaches as the paper establishes in formal ways that learning agents may perceive modularity and correlations among variables in their environments, even when such variables are actually dependent on others.

Since the concept of autopoiesis was proposed as a model of minimal living systems by Maturana and Varela, there has been still few mathematically strict models to represent the characteristics of it because of its difficulty for interpretation. This paper proposes a formal description of autopoiesis based on the theory of category and Rosen’s perspective of “closure under efficient cause.”

One major criticism of direct or active perception (and other forms of embodied action) from the perspective of cognitive psycology is that, according to common sense, there are some actions that require strictly symbolic information – for example, to stop a car in response to a red traffic light – which fall outside the realm of a perception-action cycle. Although such cognitive responses are not necessarily a goal of artificial life, they must necessarily be included within the embodied paradigm if it is to encompass the cognisant individual, the self-aware individual, or, potentially, the conscious individual. This paper will address the question, ‘can an animat appreciate art?’ Although this may seem very different to the example of a prosaic response to a traffic light, it will be argued that a common framework for establishing the meaning of an object is needed. It will also be argued that clarification to previous philosophical models of artistic engagement is required: in particular that the process of understanding is not a dialogue between an autopoietic artwork and animat, but that there is either a unity of object (artworkanimat) which becomes self-maintaining, or a more classical Gibsonian interpretation as a fixed set of affordances offered by an object to the subject, both of which lead to the conclusion that the process of understanding becomes a resonance in the unity or animat.

Key words: Autopoiesis, hermeneutics, embodiment, phenomenology, direct perception, natural vision

The lack within AL of an agreed-upon notion of life and of a set of criteria for identifying life is considered. I propound a reflection upon the codified nature of the organization of living beings. The necessity of a guiding notion based on the coding is defended. After sketching some properties of the genetic code I proceed to consider the issue of functionalism as strategy for AL. Several distinctions ranging from plain multiple realizability to total implementation independence are made, arguing that the different claims should not be confused. The consideration of the semantic and intrinsically meaningful nature of the code leads to discuss the “symbol grounding” in AL. I suggest the principle of Semantic Closure as a candidate for confronting both problems inasmuch as it can be considered an accurate guiding notion to semantically ground Artificial Life.

This paper describes a method for a computer exploration of formation of structures based on the network of autonomous units. This method has a biological correspondence with morphogenetic processes. The interactions in the network of autonomous units are modeled by two kinds of forces: repulsive and attractive forces. When and what kind of forces are active at each unit is based on genetic information and environmental factors. Genetic information enables the use of evolutionary algorithms to evolve the interactions and thus to create new structures. Environmental factors provide the needed restrictions for the space of possible structures. Depending on what meaning is given to the units, the system is capable of simulating various kinds of emergent phenomena. For example, in the case where units are interpreted as cells, where the repulsive and attractive forces represent collision and adhesion forces, a formation of multicellular organism can be achieved.

Abstract: We argue that the significance of the spatial boundary in autopoiesis has been overstated. It has the important task of distinguishing a living system as a unity in space but should not be seen as playing the additional role of delimiting the processes that make up the autopoietic system. We demonstrate the relevance of this to a current debate about the compatibility of the extended mind hypothesis with the enactive approach and show that a radically extended interpretation of autopoiesis was intended in one of the original works on the subject. Additionally we argue that the definitions of basic terms in the autopoietic literature can and should be made more precise, and we make some progress towards such a goal.

Key words: Operational limit, operational closure, physical boundary, relational domain, spatial boundary

In this paper, I propose a neo-Heideggerian framework for A-Life. Following an explanation of some key Heideggerian ideas, I endorse the view that persistent problems in orthodox cognitive science result from a commitment to a Cartesian subject-object divide. Heidegger rejects the primacy of the subject-object dichotomy; and I set about the task of showing how, by adopting a Heideggerian view, A-Life can avoid the problems that have plagued cognitive science. This requires that we extend the standard Heideggerian frame-work by introducing the notion of a biological background, a set of evolutionarily determined practices which structure the norms of animal worlds. I argue that optimality/ESS models in behavioural ecology provide a set of tools for identifying these norms, and, to secure this idea, I defend a form of adaptationism against enactivist worries. Finally, I show how A-Life can assist in the process of mapping out biological backgrounds, and how recent dynamical systems approaches in A-Life fit in with the neo-Heideggerian conceptual framework.

Key words: adaptationism, Cartesian, enactivism, Heidegger, significance, subject-object dichotomy.