Boundary Questions Between Ontology and Biology

Abstract

This chapter deals with some problems linking biology and ontology. After a general survey of the most prominent ontological questions lying behind biology, the study case of biological boundaries is addressed. The scrutiny of the relevant literature shows that biologists speak of various types of boundary: perceptual, compositional, epithelial, cellular and sensu lato processual boundaries; all of them appear to be, in a way or another, flawed by some theoretical inconsistencies. So, a new concept of organismic boundary is introduced and discussed, by which the organismic boundary is the (concrete) part of an organism which spatially encompasses all and only the other (concrete) parts of that organism.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

8.1 Introduction: Ontology and Biologists

Many biologists do not care at all for ontology¹: either they simply choose not to tackle the ontological premises and underpinnings of biological phenomena, or they firmly claim that ontology is merely a residual of superseded superstitions, if not a true obstacle to research.

In my opinion, however, not to have an ontology is to have a bad ontology; so, I think it preferable to positively address the ontological questions whenever and wherever they appear in the course of biological research, instead of obscuring or denying them.

Obviously, it is the general and theoretical biologist who must, in a way or another, explicitly take into consideration such questions, while the experimenter may simply accomplish his work without forgetting how many ontological problems hide behind it, with but a hope at a glimpse beyond his specialised field of investigation from time to time.

8.2 Ontological Questions in Biology

Now, what are such most prominent ontological questions lying behind biology?

If we skim some papers and treatises of ontology, looking for those chapters which would be worth the attention of a biologist, we can easily discover many links between these two fields of knowledge. I shall thus list some of them, mainly in the field of organismic biology and along the lines of the eighth chapter of Mahner and Bunge (1997), which is explicitly devoted to the ontological fundamentals of biophilosophy²; however, a glance at other handbooks (e.g. Jacquette 2002) would have yielded – if surely not the same ontological framework – a similar list of ontological items.

Mahner and Bunge’s starting point is given by the analysis of (concrete) things, and actually in most cases biologists deal with (concrete) bodies, their parts and/or the bodies they are part of. Here, the conceptual opposition³ part/whole is involved, with all the subtle problems posed by the dissection (be it factual or conceptual) of living bodies into their parts, or by the relations between part and part (e.g. horizontal coordination) and between part and whole (vertical subordination); a very delicate question, which still lies largely unsolved, is that of an organism’s proper parts: given a pregnant woman, it is rather easy to recognize that the foetus is not a proper part of its mother, but in some cases it is very difficult to distinguish an organism’s proper parts from bodies having accidentally or temporarily penetrated from outside, like some symbionts (Ramellini 2006a).

Things are characterized by their properties (accidents, qualities), whose study is linked to the biological analysis of characters, states of characters or traits of phenotypes (Laubicher and Wagner 2000). Relational properties are particularly important in biology (Rashevsky 1961, Rosen 2000), to the point that Mayr (1982) has qualified it as quintessentially a science of relations.

When properties hold constant relations among themselves, Mahner and Bunge claim laws obtain, and everybody knows that biological laws or quasi-laws, as well as their extension and generality, often stir hot debates in the philosophy of biology (Brandon 1997).

The totality of properties of a thing at a given instant constitutes its state, a concept which also links biology to physics, particularly to thermodynamics, with the endless and somewhat boring debate on an (alleged) steady or non-equilibrium state of living bodies (see Ageno 1986 for a convincing setting-up of the question).

Changes in/of things are of the utmost importance for biology (Salthe 1993), as one can assess by merely listing the questions involved:

• the close association usually asserted between life and dynamism or metabolism, with the ensuing problems in cases of metabolic reduction or standstill;

• the importance of both ontogenetic (please note the etymological root onto-) and phylogenetic change;

• the debate on the historical character of living bodies;

• the question whether biological change is merely quantitative or also qualitative.

From a traditional vantage point, change also implies a passage from potency to act; this conceptual opposition has obvious counterparts in biology, like all developmental potencies (from totipotency to unipotency, or even to the ontologically problematic notion of nullipotency; see NIH 2006), or the distinction between a biological possibility/capacity and its actual implementation (as with the problematic notion of the metabolic capacities of a cryoconserved tissue).

Finally, change implies a reference to a spacetime framework (for a Whiteheadian viewpoint see Woodger 1967, 299 ff.). In this respect, the synchrony/diachrony couple is spread everywhere in biology, sometimes rather hidden as in the opposition between (spatial) anatomy and (temporal) physiology, sometimes patent as in the nondimensional/multidimensional concept of biological species, or in the debate between historical and ecological biogeography.

Differences between things then call for the notion of individuality, notoriously at the core of hot disputes about the ontological status of the human embryo (see for instance the debate after the publication of Ford 1988), or of taxonomic species (since Ghiselin 1974).

In turn, individuality involves questions of identity (Wilson 1999):

• the numerical identity of biological entities: for instance, what we call human body actually is a symbiotum, where human cells are numerically a small minority of all cells present in that body’s place (see below);

• qualitative identity, with the huge biodiversity of organisms and taxa, and problems of natural kinds, species and classification (Schuh 2000, Coyne and Allen Orr 2004); diversity takes also the appearance of anomalies-abnormalities, opening the door to pathology (which after all is a chapter in general biology) and its discussions on the ontological status of diseases (Thagard 1999);

• diachronic identity, with problems of identity maintenance through time during development, especially when a metamorphosis occurs, or during processes of body division or fusion, like in autotomy or fertilization (Boniolo and Carrara 2004).

Things do not occur alone, but compose either aggregates or systems. The above mentioned debate on the ontological status of embryos mainly lies in considerations about its being a system: is a human morula a mere aggregate of blastomeres (i.e. not an embryo but at most a pre-embryo), or since the very fusion of gamete plasmalemmas does it constitute a system (i.e. first a unicellular and then a multicellular organism)?

A system is always set apart from a background, i.e. its environment (Ramellini 2007): thus, we find here the mereological and topological problems posed by boundaries between living bodies (be them unicellulars, cells inside multicellulars, or multicellulars) and their surrounding environment (more on this below); boundaries are even more problematic in the case of systems composed of spatially scattered components, like ecological communities, biological populations and ecological niches (Smith and Varzi 2002).

Other boundaries are found in respect to time, when system emergence or submergence occurs. In this respect, two fundamental ontological concepts appear to me largely underestimated by biophilosophy, namely:

• the concept of genesis, which lies at the root of primary biological notions like epigenesis, genetics, phylogeny and abiogenesis (Ramellini 2003);

• the concept of corruption (in the sense of the Greek phthrà), which results nearly forgotten, obviously apart from the case of organismic death.

Often emergence occurs by assembly and self-organization, two concepts that intervene in developmental biology and above all in the origin(s) of life, one of the greatest unsolved problems in biology (Fry 2000, Luisi 2006). The term organization contains the Aristotelian etymological root organ-, which is very fertile in biology, from organelles to superorganisms, or from organs to organisms and, indeed, to organization and self-organization themselves (Schiller 1978). In general, self-properties are of the utmost importance in biology,⁴ and have prompted in the last decades a vast literature, especially about self-organization (Kauffman 2000).

Processes of organization may lead to the emergence of new levels of reality (Poli 2001). In biology, handbooks and treatises are usually arranged according to a hierarchy of levels of organization, but the question stands as to which and how many levels to recognize or establish (Ramellini 2001).

The concept of cause or causation is to be found in the old question of final causes and teleology, recently revived by the concept of teleonomy, and undoubtedly associated with the notion of physiological function (Allen et al. 1998); also the couples proximate/ultimate causes in evolution (Mayr 1982), and downward/upward causation (Campbell 1974) stir countless debates in literature and congresses.

The possibility of uncaused events is at the core of the concepts of chance and probability: the role of chance in biology is often underlined, be it in terms of chance and necessity (Monod 1970), order from noise (Foerster 1960) or complexity from noise (Atlan 1979). However, equally often biologists underestimate the philosophical troubles raised by the concept of chance (Boniolo 2003).

Last but not least, the very concept of life has important links with ontology (Ramellini 2006a). Some biologists have asked themselves why the definition of life, though sometimes considered the central problem in theoretical biology (Rizzotti 1996), is so scarcely palatable to their colleagues. Unquestionably, a part of the answer is that life is not a truly biological concept, being on the contrary a scientific-ontological question: as Bunge correctly put it, regarding concepts like life or time, ‘science just borrows them leaving them in an intuitive or presystematic state. Only ontology is interested in explicating and systematizing concepts which, since they are used by many sciences, are claimed by none. For example, physics asks not “What is time?”, biology “What is life?”, psychology “What is the mental?”, and sociology “What is sociality?”. It is the task of ontology, jointly with the foundations of science, to try and supply answers to such questions and, in general, to clarify whatever idea science takes for granted or leaves in the twilight’ (Bunge 1977, p. 20).

As it can easily be seen, biology (as indeed any other science, i.e. – if you want – any other regional ontology) offers a rich crop of questions in material and formal ontology, many of which closely related to each other, and reciprocally overlapping.

Being impossible to even survey all these questions, I shall now focus on the subject of the boundary of a living body.

8.3 A Case Study: Biological Boundaries

The topic of boundaries has always raised some interest in both ontology (see Varzi 2004 and the references therein) and biology. The difference is that biologists have rarely tackled this question theoretically, so to the best of my knowledge the following survey is a pioneer work, which explains its conjectural and provisional character.

To begin with, if we want to address the question of the boundary of a living body, we previously need to define the term ‘living body’, a notoriously puzzling topic. To cut a long story short, I shall simply suggest here a definition developed elsewhere (Ramellini 2006a): a living body is a macroscopic body possessing a canalizing capacity largely determined by those carbon polymers which largely compose it (above all, its proteins and its deoxyribonucleic acids, whose sequences almost completely determine the sequences of its proteins). By canalizing capacity I mean the capacity of a macroscopic body to canalize an exchange (with its surrounding environment) and a largely assimilative-dissimilative replacement (inside it) of material particles such that the body maintains itself (with its canalizing capacity). The life of a living body is its possessing the canalizing capacity, while its death is the irreversible cessation of its life.

So, we are interested in the boundary of a (concrete) body which is macroscopic, i.e. showing a conduct largely following the laws of classical physics. I claim that the vast majority of biologists, when thinking of, speaking about and interacting with the living, refer to living (macroscopic and concrete) bodies. And when they tackle the question of the boundary of such bodies, biologists refer to different types of boundary, namely: perceptual (above all visual ones), compositional (above all at molecular levels), epithelial (both as epidermises and mucous membranes), cellular (above all biomembranes) and processual (in their various versions) boundaries.⁵ As often happens, also in biology the question of boundaries is strongly linked to the psychology and philosophy of perception (see e.g. Gibson 1950), to problems of spatial properties and deformations (i.e. to topology; see e.g. von Foerster 1982 and Edelman 1988 for biological development; Stroll 1988 for surfaces) and to part-whole relations (i.e. to mereology; see e.g. Woodger 1937 for biological applications, with an appendix by Alfred Tarsky; Simons 1987 and Casati and Varzi 1999 for a general framework); this obviously implies that the different viewpoints are to be carefully distinguished, without confusing ontological with biological or mereological arguments.

8.3.1 Perceptual Boundaries

The term ‘perceptual boundary’ mainly refers to visual and haptic, i.e. tactile or better somatosensory, boundaries (Ramellini 2002).

Though many biologists and, indeed, laymen, found their naive notion of boundary on optical surfaces and visual perception, rarely this preconception is explicitly dealt with.⁶ Apart from experimental studies on mimicry, camouflage and vexillary functions, only here and there do we find in literature theoretical hints on visible surfaces; for instance, according to Portmann (1965), an animal’s surface – when opaque – becomes a new organ, largely independent from inner structures: not only a protective envelope or a means of exchange between body and environment, but a very window to launch messages and establish relationships.

Let us see what such a visual surface could consist of. If you look, at some distance and from various slants, at a naked man floodlighted in clean air, you will see a variously coloured surface, undergoing changes according to his moving (e.g. breathing) and being moved (e.g. by wind). Now, if you get closer to him, some problems will arise:

• some shallow parts are nearly transparent⁷: through his nails you will see the derma lying below them, with its blood vessels, though under slanting light you will perceive their shallow, light-reflecting surface; again, being his corneas and eye lenses also transparent (due to the peculiar arrangement of their collagen and crystallin polymers), an optical illusion in the form of two black discs, i.e. the pupils, will appear⁸; however, eye surfaces shine when conjunctives reflect light;

• many shady wrinkles and pits make it difficult to locate the visual boundary: nostrils look as a dark zone not further distinguishable, while at closer inspection hair turns into the surface of each single hair. In general, where there are orifices like the mouth, the visual boundary seems located at the rima separating an «external» horny epithelium and an «internal» mucous membrane (except in cases like the hard palate, which is lined by horny epithelium, or the glans penis and the inside of the prepuce, which are covered by a mucosa). Needless to say, these mucocutaneous rimae can be treated of as parts of their own, with their peculiar properties.⁹

From a physical viewpoint, the visual perception of this boundary is linked to complex interactions among light photons, the air as a medium, the chemical components of the living body and the eye of the observer, with all the theoretical problems associated with each of these items.

Passing to somatic senses, if a woman moves her right hand towards the rest of her body, sooner or later the hand will bang into it, meeting with her haptic boundary. It is a wavy surface, changing through time; if something presses it, the surface broadens or reduces in various ways. Here also some problems arise:

• following the haptic boundary, she will arrive at her mouth, where the boundary goes on with the oral mucosa and, beyond it, the digestive and respiratory mucous epithelia, as well as the ducts of the glands associated with them; and indeed, a probe performing such a stochastic walk could even find itself again on the visual surface, for instance passing from the skin on her face, through the mouth cavity, the pharynx and the nostrils, again to the skin on her face;

• here and there one meets liquid or moist materials, like mucus or gastric fluids, so the question arises as whether the haptic boundary coincides with these materials or with the epithelium beneath them; besides, some touchable surfaces pertain to bodies which are in some sense detached from the remainder of the body, like in the case of dandruff, sebum or single hairs, so we must ask ourselves if the haptic boundary includes each dandruff flake’s surface, or passes beneath them.

Here again, the haptic boundary results from complex interactions between an haptic sensor and the materials the living body consists of (and is surrounded by: think of a living body so strictly set against other bodies, be them living – as in the case of colonies – or not – as for endolithic organisms living inside stones, that the probe cannot touch it without having touched and even penetrated the others).

In general, perceptual boundaries may also be grasped through instruments. For instance, if we look at a human body with the help of an infrared viewer (i.e. one which transduces invisible inputs into visible outputs), a strange surface will appear, vanishing as it is where the body surface temperature equals that of its background; and, if our binoculars were to detect only neutrinos, that body would not appear at all, being it perfectly transparent to such particles. Another example is given by the images provided by schlieren systems, which make the layer of warm air around human skin visible, thus enlarging the boundary beyond its visible limits, and giving to it a peculiar character of incessant trembling.

8.3.2 Compositional Boundaries

Let us now consider a probe travelling through space while recording the chemical composition of what it meets. If the probe travels through a room where our naked man is, it will detect:

• first of all, a mixture of gasses, i.e. the air;

• at a certain point, it will register a sudden compositional change, finding keratin, sebum or lactic acid, i.e. the epidermis, in the place of nitrogen or argon;

• then, a complex mixture of water, ions, molecules and polymers will follow, i.e. human cells but also intercellular matrices like blood plasma;

• finally, air will come back, when the probe leaves the man.

After a stochastic walk through the room, we shall be able to resume the probe’s records by saying that there is a body plunged into air, whose boundary corresponds to the surface of maximum compositional change.

Yet, we must remember that:

• composition may be related either to elements (Ar, C, etc.) or to pure substances (Ar, but also O₂, H₂O, Na⁺, while it is controversial whether polymers are pure substances) or to different kinds of material particles (atoms, ions, micromolecules, polymers; sure, even organelles or hairs are material entities, but let us now focus on chemistry), and so on;

• the degree of resolution must be declared: the composition per cube decimetre is a rather different matter from that per cube nanometre;

• a choice is to be made between qualitative (e.g. a list of the elements present), quantitative (e.g. the number of atoms for each element, or the volume occupied by each element’s atoms) or qualiquantitative composition.

If we were to ignore such questions, curious outcomes would result. For instance, let a probe enter a room where a man is sitting in damp air and smoking, and let the probe record the qualitative composition per cube millimetre in kinds of material particles of circamolecular size. Then, its result will be that the room is homogeneously filled up with atoms, molecules and polymers: in fact, each cube millimetre of the air surrounding the man contains atoms (e.g. of argon), molecules (e.g. of water vapour) and polymers (the smoke’s soot); but quite similarly, also each cube millimetre of the human body contains atoms (e.g. sodium cations or argon atoms), molecules (e.g. of liquid water or water vapour) and polymers (e.g. proteins in his liver, or the smoke’s soot in his lungs).

So, countless compositional boundaries result from simply combining different types of composition and degrees of resolution, while in some cases no boundary will be detected by the chemical probe.

8.3.3 Epithelial Boundaries

One is tempted to merge the preceding observations by saying that, all considered and some minor inconsistencies apart, visual, haptic and compositional boundaries coincide with an epithelium lining both the surfaces exposed to an «outer» medium and those of «inner» body cavities¹⁰; with which we enter the field of biological boundaries proper.

A first problem is that all unicellulars and many multicellulars lack tissues, hence epithelia; as to unicellulars, in a moment we shall examine cellular boundaries, while in cases like multicellular fungi, either cells are directly exposed to their surrounding environment, or pseudotissues are built.

Then, epithelia are often provided with discontinuities, like orifices or even open wounds. The case of leaves is particularly challenging: here, the cuticle of the leaf is interrupted by small openings called stomata, whose rimae open on intercellular spaces filled with air (an obvious adaptation to gas exchanges during photosynthesis); so, a spongy tissue results, with cells and many air spaces among them; in other words, differently from e.g. lung epithelia, leaves do not possess a continuous sheet of epithelial cells.

Either through orifices or vesicles, epithelia perform an often intense trafficking of matter with their surrounding environment; for instance, many glandular epithelia let out whole cells or cellular parts as their secretions, raising the problem of where and when these detached parts cross the epithelial boundary.

Finally, not in all cases does an epithelium constitute a significant (functional) boundary: plant roots are composed of a central cylinder of tissues, surrounded by a monolayer of cells called endodermis, and by a peripheral cortex, surrounded by the root epidermis. Now, the endodermis separating the cylinder from the cortex shows no intercellular gaps, since its cells are strictly in contact with one another: in fact, strips of suberin (i.e., roughly speaking, cork) stick together the endodermal cells, constituting an effective barrier against uncontrolled or undesired throughputs of substances through roots; this is why, while the peripheral cortex may contain toxic chemicals, infectious bacteria or mutualistic symbionts, the central cylinder is maintained clean and sterile. Thus, the endodermis seems a stronger (functional) boundary then the root epidermis.

8.3.4 Cellular Boundaries

To account for cases where tissues are lacking, one could resort to cells and their boundaries, that is – prima facie – to cell biomembranes, i.e. plasmalemmas.

This position is held for instance by Mahner and Bunge, who write that ‘[a]lthough every system has a more or less definite boundary separating it from its environment …, the boundary of living systems is peculiar in that it ultimately involves a biomembrane – even if it is overlain by a cellulose wall, a horn or wax layer, a shell, or what have you. As this comparatively sharp boundary restricts the exchange of substances with the environment, biosystems are semi-open systems, although they are usually said to be open systems. In general, a semi-open system is a system which has a boundary that restricts the class of exchanges between the components of the system and the items in its environment. This is why biosystems interact selectively with environmental items’ (1997, p. 143).

Let us take a closer look at the suggestion that the ultimate boundary of a living body is a biomembrane: namely, a biomembrane boundary. Obviously, here we are thinking of plasmalemmas, which however do not exhaust the vast array of biomembranes, leaving for instance apart all intracellular biomembranes.

If we consider a unicellular, the theoretical problems posed by its plasmalemma are more or less the same as for a multicellular in respect to its epithelia: for instance, a cell may produce and let out vesicles (a process called exocytosis), or it may let in environmental materials again through vesicles (endocytosis); or there are discontinuities in the plasmalemma, like temporary pores or stable protein channels. Particularly challenging is the envelope of some bacteria, consisting of an «inner» biomembrane, a middle space containing a thin cell wall, and a second, «outer» biomembrane; here, one could ask Mahner and Bunge what the ultimate biomembrane boundary is.

As to multicellulars, let us consider one more time an adult living body B belonging to Homo sapiens, developed from a human fertilized egg, i.e. a zygote Z. Let us call human eczygotic cell (EC) every cell of Homo sapiens derived from Z ¹¹: that is, for the sake of discussion let us exclude all cells of other species (like the bacteria living in symbiosis in the gut) and all cells of Homo sapiens coming from other humans (like the cells of a foetus living in B’s womb, or the white blood cells implanted into B by blood-transfusion). An EC’s plasmalemma, or at least a part of it, may be either in contact with another EC’s plasmalemma (like for two epithelial cells in contact), or free (like for a white blood cell inside blood plasma).

We have at least the following possibilities:

• 1. A first possibility is to locate the biomembrane boundary in correspondence with the surface resulting from all free EC’s plasmalemmas.

  According to this option, B’s boundary proves to be extremely fragmented, given the huge number of environmental prolongations and exclaves, as well of body enclaves. In fact, let us follow the record of a biomembrane probe: starting to move from the air surrounding B, in order to meet the cellular boundary, the probe must first pass through B’s epidermal keratinized layers (which are made of cell corpses), to reach the first epithelial (living) cells with their free plasmalemmas; then it will meet a mass of (living) cells with their plasmalemmas in close contact, often reinforced by intercellular junctions; so, the probe will travel through the body, remaining inside its cellular boundary, until it will meet other epithelial cells, whose free plasmalemmas constitute another part of B’s boundary (either because the probe will have completely crossed the body, emerging from the skin on the other side, or because it will still be «inside» B, but where a mucous membrane lines a body cavity).

  Now, this is only a rough account of the situation, since on closer inspection the biomembrane boundary appears to be much more fragmented. For instance, wherever there is a connective tissue, its cells are more or less free inside a matrix (like blood plasma or cartilage matrix), hence their plasmalemmas are free and constitute part of the biomembrane boundary: so, in a blood vessel the biomembrane boundary is located in correspondence with the free plasmalemmas of its endothelium (i.e. the wall of the vessel, directly in contact with blood), blood plasma being outside such boundary; but plunged into the plasma there are countless blood cells, which are ECs and whose countless free plasmalemmas are parts of B’s boundary … So, not only the lumina of B’s cavities, but even all B’s intercellular spaces and mediums constitute countless environmental prolongations or exclaves that fragment to excess B’s biomembrane boundary.¹²

  Another problem is presented by those body parts which actively leave, or are passively detached from, B: what about the plasmalemma of a B’s spermatozoon, when it is ejaculated outside B: does it become an exclave of B? And if a blood droplet falls to the ground from a wound, do its white blood cells’ plasmalemmas still constitute a part of B’s boundary?¹³

• 2. A second possibility is to consider, as parts of the cellular boundary, also all the remnants of plasmalemmas (i.e. all the limiting biomembranes of the corpses of what had been ECs): in such case, among the plasmalemmas which make boundary we will also find the free plasmalemma remnants of cell remnants (e.g. cell corpses and apoptotic bodies, see below); so, the boundary will be located in correspondence with the epidermal surface and its cell remnants (horny epithelium, hairs etc.). However, many enclaves would still persist, now also surrounding cell remnants like blood platelets; besides, the question of the detached parts would become complicated, since the detachment of «dead» parts (dandruff, hairs, etc.) is more frequent than for living parts.

• 3. Let us then also add all the derivatives of ECs: thus, the «cellular» boundary will shift to encompass all cells and/or their remnants and/or their derivatives: gut fluids, glandular secretions like milk, sweat or insulin, intercellular matrices, gasses (either those about to be exhaled or those otherwise produced by the body, like in the swim bladder of fishes) and extracellular annexes of some organs (like the gelatinous matrix in the inner ear, with the little «stones» or otoliths it contains).

  Also in this case, some problems will follow, like the fact that usually gas masses do not show spatial boundaries.

• 4. Finally, we could extend the «cellular» boundary to include not only possible non-eczygotic human cells but, above all, all nonhuman (and a fortiori non-eczygotic) cells, and/or their remnants, and/or their derivatives. Actually, we must not forget that human cells, in a human body, are in number a very small minority of all the cells there present: on a total number of about 10¹³ human cells, even in perfectly healthy humans we find about 10¹⁴ bacteria, just to leave apart the members of hundreds of species of protists (e.g. gut protozoa), fungi (e.g. spores), plants (e.g. pollen grains) and animals (e.g. the tiny mites happily living in our eyebrows), be them passively or actively, temporarily or permanently installed in our «human» bodies (Wilson 2005, De Rossi 2006).

  But in such case, could we still consider such an expanded boundary, which more or less coincides with the optic-haptic boundary, as the boundary of a human body?

8.3.5 Sensu Lato Processual Boundaries

Maybe as an attempt to circumvent all these difficulties, many biologists appeal to various types of sensu lato processual boundaries, often called processual proper, dynamic, functional or operational boundaries.

According to Foucault (1966), it is quite the conceptual shift from the visible boundaries of plants and animals to their organic unity of processes and functions to mark – with Cuvier – the epistemic breakthrough from natural history to biology at the end of the eighteenth century: the object of natural history ‘is given by surfaces and lines, not by functioning or invisible tissues. The plant and the animal are to be seen less in their organic unity than through the visible carving of their organs’ (1966, p. 149, my transl.); on the contrary, Cuvier opens the doors to biology when he ‘gives large prominence to functions in respect to organs, and subjugates the disposition of the organ to the sovereignty of the function’ (ivi: 276, my transl.). In sum, ‘during the classical age life was standing upon an ontology which concerned the same way all material beings, subjugated to extension, weight, movement; … from Cuvier onwards, the living escapes, at least at first sight, the general laws of the extended being’ (ivi: 286, my transl., emphasis added).¹⁴

After Cuvier, processual boundaries have been proposed repeatedly, either as a negation of static-spatial boundaries (e.g. Piaget 1967, p. 63, Bertalanffy 1968, p. 215) or as a positive claim (e.g. Maturana and Varela 1980, pp. 81, 90–91, Kauffman 1995, p. 62).

There is probably a strict link between the argument for processual boundaries and the concept of living system (as opposed to the concept of living body): in fact, contrarily to a body, a system may be composed of spatially scattered components, in which case it would be difficult to speak of a topologically continuous boundary (much like in the case of the boundary rigorously established on plasmalemmas). In this respect, Wimsatt (1976) has even claimed that evolutionary increases of complexity are characterized by ‘a trend away from 1 to 1 mappings between functions and recognizable physical objects’, whence a ‘failure of functional systems to correspond to well-delineated and spatially compact physical systems’ (1976, p. 185).

Again, processual boundaries may not coincide with other boundaries, like the cellular ones. For instance, the propagation of action potentials in a neuron depends, among other items, on the salt solution surrounding its plasmalemma; thus, a «processual» neuron will be composed of the «cellular» neuron and the muff of aqueous solution containing the ions co-responsible for impulse propagation; that is, a neuron’s cellular boundary will be located in correspondence with its plasmalemma, while its processual boundary will correspond to that layer of solution around its plasmalemma where certain ionic concentrations begin to conform to a particular chemical equilibrium called the Gibbs-Donnan equilibrium.

To see which problems emerge from the processual approach, let us scrutiny Ageno’s position, by which the boundaries of a coherent system are the boundaries of the coherence of its inner processes (1986, p. 407), so that coherent systems are delimited by the extension of their coherent processes in space and time (Ageno 1992, p. 143). Here Ageno makes reference to his distinction between bound and coherent systems: while a bound system (like an atom, a stone or a cluster of galaxies) is kept united by the attractive forces among its parts, a coherent system (like a Bénard cell, a tornado or a man) owes this to a reserve of internal energy granting the coherence of its molecular movements.

Apart from the difficulty of understanding what exactly Ageno means by coherence (see Ramellini 2006a), here the main problem relates to the very concepts of boundary of a coherence, or boundary of a process. Briefly, while we can surely speak of the temporal boundaries of a process, its spatial boundary cannot be but the boundary of the body (or bodily or concrete system) undergoing that process; consequently, what Ageno holds is that the boundary of a coherent system S is the boundary of that (concrete) system S whose processes are coherent, which is a true but rather unfruitful assertion.

In other words, despite the appeal that various process ontologies have on biologists, it seems to me that only an ontology based on the bearers of processes can satisfy their needs, be them experimental or theoretical (Mahner and Bunge 1997, pp. 20–21).

8.3.6 The Organismic Boundary

Despairing of the prospects to find out the boundary of a living body, some authors have declared that simply it does not exist, or that it is at most a fiat rather than a bona fide boundary.¹⁵

For instance, Haldane (1931) wrote that the organism and the external environment are so intimately entangled that ‘[t]here is no spatial limit to the life of an organism’ (1931, p. 74). A peculiar version of this position holds that all what a living system needs to live or develop is part of it, hence inside its boundary: if even a mathematician like Thom (1998) claimed that a prey, though outside the predator, is an integrating part of its vital dynamic totality, so that the border between predator and environment is rather fluid (1998, p. 279), it is undoubtedly the Developmental Systems Theory (DST) to have stressed that the bearer of development is not an organism, but a developmental system encompassing ‘not just genomes with cellular structures and processes, but intra- and interorganismic relations, including relations with members of other species and interactions with the inanimate surround as well’ (Oyama 1985, p. 123).

Others authors have however (correctly) criticized these positions. For instance, Needham criticized Haldane saying that ‘if no line can be drawn between organism and immediate surroundings, no better line can be drawn between immediate surroundings and far-off surroundings’, so we can but contemplate the whole universe, the ‘analysis of living things being laid aside’ (Needham 1936, p. 11); and exactly the same comment has been advanced about DST by Mahner and Bunge, when they claim that expanding ‘further and further nested developmental systems would lead us directly to holism, that is, to the assumption that the entire universe is the developmental system’ (Mahner and Bunge 1997, p. 301).

On the opposite side, some have maintained that boundaries are at the very core of life. Jonas (1966), for instance, claims that a most deep characteristic of life is ‘its being self-centered individuality, being for itself and in contraposition to all the rest of the world, with an essential boundary dividing “inside” and “outside” – notwithstanding, nay, on the very basis of the actual exchange’ between organism and environment (1966, p. 79); Hoffmeyer (1998) says that life is organized around those nested sets of membranes we call organisms; Keller (2001) writes that boundaries like cell membranes are ‘a cornerstone of biological organization … with absolutely vital significance’ (2001, p. 301).

From my ontological and biological viewpoint, the boundary of a living body cannot be but the boundary of a (concrete) body, and of a body qua living.

To be more precise, let us first draw a distinction between a living body and an organism, or organismically living body, as set out elsewhere (Ramellini 2006a): an organism is a living body which is biologically subordinated to itself and only to itself, since it possesses the capacity to biologically regulate itself and only itself, through its concrete parts regulating both themselves and each other. Thus, a bacterium in a test-tube is both a living body and an organism; a human white blood cell in a test-tube is a living body but not an organism; a human white blood cell in a human body is neither a living (free) body nor an organism, being rather a living (concrete) part of an organism.

Now, how to decide which (concrete) parts of a living body do constitute together an organism? The methodological way is: take a living body L; take a concrete part P of L, and look for the parts P subordinates to itself (regulates) and for the parts P is subordinated to (by which P is regulated); repeat the same inspection for these last subordinated-subordinating (regulated-regulating) parts, until you will obtain a closed network of inter-subordinating (inter-regulating) parts: this will be your organism O. ¹⁶ For instance, if we consider a human hepatocyte P inside a «human body» L (actually, and more precisely, a human symbiotum), we shall discover that P subordinates – and is subordinated to – the other hepatocytes; then we shall discover that these hepatocytes subordinate – and are subordinated to – the lymphocytes, and so on, until we shall arrive to a closed system of reciprocally subordinating parts, which neither subordinate other parts of L (for instance, gut bacteria) nor are subordinated to them; this closed system will be the human organism O inside the human symbiotum L.

As it is clear, the organism O will sometimes coincide with the entire living body L from which one started, while sometimes O will be smaller than L; that is, some organisms are living bodies (like the bacterium in a test-tube), while others are living parts of living bodies (like the same bacterium inserted into the intestine of a man, or the human organism inside a «human body»).

Now, let us focus on the organismic boundary of an organism O, i.e. of a body qua organismically living.¹⁷ Let us ask ourselves whether a quantity of matter M (be it an atom or a macroscopic system, a quantity of gas or a solid, and so on) is a (concrete) part of O, hence inside its organismic boundary.

A first, provisional condition for M to be a part of O follows from the fact that every part of O is, at the same time, a part of L (while the converse not always holds). Thus M, in order to be a part of O, must simultaneously be a part of L; that is, M must be not only physically close to, or in contact with, O, but also «attached» to it, in the sense of co-moving with O. ¹⁸

For instance, a beech leaf M, fallen down to the stump of the beech from which it had budded, is not a part of that beech qua body, hence a fortiori it is neither a part of that tree qua living body nor qua organismically living body (though it may contribute, when on the ground at that beech’s stump and with the rest of its litter, to protect its roots from winter cold, thus performing some «function» for that beech). But what about the same leaf M when, though still on that beech, it is about to fall to its stump? Or what about a leaf that happens to have fallen into a hole in a branch of that beech and now is rotting there?

Another example is provided by cobwebs. Once completed, a cobweb is not a part of the spider that has woven it (pace Diderot 1769), since it is not attached to it. But what about a cobweb when it is still attached to a spider’s abdomen during its weaving? And what about those single cobweb threads, attached to the abdomen of «flying» spiders in order to make them soar thanks to air currents (sometimes as far as hundreds of kilometres)?

These considerations show that M’s being attached to O is not a sufficient condition for M to be inside O’s organismic boundary: inter-subordination (inter-regulation) between M and O’s parts is also required; in other words, the (concrete) parts of an organism are characterised not only by concreteness, but also by inter-subordination (inter-regulation).

For instance, a growing leaf budded by an organismic tree is inter-subordinated with the rest of that tree, thus it is an organismic part inside its organismic boundary (or better, the deep part of that leaf is inside the organismic boundary, while its shallow part is at the same time a part of the organismic boundary). In contrast, the petiole of a leaf about to fall usually contains a layer of substances which occlude its vessels; if so, this implies that the leaf has lost inter-subordination with the rest of the tree, then the fact that it is still attached is simply a mechanical matter, with no significance as regards its biological and ontological status of being a part of the relevant organism inside its organismic boundary. In other words, a tree in autumn carrying the last leaf M about to fall is a living body L, composed of M and of the rest R of L: if R is an organism O (i.e. if all R’s parts are inter-subordinated), then its organismic boundary will exclude M; that is, M will be inside L’s boundary but outside O’s organismic boundary.

As to cobwebs, I am prepared to consider the thread of a «flying» spider as a part of the relevant organism, while the cobweb still attached to a spider’s abdomen during its weaving seems to me insufficiently inter-subordinated with the spider’s parts as to be inside the relevant organismic boundary.

In other words, an M can enter or leave O’s organismic boundary by gaining or losing inter-subordination (inter-regulation) with O’s parts.

For instance, a sand grain on the cuticle of a Palaemon shrimp is not inside its organismic boundary, but some sand grains cross the boundary when after each moult the shrimp introduces them into its newly formed statocysts, using them as statoliths¹⁹; another illustration would be the stones that poultry ingest (as well as dinosaurs did) to aid food grinding in the gizzard, and obviously a bite of food itself. Now, when does a mouthful of bread cross the organismic boundary? When it is attacked by the enzymes inside the mouth, or when it becomes a part of chyme and then of chyle, or when glucose – the end product of its digestion – crosses the intestinal mucosa? The answer is: when mouthfuls, or better the products of their gradual digestion, become inter-subordinated with the parts of the organism, i.e. more or less when they are absorbed by intestinal villi.

To look at yet another illustration, when it is inside blood plasma, a single particle of uric acid (a catabolic substance of humans deriving from the oxidization of certain organic molecules) contributes to plasma osmotic pressure, and may have a protective role as an antioxidant²⁰; in the meantime, it is subject to the general regulation of osmotic pressure. Thus, we can admit that such a particle is inter-subordinated with O’s parts, thus being a part of O inside its organismic boundary; now, the fact is that at high concentration uric acid becomes a toxic waste, to be eliminated by kidneys. Once expelled into the urine, but before urination, a particle of uric acid is no longer inter-subordinated with O’s parts, though it is still attached to O. Hence, uric acid particles cross O’s organismic boundary when they cross the free plasmalemmas of certain kidney cells called podocytes, i.e. where they lose inter-subordination with O’s parts.

The topological catastrophes of organismic boundary generation, merging and corruption are currency in biology. I shall set aside here well known examples such as cell division or fertilization, to tackle other biological phenomena that may be more intriguing to the ontologist.

During human early development, for causal reasons that are not yet well understood, it may happen that an embryo split into two embryos, generating two monozygotic twins; before the division, there is one organism possessing one organismic boundary, while afterwards there are two organisms bounded by two organismic boundaries: in fact, though the cells of the two twins are ECs derived from the same zygote, being there no metabolism at a distance, the two twins are not inter-subordinated, hence they constitute two organisms. The fact is, however, that rarely the division is not complete, so that the embryo splits only partially, and Siamese twins result: how to consider them? Undoubtedly, the twins do constitute a single living body, but how many organisms are there? Again, it all depends on inter-subordination: if the bodily conjunction between the twins is so little that it implies no inter-subordination,²¹ then there will be one living body and two organisms, hence two organismic boundaries. On the other hand, if conjunction is so extensive as to imply inter-subordination,²² then there will be one living body and one organism (usually doomed to early death), hence one organismic boundary.

An extraordinary case of organismic boundary merging occurs in the parasitic worms belonging to Diplozoon paradoxum. Hatching from eggs, their larvae settle on fish gills as ectoparasites; when two larvae meet on the same fish, they fuse their bodies and become sexually mature adults: in fact, their hermaphrodite reproductive apparatuses develop, in such a way that the testicle ducts of the first open onto the oviduct of the second, and vice versa; the intestine too branches out in both partners. From such process onwards, the «two» partners remain permanently fused, and probably neither could survive alone (supposing it could free itself from its partner). The partners assume an X shape, since body fusion involves only their middle regions: so, there are two heads (with two mouths) and two tails; reproductive organs are doubled as well, with their ducts intercommunicating. In this case, inter-subordination seems so deep that this «diplozoon» (from the Greek: double animal) can be considered one living body and one organism, with one organismic boundary²³; to realize the difference, the fish carrying it, though attached to the parasite, is undoubtedly a distinct organism, with its distinct organismic boundary.

It is easy to assess that inter-subordination (inter-regulation) is a matter of degree, thus it is not at all easy to recognize the threshold between two organisms A and B which are merely «attached» to each other, and two organisms A and B which share so much of their life processes as to generate one (new) organism C by fusion (while A and B themselves die qua organisms). However, it is clear that neither part contiguity nor body continuity are sufficient conditions for one living body to constitute one organism; rather, inter-subordination is absolutely necessary.

That is, we know many cases where a physical link is established or maintained between two (or more) living bodies, without implying organismic continuity and organismic boundary merging: apart from the above mentioned examples, a bacterial cell conjugating through the channel called pilus with its partner does not combine with the latter sufficiently tightly as to constitute one organism; in mice, the so-called polar cell is not inside the zygote’s organismic boundary, despite its being attached to the zygote by a thin extensible tether; a slender strawberry stolon sooner or later produces an independent offshoot, which constitutes a new organism, though still attached to its mother plant.

Organismic boundary corruption occurs whenever an organism dies. Defining organismic life as the possession of the regulatory capacity by an organism, we receive the result that organismic death is the irreversible cessation of its organismic life (Ramellini 2006a). So, a bacterial cell infected by viruses appropriately called bacteriophages starts to synthesize viral components which self-assemble and build new bacteriophages; at this point, the bacterium usually dies, its plasmalemma disintegrates, and the cell content is released into the surroundings (bacterial lysis); such events obviously imply the corruption of the bacterial organismic boundary. It is worth noting that not always the corruption of the organismic boundary of an organismically living body (qua organism) coincides with the corruption of the boundary of that body (qua body): for instance, during apoptosis or «cell suicide», when the cell dies its plasmalemma forms spheroidal protrusions (blebbing), so that the cell body fragments into small spheroids (apoptotic bodies), each of them bounded by a part of what had been the cell plasmalemma; that is, the organismic boundary generates, on its corruption, the body boundaries of numerous apoptotic bodies. In other words, according to the motto corruptio unius generatio alterius, plasmalemma corruption leads either to plasmalemma debris or to the generation of biomembranes of other kinds.

And now we can pause, to sum up our considerations. The relevant boundary of an organism is neither the boundary of that organism qua percept, nor qua mixture of chemicals, nor qua biological body bounded by biological envelopes, nor qua bearer of processes, but qua an organismically living body: the organismic boundary is the boundary of an organism qua organism.

So, I stipulate the following explicit intensional definition by genus and difference:

organismic boundary of the organism O =_df (concrete) part of O which spatially encompasses all and only the other (concrete) parts of O

The boundary here addressed is the boundary of the organism, and not a boundary for it; that is, we are speaking of the maximal boundary of the organism, i.e. the sum of all the boundaries for it (Smith and Varzi 2000). The central point in this definition is that the organismic boundary of an organism is, by definition, an organismic part of that (and only that) organism, i.e. a part inter-subordinated (inter-regulated) with the other organismic parts of that (and only that) organism.

Obviously, it must not be expected that the organismic boundary always be a spatially «simple» part: it may well be very indented, or scattered into disjoint sub-boundaries, and so on. That is, while living bodies and organisms, modelled as topological spaces, are connected,²⁴ organismic boundaries, modelled as topological spaces, are almost always disjoint unions.

The suggested definition squares with the following ontological positions about boundaries in general:

• that the organismic boundary is a concrete part (rather than, e.g., a lower-dimensional geometrical surface);

• that the organismic boundary is a part (only) of the organism (rather than, e.g., of both the organism and the environment);

• that the organismic boundary involves a closed/open dichotomy (rather than, e.g. a closed/closed one), i.e. that, modelled as topological spaces, the organism is closed while its environment is open (obviously, the environment is open where it faces the organismic boundary; nothing is said about possible environmental closed boundaries elsewhere);

• that both the organism and its organismic boundary are ontologically dependent upon (though chronologically coextensive with) certain interactions between bodies. That is, it is when some bodies start to interact in a certain way that an organism is generated with its organismic boundary; equally so, it is the irreversible cessation of certain interactions between the organismic parts to constitute the death of an organism and simultaneously the corruption of its organismic boundary;

• that the organismic boundary is ontologically dependent upon (though chronologically coextensive with) the organism.²⁵ That is, ontologically, first there is the organism; second, the organism is composed by organismic parts; third, among these parts there is the organismic boundary.

I am perfectly aware that these ontological positions are not unproblematic, but, while being ontologically no more problematic than others, they appear biologically more sound to the biologist I am.

However, my definition leaves open the question about the spatial extension of the organismic boundary, whether it is an extremely thin layer or a thicker bed of the organism; however, it is reasonable to claim that the organismic boundary have a «thickness» considerably inferior to that of the organism itself: after all, though the organismic boundary is an important part and performs important functions of the organism, it is neither the only part of the organism, nor one which performs most of the organismic functions; so, it would be biologically implausible to think of an organismic boundary constituting the most voluminous part of the organism itself.