Check out my review of Ken Wilber's latest book Finding Radical Wholeness

Integral World: Exploring Theories of Everything
An independent forum for a critical discussion of the integral philosophy of Ken Wilber

Andy SmithAndrew P. Smith, who has a background in molecular biology, neuroscience and pharmacology, is author of e-books Worlds within Worlds and the novel Noosphere II, which are both available online. He has recently self-published "The Dimensions of Experience: A Natural History of Consciousness" (Xlibris, 2008).



A Response to Fred Kofman

Andrew P. Smith

Fred Kofman's article "Holons, Heaps and Artifacts" raises some very important issues surrounding our understanding of holons and holarchy. In essence, he addresses the question of what kinds of holons there are, and whether everything is in some sense a holon (as Ken Wilber, perhaps not intending to be taken literally, proclaimed in Sex, Ecology, Spirituality, 1995). I found many of Kofman's ideas stimulating, sometimes provocative, and want to respond to some of them here.

Individual vs. social holons

Closely following Wilber, Kofman draws a distinction between individual and social holons, noting that "These are two fundamentally different types of holons and need to be kept separate to avoid confusion." I agree with this statement, but believe that the subsequent discussion of holons in fact exhibits this confusion. Though this is an argument I have been through before on this site, Kofman provides in some respects a more definitive elaboration of the distinctions between these holons than I have seen in Wilber, allowing me to approach the issue in a somewhat different way. I take as my point of departure the following passages:

An individual holon is a "member" of a social holon, but not a part (junior or constitutive element) of it. The social holon is the relational space, the patterns of organization in which the individual holons find a common affiliation. The social holon is the development (senior) of another social holon (predecessor or junior), not of an individual holon...Social Holons are groups of individual holons that have a patterned mode of interaction. Social holons do not have localized interiority or consciousness; they have inter-subjectivity or non-localized consciousness. Social Holons do not have unified exteriors...

To summarize, Kofman defines a social holon as a group of individual holons that have a) a common affiliation; b) a patterned mode of interaction; c) non-localized consciousness; and d) a non-unified exterior. It should be obvious that properties a) and b) apply to what he considers individual holons as well as social holons. For example, the various atoms and molecules in a cell have a common affiliation--they all belong to the cell--and a patterned mode of interaction--they have certain definable relationships to one another. The same can be said for the cells, tissues and organs within an organism.

I find criterion c) of little use, because it's unverifiable. I don't know whether human societies have a localized consciousness or not. Most scientists and philosophers believe that they don't, but if in fact they do, how would we know? In our ordinary state of consciousness, we of course don't have access to the consciousness of other forms of life, so any statements about this consciousness must be speculative. And more compellingly, no one knows whether molecules, for example, do have a localized consciousness (as Kofman's subsequent discussion implies they should).

This leaves criterion d): social holons do not have unified exteriors. I'm not completely sure what Kofman means by this, but I take it to mean that they don't have definite boundaries in space. An atom, molecule, cell, tissue or organism--all of which are considered by Kofman as examples of individual holons--is usually considered to have a definite boundary, whereas a human society, if I understand him correctly, he thinks does not. But surely it does. Any human society that has ever existed has had some physical boundary, and even the most complex ones existing today have one, if perhaps it is now the entire surface of the earth. Of course, the boundary of a society may be fluid and changeable, but the same can be said for a cell or a tissue, or even a molecule--yes, even a developing organism.

Alternatively, perhaps, Kofman means to say that the component holons of individual holons closely cohere, while those of social holons do not. Thus atoms are physically bonded to each other in molecules, and cells are in close physical contact in organisms, while individuals are not in actual physical contact in societies. But of course sometimes individuals in societes are in physical contact, and sometimes atoms in molecules, and cells in organisms, are not. In the case of atoms, there are several types of bonds of varying strengths and distances, so the whole question of physical contact is a little fuzzy. In the case of cells, there are numerous kinds of interactions that don't involve physical contact--between neurons, e.g., or between hormone-releasing cells and their targets.

More importantly, if Kofman does mean to define individual holons by some kind of coherence, it is a fallacy to apply the notion in a physical sense to human societies, because they exist on the mental level of existence. The kinds of bonds that hold human beings together in societies are obviously not purely physical or even biological, but to a large extent mental; they result in a definite coherence nonetheless. In other words, people are constrained in certain ways by their membership in societies, just as atoms are constrained in molecules, or cells in molecules. To me, at least, the existence of these constraints implies a unified exterior.

In conclusion, I find Kofman's criteria for distinguishing individual from social holons in fact fail to do so. As I said before, though, I do agree with him that the distinction is important, so let me now offer my own criteria, which I believe are much clearer and more precise, and allow us to make these distinctions with much less ambiguity. (These criteria are drawn from my book Worlds within Worlds, and the reader is referred to that for further details. Similar, though less extensively discussed, criteria have been presented in Pettersson 1996). An individual holon (which I also refer to as a fundamental or autonomous holon) differs from a social holon in three primary ways (there are other criteria, but these will suffice):

  1. an individual holon, unlike a social holon, can reproduce itself;
  2. an individual holon, unlike a social holon, can exist autonomously outside of a higher-order holon;
  3. an individual holon, unlike a social holon, can preserve the properties of its component holons.

Let me briefly expand on these definitions, and in the process reveal the classification they imply. Reproduction is a property of only two kinds of holons that we know about--cells and organisms. Though the tissues within an organism, as well as human or animal societies, may in some sense duplicate themselves, this is done through the reproduction of their component fundamental holons. So I classify cells and organisms as individual holons, while almost all other forms of existence--including all kinds of molecules, as well as tissues and organs, in addition to human and animal societies--are social holons. I say almost, because I regard atoms as individual holons as well. Though they don't fulfill the criterion of reproduction, they do satisfy the other two criteria to be presented below.

Second, individual holons can exist outside of higher-order holons. While some atoms exist within molecules, some exist in a free, unbonded state. While cells can exist within organisms, many kinds of cells live independently of organisms. And while organisms can live within societies (which, for reasons I will discuss later, I regard as a higher form of life), they can also exist outside of them. This property of autonomous existence, again, is generally not shared by other kinds of holons. It's true that some very simple molecules, such as water and carbon dioxide (which are barely beyond atoms in the holarchy), can exist outside of cells, but the great majority of types of molecules, including any with even a moderate degree of complexity, do not. The same can be said of tissues, organs, and other intercellular holons found within organisms. These are not found outside of the organism. And while human and animal societies do seem to exist outside of any higher-order holon, I have argued in Worlds that this is because the latter has not yet fully emerged; these social holons are still in the process of evolving.

Finally, individual holons preserve, or can preserve, all the properties of their component holons, while social holons can't. For example, the properties of individual atoms can be completely preserved within a cell. There are found within cells certain atoms, such as sodium ions, that have properties identical to those of the corresponding atoms outside the cell. There are also within cells social holons composed of atoms--for example, small molecules like amino acids, and larger ones like proteins--that also preserve their properties. Though such social holons, as I just pointed out, are not ordinarily found outside of cells, they can be isolated, and they behave under these conditions in basically the same way that they do within the cell. The same argument can be made with respect to organisms, which contain individual cells as well as social holons (tissues and organs of various degrees of complexity). Thus the properties of an individual red blood cell (e.g., oxygen uptake, receptor binding activity) are essentially the same within an organism or isolated outside of it, and so on.

As discussed in length in Worlds, this property in fact is how I define transcendence. A holon can be said to transcend its component holons when it not only includes them, and exhibits properties not found in the components (emergence), but is also capable of preserving the original properties. Thus a cell, in addition to manifesting new properties not found in any of its molecular components, also retains the original properties. An organism, in addition to manifesting properties not found in any of its multicellular holons, also retains them. Another way of putting this is to say that an individual or fundamental holon contains all the social holons on the level of existence below it, in both free and bonded forms. Within any cell are found individual atoms, small molecules (e.g., amino acids), macromolecules (e.g., proteins), macromolecular complexes (e.g., receptors), and so forth. Within any organism are found individual cells, simple cell units, complex cell units, organs, and organ systems.

In contrast, social holons don't preserve the properties of their individual members. When an atom becomes part of a molecule, its properties change. For example, when a hydrogen atom becomes part of an amino acid molecule, joining to a carbon atom, it loses the ability to ionize, or give up an electron. Likewise, when a molecule becomes part of a still higher-order social holon--for example, an amino acid becomes incorporated into a protein--its properties also change. When a cell becomes part of a tissue, its properties change. Human beings who are members of societies have properties different from humans who don't or didn't (certain conceptual abilities, for example).

A social holon thus contains only bonded forms of its component holons. A protein, for example, does not contain free amino acid molecules and free atoms. All the amino acids, and all the atoms, within the protein are bonded together. Generally speaking, such holons lose some properties while gaining new ones. So molecules and tissues, while including atoms and cells, respectively, don't transcend these holons. They transform them, altering their properties. Likewise with the relationship of animal and human societies to their members. This is how I define transformation.

Societies as a higher form of life

While social holons are thus distinguishable from individual holons, it does not follow that they require a different conceptual dimension in the holarchy. In fact, if we accept that molecules and tissues are social holons--and for the reasons I gave above, and others, I think we have to--then it's clear that social holons belong on the same scale or conceptual dimension as fundamental holons. Wilber in fact puts these holons on the same scale or quadrant as cells and organisms. But since human societies are also social holons, they also belong on this scale, though Wilber, and Kofman following him, do not do this. Human (and to a lesser extent, animal) societies, just like molecules and tissues, are organizations of fundamental holons which have properties that the latter do not. Molecules can do things no atom can do (e.g., an amino acid can ionize in two positions, creating a buffer), tissues can do things no cell can do (digest food, pump blood, contract muscles, etc.) and societies can do things no individual can do (from hunting large game to creating science and technology).

Moreover, in all cases, there is a clear line of evolutionary development. Molecules emerged after atoms, tissues after cells, and societies after organisms. This point, among others, I think undercuts the Wilberian argument that societies are just a "social aspect" of human beings. If, as Kofman claims in his retelling of Wilber, societies are just the other side of a coin containing individuals, why did they not emerge in parallel with these individuals?

If the relationship of societies to humans nevertheless does not seem quite the same as molecules to atoms or tissues to cells, it must be remembered that molecules and tissues are fully-evolved holons, while human societies are not. There was presumably a time in our evolutionary history, before the emergence of cells, when the highest form of life was loose societies of molecules, and still later, before the appearance of organisms, when there were loose societies of cells. These incompletely organized holons later evolved into new fundamental holons (cells, organisms) in which these societies became the highly organized molecular and intercellular holons we know of today. I take it as a strong implication of holarchy that human societies are, or could be, in the process of a similar development.

I think a second reason why many people fail to see the deep correspondence between human and animal societies and forms of existence like molecules and tissues is because such societies, existing on a higher level of existence, do have very different dynamics. As I noted earlier, whereas molecules are held together by physical bonds, and tissues by both physical and biological (for example, various forms of signalling) bonds, societies have both these and mental bonds--i.e., thoughts or memes. Since we are situated above molecules and tissues in the holarchy, we can see their bonds fairly objectively, as concrete connections. But since we are identified with our thoughts, it's extremely difficult to appreciate (in a direct, non-theoretical way) that these function very much in the same way and to the same purposes as physical and biological bonds.

However, there is yet another reason why Kofman, and a great many other people (not all of them necessarily Wilber fans), resist the conclusion that "individual, family, community, nation, humanity" constitute a genuine holarchical relationship. It's a reason lurking behind every discussion of hierarchical relationships, and I give Kofman credit for not being bashful about stating it. He worries very much about the political implications of putting societies above individuals in the holarchy:

Dehumanizing consequences also ensue from combining individual and social holons in the same hierarchy...The danger of considering a human being as a junior holon that is transcended and included by the group, is that this opens the door for a centralized control that can rightfully obliterate any individual that does not suit its purpose. Abuse is rampant when the good of the (social) whole totally supersedes the good of the individual (member). But this is exactly the case in a well-behaved holarchy where the good of the (compound individual) whole supersedes the good of the individual (part of element). A person (seen as a whole) can operate on parts of his body (elements) without regard of any alleged "rights" of these constitutive elements. For example, one can extract blood for an analysis (thus killing the blood cells), take an antibiotic (thus killing a good portion of one's intestinal flora) or surgically remove a cancerous tissue.

When one extrapolates this model to a society, one endows the central authority with the power to cut off any of its (mistakenly considered) "constitutive elements" whenever it wants without any regards for their individual rights. The consequences of this confusion are serious. Imagine a pseudo-holarchy of the type: individual, state. According to its implied relationship of "transcend-and-include", the individual is a junior holon of the state. The senior holon, being implicitly more "conscious" has the power to impose its organization on the junior one-just like a person has the power to impose her will on its arm. This provides justification for every type of repressive regime.

I'm not going to say that Kofman is wrong or misguided to worry about this. As soon as you postulate any model in which societies are higher than individuals, you provide some justification for making the latter subservient to the former. But the case of human beings manipulating their own tissues is an unusual one (which I will return to in a moment). The normal relationship of higher holons to lower ones is not one of such dominance as Kofman seems to imply when he says that "the senior holon, being implicitly more 'conscious' has the power to impose its organization on the junior one." Imposition, so to speak, goes both ways. An organism certainly does constrain the behavior of its cells and multicellular holons. But the latter maintain a great deal of autonomy, and in turn constrain the behavior of the organism. I cannot, for example, deprive my cells and tissues of oxygen and nutrients for very long without injuring and maybe ultimately killing myself. In fact, I can't with impunity do much of anything injurious to my components, except perhaps crush some skin cells, or strain my muscles and my lungs during, for example, some intense physical exercise. Generally speaking, as I noted earlier, lower order holons, while giving up some "rights" or properties, also obtain certain new rights or properties in return. Holarchy, in other words, is usually a pretty good deal for all parties involved--that's why it evolved.

So the much more appropriate analogy here is not human beings doing what they will with extracted tissues. The analogy is a normally smoothly-running system gone awry. If Kofman wants to find, in some model of holarchy, the equivalent of a Hitler or a Stalin, I suggest a much better (though hardly perfect) one would be the growth of a cancer. Societies of cells, like societies of humans, can occasionally produce individuals whose behavior is so deviant from the norm that it threatens the health and survival of every other member. A holarchical model that recognizes and accepts the analogies between these two kinds of societies is not only not promoting or encouraging or justifying such deviant behavior. It just may be putting itself in a better position to learn something helpful about it.

But what about the case of human beings performing experiments on their own tissues? How does that fit into this discussion? At the outset, I want to point out that there is a confusion of levels here. When human beings extract body tissues, this is an act by a social holon (human society) on a holon one level below it. It's an act of a social holon, because before the emergence of societies, and all the new properties they brought with them, humans could not perfom such operations. And the act is performed on a holon within the organisms making up that society (humans), which puts it a level of existence lower than both the society and the organism. So strictly speaking, if one wanted to use the holarchical model to justify "eliminating" individuals by society--to argue that this is in some sense natural--one would have to postulate a still higher level of existence beyond humans and their societies. Societies of fundamental holons on this higher level--not societies of humans--would be performing the homocidal or genocidal "experiments" on individual humans, not societies of humans themselves.

I believe such a scenario is conceivable, given the emergence of a higher level of consciousness. I'm not shy in admitting it is an implication--albeit a highly speculative one--of my understanding of holarchy. But this is hardly a reason for abandoning this kind of model (as if any model created by a human being could influence the behavior of beings on a level of existence above humans!). On the contrary, I see it as a reason for paying attention to such a model. The underlying message is: to be a human being is not necessarily to be the highest form of life in the universe. Just as we are composed of atoms and molecules and cells that we regard as trivial, so we ourselves may be part of a still higher-order holon that views us in the same way. This is, in fact, as concise a reason I can give for pursuing higher consciousness, for being eternally unsatisfied with being an ordinary human being. If you really understand the position you are in, you know that it is not a good place to be.

This bring me to the final point. What I find most disturbing about the passage from Kofman quoted above--all the more so as it speaks eloquently on behalf of a very large number of people in our society--is that it clearly, blatantly and unashamedly reveals the author's perspectival bias (what is sometimes called specism, but in the context of this discussion I guess would be better called holonism). As a human being, Kofman quite naturally wants to defend and protect his species. Thus he regards his component holons (cells, tissues, etc.) as having no "rights", while conversely regarding larger societies of which he is a member as illegitimate impostors. As a fellow human being, I sympathize with the feeling, but as someone trying to understand the world, I regard the feeling as an obstacle to that understanding. Just because we want to be the center of the universe doesn't mean we are.

Consider: according to Kofman, all non-artifactual holons, down to atoms and below, are sentient. Thus he says "holons have interiors or some degree of prehension and consciousness". While there really is no evidence for this, and this blanket statement should not be asserted as a fact of nature, I think he's probably right about this. It's my working assumption. But if he is right, then how is our treating our cells and tissues as if they have no rights any different, in an objective sense, from some higher organization of holons treating us in the same way? Because the pain our tissues could feel is presumably much less than ours? Is it simply a matter of degree? Surely it's obvious that the real rationale for Kofman's argument is self-interest, where self means not simply him alone, but his species. In other words, he's fitting his conceptual model not to facts, but to certain goals and desires.

This becomes even more obvious to me when he points out specific examples of what he regards as pathological political systems. Again, I do agree with him--these systems are abhorrent to me, too. But I feel it's very dangerous to use a holarchical model in this way to define what is and is not a justifiable form of human organization. What if someone could actually convince Kofman that human societies are higher than individuals, in the same sense that molecules are higher than atoms, or tissues than cells? Would he then retract his statements about what kinds of societies are approvable? Would he change his political views? If we really want to teach people not to try to justify certain political systems by pointing to certain theoretical models, why not go to the root of the problem, which is not the details of any model, but the attitude that it's permissible at all to draw any conclusions from these details about the proper form of society. Holarchy models, in my view, are an attempt to understand what is actually happening, not what any particular people think should be happening.


A second distinction Kofman makes is that of heaps. He defines heaps "as a random pile of stuff. For example, a pile of sand consists of the grains of sand." Though this definition conveys the general sense of what a heap is, I find it quite imprecise, even misleading. Actually, every grain of sand, itself, should be considered a heap, and a heap does not have to be a random assortment. In fact, most (though not all) heaps are not random assortments of holons, as I will explain in a moment. A more precise definition of what I believe Kofman means by a heap is a collection of individual holons that exhibits no significant emergent properties, that is, properties not exhibited by each member of the collection. Thus a pile of sand, or a pile of rocks, has no properties different from a single grain of sand or a single rock (except for the trivial one of mass), and in fact, a very small unit (consisting of just a few atoms) of the sand grain or the rock retains these same properties.

Heaps have been disregarded in most discussions of holarchy, including my own, and thus Kofman has raised an important issue, namely, just how we are to understand their relationship to individual and social holons. Two aspects of his discussion of them, however, I find quite puzzling. First, he never comments on Wilber's definition of planets as social holons of atoms (Wilber 1995). Most planets (i.e., every one we know of except earth, and possibly, Mars) are, so far as our science can discern, little more than "a pile of rocks", and so would seem to qualify as heaps. I would also argue that another one of Wilber's putative social holons, the world-wide collection of prokaryotes we now call Gaia, is also a heap. While Gaia may seem more like a society than a pile of rocks, it has very weak emergent properties, if any at all. It's basically a very large group of bacteria, which do their individual thing while interacting with one another in minimal ways. Some thinkers, such as Lynn Margulis (Margulis and Sagan 1986) and Howard Bloom (2000) have argued that transfer of DNA among different bacteria has created an information network, but since all this network has apparently accomplished is horizontal (what Wilber calls translational) change--that is, created bacteria with different properties, but no higher forms of life--it does not seem that it qualifies as an emergent property.

This leads to the other puzzling omission in Kofman's discussion of heaps. All his examples of them consist of aggregates of physical matter. Indeed, he says, "The heap is purely [a} physical aggregation of stuff." It would seem to me that if the concept of heap is very important to an understanding of holarchy, it should apply to higher, indeed to all, levels of existence. I have just suggested that there can be heaps of cells, for example. What about heaps of organisms? Perhaps some invertebrates may exist as heaps, in the sense of groups of individual organisms that interact very minimally. For example, I would regard a swarm of mosquitoes as a heap, in much the same sense that a rock, or a pile of rocks, is. The swarm has no emergent properties that I'm aware of, except that the close proximity of individual mosquitoes to one another ensures that reproduction will take place. One small portion of the swarm is pretty much like any other portion.

Notice that while a heap may have no emergent properties, it does not follow that it is necessarily "purely a physical aggregation". This is so if the holons are purely physical, as with a rock, but not so with higher-order holons. A heap of bacteria does not consist simply of the physical bodies of these micro-organisms. Itís also a biological aggregation, with biological properties--the ability to reproduce, metabolize substances, move about, and so on. It's just that these properties are uniform throughout the heap. Any small portion of a bacterial suspension exhibits these properties. Likewise, a swarm of mosquitoes is not only a physical aggregate, or even a biological aggregate, but also to some extent a mental or behavioral one. The swarm has behavioral properties--as we are all too aware of when we are unfortunate enough to be in the middle of one--but again, these properties can be found in any portion of the swarm.

This property of uniformity is not characteristic of individual or social holons. A cell, for example, is not uniform, but consists of heterogeneous components--nucleus, mitochondria, cell membrane, and so forth (and this heterogeneity exists down all the way to the molecular level, so that the cell membrane, for example, has a heterogenous composition of proteins and lipids, which in turn have a heterogeneous compostion of atoms). Thus one cannot take one portion of the cell, and find in it the properties of the whole cell (or one portion of the protein, and find it retains the properties of the entire protein). The same is true, though to a lesser extent, of social holons. A molecule or a tissue is not completely uniform, and neither is a society of organisms. Even a herd of elephants, to use one of Kofman's examples, has some heterogeneity. Some elephants play different roles from others, so that any one portion of the herd does not have all the properties of the entire herd. (Though the simpler the society, the more nearly it comes to approximate uniformity).

In summary, I would define a heap as a group of holons that interact only very weakly. The preceding discussion in fact implies that there is no sharp dividing line between a heap and a social holon, that if the interactions between holons in a heap become stronger, the heap might in fact become a social holon. A heap could be regarded as a very primitive form of a social holon, or simply as a failure of holarchical development to occur in a particular direction.

However, we can say something more about heaps, namely, about the kinds of holons, and their interactions, that lead to their formation. One important theme, as pointed out above, is uniformity. Contrary to Kofman's assertion that heaps are a random assortment of holons, most heaps have a uniform composition, which contributes directly to their lack of further development. Consider that grain of sand, for example. It consists almost entirely of a single element, silicon. Like carbon, an atom of silicon has the ability to bond to several other atoms, so in principle it could be the basis of very complex molecules, just as carbon is. But the same property allows silicon to bond just to itself, forming three-dimensional structures of potentially limitless size and uniform composition. Since these structures contain nothing but silicon, they can have no properties not found at the near atomic level.

The same is true of carbon itself. Though carbon is of course the basis of all higher forms of life--from amino acids to human societies--like silicon, it can form extensive networks in which it is bonded only to itself, as in graphite and other minerals, for example. Again, because carbon is the only kind of atom present, such structures have no emergent properties, and are not capable of evolving further. It's only when carbon bonds to to other kinds of atoms--mostly, hydrogen, oxygen and nitrogen--that novelty begins to arise, and the possibility of truly emergent forms of life becomes possible. And to reiterate, this non-uniformity exists at all levels of the holarchy. Amino acids are non-uniform with respect to their composition of atoms; proteins are non-uniform in their composition of amino acids; cells in their composition of proteins; organisms in their cells.

So one way in which heaps form is when a single type of holon interacts with others of its type. A second method of heap formation involves holons that are unable to interact with other holons at all. As I point out in Worlds, individual or fundamental holons come in two classes, those that can interact with other holons like themselves, and those that can't. Atoms, for example, can be classified as inert and reactive. Inert atoms, such as helium, don't form chemical bonds with other atoms, while reactive atoms, such as carbon, do. Hence all higher forms of life, beginning with molecules, consist almost entirely of reactive atoms.

Likewise, cells fall into two major groups, prokaryotes and eukaryotes (we can ignore, in this discussion, a recently discovered third group, which in this respect is much like the prokaryotes). Prokaryotes, like inert atoms, generally don't form multicellular associations, while eukaryotes, like reactive atoms, do. Hence all forms of life higher than cells consist of eukaryotic cells. Somewhat analogously, organisms can be classified as invertebrates, which with some important exceptions do not form societies, and vertebrates, which can form very complex societies.

In light of this classification, we can see that heaps are often formed by holons of the non-interactive variety. Thus Gaia consists of a heap of prokaryotic cells. Since these cells are able to interact with each other only very weakly--unlike eukaryotic cells, they generally can't form the kind of intimate associations that eventually lead to the evolution of organisms--they create essentially no emergent properties. Likewise, many kinds of invertebrates interact with one another very weakly, and thus don't form societies with their associated emergent properties.

To conclude this discussion, we can simply regard heaps as existing on one end of the spectrum of interactions between individual holons (often called hetarchical interactions). These interactions are weakest in heaps. They are much stronger in social holons, and strongest of all in the individual or fundamental holons. Though the heaps we see around us today seem to be going nowhere, evolutionarily speaking, during evolution we can presume that every social holon was preceded by a heap, and social holons, in turn, preceded individual holons on the next level of existence.

In fact, the steps from evolution of heap to social or individual holon are easy to enumerate, in a very general sense. They follow directly from the two principles underlying heap formation just discussed. Consider the weakly interacting prokaryotes found in Gaia. The first step towards a higher form of life would involve developing the ability to interact, that is, for the cells to associate with one another in loose colonies. The second step would involve differentiation, that is, the development of cells specialized to carry out different functions. In this way, the heap would lose its uniform nature, and at this point would become a social holon. Finally, the social holon, in associating with other kinds of social holons, would develop the ability to reproduce itself. At this point, an individual holon has emerged.


I have saved the discussion of artifacts for last, because I believe their relationship to other holons and the holarchy is the most complex and difficult to understand. Kofman defines them in this manner: "An artifact is an entity created by a holon; its pattern (structure and function) is derived from the holon's agency." Artifacts are most common and most obvious in the inventions of humankind, but he notes that other forms of life also create them, providing the examples of a beaver's dam, and a metabolic transformation catalyzed by an enzyme molecule.

The last example is an interesting insight or analogy that had not occurred to me before. I find it ironic coming from Kofman, though, because at the outset of his article he claims that artifacts are distinct from both individual and social holons, and he later argues that confusing this distinction (much like regarding societies as higher forms of life) has potentially undesirable consequences:

creating a pseudo-holarchy such as (human + computer = workstation) blurs the distinction between a human (holon) and a computer (artifact). This has a subtle de-humanizing effect since the person "loses" its interiority and becomes "a cog in the machine", being identified herself as an artifact, a (disposable) sub-system whose intentionality is imprinted from the outside by a supra-system.

But if a molecule created by an enzyme is to be viewed as both an artifact and a holon, then it's clear that Kofman himself is guilty of blurring this distinction. And far from holding it against him, I believe this is the proper approach to take in understanding the role of artifacts in the holarchy. That is to say, artifacts don't (necessarily) create a pseudo-holarchy; they can be all part of the one holarchy of life. If a molecule created by enzyme catalysis is to be considered an artifact, then this artifact, together with the enzyme, becomes part of the holon that is the cell. And in fact, if a molecule created by an enzyme is an artifact, it's very difficult to say what is and what is not an artifact. Is the enzyme molecule itself an artifact? It's created by the agency of a gene, working with some other holons in the cell. Is the gene an artifact? It was created by the agency of another gene, which duplicated itself. What about the whole cell? It was created by the agency of some other holon. Clearly, following Kofman's definition leads to the conclusion that everything is an artifact--and in some sense, it is. Everything is made by something else.

But that is too simple, and too extreme, a conclusion. We look at some artifacts--such as the beaver dam or computer workstation--and intuitively understand that artifacts are in some manner different from other holons. How? If all holons are made by other holons, how can we define artifacts to preserve this distinction? The key is simply to realize that artifacts, unlike other holons, are not created by a process of reproduction. A gene, a cell, a tissue, an organ, an organism are all created by reproduction of some fundamental holon, either a cell or an organism. A fundamental holon directly reproduces itself, while a social holon like a tissue reproduces, or grows and maintains itself, indirectly through the reproduction of its component fundamental holons. A metabolite is not created in this way, however, nor is an enzyme or other kind of protein. Similarly, there are holons within organisms that are not created by reproduction--certain portions of the bones, teeth and hair, for example. These all fit our definitions of artifacts.

Again, though--as we saw in the earlier discussion of heaps--it's important to realize that when we move up a level of existence, our definition of a phenomenon expands. Just as a cell heap, such as a group of prokaryotes, is both a physical and biological aggregate, so an organism, which completes the biological level of existence, can have biological as well as physical artifacts. Bones, teeth and hair are examples of purely physical artifacts. Examples of biological artifacts include muscular movement, digestion, breathing, circulation of blood--in fact, almost all the functions, internal and external, of the organism. These are artifacts, to repeat, because they do not involve reproduction of cells in order to occur. They are to be contrasted with those functions of the organism that do involve reproduction, which include reproduction of the organism itself, as well as maintenance of tissues, and some elements of the immune response. In other words, most of the functions of an organism represent the creation and application of artifacts--defined, again, as the non-reproductive creation of holons by other holons.

This is obviously a counter-intuitive conclusion, but it follows directly from our definition of artifacts. Notice also that this view of the organism corresponds closely with that of the cell. For if both enzymes, and their products, are artifacts, clearly most of the metabolic processes of the cell represent the creation and operation of artifacts. Indeed, in my model of the holarchy, as discussed in Worlds, enzymatic catalysis, the functional unit of the cell, is analogous to muscular movement, the functional unit of the organism. Just as almost all metabolic processes in the cell depend on the actions of various enzymes, almost all functions of the organism depend on the actions of various kinds of muscles which control both internal organs and external movements.

Now let's move up to the mental level of existence. Here we find artifacts created by individual organisms--for example, a spider's web--as well as by societies of organisms--almost all human technology. These artifacts may now include ones with mental as well as physical and biological properties, as Kofman notes with some of his examples. Thus all forms of human thought can be considered to be artifacts. In addition to this additional conceptual dimension, however, artifacts at this level differ from those at lower levels in another important way. We have seen that physical and biological artifacts exist within cells and organisms, respectively. At the mental level, however, artifacts seem to exist outside of any higher-order holon. Thus the various creations of organisms ranging from insects to our own species are experienced as separate from ourselves. This gives these artifacts a very different flavor from those that were previously discussed; they seem less natural, and more removed from living systems. Obviously, these are the kinds of things we ordinarily mean by the term artifact.

If one believes, as I do, however, that a new higher-order holon is emerging, these mental level artifacts will eventually be integrated within this holon, just as enzymes and small molecules have become integrated within cells, and the physical and biological artifacts of organisms have become integrated into them. This leads us directly back to Kofman's worry of "pseudo-holarachies", as exemplified by the human-computer workstation model. This arrangement may be a pseudo-holarchy now, but it's a strong implication of holarchy as I view it that it will not necessarily be at some time in the future. If a new, higher level of existence is created, we should expect, based on our understanding of lower levels, that all or most of our artifacts will in fact become fully integrated with ourselves, to the point where, viewed from the higher-level holon, the distinction between artifact and holon breaks down--just as we have seen it already has on the physical and biological levels. When we look at cells, we don't say that some parts of them are living, because they reproduce themselves, and some are not, because they were created by a non-reproductive process. Likewise, when we look at organisms, we don't say that some functions are living, because they result from reproducing cells, and some non-living, because they result from non-reproductive agency. In the same manner, viewed from a higher level, a fully integrated holon would not say that human beings are living, while their technology is not. It may matter to us, but it would not matter to It. All of it goes on together.

Reproduction of the human brain

Of all the artifacts created by human beings, none is more complex, more powerful, and more potentially revolutionary than the computer. Though Kofman, and many others, seem to fear the merging of human beings with computers, I believe (even as I to some extent share these fears) that it's very likely inevitable. Indeed, itís conceivably essential to the emergence of a new, higher level of existence beyond our own, and the way in which it's essential illustrates just how, as a new level appears, the distinctions between holons and artifacts break down. I have presented this scenario at length in Worlds, and also in a shorter piece, Darwinism Evolving, also available at my website, but I would like to close this response with a briefer discussion of it here. This requires some further discussion of the holarchy as I view it.

Every individual or fundamental holon--atom, cell or organism--contains what I call an informational holon as one of its components, which specifies the entire holon. In the atom, this informational holon is the nucleus, which contains protons and neutrons. If one knows how many protons and neutrons an atom has, one can know what kind of atom it is, including how many electrons it has, and where these are situated with respect to the nucleus. In the cell, the informational holon is again the cell nucleus, more specifically, the genetic material. This material specifies the proteins the cell contains, which in turn synthesize its other components. In the organism, the informational holon is the brain, which specifies the functions of all the other major organs of the organism (through nervous control of them), as well as the behavior of the organism.

Every informational holon has both a deep structure and a surface structure. The deep structure of the genome is all the genetic information it contains, all the proteins it's capable of synthesizing. The surface structure of the genome is the genes that are actually expressed--that actually synthesize their proteins-- in any particular cell at any particular time. The deep structure of the brain is its "hard-wired" connections, which contain the potential for all the various kinds of behavior of which a particular species of organism is capable. The surface structure, again, is all the behavior that is actually expressed, in any individual organism at any given time.

As I discuss at great length in Worlds, evolution occurs at all levels of the holarchy, and by at least two distinct processes, one involving changes in the deep structure of the informational holon on that level, and one involving changes in the surface structure. Darwinian evolution, for example, involves changes in the deep structure of the genome. A mutation changes a particular gene, and the altered gene may be passed on to progeny cells or organisms, and become established in a population of them. What is usually called cultural evolution, on the other hand, involves changes in the surface structure of the brain. A new form of behavior, or idea, or some other mental aspect (usually referred to as a meme), arises, and is passed on to other individuals. Darwinian evolution and cultural evolution are considered by most scientists and philosophers to be quite different processes, and indeed they are in some respects. However, when we understand and respect the distinction between deep and surface structure, on the one hand, and that between different levels of existence and their corresponding informational holons, on the other, we find that these two types of evolutionary processes share some very deep analogies. Specifically, both involve random variation and natural selection. The difference lies simply in what varies, and the level of selection.

This understanding, then, makes it possible to bring together Darwinian evolution and cultural evolution into a single, broader evolutionary theory that includes both. In addition, though, it implies that there are at least two other evolutionary processes that have been going on, or could be going on. One of these processes would involve changes in the surface structure of the genome. As discussed in Worlds, this kind of evolutionary process is illustrated by changes in the proteins expressed by individual cells, and could very conceivably have played a major role in the evolution of the first, very simple organisms from these cells. The second type of evolutionary process predicted by this theory involves changes in the deep structure of the brain. In analogy with Darwinian evolution, mutations would occur in the hard-wired connections of the brain, so that entirely new forms of human behavior would become possible.

It seems clearly impossible that this kind of evolution could occur with human beings as they are presently constituted. There is no known mechanism by which the permanent nervous connections of the human brain can be altered, except in cases of pathology, which almost always result in an evolutionarily disadvantageous form of behavior. Moreover, even if such changes did occur, there is no known way in which they could be passed on to succeeding generations of people in any manner, let alone in a faithfully inherited manner as required by this kind of evolution. For evolution of this kind implies that all changes occur at the level of the brain, independent of the genome. These changes would not be specified by the genome, and so would not be transmittable by human reproduction.

A computer, however--or more precisely, a form of artificial intelligence--is a different story. Such artifactual brains might be originally developed to carry out many of the functions in society now carried out by humans, and thus be designed to be very much like human brains. Since the essence of the brain would be software, it could reproduce itself promiscuously, and might, accidentally or by design, be subject to errors which could be inherited during the reproduction process. As I've argued in Worlds, an artificial brain of this kind, working with or independently of a human brain, could fulfill the basic requirements demanded of a system in which the deep structure of the brain was subjected to an evolutionary analog of Darwinism. Thus just as the latter process has played a major role in the evolution of a new level of existence (first, through the creation of different kinds of cells needed for an organism, and second, the creation of different kinds of organisms), so could evolution of deep structure of the brain be a driving force bringing about the next level of existence.


Fred Kofman has drawn distinctions between individual and social holons, and between them and what he calls heaps and artifacts. While all of these entities are important to our understanding of holarchy, and existence in general, I find his definitions of them inadequate in some respects. I have offered here alternative definitions, and have used these definitions both to reclassify these types of existence, and show how all are very closely related to each other. Individual holons can be distinguished from social holons by their ability to reproduce, exist autonomously, and preserve the properties of their component holons. Though social holons can't reproduce themselves, they can create other social holons, known as artifacts. Heaps are collections of individual holons which, because of their uniform composition and/or weak interactions, do not possess emergent properties not found in their individual members.

I believe we can regard individual and social holons, heaps and artifacts as existing on a spectrum, determined by the strength of interactions between them. Individual holons are at one end of this spectrum, heaps at the other, and social holons between them. Evolution of the holarchy proceeds from heap to social holon to individual holon, which then creates (or may create) new heaps and social holons on the next level of existence. This process is driven by both reproduction of individual holons, and the creation of artifacts by social holons.

Though this response has been quite critical, I don't want to imply that I didn't find a great deal of value in Fred Kofman's article. There are very large areas of agreement between us which I didn't bother to discuss (as with newspapers, it's the bad news that sells). Moreover, he brought to my attention several key ideas that had not occurred to me before, and most important, articulated them in ways that further highlighted the differences between my understanding of the holarchy and the Wilber model, which Kofman obviously adopts more or less wholesale. In this respect, I think he has been clearer in explaining, or developing, some aspects of this model then Ken himself has.

Nevertheless, Kofman's exposition can't obscure, for me, the view that the Wilber model has some significant flaws, including:

  1. it's blatantly inconsistent in its definition of social holons;
  2. it ignores a vast body of scientific data demonstrating beyond any reasonable doubt that cells and organisms share essential properties which are qualiltatively different from those of all other forms of existence that we know about;
  3. it fails to make a distinction between mental phenomena that have been, can be or conceivably might be explained in terms of causative processes at the neuronal level (functional or soft problems), and those that apparently can't be (qualia or the hard problem); and
  4. does not so much synthesize the data from a vast body of literature that Wilber has mined as simply juxtapose it, as though saying that something has four aspects brings natural and social scientists any closer.

I give Ken his due for pointing out the importance of holarchy in our understanding of existence, and for creating a model of it broad enough in scope and accurate enough in details to stimulate both interest in the public and further work by others. But I continue to maintain that this model needs to be revised.


Bloom, H. (2000) Global Brain (NY: John Wiley)

Margulis, L., and D. Sagan (1986) Microcosmos (Berkeley, CA: University of California Press)

Pettersson, M. (1996) Complexity and Evolution (Cambridge: Cambridge University Press)

Wilber, K. (1995) Sex, Ecology, Sprituality (Boston: Shambhala)

Comment Form is loading comments...