Monday, November 26, 2012

An Ecological Account of Parts and Wholes Preparatory to a Comparison with Husserl's Account of Same Issues


Systems are defined by Karl Ludwig von Bertalanffy as “sets of elements standing in interrelation”.[1]  Acknowledging that the definition may seem vague, von Bertalanffy argues that when the idea is mathematized using differential equations, novel properties can be adduced in systems in general and in more specialized applied situations.  Although von Bertalanffy has quite concrete objects in mind, for instance “a galaxy, a dog, a cell, and an atom are real systems” [von Bertalanffy’s emphasis], he also recognized as systems those “conceptual systems such as logic, mathematics (but e.g. also including music) which essentially are symbolic constructs; with abstracted systems (science) as a subclass of the latter, i.e. conceptual systems corresponding with reality.”  [von Bertalanffy’s emphasis][2]  There is, it would seem, an immediate parallel between von Bertalanffy and Husserl in their recognition that thinking of parts and whole (Husserl) or elements and systems (von Bertalanffy) can refer to quite concrete objects as well as more essential rules.  It is pretty clear though that whereas Husserl has more commitments to the ideal over the empirical von Bertalanffy’s emphasizes the converse.
Hierarchy theory is a component of this more general systems theory that is applied to understanding the “architecture” of complex systems.[3]  “Nature loves hierarchies”, Herbert Simon, the social scientists, who pointed out that natural objects can be seen as arranged like Chinese boxes, each level inside a progressively larger box.  Herbert Simon recognizes four intertwining sequences: chemical, organismic, genetic, and human social organizations.[4]  This fourth hierarchy includes “the “programs” and other components called elementary information processes”.[5]  We might like to think of this as “mind”, but in this fourth hierarchy Simon also includes those programs which “have been occurring with growing in the artificial complex systems called digital computers.”
The tenets of hierarchy theory have been attractive to ecologists since observations of the nestedness of ecological levels, organisms, populations (of a single species), communities (of several species), ecosystems (the biotic community combined with the abiotic environment) and so on.  This hierarchy in natural systems is referred to as the “level of organization” concept.  Ecologists have proceeded with the assumption that subsystems on the same level can be studied without reference to one another.  For instance, we might study prairies, making the assumption that we do not simultaneously have to include tropical rainforests in our investigation.[6]  This methodological assumption relies upon the supposed “near-decomposability” of all medium-number systems and is rooted in the observation that “most interactions in nature, between systems of all kind, decrease in strength with distance.”[7] However, there are some dangers in simply conflating ecological hierarchy with “levels of organizations” concept since natural systems are comprised of more than just simple entities (organisms with clearly defined boundaries, biotic communities that are spatiotemporally reasonably well designed etc.).  They are also comprised of more diffusely defined sets of processes, and, depending upon the research question, there is more than one “n-1” level that might be examined.[8]

We need not detain ourselves with details of how ecologists have applied their analyses of wholes and parts to questions of environmental concern.  Rather, we note, quite briefly, some of the theoretical assumptions and conclusions derived from their work, especially those aspects that either parallel or diverge from Husserl’s perspective.  Though some of what follows is the result of empirical generalizations, amenable to experimental verification, some conclusions, however, are based upon a mathematical analysis thus having a more abstract status.
Systems can be organized vertically as we have been doing above – cells, tissues, organs and so forth.  Behaviors at increasingly higher levels are characterized as occurring at slower rates.[9]  Tree leaves, for example, have the physiological mechanisms to respond to momentary changes in atmospheric content of gases, whereas the entire biosphere responds only to decadal long changes in atmospheric changes (a root of issues concerning atmospheric carbon dioxide and climate change).  There is a level of buffering between elements in the vertical hierarchy.  This is a consequence of their operation at distinct rates.  If this were not so ever small oscillation at lower levels could ramify catastrophically throughout the systems.  This places, therefore a limitation on connectivity within natural systems.  This is counterintuitive in ecology where the connectivity between systems elements had been a default position.[10] In contrast, the conclusions above point to a “loose vertical coupling” in ecological systems.  As O’Neill and his colleagues have pointed out: “The old imagery of the natural world as having everything connected to everything else is shortsighted."[11]
Decomposing systems into parts and levels by recourse to rate processes is described by O’Neill and colleagues as more “fundamental” than any other method of identifying systems levels.[12]  Alternatives such as basing a hierarchy of “tangible components” or indeed basing analysis on the traditional levels of organization are merely a “special case of surfaces defined on differences in rate”. Within a single level distinct parts can be thought of a loosely coupled but this time in a “horizontal” sense. Subsystems (Husserl’s parts with parts), often referred to as “holons” in systems literature, are defined by frequency of internal interaction frequency.  Interactions between holons are relatively infrequent.  Although the boundary of a holon can be tangible (a cell wall, for example) it does not have to be (a rhizosphere – a plant root with the small soil volume surrounding it – is a holon but does not have a visible boundary).  Thus a system with a very large number of potential components (parts of a holon) are generally organized into a more tractable number of units, interactions among which define system behavior.
We can mention one final item of importance in systems theory the implications of which has been discussed extensively in ecology: the notion of emergence.  Briefly, emergence refers to a property of self-organization in systems whereby large-scale features at one level of the hierarchy are manifested but are not predicable by inspecting the behavior of their component parts.  Nonetheless, the emergent features can impose constraints on lower levels in the hierarchy (downward causation).  In ecology emergence is often summarized by the expression “The whole is greater than the sum of its parts”.  In fact, this was a frequently employed catchphrase of E P Odum, the influential 20th C American ecosystem ecologist.  In the fifth and final edition of his influential textbook Fundamentals of Ecology Odum defines emergent property as those that emerge from components are combined to form larger wholes.[13] Such properties are, he claimed, “nonreducible”. To illustrate he suggests that when the components of water of combined its fluid properties are unpredictable from an inspection of its gaseous components.

We can now summarize that the characteristics of the theory of wholes and parts in hierarchy theory are that entities are defined by the rates of processes associated with various parts arranged as progressively larger wholes in a hierarchical sequence.  A consequence of rate defined entification is 1) that parts are only loosely coupled with higher and lower levels in a hierarchy and 2) that holons are loosely coupled horizontally with other holons.  A third component, emergence, has been incorporated into systems analysis in computer science, psychology, ecology, and even into the traditional physical sciences. 
The view of parts and wholes emerging in systems theory has quite clearly got parallels in Husserl’s account.  I will have to defer a full treatment of these parallels for another time and remain satisfied that by lining these two accounts up side-by-side that I have at least provided an entry-point into a larger discussion.  It is apparent that hierarchy theory is concerned especially with an analysis of concrete objects and with the way in which they may be (psychologically) encountered by a scientist.  This corresponds somewhat with Chapter 1 of Logical Investigation III The difference between independent and non-independent objects, even if there remain striking differences in the specifics of the analysis.  However, it is also apparent that there is no development in hierarchy theory that concern itself in an analysis of Ideals.  It is the tracking back of the concrete analysis into a priori relations that makes Husserl’s account especially provocative.  Even if we must leave such a comparison of Husserl’s theory and hierarchy theory aside for now, I will say briefly say something about the third aspect of hierarchy theory – emergence – in relation to Husserl’s work.

Other posts of mine on Husserl here, here, and here.




[1] von Bertalanffy, Karl Ludwig (1968) General System theory: Foundations, Development, Applications, New York: George Braziller, revised edition 1976 p38
[2] Ibid, xxvi
[3] Simon, Herbert A. (1973) The Organization of Complex Systems in Hierarchy Theory: The Challenge of Complex Systems (The International Library of Systems Theory and Philosophy), editor H H Pattee, p5
[4] Ibid,
[5] Ibid., 6
[6] O'Neill, R. V., D. DeAngelis, J. Waide and T. F. H. Allen. (1986) A Hierarchical Concept of Ecosystems. Princeton University Press.
[7] Simon, 9
[8] In the ecological hierarchy literature the fairly reasonable approach of calling the level of proximate interest in a given system “n”, the level below it is then n-1, the level about n+1 etc.  Most ecologists restrict themselves to a “triad” of n-1, n, n+1.  Explanations of behavior are typically sought in the n-1 level (and though one might reduce the system future it is usually deemed not useful to do so) and the systems constraints are located at the n+1 level.
[9] O’Neill et al, 76
[10] A popular expression of this sentiment is from John Muir, who in My First Summer in the Sierra (1911) wrote “When we try to pick out anything by itself, we find it hitched to everything else in the Universe.”
[11] O’Neill et al, 86
[12] Ibid., 93
[13] Odum Eugene and Barrett Gary W. Fundamentals of Ecology Brooks Cole; 5 edition (July 27, 2004)

No comments:

Post a Comment