Printer Friendly

Towards a science of systems.


For many years Jim Miller encouraged me to publish our preliminary Institute research that attempted to use empirical data from the natural sciences to determine the non-anthropomorphic 'hierarchical levels' of commonly recognized physical and biological systems. Due to our periodic discussions and meetings, he was well aware that this data might disagree with his own hierarchical taxonomy, but ever the scientist, he wanted to move the field along more than he wanted mere reaffirmation of his personal approaches. His greatest wish was to see systems theory develop into a rigorous 'science of systems', and so the title and purpose of this review explores that possibility. I am deeply saddened by his death, and in his memory, I write this article to begin fulfilling promises I made to him long ago. I only regret that now he will not be able to respond directly to these observations. I trust that the many co-workers he left behind who continue to work enthusiastically on Miller's Living Systems Theory (LST) will rise to the occasion to respond to these results and analysis so that both LST and our Institute work continues to advance. Nothing would please Jim Miller more. He knew that a science evolves only when its theories face the most exacting scrutiny and negative selection pressure, something all too often missing in the methods and traditions typical of current systems investigations. This comparison of LST and a system of systems processes (SSP) tries to be true to Miller's vision and to his preferred method of inquiry. He believed in and tried to promote a critical 'science' of systems.


Very few systems workers have been as consistent, systematic and detailed as the husband and wife team of Jim and Jessie Miller. Virtually none have left us a theory as thorough and organized as that presented in the monumental 1102-page treatise Living Systems, clearly a lifework of significance (Miller, 1978). Such an extensive work invites multiple interpretations. Here we focus on three key themes that Miller embedded in the text and further developed in his many lectures and presentations.

Unifying the Social and Biological Sciences

Miller held the strong intuition that just as biological systems had been proven during his lifetime to be grounded in and emergent from chemical systems, social systems were as surely grounded in and emergent from biological systems. This would suggest that many of the properties and pathologies of social systems could be understood by analysis of their underlying biological subsystems, just as medicine had shown that many biological successes and maladies had biophysical and biochemical precedents.

Miller was either far ahead of his times and correct, or controversial and in error depending on the observer. Powerful co-workers of his time, such as Russ Ackoff, argued that there was little or nothing common or similar between natural and social systems (personal communication). Therefore, study of natural systems had little to do with social and behavioural systems. One could not usefully inform the other. Miller's text (Miller, 1978) was a detailed rebuttal to this thesis. It is important to recognize that Miller's training was unique. He studied the medical sciences and earned an M.D. degree in Psychiatry practicing that discipline even until his last days at UCLA and the nearby Veterans Hospital. Medical science exists clearly at the nexus between the natural and socio-behavioural sciences. So his training placed him in the enviable position of knowing a great deal about both natural (biological, biochemical) and social (behavioural, psychological) systems. Through his lifelong, in-depth study of both literatures, and because of the rigor of his medical training, he was uniquely qualified to examine and to judge whether or not natural and social systems could be usefully compared for isomorphic relationships.

Making Systems Theory More Scientific

Miller consistently emphasized the need to put the ideas and insights of systems theory into more precise formulations using mathematics and to the test them by experiments based on the 'cgs' (centimeter/gram/second)(measurement) system. During his time it was accepted for highly placed academics to express their comparisons of different systems in terms of language presented in 'conjectures', that were, and still are, often mistakenly termed 'laws' (such as Ashby's Law, or Deutsch's Law). Patterns and processes of one system were related to those of other systems in descriptive terms. Miller recognized that this might be acceptable for initial explorations, but systems studies could never be appreciated, and more importantly achieve utility, unless they were expressed and tested with more formal methods. This was the motivation behind his formulation of so many cross-level hypotheses. By naming these putative similarities as 'hypotheses' and expressing them in as 'cgs'-accessible a form as possible, he was potentially making theory more scientific.

Describing a General Theory of Systems

Miller was perhaps the most dedicated of the early founders of general systems theory in pursuing an explicit generalized theory that was linked to conventional science results as expressed in the various disciplinary literatures. He used the peer-reviewed literature as his base weaving the tapestry of general theory from the supposedly real threads each science produced. The result was a general theory that was characterized by greater diversity or possessed many more key themes and ideas than competing theories. Other systems investigators, even founders, would base their entire general theory on one common process or structure. While this oversimplification was attractive (and still is attractive) to newcomers to the field, it had the exact opposite effect on experienced critics. They attacked the oversimplification of some frameworks. The same observation holds for methods. Some workers relied almost entirely on isomorphisms that could be expressed in mathematics. Miller used both descriptions of common structures or processes and mathematics. But he also saw and responded to the need for a simple 'framework' on which to organize the great diversity of findings that buttressed his general theory. That framework consisted of 'hierarchical levels' and 'common subsystems'. Thus, Miller used all three strategies; mathematics, literature-supported descriptions and frameworks.


LST and the System of Systems Processes (hereafter SSP) are both very detailed models despite being candidates for a highly generalized theory of systems. To compare them, we will need to focus only on the key features of each model or theory. First, we introduce an overview of LST, breaking the mammoth work down to three major contributions.

Hierarchical Levels of Systems in LST

Miller consistently uses seven hierarchical levels as an organizing framework for his cross-level hypotheses, namely, the cell, organ, organism, group, organization, society, and supranational systems. He completely reinterprets and reorganizes the huge literature that he cites from biology and behavioural science in terms of these seven levels. So in our comparisons here, we will deliver an earnest critique of the levels he selected to evaluate the strength and weaknesses of this hierarchical taxonomy and whether or not future work should merely follow this framework or needs to modify and surpass it.

Key Subsystems in LST

Miller is also constant in his use of 20 subsystems as an organizing framework for his model. These are: boundary, ingestor, distributor, converter, producer, storage, extruder, motor, supporter, timer, input transducer, internal transducer, output transducer, channel and net, decoder, associator, memory, decider, encoder and reproducer. The 'timer' subsystem was later added to the original 19 cited in Living Systems. This paper will show that the SSP does not use this subsystems classification and will discuss the strengths and weaknesses of its use.

Cross-Level Hypotheses in LST

Miller included another feature unique to LST. He actively combed the biological and sociological literature, and the many years of discussion of his general systems group at the University of Michigan for 'hypotheses' that could conceivably be tested in the future. He looked for statements of relations that could be true across two or more of the hierarchical levels of living systems that were the subject of his analysis. Given this characteristic, he called the results 'cross-level hypotheses'. Because they were 'cross-level', these conjectures were clearly parts of a presumed general theory of systems. He then organized and presented the 173 hypotheses or conjectures presented in Living Systems on a grid using the aforementioned seven hierarchical levels and 20 subsystems as axes. It is important to recognize that in Miller's LST all three of these features were themselves totally integrated and inseparable. To some degree, both the validity and utility of LST depends on these three features.


The SSP has generally been presented in conference proceedings and edited collections (Troncale, 1978a, 1982a, 1985b, 1986a, Troncale and Voorhees, 1983). We present here a very condensed overview to enable its comparison with LST. Like LST, SSP makes three major contributions, but these are substantially different from Miller's approach in extent and kind.

Network or Mapping of Systems-Level Interactions

The SSP essentially is the detailed identification, representation, encoding, and simulation of the vast number of interactions that occur between 'systems mechanisms' to achieve key systems functions. Each system's mechanism is a process alleged to be isomorphic across many disciplines and real systems. At first inspection, the graphic portrayal of SSP has the appearance of most networks; there are a multitude of 'nodes' connected by a multitude of 'connections'. Each individual node of the SSP (shown as boxes in Figures 1 and 2) is one of these 'isomorphies'. Each individual edge (connection or interaction) in the graph is a 'linkage proposition' (abbreviated to LP for the rest of this paper). LP's are statements in English (or any language), or in computer languages like Prolog or LISP of how one systems process influences another. Figure 1 shows only a sample of interactions between a single isomorphic process (Boundary Conditions) with 15 other isomorphic processes.


The isomorphic systems processes shown in Figures 1 and 2 are grouped into clusters shown by some major Systems Functions. Thus, the SSP model suggests that 12 known systems processes contribute to any system's maintenance through time. Not all of these systems processes are included in Figures 1 and 2 to simplify the diagrams. The sequence of Major Functions from the upper left hand corner moving clockwise is not arbitrary. These 12 Functions are suggested as Stages in a generalized life cycle for any 'mature' system. The semi-bold arrows from the inner small boxes (systems processes) into the larger boxes (Life Cycle stages) suggest that the set of systems mechanisms contribute as a group to achieving the Major Function. The bold arrows around the periphery suggest the succession of stages. Note that the lines for the tenth through twelfth stages are dashed indicating that they are a special set of stages that only occur when one scalar level of natural systems is giving rise to a completely new scale of systems (emergence). Otherwise the Systems Evolution stage merely gives rise to new variants (taxons) of an already established scale of systems. Thus, in most cases in nature, Stage Nine jumps to Stage One, unless Stages 10, 11 and 12 result in a new Field of Potential (or new scale of magnitudes in nature).

Work is underway to make this SSP graphic more fully interactive. For example, one should be able to click on any LP connection (line) to see the Linkage Proposition statement, or right click and get a menu allowing one to view the literature from which the LP was derived, or to see its expression in math symbols or Prolog, and other variations. The same would be available for each of the nodes (isomorphic systems processes).

Figure Two increases the detail by showing these LP's on Boundary Conditions, plus samples of LP's for an additional four isomorphic systems processes, namely Hierarchical Form, Synergy, three Feedback mechanisms taken together (Positive, Negative, and Coupled), and Cycles/Oscillations. This image shows 42 isomorphic systems processes that are influenced by or influence the five just mentioned. Already one can see that the complete SSP would be a very complex network. That is why it is termed a "system of systems processes'. Imagine Figure 2 multiplied by the lines for a full 100 nodes (systems processes) and several hundred Linkage Propositions and you glimpse the true nature of the SSP.

So the resulting network is proposed as an expression of all known interactions that are required for a system to survive sufficiently long for humans to recognize it. When first presented 28 years ago (Troncale, 1978a) such highly detailed networks were not as common or popular as they have become today given the appearance and analysis of the cell interactomes of thousands of specific proteins, the proteome, genome, social interaction networks, neural nets and the Internet. At least four books and many research articles are currently exploring these types of complex networks with some success in finding similarities between networks in physical systems, living systems and social systems. But while the latter represent comparison of particular systems, the SSP allegedly represents what is common between all compared particular systems, even those humans have not yet witnessed or have not yet engineered.

Comprehensive List of Systems Mechanisms or Processes

The SSP is designed to overcome a key obstacle that inhibits recognizing a general theory of systems (or application of a practical systems theory). Individual researchers tend to favour just one or only a small, restricted set of systems mechanisms in their attempt to describe systems behaviour. If you compare Boulding with Klir with Odum with Forrester with Miller, etc., you find omissions in each covered in the other, which has its own omissions. While each is more systems-oriented than conventional approaches, they have not gone quite far enough if they leave out the impacts of excluded processes. The SSP incorporates as many as 100 discovered and documented mechanisms for describing systems in general and more are added each year. SSP holds that these mechanisms are axiomatic. Each is necessary, but not sufficient. This means that every 'mature' system must possess the full set of these mechanisms to be a successful system. They are the most basic or fundamental level of systems formation and origin, and so of its dynamic and description. Some workers would prefer to refer to 'processes' rather than mechanisms to avoid the emphasis on the machine metaphor. Indeed, while natural systems scientists might be comfortable with 'mechanisms', social systems workers might be more comfortable with 'processes' as a descriptive term. Processes can be described as algorithms, or required sequences of transformation, or steps that result in structure or dynamics. The nodes in the SSP represent these axiomatic systems processes. Please see past articles listed in the references (for example, Troncale, 1978a, 1982a, 1985b, 1986a) for a working list of these systems processes (that serve as nodes in the SSP graphic).

Linkage Propositions (LP's)

On one level of inquiry, each node of the SSP graph is already isomorphic as it is a process found across many real systems. But each does not act in isolation. They influence each other. A 'linkage proposition' is a semantic statement of a specific, single, directional or mutual influence between two systems processes. The LP always takes the form (systems process A)[influence] (systems process B). Each 'line' or connection between nodes in the network graphics of SSP is a specific linkage proposition. Evidence for these directional, mutual or bi-directional LP's are taken from the refered literature of various disciplines that represent different kinds of systems. Some workers would argue that the LP is proven for their discipline. But the LP is called a 'proposition' because it is here suggested that it could also be proven in other disciplines. Until a sufficient number of such disciplines are studied, and the LP is found in each, the LP must continue to be called only a proposition. Because it is one giant level of abstraction greater than most theories (interactions of interactions), an individual linkage proposition achieves an additional level of isomorphic comparison that is often missing even in general systems theory discussions. The large set of them achieves a descriptive detail unusual for these theories.

Computerized and Formal Tools for Ease of Use

Just as panomics, systems biology, systems neuroscience and earth systems science are being overwhelmed by vast petabyte datasets, so also the SSP is an overwhelming level of detail for humans to interpret and use. So we are computerizing the SSP network so that someday a user could click on any node (systems process) or connection (linkage proposition) to ask for its information (background literature, examples, limits, measures--fully a dozen information categories). We are trying to find the manpower to make this into a GENSYSML database for use by systems workers (Troncale, 2002). Earlier presentations of SSP (then called LPTM--the Linkage Proposition Template Model) showed that there were many classes of LP influence that were isomorphic across wide ranges of systems. These classes could be represented by 'formalisms' becoming a new math of its own (Troncale and Voorhees, 1983), or expressed in computer languages such as Prolog and LISP. New propositions could be derived from established LP's. In these ways, the SSP could become an environment of valuable detail for systems investigation or systems consulting.


Since there are so many concepts, processes and theories extant in the area of systems, one of the most urgent necessities is rigorous comparison of those theories with the intent of their eventual reinforcement, improvement, elimination, absorption, or unification. It is popular for systems theorists to criticize the natural scientists tendency to defend conventional disciplinary territories or 'turf'. It is amusing (if not tragic) to this observer, that while criticizing this in reductionist science, the systems scientists are guilty of exactly the same fault in different form. They tend to devise a general theory and then defend it against all variants, heresies, and alternatives. It is quite common in professional societies to have sessions completely on one approach or another with only sympathizers and devoted followers of one model or approach talking to each other. In the interest of moving towards unification and improvement, we here compare the two theories discussed in this article using 20 initial criteria. We juxtapose 7 similarities and 13 dissimilarities between the Linkage Proposition and Living Systems Models. At first glance, both linkage propositions and cross-level hypotheses may seem entirely similar. Indeed, Miller states 'As employed in this chapter "hypothesis" and "proposition" are essentially synonymous'. But consider the following insights derived from closer analysis.

Similarity: Widespread Use of 'Unit' Concepts

Both LST and SSP use explicit, stand-alone statements. Linkage Propositions (LP's) and Cross-Level Hypotheses (hereafter CLH's) are distinct influences or relations that are very concise in their expression or description. They point to a very specific isomorphic relation between two or many systems. What the founders of both have attempted is making 'theory' more concrete. Despite the effort of both theories to describe a 'general theory" that explains the many specific manifestations of real systems, they have 'reduced' the generality to an elaborate matrix of particular relations. Some systems theorists expect that the best expression of a systems theory would be best captured in a general framework or a few equations that unify everything. But these two theories emphasize both many-to-many mappings and many-to-one mappings, not just the latter. Rather than long descriptions of similarity, intangible metaphors or vague and elusive analogies that express holism in the abstract, they both prefer specific, tangible and delimited expressions of isomorphic relation.

Similarity: Empirically-Based/or/Falsifiable

Both LP's and CLH's are considered conditional. Both are designed for testing. Each has some empirical basis that enables testing. Both predict specific behaviours and/or consequences that could be tested. As such they could be falsified, indeed, are intended to invite falsification. Both LP's and CLH's are proposed to fill the same function that 'conjectures' do in mathematics. They are devised to stimulate testing or proof in the future. It is important to note that evidence already exists for many LP's and CLH's, at least in one system; that is why they are proposed. But evidence that they exist across many systems in addition to the one's in which they have been identified is the higher objective for falsification in these two theories.

Similarity: Orthologous Character

The word 'orthologous' is used in evolutionary and systems biology to describe similarity in information sequences that are related by vertical descent from a common ancestor. Both CLH's and LP's are expressed in language, and so are information sequences. There is a good possibility that the independently derived languages of both theories bear similarities, since they both try to capture expression of similarities across apparently different systems in their language sequence. But since the real systems examined are, in many cases, the same systems, the expressions of the similarities might be usefully considered 'derived by vertical (independent) descent from common ancestor systems'. They are direct evolutionary counterparts. Some of the cross-level hypotheses and linkage propositions appear to refer to each other, can be transformed into each other, or could be derived from each other. In this sense, they are 'orthologous'.

Similarity: Tough Selection and Evolving Theories

Because of their falsifiability, both LST and SSP are designed to achieve gradual and incremental improvement. These theories do not exist statically in the past; they are meant to exist into the future with constant upgrading. Their specific description of components and the accessibility of those components for testing are meant to demand or encourage their modification by many workers. This is their mutual strength and weakness, because the 'harvest is (potentially) great, but the workers are few'. It is important that both be institutionalized to ensure future negative and positive selection. LST has been institutionalized in the ISSS-SIG dedicated to its improvement, not just its perpetuation. Both would be institutionalized in the new SIG on Systems Pathology. SIG's are Special Integration Groups of the International Society for the Systems Sciences (see ISSS homepage).

Similarity: Objectives

Both LST and SSP have similar objectives. Both are designed to enable a science of systems by subjecting the unit concepts to testing and verification. Both are a candidate general theory of systems. Both try to unify the particular manifestations of known classes of biological and social systems. Both try to simplify the study and understanding of these diverse classes of systems. Both can be used to discover systems pathologies, their aetiology, and how pathologies could be 'cured'.

Similarity: Predictions

The ability of laws and facts to enable predictions is clearly one of the foundations of rigorous science. We are used to culling good from poor hypotheses on the basis of whether or not they facilitate the recognition and testing of specific predictions. Both cross-level hypotheses and linkage propositions describe relations, predictions or mechanisms that readily admit or could be transformed into predictions.

Similarity: Utility for Studies of System's Pathology?

The manner in which both LST and SSP are constructed enables their use in the elaboration of a new specialty of System's Pathology (Troncale, 2001b), which is covered in more detail later in this paper. It is an application of the successes of medicine concerning the healthy and diseased states of humans to the healthy and diseased states of systems in general. The new field depends on precision of expression, testing, and documentation of the specific mechanics that result in systems stability and behaviours. Both the set of linkage propositions and cross-level hypotheses provide the level of specificity needed as building blocks for evaluation of a System's Pathology. The vague and overly abstract nature of many system's theories do not serve the new field as well. G.A. Swanson has recently analysed LST for its potential contributions to Systems Pathology (Swanson, 2004). Troncale has analysed the SSP for its contributions (Troncale, 2004b). Indeed, the SSP was designed, in part, to enable a rigorous Systems Pathology.

Dissimilarity: Scope of Systems Analysed

LST clearly focuses on its specified seven levels of sociological and biological systems. That is why it is called Living Systems Theory. It essentially excludes all physical systems as entities for description themselves. In contrast, the SSP is formulated to include all natural systems, including all physical, biological and social systems. Thus, there are many scalar levels of real systems included in the SSP (for example, the cosmos, galaxies and their clusters, stars and their clusters, man-made physical systems, planetary systems, geological systems, chemical systems, etc). that are not included explicitly in the LST except for their impact on living systems.

Dissimilarity: Levels Required for Cross-Level Hypotheses versus Linkage Propositions

Miller was modest in his claims for his LST systems hypotheses. Influenced by the four original goals for a general theory described by the Founders of the Society for General Systems Research (which became the ISGSR and then the ISSS), he suggested that any hypothesis that expressed an isomorphic relation between two or more of his seven levels was cross-level. The SSP, however, suggests that any linkage proposition is a similarity across all known or knowable hierarchical levels. The strength of the more modest LST claim is that it is more achievable. But this limit on span weakens its claim as the more general or comprehensive theory. The linkage propositions are posed as required or obligate mechanisms of any mature system, living or non-living, if it is to exist long enough for human perception and study. This is their strength; but it comes with the weakness of being more difficult to document or prove.

Dissimilarity: Testing Methodology: Use of cgs System

Throughout his magnum opus, Miller emphasized using the same cgs system to test his cross-level hypotheses that were used by his beloved discipline of medicine to such good effect. He associated the use of data measured in centimeters, grams, seconds with scientific testing of relations. As a result, many of his cross-level hypotheses are stated in such a way that they are amenable to applying cgs tests. By contrast, LP's are not stated as dependent on cgs testing. In fact, many LP's were derived from just such measurement of real systems. But also included are many relations that were derived from systems simulations. Many LP's express relationships that are documented for interactions several levels more abstract than the immediate cgs manifestations. The range of tests for LP's goes beyond cgs tests and admits that some relations are not cgs feasible.

Dissimilarity: Dominant Classification Framework

LST uses either the aforementioned seven hierarchical levels or the 20 critical sub-systems as the dominant framework for perceiving, identifying, and stating the cross-level hypotheses. In contrast, the SSP has hierarchical clustering as a key systems process, but not as an over-riding framework. Some have criticized LST because of its dependence on the specified hierarchical levels. But others would describe this weakness as strength because it simplifies and unifies treatment of such a vast literature. The key framework for the SSP is the internal systems processes themselves, not external, human-based classifications.

Dissimilarity: Level of Abstraction or Generality

The Linkage Proposition theory is at least one level more abstract than the LST model. LST tries to identify cross-level and sub-system organization similarities that are isomorphic between specific, compared biological and social systems. So it is at the second or third level of abstraction from the particular real system depending on whether you are looking at the hierarchical or subsystem framework. The SSP is at fourth level of abstraction. The nodes (or processes) linked LP's are similar mutual influences between isomorphic comparisons. So these are actually isomorphies on isomorphies, or similarities between similarities constituting an extra layer of abstraction and so generality.

Dissimilarity: Position of Systems Mechanisms

While the LST describes systems processes like dynamic equilibrium, feedback, cycles, flows of information, oscillation etc. in its explanation of living systems behaviours, these systems processes are not the major framework. It is Miller's organizational framework of sub-system types that constitutes the framework. In the SSP, the isomorphic systems processes are the primary framework.

Dissimilarity: Sources

Miller's Living Systems is the epitome of detailed, extensive, and careful documentation. It contains a staggering 3335 references and footnotes at the ends of its 13 chapters. While most of these sources are part of the sociological, biological or early general systems literature, Miller has also included occasional sources and insights from mathematics, information theory, and operations management. As such LST is far better documented than the current, very limited state of the SSP citation foundation. But by definition, SSP will eventually have to include a much larger citation foundation than LST because it must include all of the Living Systems synthesis plus an equally detailed synthesis of the physical sciences (astronomy, physics, chemistry, geology) molecular biological sciences, the new interdisciplinary natural sciences (astrobiology, bioinformatics, genomics, proteomics, earth systems science, systems engineering), and the new literature of the complex systems movement.

Dissimilarity: Central Role of Information and Reproduction

Miller's selection of themes and frameworks follows from his focus on unifying the biological and social sciences. The themes are therefore dependent on the 'levels' and 'real subsystems' taxonomy adopted. These subsystems favour the 'information' rich content and basis of living systems. This omits virtually all physical systems where 'information', if present at all given the ongoing debate on the nature of entropy, has a quite different connotation and expression. In contrast, the primary focus of SSP is the set of systems mechanisms themselves, not the limited taxonomy of particular subsystems recognized by LST. SSP does not recognize the importance of all of the 20 subsystems, in fact, it disputes raising certain subsystems important to living systems, such as decider, memory, associator, or reproducer to the level of general theory as they are restricted to purposeful, information-dense living systems without counterparts in physical systems.

Dissimilarity: Dissemination Mode

The interactive graphics proposed for the SSP (Troncale, 1985b, 1986a) would be an electronic-based, Internet-based documentation complete with referrals to key Internet sites, while the documentation for LST currently is limited to the extensive references in the text Living Systems and follow-up literature. This is clearly due to the text appearing and having an impact before the Internet was even born. (It is interesting to note in this context that Miller was a co-Founder of EDUCOM, an early recognition that computer networks were going to be essential for colleges and universities and their mission of education). So while the dominant mode of dissemination is different between SSP and LST, it is only because of the difference in generations of their authors. Will colleagues of Miller extend the Living Systems text to become a 'living document' by conversion to electronic modes of dissemination? Let this be an explicit challenge to them. Because of its detail, LST needs this development or its audience of users and practitioners will remain limited.

Dissimilarity: 'Usability' Tools

Once disseminated, detailed theories need to be usable. Because of their great diversity and detail (almost 200 cross-level hypotheses, and hundreds of linkage propositions), both the LST and the SSP are very difficult to use. The Linkage Propositions of the SSP are expected to be available to the public as language-based expressions (Troncale, 1978a), as formal (symbolic, logic) expressions (Troncale, 1982a) as statements in Prologue and/or LISP (Troncale, 1986a), as XML protocols in GENSYSML (Troncale, 2002), and as computer-based simulations (Troncale, 2004a). The rich set of cross-level hypotheses of LST exist mainly as the first of these forms unless the Living Systems Theory SIG of the ISSS accepts the challenge of converting them to these other forms. It is important that those dedicated to Living Systems Theory provide more tools to increase the usability of the theory.

Dissimilarity: Emphasis on Functionality

Because of their formulation, cross-level hypotheses and linkage propositions are similar in their intention to be functional, yet dissimilar in their domain of functionality. I define 'functionality' in this context as 'praxis'. To be functional is to stimulate or enable the practical application of features of a theory of systems to the improvement or maintenance of real, particular systems. Development of the field of systems application has far outpaced development of adequate or proven systems theories. Witness the explosion in books, centres, education programmes, and guru's for systems application, especially in social systems and business. I once asked Peter Checkland, who might now be considered a grandfather of the soft systems application movement (SSM), about one small box in his overall diagram of how to do a successful systems intervention (Checkland, 1981). It turned out that the box was supposed to contain the guidelines resulting from systems theoretical studies that informed those doing the intervention concerning how systems worked best. He agreed that the box was virtually empty in most SSM work. It is difficult to determine who is most culpable, the theoreticians for not filling the box, or the arrogance of those attempting interventions without filling the box, or recognizing the cautions needed for using such a restricted knowledge base. In either case, the cross--level hypotheses and linkage propositions are prime candidates for helping fill that box to improve and justify systems applications. As noted before, they would do this in different domains, CLH's for some living and social systems, and LP's for the wider domain including physical systems.

Dissimilarity: Readiness for Systems Simulation?

Which approach, LST or SSP, is closest to enquiry using systems simulation tools. Howard Odum, also a Past President of the International Society for Systems Science (ISSS) like Miller, did model Living Systems Theory in the last Chapter of his text on Systems Ecology (1983). Odum's tool generally uses BASIC as the programming language and, if it modelled a real system rather than the abstract LST, would have calculus equations to define each of the flows between components. However, there were no stocks and flows to associate with the general entities of Miller's critical subsystems, so the equations would be meaningless unless restricted to particular real systems. So neither Miller's cross-level hypotheses, nor his hierarchical levels and critical subsystems have been subjected to simulation. The same is true for the SSP, except that coding of the Linkage Propositions in Prolog will permit its direct use as a simulation that potentially could generate new LP's from the full set of known LP's. This is a different type of simulation than that prescribed by other tools, but SSP is closer to readiness for this approach than LST.

Similarity or Dissimilarity: Robustness?

Another important criteria for comparison between LST and SSP is their ability to give similar results across a proscribed range of perturbations. This 20th criteria leads to a stand-off; it is neither a similarity or a dissimilarity at the present time. If a model gives wildly different results with small changes in variables, it may be thought not to model natural situations adequately. This is a measure of whether or not a model is robust. A model's 'robustness' is often used to distinguish between different simulations. In their current form, neither LST nor SSP can be tested for which is the most robust. But the recent developments in complexity theory wherein small differences in initial conditions can lead to dramatically different and unexpected outcomes suggests that 'general' models of systems, which are essentially models of complexity, may be difficult to compare in terms of the standard measure of robustness.

Undoubtedly, there are other similarities and dissimilarities between LST and SSP whose elucidation could be instructive to both. Hopefully, additions to the list, more extensive discussion and debate on these 20 criteria, and follow-up insights will contribute to the objective they hold in common, that is, moving theory to a science of systems.


Empirical challenges to the hierarchical framework of Miller's LST is one example of testing a theory, extension of a theory, and a case study of what a science of systems might look like. Each of Miller's 'hierarchical levels' becomes a Chapter title and theme covered in 50 to > 150 pages in the Living Systems text (Miller, 1978). These levels are one of his two major frameworks. Miller reviews the literature for facts about each of his proposed 20 subsystems for each hierarchical level. Only three of these hierarchical levels are those usually cited in biological science textbooks. These texts normally cite nine levels for biology, namely molecule/polymer, organelle, cell, tissue, organ, organism, population, community, ecosystem/biome. Recent studies add a 10th level called 'molecular machines' between polymer and organelle. The absence of eight of these levels shows Miller's emphasis on social systems as a psychiatrist given that five of his seven levels are human based (as his treatment of 'organism' favours human behaviour).

Neither Miller, nor the biologists cite empirical evidence that would support the levels they recognize even though both emphasize the need for testing of hypotheses throughout their work. Our Institute classifies these 'putative levels' as 'anthropomorphic' levels, that is, hierarchical levels derived from untested human assumptions and simple human observation and description. Such 'anthropomorphic' levels are useful, but we decided that they should be subjected to empirical tests, as are most conclusions and theories in the natural sciences. A series of papers on the new specialty of Systems Allometry (Troncale, 1986b, 1987, 1988, Troncale et. al., 1990a) and the more traditional field of Hierarchy Theory presented the original tests that are only briefly summarized here (Troncale, 1981, 1982b, 1982c, 1987, Troncale et. al., 1990a).

Testing for Non-Anthropomorphic, Natural Hierarchies

Tests require hypotheses, and hypotheses should enable predictions. In this case, our prediction is that the levels cited by Miller or the biologists must exhibit the predicted characteristics of hierarchical levels. Among these are (i) clear clustering of objects on a scalar 'level' distinct from and quantal (that is, discontinuous) from other levels, (ii) increases in scalar magnitudes across the series of levels, and (iii) decreases in numbers of objects per level across the series. There are several other such 'identifying features' of hierarchies, but we decided to test for prediction (i) (that is looking for distinct clustering as seen in 'level, gap, level, gap' structure) for the natural science levels in the most empirical manner possible. If graphical and statistical analysis of the raw data demonstrated the level, gap, level feature for the levels named, then the hierarchy would be confirmed. This required collection of unusual relational databases from a wide range of natural science disciplines. It is important to note that this is a case study of rigorous scientific testing of a clearly cross-disciplinary prediction. It is a truly general theory of systems or systems science exemplar. The hypothesis, prediction, and testing could not occur within the domain of any one of the sciences, since all three must cross the domains of all of the natural sciences. So it is clearly systems-level science, or cross-disciplinary in Miller's terminology.

Empirical Data from Refereed Natural Science Journals

Data was collected on 15 parameter trends, including 10 Newtonian and five Information-based parameter sets. The data collected were entirely from accepted professional natural science journals. Each datum was the result of experimentation (often by repeated experiments) that measured the result in a manner consistent with the requirements of that natural science discipline. In this way, we were using natural science results for a systems science test, hopefully providing a case study of cooperation, rather than dispute, between reductionist and systems disciplines. The data ranged across an incredible 34 orders of magnitude (from [10.sup.-22] to 1012) given all the parameters used. Each datum was for a specific, natural-science-recognized entity of one of the conventional natural science disciplines (usually attributed to one of the anthropomorphic level labels) for one parameter. For example, 147 published mass values for known individual polymers, organelles, cells, tissues, organs, organisms, populations, communities and ecosystems were collected. As much as possible, 24 items of documentation were collected for each datum. Next, 355 cycle time (lifespan) values were collected and compared, and so on. All of the mass data for one test of levels was included in order to look for the predicted quantal or discontinuous distribution of the mass scalar values across those entities for that parameter. And so on to confirm or deny hierarchical structure using the 15 different parameter sets.

Preliminary Results

The results summarized here are only for the biological levels cited above and only for that set of parameters that provided adequate data across the range. (More extensive tests were performed including astronomical, physical, and chemical systems). I include here, for illustration, only two of the simplest tests performed. First, let me explain the general format for Figures 3A and 3B. Each has one key axis at the bottom that shows the scale of either mass or lifespan values expressed in scientific notation for all the suggested hierarchical levels studied using that parameter. On the left are the names of the putative biological hierarchical levels. The lines and numbers within the graph space are proportional and computer generated. They represent the statistical distribution of values for the sample of either mass values for each level, or lifespan values for each level taken from the scientific literature. The dark line shows the base of each sample distribution; that is, the full range of variation of the asymmetrical curve of that sample (which would be a bell-shaped curve if the distribution was normal; but is not for these real samples. In fact, you can see that the variation of values is quite different for each level sample as reflected in the width of the base distribution and the positions of the M's, l's, and 2's for each). The numbers 21M12 are computer generated to show the nature of the statistical distribution of the sample; they denote the exact positions of the M = mean, 1 = the first standard deviation, 2 = the second standard deviation for the sample. About 95% of the values lie within the second deviation (the #2's for each sample's bar). Now statisticians tell us that when two samples have overlap between their second deviations, they may be considered indistinct (not separate) samples. It doesn't matter if the extreme values overlap; just as long as the second SD's don't overlap. So the reader may want to examine each parameter set for each hypothesized level to see if it has a #2 anywhere within the 21M12 distribution of another parameter set. If there is overlap, the two putative levels may be considered as not separate; they are part of the same hierarchical level.


Figure 3A shows the statistical distribution of values for mass covering an incredible 40 orders of magnitude. Please note that the second standard deviations (SD's) clearly overlap for both molecules and organelles, but not for that group and cells. And the cell SD's are distinct from the next scalar magnitude which, because of overlaps, groups tissues, communities, and organisms. Ecosystems are clearly a different level from all others. So the seven hierarchical levels here appear to actually be four or five, given just the mass data, and depending on your confidence in how near to each other are the molecular and organellar levels.

But each parameter set does not yield the same conclusion. In Figure 3B, for example, lifespan data measured in seconds suggests molecules, cells, and organelles are very similar and may be considered one level, that is distinct from tissues, and tissues from organisms. Using this parameter across biolevels, five popularly accepted levels seem to be three distinct levels. Notice also that this parameter only varies across nine orders of magnitude, not 40 as in the case of mass values. Does the number of orders of magnitude represented by a particular parameter measure influence the test? Use of individual parameters is suspect if the parameter used (as well as the sample size included) significantly changes the non-anthropomorphic levels demonstrated. So perhaps one needs to use many parameters simultaneously to capture the real nature of a set of entities.

Figure 4 shows a multi-parametric test. It combines four of the parameter trends using the statistical clustering programme CLUSTAN. This shows that the supposed molecular, organellar, and cellular levels of the biohierarchy are in fact all on one level of magnitude, so are not three hierarchical levels in a series. The multi-parametric test is considered the stronger of the tests. Although the data sets contain thousands of data points, I believe these tests are insufficient until much larger data sets can be collected and analysed with still stronger statistical programs and strategies. It is not the purpose of this paper to present the entire picture, but only to show that it is possible to test a system's-science-based hypothesis in the sense that Miller long advocated. These tests present results that may be relevant to his LST hierarchical framework in the hopes of improving, extending, or modernizing that framework.


Impacts on LST and the Search for A Theory of Emergence

Even as preliminary results, this exercise in using real data from real systems to eliminate alternative hypotheses indicates that there could be benefits from a more scientific pursuit of theory on the systems level. It also indicates that we cannot assume from mere human-based (anthropomorphic) description and inspection that hierarchical levels are the ones we traditionally have named and popularized. Results such as these can be used in various ways. LST workers might want to alter the identity and naming of the hierarchical levels they use as a framework. Or they may want to extend the number of hierarchical levels they cite as their framework. They may want to apply similar tests, if possible, to their putative social systems hierarchical levels. They could use and extend the more rigorous methodology advocated here, or correct and alter it for adaptation to their levels.

Perhaps the most exciting development, the one with the greatest potential, would be the application of such empirical knowledge about non-anthropomorphic levels to elucidation of a theory of emergence (Troncale, 1972, 1978b, 1989). One of the clearest examples of emergence, whether in natural or social systems, is the appearance of new hierarchical levels of natural (non-human) system in the hierarchical series. This natural system's-based 'scalar emergence' offers many empirical clues about the emergence mechanisms that operate at the point of emergence-that is when the new scalar level of natural object emerges from the former. All other attempts to formulate a general theory of emergence have been pure theory or simulation and not based on natural systems or real hierarchies. It is interesting to note that while Miller mentions 'emergence' in five places, and has included one related cross-level hypothesis, he generally described the results of the emergence (newly emergent characters), not the process or mechanism that causes the emergence. Combining this work with Miller might lead to some additional hypotheses on this key systems-level process.


Perhaps it is not surprising that an MD would include a two-page section explicitly on 'pathology' in his Living Systems text (Miller, 1978, pp. 81-83). But given the extent of development of LST in the text, it is surprising that pathology isn't covered in a more explicit and extended manner. Only about 1% of his 173 Cross-Level Hypotheses in the Hypothesis Chapter and 1% of the List of Hypotheses (~600) may be considered as related to pathology. G.A. Swanson notes, 'If implications for the study of pathology permeate LST, that penetration is not obvious'. Swanson's paper is an authoritative and highly informative condensation that focuses 1100 pages of Miller's indirect 'implications for systems pathology' into 10 pages that directly show the potential utility and contributions of LST for a new effort in Systems Pathology (Swanson, 2004). The same can be said for the current stage of development of SSP. The Linkage Propositions are written to describe the mechanisms or processes by which 'healthy" systems come into being, exhibit systems behaviour, and evolve. However, the existence of so much detail on how healthy systems work, can be readily used to show the reoccurring cases of 'disease' when they do not work. In fact, is not that exactly the reason for the successes of modern medicine? Note that this strategy does not work unless there is an adequate amount of detail in the description of 'healthy' that is supported by extraordinary documentation. Both the LST and the SSP can provide that level of detail.


So how do both of these models compare with other extant attempts at unification? Since this is not the focus of this paper, only a couple of observations must suffice. LST is actually an extension of the Framework offered by Gerard who worked on synthesis of mostly the first three biological levels of Miller's hierarchy (and from which Miller borrowed that part of his hierarchical taxonomy). Both the LST-SSP are incredibly more detailed than the simple frameworks offered by Gerard and other Founders such as Boulding whose framework is based on emergence of increasing importance of use of control and feedback on new levels. These simple, all-encompassing frameworks attract some workers who are searching for a 'holy grail' of simplification that is easy to comprehend. But it actually deters many others who are looking for a model equal to the complexity of the systems included. LST-SSP also include a much wider diversity of findings than many systems workers who focus their life work on only one of the systems processes of the SSP or one of the domains of the LST to the exclusion of many of the others. Examples would be the fine work of Peter Coming on 'synergy' (Corning, 2003), and the work of physical scientists and mathematicians on 'self-organized criticality' (Jensen et al., 1998; Per Bak, 1999). There are many such workers. The 'scope of unification' for these theories is more circumscribed. They may cite other systems processes, but usually only in subservience to their single organizing principle. The processes are not axiomatic and co-equal as hypothesized in the SSP. I do not intend this as a criticism of these workers. Without their work and the work of many natural scientists, the wider unification attempted by LST and SSP would not be possible. Both LST and SSP approaches are clearly less mathematical than the works of Klir and colleagues (Klir, 1969; see, for a more comprehensive view of systems science, 1991). They are not as obligate computer simulation-based as is the work of Howard Odum (Odum, 1983) or of Jay Forrester (Forrester, 1968, 1973), although the SSP is intended for later computer simulation and even artificial systems research in that domain. But, with the exception of Odum, most of these attempts at synthesis do not include as many systems processes, on such an equal basis, as LST and SSP.


Bridging the Gaps: Natural to Living to Social Sciences

Across human history, we have been stymied by the mystery of the 'differences' between living and non-living systems. The huge, apparent gaps between physical systems and living systems, or between living systems and conscious systems have been explained by metaphor, then mythologies, then religion. For most humans today, the three domains are still completely separate in characteristics, meaning, and potential. Modern science has been able to explain how the 'gaps' are more gaps in our knowledge than real differences. Inexorably, each generation has been able to add more knowledge to how things work in order to explain how physical systems could have given rise naturally to living systems, and these to conscious systems in an unbroken sequence of natural origins (Troncale, 1990b, 2000, 2003). Recently, systems science has joined the natural sciences in showing that there are key similarities to these systems at first thought to be so utterly different. Clearly, SSP has a wider span than LST because it includes natural, physical systems handled in LST only in terms of how they affect or are affected by living systems. SSP handles them directly as candidate systems in themselves that must be compared for synthesis. If LST and SSP could be joined, the resulting similarities would bridge all of the know systems from the first physical origins to our halting attempts to establish a lasting civilization. It is not clear how the 'artificial' systems of manmade objects other than social systems would be handled by this synthesis. But bridging the gap between natural and social systems is a worthy enough goal for this first stage of synthesis.

Hierarchies as Frameworks versus as Clues to Emergence

If the specific hierarchical levels that Gerard and Miller used are superseded by some more non-anthropomorphic set of levels, the essence of their theories are not falsified so much as improved. The empirical demonstration of natural hierarchical levels significantly supports their formulations and update to current standards. As pointed out above, the recognition of new scalar levels that emerged in nature, and their close association with accessible measures and parameters that can be manipulated and analysed, may help define and prove an empirically-based theory of emergence for the first time. This would move key tenets of systems theory closer to systems science.

Hierarchies of General Theories

The detail included in both LST and SSP reminds me of conversations with G. Klir, Past President of ISSS, concerning the need for a priori descriptions of the characteristics of, or performance specifications for, a general theory of systems (Troncale, 1984). We concluded then that it was more reasonable to expect a hierarchy of partial theories than to expect one overarching general theory. But the partial theories could be arranged in a hierarchy of general theories based on the span of real systems each covered and each's degree of abstraction from those real systems. This conception of systems theory would free its workers from constant arguments over 'the' theory that unifies everything and focus them on describing or proving which theories best serve which cohorts of systems. We felt the concept of hierarchies of general theory would also be more acceptable to the established science disciplines. Both LST and SSP, since each includes hierarchies as a key framework or process, could be used as a start toward description of a hierarchy of general theories.

Dramatic Expansion of Systems or Systems Science Hypotheses

If the differences of SSP and LST formats can be overcome or synthesized, the very specific 'conjectures' and/or 'predictions' both offer in the cross-level hypotheses and linkage propositions could be integrated. The result would be a network of unprecedented detail to define either how healthy systems work, how new systems can be best designed, or how pathologies of systems could be best cured. It is my opinion that the well-established techniques and practices of SSM (soft systems methodology) need this type of prescriptive pool of guidelines to better enhance their systems interventions. A very important caveat is that CLH's are defined as 'hypotheses', and LP's as 'propositions' precisely because they have not yet been proven true across most systems. That step is necessary before they can be applied with confidence. Still, the level of detail provided by both theories make that kind of empirical refinement possible, perhaps for the first time.


Context: Recent Developments in the Natural Sciences

New interdisciplinary specialties have recently arisen from the conventional natural sciences driven by their natural development, their production of vast quantities of data, and by a flood of new funding. Systems Biology struggles to make sense of 40 000 bits of data produced by one microarray experiment, by the interactions of thousands of genes and proteins in the 800 species genomes now fully sequenced. The complete new genomes of new species can now be sequenced in less than a week. New centres, departments, and institutes are appearing in Systems Biology at major R1 universities with start-up funds of $50 to $100 million dollars. Neuroscience using tools that produce 1 mm resolution brain scans strains to understand terabyte amounts of data on the human brain. Earth Systems Science using remote sensing tools labours to make sense of petabytes of data on vegetation distributions, or climate or ocean flows. Astrobiology opens the local solar system and the entire cosmos to questions of how life began or could be sustained. These natural sciences have discovered their system's dimension (Troncale, 2001a). But they are the first to admit that they do not have the concepts and tools for dealing with these vast databases. Their dilemma is how to put these vast numbers of 'parts' together again into systems that work. Living Systems Theory, with its level of detail, its cross-level hypotheses, and its goal of unification of the natural biological and social systems is well placed to make some useful suggestions and contributions to these more well-established natural science disciplines. LST may be decades old, but it is still very timely, perhaps now more than ever.

The New Sciences of Complexity

It is interesting to see a recent trend in the literature that is emerging from new developments that were originally highly critical of past attempts at general theories of systems. The new Sciences of Complexity (chaos theory, fractals, self-criticality, self-organization, artificial life research, emergence) have inevitably moved to the position of feeling the need to find out the range of action of these phenomena they study. They are now asking the question, 'How widespread are these processes?" 'Are there general or underlying principles and mechanics we can discover?' This is also true for the above-cited new disciplines of astrobiology, bioinformatics, earth systems science, the brain sciences, and systems biology. Again, LST could provide timely clues with its many cross-level hypotheses, detailed literature citations and its hierarchical framework.

Developments in Systems Education

Given this burst of funding and interest in the systems-level of established and respected disciplines, it is also interesting to note how far behind they are in teaching their students to do interdisciplinary work. So critical is the need that major government programmes such as NIH and NSF are starting new programmes to fund more interdisciplinary teams. Yet most bemoan that our academic institutions are still structured along strict disciplinary lines. Competition, solely disciplinary-based reward systems, and isolation of vocabularies are built into the educational establishment. Both SSP and LST provide a framework for studies in systems education that these older disciplines might want to explore. They both provide a good framework for systems studies, exceptional detail, and ties to the disciplinary literature. It is interesting to note that Miller initiated an Institute for Systems Studies (see Troncale, Recollections, this issue) circa 1975, anticipating its need well before these very recent developments in the sciences. It is ironic to note the termination of that Institute (and many others, see Troncale, Recollections, this issue) by the very establishment that now needs to meet the new challenge of systems education. Perhaps now others will recognize how timely LST is today for today's challenges.


The Concept of "Empirical Refinement:' Need for a Unique Systems Methodology

The use of 'correspondence principles' in the natural sciences is well understood. The evidence for a cause-effect or formulaic relationship among variables depends on limiting the variables in a controlled fashion, on a deep, prerequisite understanding of the system perturbed, and by measurements that are designed to be consistent or inconsistent with predictions. Empirical measurement and statistical analysis are the foundation for experimental verification or falsification of hypotheses. This has, in the past, only occurred in very reduced subsystems of physics, geology, biology, chemistry, etc. If systems science is to be more scientific, it needs to develop a new methodology well beyond its present reliance on endless philosophical discussion and unaided description. It needs to have a 'wetlab & realworld' correspondence to computer simulations and pure mathematical modelling. But comparing systems across great spans of scalar magnitude and type exceeds the reach of the correspondence principles, tools, measurements, and knowledge base of standard natural disciplines. The mutual interactions, processes, and patterns of systems theory are, by definition, those that can be exhibited only across many such systems. What the cross-level hypotheses of LST and the linkage propositions of SSP make possible is the empirical refinement of the interdisciplinary patterns and processes. What do we mean by 'empirical refinement' (Troncale, 1985a, 1989)? We use 'refinement' to distinguish this type of test from those of normal science that are capable of direct falsifiability. Tools and measurements that are accepted within each discipline can be used to rigorously examine and verify that a systems process exists within that discipline and then the same tests can be made within each discipline, according to its tools and standards, and the results compared for isomorphic relationships. If the process is the same across all systems as measured by each discipline, the resulting expression of the similarity of process or mechanism can be used to construct a general theory for that cohort of system. This would be a unique methodology specific to a science of systems. The LST and SSP are uniquely designed to enable development and application of this new method of empirical refinement across sciences leading to a new science of systems.


Checkland P. 1981. Systems Thinking, Systems Practice. Wiley: New York.

Corning P. 2003. Nature's Magic: Synergy in Evolution and the Fate of Humankind. Cambridge University Press: Cambridge, UK; 464.

Forrester J. 1968. Principles of Systems. Wright-Allen: Cambridge.

Forrester J. 1973. World Dynamics. Wright-Allen: Cambridge, 144.

ISSS Homepage, International Society for the Systems Science,

Jensen HJ, Goddard P, Yeomans J. (eds). 1998. Self-Organized Criticality. Cambridge University Press: Cambridge, UK; 167.

Klir JG. 1969. An Approach to General Systems Theory. Van Nostrand: New York; 323.

Klir JG. 1991. Facets of Systems Science. Plenum: New York; 664.

Miller JG. 1978. Living Systems. McGraw-Hill: NY.

Odum HT. 1983. Systems Ecology: An Introduction. Wiley: New York; 644.

Per Bak. 1999. How Nature Works: The Science of Self-Organized Criticality. Copernicus Books; 212.

Swanson GA. 2004. Living systems theory and the study of pathology. In Proceedings for the 50th Annual Conference of the International Society for the Systems Sciences, Asilomar, CA, July 4-9.

Troncale LR. 1972. Origins of hierarchical levels by the action of systems field axioms. In the Proceedings of the ISGSR 16th Annual Meeting. Published by the Society for General Systems Research, Louisville, KY, 35 pages.

Troncale LR. 1978a. Linkage propositions between fifty principal systems concepts, in Applied General Systems Research: Recent Developments and Trends: N.A.T.O. Conference Series II. Systems Science, Klir GJ (ed.). Plenum Press: New York; 29-52.

Troncale LR. 1978b. Origins of hierarchical levels: an emergent evolutionary process based on systems concepts. In Proceedings of the 22nd Annual Meeting of the S.G.S.R., Ericson R (ed.); 84-94.

Troncale LR. 1981. Are levels of complexity in biosystems real? Applications of clustering theory to modeling systems emergence. In Applied Systems and Cybernetics. Systems Concepts, Models, and Methodology, Vol. II. Lasker G (ed.). Pergamon Press: New York; 1020-1026.

Troncale LR. 1982a. Linkage propositions between systems isomorphies. In A General Survey of Systems Methodology: Conceptual and Mathematical Tools, Vol. I. Troncale L (ed.). Intersystems Publ: Seaside, CA; 27-38.

Troncale LR. 1982b. Testing hierarchy models with data using computerized, empirical data bases. In A General Survey of Systems Methodology: Conceptual and Mathematical Tools, Vol. I. Troncale L (ed.). Intersystems Publ: Seaside, CA; 90-102.

Troncale LR. 1982c. Some key, unanswered questions about hierarchies. In A General Survey of Systems Methodology: Conceptual and Mathematical Tools, Vol. I. Troncale L (ed.). Intersystems Publ: Seaside, CA; 77-81.

Troncale LR. 1984. What would a general systems theory look like if I bumped into it? General Systems Bulletin XIV (3): 7-10. (editorial).

Troncale LR. 1985a. On the possibility of empirical refinement of general systems isomorphies. In Systems Inquiring: Theory, Philosophy, Methodology, Vol. I. Banathy BA (ed.). Intersystems Publications: Seaside, CA; 3-6. (editorial).

Troncale LR. 1985b. Two practical tools that facilitate general systems education and research: personal, microcomputer relational data bases on 75 systems isomorphies and 175 linkage propositions. In Proceedings of the International Conference on Systems Research and Education. Helsinki, Finland, June, 1985. 11 pages.

Troncale LR. 1986a. Knowing natural systems enables better design of man-made systems: the linkage proposition model. In Power, Utopia and Society: New Approaches to Complex Systems, Trappl R (ed.). Plenum Press: New York; 43-80.

Troncale LR. 1986b. Allometry in biology: allometry in systems science-will general systems form always be a mystery? In Mental Images, Values, and Reality: SGSR Proceedings, 30th Annual Meeting, Dillon J (ed.). Intersystems Publications: Seaside, CA; D51-D61.

Troncale L. 1987. Hierarchy theory VII. systems allometry II.: further tests of quantitative correlations across levels of systems organization. In Problems of Constancy and Change: The Complementarity of Systems Approaches to Complexity: Proc. of the 31st Intern. Conf. on Systems Theory, Vol. III. Kiss I (ed.). Publ. by the SGSR from Budapest, Hungary; 97-107.

Troncale L. 1988. The new field of systems allometry: discovery of empirical evidence for invariant proportions across diverse systems. In Cybernetics and Systems 1988 Part I, Trappl R (ed.). Kluwer Academic Publ: Boston; 123-130.

Troncale L. 1989. Is an empirically-based theory of emergence possible? In Proceedings of the 33rd Annual ISSS Conference, Edinburgh, Scotland. Ledington PWJ (ed.), Vol. III; 69. (Abstract)

Troncale L. 1990b. Theory of Emergence for and Unbroken Sequence of Origins. Presidential Address. 35th Annual Meeting of the International Society for the Systems Sciences, Oestersund, Sweden.

Troncale L. 2001a. The future of the natural systems sciences. In Understanding Complexity, Ragsdell & Wilby (eds). Kluwer Academic/Plenum Publishers: New York; 219-237.

Troncale L. 2001b. The new field of systems pathology, July 10th. In the Proceedings of the 45th Annual Conference of the International Society for the Systems Sciences (ISSS). Asilomar: Pacific Grove, CA. Published by the ISSS. (Abstract)

Troncale L. 2002. SYSML: An open source computer-based tool for research and education in the systems sciences biannual meeting of the international conference on complex systems. Published by NECSI, Cambridge, MA. Meeting in Nashua, NH, June 9-14. 108 (abstract).

Troncale L. 2003. Imminent integration of the natural and systems sciences: the unbroken sequence of origins section complex systems lecture series-Talk #1. Complex Systems Institute, University of Anchorage, Alaska, April 22. (see http://www.uaa.alaska. edu/complexsystems/events.htm)

Troncale L. 2004a. Is Artificial 'Systems' Research Possible? In Section on Concepts, Formalisms, Methods, and Tools in the Abstract Workbook for the International Conference on Complex Systems' 04. Published by the New England Complex Systems Institute (NECSI), Cambridge, MA. pp. 13-14.

Troncale L. 2004b. The systems pathology handbook and manifesto. In Proceedings for the 49th Annual Conference of the International Society for the Systems Sciences, Asilomar, CA, July 4-9.

Troncale LR, Voorhees B. 1983. Towards a formalization of systems linkage propositions. In The Relation Between Major World Problems and Systems Learning, Lasker G (ed.). Intersystems Publications: Seaside, CA, Coutts Library Services; 341-349.

Troncale L, Williams D, Putinier J, Cramer J. 1990a. What empirical testing of hierarchies & systems allometry says about general systems models of evolution/emergence. In Proceedings of the 34th Annual Mtg. of the International Society for the Systems Sciences (ISSS), Banathy BA (ed.). NECSI: Cambridge, MA; 1151-1156.

Len Troncale *

Institute for Advanced Systems Studies and Biology Department, California State Polytechnic University, USA

* Correspondence to: Len Troncale, Institute for Advanced Systems Studies and Biology Department, California State Polytechnic University, 3801 W. Temple, Pomona, California, USA, 91768. E-mail:
COPYRIGHT 2006 John Wiley & Sons, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Research Article
Author:Troncale, Len
Publication:Systems Research and Behavioral Science
Geographic Code:1USA
Date:May 1, 2006
Previous Article:Living Systems Theory and Social Entropy Theory.
Next Article:Bios, a process approach to living system theory. In honour of James and Jessie Miller.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters