System Conceptualisation is the grand name for forming an idea of what the systems or system of systems are that are of interest in addressing a problem situation. Many people with positivist or empiricist leanings think that the concept of the system or systems of interest is derived simply and straightforwardly from "observation". However, as Thomas Khun and Karl Popper and various other philosophers and systems thinkers pointed out the idea of "objective observation" is something of a myth; every observation is "theory-laden". This means the process of system conceptualisation begins with a theory of what the system is that is used as the basis of observation and experiment planning - and the results of the observations help to evolve the concept of the system (or systems). The conceptualisation of the system therefore occurs very early on in any Systems Analysis or Systems Modelling exercise. As such it is profoundly influenced by the philosophical predispositions of the observer / analyst.
This page explains how STREAMS approaches System conceptualisation - and distinguishes three different approaches, based on three different philosophical predispositions to the production of a system (of interest concept and its models.
- 1 The Base Model of Reality
- 2 The Positivist View
- 3 The Interpretivist View
- 4 The Critical Realist View
- 5 Relation to Soft Systems' Notions of System Conceptualisation
- 6 Perspectives and Positions: The Possibility of Objectivity
- 7 Natural, Analytic and Engineered System Boundaries
- 8 Navigation
The Base Model of Reality
In line with the Principle of Pragmatic Ontological Pluralism (POP) the model distinguishes three fundamentally different categories or levels of thing which are labelled here as "Physical", "Mental" and "Abstractal". They are represented as 'levels' because, in line with the Principle of Naturalistic Closure, STREAMS assumes that the mental emerges from the Physical (in brain-like physical structures), and that the Abstractal emerges from the Mental (in mathematics-like and science-like cultures (ie communities of mentality)).
There are levels of composition, levels of typology (categorisation) and levels of description in each of the three primary layers. Hence in The Physical we can distinguish the microphysical world composed of nucleons and electrons and the forces that affect them. At the opposite end of the scale there is the macrophysical world comprising Space and the stars and planets, and cluster and galaxies etc. And in-between the mesophysical world of everyday objects that are "human-scale" - cars and buildings and light and heat, and gasses and fluids etc. Levels of composition asserts that molecules are made of atoms and compounds made of molecules and these make up substances that can take on various forms and shapes with chemical, mechanical and electrical properties that can construct objects like trees and dogs and motor-cars and aeroplanes. Levels of description asserts that people can describe (and model) these things in different ways, depending on their purposes, that can be high-level (and neglectful of irrelevant details relative to the purposes) or low-level (neglectful of gross features not relevant to the purposes). So for example you would not describe cars on the motorway in terms of atoms of steel and copper and compounds of plastic and rubber - the low level description is not relevant for traffic modelling purposes.
The most fundamental layer and category is "The Physical" - and it includes all the entities and fields or forces that Physics identifies and considers to make up the Universe. Hence, this is the "base reality" - out of which everything else - everything non-physical emerges. To be explicit, the physical layer includes "ordinary" matter composed from nucleons - protons and neutrons - and electrons; but it also includes all the theorised particles of the Standard Model of particle physics. Consequently it includes all the particle picture bosons that describe or explain the usual fields and forces - including the Strong Force, Weak force, electric and magnetic fields (and photons). But it also includes 'mass' - and the gravitational and Higgs field.
However, the particle physics perspective is a microphysical level and description - and there are higher levels - the level of solids, liquids, gasses and plasmas composed of the 'normal elements' - as tabulated in the Periodic Table. There is then, of course, the level of description in terms of categories of materials: metals and ceramics, plastics and glasses, semiconductors - and all the things of materials science and chemistry. Ultimately these result in all the everyday objects of the natural and engineered world - everything from trees and animals through to buildings and aircraft.
The mental is defined simply as all the things that exist only in the context of the mind (human or animal). The mental does not exhaust the category of the "non-physical" - that would be dualism; there are non-physical things that are not mental (see The Abstractal below). To be explicit the mental includes perceptions, conceptions, thoughts, ideas, pains, experiences, dreams, feelings etc. - all the things that can occur between the ears of a (normal) human being. The mental is necessarily subjective, 'ontologically subjective' - meaning it can only be perceived, experienced, be known about and exist in relation to a consciousness or mind that does the perceiving, experiencing, knowing etc; no mind, no mental 'object'. It therefore follows that 'direct experience' is necessarily private - only I can have my experiences; and this goes for all things mental - only I can think my thoughts, dream my dreams etc.. Minds are assumed to emerge from certain physical structures (and actual or potential information flows inherent in those physical structures) and comprise all the mental 'objects', all the processes to which objects can be subject (like imagining, deciding, resolving, analysing, understanding etc.) and the context (mental models) in which these objects and processes occur. [A mental 'object' is a fundamentally different type of 'object' from the notion of a physical object but there is often a 'causally-generated' correspondence; the laptop in my head (my perception of my laptop) is different from the laptop on my desk that my fingers flitter over - though the one causes the other (I perceive it (have the perception) when I look at the thing while typing). Again there are levels of description and composition in the mental world - my laptop (appears to) have a keyboard, mouse, power supply, screen, fan - and heat output, etc. all of which have discernibly separate perceptions but which compose together to make the perception of a laptop; and so it goes with all perceptions that together compose the "subjective field" - ie the totality of my current experiences of the world. These are however, all in my head although there is a strong correspondence with the 'real' physical world - a correspondence which underpins Direct Realism (or Naïve Realism). [And the fact I can assert that there is a correspondence - and that it is not only plausible and credible but commonly and easily understood - shows that there is an 'approximate truth' in Direct Realism. If there were not any truth that correspondence notion would be incomprehensible, but luckily everyone unpolluted by sophisticated philosophy knows it is true from their own experience.]
Subjective and Inter-Subjective
It was observed above that mental things are necessarily subjective and private. This does not mean that they cannot be shared and common. People can have the same, or very similar, experiences, thoughts, ideas, pains etc. - in the type sense, not the token sense. [That is you cannot experience my pain and I cannot experience yours - but we can both have the same (or very similar) back pain, perhaps following an injury.] Ideas, and mental objects generally, can be 'copied' from one mind to another in the same way that a slightly distorting photocopier can copy an image from one sheet to paper to another. It won't be precisely the same image - but it will be approximately the same image and everyone would recognise it as the same image. This 'copying' is not through any mystical, clairvoyant mental 'radio transmission' directly from one mind to another; rather it is a much more prosaic indirect mechanism. One mind creates a series of symbols, signs and images in the shared common physical reality which, upon being perceived by the other mind causes the idea that started out in the first mind to occur in the second. When 'the same' idea/thought/perception/experience is shared by a number of minds it is said to be 'intersubjective'.
For this account of intersubjectivity and communication to work there are a number of assumptions that have to be made: 1) that multiple minds exist 2) that they share a common physical reality, 3) that there is mental-to-physical causality that enables symbols, signs and images to be created in the common physical reality 4) that there is physical-to-mental causality because 'reading' those signs and symbols causes the idea to occur in the receiving mind and 5) that there is/are some conventions on the signs and symbols (including acoustic signs and symbols called 'speech') that is understood by both minds. The last assumption is clearly related to Wittegenstein's (impossibility of a) Private Language Argument. It also raises the question of how such conventions might arise. Perhaps the only plausible explanation of why any shared languages exist is that a) they arise naturally and inevitably through behavioural communication - as, for example, in dogs and wolves, and horses and chimpanzees - and almost any intelligent animal b)they are useful - allowing group action to achieve mutually beneficial aims and purposes and c) they can easily develop from simple references to a common, physical reality, with the simplest being the 'point-and-enunciate' method of lore. The latter of course implies the approximate truth of Direct Realism. So being a Realist about the Mental implies rejection of both Berkelyan and Kantian Idealism and Subjectivism - on some very plausible assumptions. In 'traditional' Soft Systems Thinking there is a distinct subjectivist-idealist (Kantian) strand and tenor - which from a STREAMS perspective looks misguided and logically incoherent.
The Abstractal is a difficult concept - partly because it is so unfamiliar. Since the time of Descartes both philosopher and layman have been 'conditioned' to think in dualistic terms - that "physical" and "mental" together is all there is. Despite the increasing - indeed some would say overwhelming - volume of evidence and argument that suggests both simple monisms and any form of dualism cannot be correct people have psychological difficulty in letting go the (supposedly intuitive but actually learned) idea that something is either physical or mental or illusory. A small minority of people end up with Ontological Pluralism.
One way to intrinsically define the abstractal is counterfactually: The Abstractal comprises those abstract things that would be (reflected in) common, ubiquitous concepts, systems and structures that any sufficiently large group of sufficiently intelligent beings would come to given sufficient time.
The Abstractal is almost the same in concept as Popper's World 3 but there are subtle differences. There are three criteria for something to be counted as being included in the Abstractal category (or real things): 1) causal efficacy - it must be the cause of or caused by some events or interactions 2) it is abstract and non-physical 3) it is logically coherent with the existing body of knowledge of the Abstractal. Logic and mathematics are paradigm 'objects' in the Abstractal - but so are well established scientific theories, for example the theory of Special Relativity. [Contrary to the common misconception scientific theories are not proven - in the sense of confirmed or verified - by experiment; they are only proven in the sense of "tested" and falsified or disconfirmed - or not. This is Popperian falsificationism.]
The Positivist View
The 'positivist' view of System Conceptualisation derives from traditional positivist philosophy that underpins not only the science of the 18th and 19th centuries (by the end of the 19th century the limitations of positivist philosophy in Physics were becoming apparent) but also tradition Hard Systems Thinking.
The key points are that for the Positivist "systems" are things in the "real world" and observing them is unproblematic. Hence there can be mistaken concepts of the system or different veridical concepts mistakenly using the same name but two non-mistaken observers must 'see' the same, objective system and form the same concept of it. If both observers are really talking about the same system - they must have the one right way of conceiving it (and therefore representing in models). The models are 'correct' to the extent that the model explains, describes, predicts and mimics the 'real-world' system (in all important aspects for the purposes in hand). To the extent the model differs from the real-world (in its predictions/explanations), it is wrong or incomplete. If two observers or analysts disagree about "the system" (so long as they genuinely are talking about the same system) then one of them must be wrong (or maybe both of them) and the arbiter is reality; the models (or concepts) must be tested against reality - and the one that does not predict correctly or explain well is the wrong one.
The diagram above illustrates the positivist idea of System Conceptualisation. The red line traces the causal interactions between the elements or components of the system (identified as such by the observers) and he or she forms the system concept by logically and notionally putting a boundary around those elements. Hence the direction-of-observation is world-to-mind and the direction-of-fit is mind-to-world; the model is good if it fits - correctly describes - the world. The model expresses truth if the world behaves as the model says it does.
The positivist assumes the system concept she arrives at is correct and veridical because objective observation is unproblematic. Unfortunatley observation of real-world practice shows that this assumption is not valid. People's concepts of systems vary according to their philosophical predispositions, their 'social location' and their prejudices and biases; the ideal of objective observation is in practice unattainable. [In Science somewhat extreme measures have to be taken to try to exclude all biasses - as for example with double-blind trials; in general in the situations where Systems Thinking is applied these measures are not practical or economic.]
Positivist philosophy is especially popular with Physicalists - who seek to reduce all phenomena and theories of different phenomena to physical effects and causes. They deny the 'real' existence of psychological or social causes - and seek to explain their plainly apparent reality in physical terms. A consequence of the Physicalist's philosophical commitment to a reductive Realism form of monism she is unable to either a) conceptualise systems involving mental, social or abstractal causes or entities or b) describe systems in using psychological, social or abstractal terms. So for example, cars stopping at red lights because Drivers consciously follow the Law in regard to driving is an explanation for an observed phenomenon that is unavailable to the Physicalist. This latter is the restriction in Traditional Systems Engineering or Hard Systems Thinking at it operates at the very early stage of System Conceptualisation to exclude huge swathes of different types of system (systems with non-physical components or interactions) from analysis.
The Interpretivist View
The main tenet of the Interpretivist View is that "Systems" per se do not exist in the "real-world" but are mental constructs that we 'project' onto the world. The Interpretivist rejects the notion of objective perception / conception of an external common, public "real-world" system. For the Interpretivist therefore different systems conceptions, by observers are all equally valid, and expressions of their personal, subjective mental constructs (barring the scenarios of two observers talking about different systems under the one name). Obviously two observers or analysts can discuss their subjective conceptualisations and, if they are willing, drive towards a consensus - and form group or stakeholder grouping views about what the systems are. In this sense the Interpretivist view leads naturally to epistemic relativism since the conception, understanding and explanations of the system are relative to the group holding and discussing them. For the Interpretivist there is no problem of objective observation, like the Positivist's problem, because systems are not observed from the 'real-world' but actually subjective creations that are projected onto it. Hence Weinberg is able to assert "..., a system is a way of looking at the world. The system is a point of view - ..." in "An Introduction to General System Thinking".
This is illustrated by the above diagram - the red and green lines represent different conceptions (and hence different shapes) of the notionally same reality projected onto that reality by different observers / analysts. These are different 'understandings' of the system - and, in principle, lead to different explanations of the observed real-world phenomena. For the interpretivist and committed epistemic relativist these 'understandings' are equally valid and the observers are equally well-justified in holding them; they literally and metaphorically see things differently.
The Interpretivist view immediately solves the problem of the observable epistemic relativity of system conceptualisation - an epistemic relativity that has, in the past been used coercively to force a particular worldview and understanding onto the problem situation. [This has been done both consciously and unconsciously.]
However, despite its popularity (in a postmodernist world) with some Systems Thinkers and many students of Systems Thinking, the Interpretivist View of System Conceptualisation is problematic. Among the problems of the Interpretivist view are that it is founded on a confusion and conflation of the appearance of things with the reality of things, it is logically incoherent and based on a fallacy, it prioritises the epistemic (what people think) over the ontic (how things are), it is anthropocentric or even egocentric, it is accepting of bias and prejudice, and it is not so useful as a scientific approach to systems thinking. This is unfortunate as much of Soft Systems Thinking assumes it is based on an Interpretivist, subjectivist worldview and philosophy -and if this were true it would cast severe doubt on the methodologies and methods of Soft Systems Thinking. However, in STREAMS, we argue that the right philosophical basis for Soft Systems Thinking is not some form of watered-down neo-Kantian Idealism but simply a postpositivism that incorporates the principle of epistemic fallibility. The confusion at the heart of the Interpretivist view is that of confusion the idea / conception / perception / appearence of a thing (the mental construct of "System") with the thing that causes the idea /conception / perception / appearance (the real-world System). The logical incoherence is that in order to discuss things - and do science about them - there must be a real-thing that is common to and causes the observers' perceptions and cognitions of the things (and a common language that incorporates well-understood and widely-understood references to the real-world things). John Searle call this "The Bad Argument" and identifies the fallacy as an ambiguity in the notion of "aware of" (see "Seeing Things As They Are") while Roy Bhaskar calls the fallacy the "Epistemic Fallacy" (see "A Realist Theory Of Science").
The Critical Realist View
The Critical Realist View transcends and goes beyond the false assumptions of both traditional positivism and interpretivism (as described above) - and attempts to present a logically coherent philosophical basis for the doing of science that incorporates the later 20th Century developments in the philosophy and sociology of Science. In this sense the Critical Realist view is a postpositivist one. And the foundations of Systems Thinking can be revised on the basis of this postpositivist view.
The critical realist and the ontological pluralist argue that both the traditional positivist and the subjectivist interpretivist make the same mistake: that of assuming that their perception / cognition is infallible. In the positivist case, this assumption along with the (anti-subjectivist) assumption of a mind-independent reality implies that observation is unproblematic and therefore his or her cognition creates a veridical, truthful concept of some segment of reality; and anyone with a different concept must be mistaken. In the subjectivist interpretivist case this assumption along with the subjectivist assumption that there is no mind-independent reality implies that everyone has their own personal, private truthful conception - and therefore any common reality must be socially-constructed by interaction between minds. However, the lesson of 300 years of science - and 3000 years of general cultural development - is that perception / cognition is not infallible. Both the Idealist (Interpretivist) and the Empricist (Naive Realist) make the same invalid assumption - that their perception is 100% veridical and reliable; recognising that this does not accord with actual experience the Idealist theorizes that perceptions are veridical and therefore have nothing to do with external reality - and therefore knowledge is relative and subjective, the empiricist theorizes that perceptions are veridical and therefore anything subjective cannot be 'real'; both conflate their inner, private subjective reality with external, public objective reality rather than question that their perceptions and knowledge can be 'fallible'. This is nothing but philosophical arrogance.
For the Critical Realist and Ontological Pluralist therefore "Systems" exist in the Physical World and "System Concepts" exist in the Mental world and the process of getting from perception of a "System" to the construction of a "System Concept" (a mental model) is fallible. To this must be added the fact that, as both Kuhn and Popper observed, all observation is 'theory-laden'. This means that in order to form any perception or cognition (of anything) a person (or other animal) actually begins with an 'intuitive' concept (or 'theory') of the thing. [In fact while many such concepts are labelled as 'intuitive' they are really not; they are learned uncritically as infants and reflect more the cultural prejudices of the past; however, there is a level of 'intuitive concepts' built into our genetic inheritance - where information about the world in which terrestrial life has lived for millions of years is encoded into our genes and expressed in the construction of plant and animal bodies. Hence humans see a narrow spectrum of electromagnetic radiation between the infra-red and ultra-violet. In passing, this simple observation also shows the falsity of naïve empiricism - there is much more in the world than plant and animal bodies (let alone human ones) are able to perceive.]
The question then arises as to how any knowledge can be objective - since personal perception and cognition are fallible. STREAMS follows Karl Popper's path in this: individual and small-group 'knowledge' - it hardly qualifies as 'knowledge' be any reasonable, rational criteria - is to be regarded as conjecture that needs to be subject to testing and refutation. Hence any concept, mental picture or model held by, or that is the product of, any individual or small group (whatever 'small' means in context) is regarded not as knowledge but as conjecture and hypothesis - and this applies to any system conceptualisation too.
For Popper the method of transforming group-relative conjecture into more objective knowledge is to work through the implications of the conjecture into empirical tests - experiments that can be performed or observations made - and anything failing the tests is not knowledge and anything not in-principle testable is 'unscientific' or 'pseudo-science' and not even a candidate for knowledge. This is Popper's 'Demarcation Criterion'. In STREAMS things are not as simple as Popper imagined - and his simple criterion has been criticised for being simplistic and not reflecting the fact that it takes more than a simple experimental failure to falsify a theory, and that theories don't change that way. In STREAMS we recognise that there is a good deal of 'logical distance' between any simple conjecture (any perception or preliminary system conceptualisation) and the full range of its implications - in all three worlds - the Physical, the Mental and the Abstractal. This logical distance cannot be traversed by any single individual or small group -it requires the attention of a group of minds with differing worldviews and 'ways-of-thinking'. Hence the criteria for a conjecture surviving falsification attempts and becoming a part of a body of knowledge expands to include: a) logical coherence with the existing body of knowledge - including justification by logical argument from well-established premises b) participating or featuring in a number of explanatory models and theories and c) having no implications that are counter to demonstrated empirical reality. The process is one of critical discourse among the stakeholders who have an interest in the systems and the system concepts involved.