by Hugh M. Lewis

For every particular problem, there is a single ideal solution, no matter how complex, even though in reality an ideal solution can never be absolutely achieved. For any general class of problem sets, there is some general class of ideal type solutions specific to that class, no matter how complex and even though in reality such ideal solutions cannot ever be achieved. The problem of fitting solutions to problems, of defining problems and then designing solutions to address to such problems, represents the realm of applied systems. 

From a human standpoint, the primary function of any applied system is to serve the adaptive or alternatively the reproductive success of human beings, in the structure of the long run. This sets up a strange dilemma of long-term objectives and possible alternative strategies defining and constraining the development of applied systems--for it is entirely possible that a relative, "local" solution may work for a specific problem or kind of problem in the immediate term, but not represent the best possible or most optimal solution for the general class of problems represented in the larger frame of reference. We must therefore admit a certain systems based relativity of application, that the value of any solution must be measured ultimately by standards that are contextually defined by a larger frame of reference, and not by standards that are relatively defined by the immediate context of application. We would admit for instance that killing the last remaining pair of deer on earth might save ourselves from immediate starvation if we were snowbound on a mountain top, but that in the larger frame of reference of the extinction of an entire species, our own acquiescence to our cold fate on the mountain top might be a small sacrifice to pay for the good of the larger good of biological systems, knowing that there would be many other humans to replace us, but no other deer to replace the ones we would eat.

All human development issues, whether we are talking about global political-economic development, or we are talking about social and individual human development, remains centrally in the realm and strategic/tactical focus of applied systems theory and practice. Approaching the general problem of development therefore, as a cross-disciplinary academic concern, leads us invariably to the problematics of applied systems and the application of general or specific systems to diverse kinds of problem sets.

From a general systems standpoint, we might approach the problem as attempting to understand what common factors and issues arise in the resolution of a broad range of problem sets, and trying to distill from such an inventory a common structure of systems pattern in the application of general solutions to general problems.

I define applied systems as a subclass of alternative or artificial systems, or those systems that are human-made or that are constructions or products of our construction efforts. There are various kinds of applied systems, but we may say in general that they tend to be heterogeneous and they tend to the solution of certain specific kinds of problem sets at whatever level of their functional engagement in the larger scheme of human systems. As artificial systems these applied systems tend to be complex, to tackle problems without finite or clear solutions, and to function in the manner of all working systems, to obtain to some level of optimal operational efficiencies that are permitted by their design.

I would not now claim that all applied systems are isomorphic with all artificial or human-made systems. Artificial systems are a subset of a broader range of applied or alternative systems, and these encompass other possibilities than just human-made ones. One can make the argument for instance that biological systems are strictly speaking "applied" systems, and that evolution provides alternative systems to the general problems of the survival and reproductive success of living systems. The intentionality structures and deliberateness of such natural systems is obviously lacking, unless one posits the existence of a divine creator, but it is clear they are following an inherent logic and a form of informational transmission (genetic, environmental) that results in the creation of new systems previously unknown or non-existent. Certainly the primates alone would be enough to test our convictions about the exclusive human prerogative over applied systems.

Thus applied systems are by definition almost developmental systems that are subject to growth and possibly regeneration, whether or not they are self-organizational or deliberate. Just as real system are subject to change and eventual termination, so are applied systems, as longer-term frameworks featuring succession and replication, subject to comparable constraints of change and eventual extinction of types coinciding with the creation of new types.

We may stratify applied systems on the basis of their functionality and relative specificity of focus, as well as upon the basis of the level at which they are intentionally operational. An automobile engine is a kind of applied system, and serves as a archetypical model of such an applied system. 

At another level we have the car itself, including the engine, as an applied system at a different level of consideration or functional application. If we back away by one more degree, we see that the car that travels on roads, containing a driver and so many passengers, constitutes part of a larger system of highways and roads that is occupied by numerous cars. We may look at each of these systems as nested within the other, and as occurring at different levels of pattern and analysis. Within the framework of a larger highway system, the individual car, and the engine that is central part of the car, constitute subsystems at their respective levels of analysis, and we know well from experience that a network as large as a highway system contains multiple, indeed a myriad, of subsystems and sub-subsystems.

All applied and real systems exhibit non-random pattern or order that permit some level of operational efficiency to occur on a regular basis. Such a system obtains a state of dynamic equilibrium with its environment when it is said to be functioning normally at its own levels of relative efficiency.

Applied systems may be defined, therefore, as alternative real systems that deal in a functional manner with delimited problem sets, with the application system serving as a partial or imperfect but reasonable solution to such a problem set.

Each applied system, seen from the standpoint of the problem set it is intentionally designed to resolve, may be said to have an ideal prototype that represents the most optimal solution, the single best solution, to the problem set in consideration, if not in a specific sense, then in a general sense of the larger problem set that an individual problem or situation represents and stands for. Therefore all improvements to applied systems to certain delimited problem sets may be said to be developmentally streamlined towards the single best, most optimal solution set that is ideally applicable to that kind or type of problem set. It is predicted therefore that as applied systems develop and become developmentally streamlined over time, different instances or kinds of applied systems dealing with the same general kinds of problem sets will tend towards evolutionary or developmental convergence to a common prototypical form that represents such an optimal solution set.

Because of the complexities involved in all systems, even seemingly simple ones, neither the problem nor the solution set may be said to be simple to define, and thus are themselves complex at a level that may be said to be "infinitely complex." There are two outcomes of this general constraint. First, no matter how much development is involved in an applied systems framework, no matter how much streamlining and developmental convergence that may be achieved in the form and function of an applied system to a particular kind of problem, no perfect or one ideal solution will ever be obtained that will be sufficient for a particular kind of problem, much less for all variants of the general problem set. Secondly, the developmental trajectory of any particular design pathway will tend to achieve some optimal equilibrium state in adaptation to the problem set, that will be the average efficiency that is realizable by that particular problem set over the long run.

The problem of applied systems leads us automatically to consider the process of the developmental elaboration and increasing differentiation of such systems, which is complementary to the process of the developmental convergence of different systems upon a common streamlined prototype. Convergence and developmental divergence through differentiated elaboration are fundamentally, dialectically contraposed processes of all applied systems.

Differentiation must be considered a natural outcome of all applied systems development, since no two real systems are ever exactly alike, and there is a continuous process of reproduction of systems leading to variation of form and function. This can be considered to be inherent to the underdetermined complexity of such systems that permits multiple variation to occur within the same form and framework.

General Systems Essays, Vol. I