2.3 THE SEMIOTICS OF ENGINEERING DESIGN
Preliminary design and formal design can be considered, respectively, phenomenological and scientific. A phenomenological model is holistic and works with the entire breadth of the engineering environment. A scientific model works in depth with a particular discipline. A phenomenological model attempts to make sense of a particular situation in its context, without abstracting to make generalizations. In contrast, scientific engineering models are formulated with measurable facts objectively obtained through repeatable methods. A phenomenological model is developed out of practice while a scientific model is based on theory.
Both phenomenological and scientific design of practical models can be considered as the application of knowledge, with knowledge constructed and accepted through a social process. This knowledge can be expressed specifically in the form of agreed-to facts and specifications, or in the form of standard operating procedures.
Purely scientific or purely phenomenological models have their drawbacks. The scientific model may become ossified with the assumption of the immutability of the data from which the model's theory was developed. The phenomenological model can be a closed system, depending on shared understanding which, once reached, may not be sensitive to changes in the world [de Neufville 86].
The resolution of the phenomenological and scientific dichotomy of modeling is attempted through the ideas of "critical theory." Critical theory contends that knowledge is developed through the interplay of theory and practice. At most, critical theory attempts to provide a reconciliation of the experiential base of values with the rigid and reproducible doctrines of science. At minimum it juxtaposes or combines empirical science and moral evaluation.
In the formulation of public policies that require the use of scientific model, the disparities between the phenomenological world within which the politician operates and the positivistic world of the expert scientist become more pronounced. Jurgen Habermas proposed the three models of the relationship between the politician and the expert: the Decisionistic Model, the Technocratic Model, and the Practical-Dialectical model [Dallmayr 86].
In the Decisionistic Model the politician has the ultimate authority in defining policy. This is undoubtedly the model in effect in most societies. Research is often at the mercy of the politician, and the science behind the research is often in a form unavailable for ready assimilation by the politician. It is also possible that applied analysis may be instrumental to the ends of the politician. It becomes a "handmaiden" of organizations with the analysis adjusted to meet fixed ends [MacRae 86].
In the Technocratic Model, the politician is an agent of the scientific intelligensia, with rational and pure scientific methods defining policy. Such a model, even when employed in limited cases, often runs afoul of the communities upon which the policy is imposed because pure science is often divorced through abstraction from the values and intents of the community. Separation of values from the formal model strengthens the science but weakens the applicability back to the original culture. It is argued by [MacRae 86] that without a value-based set of questions underlying the science, the application of this science is deprived of rational ethos and thus its strength.
The Practical-Dialectical model is also known as the Pragmatistic model and involves interaction between expert and politician. Practical discourse attempts to provide a method for an organized and disciplined dialogue between the politician and expert and is the subject of seminal research in the field of policy analysis.
Practical discourse advocates disciplined normative deliberation. There is an underlying recognition that "where scientific judgment can be based on a computational algorithm, the validity of a normative policy argument is in the final analysis determined by the communicative power or persuasive force it has for its audience." [Fischer 86]. In practice, the technical interests of empirical science must be evaluated in terms of the ethical values of society.
Most research in engineering design diligently applies the positivistic approach. Engineering design, with its fixed goal of producing viable artifacts, is in fact applied analysis. In the positivistic approach, engineering separates the values of the system context from the facts. However, in the context of engineering design, there is the normative, experiential base of knowledge, some of which is not easily captured or expressed as computable facts. This base may have legitimate technical content or it may be more of a contextual nature.
Taking Habermas's models of relations between politician and expert, let the designer, with the capability for a breadth of understanding, be posed as a politician. Other agents of the system stand as experts, each with a depth of understanding beyond that of the designer.
In the decisionistic model, designers take advice from the experts, make their own decision, and incorporate the results into design. In a technocratic model of the CAD environment, the designer is ancillary to the incorporation of recommendations into the design; technology drives the decision-making process.
The pragmatistic model mediates between the politics of the designer and the autocratic technical decisions of the machine, providing for critical interaction between the designer-as-politician and the advising-agent-as-expert. The pragmatistic model is more than a hybrid; it shifts the use of computers from the computational nature of decision-making to the dialectical nature of communication and consensus.