Fil Salustri's Design Site

Site Tools


Table of Contents

Designing as Specifying

What is the difference between designing and creating specifications?

Designing usually starts with broad qualitative problem elements that need to be reduced to structured, solvable elements.

  • i.e. a problem specification.

Designing also results in a set of artifacts that specify how to implement something that will cause a (presumably beneficial) change in each dimension of the problem.

  • i.e. a solution specification.

Thus, designing is (partly) specifying.

The dictionary gives this general definition: “an act of describing or identifying something precisely or of stating a precise requirement.”

  • Thesaurus entries include: “statement, identification, definition, description, setting out, framing, designation, detailing, enumeration; stipulation, prescription.”

A design is, however, rarely as antiseptic as what is commonly thought of as a “specification.”

  • A design is just a model of something else, and thus imperfect by definition.
    • Therefore, the design specification will also be imperfect.
  • By capturing the qualitative as well as the technical aspects of a design, a specification provides a better sense of the design intent to the agents who will have to manifest/implement/manufacture it.
  • We mean qualitative in the sense of a thing having certain recognizable qualities.
    • Strictly speaking, such qualities are not subject to the “hard” quantifications that engineers prefer.
    • However, at a coarse level, some kind of quantification is possible.
      • For instance, many users have commented favourably on the aesthetics of the new Apple iPhone.
        • Though its aesthetic cannot be quantified in the usual scientific way, there are relative measures that can be applied.
          • e.g. as when significantly large groups say that the device is the “most beautiful” of all similar products.

Depending on the nature of the design problem, the qualitative design information may more accurately capture the spirit and intent of the design than might the technical information.

  • If errors in the technical specification are found, they might be addressed by changes that bring them more in line with the qualitative specifications.

However, the qualitative information that must be captured to truly specify a design cannot just be specified.

  • There is no method of which we are aware that can accurately and reliably capture this sometimes ethereal sense of a particular design.
  • Nonetheless, if we could specify this kind of information, it seems quite certain to result in better manifestations of those designs.

Therefore, reconciling the dry and technical notion of specification with the broader and more informal articulation of qualities would be an important achievement.

In this regard, the technical aspects of specification might be used as grounding for other kinds of specification.

  • Compared to the broader kinds of design specification, technical specification is quite well understood, and there is no doubt of the necessity of technical specifications for the sake of implementing designs.
  • At very least, methods of technical specification allow one to identify those aspects of a design for which implementation is relatively well understood.
  • Knowing this allows us to focus, by elimination, on those aspects of a design not otherwise covered and that go beyond the technical.

Open research questions here include:

  • What information is missing from the technical specification of a design that is necessary nonetheless to convey the design in toto to others?
  • How can one organize the non-technical specification in relation to the (relatively) well-understood technical information?
  • What mismatches or apparent contradictions arise from such a categorization?
  • How does non-technical information help to capture the “spirit” of a design for the sake of better implementing it?

We suggest that a boundary layer occurs between designing and specifying with respect to the amount of qualitative information that must be captured to represent the design accurately and completely.

  • In this view, specifications that have no qualitative elements were not an artifact of a design process.

See Also


<refnotes> notes-separator: none </refnotes>

research/designing_as_specifying.txt · Last modified: 2020.03.12 13:30 (external edit)