Design: One, but in different forms

Willemien Visser

Source-linked AI summary

Design: One, but in different forms

Willemien Visser

arXiv:0708.1725v2Aug 13, 2007cs.HC

TL;DR

The paper addresses whether design has common cognitive characteristics across situations while also taking different forms. It reviews cognitive design research and proposes candidate variables related to the design process, designers, and artefact. The review supports distinctive commonalities across design activities, while the claim about a fourth member—different forms of design—remains hypothetical and requires further research.

Problem
The paper examines whether differences between design activities across situations are compatible with treating design as a distinctive generic cognitive activity.
Method
The paper reviews cognitive design research and examines candidate variables underlying differences between design situations.
Results
The review provides support for the hypothesis that design has distinctive commonalities across forms, while the claim about different forms of design remains hypothetical.
Takeaways & Limitations
Design can be considered a specific generic cognitive activity while recognizing that its implementation may vary with the design situation.
Takeaways & Limitations
The hypothesis requires more research because empirical evidence is scarce, and the proposed fourth member remains completely hypothetical.

Abstract

from arXiv · show

This overview paper defends an augmented cognitively oriented generic-design hypothesis: there are both significant similarities between the design activities implemented in different situations and crucial differences between these and other cognitive activities; yet, characteristics of a design situation (related to the design process, the designers, and the artefact) introduce specificities in the corresponding cognitive activities and structures that are used, and in the resulting designs. We thus augment the classical generic-design hypothesis with that of different forms of designing. We review the data available in the cognitive design research literature and propose a series of candidates underlying such forms of design, outlining a number of directions requiring further elaboration.

activity unrecognisable as design."

The paper distinguishes a science of design from a design science, framing design research as systematic investigation while leaving the nature of design open. It situates this distinction within ongoing debate about design’s relation to science.

The design-research community has debated whether design science should be developed or design and science should remain clearly distinguished.
Design research distinguishes a science of design from a design science.A science of design uses systematic, reliable methods to improve understanding of design, whereas a design science concerns regular phenomena of systems and the design process.
The paper treats design as a subject whose nature remains open to interpretation rather than assuming that design itself is necessarily a scientific activity.

also Cross, 2001b)

The paper replaces domain-based generality with an augmented generic-design hypothesis: design situations differ, and those differences can produce different forms of design without eliminating commonalities.

The paper distinguishes domain independence from its broader focus on design situations.Here, domains are generally equated with practice disciplines such as engineering, architecture, computer science, or product design.
A design situation is characterized along three main dimensions: the design process, the designers, and the artefact.
The augmented generic-design hypothesis predicts different forms of design in different design situations when design context is retained.

1. Design is one: commonalities between designing in different situations

Cognitive design research characterizes design as a distinctive cognitive activity with recurring features across domains, while later work revises and extends parts of Simon’s account.

The literature extends or revises Simon’s characterization of design across a series of cognitive features rather than uniformly contradicting it.
Design is treated as a cognitive activity rather than a professional status, applying to anyone who devises actions to change existing situations into preferred ones.
The paper argues that designing is more appropriately qualified as constructing representations, although design can formally be considered problem solving.
Design problems are commonly characterized as ill-defined, with no definitive formulation and potentially multiple solutions.
Design activity is described as opportunistically organized, and designers do not systematically follow prescriptive distinctions such as analysis and synthesis.
Designers often work by satisficing, accepting satisfactory solutions rather than calculating an optimum.

(vi) Design generally involves complex problems that are rarely decomposable into independent

Design commonly involves complex, interdependent problems that resist systematic decomposition, while designers use early solution ideas, multiple acceptable outcomes, negotiated evaluation, and knowledge reuse.

(vi) Design generally involves complex problems that are rarely decomposable into independent: Design problems are rarely decomposed systematically into independent subproblems, and one component may admit multiple decompositions.
(vi) Design generally involves complex problems that are rarely decomposable into independent: Architects may use idiosyncratic decomposition and integration strategies, whereas electronic and mechanical design can have theoretically determined part interactions.
(vi) Design generally involves complex problems that are rarely decomposable into independent: Designers often begin with a few objectives that form an initial solution kernel, also called a primary generator or guiding theme.
(vi) Design generally involves complex problems that are rarely decomposable into independent: Design problems typically have several acceptable solutions rather than one correct solution.
(vi) Design generally involves complex problems that are rarely decomposable into independent: Design proposals are evaluated through technical criteria, negotiation, compromise, and possible revision of the criteria and procedures themselves.
(vi) Design generally involves complex problems that are rarely decomposable into independent: Analogical reuse of knowledge from previous projects is a central approach reported across cognitive design studies, combined with generic knowledge.

(xi) Design activity is mostly opportunistically organised: designers proceed in a non-systematic,

Across design domains, studies support important commonalities in design activity, especially its opportunistic rather than uniformly systematic organization.

Designers combine top-down, bottom-up, breadth-first, and depth-first modes rather than following one fixed strategy.These modes may be mixed locally and occasionally throughout a process.
Opportunistic design can include hierarchical local episodes even when the activity is not globally hierarchical.
Designers’ paths through design phases vary, and problem structuring may recur periodically rather than occurring only at the beginning.
Evidence for generic design remains limited because systematic empirical research across domains is scarce.
The literature identifies commonalities across design activities in software, architecture, naming, and letter writing.

software and mechanical design—in order to develop domain-independent design knowledge. The authors

Comparisons across software, architectural, mechanical, and product design suggest shared features, while the evidence base is constrained by how studies define and observe design situations.

Studies compare design domains to identify knowledge that may apply beyond a single discipline.
Some authors report that differences between design disciplines concern mainly differences in terminology.
The cited cross-domain study did not observe designers at work, instead using interviews and project documentation.
Existing comparisons treated different situations as different disciplinary domains, not differences in age, sex, expertise, or working conditions.

2. Design is different from nondesign

The paper argues that design has characteristics distinguishing it from other cognitive activities, while acknowledging that direct empirical evidence for this distinction is limited.

The classical novice–expert depth-first versus breadth-first distinction is not displayed systematically in design.
Design expertise includes problem framing, solution-focused reasoning, frequent shifts between cognitive activities, and commitment to early solution concepts.
Lawson’s comparison found science students analyzing problem structure while design students generated successive acceptable solutions.
Comparisons with science and discovery reveal substantial cognitive similarities, including problem solving, pattern recognition, and analogical reasoning.
The paper distinguishes design from science by its goal of creating a new artefact, whereas science seeks explanation of an existing function.
Except for expertise, the paper finds little direct evidence that design differs significantly from nondesign.

3. Design is one, but takes different forms

Design shares general cognitive characteristics but takes different forms shaped by the design process, the designer, and the artefact.

3. Design is one, but takes different forms: Akin proposes that architects tend toward creative strategies while engineers tend toward routine design, but product designers may also act creatively and flexibly.
3. Design is one, but takes different forms: Architects may use individual decomposition and ad hoc complexity-management strategies, contrasted with standardized schemata and predetermined engineering procedures.
3. Design is one, but takes different forms: Evidence on premature commitment varies with designers’ expertise, collaboration, and whether situations are natural or artificially restricted.
3. Design is one, but takes different forms: Professional design projects tend to be organized opportunistically, whereas simpler restricted situations more often permit systematic decomposition.
3. Design is one, but takes different forms: Reuse of elements from previous projects appears to be a specific professional design approach, though it also occurs experimentally.
3. Design is one, but takes different forms: Design setting and project complexity influence activity, and user considerations can guide professional design decisions.
3. Design is one, but takes different forms: The paper proposes three dimensions underlying differences between forms of design: the design process, the designer, and the artefact.

using rough guidelines, intuition and experiential knowledge."

Design activities differ across situations because representations, evaluation methods, users, expertise, and task structure shape designers’ cognitive processing and resulting designs.

Different representation formats activate different perceptual and cognitive processes, contributing to differences between design situations.Physical and symbolic artefacts may rely on different representations, while sketches can facilitate visualisation, manipulation, simulation, and evaluation.
Design domains differ in how users understand intermediate and final representations, producing different collaboration, testing, and accountability patterns.Architectural drawings may be understandable to end-users, unlike more invisible design representations.
Evaluation ranges from relatively objective performance measures in engineering to more subjective criteria such as aesthetics in other domains.Simulation, both physical and mental, occupies an intermediate position between objective and subjective evaluation.
Designers’ expertise and project routine character influence how knowledge and analogies are organised and used, affecting opportunities for reuse.Comparative studies linked routine character to analogy use at both action-execution and action-management levels.

given profession, its educational philosophy and the predisposition of its participants."

Forms of designing vary with characteristics of the artefact and its social context, including social embeddedness, artefact type, evolution, and temporal behaviour.

The paper identifies social embeddedness, artefact type, and artefact evolution as three variables distinguishing forms of designing.Artefact type is illustrated through the contrast between structures and processes.
Social embeddedness concerns designers’ or stakeholders’ representations of an artefact’s insertion and future position within its environment.It is not an externally defined task property but a characteristic of how people represent the project.
The paper treats social embeddedness as dependent on designers’ or stakeholders’ views, so it is not restricted to conventionally social or societal problems.Even activities such as meal planning can become socially embedded depending on views of guests and consequences.
Socially embedded problems may have multiple solutions judged as more or less appropriate or acceptable according to adopted criteria.The influence of social embeddedness on activity is described as probably similar to that of ill-definedness.
Structures and processes may be represented and processed differently because processes involve spatial and temporal constraints and behaviour over time.Interactive systems are contrasted with houses on whether they are designed to have behaviour over time.

4. Conclusion

The paper concludes that design has distinctive cognitive characteristics, shares commonalities across situations, and nevertheless takes different forms. It presents this augmented hypothesis as promising but still requiring systematic empirical testing.

4. Conclusion: The hypothesis remains hypothetical because empirical evidence is rare and dispersed, requiring new systematic research, preferably in professional design situations.The paper explicitly calls for research comparable to Goel and Pirolli’s work but conducted in real design contexts.
4. Conclusion: The review proposes candidate variables spanning design situations, designers, artefacts, representations, expertise, and project routine character.The paper synthesises roughly 25 years of cognitive design research and reports precise data for several comparisons.
4. Conclusion: Comparative evidence concerns domains, professional versus artificially restricted settings, design methods, expertise, analogy use, and designer idiosyncrasies.These findings provide empirical material relevant to the proposed differences between forms of designing.
4. Conclusion: Representational structures and activities are proposed as particularly influential because design is viewed as the construction of representations.The paper hypothesises that sketches and drawings may facilitate simulation and evaluation through easier visualisation and manipulation.
4. Conclusion: Design support may need modalities tailored to the representational activities and structures used in specific tasks and situations.The paper identifies technological or methodological support as possible directions for further development.
4. Conclusion: The augmented generic-design hypothesis combines design’s distinctive cognitive characteristics, cross-situation commonalities, and situation-dependent differences.Its fourth component proposes that commonalities across forms remain sufficiently distinctive to treat design as a specific generic activity.

574). Berlin: Springer.

This section lists references spanning cognitive design research, design methodology, expertise, representations, and human-computer interaction.

The bibliography includes work on representations, diagrams, sketches, and external cognition in design.
The references cover cognitive studies of design activities and design processes.
Several entries address expertise, including expert knowledge, expertise variation, and expertise in design.

Kant, E. (1985). Understanding and automating algorithm design. IEEE Transactions on Software

This section lists references on design cognition across software, architectural, engineering, and interactive-systems contexts.

The references include studies of cognitive processes in logical, architectural, and software design.
Several entries examine design strategies, methods, paradigms, and problem-solving actions.
The bibliography also covers analogy, co-evolution, and interactive systems in design.

Méhier, C. (2005). Intégrer des contraintes environnementales dans la conception. L'éco-conception [Integrate

This section gathers references on design theory, ill-structured problems, design processes, representations, and empirical studies across engineering and cognitive design.

The references address systematic engineering design and design-process models.
Several works concern ill-defined or ill-structured problems and problem solving in design.
The bibliography includes research on representations, sketching, analogy, and design knowledge.

Visser, W. (1994a). Organisation of design activities: Opportunistic, with hierarchical episodes. Interacting

This section lists references on organizing design activities, professional design, design representations, expertise, and cognitive approaches to design.

The references include studies of planning, organization, and opportunistic behavior in design activities.
Several entries focus on professional programming, software design strategies, and user-centred requirements.
The bibliography also covers collective design, design knowledge modeling, and cognitive representations.