paper: interface metaphors

Interface Metaphors and User Interface Design
John M. Carroll, Robert L. Mack, Wendy A. Kellogg
Handbook of Human Computer Interaction, 1988

This paper examines the use of metaphor as a device for framing and understanding user interface designs. It reviews operational, structural, and pragmatic views on metaphor and proposes a metaphor design methodology. In short the operational approach concerns the measurable behavioral effects of applying metaphor; structural analyses attempt to define, formalize, and abstract metaphors; and the pragmatic approach views metaphors in context – including the goals motivating metaphor use and the affective effects of metaphor. The proposed design methodology consists of 4 phases: identifying candidate metaphors, elaborating source and target domain matches, identifying metaphorical mismatches, and finally designing fixes for these mismatches. Strangely, this paper makes absolutely no mention whatsoever of George Lakoff’s influential work on conceptual metaphor, which I’m almost certain had been published prior to this article. My outlined notes follow below.

• Introduction
• Design interface actions, procedures and concepts to exploit specific prior knowledge that users have of other domains
• Metaphors as alternative to reducing the absolute complexity
• Metaphors, by definition, must provide imperfect mappings to their target domains
• (otherwise, it would be the item it mapped to)
• Inevitable mismatches are a source of new complexities for users
• Metaphors often apply unevenly within a software domain
• Composite metaphors are common
• Desktop metaphor x Direct manipulation
• Learning by analogy, one of the most basic approaches to learning
• Operational Approaches to Metaphor
• Focus on demonstrating measurable behavioral effects from employing metaphor
• Raise questions of precisely how metaphor operates in the mind
• Offers no principles that predict “good” and “bad” metaphors extensibly
• Offers no principled definition of what a metaphor actually is
• Structural Approaches to Metaphor
• Develop representational descriptions of metaphors
• … as relations among primitives in the source and the target domain
• Douglas and Moran - 1983
• Structural analysis of typewriter metaphor
• Domain operators as structural primitives
• type character -> type character
• space bar -> insert blank space
• Problems causes by mismatches – 62/105 collected errors
• Gentner’s structure mapping theory
• Holistic mappings between graph theoretic representations
• Doesn’t consider individual entities piece-meal
• Rutherford example – planet/sun -> electron/nucleus
• Attribute predicates generally fail to map
• Characteristics of structural formulation
• Base specificity – how well understood is source domain (bounds usefulness)
• Clarity – precision of node correspondences across mapping (e.g. 1-1 or 1-many)
• Richness – density of predicates
• Abstractness – level relations compromising mapping are defined (node mappings vs. predicate mappings)
• Systematicity – extent mapped relations are mutually constrained by membership in structure of relations
• Exhaustiveness – directional surjectivity
• Transparency – how easy to tell which source relations get mapped to target
• Scope – extensibility of the mapping
• Relate metaphors to cognitive aspects of use
• Expressive (literary) metaphors
• Explanatory (scientific) metaphors
• Both rated better when clarity is higher, but richness more important to expressive metaphors
• What is appropriate grain for mapping?
• What operators, what relations should be defined for a particular metaphorical mapping?
• Structure-mapping analysis can prove insufficiently objective
• Misses external consequences of metaphor
• E.g. interpersonal attraction -> ionic bonding in chemistry may make chemistry more interesting
• Pragmatic Approaches to Metaphor
• Focus on use in the context and complexity of real-world situations
• Emphasize the intentional use of metaphor to an end (i.e. aid users learning less familiar domain)
• Context
• Goals associated with metaphor
• Characteristics beyond similarity basis
• Incompleteness
• Involvement in compositions
• Acts as a filter on structural analyses
• Suggest how inevitable structural flaws can play a useful role
• Context of metaphor in use
• Analysis of metaphors must rest on empirical task analysis of what users actually do
• Metaphors can be used to draw attention to specific features of target domain
• Thereby motivate further thought about comparison with source
• Often primary purpose of literary metaphor
• Interface metaphors can pose questions and open new possibilities
• Metaphor mismatches
• Mismatch and its resolution can elaborate an accurate conceptual understanding of the system
• As negative exemplars can help to clarify a new concept
• Composite metaphors
• Mismatched of incomplete correspondence sometimes addressed by composite metaphors
• Useful beyond increasing coverage of a target domain
• May help generate more and different kinds of inferences about target domain
• May aid quick convergence of integrated understanding of target domain
• Toward a Theory of Metaphor
• Competence theories of metaphor: operational, structural
• Performance theories of metaphor: pragmatic analyses
• Suggests need for integrated theory of metaphor
• 3 phases of metaphorical reasoning
• instantiation
• recognition or retrieval of something known (potential source analog)
• automatic and holistic activation process, analytically incomplete
• elaboration
• generation of inferences about how source can be applied
• pragmatically guided structure mapping, identifying relevant predicates
• consolidation
• consolidate elaborated metaphor into a mental model of the target domain
• integrates partial mappings into a single representation of target domain
• Integrated understanding of target is not couched in a metaphor, but a mental model
• “Distinction between models and metaphors one of open-endedness, incompleteness, and inconsistent validity of metaphoric comparisons versus the explicitness and comprehensiveness and validity of the models which the successful learner will ultimately obtain.”
• Designing with Metaphors
• Use of metaphors is currently haphazard. Can we systematize it?
• Structured methodology for interface metaphors
• Identify candidate metaphor or composite metaphors
• Sources include
• Predecessor tools and systems
• Human propensities
• Sheer invention
• Levels for metaphor
• Tasks – what people do (goals and subgoals)
• Methods – how tasks are accomplished
• Appearance – look and feel
• Detail metaphor/software matches w.r.t. representative user scenarios
• Match metaphors against the levels listed above
• Can help to enumerate the objects of the metaphor domain
• Can begin assessing ‘goodness’
• Identify likely mismatches and their implications
• Discrepancy in the software domain must be interpretable
• Mismatch should be isolable
• Salient alternative course in the target domain should be available
• Identify design strategies to help users manage mismatches
• Creating interface designs that encourage and support exploration is a key approach
• Users must be able to “recover” from metaphor mismatches
• Progressive disclosure of advanced functions
• Iterative design
• Composite metaphors can be a design solution for resolving mismatches
• Help system can anticipate mismatches as well
• Conclusion
• Metaphors draw correspondences, and have motivational and affective consequences for users
• They interact with an frame user’s problem-solving efforts in learning the target domain
• Ultimate problem is for user to develop a mental model of the target domain itself.
• There is no predictive theory of metaphor – they must be designed on a case by case basis, and carefully analyzed and evaluated
• This endemic not just to interface metaphors, but user interface design itself