Design Systems as Enabling Contexts for Designerly Application: A Case Study after Designer Withdrawal

Slides from a 15-minute presentation at dmi:Academic Design Management Conference 2026 (Georgia Tech College of Design, Atlanta, 12 August 2026).

Design Enablement, the cultivation of designerly ways of knowing within organizations, has been theorized through a designer-led pattern: professional designers activate three directional functions (Pulling, Pushing, Demonstrating) among non-designers through co-presence. Whether the design system produced in such an engagement keeps functioning as an enabling context after the designer withdraws has not been examined empirically.

This exploratory case study takes a mid-sized Japanese software company (about 100 employees) with no professional designers, where a design system was built through an external designer engagement in 2024. About 14 months after withdrawal, semi-structured interviews with the design system owner and with an engineer maintaining it were analyzed using SCAT (Steps for Coding and Theorization).

The design system appears to function as an enabling context for designerly application, reasoning competently within the solution space the system has modeled. Under the conditions observed it does not support designerly discovery, reshaping the context itself when problems exceed its scope. The feedback loop circulated at the how layer but not at the why layer. The talk proposes three hypotheses, two further properties of the enabling context, and a revision of Design Enablement that distinguishes two approach patterns (designer-led and environment-led) and two capabilities (designerly application and designerly discovery).

Paper: Akira Motomura, "Design Systems as Enabling Contexts for Designerly Application: A Case Study after Designer Withdrawal," dmi:Academic Design Management Conference 2026.

dmi:Academic Design Management Conference 2026（米国ジョージア工科大学デザイン学部、2026年8月12日）での15分の口頭発表のスライドです。

デザイン・イネーブルメント（組織のなかにデザイナー的な知の様式を育てること）は、専門デザイナーが同じ場にいることを前提に、非デザイナーに対して三方向の機能（Pulling、Pushing、Demonstrating）を働かせる「デザイナー主導のパターン」として理論化されてきました。その関与のなかで作られたデザインシステムが、デザイナーが去ったあともイネーブリング・コンテクストとして機能し続けるのかは、実証的に検討されていません。

対象は、専門デザイナーを持たない日本の中規模ソフトウェア企業（従業員約100名）で、2024年に外部デザイナーの関与によってデザインシステムを構築した事例です。関与終了から約14か月後に、デザインシステムのオーナーと保守を担うエンジニアへ半構造化インタビューを行い、SCAT（Steps for Coding and Theorization）で分析しました。

デザインシステムは、システムがすでにモデル化した解空間のなかで筋の通った推論を行う「デザイナー的な適用（designerly application）」については、イネーブリング・コンテクストとして機能していると読めます。一方、観察された条件のもとでは、問題がその範囲を超えたときに文脈そのものを組み替える「デザイナー的な発見（designerly discovery）」は支えていませんでした。フィードバックループは how の層では循環していましたが、why の層では循環していませんでした。発表では3つの仮説と、イネーブリング・コンテクストの2つの性質、そして2つのアプローチのパターン（デザイナー主導・環境主導）と2つの能力（適用・発見）を区別するデザイン・イネーブルメントの改訂案を示します。

論文：本村章「Design Systems as Enabling Contexts for Designerly Application: A Case Study after Designer Withdrawal」dmi:Academic Design Management Conference 2026

Transcript

1. dmi:Academic Design Management Conference 2026 ｜ Georgia Tech, Atlanta Design Systems as Enabling Contexts for Designerly Application A Case Study after Designer Withdrawal Akira Motomura, Spice Factory Inc. ｜ 12 August 2026
2. Speaker Akira Motomura Practice Studied communication design in the United States, worked at Dubberly Design Office in San Francisco, then led a fifty-person design organization at Yumemi. Now Chief Design Officer at Spice Factory, a design and engineering consulting firm in Tokyo, and a visiting researcher at the Research Center for Design Science, Ritsumeikan University. Research I move between practice and research on Design Enablement: the engagements I run become the cases I study. Four papers so far, and two ADMC presentations including this one. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 02
3. Contents Structure of this talk Five parts, following the paper. The first merges the introduction and the theory, because the theory is what makes the question askable. 01 Introduction and Theory 02 Case and Method 03 Findings 04 Discussion 05 Conclusion © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 03
4. Introduction and Theory © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 04
5. Introduction and Theory The three directional functions of Design Enablement Motomura and Yaegashi introduced Design Enablement at ADMC 2024: a set of directional functions through which professional designers support the development of design capability among non-designers. Pulling Eliciting from non-designers what they seek to realize. This function requires trust. Pushing Transferring explicit knowledge through documentation, guidelines and other structured means. Demonstrating Transferring tacit knowledge through practice-based co-work, internalized by direct experience. Motomura and Yaegashi (2024). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 05
6. Introduction and Theory The design system as practitioners meet it Components, styles and patterns held in one place. Industry discourse has framed this as a productivity tool, and academic inquiry into it remains limited. Drawn for this talk, not a screenshot of the case. Industry discourse frames design systems as repositories of reusable components (Handal et al., 2022; Lamine & Cheng, 2022). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 06
7. Introduction and Theory The operational view of design systems Research has no agreed definition. What the accounts share is that the design system is infrastructure, valued for efficiency and criticized for narrowing what practitioners imagine. Design systems have become pervasive infrastructure in digital product development, yet academic inquiry remains limited and there is no agreed definition. Industry discourse has largely framed them as productivity tools: repositories of reusable components (Handal et al., 2022; Lamine & Cheng, 2022). The parts a design system holds — A shared repository of the assets, code, documentation, The gains and the costs — Valued as infrastructure, for the efficiency and consistency it guidelines, and principles a team uses to keep products consistent at scale (Werle, 2021) brings (Gustafsson, 2021) — Criticized for homogenizing design, at the cost of making unusual — Handal et al. (2022) and Lamine and Cheng (2022) add the design designs more difficult to create (Goree et al., 2021) philosophy, the interaction patterns, and the people who sustain the system Werle (2021); Handal et al. (2022); Lamine and Cheng (2022); Gustafsson (2021); Goree et al. (2021). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 07
8. Introduction and Theory Three primary elements of a design system Motomura and Takeda proposed these three inductively from the studies above. The list under each element is what those studies name, before the abstraction. 01 02 03 Design principles Guidelines and repositories Application processes Articulated intent and stance. Reusable assets and references. The practices that carry the system into — Motive and goals for adoption (Kholmatova, — UI assets, code, documentation (Werle, 2017; Mall, 2023) 2021) — Ideals and values, as philosophy (IBM, n.d.; Handal et al., 2022) — Reusable components (Handal et al., 2022; Lamine & Cheng, 2022) — Principles a team keeps (Werle, 2021) product work. — Design patterns as shared solutions (Alexander et al., 1977) — Plasticity, to fit a local context (Kholmatova, 2017) — Workshops and reviews (Motomura & Takeda, 2025) Motomura and Takeda (2025), Section V of the research note. Kholmatova, Mall, IBM and Alexander et al. are cited only in the note. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 08
9. Introduction and Theory Design Enablement mapped onto the design system Each element carries one enablement function. This correspondence is what makes it worth asking whether the artifact alone can do what the designer did. Design Enablement function Design system element Pulling Design principles elicit aspirations and local knowledge articulated intent and stance Pushing Guidelines and repositories transfer explicit knowledge reusable assets and references Demonstrating Application processes transfer tacit knowledge how the system reaches product work Motomura and Takeda (2025) name the three layers purpose and design principles, guidelines, and application to products (Figure 5-1 of the research note). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 09
10. Can design systems function as enabling contexts that allow non-designers to practice designerly ways of knowing, even in the absence of embedded professional designers? An exploratory question. This paper generates hypotheses rather than establishing causation. The three elements line up with the three functions, yet every prior case ran through designer co-presence. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 10
11. Introduction and Theory The enabling context and its five enablers von Krogh, Ichijo and Nonaka theorize knowledge enabling through five enablers. The five are not equivalent in kind: the fourth produces the space the other four work in. The definition The five enablers von Krogh, Ichijo and Nonaka (2000) define it as a shared von Krogh, Ichijo and Nonaka theorize knowledge enabling organizational space in which tacit knowledge can surface, be through five enablers. The five are not equivalent in kind. shared, and be converted into organizational knowledge. — Sustained by care-based relationships and institutional conditions — Knowledge activists are the members who catalyze and sustain the space 01 Instill a knowledge vision 02 Manage conversations 03 Mobilize knowledge activists 04 Create the right context, which produces the space itself 05 Globalize local knowledge von Krogh, Ichijo and Nonaka (2000, p. xi). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 11
12. Introduction and Theory Design Enablement, the design system, and the enablers This paper reads the design system as the space enabler 4 produces, so the three elements give that space a tangible form and the three functions run through them. Design Enablement function Design system element Knowledge enabling enabler Pulling: elicit aspirations and local Design principles Instill a knowledge vision (1) Pushing: transfer explicit knowledge Guidelines and repositories Globalize local knowledge (5) Demonstrating: transfer tacit knowledge Application processes Manage conversations (2) and mobilize knowledge knowledge activists (3) Table 1 of the paper. Motomura and Yaegashi (2024); Motomura and Takeda (2025); von Krogh, Ichijo and Nonaka (2000). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 12
13. Introduction and Theory With and without embedded professional designers Prior work documented the designer in the mediating position. This paper puts the artifact there instead, which is the hypothesis the case study is built to test. Designer-led: documented in prior work Environment-led: the hypothesis of this paper The designer's co-presence activates the three functions The design system itself takes the mediating role, with the three directly. functions operating through its own layers. — Pulling through stakeholder interviews — Pulling through design principles — Pushing through workshops formulating design principles — Pushing through guidelines and repositories — Demonstrating through joint construction of the component — Demonstrating through application processes library — Two engineers take the knowledge activist role instead — The designer is the knowledge activist who sustains the context Motomura and Yaegashi (2024); Takeda and Motomura (2025); von Krogh, Ichijo and Nonaka (2000) for knowledge activists. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 13
14. Case and Method © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 14
15. Case and Method The case details A mid-sized software company with no professional designers, and the six-month engagement that built the design system this study examines. A mid-sized software company The engagement, July to December 2024 About 100 employees, serving public-sector clients. The The author served as the external design consultant. The portfolio includes a flagship portal system and a specialized engagement followed the Design Enablement framework domain of map-based information systems. All product through three interwoven activities, which operated as layered development is carried out by engineers and project managers. modes rather than discrete phases. — Stakeholder interviews surfacing latent design needs and existing design legacies (Pulling) — Workshops formulating design principles (Pushing) — Joint construction of the component library and style layer (Demonstrating) Motomura and Yaegashi (2024) for the framework. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 15
16. Case and Method Reasons why this case was chosen Three conditions have to hold for the environment-led pattern to be observable at all, and this organization is where all three did. 01 02 03 A system built to the framework The designer then left Fourteen months of operation The author withdrew on completion. From By the primary interview in February 2026, The organization implemented a design that point the organization operated the the organization had run the system for system through a designer-led system with no professional designer, about 14 months, with two engineers engagement that followed Design which is what makes the environment-led acting as its knowledge activists. Enablement, so the artifact under study pattern observable at all. was built to carry the three functions. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 16
17. Case and Method The informants, the interviews, and the positionality Two engineers acting as the organization's knowledge activists, interviewed 14 months after the author, who had built the system, left the field. The informants The interviews Positionality Two engineers held the design system Both interviews were semi-structured The author built the design system under after the author left. Neither holds a and recorded with consent. Interviews, study. The dual role affords deep design title. transcripts and the SCAT analysis were all contextual understanding, and introduces in Japanese; the English came at the the risk of confirmatory bias. — Informant A: took part in the original construction, then held maintenance and lateral deployment writing stage. — A: 19 February 2026, about 100 minutes, in — Informant B: co-steward and day-to-day maintainer of the system person — B: 10 April 2026, about 60 minutes, online — The application/discovery distinction was disclosed 45 minutes in © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 — SCAT documents every analytical step — The interviews covered the postwithdrawal period only — Instances where the design system failed as an enabling context are reported, not set aside 17
18. Case and Method Method: SCAT (Steps for Coding and Theorization) SCAT was selected for its transparent, auditable procedure and for the invitation in step 3 to bring in concepts from outside the data, which suits an exploratory study. SCAT step Result for segment [B-16] Step 1: Noteworthy phrases how-to-use questions; feedback on the system itself; the principles underneath Step 2: Paraphrase usage-level questions arrive, but feedback on the system’s own principles does not; the loop returns at the level of application but not underlying rationale Step 3: External concepts first- and second-order feedback; how layer and why layer; feedback loop impairment Step 4: Thematic construct feedback circulating at the how layer but not the why layer Table 2 of the paper. Otani (2019). Segment [B-16] reads: “Questions about how to use the component library do come in. But feedback on the design system itself, on the principles underneath, doesn’t really reach us”. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 18
19. Findings © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 19
20. Findings The constructs, and where the two accounts meet Two informants produced 37 coded segments and 31 thematic constructs. Eight constructs pair across the two accounts; fourteen from Informant B have no counterpart. Figure 1 of the paper, split in two. Eight of the nine CA constructs pair with a CB construct; fourteen CB constructs extend the analysis instead. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 20
21. Findings 4.1 Finding 1: Preconditions from the Designer-Led Phase Before the findings about what the artifact did, a precondition: whether a design system capable of carrying designerly reasoning was produced in the first place. The baseline before the engagement The artifact the phase produced Design practice consisted largely of simply The system was completed and released for the carrying over the previous year's materials, without flagship portal, and the organization viewed the articulated intent [B-3]. Individual design interest rationale for building it favorably [A-1]. Against that remained sporadic and disconnected from any baseline it was a newly available artifact encoding organizational identity [B-4]. intentional design decisions. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 21
22. Findings 4.2 Finding 2: Partial Evidence of an Environment-Led Pattern Five dynamics appeared after the designer withdrew. Two of them are evidence for the pattern, and three of them mark where its reach stopped. 4.2.1 Emerging Designerly Reasoning 4.2.4 A Domain-Specific Boundary Condition Engineers argued decisions from what an action means, not A second team building map-based systems stayed outside [A- preference [A-8]. Informant B said the same [B-20]. 13]: a different component library [B-5] and contractual deadlines [B-17]. 4.2.2 Shared-Vocabulary Function under Organizational Scaling As the team grew more diverse, the system became the shared 4.2.5 Additional Dynamics from the Supplementary Interview vocabulary for design decisions [A-9], [B-13]. Client conversations became easier [B-14]. Templates were hard to locate [B-12], and AI-first workflows bypass the system [B- 4.2.3 Persistent Barriers 22]. Awareness near-universal, adoption about 50% [A-3], [B-9] © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 22
23. Findings 4.3 Finding 3: The Feedback Loop as a Structural Challenge Questions about how to use the components arrive; feedback on the principles underneath does not. Two structural conditions account for the unevenness. 4.3.1 Non-Institutionalized Knowledge Activists Informants A and B maintained the system and mediated between teams without formal mandate, dedicated time, or recognition. The work was done in the time left over [B-11]. 4.3.2 Hierarchical Invisibility of Design Practice Within engineering the work was recognized as design; at management level it was not [A-12]. An earlier attempt to establish a design role ended in reassignment to programming Figure 2 of the paper. Feedback reaches the how layer but not the why layer, where the system's scope and principles would need reconsidering [A-4], [B-16]. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 [A-6]. 23
24. Discussion © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 24
25. Discussion 5.1 Two Patterns of Design Enablement The case supported reasoning inside the solution space the system had modeled, and did not support reshaping that space when problems fell outside it. Designerly application Designerly discovery The capability to apply existing design principles to recognized The capability to reshape the enabling context itself when a problems through evidence-based reasoning: operating problem falls outside the scope it was built for. competently within the solution space that an existing enabling context has already modeled. — Engineers reasoned through action-meaning semantics [A-8] — The evidence is mixed: younger members' reuse [A-15] resembles senior copying — The system encoded the flagship portal's conventions, so the map-based domain fell outside its scope — Closing that gap meant reshaping the system, not applying it [A13], [B-5] — Informant B answered that this is not the role they are asked to fill [B-18] Dubberly and Pangaro (2023) for the second-order perspective. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 25
26. Discussion 5.2 Hypothesized Properties of Design Systems as Enabling Contexts 5.2.1 Visibility of Reasoning (H1) The enabling effect comes from visible reasoning, not from 5.2.3 Institutional Authorization of Design Work (H3) components alone. The effect seemed to come from seeing the Both feedback loop activation and designerly discovery require reasoning [A-8]. that reshaping the system be recognized as legitimate organizational activity [B-18], [B-20]. 5.2.2 The Design System as a Boundary Object (H2) The system mediates design-oriented communication (Star & 5.2.4 Additional Properties: Meta-Usability and Workflow Integration Griesemer, 1989), activated by organizational scaling and A context that cannot be located and entered cannot function bounded by domain fit [B-17]. [B-12]. Contexts built on human-mediated design work can be bypassed by AI-driven workflows [B-22]. Star and Griesemer (1989); Nonaka and Takeuchi (1995); Böhm and Durst (2025). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 26
27. Discussion 5.3 Limits of the Environment-Led Pattern Two limits the artifact alone does not overcome, both of which point back toward the designer-led pattern or a hybrid arrangement. 5.3.1 Real-Time Dialogue and Professional Legitimation 5.3.2 Domain-Specific Reframing and Feedback-Loop Governance Where practitioners met situations the system did not Extending the system to the second domain required address, it offered no mechanism for collaborative reconsidering its scope, which is designerly discovery. problem-solving; Pulling cannot be fully delegated to an Without institutional space for synthesis, the insights artifact. Informant B's capability to lead adaptation was generated through daily application remained constrained by how he was seen: “We're seen as fragmented and unintegrated. engineers” [B-19]. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 27
28. Discussion 5.4 Revisions to the Design Enablement Framework The definition generalizes from designers infusing non-designers, to cultivating designerly ways of knowing within organizations, along two patterns, two capabilities and three functions. HOW LAYER WHY LAYER Designerly application Designer-led Designerly discovery Co-presence activates Pulling, Pushing and Demonstrating through direct dialogue. The feedback loop closes in real time. Designer catalyzes reframing within the ongoing work; the loop closes through sustained interaction with the organization. prior work Prior work · the two capabilities were not distinguished Environment-led The system encodes reasoning and enables evidence-based application at the how layer. The feedback loop fails to reach the why layer; reshaping the context stays unavailable under the conditions observed. this study This study · partial support This study · not supported Figure 3 of the paper, redrawn wide. Rows are the two approach patterns, columns the two capabilities, as in the paper. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 28
29. Conclusion © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 29
30. Conclusion 6.1 Summary The environment-led pattern sustained designerly application but not designerly discovery, and the feedback loop it depends on was impaired unevenly. 01 02 03 04 Designerly application A boundary object The feedback loop Designerly discovery Non-designers began to The design system might It reached the how layer but Reshaping the system itself reason about design serve an increasingly diverse not the why layer, where was not supported under decisions through evidence. engineering team as a scope and principles would the institutional conditions boundary object. need reconsidering. observed. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 30
31. Conclusion 6.2 Theoretical Contributions Four contributions to Design Enablement theory, the first two of which decouple the framework from the assumption that a professional designer is present. 01 02 03 04 Decoupling from copresence Application and discovery Institutional conditions Two emergent themes Documents an environment- Introduces a formal distinction Proposes role authorization for context, and integration with led pattern in which designed between designerly application knowledge activists and AI-driven workflows, warrant artifacts may carry enabling and designerly discovery, two organizational visibility of future investigation. potential, decoupling the modes the pattern supports design practice as candidate framework from designer co- unevenly. necessary conditions. Meta-usability of the enabling presence. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 31
32. Conclusion 6.3 Limitations and 6.4 Future Research The study generates hypotheses rather than establishing them, and the four directions that follow start from what a single, insider, retrospective case left open. 6.3 Limitations — A single case, drawn primarily from one informant — The insider position: designer and researcher in one role — Retrospective accounts, not longitudinal observation — One setting: a Japanese firm serving public-sector clients 6.4 Future Research — Comparative cases that vary the institutional conditions and domain fit — The distinction theorized through design legacies and feedback — Meta-usability, and integration with AI-driven workflows — Non-designers' route from application to discovery Dubberly and Pangaro (2023); Junginger (2015); Böhm and Durst (2025). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 32
33. Thank you Akira Motomura ｜ Spice Factory Inc. ｜ akira_motomura@spice-factory.co.jp
34. Appendix © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 34
35. Appendix Design as a third area of knowing Five works established design as its own way of knowing. Source Established Archer (1979) Positioned design as its own discipline Cross (1982) A third area of human knowing, with reasoning and rigor peculiar to it Buchanan (1992) Design thinking engages indeterminate, wicked problems Manzini (2015) Diffuse design and expert design stand in concert, not in hierarchy Cross (2023) Designerly thinking: co-evolving problem and solution Archer (1979); Cross (1982, 2023); Buchanan (1992); Manzini (2015), who places design exercised by everyone alongside expert design. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 35
36. Appendix Design against the sciences and the humanities Design is its own way of knowing, not a lesser science. Sciences Humanities Design Phenomenon of study The natural world Human experience The artificial world Appropriate methods Controlled experiment, Analogy, metaphor, criticism, Modelling, pattern-formation, classification, analysis evaluation synthesis Objectivity, rationality, Subjectivity, imagination, Practicality, ingenuity, empathy, and neutrality, and a concern for commitment, and a concern for a concern for appropriateness truth justice Values Cross (1982, pp. 221–222). Archer (1979) positioned design as a discipline, and Manzini (2015) places design exercised by everyone alongside design exercised by professionals. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 36
37. Appendix The co-evolution of problem and solution Designerly thinking moves in both directions. Drawing, now cheap at scale The problem The framing The solution Co-evolution The answer drawn Thinking, which stays scarce Cross (2023) describes designerly thinking as proceeding through the co-evolution of problem and solution. Generative AI realized the drawing side at scale, leaving the thinking side scarce. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 37
38. Appendix The shift in what is scarce Generative AI made generation cheap, not reframing. What stays scarce What the experiments show What changed Evaluating, reframing and directing: the thinking side of the cycle Hou et al. (2025) Shin et al. (2025) More ideas, but no gain in the ability to reframe the problem The gap between more and less experienced practitioners may widen Anthropic (2026) Anthropic (2026) Generative AI realizes the drawing side of the cycle at scale Generation is open to anyone who can articulate intent Anthropic (2026); Hou, Wang, Wang, Wang and Yang (2025); Shin, Polyanskaya, Lucero and Oulasvirta (2025). © 2026 Spice Factory Inc. dmi:Academic Design Management Conference 2026 ｜ 38
39. Appendix Language, translation, and AI use The study was carried out in Japanese. Only the quoted excerpts and the construct names became English, and that happened at the writing stage. The Japanese original The interviews, the transcripts and the SCAT The English translation, and the AI use analysis were all prepared in Japanese. The 37 AI tools assisted with transcript preparation and coded segments and 31 thematic constructs were the English translation of quoted excerpts and named in Japanese first. thematic constructs. The same applies to this presentation. All analytical judgments are the author's. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 39
40. Appendix Segment [A-8] traced through the four SCAT steps The other segment Table 2 traces, from raw text to a thematic construct. SCAT step Result for segment [A-8] Step 1: Noteworthy phrases button color; this action; reasoned discussion Step 2: Paraphrase a design decision grounded in the semantic rationale of the UI; a shift from intuitive preference to logical reasoning Step 3: External concepts Internalization; evidence-based design; designerly reasoning Step 4: Thematic construct evidence-based reasoning about design decisions Table 2 of the paper. Segment [A-8] reads: “this button should be this color because this action has this meaning ... it’s become a reasoned discussion”. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 40
41. Appendix Barriers to adoption across roles and generations The gap is structural rather than a matter of viewpoint. Organizational awareness of the design system reached near-universal levels while actual adoption remained at approximately 50%. Both informants reported the same figures without prompting, which is why the gap is read as structural. — Senior non-engineering staff copy and reuse legacy document formats rather than the templates [A-2] — Informant B reads this as unresolved individual judgment: “my design is better than this one” [B-10] — Within engineering, mid-career engineers use the system actively [A-10] — Some junior engineers resist it on aesthetic grounds, so adoption appears mediated by professional identity [A-10] © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 41
42. Appendix Two further properties of the enabling context Both properties surfaced in the second interview. Meta-usability Integration with AI-driven workflows If the artifacts that constitute the enabling context cannot be Enabling contexts built on the assumption of human-mediated located and entered by their users, the context cannot function design work can be structurally bypassed by workflows that regardless of its internal content. delegate generation to a machine. — Templates sit on a file server in locations users find hard to locate [B-12] — AI coding tools generate prototypes directly from naturallanguage prompts [B-22] — Time spent searching turns the system into friction rather than acceleration [B-12] — “It becomes friction in the very motion we’re trying to accelerate” [B-22] — Access logs show members, including some at section-head level, had not opened the shared workspace for 6 months or — Delegating combination to the machine can erode users' own capacity to integrate knowledge (Böhm & Durst, 2025) more © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 42
43. Appendix The alternative reading: structural bias toward reuse Goree et al. (2021) read the asymmetry differently. Investigating two decades of web design, Goree et al. (2021) found that reliance on shared libraries, frameworks, and templates has measurably homogenized design, at the cost of making unusual designs harder to create. On that reading, application without discovery is structural rather than a failure of support. — The case fits part of this: engineers did lean on existing outputs [A-15] — But the reasoning worked from the meaning behind design choices, not only their form [A-8] — Informant B came to think about why the system was built, not only how to apply it [B-20] — We read the asymmetry as shaped by both: the system leans toward application, and institutional support shapes whether discovery becomes possible Goree, Doosti, Crandall and Su (2021). © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 43
44. Appendix References (1 of 3) Every work this talk names, in full. The four marked with an asterisk are cited in the research note rather than in the paper, so they are not in the proceedings. — Alexander, C., Ishikawa, S., & Silverstein, M. (1977). A pattern — Anthropic. (2026). Introducing Claude Design by Anthropic Labs. language. * — Archer, B. (1979). Design as a discipline. Design Studies, 1(1). — Böhm, K., & Durst, S. (2025). Knowledge management in the age of generative AI. VINE, 56(1). — Buchanan, R. (1992). Wicked problems in design thinking. Design — Cross, N. (1982). Designerly ways of knowing. Design Studies, 3(4). Issues, 8(2). — Cross, N. (2023). Designerly ways of knowing. Birkhäuser. — Dubberly, H., & Pangaro, P. (2023). How might we help designers understand systems? She Ji, 9(2). — Goree, S., Doosti, B., Crandall, D., & Su, N. M. (2021). Investigating the homogenization of web design. CHI '21. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 44
45. Appendix References (2 of 3) Every work this talk names, in full. The four marked with an asterisk are cited in the research note rather than in the paper, so they are not in the proceedings. — Gustafsson, E. W. (2021). Defining and evaluating design system usability. KTH. — Handal, S., Koo, C., & Adelakun, O. (2022). Design systems implementation for digital transformation. ICITS 2022. — Hou, J., Wang, L., Wang, G., Wang, H. J., & Yang, S. (2025). The — IBM. (n.d.). Principles. In IBM design language. * double-edged roles of generative AI in the creative process. ISR. — Junginger, S. (2015). Organizational design legacies and service — Kholmatova, A. (2017). Design systems. Smashing Magazine. * design. The Design Journal, 18(2). — Lamine, Y., & Cheng, J. (2022). Understanding and supporting the — Mall, D. (2023). Design that scales. Rosenfeld Media. * design systems practice. arXiv. — Manzini, E. (2015). Design, when everybody designs. MIT Press. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 45
46. Appendix References (3 of 3) Every work this talk names, in full. The four marked with an asterisk are cited in the research note rather than in the paper, so they are not in the proceedings. — Motomura, A., & Takeda, G. (2025). Design enablement through design systems. Journal of Design Science, 5(1). — Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating company. Oxford University Press. — Shin, J., Polyanskaya, A., Lucero, A., & Oulasvirta, A. (2025). No evidence for LLMs being useful in problem reframing. CHI '25. — Takeda, G., & Motomura, A. (2025). A case study on the construction of a design system. Journal of Design Science, 5(1). — Motomura, A., & Yaegashi, K. (2024). Enabling non-designers to design. DMI:ADMC 24. — Otani, T. (2019). Qualitative research: from methodology to analysis using SCAT. Nagoya University Press. — Star, S. L., & Griesemer, J. R. (1989). Institutional ecology, translations and boundary objects. SSS, 19(3). — von Krogh, G., Ichijo, K., & Nonaka, I. (2000). Enabling knowledge creation. Oxford University Press. — Werle, D. R. (2021). Design systems: definitions and main elements. Tallinn University. © 2026 Spice Factory Inc. ｜ dmi:Academic Design Management Conference 2026 46