What Is Happening
In Japan's information environment, voices that fail to reach decision-makers are produced structurally. In administrative advisory councils, evidence from affected parties is dismissed as "insufficient sample size." On social media, minority viewpoints are buried under harassment. Regional issues fall outside the agenda of national newspapers. Ignorance does not arise naturally. It is manufactured institutionally.
Conventional fact-checking and awareness campaigns have been forced into asymmetric warfare against this structure. The attritional labor of correcting individual pieces of misinformation, one by one. The individualist ethos that reduces literacy to personal capacity. Neither approach has the power to move the underlying structure.
At the same time, practices that embed epistemic justice into institutions are emerging in Japan. Attempts to modify the internal design of platforms. Attempts by independent media to accumulate resources for monitoring power. Attempts by citizens to intervene in administrative data. This essay frames these as "Counter-Design."
Counter-Design is a practice aimed at redesigning the information environment itself, not at reforming individual consciousness. Examining the framework of "design justice" proposed by Sasha Costanza-Chock (2020) in Design Justice within the Japanese context brings emergent cases into view.
Background and Context
What Counter-Design Means
Counter-Design assumes that dominant design reproduces structural inequality. Costanza-Chock (2020) analyzed how standard UX design excludes anyone outside the "assumed default user." Connecting this critique to the tradition of agnotology yields the following working hypotheses.
- The standard design of the information environment carries a distributive function that makes certain voices easier to hear and others harder to hear
- This distribution is not neutral; it reflects existing power relations
- Counter-Design is an intentional intervention to reorganize this distribution
The theoretical companion, "Epistemic Injustice and Information Access Gaps in NPOs," draws on Fricker (2007) on epistemic injustice. The present essay is its implementation-focused counterpart.
Selection Criteria
Candidate cases of counter-design in Japan were assessed against five criteria.
- Implementation of epistemic justice — Is the design intent to reorganize the distribution of voice clearly articulated?
- Collective intervention — Does it operate as a collective mechanism, not merely individual effort?
- Analyzability of structure — Is there sufficient disclosure to decompose and describe the mechanism?
- Continuous operation — Is it organized as an ongoing institution rather than a one-off project?
- Transferability — Does it contain principles other domains or organizations can adopt?
Five cases satisfy these criteria. Yahoo! News Comments (as internal platform design). Tansa (as investigative journalism). Choose Life Project (as deliberative media). Code for Japan (as civic tech). SYNODOS (as an expert platform). Educational design is set aside for a separate essay.
Reading the Structure
Case 1: The Internal Design of Yahoo! News Comments
This is a rare case in which a large platform intervened in its own architecture.
The comment section of Yahoo! News had a structure in which the quantitative dominance of anonymous posts made harassment easy to surface. The operator introduced multiple AI models in stages. It published a Comment Policy and introduced a Comment Revision Model that suggests rewording before submission. Combined with 24-hour human monitoring, a "Constructive Comment Ranking Model" changed the weighting of display order.
The targeted structure of ignorance production is a simple logic of numbers. When harassment is posted collectively, it fills the screen as "majority opinion," silencing a minority of constructive comments. The counter-design intervention operates at a single point: reorganizing the weighting of display order from "vote count" to "constructive score."
Reduction of violating comments through pre-submission intervention has been reported as an outcome. The limits are that the definition of "constructiveness" is left to the platform operator, and the transparency of the scoring model is not open to external audit.
Case 2: Tansa (formerly Waseda Chronicle)
Founded in 2017 as a project of the Waseda University Journalism Institute, it became independent from the university in 2018 and was renamed Tokyo Investigative Newsroom Tansa in 2021. It is a nonprofit investigative journalism organization.
The targeted structure of ignorance production is the attenuation of criticism of power under advertising-dependent models. Existing media face structural constraints in investigative reporting on advertisers, government officials, and large corporations. Tansa adopts a donation-supported model that keeps it independent from advertisers, sustaining long-term investigations on tightly scoped subjects.
The counter-design intervention lies in the funding model. A structure in which donors fund "investigative capacity itself" rather than "topics" enables a temporal horizon incompatible with the short-term expectations of advertising revenue.
Achievements include multi-year investigations and receipt of the Journalism X Award from the Citizen Journalism Support Fund. Limits include the constraint that donation scale caps the number of investigations, and the scale problem in a Japanese society where donation culture is not fully mature.
Case 3: Choose Life Project
Founded in 2016 by former TV news producers as a video-media project, incorporated in 2020. It distributes deliberative programs on political and social issues via the Web.
The targeted structure of ignorance production is the programming logic of terrestrial broadcasting. Under time-slot constraints and audience-rating constraints, complex political deliberation is simplified in the name of "accessibility." CLP removes the time-slot constraint, sustaining long-form deliberation through Web distribution.
The counter-design intervention lies in carving out a deliberative space outside broadcast regulation and advertising revenue. Pre-election deliberation, Diet-session explanation, and debates that seat multiple parties together — items difficult to program on terrestrial television — become structurally possible.
Achievements include continuous distribution of deliberative content during electoral periods and expansion of the viewer base. A limit is the fragility of the operating base, which surfaced in 2022 as "criticism concerning transparency"; the organizational governance design of collective intervention remains an open question.
Case 4: Code for Japan
Founded in 2013. A general incorporated association promoting collaboration between government and citizens in the civic-tech domain. Its Slack community exceeds 8,000 members.
The targeted structure of ignorance production is the closed design of administrative data. Administrative data are "publicly released" in name, but are not actually provided in forms citizens can use. This structurally narrows the paths by which citizens can intervene in administrative decisions.
The counter-design intervention builds citizen-side infrastructure for acquiring, processing, and presenting data. Under the slogan "Think together, build together," Code for Japan sustains a fellowship program that embeds personnel into government, a nationwide expansion of regional brigades, and collaborative hackathons.
Achievements include cooperation with Taiwan's g0v during the COVID-19 pandemic, influence on the design of government digital services, and the fellowship track record. Limits include the tendency for civic-tech participants to be a limited demographic (a stratification of technical skill), and the uneven receptivity of government partners.
Case 5: SYNODOS
Founded in 2007. A portal of expert commentary and liberal education. It gathers commentary from researchers, professionals, NGO/NPO staff, journalists, and affected parties. Serizawa Kazuya serves as representative and editor-in-chief.
The targeted structure of ignorance production is the disconnect between expert knowledge and popular discourse. Academic papers do not reach general readers because of technical vocabulary and peer-review constraints. Social-media discourse spreads without expert vetting. The middle-tier media that sustained "careful expert commentary" in the gap between these two were thin in Japan.
The counter-design intervention lies in an editorial design that selects appropriate experts by theme and publishes at a speed faster than peer review but slower than social media. It operates through a combination of paid membership and advertising.
Achievements include nearly two decades of continuous operation and an archive of cross-disciplinary commentary that has grown over time. Limits include a level of author compensation dependent on the goodwill of experts, and the access stratification of the expert-platform format itself.
Common Patterns and Differences
Contrasting the five cases brings out a typology of counter-design.
Collective intervention divides into three types. Internal platform design (Yahoo), independent media (Tansa, CLP, SYNODOS), and civic participation infrastructure (Code for Japan). The first is "local reorganization within the dominant architecture." The second is "exit from the dominant architecture and construction of alternatives." The third is "external intervention into the dominant architecture."
Funding models also diverge. Advertising revenue (Yahoo), donation (Tansa, CLP), business plus subsidy (Code for Japan), paid membership plus advertising (SYNODOS). The rigor of epistemic-justice implementation correlates loosely with the independence of the funding source. The higher the advertising dependence, the narrower the scope of critical intervention tends to become.
Measures of epistemic justice differ across cases. Yahoo has the quantitative indicator of reduction in violating comments. Tansa signals it through the number of investigative articles and awards. CLP signals it through viewer numbers and viewing time. Code for Japan signals it through community scale and the number of government partnerships. SYNODOS signals it through the article archive and membership. None of these indicators directly measures "how much the distribution of voice has been reorganized"; they remain indirect. This is a shared limit.
Implications for Practitioners
Which type is applicable to one's own organization can be judged along three axes.
First, the choice between reorganizing within the dominant architecture, intervening from outside, or exiting to construct an alternative. The Yahoo type (internal reorganization) is available only to platform operators. The Code for Japan type (external intervention) presupposes a degree of government information disclosure. The Tansa/CLP/SYNODOS type (alternative construction) requires a sustained period of investment to secure the independence of the funding model.
Second, the institutional barriers of the funding model. Japan's tax treatment of donations is less favorable than that of the United States, constraining the scale of donation-based models. This is a structural barrier that cannot be overcome by individual organizations, and requires intervention at a separate layer — tax reform.
Third, the design of outcome indicators. Because no indicator directly measures "reorganization of the distribution of voice," one must rely on indirect measures. This indicator design itself remains an open question for counter-design research.
The position of ISVD (Institute for Social Vision Design) differs from all five cases. It is neither Tansa-type (investigative journalism) nor Code-for-Japan-type (civic tech), but a hybrid of research and practice that connects the theory of epistemic justice to its implementation. There is room to approach the shared question of the five cases — outcome-indicator design — from the framework of agnotology.
Questions for This Laboratory
- Can we design an indicator that directly measures the degree of implementation of epistemic justice?
- How do the five counter-design types apply in other domains such as education, judiciary, and medicine?
- Between the internal-reorganization type (Yahoo) and the external-exit type (Tansa), which possesses greater structural resilience?
- How can the relationship between Japan's donation tax regime and the scalability of counter-design be quantified?
These questions connect directly to the theory of "Epistemic Injustice and Information Access Gaps in NPOs." They also engage, at the level of concrete cases, with the institutional-design questions raised in "The Obstructive Effect of Strategic Ignorance in EBPM."
References
Epistemic Injustice: Power and the Ethics of Knowing — Fricker, M.. Oxford University Press
Agnotology: The Making and Unmaking of Ignorance — Proctor, R. N. & Schiebinger, L.. Stanford University Press
Yahoo! News Launches Introduction of 'Comment Revision Model' Where AI Proposes Rewording of Expressions — LY Corporation. LY Corporation Press Release
About Tansa — Tokyo Investigative Newsroom Tansa. Tansa
About Choose Life Project — CLP Inc.. Choose Life Project
Code for Japan — Code for Japan. Code for Japan
Company Overview of SYNODOS Inc. — SYNODOS. SYNODOS
