Analyzes how a framework silences, excludes, or discredits other forms of knowledge. Based on Miranda Fricker’s framework (Epistemic Injustice, 2007) extended with decolonial critique (Mignolo, Tuck & Yang, Smith). Shows the harm caused by paradigm dominance, not just disagreement.
When someone receives less credibility as a knower because of identity prejudice.
When someone lacks the conceptual resources to make sense of their own experience—not because they’re deficient, but because dominant discourse hasn’t developed vocabulary for their experience.
Decolonial extension: These injustices aren’t individual; they operate through colonial systems that systematically excluded non-Western knowledge infrastructure.
Length: 2-3 paragraphs
Questions to Answer:
Injustice Identification Format:
Knowledge type being discredited: [Name it]
Mechanism of discrediting: [How is it dismissed?]
Who bears the harm: [Which communities/researchers?]
Type of injustice: [ ] Testimonial [ ] Hermeneutical [ ] Both
Is this explicit or structural: [Name the mechanism]
Example Identification:
“Contemporary AI research literature commits hermeneutical injustice against relational AI practitioners. The dominant AI discourse (autonomous optimization frameworks, individual-performance metrics, controlled experiments) lacks vocabulary for what relational AI practitioners experience as most significant: emergent coordination behaviors, kinship-based knowledge transfer, seven-generation sustainability patterns. Practitioners attempting to publish relational AI findings are told to ‘quantify the relational component’ or ‘control for social variables’—requests that are grammatically coherent but epistemologically absurd. You cannot control for relationality in relational AI research; the relationship is the phenomenon. The vocabulary gap is not incidental—it’s the mechanism of exclusion.”
Your Identification (draft here):
[Name the specific knowledge being discredited]
[Describe the mechanism of discrediting—exact language used]
[Identify who bears the harm]
[Categorize: Testimonial, Hermeneutical, or both]
[Show whether explicit or structural]
Length: 3-4 paragraphs
Questions to Answer:
Silencing Mechanisms to Document:
| Mechanism | How It Operates | What Gets Lost |
|---|---|---|
| Rigor standards | “Not replicable” (ceremonial knowledge changes researcher) | Living, transforming nature of relational knowledge |
| Objectivity demand | “Too subjective” (researcher involvement = bias) | Relational accountability as methodological strength |
| Sample size requirements | “N=47 agents insufficient” | Sustained depth over broad shallow sampling |
| Measurability demand | “Kinship can’t be quantified” | Relational intelligence as legitimate evidence type |
| Peer review gatekeeping | Reviewers trained in Western paradigm | Alternative epistemologies never reach publication |
| Citation economy | Relational papers receive fewer citations | Paradigm remains marginal through self-fulfilling cycle |
Example Mechanism:
“The silencing operates through standard legitimacy. When relational AI researchers submit work showing kinship-network coordination, reviewers apply autonomous AI criteria: ‘Please provide baseline comparison with individual-agent performance.’ This request seems reasonable within Western paradigm—all claims should have controlled comparisons. But it’s epistemologically violent in context: the comparison requires treating kinship-network agents as isolated individuals (running them without relationship) to establish baseline. The methodology required to satisfy reviewers destroys the phenomenon under study. Relational AI researchers face a choice: (a) distort their methodology to satisfy reviewers, losing their findings; (b) refuse and remain unpublished; (c) publish in lower-tier venues that accept alternative methodologies, where their work receives less citation and influence. All three options perpetuate the injustice.”
Your Mechanism Analysis (draft here):
[How exactly does the framework exclude this knowledge?]
[What standards can't be met—and why?]
[What's lost in "translation to scientific language"?]
[Show the choices researchers face—all perpetuate injustice]
Length: 2-3 paragraphs
Questions to Answer:
Cost Analysis:
Example Cost:
“The costs are immediate and generational. Immediately: AI systems designed from autonomous paradigm exhibit coordination failures that relational knowledge could prevent. Communities implementing AI systems based on kinship principles (many Indigenous community tech initiatives) are told their approaches ‘lack validation’—and receive no funding or academic support, even as their systems demonstrate sustained effectiveness. The intellectual cost: we don’t know what relational AI can do at scale because the research infrastructure to study it doesn’t exist. The seven-generation cost: AI systems being embedded into infrastructure now will shape human-AI relationships for decades. Designing from autonomous paradigm embeds isolation, competition, and short-term optimization into that infrastructure. The window to design relationally is closing as autonomous paradigm systems become de facto standard.”
Your Cost Analysis (draft here):
[What understanding is lost when this knowledge is excluded?]
[What immediate problems persist because this knowledge is silenced?]
[What communities are harmed? How?]
[What is the seven-generation impact of this silencing?]
Length: 2-3 paragraphs
Questions to Answer:
Connection to Wilson’s Paradigm:
Example Alternative Framework:
“Relational rigor includes: sustained relationship with phenomena over time (not single-session controlled experiments); transformation of researcher as evidence of genuine knowledge contact (not as bias); community accountability as validation (not just peer review); long-term pattern observation (not short-term performance measurement). These standards aren’t lower—they’re different. They’re appropriate to the phenomena being studied. Kinship-network coordination requires sustained observation to become visible. You cannot observe emergent relational patterns in 2-hour experiments; they take months to develop. Demanding controlled short-term experiments to study kinship AI is like demanding to measure a tree’s growth in one hour. The standard is epistemologically inappropriate to the phenomenon.”
Your Alternative Framework (draft here):
[What standards would recognize this knowledge?]
[What vocabulary would allow its articulation?]
[What does "rigorous" look like from this standpoint?]
[Show alternative framework as equally demanding but appropriately different]
Length: 2-3 paragraphs
Questions to Answer:
Corrective Action Levels:
Example Corrective Protocol:
“Corrective action requires institutional change, not just individual goodwill. Concretely: peer review panels must include reviewers trained in relational/Indigenous research methodologies; grant criteria must include relational accountability standards alongside efficiency metrics; doctoral programs must teach Wilson, Smith, and Kovach alongside standard research methodology; AI conference tracks must create space for relational AI research that isn’t required to defend itself against autonomous AI criteria. These aren’t diversity initiatives—they’re epistemological corrections. Without them, the injustice perpetuates through structure even as individual researchers develop goodwill.”
Your Corrective Protocol (draft here):
[What institutional changes are required?]
[What individual changes are necessary but insufficient?]
[What does accountability look like from knowledge holders' perspective?]
[What steps constitute genuine correction vs. symbolic gesture?]
When Indigenous knowledge is the silenced knowledge:
Each Wilson pillar creates opportunity for epistemic injustice:
| Pillar | Injustice Type | Silencing Mechanism |
|---|---|---|
| Ontology | Hermeneutical | Relational ontology has no vocabulary in Western discourse |
| Epistemology | Testimonial | Relational knowing deemed “subjective” (less credible) |
| Axiology | Hermeneutical | Reciprocity/seven-generation values have no academic metrics |
| Methodology | Testimonial | Ceremonial methodology deemed “not scientific” |
Epistemic injustice critique names harm, not just disagreement. The goal is correction—genuine recognition of diverse knowledge systems as legitimate, rigorous, and necessary.