Power and Epistemology - Who's Knowledge Counts? Farhana Sultana of Syracuse University

Summary

Farhana Sultana argues that knowledge used in science and development policy is never wholly neutral. What is accepted as “objective” evidence is shaped by power: historical and institutional hierarchies influence which questions are asked, whose data are valued, and whose knowledge is treated as authoritative.

Main Ideas

• Science is not an absolute, neutral truth. Scientific knowledge changes, and there are multiple ways of knowing and understanding the world. Treating science or evidence as universally objective can conceal the political choices and power relations behind it.
• Knowledge is situated. Sultana critiques the imagined “view from nowhere”—the idea that researchers can produce knowledge entirely detached from their own position. Drawing on feminist critiques, she favors approaches such as standpoint theory and situated knowledge, which account for both the people producing knowledge and those being studied.
• Scientific authority is unequal. Gendered, racialized, and other structural barriers affect who conducts research, who can participate, which forms of evidence are accepted, and which disciplines or methods are privileged. Quantitative and physical-science approaches may receive more weight than qualitative research, social sciences, or other knowledge systems.
• Evidence and measurement involve choices. Deciding what counts as evidence—and what is worth measuring—is not a neutral technical exercise. Overreliance on metrics, reproducibility, and observable results can reduce complex problems to narrow measures and constrain creative responses.
• Development carries colonial legacies. Sultana links the rise of development to colonial history and argues that Western and Eurocentric assumptions can persist in contemporary planning, financing, and expertise. She warns that older racialized ideas, including forms of environmental and biological determinism, can continue to shape development and environmental policy.
• Plans can fail when they disregard local realities. A disconnect between Western scientific frameworks and local cosmologies, lived experience, and needs can lead to projects that miss their intended goals. Better data collection alone will not fix this if the underlying assumptions about what matters and what counts as knowledge remain mismatched.
• Representation alone is not enough. Token inclusion can leave institutional hierarchies and dominant ways of thinking intact. Consultation can also reproduce exclusions—for example, when a community is represented mainly by men, producing a skewed account of its perspectives and needs.
• Policy work requires accountability and reflection. Sultana urges people involved in development to examine their own positions, assumptions, and institutional power, and to take seriously the risk of reproducing a “white savior” role while intending to help.

Approaches and Lessons Proposed

• Ask critical questions about evidence:
  • What is being treated as evidence, and why?
  • Who produces, holds, and validates it?
  • Whose knowledge is excluded or given less weight?
  • What political interests or power relations does its use serve?
• Use transdisciplinary collaboration. Work across academic disciplines and with citizens, stakeholders, and the wider public, rather than relying only on expert or institutional perspectives.
• Co-create knowledge. Include different forms of testimony and local knowledge, and examine how authority and credibility are distributed among participants.
• Apply intersectional analysis. Consider how gender, class, race, and other relations shape whose experiences are heard; do not assume that a small set of community representatives speaks for everyone.
• Address structural inequalities directly. Examine ideological assumptions, systemic sexism, institutional racism, and historical scientific racism—not only the technical quality of data.
• Build reflexivity, accountability, and inclusion into policy processes. Consider who devises policies, for whom they are designed, and how monitoring, evaluation, and learning are carried out.
• Move beyond tokenism. Assess whether participation changes whose perspectives and knowledge shape decisions, rather than simply adding representation.

Speakers and Sources Mentioned

• Farhana Sultana — the speaker presenting the talk.
• Other panelists — Sultana refers to remarks from an earlier part of the panel, including a person she calls “Olivia” and other unnamed speakers; their full identities cannot be established from the subtitles.