Sources and References

Primary texts

• Longino, Helen. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry. Princeton University Press, 1990.
• Longino, Helen. “Values and Objectivity.” Science, Technology, and Human Values 19.4 (1994): 382–400.
• Longino, Helen. The Fate of Knowledge. Princeton University Press, 2002.
• Harding, Sandra. The Science Question in Feminism. Cornell University Press, 1986.
• Harding, Sandra. Whose Science? Whose Knowledge? Thinking from Women’s Lives. Cornell University Press, 1991.
• Keller, Evelyn Fox. Reflections on Gender and Science. Yale University Press, 1985.
• Nelson, Lynn Hankinson. Who Knows: From Quine to a Feminist Empiricism. Temple University Press, 1990.
• Wylie, Alison. Thinking from Things: Essays in the Philosophy of Archaeology. University of California Press, 2002.
• Anderson, Elizabeth. “Feminist Epistemology and Philosophy of Science.” Stanford Encyclopedia of Philosophy, 2020.
• Kitcher, Philip. Science in a Free Society. Princeton University Press, 2001.
• Haraway, Donna. “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective.” Feminist Studies 14.3 (1988): 575–599.
• Collins, Patricia Hill. Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment. Routledge, 1990.
• Fehr, Carla. “Feminist Engagement with Evolutionary Psychology.” Hypatia 27.1 (2012): 50–72.

Online resources

• Anderson, Elizabeth. “Feminist Epistemology and Philosophy of Science.” Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/feminism-epistemology/
• Intemann, Kristen. “Feminist Philosophy of Science.” Internet Encyclopedia of Philosophy. https://iep.utm.edu/fem-sci/
• Solomon, Miriam. “Social Empiricism.” Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/scientific-knowledge-social/

Chapter 1: What Is Feminist Philosophy of Science?

1.1 The Problem Space

Feminist philosophy of science is a branch of epistemology and philosophy of science that investigates how gender — as a social category, a cultural system of meaning, and a dimension of power — shapes the production, validation, and application of scientific knowledge. It does not merely ask whether women are represented in science; it poses the deeper question of whether the very norms, methods, and self-understanding of scientific practice are gendered in ways that distort inquiry and marginalize certain knowers.

Feminist philosophy of science (女性主义科学哲学) thus operates at the intersection of two projects: it inherits from mainstream philosophy of science the questions of confirmation, explanation, realism, and objectivity, while simultaneously bringing feminist theory’s attention to the politics of knowledge, the situatedness of inquiry, and the epistemic consequences of social structure.

The field emerged in the late 1970s and early 1980s through the convergence of the women’s liberation movement, the growth of women’s studies in academia, and critiques internal to the philosophy of science that had already destabilized the received view of science as a pure, value-free activity. Figures such as Evelyn Fox Keller, Sandra Harding, Helen Longino, and Lynn Hankinson Nelson each brought distinct approaches — but all agreed that the question “what counts as science done well?” could not be answered without attending to who is doing it and under what social conditions.

1.2 The Three Waves of Feminist Philosophy of Science

The development of feminist philosophy of science can be usefully organized into three broad phases, each building on and critiquing the previous one.

The first wave: women in science (1970s–mid-1980s). The earliest feminist interventions in science studies were primarily historical and sociological. Scholars documented the systematic exclusion of women from scientific institutions, the erasure of women’s scientific contributions, and the ways in which formal barriers — restrictions on university attendance, denial of laboratory positions, exclusion from professional societies — had prevented women from participating in the production of scientific knowledge. Key figures here include Margaret Rossiter, whose historical work Women Scientists in America (1982) documented the mechanisms by which women’s scientific contributions were systematically attributed to male colleagues or simply ignored (a pattern she called the “Matilda effect,” analogous to the already-named “Matthew effect” for men). The first wave established that the sociology of science was deeply gendered, but it largely left untouched the epistemological question of whether the content of science was affected.

The second wave: bias in the content of science (mid-1980s–mid-1990s). The second wave moved from asking “where are the women?” to asking “how has the absence of women shaped what science studies, how it studies it, and what it concludes?” The key move was to argue that the exclusion of women from science had not merely been unjust to individual women; it had produced distorted knowledge. Evelyn Fox Keller’s analysis of gender metaphors in the language of science, Sandra Harding’s critique of androcentrism across the natural and social sciences, and Helen Longino’s work on values in scientific reasoning all belong to this phase. The second wave was also the period when feminist science studies became philosophically sophisticated — engaging directly with the mainstream philosophy of science literature on underdetermination, confirmation, and the demarcation problem.

The third wave: social epistemology and structural reform (mid-1990s–present). The third wave is characterized by a turn from critique to construction — from identifying the ways science fails to epistemically inadequate forms of diversity and objectivity, to developing positive accounts of what epistemically adequate, democratically accountable science would look like. Longino’s later work, Kitcher’s social epistemology, Wylie’s methodological innovations in archaeology, and Fehr’s empirical work on demographic diversity in science all belong here. The third wave also engages more directly with questions of global justice — asking whose science counts, whose knowledge is recognized, and how the global distribution of scientific authority reflects historical patterns of colonialism and extraction.

1.3 Historical Context: Women in Science

The historical exclusion of women from scientific institutions is well documented and philosophically significant. It is not merely background context; it is directly relevant to the epistemological arguments of the field, because it shows that the composition of scientific communities is not epistemically neutral.

Throughout most of the history of modern science — roughly from the Scientific Revolution to the late twentieth century — women were systematically excluded from the formal institutions of scientific inquiry: universities, learned societies, and professional laboratories. When women did science, it was often in informal, domestic, or auxiliary roles. Caroline Herschel assisted her brother William and made independent discoveries of comets that were often attributed to him. Rosalind Franklin’s X-ray crystallography data — produced with technical virtuosity and interpretive skill — was used without her knowledge or credit by Watson and Crick in their structural analysis of DNA. Cecilia Payne-Gaposchkin discovered that stars are composed primarily of hydrogen and helium — a fundamental finding in astrophysics — but was persuaded by her supervisor Henry Norris Russell to suppress this conclusion in her dissertation; Russell later published the same finding himself and received the credit for it.

These cases are not merely anecdotes. They illustrate a systematic pattern — what Rossiter calls the Matilda effect — in which women’s scientific contributions are attributed to men, minimized, or ignored, regardless of their quality. The pattern is not explained by women’s lack of ability or interest but by institutional structures, social norms about gender and intellectual authority, and the intertwined dynamics of prestige and power in science.

1.4 Why Feminist? Why Science?

One might ask whether a political commitment — feminism — has any legitimate role in epistemology, the discipline traditionally concerned with the nature and grounds of knowledge. The worry is that political commitments introduce bias and compromise rational inquiry.

Feminist philosophers of science turn this worry on its head. Their central contention is that the relevant question is not whether values enter science, but which values, whose values, and to what effect. The pretense of value-freedom, they argue, has historically served to insulate sexist and androcentric assumptions from critical scrutiny rather than to eliminate them.

1.5 Three Entry Points in the Literature

Scholars in this field typically enter through one of three problems, each of which will recur throughout this course:

1. The bias problem: Scientific research on sex, gender, race, and sexuality has historically encoded cultural stereotypes as empirical findings. How do such biases enter, and how can they be detected and corrected?

2. The objectivity problem: If values shape inquiry, what becomes of the ideal of objectivity? Is objectivity possible? Desirable? How should it be reconceptualized?

3. The standpoint problem: Do socially marginalized groups have distinctive epistemic resources — a different vantage point that reveals truths invisible from dominant positions?

These problems are related but analytically distinct. The course traces them through the central contributions of Helen Longino, Sandra Harding, Evelyn Fox Keller, Donna Haraway, and others, before turning to case studies in which these abstractions become concrete and politically urgent.

Chapter 2: Androcentrism and Bias in Scientific Research

2.1 Defining Androcentrism

Androcentrism (男性中心主义) is the practice, often unconscious, of treating the male perspective, male experience, or male-associated traits as the default, the norm, or the ideal. In science, androcentrism manifests when:

• research questions are framed around phenomena salient to men’s lives;
• male subjects are used as the standard for medical or psychological research;
• explanations appeal to distinctively masculine values such as competition, aggression, or dominance while ignoring cooperation, care, or nurturance;
• the labor and knowledge of women (including indigenous women and women of color) is appropriated without credit.

2.2 A Taxonomy of Bias Types

Feminist philosophers of science have developed a rich taxonomy of the mechanisms through which gender bias enters scientific inquiry. It is important to distinguish these mechanisms precisely, because different mechanisms require different corrective strategies.

Feminist philosophers argue that recognizing these distinct mechanisms is important because simply adding women to a study does not correct conceptual or theoretical-framework bias; it requires reconceptualizing the research questions themselves.

2.3 Historical Examples of Bias

2.3.1 Primatology and Female Agency

One of the most discussed cases in feminist philosophy of science concerns primatology. Donna Haraway’s analysis in Primate Visions (1989) and related work by Sarah Blaffer Hrdy showed that mid-twentieth-century primate research systematically underestimated the agency, social complexity, and strategic behavior of female primates. Dominant males were treated as the organizing social units; female behavior was described in terms of passivity and receptivity.

When women began entering primatology in larger numbers in the 1970s and 1980s, different questions were asked, different behavioral patterns were observed and recorded, and the dominant narrative of male-driven primate social organization was substantially revised. Female primates turned out to be strategically active, to form coalitions, to play central roles in transmitting cultural practices across generations, and to exercise mate choice in ways that shape the social structure of primate groups.

Sarah Blaffer Hrdy’s work on infanticide among langur monkeys is paradigmatic. The prevailing account held that female langurs were passive victims of male infanticide; Hrdy showed through careful field observation and evolutionary analysis that females actively and strategically resisted infanticide through coalition formation, sexual solicitation of potentially infanticidal males, and other behaviors. Her account reconceptualized female langurs as strategic agents whose behavior shapes their social environment — a finding invisible within the androcentric conceptual framework.

2.3.2 Cardiovascular Medicine and the “Default Male”

Until the late twentieth century, clinical trials for cardiovascular drugs routinely excluded women as subjects on the grounds that hormonal variation would complicate the data. The National Institutes of Health’s 1986 policy statement on women as subjects in clinical research documented the systematic exclusion of women and the unexamined assumption that findings from male subjects would generalize to women without modification.

The result was a set of treatment protocols calibrated to male physiology and applied to women without adequate empirical basis. Heart attack symptoms, for example, present differently in women than in men: while men typically present with acute chest pain, women more commonly present with atypical symptoms — nausea, jaw pain, fatigue, shortness of breath — that are less likely to be recognized as cardiac in origin. Because the male presentation was the one studied, taught, and codified in diagnostic criteria, women’s symptoms were regularly misidentified, attributed to anxiety or gastrointestinal problems, and treated less aggressively. This contributed to demonstrably higher in-hospital mortality for women following heart attacks in the period when the male-default assumption dominated clinical practice.

This case shows that androcentrism has direct consequences for human welfare — it is not merely an academic problem about how we describe the world, but a practical problem about how interventions are designed and applied. It also shows that the bias was invisible to the researchers who perpetuated it: they were not consciously discriminating against women; they were following the norms of their field, norms that incorporated androcentric assumptions so deeply that they did not register as assumptions at all.

2.3.3 Simon Baron-Cohen and Systemizing-Empathizing Theory

Simon Baron-Cohen’s theory of the “extreme male brain” and the systemizing-empathizing (S-E) distinction provides a contemporary example of bias in the study of cognitive sex differences. Baron-Cohen argues that, on average, males tend toward systemizing (analyzing and building rule-based systems) and females toward empathizing (identifying and responding to the mental states of others), and that autism represents an extreme version of the male cognitive profile.

Feminist critics have raised a number of methodological objections. First, the S-E framework operationalizes the concepts via self-report questionnaires that are themselves susceptible to stereotype threat and social desirability effects: people who know that they are being assessed on “male” or “female” traits report responses consistent with gender stereotypes at higher rates. Second, the theory builds cognitive sex differences into its foundational categories — defining systemizing and empathizing in ways that make them by construction correlated with gender stereotypes, making the finding that males score higher on systemizing in part an artifact of the measure. Third, the inference from correlational data to evolutionary and neurological explanation is underdetermined: the observed differences in self-reported cognitive style are consistent with socialization explanations without requiring any hypothesis about evolved sex differences in neural architecture.

Chapter 3: Contextual Empiricism — Values and Objectivity (Longino)

3.1 Longino’s Central Project

Helen Longino’s Science as Social Knowledge (1990) is the foundational text of this course. Its central aim is to develop a philosophy of science that can acknowledge the role of values in scientific inquiry without abandoning the ideal of objectivity. Longino calls her position contextual empiricism (语境经验主义).

The name signals two commitments. “Empiricism” signals that observation and evidence remain central — science is not reducible to ideology or power. “Contextual” signals that evidence never speaks for itself; it is always interpreted within a background of contextual assumptions that include, inevitably, social values.

3.2 The Underdetermination Argument

Longino builds on the thesis, familiar from Duhem and Quine, that any given body of evidence is compatible with multiple theories. The underdetermination of theory by evidence (理论被证据的不充分决定) is a deep epistemological fact with direct consequences for the relationship between science and values.

The classical Duhem-Quine thesis holds that individual scientific hypotheses cannot be confirmed or disconfirmed in isolation: any test of a hypothesis H requires auxiliary assumptions A (about instruments, background conditions, measurement procedures), and what experience directly tests is the conjunction H & A. If the conjunction fails — if the experimental result is not what H & A predicted — we can respond by abandoning H or by abandoning one of the auxiliary assumptions. The choice between these options is not itself determined by the experimental result.

Longino extends this insight in a distinctively feminist direction. She argues that the underdetermination is not merely a logical possibility but a permanent feature of actual scientific practice, and that the background assumptions which mediate the inferential relationship between evidence and hypothesis regularly encode social values. The choice of which auxiliary assumptions to treat as unproblematic — which features of the experimental setup to trust, which background regularities to accept as established — reflects judgments shaped by the social context of inquiry.

A concrete illustration: consider research on sex differences in mathematical ability. The evidence — observed differences in test scores between male and female populations — is consistent with several hypotheses: that there are innate cognitive differences, that differential socialization produces the observed differences, that stereotype threat depresses female performance, or that the tests themselves are biased. Choosing among these hypotheses requires background assumptions about the reliability of the test as a measure of mathematical ability, about the extent to which social conditions can be controlled for, and about which causal factors are most plausible given broader theoretical commitments. These choices are not determined by the test score data alone, and they regularly track the social values and political commitments of researchers.

3.3 The Value-Ladenness of Science

Longino distinguishes between two types of values that operate in science:

• Constitutive values (构成性价值): the internal norms of scientific practice — predictive accuracy, internal consistency, simplicity, scope. These are values that define what counts as good science, and they are generally endorsed even by those who espouse the value-free ideal.

• Contextual values (语境性价值): the social, moral, and political values that scientists bring to inquiry from their broader cultural context — values concerning gender, race, class, efficiency, and social utility.

The value-free ideal, on Longino’s account, is the view that contextual values can and should be excluded from scientific reasoning. Longino argues that this ideal is unachievable because contextual values inevitably shape the choice of research questions, the interpretation of evidence, and the assessment of theories through their influence on background assumptions.

3.4 Reconceiving Objectivity: The Four Conditions

If contextual values inevitably shape inquiry, what becomes of objectivity? Longino’s answer is to reconceive objectivity as a property not of individual scientists or individual research acts, but of scientific communities and their social practices.

Objectivity (客观性), on Longino’s view, is achieved when a community has in place social structures that enable genuine critical interaction with scientific claims — structures that allow background assumptions to be surfaced, challenged, and revised in light of criticism.

She identifies four features that a scientific community must exhibit to count as producing objective knowledge:

These conditions function as normative standards by which existing scientific communities can be evaluated and found wanting. Longino emphasizes that the conditions are difficult to satisfy and that actual scientific communities only partially realize them. The project of reforming science is, in part, a project of creating the social conditions under which these norms can actually be met.

Chapter 4: The Social Structure of Scientific Knowledge

4.1 Science as a Social Practice

One of Longino’s most important contributions is to shift the locus of epistemological analysis from the individual scientist to the scientific community. This move connects her work to the broader field of social epistemology (社会认识论) associated with figures such as Philip Kitcher and Alvin Goldman.

The traditional picture in philosophy of science — associated with logical empiricism — treated science as an idealized individual rational agent confronting evidence. Longino argues that this picture misrepresents how science actually works and how it achieves whatever epistemic virtues it has.

Real science is conducted by communities: research groups, disciplines, journals, grant-funding bodies, and universities. The epistemic virtues of science — the capacity to correct error, to resist idiosyncratic bias, to converge on well-supported conclusions — are properties of these communities and the norms that govern their interaction, not of any individual.

4.2 Kitcher and the Division of Cognitive Labor

Philip Kitcher’s work on the division of cognitive labor (认知劳动的分工) complements and complicates Longino’s account. Kitcher argues that a scientific community benefits from having its members pursue different research strategies, even when some of those strategies are less well-supported by current evidence, because the community as a whole explores the space of possibilities more efficiently.

Kitcher formalizes this insight using models from economics and decision theory. He shows that if all scientists follow the same strategy — pursuing the currently most-favored hypothesis — the community is vulnerable to getting locked into a false paradigm if that hypothesis turns out to be wrong. Conversely, if a minority of researchers pursue less favored hypotheses, the community retains the capacity to detect and correct error. The optimal distribution of cognitive labor is therefore not one in which all researchers maximize expected utility given their individual evidence, but one in which different researchers follow different strategies, generating what Kitcher calls “productive heterodoxy.”

This has a striking implication: the epistemic interests of a scientific community may be served by having members who hold different background assumptions — including values — as long as the social processes of the community enable productive exchange and convergence. Diversity of perspective is not merely a matter of social justice; it has epistemic value.

4.3 Goldman’s Social Epistemology

Alvin Goldman’s reliabilist social epistemology provides another angle on the relationship between social structure and scientific knowledge. Goldman holds that the epistemic quality of beliefs depends on the reliability of the cognitive processes that produce them. Applied to social settings, this means that the epistemic quality of the outputs of a scientific community depends on the reliability of the community’s collective cognitive processes — its methods of evidence gathering, theory evaluation, and consensus formation.

Goldman has argued for what he calls veritistic social epistemology: an account of social epistemic practices evaluated in terms of their tendency to produce true beliefs. On this view, the value of institutional arrangements in science — peer review, replication requirements, blind review — is their contribution to the reliability of the truth-tracking processes of the community.

Feminist philosophers have engaged critically with Goldman’s approach, arguing that it is insufficiently attentive to the ways in which power differentials within scientific communities undermine the reliability of collective processes. Peer review is not veritistically reliable if reviewers systematically discount the work of women and minorities — even if each individual reviewer is making what they take to be an honest assessment. The reliability of the community’s truth-tracking depends on the community’s freedom from structural biases that distort credibility assessments, and Goldman’s framework needs to be supplemented by the social and political analysis that feminist philosophy of science provides.

4.4 Diversity and Convergence: A Tension

A genuine tension runs through the literature on diversity in scientific communities. On one hand, the arguments from Longino, Kitcher, and feminist standpoint theory all support the value of cognitive diversity — of having different perspectives, different background assumptions, and different research strategies represented in the community. On the other hand, science aims at convergence — at reaching shared, well-supported conclusions that represent the state of knowledge on a question.

How can a community simultaneously value diversity and aim at convergence? The tension is real but not irresolvable. The key is to distinguish between diversity of background perspective and convergence on evidential conclusions. The ideal is a community in which diverse perspectives interact through the norms Longino identifies — genuine critical engagement, uptake of criticism, shared standards, equal intellectual authority — and in which this interaction produces convergence that is genuinely earned through argument and evidence rather than secured by power. This is the ideal of earned consensus: a consensus that has been achieved by a community that was diverse enough to surface and challenge its own background assumptions, and that has responded to criticism through rational persuasion rather than social pressure.

4.5 Critical Discursive Communities

Longino develops the notion of a critical discursive community (批判性话语共同体) to describe what a scientific community must be like to produce knowledge that is genuinely objective. The concept combines two ideas: the sociological observation that science is a communicative practice, and the normative claim that the epistemic quality of science depends on the quality of its internal discourse.

The four norms she identifies function as normative standards by which existing scientific communities can be evaluated and found wanting. Many real scientific communities fail the equality condition: the voices of women, of researchers from the global South, of those who work on topics associated with marginal groups, are systematically discounted in ways that cannot be justified by the quality of their arguments.

In The Fate of Knowledge (2002), Longino refines her position in response to critics who charged that her social epistemology was insufficiently “social” — that she retained too much of the individual-cognizer framework even after the sociological turn. She responds by more explicitly endorsing an account of knowledge as a collective achievement, not merely the aggregate of individual justified beliefs. She also engages more directly with the tension between pluralism and convergence. A thorough social epistemology must explain how communities with diverse starting points can converge on shared conclusions, and must specify which kinds of convergence are epistemically significant and which are merely the product of power asymmetries that suppress dissent.

Chapter 5: Standpoint Theory and Scientific Inquiry

5.1 Origins in Hegel and Marx

Standpoint theory (立场理论) holds that one’s social location — particularly one’s position within structures of domination — shapes what one can know, and that marginalized social locations may confer distinctive epistemic advantages. The roots of this view lie in Hegel’s master-slave dialectic, which suggested that the slave, because forced to engage with the material world through labor, achieves a more concrete understanding of reality than the master who depends on the slave’s mediation. Marx transposed this insight into claims about the epistemic privilege of the proletariat.

Sandra Harding and Dorothy Smith applied this framework to the social position of women, arguing that women’s experiences — especially their experiences of work, both paid and unpaid, and of embodiment — provide a vantage point from which aspects of social reality invisible from dominant positions can be perceived.

5.2 Harding’s Strong Objectivity

Sandra Harding’s most influential epistemological contribution is the concept of strong objectivity (强客观性). She argues that the conventional ideal of objectivity — which she calls “weak objectivity” — demands only that scientists apply their methods consistently to the objects under study. Weak objectivity leaves unexamined the values that shape the choice of research questions, conceptual frameworks, and background assumptions.

Strong objectivity demands that the critical gaze be turned on the inquiry process itself — on the social, political, and cultural conditions of knowledge production. This requires including as objects of study the standpoints of those who do the inquiry. Far from compromising objectivity, Harding argues, this reflexivity enhances it by surfacing assumptions that would otherwise remain invisible and unchallenged.

5.3 Wylie’s Feminist Standpoint in Archaeology

Alison Wylie offers perhaps the most philosophically careful development of standpoint epistemology. Working in the philosophy of archaeology, Wylie develops what she calls a tacking methodology — the practice of moving back and forth between background theories (drawn from biology, social theory, and feminist scholarship) and the material evidence uncovered in the field, refining each in light of the other.

Wylie distances feminist standpoint theory from what she calls its “strong” or “romanticist” formulations, which imply that marginalized knowers have automatic epistemic privilege simply by virtue of their social location. Instead, she argues that standpoint positions are:

• achieved rather than given: they require deliberate political and intellectual work, not merely the experience of oppression;
• resource-generating rather than truth-guaranteeing: they open up productive questions and help identify systematic blind spots, but do not automatically yield correct answers;
• defeasible: standpoint-derived insights are subject to empirical test and critical challenge like any other knowledge claim.

5.4 Collins and Black Feminist Thought

Patricia Hill Collins’s Black Feminist Thought (1990) extends standpoint epistemology by insisting on the intersection of gender, race, and class as axes of social location that generate distinctive epistemic resources. Collins argues that Black women’s experiences of domination along multiple dimensions — race, gender, and often class — produce a particular vantage point: what she calls the “outsider-within” (局外人-内部人) position.

Black women who work in white-dominated institutions are insiders in the sense of having access to the norms and knowledge of those institutions, and outsiders in the sense of never being fully integrated into them. This double consciousness — to use Du Bois’s term that Collins invokes — enables a distinctive form of critical perception, one able to see both the workings of the institution and its blind spots simultaneously. The outsider-within is not merely a witness to the gap between official claims and actual practice; she is epistemically positioned to perceive regularities and structures that insiders, precisely because they benefit from them, are motivated not to see.

Collins identifies several features of Black feminist epistemology that distinguish it from mainstream epistemological norms:

5.5 Sandoval and the Methodology of the Oppressed

Chela Sandoval’s Methodology of the Oppressed (2000) extends standpoint theory in a direction that emphasizes the tactical dimension of marginalized knowledge. Sandoval argues that the experience of U.S. Third World feminism — feminism produced from within multiple, overlapping systems of domination — has generated a distinctive methodology characterized by what she calls “differential consciousness”: the capacity to move strategically between different identity positions, ideological frameworks, and modes of political engagement.

Differential consciousness is not the same as standpoint in the Hardingian sense: it is not a fixed location that confers privileged access to certain truths. It is rather a cultivated capacity for fluidity — the ability to adopt different epistemic and political positions as the situation requires, without commitment to any single fixed standpoint as definitive. Sandoval associates this capacity with a form of tactical agency that is epistemically productive precisely because it is resistant to the reification of any single perspective as the authoritative one.

5.6 Critiques of Strong Standpoint Claims

Standpoint theory has attracted both sympathetic revision and sharp critique. Three main objections deserve attention.

The essentialization objection: Critics argue that standpoint theory risks essentializing social groups — treating women, or Black women, as if they share a unified experience and therefore a unified standpoint. But experience varies enormously within groups, and the assumption of a shared standpoint may suppress differences within the group while exaggerating differences between groups.

The privilege-without-warrant objection: Some critics argue that the inference from oppression to epistemic advantage is unwarranted. Being oppressed may sensitize one to certain features of social reality, but it does not automatically produce accurate beliefs. The history of social movements contains as many epistemic failures as successes among oppressed groups. Wylie’s distinction between achieved and automatic standpoint is an attempt to address this objection, but critics argue that it merely defers the question: what makes a particular achieved standpoint epistemically reliable?

The self-undermining objection: If standpoints are constituted by social location, and if the standpoints of dominant groups are epistemically distorting, what is the epistemic status of standpoint theory itself? Standpoint theory is produced by academic feminist philosophers who, whatever their social identities, occupy a relatively privileged position within the academy. Does the theory undermine itself?

Collins and Harding have both argued that standpoint theory is not self-undermining because it does not claim that standpoints automatically confer privilege — only that they generate resources that require further refinement through critical inquiry. The epistemic value of a standpoint is never automatic; it is always a product of the combination of structural position and the critical work of examining that position.

Chapter 6: Gender and the Culture of Science

6.1 Keller’s Psychoanalytic Critique

Evelyn Fox Keller’s Reflections on Gender and Science (1985) approaches the relationship between gender and science through a psychoanalytic lens, drawing on object relations theory — particularly the work of Nancy Chodorow — to argue that the cultural norms of scientific practice encode a specifically masculine conception of the relationship between self and world.

Keller observes that the language of scientific mastery — “conquering” nature, “penetrating” secrets, “dominating” phenomena — is not incidental rhetoric but reflects a deep-seated cultural script linking scientific knowledge with masculine control over a feminized nature. This script shapes not only the language but the goals and self-conception of inquiry.

6.2 Static and Dynamic Objectivity

Keller distinguishes between two conceptions of objectivity:

• Static objectivity (静态客观性): the ideal of knowing the world by distancing oneself from it completely, effacing all personal connection, and treating the object of inquiry as wholly separate from and subordinate to the knowing subject. Keller associates this with a masculine developmental trajectory characterized by rigid ego boundaries and anxiety about connection.

• Dynamic objectivity (动态客观性): an alternative ideal that Keller finds exemplified in the practice of Barbara McClintock. Dynamic objectivity involves deep engagement with the object of inquiry, a willingness to attend to its particularity and complexity, and an empathic responsiveness that allows the object’s own structure to guide inquiry.

6.3 The McClintock Case

Barbara McClintock’s career is Keller’s central case study. McClintock spent decades studying the genetics of corn (maize), developing an extraordinarily intimate knowledge of individual plants and their chromosomal behavior. Her discovery of genetic transposition — the movement of genetic elements from one position in the genome to another — was made possible by this intimacy: she noticed anomalies in the inheritance patterns of individual plants that other researchers had dismissed as experimental noise.

McClintock’s discovery was not recognized for approximately thirty years. The mainstream geneticists of her time operated within a conceptual framework that assumed genes were fixed elements of the genome, and her claims about transposable elements were treated as implausible or even incoherent. She was not ignored because her evidence was poor — her experimental methodology was meticulous — but because her findings did not fit the dominant theoretical framework, and because she was a woman working alone in a field that had become increasingly dominated by the molecular biology paradigm and its associated values of large-scale, collaborative, reductionist research.

6.4 Haraway and Situated Knowledges

Donna Haraway’s 1988 essay “Situated Knowledges” is one of the most cited texts in feminist philosophy of science, and one of the most challenging. Haraway argues against two symmetrical errors:

• God-trick (上帝把戏): the pretense of a view from nowhere, a perspective that claims universal validity by effacing its own situatedness. This is the error of traditional claims to scientific objectivity.
• Radical constructivism: the view that there is no world beyond discourse, that all knowledge claims are equally constructed and therefore equally valid or invalid.

Against both, Haraway proposes the concept of situated knowledges (情境性知识): all knowledge is produced from a particular location — an embodied, historically situated standpoint — and this situatedness is not a defect but the condition of the possibility of genuine knowledge. Feminist objectivity, on Haraway’s account, means claiming knowledge produced from specific, accountable locations, not from nowhere.

Haraway extends this insight with the concept of the “partial perspective”: because all knowledge is produced from somewhere, no single perspective provides complete knowledge of anything. Genuine objectivity is a collaborative, multi-perspectival achievement — an “objectively partial view” that acknowledges its own limitations and seeks to supplement itself through critical engagement with other partial perspectives. This is Haraway’s version of the feminist commitment to diversity: not as a political concession but as an epistemic requirement.

6.5 Barad’s Agential Realism

Karen Barad’s Meeting the Universe Halfway (2007) develops a more radical position she calls agential realism. Drawing on Niels Bohr’s philosophy-physics and feminist science studies, Barad argues that the dichotomy between subject and object that underlies both traditional scientific realism and the social constructivist critique is itself philosophically untenable.

For Barad, matter and meaning are not separate domains; they are intra-actively constituted — they come into being together through what she calls “material-discursive practices.” Scientific apparatuses are not passive instruments that transparently reveal a pre-existing world; they are material-discursive configurations through which phenomena are produced. The “phenomenon” — the object of scientific inquiry — is not something that exists independently of the apparatus; it is the whole configuration of apparatus, scientist, and world that produces measurable outcomes.

Barad’s agential realism is relevant to feminist philosophy of science because it provides a framework in which the social conditions of inquiry are not merely context for knowledge production but are constitutively involved in what gets produced as knowledge. The laboratory, the scientist’s body, the funding structure, and the theoretical framework are all part of the material-discursive apparatus through which phenomena are enacted, not merely observed.

Chapter 7: The Value-Free Ideal and Its Critics

7.1 The Value-Free Ideal Stated

The value-free ideal (无价值理想) holds that scientific reasoning, at least in its context of justification, should be governed solely by epistemic norms — evidence, logical consistency, simplicity — and that moral, political, and social values should play no role in evaluating hypotheses or theories. This ideal is associated with the logical empiricist tradition and with the popular self-understanding of science as a uniquely rational enterprise.

The ideal captures something important: we do not want scientific conclusions about climate, vaccines, or drug safety to be determined by commercial or political interests. The value-free ideal is meant to protect against exactly this kind of corruption.

7.2 Arguments Against the Ideal

Feminist philosophers of science, along with philosophers like Heather Douglas and Philip Kitcher, have mounted several distinct objections.

7.2.1 The Argument from Underdetermination

As noted in Chapter 3, no body of evidence uniquely determines a theory. Choices among underdetermined options must be made, and these choices inevitably draw on considerations beyond evidence — including social values encoded in background assumptions. Values are not merely external contaminants; they are part of the inferential machinery of science.

7.2.2 Heather Douglas’s Argument from Inductive Risk

Inductive risk (归纳风险) is the risk of being wrong — of accepting a hypothesis that is false, or rejecting one that is true. Douglas’s book Science, Policy, and the Value-Free Ideal (2009) develops the most sustained philosophical case for the ineliminability of values in science, centered on the concept of inductive risk.

Douglas argues that the standard philosophical picture treats value judgments as appropriate only in the context of application — deciding how to use scientific findings — and inappropriate in the context of justification — evaluating whether the evidence supports a hypothesis. Her argument is that this clean separation is impossible, because decisions about what constitutes sufficient evidence for a conclusion are always decisions about how to balance the risk of false positives against the risk of false negatives, and this balance depends on the consequences of being wrong.

Douglas argues that the role of values at the level of evidence evaluation is indirect but genuine: values determine the acceptable risk of error, which in turn determines the evidential threshold for accepting hypotheses. This is not the same as allowing values to override evidence; it is recognizing that the translation from evidence to conclusion is never purely formal but always involves risk management, and risk management is inherently evaluative.

7.2.3 Anderson’s Argument for Value-Laden Science

Elizabeth Anderson argues not merely that values inevitably enter science but that some forms of value-ladenness are epistemically desirable. Her argument proceeds in two stages.

First, she argues that the choice of research questions is irreducibly value-laden and should be. Science does not choose its questions from a value-free algorithm; it chooses them from among the practically and theoretically important questions that scientists and their funders consider worth pursuing. This choice reflects values, and there is nothing wrong with that — on the contrary, a science that chose its questions randomly, or according to purely aesthetic criteria, would be epistemically irresponsible.

Second, Anderson argues that social diversity in science — diversity in the values, perspectives, and life experiences of scientists — is epistemically valuable because it generates diversity in research questions, methods, and interpretive frameworks. A community of scientists all of whom share the same values, class position, and social experiences will generate systematic blind spots. Feminist values, in particular, have prompted the formulation of research questions that have yielded genuine empirical discoveries — about female primate behavior, about sex differences in cardiac presentation, about the social determinants of health — that androcentric science had failed to pursue.

7.2.4 Responses Defending the Value-Free Ideal

Not all philosophers are convinced that the value-free ideal should be abandoned. Several responses have been offered.

The corrupting values distinction: Defenders of a modified value-free ideal argue that the relevant question is not whether values enter science but whether they enter in epistemically corrupting ways. Some values — the value of accurate prediction, the value of explanatory depth — are constitutive of good science and should be included in the “epistemic” category. Others — commercial interests, political commitments — distort evidence evaluation in ways that undermine the epistemic goals of science. The revised ideal is not that science is value-free, but that it should be free of epistemically corrupting values.

The transparency response: A second defense holds that the problem identified by feminist critics is not values per se but hidden values — values that operate without acknowledgment and therefore cannot be subject to critical scrutiny. The remedy is transparency: scientists should make their value-laden background assumptions explicit, subject them to critical review, and ensure that different value perspectives are represented in the critical process. This response is broadly consistent with Longino’s position and accepts much of the feminist critique while retaining a modified version of the value-free aspiration.

The institutional solution: Some defenders argue that the problem of values in science is best addressed not by abandoning the ideal but by constructing institutions — peer review, replication requirements, adversarial collaboration — that make the influence of idiosyncratic values on conclusions less likely. The value-free ideal functions as a regulative ideal, not an achievable standard, and the appropriate response to its non-attainment is institutional reform, not philosophical revisionism.

Chapter 8: Case Studies in Sex, Sexuality, Race, and Medicine

8.1 The Case Study Method in Philosophy of Science

Philosophy of science, including feminist philosophy of science, has increasingly moved toward case study methodology (案例研究方法). Rather than deriving epistemological conclusions from armchair reflection on an idealized image of science, case study methodology begins with the detailed examination of actual episodes in the history or sociology of science.

This methodology serves feminist philosophy of science well for several reasons. First, it makes the abstract arguments about bias and values concrete and testable. Second, it demonstrates that the feminist critique is not merely ideological complaint but is responsive to how science actually works. Third, case studies in politically sensitive areas — research on sex differences, sexual orientation, and race — directly reveal the human consequences of epistemically defective science.

8.2 Research on Sex Differences: Simon Baron-Cohen and the Extreme Male Brain

The study of sex differences is one of the most contested areas in feminist philosophy of science. Simon Baron-Cohen’s theory of the “extreme male brain” and the systemizing-empathizing (S-E) dimension provides a detailed case study in the feminist critique of this research tradition.

Baron-Cohen’s theory claims that male brains are, on average, oriented toward systemizing (analyzing, building, and manipulating rule-governed systems) while female brains are oriented toward empathizing (identifying and responding to emotional states). Autism spectrum disorder represents, on this view, an “extreme male brain” — a profile marked by intense systemizing and impaired empathizing.

Feminist critics have raised multiple levels of objection.

At the methodological level, the primary instruments measuring systemizing and empathizing are self-report questionnaires — the Systemizing Quotient and the Empathizing Quotient. These instruments are susceptible to the same stereotype threat effects documented in cognitive testing: subjects who are primed to think about gender differences in cognitive style report responses more consistent with gender stereotypes than subjects who are not so primed. The measures therefore do not straightforwardly assess the cognitive dispositions they claim to measure; they measure the interaction between those dispositions and the social context of assessment.

At the conceptual level, the categories of systemizing and empathizing are defined in ways that build in gender-coded associations from the outset. “Systemizing” is characterized through examples from engineering, physics, and rule-governed games; “empathizing” through examples from social care and emotional responsiveness. This means that individuals socialized into different gender roles will score differently not necessarily because of differences in underlying cognitive capacity but because of differences in exposure, practice, and motivation in gender-coded domains.

At the interpretive level, even if real average differences in S-E scores exist, the inference from cognitive profile to neural architecture to evolutionary selection pressure is multiply underdetermined. The observed differences are consistent with socialization hypotheses, with the effects of different activity environments on neural development, and with measurement artifacts — without requiring any hypothesis about evolved sex differences in brain architecture.

8.3 Race Science: Category, Confounding, and Political Context

Race science (种族科学) refers to research that treats race as a biologically meaningful category and uses it to explain differences in social outcomes — in cognitive ability, in health, in behavior. Feminist philosophers of science, along with philosophers of race such as Naomi Zack and philosophers of biology such as Philip Kitcher and Evelynn Hammonds, have subjected this research to sustained critique.

The central philosophical issue concerns category validity: does “race,” as used in scientific research, pick out a biologically coherent grouping? The consensus in genetics and population biology is that it does not. The genetic variation within so-called racial groups is substantially greater than variation between them — the human species is relatively genetically homogeneous, and the visible phenotypic differences that underlie folk racial classification (skin color, hair texture, facial morphology) represent a tiny fraction of total genetic variation and correlate poorly with other genetic characteristics.

This does not mean that race is a biologically meaningless category for all purposes: race as a social category has real biological consequences because it affects exposure to environmental stressors, access to medical care, the physiological effects of chronic discrimination, and the distribution of environmental toxins. But these effects flow from the social reality of race, not from any biological essence. Using race as an independent variable in biomedical research without controlling for these social determinants conflates social causes with putative biological ones — a fundamental methodological error with significant consequences for health policy.

The political context of race science compounds the philosophical problems. Research on race and cognitive ability has been systematically mobilized by particular political interests — interests in restricting immigration, reducing welfare expenditure, and opposing educational equity programs. The patterns of funding, publication, and media amplification of race science reflect these interests and are not independent of them. Feminist philosophers have argued that this political context is not merely background; it shapes which hypotheses are considered worth testing, which datasets are collected, and which findings are deemed worthy of publication, in ways that introduce systematic biases difficult to detect from within the research paradigm.

8.4 Medicine: Cardiovascular Disease and the Persistence of the Default Male

The cardiovascular medicine case, introduced in Chapter 2, illustrates the persistent consequences of androcentric research paradigms even after the problem has been formally recognized. The 1993 NIH Revitalization Act required the inclusion of women and minorities in federally funded clinical trials — a formal response to the documented exclusion of women from earlier research. Yet feminist philosophers of science and medical researchers have documented that the inclusion mandate has not fully corrected the underlying epistemological problems.

First, inclusion without disaggregated analysis does not generate usable knowledge about sex-specific effects. If clinical trials include women but do not analyze data separately by sex, the resulting findings may be dominated by the male subjects (who still constitute the majority in many trials) and may fail to detect effects that are present specifically in women.

Second, the conceptual frameworks used to interpret cardiovascular research remain shaped by the male-default assumption. Diagnostic criteria for acute myocardial infarction, risk stratification models, and treatment guidelines were developed on predominantly male populations and continue to be calibrated to male physiological norms. Correcting this requires not merely adding women to existing studies but redesigning the research questions — asking what cardiovascular disease looks like in women, what risk factors are most predictive in female populations, what treatment protocols work best for women.

Third, the hierarchy of clinical evidence — the privileging of randomized controlled trials over observational studies — may itself be gendered. Conditions that affect primarily women (such as lupus, fibromyalgia, and many autoimmune disorders), or that present differently in women, have historically received less RCT attention because they were less commercially and professionally prioritized. The hierarchy of evidence forms produces systematic blind spots that track gender and race.

Chapter 9: Diversity, Scientific Communities, and Epistemic Justice

9.1 The Epistemic Case for Diversity

The arguments developed in previous chapters converge on a strong claim: diversity (多样性) in scientific communities is not merely a matter of fairness or legal compliance — it is an epistemic requirement for the production of objective knowledge.

The argument has multiple strands:

1. The Longino argument: Genuine critical interaction requires that the perspectives of community members be diverse enough to surface each other’s background assumptions. A homogeneous community will share too many background assumptions to challenge the values encoded in them.

2. The Kitcher argument: A division of cognitive labor in which different researchers pursue different strategies is more efficient than one in which all pursue the dominant approach. Diversity of background perspective is one mechanism by which diversity of research strategy is generated and maintained.

3. The standpoint argument: Researchers who occupy marginalized social positions have access to experiences and observations that are invisible from dominant positions, and these observations can generate productive research programs.

4. The Wylie-Fehr argument: Diversity in who asks the questions influences which questions get asked, and this in turn influences which phenomena get studied and which explanatory frameworks get developed and tested.

9.2 Fehr’s Research on Demographic Diversity

Carla Fehr’s empirical research on the relationship between demographic diversity and epistemic outcomes in science provides quantitative grounding for the philosophical arguments above. Fehr has argued that the epistemic benefits of diversity are not merely theoretical but demonstrable through the sociology of science — and that current patterns of underrepresentation of women and minorities in many scientific fields represent a genuine epistemic loss, not merely a social injustice.

Fehr’s work focuses particularly on the relationship between diversity and the detection of androcentric bias. She shows that in fields with higher proportions of women researchers, research on sex differences is more likely to: (a) examine female subjects as the primary population rather than as controls for male subjects; (b) operationalize concepts in ways that are not gender-coded from the outset; (c) attend to socialization and environmental explanations alongside biological ones; and (d) be cited and built upon by researchers across a range of methodological traditions.

This empirical pattern is what one would predict on the basis of the philosophical arguments about standpoint and background assumptions: researchers who have personal experience of the phenomena being studied, or who occupy social positions from which certain patterns of inquiry appear natural, will be more sensitive to the androcentric assumptions embedded in standard research protocols. The empirical finding that diversity correlates with improved epistemics supports the philosophical thesis that the social composition of scientific communities is an epistemological question.

9.3 Structural Versus Perspectival Diversity

A crucial distinction in this literature is between structural diversity and perspectival diversity.

The relationship between structural and perspectival diversity is important but not deterministic. Structural diversity tends to generate perspectival diversity because people from different social locations tend to develop different perspectives on the phenomena their fields study. But the relationship can be undermined by institutional pressures toward conformity: a scientific community can increase structural diversity while maintaining perspectival homogeneity if the institutions through which scientists are trained, evaluated, and promoted systematically reward conformity to existing norms and penalize heterodox approaches.

This means that the epistemic case for diversity is not merely a case for changing the demographic composition of science; it is a case for changing the institutional structures through which scientific credibility is produced and maintained. Increasing structural diversity without reforming those structures may achieve justice goals without achieving epistemic goals.

9.4 Justice and Epistemic Value: The Relationship

The feminist philosophy of science literature reveals a deep relationship between epistemic value and social justice, but the nature of this relationship requires careful analysis. Three positions are worth distinguishing:

The contingent convergence view holds that justice and epistemic value typically (though not necessarily) point in the same direction: including marginalized perspectives tends to be both just and epistemically beneficial, while excluding them tends to be both unjust and epistemically harmful. The convergence is contingent because, in principle, the best epistemic outcomes might in some cases be achieved by arrangements that are unjust.

The constitutive view holds that justice is constitutively connected to epistemic value: the epistemic norms that Longino identifies — equality of intellectual authority, genuine uptake of criticism — are both epistemic and justice norms, and satisfying them requires both epistemic and social change. On this view, the line between epistemic and social reform is not sharp.

The priority of justice view holds that justice takes priority: we should reform science because it is unjust to exclude certain voices and perspectives, and the epistemic benefits are a welcome but secondary consequence. This view is defended by those who worry that framing diversity primarily in epistemic terms reduces marginalized researchers to instruments of epistemic improvement, rather than recognizing their independent standing as persons and knowers.

9.5 Structural Barriers to Epistemic Diversity

Carla Fehr’s research on women and minorities in science and philosophy addresses the gap between the epistemic case for diversity and the sociological reality of scientific communities. She identifies several structural mechanisms through which diversity is suppressed:

• Implicit bias (内隐偏见): well-documented in studies of hiring, promotion, peer review, and citation, implicit bias against women and racial minorities systematically discounts their work and reduces their authority within scientific communities.
• Climate effects: the social climate of scientific environments affects retention of minority researchers independently of bias in formal evaluations; environments that are overtly or subtly hostile to certain groups lead those groups to exit the field at higher rates.
• Citation practices: patterns of who cites whom reinforce hierarchies of intellectual authority. Work by women and minority scholars is systematically undercited relative to its quality and influence, reducing its visibility and its authors’ career advancement.
• Gatekeeping: the norms of what counts as legitimate science — what counts as a researchable question, what methodologies are acceptable, what counts as a significant finding — are shaped by the values and interests of those who design and dominate scientific institutions.

9.6 Epistemic Injustice and Its Forms

Miranda Fricker’s concept of epistemic injustice (认识论不公正) provides a framework for understanding these structural problems. Fricker distinguishes two forms:

9.7 Toward Epistemically Just Scientific Communities

The normative conclusion that emerges from this analysis is not simply that more diverse scientists should be recruited — though that is necessary. It is that the norms and practices of scientific communities must be reformed so that they can function as the critical discursive communities that Longino’s epistemology requires.

This means:

• Reforming peer review to reduce the influence of implicit bias on credibility assessments.
• Expanding the range of questions and methodologies that are recognized as scientifically legitimate.
• Creating institutional mechanisms by which community members can surface and challenge background assumptions without career penalty.
• Attending to whose testimony is heard and credited in disciplinary debates, not only who is nominally included.

9.8 Nelson’s Feminist Empiricism

Lynn Hankinson Nelson’s Who Knows (1990) offers a complementary perspective by extending Quine’s naturalized epistemology in a feminist direction. Nelson argues that the unit of epistemic activity is not the individual scientist but the community (共同体) — communities are the primary knowers, individuals know only derivatively as members of communities.

This communitarian epistemology has significant feminist implications. If communities are the primary knowers, then the composition and norms of communities are epistemological questions, not merely sociological ones. Excluding women from epistemic communities is not merely unjust; it produces worse knowledge, because the community’s epistemic capacities are diminished.

Nelson’s Quinean commitment to naturalism means that her feminist epistemology is continuous with empirical science rather than in tension with it. She sees feminist science studies — the empirical study of how gender affects scientific practice — as part of the naturalized epistemological project of understanding how knowledge is actually produced.

9.9 Conclusion: A Research Program, Not a Settled Doctrine

Feminist philosophy of science is best understood not as a set of settled conclusions but as a research program (研究纲领) organized around a cluster of related questions:

• How do social values enter scientific inquiry, through what mechanisms, and with what consequences?
• What does it mean for science to be objective, and what social conditions are required for objectivity to be achieved?
• How do the social positions of scientists — their gender, race, class, sexuality — shape what they can and cannot know?
• What is the relationship between epistemic justice and social justice?
• How should scientific institutions be reformed to produce knowledge that is both more reliable and more genuinely representative of human experience?

The work of Longino, Harding, Keller, Haraway, Collins, Wylie, Nelson, Anderson, Kitcher, and Fehr, taken together, does not provide a unified theory. But it constitutes a rigorous, empirically informed, and politically engaged body of work that has permanently altered how philosophy of science understands the relationship between knowledge and power. The task now is to extend, refine, and apply these insights — including through the kind of careful case study analysis and institutional reform that this seminar takes as one of its organizing occasions.