3.1 What is feminist standpoint theory?

Feminist standpoint theory combines the idea of situated knowledge (from above) with claims about systems of social oppression. On this view, subjects have different ‘social locations’ – meaning different statuses as socially oppressed or socially privileged – and these different social locations come with different experiences, which have the potential to enable different epistemic perspectives.

This idea has its roots in Marxist historical materialism. On Lukács’ (Reference Lukács1971) interpretation, Marx argued that the different social locations of the bourgeoisie and of the proletariat lead them to have different perspectives on economic exchange and the social relations that hold between the two groups. From the perspective of the proletariat, the oppressive nature of these social relations is, or can be made, visible, whilst from the perspective of the bourgeoisie the oppressive nature of these social relations is obscured.

Feminist standpoint theorists have taken the central idea behind this and applied it to gender oppression, and to multiple intersecting dimensions of oppression (e.g. oppression based on both race and gender). This central idea is that subjects who are socially oppressed have distinct experiences, and after critically reflecting upon these experiences they can turn their perspective into a ‘standpoint’ – an epistemically privileged perspective from which the nature of relevant social relations is visible. Subjects who are not oppressed don't have these experiences, and as a result are less likely to achieve a standpoint, and more likely to continue to have a distorted understanding of social relations. This amounts to the socially oppressed group having what is often called an epistemic advantage:

Different authors cash this idea out in different (though related) ways. We'll discuss some specific proposals in the next two subsections, but before doing so it's worth emphasising several caveats that feminist standpoint theorists frequently make about the thesis of epistemic advantage. The thesis is often misunderstood by those who don't acknowledge these caveats, and who then think the thesis is obviously implausible, or objectionable on feminist grounds.

First, the epistemic advantage is not supposed to depend on essential categories. There needn't be any properties which are essential to membership of the groups which have the epistemic advantage (Hartsock Reference Hartsock1997; Smith Reference Smith1997; Wylie Reference Wylie, Figueroa and Harding2003). As Fricker puts it, standpoints don't depend on oppressed people (or even particular subsets of oppressed people, such as Latina women or gay men) being the same. It only requires that their experiences are similar in certain ways (Fricker Reference Fricker1999: 201).

Second, this point is true of the property of possessing epistemic advantage, too. Possessing epistemic advantage is neither a necessary nor a sufficient condition on membership of a particular group. Where it is present in oppressed people it is dependent on the fact that they live in a world containing oppression, and it is also possible that it can be achieved by non-oppressed people (Medina Reference Medina2012).

Third, this is because epistemic advantage is not automatic, but must be achieved through critical reflection. And this achievement is collaborative; it is not the work of an individual (Fricker Reference Fricker1999: 202–3; Wylie Reference Wylie, Figueroa and Harding2003; Medina Reference Medina2012).

Fourth, it needn't be an advantage with regards to all objects of knowledge. Standpoint theorists are usually clear that the argument for advantage will need to be made on a domain-by-domain basis, and that it is easiest to find in domains of knowledge which involve social relations (e.g. Harding Reference Harding1991: 46; Wylie Reference Wylie, Figueroa and Harding2003: 37).Footnote ⁸ Fricker suggests that knowledge of “the social world” should be “fragmented” even further into knowledge of “relevant areas of the social world” (Fricker Reference Fricker1999: 203).

With these caveats in place, we'll spend the remainder of this section discussing some specific standpoint theory proposals, and evaluating whether they render justification dependent on social factors in a constitutive sense, or just in a causal one.

3.2 Access to evidence

First, let's look at an account which only renders justification as dependent on social factors in the causal sense, and so is compatible with the classical conception of knowledge.

Nancy Hartsock (Reference Hartsock, Harding and Hintikka1983) focussed on the “sexual division of labour” as the source of epistemic advantage. Her paper on this is a classic, which sparked off the rest of the literature on standpoint theory in philosophy. Her aim was to adapt the Marxist framework of economic oppression to explain the epistemic benefits that can come along with gender oppression. Marx argued, or is sometimes interpreted as arguing, that proletarian workers are better positioned to understand the social relations underpinning society and economics than their economic superiors are, because they are more in touch with ‘material necessity’ – the concrete, physical resources that are required for human safety and survival.

Hartsock argued, similarly, that women's roles as wives and mothers make them responsible for, and so particularly familiar with, lots of things which are essential to the functioning of society – such as cooking, cleaning, and child rearing. But she added that, for women, this connection with material necessity is even more pronounced than that of male workers. Women aren't just expected to produce and maintain objects (like chairs and boots), but also people – which is a more repetitive, extended and involved process. And their roles as producers encompass more of their lives than those of male workers – at the end of the day the male worker goes home and leaves behind his role, to some extent, but wives and mothers don't have this opportunity. This amounts to women having much more evidence, or many more justified beliefs, about what is required to satisfy our needs than men (who are not typically responsible for these things) usually have.

As well as affecting the evidence that they have access to, Hartsock argues that women's experiences can also affect the way they weight the importance of different questions and activities. One example Hartsock provides, from Marilyn French's novel The Women's Room, makes this point especially vividly:

The idea being captured here is that the roles women play in society affect the weight they give to different concerns and questions. There are some questions and topics that they don't have the luxury of spending time and effort on (and others that they are vividly aware of the material importance of) because they have to clean up piss on a regular basis.

This is the kind of feminist point that classical epistemologists often seem to be most resistant to – the idea that women's subjective experiences as caregivers can affect what they know is seen as somewhat radical. But interestingly, this point is completely compatible with the classical conception. Remember that the classical conception allows that social factors can affect justification in the causal sense, by leading subjects to have access to different evidence, or to weight the importance of various questions differently. So Hartsock's points about women's social roles affecting justification via the evidence they have access to and the weightings they attribute to different questions is, perhaps surprisingly, compatible with the classical understanding of objectivity about justification.Footnote ⁹

3.3 Conceptions of evidence

Other aspects of standpoint theory can, we think, be seen to understand justification as dependent on social factors in the constitutive sense, as well as the causal one. The other hugely influential, classic paper that sparked off a lot of standpoint theory is Patricia Hill Collins’ paper ‘Learning from the Outsider Within’ (Collins Reference Collins1986). Collins is a sociologist, and in this paper she produced a detailed case study of the epistemic advantages of black women with regard to the discipline of sociology.

She identifies a number of different epistemic advantages that black women may develop. One in particular is of interest to us here: she argues that black women – because of the interlocking nature of the oppression that they face – are in a better position to compare and evaluate the epistemic resources that the discipline relies on than those who don't face oppression, or who face oppression only along one dimension:Footnote ¹⁰

Collins’ claim is that there are assumptions in the sociological standpoint which black women stand in a unique relation to – sexist and racist assumptions which are harmful to black women – which they are therefore less likely to accept and more likely to criticise. This is a claim about the effect of social factors on the component parts of a justificatory framework, and so seems to assume constitutive dependence.

This point about constitutive dependence has since been made more explicit by Sandra Harding (Reference Harding1991, Reference Harding1995), in her work on values in science. Harding defends a standpoint theory on which epistemic advantage lies in the ability to identify a particular set of values that play a role in knowledge production (or, more specifically, in science). Tellingly, she refers to this set of values as constitutive values.Footnote ¹¹ She argues that as women and other groups who are socially oppressed have typically been left out of science in the past, they are better able to identify constitutive values that have an effect on science from within (such as culture-wide assumptions about race and gender), whilst those from groups who have traditionally made up scientific research teams can only identify overt values (such as the interests that particular businesses might have in funding research into the products that they sell) that affect science from outside (Harding Reference Harding1995: 335). This is because, as Collins argued, those who have been excluded from a discipline due to oppression are less committed, and more resistant, to the oppressive assumptions embedded within that discipline's justificatory resources.

A real-life illustration of this claim can be found in the field of behavioural endocrinology. Researchers interested in the impact of hormones on sexual behaviour discovered that the sexual behaviour of rhesus monkeys peaks at the time when the females are ovulating, and wanted to determine how the male monkeys ‘knew’ when to initiate sex at this optimal (in terms of procreation) time. It's now accepted that the peak in sexual activity during ovulation is explained by female rhesus monkeys initiating sex, but for a long time the research carried out overlooked this fact and instead focused on the behaviours and capacities of the male monkeys.

Kim Wallen (Reference Wallen1990, Reference Wallen, Wallen and Schneider2000)Footnote ¹², a behavioural endocrinologist specialising in female sexuality, tracks how this situation eventually changed. He says that the crucial turning point was in 1976 – over 30 years after examples of female initiation were first recorded – when one of the “patriarchs” who had “dominated” the field published a paper distinguishing proceptivity (the active solicitation of sexual activity) from the more passive receptivity. According to Wallen this reflected a cultural shift due to “the rise of the women's movement, or to the dramatic increase in female graduate students in endocrinology, or to both” (Reference Wallen, Wallen and Schneider2000: 296).

This seems, then, to be an example of epistemic advantage uncovering constitutive sexist values. An influx of female, or feminist, researchers led to greater scrutiny of the dominant perspective and its justificatory resources, and enabled sexist assumptions about male sexual proactivity and female passivity to be uncovered. And, importantly, this is not an example of justification (or evidence) being dependent on social factors in the causal sense. The male researchers who initially failed to explain the timing of the rhesus monkeys' sexual contact had access to the same bodies of evidence as later researchers did – the behaviour that would later be treated as evidence of proceptivity in female rhesus monkeys (their slapping the ground in front of males) had been observed and published on over 30 years before the term ‘proceptivity’ was coined.Footnote ¹³ Something more complicated was going on here.

At this point, Boghossian's simplified understanding of evidence begins to cause confusion. If we think that access to observations and data is equivalent to evidence and justified belief, then it's hard to explain why the early researchers went wrong. But if we recognise that there's an additional step required to turn observation into evidence, and that this step involves value-laden processes like interpretation, then it becomes much easier to explain what happened. The researchers had access to the relevant observations, but didn't conceptualise them, or what they represented, as evidence because of social factors (such as the sexist assumptions about female passivity). In offering this account – as standpoint theorists following Harding would do – the classical conception of knowledge is rejected in favour of a form of social constructivism (as we have defined it).

In this section, we have explored which sense of dependence feminist standpoint theorists think is relevant when it comes to the relation between social factors and justification. We saw that, in some cases, the dependence described is causal, and so doesn't contradict the classical conception of knowledge. However, some accounts involve a more controversial constitutive dependence, which does contradict the classical conception. These accounts are forms of social constructivism.

4.1 What is feminist empiricism?

Feminist empiricism combines the empiricist view that experience and observation provide the least defeasible evidence we have about how the world is with the idea that feminist social values can play a legitimate role in the process of inquiry. This combination fits naturally within the framework of social naturalised epistemology. Here's Elizabeth Anderson:

Like Quine (Reference Quine and Quine1969), feminist empiricists look at the causal forces that shape what we know and how we know it. But, where Quine imagined studying the reactions of experimental subjects to stimuli in controlled laboratory settings, feminist empiricists study the process of knowledge production in concrete social contexts, with a focus on how conceptions and norms of gender impact on that process. For instance, Anderson cites studies which show that the results of social science survey research depend on the perceived gender or race of the interviewer and that the data produced in anthropological fieldwork depends on the gender of the anthropologist (Reference Anderson1995a: 60–1). The conclusion she draws is that, the more diverse a social scientific or anthropological research team is, the more likely they are to obtain accurate results from surveys and fieldwork. The upshot is that diversity is desirable from the scientific point of view as well as the moral point of view. There is an ‘internal’ scientific rationale for favouring diversity, as well as an ‘external’ moral rationale.

The more fundamental difference between feminist empiricism and Quinean naturalised epistemology is that, where Quine accepted a basic distinction between facts and values, feminist empiricists question this distinction. Here's Anderson again:

Anderson is making two points here. Her first point is that, if we can demonstrate that there are gender biases in currently accepted scientific methods, then it is legitimate to modify those methods. We don't consider this to be much of a departure from Quinean naturalised epistemology. While Quine talks about epistemology as a ‘branch’ of psychology, he still thinks it can serve an ameliorative purpose. The whole point of studying the actual process of knowledge production is to improve it (see Quine Reference Quine, Hahn and Schilpp1986; Roth Reference Roth1999).

Anderson's second point is that values, specifically feminist values, can legitimately inform the actual conduct of inquiry. This is clearly a departure from Quinean naturalised epistemology. For Quine, our values set the goals of inquiry, but they play no role in its actual conduct; for feminist empiricists, our values both set the goals and play a role in the conduct of inquiry. But what is this role? We will now turn to this question.

4.2 Feminist social values

We will start with an answer to this question that has been proposed by Helen Longino (see Longino Reference Longino1990, Reference Longino1994, Reference Longino1997, Reference Longino2002). Once we have looked at her answer, we will argue that it is tantamount to rejecting objectivism about justification, and accepting a form of social constructivism. We will finish by commenting on a dispute within feminist empiricism about what legitimises the feminist empiricist appeal to feminist values in theory choice in science. As we will see, this debate has consequences for whether feminist empiricism is committed to a form of social constructivism.

Anderson summarises Longino's account of the role of feminist values as follows:

Here is our reconstruction of Longino's argument:

1. 1. There is a ‘logical gap’ between empirical data and theory in the sense that the data only provide evidential support for a given theory in conjunction with certain background assumptions.

2. 2. In the philosophy of science literature, the standard way to bridge this logical gap is by appealing to the theoretical virtues (consistency, simplicity, explanatory power, fruitfulness). It is these virtues that secure the evidential link between data and theory.

3. 3. Appealing to feminist social values to bridge the gap is as legitimate a way (or maybe even a more legitimate way) of securing the evidential link between data and theory as appealing to the standard theoretical virtues.

4. 4. Therefore, feminist social values can legitimately secure the evidential link between data and theory, and so figure in the justification for accepting scientific theories.Footnote ¹⁴

The first premise articulates a familiar idea in the philosophy of science: the available empirical data always underdetermine theory choice in the sense that two observers can always disagree about what theory the data support by making different background assumptions (see Quine Reference Quine1951; Duhem Reference Duhem1954). Some took the failure to observe the stellar parallax in the 17th century as evidence that the earth did not revolve round the sun because they assumed that the stars are not far away. Others took it as evidence that the stars are far away because they assumed that the earth revolved round the sun (see Anderson Reference Anderson1995b: 28). As this example might suggest, it is sometimes possible to resolve these disputes by examining the background assumptions themselves. But this is not always possible. The second premise states the standard response to this problem in the philosophy of science literature: the theoretical virtues can be invoked to argue that one theory is preferable to the alternatives because of its consistency with other theories, its simplicity, its explanatory power or its fruitfulness (see Kuhn Reference Kuhn and Kuhn1977).

The third premise is the crucial premise in the argument. It states that it is as (if not more) legitimate to appeal to feminist social values to bridge the gap between observation and theory as it is to appeal to the standard theoretical virtues. For instance, it can be as legitimate to appeal to the relative complexity of a theory as a reason to prefer it over a simpler theory as it is to appeal to the relative simplicity of a theory as a reason to prefer it over a more complex theory. Whether the appeal is legitimate will depend on the case at hand. The conclusion follows from these premises: if feminist social values secure the evidential link between observation and theory, they are part of what justifies accepting the theory. They determine what ‘counts as’ evidence for the theory.

While the first two premises raise interesting questions in the philosophy of science, we lack the space to deal with them here. Accordingly, we will focus on the third premise. We will start by looking at Longino's defence of this premise. We will then show that, if her argument is successful, it gives us good reasons to reject objectivism about justification, and accept social constructivism.

Here is Longino's proposed list of feminist social values (see Longino Reference Longino1997: 21–3):

• • Empirical adequacy.

• • Novelty of explanatory framework. (Contrast: consistency.)

• • Ontological heterogeneity. (Contrast: simplicity.)

• • Explanatory complexity or mutuality of interaction. (Contrast: explanatory power or generality.)

• • Applicability to human needs and potential to decentralise power.Footnote ¹⁵

This list includes the familiar value (or virtue) of empirical adequacy: it is never legitimate to accept a theory that conflicts with observation. The final value (applicability to human needs) can be seen as a slightly different take on the traditional virtue of fruitfulness. But the other three values contrast with traditional virtues. In what follows we will focus on these contrasting values.

Let's look at each virtue in turn (what follows is based on Longino Reference Longino1997: 23–7). First, take consistency. The thought behind appeals to consistency is that it is a mark in favour of a theory if it is consistent with our other theories. But the problem with this is familiar from discussions of coherentism in epistemology: bodies of beliefs can be internally consistent yet entirely false. Thus, the legitimacy of appeals to consistency depends on the status of our other theories. Second, take simplicity. While theories shouldn't posit entities for no reason, this hardly means that we should in general prefer theories that posit fewer entities to theories that posit more. We have no reason to think that the universe is particularly simple, and plenty of reason to think it is enormously complex. Thus, we need to evaluate, on a case-by-case basis, whether there is a sound rationale for the entities posited by particular theories. Finally, take explanatory power. Longino follows Cartwright (Reference Cartwright1980) in arguing that explanatory power and generality usually comes at the cost of literal truth: powerful and general sets of scientific laws are usually false, and true scientific laws are usually full of ceteris paribus clauses that rob them of any generality. This is not to say that it is never legitimate to appeal to explanatory power. But it does show that the legitimacy of appeals to explanatory power depends on the specifics of the case.

Longino takes this to show that whether an appeal to feminist social values or an appeal to the standard theoretical virtues is legitimate is going to depend on the specifics of the case at hand. But in what sort of case would it really be more legitimate to appeal to feminist social values? Here, Longino cites cases where scientific advances have resulted from – or would have resulted from – a preference for theories that are novel, heterogeneous or explanatory complex. Take, for instance, Barbara McClintock's ground-breaking work on genetic transposition, which involved looking at the differences between individual cobs of corn (see Longino Reference Longino1994: 477; Anderson Reference Anderson1995a: 60). Longino argues that McClintock's work reflected a preference for complex explanations that only work for very specific phenomena over unifying explanations that cover a diverse range of phenomena. From the fact that scientific advances have resulted from a preference for theories satisfying her feminist values rather than the traditional theoretical virtues, she concludes that, at least in these cases, it was more legitimate to appeal to these feminist values than it would have been to appeal to the standard theoretical virtues.

While more could be said about Longino's argument, we will now turn to its implications for objectivism and social constructivism about justification. An initial difficulty here is that, where Boghossian talks about beliefs, Longino talks about scientific theories. This reflects a sub-disciplinary divide; Boghossian is using the standard terminology in epistemology, whereas Longino is using the standard terminology in philosophy of science. We are going to bridge this divide by assuming that, while scientific theories don't simply consist in sets of beliefs, sets of beliefs are an important component of scientific theories, and that, while there is more to justifying a scientific theory than justifying the component beliefs, justifying a scientific theory requires justifying its component beliefs. Thus, if we can show that the justification for accepting a scientific theory (constitutively) depends on social factors, then we have ipso facto shown that the justification for accepting the various beliefs that are part of that theory depend on social factors.

On Longino's view, the available observational data will always logically underdetermine theory-choice. In order to bridge this gap, we need to appeal to a set of virtues or values. But here Longino thinks it is as (if not more) legitimate to appeal to feminist values to bridge the gap as it is to appeal to the traditional theoretical virtues. So feminist values bridge the gap between observation and theory, and so secure the evidential link between observation and theory. They determine whether the observation counts as evidence for a particular theory in the first place. Values therefore play a crucial role in the process of justifying scientific theories. But the idea is not that our values merely lead us to pay attention to particular bits of evidence, and therefore cause us to accept one theory rather than another. Rather, the idea is that they partly constitute our justification for accepting one theory rather than another. Thus, on Longino's view, feminist social values can partly constitute our justification for accepting scientific theories, and therefore a form of social constructivism is true.

We have now shown that Longino assigns a role to feminist social values on which they partly constitute our justification for accepting scientific theories. We want to end this section by considering a debate within feminist empiricism about the status of feminist values.Footnote ¹⁶ Maya Goldenberg (Reference Goldenberg2015) distinguishes between two ‘streams’ in feminist empiricism: the ‘community-based social knowledge’ stream (of which Longino is representative) and the ‘values as evidence’ stream (of which Anderson is representative).Footnote ¹⁷ Both streams agree that values (feminist values included) play a legitimate role in inquiry. But they disagree about what legitimises their use. Put very roughly, for Longino, values are legitimised by community criticism. An appeal to values is legitimate if it withstands a public process of critical scrutiny, where the process satisfies certain conditions:

• • There must be publicly recognised forms for criticism (e.g. peer-reviewed journals).

• • There must be uptake of criticism (e.g. engagement with comments from peer review).

• • There must be public standards for evaluation of criticism (e.g. discipline-wide norms, like a requirement of replicability).

• • There must be tempered equality of intellectual authority (see Longino Reference Longino2002: 128–35).

In contrast, for Anderson values are legitimate insofar as they are evidence-based. Anderson (Reference Anderson2004) considers two different ways in which someone researching the effects of divorce could frame things. One option would be to frame divorce as the traumatic breakup of a family. Another option, taken in some feminist work on divorce, would be to frame divorce as a transformation of the family unit that presents an opportunity for personal growth for the divorcee. Both framings are value-laden in the sense that they involve taking an evaluative stance towards divorce. But we can assess which framing is more fruitful and illuminating without putting them both through a public process of critical scrutiny. Anderson argues that the assumption that divorce involves traumatic breakup is problematic because it is difficult (to say the least) to isolate harms caused by the divorce from harms caused by the relationship prior to the divorce. If this is right, then the feminist way of framing divorce is legitimate because it is more fruitful and illuminating than the alternative.

One might wonder whether this difference between Longino and Anderson is reflected in a difference with respect to their views about social constructivism: views like Longino's are committed to a form of social constructivism, whereas views like Anderson's are not. If this is right, there is a similar divide within feminist empiricism to the divide within feminist standpoint theory: for Longino feminist values determine what counts as evidence for or against a given theory, whereas for Anderson feminist values don't ‘stand outside’ of evidential relations. Goldenberg appears to endorse this interpretation:

If values are amenable to the same empirical modes of inquiry as factual statements then it must be possible to offer evidence for and against them, and if it is possible to offer evidence for and against them then they can't determine what counts as evidence in the first place. This suggests that, while some feminist empiricists (e.g. Longino) endorse a form of social constructivism, and so reject the classical conception, others (e.g. Anderson) aim to fit feminist empiricism into the classical conception of knowledge.