Introduction
Looking at specific populations of knowers reveals that the presumption of sameness within knowledge communities can lead to a number of epistemological oversights. A good example of this is found in the case of survivors of sexual violence. This group of individuals demonstrates the epistemic significance of difference by providing us with insight into some of the central issues that arise for our theories of justification and knowledge. In this paper I argue that this case offers a new perspective on the debate between the epistemic internalist and externalist. In particular, I argue that it demonstrates that internalist justification is not a necessary condition for knowledge, even if there are cases in which, when combined with true belief, it is sufficient. But it is not the only sufficient condition for knowledge. The case of survivors of sexual violence supports the externalist idea that an individual can know even if she cannot provide a reasoned defense of her non-accidentally true beliefs about the world, so long as her beliefs were formed using methods that tend to produce true beliefs; hence reliable belief- forming mechanisms, when combined with true belief, are also a sufficient condition for knowledge. While arguments in favor of externalist theories of knowledge are not new, looking at this group of knowers provides us with a fresh insight into the complicated psychological dimensions of belief formation and the implications that this has for an epistemology that demands reasons that are first-person accessible. This case study exposes the everydayness of knowledge without citable reasons and the significance of this phenomenon in our epistemic lives, a fact that is overlooked in the current debate by internalists and externalists alike.
Epistemic Success and the Standard Analysis
Knowledge is a success term. To say that a subject S knows that p is to pay S a compliment: to acknowledge that S is epistemically successful with respect to her belief that p. But what, precisely, does this mean? At the very least, it means that the subject’s beliefs are true. But, as Plato suggested long ago, we might come across a true belief by chance. Since success is a measure of achievement, something more is needed here. In order to get at the idea that knowing represents an accomplishment, epistemic success demands that an individual’s true beliefs about the world be non-accidentally true for her.Footnote 1 This idea is captured by the standard analysis of knowledge, according to which S knows that p if S has a justified true belief that p, provided there are no defeaters of the kind that can prevent a non-accidentally true belief from being knowledge. The case study below illustrates, however, that an individual can achieve epistemic success with respect to her beliefs about the world even if those beliefs are not justified in the traditional evidentiary sense of the term; even if, in other words, that individual cannot defend her true beliefs about the world. What this means is that justification, understood in this traditional sense, is not the only way of ensuring that our true beliefs about the world are non-accidentally true. As this case study implies, another way to fulfill the non-accidental clause for knowledge is the use of reliable belief-forming mechanisms. This suggests that the traditional analysis of knowledge as justified true belief depends on an inflated use of the “non-accidental” clause, and this in turn results in the conflation of the idea of “non-accidental” with the notion of “justification.” In what follows I argue that we could go a long way towards resolving the stalemate between internalists and externalists by sticking to Plato’s insight that what is needed for knowledge is true belief plus an account (Theaetatus, 201c-210a), and leaving open just what this account might be. If this is right, then knowledge requires only that a subject’s true beliefs be true for her non-accidentally. Put formally:
S knows that p if and only if S’s true belief that p is true for S non-accidentally
As I hope to show, this deflated definition of knowledge is robust enough to explicate the concept of knowledge while at the same time allowing us to recognize cases of epistemic success where justification, understood in the traditional evidentiary sense, is absent.
Internalism and Externalism
The precise meanings of epistemic internalism and externalism are a matter of some debate, but there is a measure of agreement on a number of central points.Footnote 2 While both concepts are typically construed as fulfilling the justification condition for knowledge, as I suggest above I think it is better to see them as each articulating a standard that a true belief must meet in order to be considered non-accidentally true. One way of understanding internalism and externalism is to see them as offering different ways of securing this third condition for knowledge. The internalist formulates this non-accidental clause through the traditional notion of justification, which involves reasons that people have and have access to. What this means is that for the internalist, beliefs are justified in virtue of the evidence that we have in support of them. This evidence might consist of facts about our own mental states, coherence with other beliefs that we have, or facts about the external world.Footnote 3 But whatever it is that justifies a subject’s beliefs about the world, for the internalist, this evidence must be cognitively accessible to the subject such that it can serve as a reason for the subject to adopt the beliefs that she does. So, for example, it might be true that George Bush is in China right now, but that fact does not provide me with a reason for believing that Bush is in China unless it is a fact that I am aware of. Of course, no amount of evidence is a guarantee of truth; despite our best efforts we may be tricked or innocently misled into believing a well-supported falsehood.Footnote 4 But so long as I am doing the best that I can do, qua knower, in terms of gathering evidence for my beliefs, then according to internalism I will have fulfilled the non-accidental clause for knowledge. Accordingly, on this view, if a subject S has a non-accidentally true belief that p then S will be able to defend p, to stand behind p in the game of giving and taking of reasons. In these cases, according to the standard definition of knowledge as justified true belief, we conclude that S knows that p.
This characterization of internalism makes it clear how internalist justification is connected to the idea of deontology, in this case epistemic duty.Footnote 5 I will return to this point later, but for now let me make the connection explicit. The idea just is that we fulfill our duties, qua epistemic agents, so long as we accept only those beliefs that we think are likely to be true, in virtue of our access to the evidentiary connections presented upon reflection. On this account, then, to be justified is to be epistemically blameless.Footnote 6 Thus, for any subject, mirroring the gap between having justified beliefs and having true beliefs is the gap between fulfilling our epistemic duties and having true beliefs. That is to say that a subject can be epistemically praiseworthy even if her beliefs about the world are false.
The externalist intuition, on the other hand, is that there are circumstances where S knows that p even though S cannot defend her belief that p, cases where justification, understood in this traditional sense, drops out. These are cases where S has acquired p in a reliable way such that S’s belief that p is, again, non-accidentally true. In these cases, according to externalism, S’s reliable belief-forming mechanism satisfies the third condition for knowledge even if S lacks cognitive awareness of this reliable process; what matters is that it is in fact reliable, such that it leads to more true beliefs than false ones, and not that a subject sees it as such. A familiar example from the literature that illustrates this case is that of the industrial chicken-sexer. This is an individual who reliably sorts hatchlings into males and females by inspecting them without, apparently, having a hot clue as to how she does it. There is no explicit recipe that chicken-sexers learn, but with enough practice they just catch on. It is thus a well-developed skill, but not one that is available to the chicken-sexer through introspection. In fact, although the chicken-sexer suspects that she makes her discriminations according to visual signs, tests have shown that they depend on olfactory ones. Finally, not only does the chicken-sexer reliably discriminate between male and female chicks but she has a strong conviction about which chicks are male and female, even though she cannot say exactly why.
At first glance this case appears to be ripe for an externalist analysis, for while in some sense chicken-sexers respond blindly, we want to say that their beliefs are nevertheless non-accidentally true; they have been formed in a way that routinely leads to true beliefs about the sexes of chicks, whether or not the chicken-sexer is clued into how she is doing this. However, at least as I have described it thus far, the case of the chicken-sexer fits comfortably within an internalist framework. For as critics of externalism have pointed out, if S acquires p in a reliable way and S is able to see herself as using reliable methods, then S can offer up these methods as a reason for why she holds the belief that p. In other words, even if the chicken-sexer does not know what her methods of success are, so long as she is aware that she is good at sorting chicks into sexes then she can use that fact as a reason for why she holds the beliefs that she does and justification, in the traditional evidentiary sense, is preserved.
What this illustrates is that pressure is applied on the traditional understanding of knowledge as justified true belief only when a subject S cannot use her reliability as a reason for why she holds the beliefs that she does. We are forced to move beyond internalism in the case where the subject not only is blind insofar as she lacks awareness of her reliable belief-forming mechanisms, but is “superblind” (to borrow a phrase from Robert Brandom) in that she does not even see herself as employing said mechanisms (Brandom Reference Brandom1998, 376). The importance of this point cannot be overstated. The plausibility of the kinds of cases that demand an externalist account depends on it, and so consequently does the need for the kind of disjunctive theory of knowledge that I am advocating for here. Thus, with respect to the chicken-sexer, we are to imagine that while she sorts chicks with convincing reliability, and thus rightly believes that this chick is a male chick and that one is a female, she lacks access to both how she does this and to the very fact that she does this. Accordingly, these individuals are not able to use their reliably formed beliefs inferentially; that is, as premises or conclusions in arguments; they cannot themselves depend on their own reliably formed true beliefs. They exhibit an appropriate response to stimuli, like bulls to flapping red material, but they are unable to rely on their own responses. Still, Brandom, for one, argues that these cases of superblindness should count as genuine instances of knowledge, but he also claims that they are, like the chicken-sexer case, “delicate and special,” essentially “fringe phenomena” (1998, 375 and 381). I think that Brandom is right that these cases should count as knowledge, but he is wrong that they are fringe phenomena. On the contrary, they are run-of-the-mill. Our failure to recognize this can be blamed, at least in part, on the kinds of examples found in the literature and which are taken as prototypical for an externalist theory of knowledge. These examples strain our imagination; they range from the eccentric case of the chicken-sexer to the otherworldly case of clairvoyance.Footnote 7 This emphasis on oddball examples belittles the complex emotional and psychological dimensions that feature in our epistemic behaviour, and in doing so detracts from our ability to see the significance of the phenomenon of superblindness in our ordinary epistemic lives. The everydayness of knowledge without citable reasons becomes plain when we turn our attention away from the chicken-sexer and to the case of the survivor of sexual violence and her beliefs about the world that are formed in the aftermath of a traumatic experience.
Survivors of Sexual Violence
In what follows I shall argue that the phenomenon of superblindness is one common among survivors of sexual violence. This phenomenon is widespread within this population of knowers as a result of the experience of psychological trauma. Indeed, I think that the results argued for here can be generalized to the broader population of trauma survivors; I am focusing on this particular subclass because I myself am a rape survivor and as such I have first-hand experience of the aftermath of sexual violence. This experience leaves me well placed to consider the sorts of beliefs formed by survivors in the aftermath of trauma as they face upheaval with respect to some of their most deeply entrenched beliefs about the world. These beliefs range from the very general, for instance about the preponderance of sexual violence in the world, to the specific, for instance about the character of one’s assailant. And this disruption in one’s belief set is typical whether we are talking about adult survivors of childhood sexual abuse, adult survivors of adult rape, or child survivors of sexual abuse.
As an example, let us take the case of the adult survivor of adult rape. It is commonly understood that following a traumatic event such as rape an individual undergoes intense personal suffering, which impacts her emotional, psychological, and physiological well-being. The standard view in psychiatric medicine today is that this condition may, depending on a number of factors (including but not limited to the severity of the attack), develop into Post-Traumatic Stress Disorder (PTSD).Footnote 8 Shortly I will mention a number of theoretical interpretations of PTSD, but for now I want to suggest that this lingering symptomatology is only one side of the aftermath of psychological trauma — namely, the shattered self.Footnote 9 The other side of this aftermath is a shattered world view, which is the consequence of sexual violence on the survivor’s beliefs about the world. Elsewhere I have argued that survivors experience cognitive dissonance as they are faced, often quite suddenly, with evidence that flies in the face of some of their core, implicitly held beliefs about the world (Freedman Reference Freedman2006).Footnote 10
So, for example, prior to my rape I believed in what has been called the “just-world philosophy,” according to which the world is basically fair, and so long as “you are sufficiently careful, intelligent, moral, and competent, you can avoid misfortune” (Matsakis Reference Matsakis1998, 26). I believed, in other words, that it was within my power to protect myself against serious personal harm. I struggled with this belief in the aftermath of my rape. What evidence there was to support it had been overturned in light of this violent experience. The bare facts of the rape were enough to cast serious doubt on the myth of my own immunity.Footnote 11 For the survivor, the idea that the world is basically safe is, like a Popperian conjecture, falsified in one bold test. Most survivors of sexual violence struggle, as I did, with a smattering of inconsistent beliefs about their safety and place in the world. They wonder: “Am I not able to keep myself from serious harm? Is the world not a safe place, or was I somehow incautious? Was my attack a sign of an individual failing — mine — or a social one? Am I responsible for the attack? But how could I be responsible for the brutalizing actions of a knife-wielding rapist?” And so on. After a violent and traumatic experience, survivors wrestle with the question “What am I to believe?” In some cases, the answer to this leads to a rejection of the old belief set and the so-called “just-world philosophy” in favor of an altered world view, one which is more in keeping with the reality of their traumatic experiences.
Statistical Evidence
As it turns out, statistics confirm that the beliefs that comprise this altered world view tend to be true.Footnote 12 It follows that the extreme acts of violence that make up a traumatic experience tend to have a veridical impact when it comes to facts about sexual violence against women. I think it is important to stress that this is a contingent truth and not a necessary one. It is a social fact. The experience of sexual violence is currently a reliable way of forming beliefs about the world because of the way the world currently is. Another key point to emphasize is that the kind of cognitive shift that follows from a traumatic experience does not result in true beliefs about just anything; for instance, how old Churchill was when he died or the number of tigers in the Toronto Zoo. The experience of sexual violence tends to result in true beliefs qua survivor; that is, beliefs regarding the preponderance of violence against women and women’s relative safety in the world.
Of course, not all beliefs formed in the aftermath of extreme violence are true. For example, there is the case of the survivor who generalizes too broadly from her experience and forms the false belief that “all men are predators,” or the survivor who forms precisely the wrong belief, such as “If I don’t go out at night, I’ll be safe from rape.”Footnote 13 Thus, the trauma of sexual violence does not result exclusively in true beliefs. But that is not a claim I want to defend. Rather, my point is that a traumatic experience has a veridical impact when it comes to facts about sexual violence against women, which is to say that it results in a greater proportion of true beliefs to false ones.Footnote 14 Importantly, what this establishes is that beliefs formed in the aftermath of a traumatic experience, when true, are not true by accident. And indeed, these sorts of beliefs are borne out by the universality of the everydayness of sexual violence, in particular against women and girls.Footnote 15 As Amnesty International states in its 2004 report It’s in Our Hands: Stop Violence against Women, violence against women is “the most outrageous human rights scandal of our times,” and sexual violence against women is universal (2004, iii and 6).Footnote 16 This is true both in war-torn societies, where rape is a common practice of warfare used by both governments and armed groups,Footnote 17 and in times of relative peace. For example, 17.6 % of women in the United States have survived a completed or attempted rape. Of these, 21.6% were younger than age 12 when they were first raped, and 32.4% were between the ages of 12 and 17 (National survey, 2000). Unsurprisingly, the picture is even worse in less-developed countries. The World Health Organization, in a recent report, suggests that women living in the third world and women living in poverty suffer disproportionately (2002, 99). The Mohawk writer Patricia Monture-Angus speaks eloquently of this in her discussion of Aboriginal women in Canada, a demographic group whose living conditions are often compared to those of women in the third world. According to Monture-Angus, these women experience violence of a “non-incidental” sort, which she describes as not just one rape or assault or battering, but a lifetime of it. Thus, as she notes, in this case statistics are misleading for they play down the routineness of violence and in doing so disguise what she refers to as “the utter totality of the experiences of violence that Aboriginal women face” (1995, 170-171).
Theories of Trauma
If the account that I have presented thus far is correct, then following a traumatic experience it is typical for survivors of sexual violence to revise a number of their core beliefs about the world. Moreover, these revised beliefs tend to be true, as evidenced by the statistics on sexual violence. What I now want to suggest is that while a traumatic experience of violence can trigger a change in one’s belief set, according to our current best theoretical understandings of the post-traumatic experience it also represents a major obstacle which can undermine a survivor’s ability to see herself as a reliable indicator of these same facts. The problem is that survivors of sexual violence often have trouble accessing the details of their traumatic experiences, and so a fortiori are unable to see that experience as a reason for why they hold the beliefs that they do. Thus, what we have in the case of the survivor of sexual violence is a subject who, qua survivor, has certain reliably formed, i.e., non-accidentally true, beliefs about which she has deep conviction, but which she cannot defend.
For my purposes here, the details of current theories of trauma are not important; a rough sketch of the main accounts should suffice. The basic idea and one which is shared by most competing theories of trauma is that, in the language of folk psychology, survivors “block out” some or all of the details of their traumatic experiences. Depending on which interpretation of trauma we are looking at, the memory of the traumatic event might be forever repressed or temporarily dissociated from.Footnote 18 In either case, the traumatic event is out of the cognitive reach of the traumatized individual, and the subject thus lacks easy access to the details of the traumatic event. There is some general agreement that the psychological blocking is caused by the fact that the memory of the attack is too distressing for the survivor to consciously process and recall. Some trauma theorists argue that in the moment of trauma an individual’s cognitive and perceptual capacities are shattered, rendering the traumatic scene literally inaccessible to the survivor for recollection (Leys Reference Leys2000, 9).Footnote 19 A less extreme (and more plausible) view, one which dates back to the late nineteenth century and the work of Pierre Janet, is that in the moment of trauma the victim distances herself from her own traumatic experience; she splits, or dissociates. As a result she ends up witnessing her own attack, but as a kind of spectator. Accordingly, the memory of the trauma is not integrated into conscious memory but is instead split off from conscious recall.Footnote 20 Thus, on this view, the traumatic scene is in principle accessible to the subject but the activity of remembering it involves a deliberate attempt at reintegrating the unconscious traumatic memory with the conscious memory. This process of memory consolidation can be long and difficult and may likely depend on a variety of therapeutic interventions, i.e., the “talking cure” or some other form of therapy.Footnote 21, Footnote 22
Do survivors dissociate in the moment of trauma; do they “split,” as Janet long ago suggested? Is the traumatic scene repressed and held in the unconscious memory? Is it in principle unavailable to the trauma survivor, or principally accessible but only after a therapeutic recollecting and integration of the traumatic memory? The emergence of brain imaging technologies and other developments in cognitive neuroscience over the past few decades may one day help us find an answer to these difficult questions. Indeed, these developments might also help us to better understand precisely how a memory that is inaccessible to a subject might nevertheless deeply impact her beliefs about the world.
This is a pressing concern for my account, and while I will not elaborate an answer to it here I want at least to gesture at a plausible solution. A long-standing view in cognitive neuroscience is that the memory is not a unitary system — an idea that was fleshed out by Tulving (Reference Tulving, Tulving and Donaldson1972) but which dates back at least to William James (1896). The neuropsychological evidence supports a distinction between three components of the memory: procedural memory, which is our memory for habits and dispositions; semantic memory, which stores concepts and propositions; and episodic memory, which stores memories of events (Tulving, Reference Tulving1983). So one idea that renders plausible the suggestion that trauma survivors revise their beliefs in light of a violent experience even if they cannot quite recall that experience is that the extreme stress impairs their episodic memories, which they dissociate from, while leaving their semantic memories intact. This would explain how a traumatic event can be informational, even if the trauma survivor lacks conscious access to that information.
In any case, it seems clear that whatever interpretation of the post-traumatic experience we prefer, only a naïve understanding of trauma would posit the traumatic memory as something that is readily accessible to the trauma survivor. Even if full-blown repression of trauma is rare (McNally Reference McNally2003), trauma certainly is not, and it affects the beliefs of traumatized individuals in ways that they often do not fully understand. This fact alone is enough to support my conclusion as to the prevalence of superblindsightedness while remaining neutral with respect to the precise nature of psychological blocking.
The Limits of Internalism
The point I have been arguing for is that in the aftermath of a traumatic experience, survivors of sexual violence form beliefs about the world that they are not always able to defend. Because of the psychological factors just described, these survivors often fail both to see themselves as reliable reporters of their traumatic experience and also to see the implications of that experience with respect to their relative safety in the world.Footnote 23 Real-life examples of this sort saturate the literature on trauma and are retold most convincingly by the psychologists and psychotherapists who work with trauma survivors.
For example, I once heard a story of a woman who refused to leave her daughter’s bedroom door unguarded during family gatherings that included her dad’s brother; she used to hire a babysitter to sit outside the girl’s room. The woman, call her Sue, had a belief that p, where p is “my daughter is not safe around her uncle.” Accordingly, she refused to leave her daughter unattended during these family events, but when pressed could not say exactly why.Footnote 24 Later, after much psychotherapy, Sue was able to remember being assaulted by this uncle when she herself was a child. This traumatic experience was the reason for Sue’s belief that p. I take this to be an exemplary case of superblindness. Sue’s belief that her daughter was not safe around her (Sue’s) uncle was a strongly held, non-accidentally true belief that, at least for a period of time, lacked justification. Because p satisfied the condition of being a non-accidentally true belief for Sue, I think what we want to say is that Sue knew that p, for what we have here is an individual who had not just a non-accidentally true belief about the world, but one that she had come to trust. This individual, in other words, had a true belief that was non-accidentally true for her and which she was rightly convinced was true, even though she could not say exactly why. We can see that there were good reasons for Sue’s belief about her uncle; it is just that she could not see them. But if that is right, then we have to make room in our theory of knowledge for reliably formed beliefs to count as a sufficient condition for knowledge, when those beliefs are true. Because Sue’s belief lacked internal justification it seems fair to say that Sue did not exercise cognitive responsibility in forming her belief, but we simply beg the question if we assume that this denied her knowledge. This is a point worth slowing down for.
As we saw earlier, the internalist conception of justification comes bundled up with the notion of deontology. As a result, to know, on an internalist account, implies doxastic responsibility. But the very question at stake is whether or not this is a precondition for knowledge. And I think the answer to this is, No. After all, I too am cognitively irresponsible when it comes to my beliefs about the importance of milk in one’s diet or the health benefits of broccoli, but I want to say that I know these facts.Footnote 25 Nor do I remember how it is that I know that next year is a leap year, but I am certain of this fact, and I do not see why the epistemic status of my belief will change once I Google the information to confirm it. Indeed, cases of forgotten evidence present to internalist theories a kind of challenge similar to the one I have sketched here; they help to illustrate the dilemma we face with cognizers who apparently know, but cannot defend their non-accidentally true beliefs about the world. These cases of knowledge without citable reasons are similar to the case of the trauma survivor, but represent a less extreme example. Still, both cases illustrate the phenomenon of repressed reasons for belief and consequently both cases reveal the shortcoming of an epistemology that demands first-person accessibility to those reasons.
Entitlement
The case of the survivor of sexual violence is meant to exemplify the idea that there is more than one way for individuals to achieve epistemic success, which is to say that there are at least two ways for our true beliefs about the world to be non-accidentally true for a subject, and thus two ways for a subject to know (again, provided there are no defeaters of the kind that can prevent non-accidentally true belief from being knowledge). One way is by gathering evidence for the belief; the other way is by using reliable belief-forming mechanisms. Both methods have positive force in providing rational support for propositional attitudes. I think the notion of entitlement may help to make this idea more clear. If we understand entitlement as broader than justification, then we can say that those beliefs that are non-accidentally true for a subject are just those beliefs that the subject is entitled to. This understanding of entitlement fits well with our everyday use of the term, and it helps to give shape to the idea that epistemic success demands a kind of achievement while leaving open the particular character of that achievement. Tyler Burge (Reference Burge1993) has suggested that “entitlements are epistemic rights or warrants that need not be understood by or even accessible to the subject” (Burge Reference Burge1993, 458). His characterization of an epistemic right fits well with my analysis here. Individuals who have reliably formed beliefs about the world are entitled to those beliefs; they have an epistemic right to them, even if they lack justification for them.
I think this point needs extra emphasis in the case that I have described, for the survivor of sexual violence faces more hurdles than are faced in other typical cases of knowledge without citable reasons. This is because the sorts of beliefs formed by the survivor, qua survivor, paint a picture of the world that is routinely dismissed by our contemporary culture, a culture that keeps well hidden the universality of sexual violence against women. What this means is that even if the survivor can access the details of her own traumatic experience, she faces further external hurdles — social, political, economic, and legal ones — that give her further reason to doubt her own reliability. Indeed, despite increased awareness and public concern over the past few decades about violence against women, the everydayness of sexual violence remains a social secret. It is thus no surprise that even in the absence of superblindness, women tend to doubt their own credibility as reliable reporters of their own traumatic experiences. Indeed, I think it is plausible to suppose that the inconsistency between what survivors experience and the public narrative provides an incentive for survivors to dissociate, as a strategy to combat the dissonance and distress brought on by the conflicting stories. One important step in assuring the dissonance will be resolved for the survivor is by articulating a theory of knowledge that recognizes her epistemic entitlement to her reliably formed true beliefs.
Conclusion: The Limits of Externalism
My main purpose here has been to provide support for the externalist intuition that an individual can know, even if she cannot provide a reasoned defense of her non-accidentally true beliefs about the world by offering a case study that makes that intuition plausible. The current literature on externalism, with its focus on eccentric examples such as that of the chicken-sexer, persuades us to think of the reliabilist insight as relating to a small minority of oddball cases. This distorts our understanding of the psychological and emotional dimensions of belief formation and in doing so presents an overly simplistic view of the cognitive access that individuals have to their own beliefs. The case study I have examined here shows the limits of internalism with its insistence that reasons be first-person accessible by exposing its overly rationalistic bent, which deprives the honorific of knowledge from those individuals who are entitled to it. One conclusion that we can draw from this analysis is that internalist justification is not a necessary condition for knowledge, even if there are cases in which, when combined with true belief, it is sufficient. Moreover, if a theory of knowledge is to be accountable to epistemic communities like the one described here, then it will need to recognize the epistemic success of those individuals who have reliably acquired beliefs about the world. The case of the survivor of sexual violence dictates that reliabilism, when combined with true belief, will also be a sufficient condition for knowledge. But there are limits to externalism as well. While there may be individuals who, with respect to some of their core beliefs about the world, know, even though they cannot defend those beliefs, knowledge without citable reasons is not possible as a global phenomenon; that is, with respect to all the beliefs that we hold. Individuals who are never in a position to see why they hold the beliefs that they do hardly resemble concept users in a cognitive practice. These individuals could never rely on their own responses, never use their reliably formed beliefs inferentially; that is, as premises or conclusions in arguments. It is hard to even imagine these subjects as individuals; they would be less like cognitive agents than they would be like thermometers to temperature, exhibiting the appropriate responses to stimuli. Even with this restriction on externalism in place, there is nevertheless an important difference between the reliable responses of the survivor of sexual violence and the reliable responses of the thermometer, since the thermometer will never be in the position of being able to defend its responses. In the case of the trauma survivor, however, there is at least the possibility that she can enter into the cognitive game of giving and taking of reasons. In the meantime, an externalist theory of knowledge allows us to recognize the epistemic success of her reliably acquired true beliefs about the world, and in doing so implicitly acknowledges that there are different ways of acquiring beliefs which are germane to different populations of knowers.
