In this provocative and important paper, two starkly opposing theories of the relationship between knowledge and belief are pitted against one another. View 1 represents the way knowledge has long been understood in epistemology: Representations of knowledge depend on prior representations of belief, and therefore the attribution of belief is more psychologically “basic” than the attribution of knowledge. View 2 is the surprising idea that representations of knowledge are more basic than representations of belief.
The authors argue that a wealth of evidence from developmental, comparative, and social psychology, as well as from experimental philosophy, supports view 2 and refutes view 1. However, I suggest that both of these views are actually correct, and that the evidence supporting view 2 doesn't undermine the longstanding view of knowledge captured by view 1. This is because the word “knowledge” in English corresponds to two concepts (indeed, other languages use multiple words for different types of knowing that capture some of this polysemy); the concept of knowledge referred to in view 1 is a different concept of knowledge than the one referred to in view 2.
View 1 concerns the concept of propositional knowledge that has long been a topic of interest in epistemology. To know that P requires a justified belief that P, and it also requires that P is true. (It also likely requires additional factors; see Gettier, Reference Gettier1963; Starmans & Friedman, Reference Starmans and Friedman2012.) The appropriate contrast class for this concept of knowledge is belief. That is, a question about knowledge-belief asks: “Does she know that P, or does she merely believe that P?” Hence, when considering propositional knowledge, it's impossible to attribute knowledge without first attributing belief, because having a belief is just one component of having knowledge.
View 2 concerns knowledge in the sense of “awareness.” To have knowledge in this sense only requires not being ignorant. A question about awareness asks: “Does she know that P, or is she ignorant of P?” When considering whether someone is ignorant, the issue isn't the presence or absence of a belief; it is whether or not there is awareness.
The research reviewed in the target article reveals that infants, young children, and nonhuman primates attribute knowledge in this latter sense. And, as carefully illustrated by Phillips et al., this does not require attributing belief. Infants successfully predict how others will behave when aware of an object's location (usually, although not always, through visually observing it). However, when an observer's perceptual access to the location of an object has been disrupted, although adults would expect that they still have a true belief, infants now act as if they are ignorant, and have no expectation as to their behavior. Similarly, chimpanzees have demonstrated an understanding of awareness and ignorance, but there is no clear evidence that they are able to represent the beliefs of others, whether true beliefs or false beliefs. Furthermore, there are circumstances in which adults more readily attribute knowledge than belief (e.g., Myers-Schulz & Schwitzgebel, Reference Myers-Schulz and Schwitzgebel2013; Phillips et al., Reference Phillips, Strickland, Dungan, Armary, Knobe and Cushman2018), suggesting that this notion of knowledge persists through development. All of this supports view 2.
However, there is another sense of knowledge that adults are concerned with. Perhaps the most famous quest for knowledge of this sort was that of Descartes, who asked how he could know whether the things he believed to be true were actually true. Descartes had plenty of beliefs that he was very confident about, such that he was sitting in his dressing gown by the fire, holding writing papers in his hands. But he worried that he might be dreaming, or insane, or deceived by a malignant demon. He wanted to know whether these beliefs rose to the level of knowledge, and ultimately concluded that there was one thing he did know: “I knew that I was a substance whose whole essence or nature is simply to think, and which does not require any place or depend on any material thing, in order to exist” (Descartes, Reference Descartes1980/1637, p. 18). Here, Descartes is clearly referring to propositional knowledge, not simply awareness.
Phillips et al. might argue that this notion of knowledge is special to philosophers; knowledge in the “awareness” sense is the only concept that exists in everyday use. But this clearly isn't the case. Ordinary people ask all the time whether someone knows something or just believes it. This might mean asking whether someone is certain of their belief, whether someone's belief is true, or whether they have the right kind of evidence for their belief. In all these cases, however, we are first attributing a belief to someone, and then asking whether this belief rises to the level of knowledge. We might also withhold attribution of knowledge even in cases where there is awareness. Someone in the desert, half mad with thirst, might see a (real) lake in front of him – might be fully aware of it – but worry that it's a hallucination. “I believe there is water over there, but I just don't know for sure!” he might cry out to himself.
Properly distinguishing these two concepts of knowledge reveals two things. First, representations of knowledge in the sense of awareness are more basic than representations of belief. Second, representations of belief are more basic than representations of propositional knowledge. View 2 is correct, but so is view 1.
In this provocative and important paper, two starkly opposing theories of the relationship between knowledge and belief are pitted against one another. View 1 represents the way knowledge has long been understood in epistemology: Representations of knowledge depend on prior representations of belief, and therefore the attribution of belief is more psychologically “basic” than the attribution of knowledge. View 2 is the surprising idea that representations of knowledge are more basic than representations of belief.
The authors argue that a wealth of evidence from developmental, comparative, and social psychology, as well as from experimental philosophy, supports view 2 and refutes view 1. However, I suggest that both of these views are actually correct, and that the evidence supporting view 2 doesn't undermine the longstanding view of knowledge captured by view 1. This is because the word “knowledge” in English corresponds to two concepts (indeed, other languages use multiple words for different types of knowing that capture some of this polysemy); the concept of knowledge referred to in view 1 is a different concept of knowledge than the one referred to in view 2.
View 1 concerns the concept of propositional knowledge that has long been a topic of interest in epistemology. To know that P requires a justified belief that P, and it also requires that P is true. (It also likely requires additional factors; see Gettier, Reference Gettier1963; Starmans & Friedman, Reference Starmans and Friedman2012.) The appropriate contrast class for this concept of knowledge is belief. That is, a question about knowledge-belief asks: “Does she know that P, or does she merely believe that P?” Hence, when considering propositional knowledge, it's impossible to attribute knowledge without first attributing belief, because having a belief is just one component of having knowledge.
View 2 concerns knowledge in the sense of “awareness.” To have knowledge in this sense only requires not being ignorant. A question about awareness asks: “Does she know that P, or is she ignorant of P?” When considering whether someone is ignorant, the issue isn't the presence or absence of a belief; it is whether or not there is awareness.
The research reviewed in the target article reveals that infants, young children, and nonhuman primates attribute knowledge in this latter sense. And, as carefully illustrated by Phillips et al., this does not require attributing belief. Infants successfully predict how others will behave when aware of an object's location (usually, although not always, through visually observing it). However, when an observer's perceptual access to the location of an object has been disrupted, although adults would expect that they still have a true belief, infants now act as if they are ignorant, and have no expectation as to their behavior. Similarly, chimpanzees have demonstrated an understanding of awareness and ignorance, but there is no clear evidence that they are able to represent the beliefs of others, whether true beliefs or false beliefs. Furthermore, there are circumstances in which adults more readily attribute knowledge than belief (e.g., Myers-Schulz & Schwitzgebel, Reference Myers-Schulz and Schwitzgebel2013; Phillips et al., Reference Phillips, Strickland, Dungan, Armary, Knobe and Cushman2018), suggesting that this notion of knowledge persists through development. All of this supports view 2.
However, there is another sense of knowledge that adults are concerned with. Perhaps the most famous quest for knowledge of this sort was that of Descartes, who asked how he could know whether the things he believed to be true were actually true. Descartes had plenty of beliefs that he was very confident about, such that he was sitting in his dressing gown by the fire, holding writing papers in his hands. But he worried that he might be dreaming, or insane, or deceived by a malignant demon. He wanted to know whether these beliefs rose to the level of knowledge, and ultimately concluded that there was one thing he did know: “I knew that I was a substance whose whole essence or nature is simply to think, and which does not require any place or depend on any material thing, in order to exist” (Descartes, Reference Descartes1980/1637, p. 18). Here, Descartes is clearly referring to propositional knowledge, not simply awareness.
Phillips et al. might argue that this notion of knowledge is special to philosophers; knowledge in the “awareness” sense is the only concept that exists in everyday use. But this clearly isn't the case. Ordinary people ask all the time whether someone knows something or just believes it. This might mean asking whether someone is certain of their belief, whether someone's belief is true, or whether they have the right kind of evidence for their belief. In all these cases, however, we are first attributing a belief to someone, and then asking whether this belief rises to the level of knowledge. We might also withhold attribution of knowledge even in cases where there is awareness. Someone in the desert, half mad with thirst, might see a (real) lake in front of him – might be fully aware of it – but worry that it's a hallucination. “I believe there is water over there, but I just don't know for sure!” he might cry out to himself.
Properly distinguishing these two concepts of knowledge reveals two things. First, representations of knowledge in the sense of awareness are more basic than representations of belief. Second, representations of belief are more basic than representations of propositional knowledge. View 2 is correct, but so is view 1.
Conflict of interest
None.