To show that representations of knowledge are more basic than representations of belief, Phillips et al. draw on evidence from various branches of psychology, cognitive science, and experimental philosophy. My aim is to add support from a very different source: the history of philosophy. For it turns out that Western philosophy – according, at least, to its first major historian, Aristotle – developed theories of knowledge long before it developed theories of belief.
My evidence is drawn from Aristotle's criticism of his Presocratic predecessors' theories of cognition, in his main psychological treatise. I will show that in this discussion (De Anima III.3), Aristotle:
(1) argues that the Presocratics had a theory of knowledge;
(2) argues that they had no theory of false belief; and finally,
(3) introduces as a philosophical innovation a genus of which knowledge and false belief are both species: belief.
First, Aristotle's Claim 1: the Presocratics had a theory of knowledge.
According to Aristotle, his Presocratic predecessors held that one of the defining features of soul (psuchê) is the capacity for what he calls gnôsis (De Anima 404b9, 404b27-28). The word is sometimes translated as “knowledge,” and sometimes as “cognition.” Here, the ambiguity is significant. For Aristotle's criticisms amount to the thesis that the Presocratics attempted to give a general account of cognitive activity, but failed precisely because they construed all cognition as knowledge.
On Aristotle's own view there are two broad species of cognition: perception and thought (to noein, to dianoeisthai). He accuses the Presocratics of conflating the two. Perception is their model for all cognition. Moreover, they construe perception as physical contact between the mind and worldly objects in which the mind comes to resemble the objects. Therefore, on their view, all cognition is true (DA 427a21-b3). (For a quick reconstruction of the argument see below; for a detailed reconstruction see Lee, Reference Lee2005.)
Thus, Aristotle construes the gnôsis of the Presocratics as truth-ensuring contact with reality.
I submit that this clearly counts as a theory of what we would call knowledge (along the lines of Williamson's “most general factive mental state” (Williamson, Reference Williamson2000)). Gnôsis on this account is factive, and, because it involves direct contact and special fit between mind and object, it is more than just true belief. (It also fits Phillips' et al.'s further criteria: it is multi-modal, and allows for representations of egocentric ignorance.) Aristotle is attributing to his predecessors a theory of knowledge.
Second, Aristotle's Claim 2: the Presocratics had no theory of false belief.
Precisely because they construed cognition as they do, Aristotle goes on to argue, his predecessors cannot make sense of cognitive error (427a28-b6). His claim seems to be: If thought is a matter of the mind being made to resemble its object, then if you are thinking at all, you are thinking veridically. Indeed, some of the Presocratics simply deny that false belief exists, arguing that “everything that appears is true” (427b3; cf. Metaphysics IV.5, which explicitly equates this slogan with the relativist claim that all opinions are true).
Aristotle is aware that many Presocratics believed in cognitive error. His claim is that they failed to offer a theory of it, or even a theory on which it is possible. In constructing their epistemologies they developed accounts of knowledge, and got stuck there. The clear implication is that it takes a more sophisticated philosopher to develop a theory of false belief. (Compare Plato's criticism of Parmenides in the Sophist.)
Finally, Aristotle's Claim 3: the Presocratics had no theory of belief in general.
To account for cognitive error as well as knowledge, Aristotle thinks, we need to recognize a broader category to which both belong. This is precisely what he does in the next part of the discussion, using new technical vocabulary to introduce a new concept.
Thinking, he argues, is composed of two components: phantasia (quasi-perceptual appearance), and hupolêpsis. Hupolêpsis is a genus with several species, some factive and some anti-factive: theoretical knowledge (epistêmê), practical knowledge (phronêsis), true opinion (true doxa), and “the opposites of these” – that is, their false counterparts (427b79-11 and 24–26). Although he does not define hupolêpsis, he argues that it presupposes conviction, and suggests that it consists of taking something to be true or false (428a20-428b4). In other words – as many have recognized, and as I argue in detail elsewhere (Moss & Schwab, Reference Moss and Schwab2019) – hupolêpsis is what modern epistemology calls belief. It is generic taking-to-be-true, which can be true or false, and which when the right conditions are fulfilled constitutes knowledge.
Aristotle does not explicitly accuse the Presocratics of lacking a theory of belief. But he does take their inability to account for cognitive error to show the need for a new theory of thought, one which crucially includes a component so theoretically novel that it requires a neologism (“hupolêpsis”). The implication is that his predecessors lacked a theory of belief, and that he is the first to develop one.
Thus, according to Aristotle, in the development of Western philosophy theories of knowledge preceded theories of belief.
I leave to another occasion the question of whether Aristotle is right. A very brief defense: Plato argues that accounting for false belief is a difficult task, and only late and tentatively offers anything like an account of generic belief. (See Moss & Schwab, Reference Moss and Schwab2019; for assessment of Aristotle's treatment of Presocratic epistemology, see Lee, Reference Lee2005.)
At any rate, if Aristotle is right, then – granted the plausible assumption that we more easily theorize concepts that are more basic – his account offers further support for Phillips et al.'s contention. For evidently, it comes more easily to humans to construct a philosophical theory of knowledge than one of belief.
You have
Access
To show that representations of knowledge are more basic than representations of belief, Phillips et al. draw on evidence from various branches of psychology, cognitive science, and experimental philosophy. My aim is to add support from a very different source: the history of philosophy. For it turns out that Western philosophy – according, at least, to its first major historian, Aristotle – developed theories of knowledge long before it developed theories of belief.
My evidence is drawn from Aristotle's criticism of his Presocratic predecessors' theories of cognition, in his main psychological treatise. I will show that in this discussion (De Anima III.3), Aristotle:
(1) argues that the Presocratics had a theory of knowledge;
(2) argues that they had no theory of false belief; and finally,
(3) introduces as a philosophical innovation a genus of which knowledge and false belief are both species: belief.
First, Aristotle's Claim 1: the Presocratics had a theory of knowledge.
According to Aristotle, his Presocratic predecessors held that one of the defining features of soul (psuchê) is the capacity for what he calls gnôsis (De Anima 404b9, 404b27-28). The word is sometimes translated as “knowledge,” and sometimes as “cognition.” Here, the ambiguity is significant. For Aristotle's criticisms amount to the thesis that the Presocratics attempted to give a general account of cognitive activity, but failed precisely because they construed all cognition as knowledge.
On Aristotle's own view there are two broad species of cognition: perception and thought (to noein, to dianoeisthai). He accuses the Presocratics of conflating the two. Perception is their model for all cognition. Moreover, they construe perception as physical contact between the mind and worldly objects in which the mind comes to resemble the objects. Therefore, on their view, all cognition is true (DA 427a21-b3). (For a quick reconstruction of the argument see below; for a detailed reconstruction see Lee, Reference Lee2005.)
Thus, Aristotle construes the gnôsis of the Presocratics as truth-ensuring contact with reality.
I submit that this clearly counts as a theory of what we would call knowledge (along the lines of Williamson's “most general factive mental state” (Williamson, Reference Williamson2000)). Gnôsis on this account is factive, and, because it involves direct contact and special fit between mind and object, it is more than just true belief. (It also fits Phillips' et al.'s further criteria: it is multi-modal, and allows for representations of egocentric ignorance.) Aristotle is attributing to his predecessors a theory of knowledge.
Second, Aristotle's Claim 2: the Presocratics had no theory of false belief.
Precisely because they construed cognition as they do, Aristotle goes on to argue, his predecessors cannot make sense of cognitive error (427a28-b6). His claim seems to be: If thought is a matter of the mind being made to resemble its object, then if you are thinking at all, you are thinking veridically. Indeed, some of the Presocratics simply deny that false belief exists, arguing that “everything that appears is true” (427b3; cf. Metaphysics IV.5, which explicitly equates this slogan with the relativist claim that all opinions are true).
Aristotle is aware that many Presocratics believed in cognitive error. His claim is that they failed to offer a theory of it, or even a theory on which it is possible. In constructing their epistemologies they developed accounts of knowledge, and got stuck there. The clear implication is that it takes a more sophisticated philosopher to develop a theory of false belief. (Compare Plato's criticism of Parmenides in the Sophist.)
Finally, Aristotle's Claim 3: the Presocratics had no theory of belief in general.
To account for cognitive error as well as knowledge, Aristotle thinks, we need to recognize a broader category to which both belong. This is precisely what he does in the next part of the discussion, using new technical vocabulary to introduce a new concept.
Thinking, he argues, is composed of two components: phantasia (quasi-perceptual appearance), and hupolêpsis. Hupolêpsis is a genus with several species, some factive and some anti-factive: theoretical knowledge (epistêmê), practical knowledge (phronêsis), true opinion (true doxa), and “the opposites of these” – that is, their false counterparts (427b79-11 and 24–26). Although he does not define hupolêpsis, he argues that it presupposes conviction, and suggests that it consists of taking something to be true or false (428a20-428b4). In other words – as many have recognized, and as I argue in detail elsewhere (Moss & Schwab, Reference Moss and Schwab2019) – hupolêpsis is what modern epistemology calls belief. It is generic taking-to-be-true, which can be true or false, and which when the right conditions are fulfilled constitutes knowledge.
Aristotle does not explicitly accuse the Presocratics of lacking a theory of belief. But he does take their inability to account for cognitive error to show the need for a new theory of thought, one which crucially includes a component so theoretically novel that it requires a neologism (“hupolêpsis”). The implication is that his predecessors lacked a theory of belief, and that he is the first to develop one.
Thus, according to Aristotle, in the development of Western philosophy theories of knowledge preceded theories of belief.
I leave to another occasion the question of whether Aristotle is right. A very brief defense: Plato argues that accounting for false belief is a difficult task, and only late and tentatively offers anything like an account of generic belief. (See Moss & Schwab, Reference Moss and Schwab2019; for assessment of Aristotle's treatment of Presocratic epistemology, see Lee, Reference Lee2005.)
At any rate, if Aristotle is right, then – granted the plausible assumption that we more easily theorize concepts that are more basic – his account offers further support for Phillips et al.'s contention. For evidently, it comes more easily to humans to construct a philosophical theory of knowledge than one of belief.
Conflict of interest
None.