The authors distinguish between two cognitive processes: believing and knowing. But this is not the most basic or productive distinction to be made in the current context. The most basic distinction, in my view, is between two kinds of knowing, indicated in many of the world's languages by two different verbs: in German by kennen and wissen, in French by connaître and savoir, and in Spanish by conocer and saber. The first member of each pair is normally glossed as something like “be acquainted with,” as in “I know (am acquainted with) the restaurant to which you are referring.” The second member of each pair is normally glossed in propositional terms, as in “I know that the restaurant to which you are referring is on 9th Street.” The opposite of knowledge-by-acquaintance is ignorance: “I do not know (am ignorant of) the restaurant to which you are referring.” The opposite of propositional knowledge is error: “The restaurant to which you are referring is not on 9th Street.”
My claim is that this distinction between knowledge-by-acquaintance and propositional knowledge is most basic. So what about believing? Believing is inherently propositional. One does not believe a restaurant; rather, one believes that a particular restaurant is closed. Believing involves an agent representing some state of the world, and either she or some outside observer has some reason to doubt that this representation is accurate – as in all propositional judgments. Thus, you know that the restaurant to which I am referring is on 9th Street, and you believe (i.e., either I or you are not sure that this is the case) that its ZIP Code is 27709. The difference is that I agree with your first judgment (or else you are sure), but I disagree (or you are not sure) about your second judgment. The key point is that believing and knowing in the propositional sense both involve cognitive representations of states of affairs coupled with judgments about the accuracy or inaccuracy of the representation as compared to the real situation (i.e., one can have different “attitudes” to the propositional content). Such comparison between representation and reality is not involved in knowledge-by-acquaintance. Even if we establish that I am acquainted with some state of affairs – I am acquainted with the fact that the restaurant is on 9th Street – this does not involve an attitude or judgment about whether or not my representation of that state of affairs is accurate.
My empirical claim is that what comes first in both phylogeny and ontogeny is an understanding that agents are acquainted with things (mainly by perception). Thus, chimpanzees understand that a competitor sees, and so is acquainted with, the location of a piece of food, and they even understand that a competitor has seen food being hidden and thereby has become acquainted with its location (Hare, Call, Agnetta, & Tomasello, Reference Hare, Call, Agnetta and Tomasello2000; Hare, Call, & Tomasello, Reference Hare, Call and Tomasello2001). (NB: the apes have propositional knowledge, but they do not attribute it to others.) One might choose to gloss this as an understanding that the competitor knows that the food is in a certain location; but this would not fit with the empirical data. Studies specifically designed to distinguish between knowledge-by-acquaintance and propositional knowledge in great apes (e.g., Kaminski, Call, & Tomasello, Reference Kaminski, Call and Tomasello2008) show that when chimpanzees see a competitor witnessing the hiding of food, they understand that she is acquainted with the food's location; when they witness her not witnessing the hiding of food, they attribute to her ignorance. But the key finding is this: When chimpanzees witness a competitor being blocked from witnessing the moving of food from one location to another, they do not attribute to her an incorrect representation of the food's location based on her now outdated witnessing, but only, again, ignorance of the food's location. They can understand that their competitor is acquainted with the food's location, but they do not compare her representation of the situation either to their own or to any “objective” situation, and so there is no question of accuracy or potential error. They attribute to their competitor knowledge-by-acquaintance (or ignorance), not propositional knowledge, belief, or error.
The same analysis applies to looking studies with human infants (and apes). As argued by Tomasello (Reference Tomasello2018), in the classic looking-time studies of infants' understanding of false belief (e.g., Onishi & Baillargeon, Reference Onishi and Baillargeon2005), it may be that infants simply understand that the agent is acquainted with (has registered) the location of the object based on where she last saw it. But, to do this, infants do not need to relate or coordinate their understanding of the agent's representation with any other representation, neither their own nor some objective representation. To be surprised or to predict that the agent is searching for the object somewhere other than where she saw it disappear does not require a judgment of whether her knowledge is accurate or inaccurate; the infant's or ape's own knowledge of the location of the object is irrelevant and not attended to (and the same analysis applies to anticipatory looking studies, such as that of Krupenye, Kano, Call, Hirata, & Tomasello, Reference Krupenye, Kano, Call, Hirata and Tomasello2016).
I thus agree that an understanding that others know things is both phylogenetically and ontogenetically primary, but only if we are talking about knowledge-by-acquaintance involving simple representations. Understanding representations as propositional entails, in addition, an understanding that they either match or mismatch with the objective situation as represented by the one making the judgment. Propositional knowledge and beliefs thus involve a comparison and/or coordination of different representations of one and the same situation, which presupposes both a prior understanding of something more primitive like knowledge-by-acquaintance (based on simple perception and representation) and, in addition, an ability to compare and/or coordinate potentially different representations (at an executive level).
The authors distinguish between two cognitive processes: believing and knowing. But this is not the most basic or productive distinction to be made in the current context. The most basic distinction, in my view, is between two kinds of knowing, indicated in many of the world's languages by two different verbs: in German by kennen and wissen, in French by connaître and savoir, and in Spanish by conocer and saber. The first member of each pair is normally glossed as something like “be acquainted with,” as in “I know (am acquainted with) the restaurant to which you are referring.” The second member of each pair is normally glossed in propositional terms, as in “I know that the restaurant to which you are referring is on 9th Street.” The opposite of knowledge-by-acquaintance is ignorance: “I do not know (am ignorant of) the restaurant to which you are referring.” The opposite of propositional knowledge is error: “The restaurant to which you are referring is not on 9th Street.”
My claim is that this distinction between knowledge-by-acquaintance and propositional knowledge is most basic. So what about believing? Believing is inherently propositional. One does not believe a restaurant; rather, one believes that a particular restaurant is closed. Believing involves an agent representing some state of the world, and either she or some outside observer has some reason to doubt that this representation is accurate – as in all propositional judgments. Thus, you know that the restaurant to which I am referring is on 9th Street, and you believe (i.e., either I or you are not sure that this is the case) that its ZIP Code is 27709. The difference is that I agree with your first judgment (or else you are sure), but I disagree (or you are not sure) about your second judgment. The key point is that believing and knowing in the propositional sense both involve cognitive representations of states of affairs coupled with judgments about the accuracy or inaccuracy of the representation as compared to the real situation (i.e., one can have different “attitudes” to the propositional content). Such comparison between representation and reality is not involved in knowledge-by-acquaintance. Even if we establish that I am acquainted with some state of affairs – I am acquainted with the fact that the restaurant is on 9th Street – this does not involve an attitude or judgment about whether or not my representation of that state of affairs is accurate.
My empirical claim is that what comes first in both phylogeny and ontogeny is an understanding that agents are acquainted with things (mainly by perception). Thus, chimpanzees understand that a competitor sees, and so is acquainted with, the location of a piece of food, and they even understand that a competitor has seen food being hidden and thereby has become acquainted with its location (Hare, Call, Agnetta, & Tomasello, Reference Hare, Call, Agnetta and Tomasello2000; Hare, Call, & Tomasello, Reference Hare, Call and Tomasello2001). (NB: the apes have propositional knowledge, but they do not attribute it to others.) One might choose to gloss this as an understanding that the competitor knows that the food is in a certain location; but this would not fit with the empirical data. Studies specifically designed to distinguish between knowledge-by-acquaintance and propositional knowledge in great apes (e.g., Kaminski, Call, & Tomasello, Reference Kaminski, Call and Tomasello2008) show that when chimpanzees see a competitor witnessing the hiding of food, they understand that she is acquainted with the food's location; when they witness her not witnessing the hiding of food, they attribute to her ignorance. But the key finding is this: When chimpanzees witness a competitor being blocked from witnessing the moving of food from one location to another, they do not attribute to her an incorrect representation of the food's location based on her now outdated witnessing, but only, again, ignorance of the food's location. They can understand that their competitor is acquainted with the food's location, but they do not compare her representation of the situation either to their own or to any “objective” situation, and so there is no question of accuracy or potential error. They attribute to their competitor knowledge-by-acquaintance (or ignorance), not propositional knowledge, belief, or error.
The same analysis applies to looking studies with human infants (and apes). As argued by Tomasello (Reference Tomasello2018), in the classic looking-time studies of infants' understanding of false belief (e.g., Onishi & Baillargeon, Reference Onishi and Baillargeon2005), it may be that infants simply understand that the agent is acquainted with (has registered) the location of the object based on where she last saw it. But, to do this, infants do not need to relate or coordinate their understanding of the agent's representation with any other representation, neither their own nor some objective representation. To be surprised or to predict that the agent is searching for the object somewhere other than where she saw it disappear does not require a judgment of whether her knowledge is accurate or inaccurate; the infant's or ape's own knowledge of the location of the object is irrelevant and not attended to (and the same analysis applies to anticipatory looking studies, such as that of Krupenye, Kano, Call, Hirata, & Tomasello, Reference Krupenye, Kano, Call, Hirata and Tomasello2016).
I thus agree that an understanding that others know things is both phylogenetically and ontogenetically primary, but only if we are talking about knowledge-by-acquaintance involving simple representations. Understanding representations as propositional entails, in addition, an understanding that they either match or mismatch with the objective situation as represented by the one making the judgment. Propositional knowledge and beliefs thus involve a comparison and/or coordination of different representations of one and the same situation, which presupposes both a prior understanding of something more primitive like knowledge-by-acquaintance (based on simple perception and representation) and, in addition, an ability to compare and/or coordinate potentially different representations (at an executive level).
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflict of interest
None.