One of the most important components of core cognition is the concept object, which is understood to apply to entities that are three-dimensional, bounded, and coherent, and that persist through time and have trajectories that are spatiotemporally continuous.
Carey maintains that core cognition is part of our biological heritage (Carey Reference Carey2009). Accordingly, because the concept of an object belongs to core cognition, it is important that she be able to show that infants possess it, and that they acquire it by maturation rather than learning. As she recognizes, this brings her into opposition to W.V. Quine, who maintained (Quine Reference Quine1960) that all concepts that refer to specific objects, or to specific types of object, are associated with explicitly represented principles of individuation, and are acquired either during or after the acquisition of a language. More specifically, Quine held that if one is to represent an entity that is governed by determinate principles of individuation, one must have the conceptual sophistication to be able to express those principles, and one must therefore have acquired a logical apparatus consisting of quantifiers, an identity predicate, numerals, and devices for forming plurals. Carey rejects these claims, at least insofar as they are concerned with concepts that belong to core cognition.
To appreciate the rationale for Quine's position, we need to consider his gavagai argument. Suppose that a field linguist has found that uses of the term gavagai are highly correlated with situations containing rabbits. Can the linguist conclude that the term refers to rabbits? According to Quine, the answer is “no.” For all the linguist has thus far determined, it could be true that gavagai refers instead to momentary stages in the lives of rabbits, or to undetached parts of rabbits. In general, data concerning statistical correlations between uses of words and situations in the world leave questions about reference open, as do facts involving such phenomena as pointing gestures and the direction of gaze. In order to assign a unique reference to gavagai, Quine maintains, the linguist will have to formulate questions like “Is the gavagai that I'm pointing to now identical with the gavagai that I pointed to a moment ago?” and “Is there exactly one gavagai here or more than one?” The answers to such questions will make it possible for the linguist to settle questions that are otherwise empirically undecidable.
It is easy to reformulate Quine's argument so that it poses a direct challenge to Carey's view. Instead of asking about the reference of gavagai, we can ask whether the core concept object refers to three-dimensional entities that are bounded, coherent, persisting, and that move on continuous trajectories, or it refers instead to stages of such entities, or to undetached parts thereof. It is clear that Carey must provide considerations that favor the first possibility in order to secure her claim about the reference of object. Can she do this?
Carey presents considerable evidence that infants possess perceptual input analyzers that are able to detect three dimensionality, boundedness, coherence, and so on, and she argues persuasively that infants' use of the core concept object is governed by the influence of such devices. Thus, for example, she reviews Baillargeon's experiment (Baillargeon et al. Reference Baillargeon, Spelke and Wasserman1985) concerning the responses of 5-month-old infants to impossible scenarios in which planes rotate backwards into areas that have been seen to contain solid objects. The fact that Baillargeon's subjects registered surprise is naturally interpreted as indicating that they expected solid objects to persist through time. Doesn't this show that object stands for objects rather than temporal stages of objects?
Quine would answer “no,” maintaining that there is an alternative interpretation of Baillargeon's experiment. Instead of supposing that Baillargeon's subjects expected that an object first seen at time T would continue to exist, we can suppose that they expected that an object-stage seen at T would be followed by a number of other stages, each linked seamlessly with its predecessor. They registered surprise because the initial stage seemed to be the final stage or a nearly final stage. Equally, Quine would offer reinterpretations of the experiments designed to establish that infants can register and make use of information about spatial boundaries. Instead of showing that infants can discriminate between objects that have separate boundaries, he would say, the experiments can be understood to show that infants can discriminate between parts of objects when they fall within separate boundaries.
As far as I can see, there is nothing in Carey's empirical data, considered in isolation from theoretical considerations, that precludes Quinean reinterpretations of this sort. I believe, however, that it is possible to defend her position with a more or less a priori argument. This argument has two components. (There is a related line of thought in Wright [Reference Wright, Hale and Wright1997].)
The first component is the observation that Quine is committed to accepting Carey's claim that infants represent whole, enduring objects. This is because his alternative hypotheses actually presuppose the truth of that claim. One of his alternative hypotheses is that infants represent temporal stages of objects. In order to represent an item as a stage of an object, a subject must do three things: represent an enduring object, represent the item, and represent the latter as a stage of the former. A similar point applies to Quine's other alternative hypothesis. In order to represent an item as an undetached part of an object, a subject must represent a whole object, represent the item, and represent the latter as a part of the former. In other words, because Quine's alternative hypotheses are concerned with undetached stages and undetached parts, they imply that it is necessary to represent enduring objects in order to represent stages, and that it is necessary to represent whole objects in order to represent parts.
This brings us to the second component of the answer. In view of the foregoing considerations, Quine must change his alternative hypotheses substantially. Thus, for example, instead of the hypothesis that infants represent stages rather than enduring objects, he would have to propose the more complex hypothesis that infants represent stages in addition to enduring objects, and that they represent the former in virtue of representing the latter. But it is clear that hypotheses of this sort are ruled out by considerations of simplicity. On any reasonable theory of representation, it takes time and energy to form a representation. Accordingly, it is appropriate to dismiss claims about multiple simultaneous representations unless they are warranted by positive evidence. As far as I know, there are no experiments in the literature that provide support for baroque Quinean hypotheses of the sort we are presently considering.
You have
Access
One of the most important components of core cognition is the concept object, which is understood to apply to entities that are three-dimensional, bounded, and coherent, and that persist through time and have trajectories that are spatiotemporally continuous.
Carey maintains that core cognition is part of our biological heritage (Carey Reference Carey2009). Accordingly, because the concept of an object belongs to core cognition, it is important that she be able to show that infants possess it, and that they acquire it by maturation rather than learning. As she recognizes, this brings her into opposition to W.V. Quine, who maintained (Quine Reference Quine1960) that all concepts that refer to specific objects, or to specific types of object, are associated with explicitly represented principles of individuation, and are acquired either during or after the acquisition of a language. More specifically, Quine held that if one is to represent an entity that is governed by determinate principles of individuation, one must have the conceptual sophistication to be able to express those principles, and one must therefore have acquired a logical apparatus consisting of quantifiers, an identity predicate, numerals, and devices for forming plurals. Carey rejects these claims, at least insofar as they are concerned with concepts that belong to core cognition.
To appreciate the rationale for Quine's position, we need to consider his gavagai argument. Suppose that a field linguist has found that uses of the term gavagai are highly correlated with situations containing rabbits. Can the linguist conclude that the term refers to rabbits? According to Quine, the answer is “no.” For all the linguist has thus far determined, it could be true that gavagai refers instead to momentary stages in the lives of rabbits, or to undetached parts of rabbits. In general, data concerning statistical correlations between uses of words and situations in the world leave questions about reference open, as do facts involving such phenomena as pointing gestures and the direction of gaze. In order to assign a unique reference to gavagai, Quine maintains, the linguist will have to formulate questions like “Is the gavagai that I'm pointing to now identical with the gavagai that I pointed to a moment ago?” and “Is there exactly one gavagai here or more than one?” The answers to such questions will make it possible for the linguist to settle questions that are otherwise empirically undecidable.
It is easy to reformulate Quine's argument so that it poses a direct challenge to Carey's view. Instead of asking about the reference of gavagai, we can ask whether the core concept object refers to three-dimensional entities that are bounded, coherent, persisting, and that move on continuous trajectories, or it refers instead to stages of such entities, or to undetached parts thereof. It is clear that Carey must provide considerations that favor the first possibility in order to secure her claim about the reference of object. Can she do this?
Carey presents considerable evidence that infants possess perceptual input analyzers that are able to detect three dimensionality, boundedness, coherence, and so on, and she argues persuasively that infants' use of the core concept object is governed by the influence of such devices. Thus, for example, she reviews Baillargeon's experiment (Baillargeon et al. Reference Baillargeon, Spelke and Wasserman1985) concerning the responses of 5-month-old infants to impossible scenarios in which planes rotate backwards into areas that have been seen to contain solid objects. The fact that Baillargeon's subjects registered surprise is naturally interpreted as indicating that they expected solid objects to persist through time. Doesn't this show that object stands for objects rather than temporal stages of objects?
Quine would answer “no,” maintaining that there is an alternative interpretation of Baillargeon's experiment. Instead of supposing that Baillargeon's subjects expected that an object first seen at time T would continue to exist, we can suppose that they expected that an object-stage seen at T would be followed by a number of other stages, each linked seamlessly with its predecessor. They registered surprise because the initial stage seemed to be the final stage or a nearly final stage. Equally, Quine would offer reinterpretations of the experiments designed to establish that infants can register and make use of information about spatial boundaries. Instead of showing that infants can discriminate between objects that have separate boundaries, he would say, the experiments can be understood to show that infants can discriminate between parts of objects when they fall within separate boundaries.
As far as I can see, there is nothing in Carey's empirical data, considered in isolation from theoretical considerations, that precludes Quinean reinterpretations of this sort. I believe, however, that it is possible to defend her position with a more or less a priori argument. This argument has two components. (There is a related line of thought in Wright [Reference Wright, Hale and Wright1997].)
The first component is the observation that Quine is committed to accepting Carey's claim that infants represent whole, enduring objects. This is because his alternative hypotheses actually presuppose the truth of that claim. One of his alternative hypotheses is that infants represent temporal stages of objects. In order to represent an item as a stage of an object, a subject must do three things: represent an enduring object, represent the item, and represent the latter as a stage of the former. A similar point applies to Quine's other alternative hypothesis. In order to represent an item as an undetached part of an object, a subject must represent a whole object, represent the item, and represent the latter as a part of the former. In other words, because Quine's alternative hypotheses are concerned with undetached stages and undetached parts, they imply that it is necessary to represent enduring objects in order to represent stages, and that it is necessary to represent whole objects in order to represent parts.
This brings us to the second component of the answer. In view of the foregoing considerations, Quine must change his alternative hypotheses substantially. Thus, for example, instead of the hypothesis that infants represent stages rather than enduring objects, he would have to propose the more complex hypothesis that infants represent stages in addition to enduring objects, and that they represent the former in virtue of representing the latter. But it is clear that hypotheses of this sort are ruled out by considerations of simplicity. On any reasonable theory of representation, it takes time and energy to form a representation. Accordingly, it is appropriate to dismiss claims about multiple simultaneous representations unless they are warranted by positive evidence. As far as I know, there are no experiments in the literature that provide support for baroque Quinean hypotheses of the sort we are presently considering.