We propose conceptuality as a demarcation criterion for knowledge-that in relation to knowledge-how that is able to account for the peculiarities of both knowledge types outlined in the last section. In order to show that conceptuality can do the job, we have, in a first step, to establish a notion of concept that does not presuppose in an obvious way characteristics of knowledge-that, i.e., the notion we look for should not entail that concepts are essentially linguistic entities enabling persons to verbally express knowledge-that. What we then need, in a second step, is a notion that entails some fundamental and (hopefully) non-contentious assumptions about necessary conditions for concept possession in terms of abilities. Characterizing concepts in the form of abilities necessary for concept possession should enable us to show that having those abilities necessary for concept possession is exactly what is needed for the subject to overcome the peculiar limitations accompanying knowledge-how, and thus to gain access to the level of knowledge-that (see section 5).
In shaping the sought-for notion of conceptuality we take recourse to work by Allen & Hauser (1991), Pylyshyn (1990), and Newen & Bartels (2007). Allen and Hauser have claimed that, from the perspective of interpreting the behavior of systems including human and animal organisms as much as artificial systems, the ascription of genuine concepts requires “evidence supporting the presence of a mental representation that is independent of solely perceptual information” (Allen & Hauser 1991, p. 231). The criterion of independence, as called for by these authors, is that it enables the system to show flexible behavior, in contrast to the performance of rigid mechanisms:[9]
[I]ndependence in this sense entails that the responses of the animal to a certain stimulus are not just ‘driven by’ that stimulus, and are also not to be explained as cases of stimulus generalization, i.e., discrimination by a mechanism responsive to a single basic stimulus. (Newen & Bartels 2007, p. 287)
If the reactions of a system to a given stimulus can be modified by the presence of additional stimuli representing the peculiarities of the situation in which the reaction occurs, the system will be first able to generalize—as rudimentary as that ability may be—the information received. It is then that we can legitimately ascribe the possession of concepts: “First, an organism whose internal representations are concept-like should be able to generalize information obtained from a variety of perceptual inputs and use that information in a range of behavioral situations” (Newen & Bartels 2007, p. 287).
We thus arrive at a criterion for conceptuality, which can be called, following Allen (1999), the “transcendence of particular stimuli” or, in terms given by Pylyshyn (1990), the “criterion of informational plasticity”. Essentially the criterion requires the “possibility of the modification of a response in the light of additional information” (Allen 1999); the kind of response has to depend, crucially, on other sources of information (cf. Newen & Bartels 2007, p. 287).
The criterion considered above is still not sharp enough. As long as we do not further specify the “modification of a response” occurring “in the light of additional information”, each sort of extension of the processing capacities of an individual would count as reaching the level of “conceptuality” if only this extension enables the individual to integrate some additional source of information into its behavioral repertoire. Thus diversification of processing capacities could then not be distinguished from transition from non-conceptual to conceptual processing capacities.
To get a criterion for conceptuality it is required that the “modification of a response” mentioned above concerns classificatory behavior. The “additional stimulus”, in that case, not only has to work as a switching point, opening one or the other pathway for a response within a non-conceptual behavioral pattern, it also has to stand for a category, according to which the actual behavioral pattern can be classified.
One example of additional stimuli characteristics standing for a category has been described by Newen & Bartels (2007) with respect to the conceptual abilities of the grey parrot Alex (Pepperberg 1999). In order to be able to form elementary color concepts, for example the concept “green”, Alex should not only be able to generalize over a class of similar stimuli and thus to identify a sample of different green objects, but should additionally be able to represent green as a color. Only then could we ascribe to him the ability to classify green objects according to a well-determined class concept.
The test items by which Pepperberg examined the classification abilities of the animal were, for instance, “What color?” or “What shape?”. These questions should
[…] determine if he [Alex] could respond not only to specific properties or patterns of stimuli [e.g., to green objects], but also to classes or categories to which these specific properties or patterns belong […]. Could he, for example, go beyond recognizing what is, or is not, ‘green’ to recognizing the nature of the relationship between a green pen and a blade of grass? (cf. Pepperberg 1999, p. 52)
It happened that Alex was indeed able to classify the given “key” stimulus, e.g., a green, round object, visually presented to him, as “green” or “round” according to different dimensions (e.g., color or shape) represented by additional auditory stimuli. His choice of response (“green” or “round”) turned out to depend crucially on the “additional information” given in form of the auditory stimulus. As such, Newen and Bartels concluded that “Alex was able to represent different properties while having only one and the same visual input of an object.”[10]
With this example in view, Newen & Bartels (2007) formulated the following requirements for the possession of concepts—for instance, the concept “red”: A cognitive system has the concept “red” only if (i) it has relative stimulus independence such that it depends on some additional mechanism—which detects and weighs stimuli other than the key stimulus of redness—to determine that the system focuses on redness while perceiving a red square, in contrast to some other property; and (ii) the property of being red is represented as an instance of the dimension “color”.[11]
Note that the above-mentioned definition of conceptuality does not only require the existence of some “additional stimulus” to which the individual has to be responsive, but that there has to be some additional internal mechanism of processing by which the individual is able to “detect and weigh” a specific additional stimulus as standing for a category (e.g., “color”). The responsiveness of the individual to that stimulus shows up when it focuses its attention on those aspects of a scene, or to those items of a behavioral pattern, which exemplify the respective category.
Another example would be the balancing of coffee cups by a waiter in a restaurant. Let us assume that the waiter for some time possesses the ability to balance cups of different shapes without spilling coffee, and without consciously attending to a particular cup, or the shape of a particular cup that he is currently dealing with. At some point he is told that there are essentially two different kinds of cups, one high and cylindrical, and the other flat and bowl-shaped (this information is the “additional stimulus”). The waiter “detects and weighs” the additional stimulus by focusing his attention, from that time on, to his own specific handling of cups, depending on the sort of cup a particular exemplar belongs to. He might then detect that he had previously managed to deal with both kinds of cups efficiently and without spilling coffee without even noticing that liquids in both reacted in different ways to his movements. The waiter’s behavior has switched from a former “non-conceptual” dealing with coffee cups to a form of behavior that is “conceptual” in the sense of exhibiting an additional ability, namely the ability to classify his own performance in balancing coffee cups according to a category (in this case the cups’ shape).
How does such a notion of conceptuality relate to Evans’ notion of non-conceptual knowledge in terms of first-person dispositions that we made use of in sections 2 and 4? If the possession of concepts is constituted (in contrast to non-conceptual cognitive processing) by the gaining of additional abilities, it should be made plausible how those additional abilities connect to a non-conceptual basis in Evans’ theory.
In our treatment of his theory, we followed an interpretation of Evans’ work as implying that non-conceptual knowledge relies on the disposition to have one’s own motor reactions be determined by sensory and kinesthetic information that is mediated by either some external object or by one’s own body. Again, the waiter dealing with the coffee cups may help to illustrate the point. The waiter’s experienced handling relies on a disposition to have his motor actions determined by the multimodal sensory information that is mediated by holding coffee cups in his hands. The waiter’s knowledge-how to balance the cup might be completely independent of any conceptual reference to coffee cups. He could be the experienced waiter that he is—at least with respect to his balancing ability—without even knowing in a conceptual way “what a coffee cup is”. Reference to the objects he is dealing with was accomplished only by being able to react in a coordinated way to sensorimotor information originating from handling these objects.
At the time he is told that coffee cups come in two different shapes, his cognitive system enables him to use that information such that he begins to rely on a category (i.e., a cup’s shape) in order to refer to coffee cups, and to classify his own balancing behavior according to the objects thus categorized. He reaches, in some minimal way, the level of conceptual knowledge, since he now begins to identify both, that is, the objects and his behavior with respect to these objects, by conceptual means.