Another aspect of touch that has gone largely unnoticed (at least among philosophers) is the fact that spatial information plays a less important role in securing object reference and recognition in touch than in vision. Touch is continually contrasted with vision, and the focus is largely on the fact that both senses involve spatial representation (considerable focus has been on the representation of shape in both senses, for instance). It is easy to assume that touch and vision make similar use of spatial information in securing reference and especially recognition. This is not the case. Spatial information plays a smaller role in tactual object-recognition than it does in vision. Most important in touch are so-called intensive features — things like material composition, texture, weight, temperature, and so on. This may translate over to object reference as well. We seem to secure reference to the keys in our pocket through intensive features like metallic and cold and small than we do from the specific shapes of the individual keys. This is actually a surprising fact, but one which was discovered by careful empirical investigation (Klatzky et al. 1993, Lederman and Klatzky 1997). Restricting the availability of so-called intensive cues causes our otherwise excellent haptic recognitional capacities to suffer greatly. For instance, Lederman and Klatzky (2003) found that subjects given a range of tangible stimuli shaped like ordinary objects, but made out of the same uniform material, find object-identification tasks more difficult than when they have access to the material composition and heft of the object but not its overall shape. Other studies have shown that non-spatial surface properties are available faster in haptic processing than are spatial features (Lederman and Klatzky 1997). This means that in a typical tactual experience we can be aware of what something is — we can identify the object or a range of its most salient features — without knowing where exactly the thing is located or its spatial characteristics. This fact helps explain how one can secure tangible reference to an object through a probe or intermediary even when that intermediary supplies only limited spatial information about the object (that is, without specifying its exact shape and location). The reference can be secured via sparse spatial information so long as there is enough additional information about an object’s other salient features (solidity, roughness, weight, etc). The coordinates of peripersonal space are sparse in exactly this sense: they do not specify exact locations or precise geometric relations. Instead they label one’s surroundings with limited information about the immediate environment in action-guiding terms. These in turn allow a subject to engage with a distal object via some appropriate intermediary and thereby discover its diagnostic, intensive features.
