According to Lakoff and Johnson, all conceptions are either basic bodily concepts or metaphorical extensions of them. Concepts of bodily experience FORCE, RESISTANCE, PATH, PROPULSION, BALANCE come first, then all others are constructed by borrowing materials from this stock of basic bodily concepts.
A romantic relationship is like a journey. It can go well or get off course. People can walk through life together or go their separate ways. And so on. Some people think of time as a kind of line that we move along, with basic features that mirror the structure of space. If point a is closer to me in space than is point b, it takes less time to get to a than b. This suggests modeling the entire domain of time on the (source) domain of space and movement through it.
What is the relation between the basic bodily concepts and all the rest? Is it literal construction or composition? Is this a plausible general story about all concepts? What about concepts of palm trees, justice, and universities? How are L&J thinking about concepts? As conceptions? Does it matter?
Assume that L&J have something like the shaping of conceptions in mind. What do conceptions do? Guide inferences and processing more generally. The computational approach seems designed to model inferences and processing, so it seems to offer a natural way to model L&J s view of cognition.
Computationalism in cognitive science: all intelligent behavior can be generated by a computer (e.g., universal Turing machine) and only by a computer. Embodied view of cognition: Cognition can be understood only by investigating the particular bodily processes or states of the subject.
Computa(onalism and the embodied view don t appear to be at odds with each other. In fact, computa(onalism would seem to predict the need to look at the physical human in order to model human cogni(on. Many different algorithms can compute the same func(on, so to find out how a human computes a given func(on, we ll have to get addi(onal evidence. Behavioral studies can help, but neural and bodily evidence should help as well.
The distinctive contribution of the embodied view might be methodological. It tells us where to look when investigating cognition: Look at the role played by bodily or sensorimotor experience or at neural representations of bodily states or processes. Is this at odds with computationalism? Not in theory, but perhaps in historical practice. Recall Turing s remark about how to discover the algorithms used by humans to solve problems: just ask them, he says.
These are neurons in premotor cortex that respond selectively to transitive actions (when there is an object acted upon), whether the subject is observing someone else engage in that kind of action or is engaging in the action him or herself. Some embodied theorists take these to show that our very understanding of others is sympathetic. We understand others by acting out what they re doing in our own motor system.
They seem to show that at least some neurons in premotor cortex (PMv) are sensitive to the kind of action in general, rather than the specific way in which it s performed (which particular effector/ appendage is used to perform it). This seems to be at odds with the idea of a simulation- based representation. (They also seem to show that the brain represents (in STS) action- types as performed by specific effectors.)
Perhaps mirror neurons are the physical implementations of certain abstract concepts, for example, the concept of reaching- for. All parties to the debate agree that computational symbols must take some physical form. Whatever physical form they take, it s only to be expected that, whenever the subject activates the relevant concept, that physical form will be active, regardless of whether the concept gets activated because the subject is observing others or planning his or her own action.