Knowledge Representation: Defining Mental Models (Part III)

Mental Models Series

Packing one or more labels (expressing the same thing) together with a set of characteristics into discrete cognitive units within a context is what defines a concept. Therefore, one or more labels can be associated with a concept, while a label can be associated with more concepts when different contexts apply. 

Concepts are associated with physical objects (doors, keys, latches, knobs), subjects (doorman, locksmith), scenarios and actions (opening a door, opening a door with a key, opening a door without a key, closing a door, breaking a door). As previously discussed, this mix forms together a ‘mental model’, however more precision is needed if we want to delimit the borders of a mental model.

When one or more labels are associated with one or more actions, or a chain of causality exists, then a mental model can be more likely extrapolated (aka build). Minimally we can consider only one object (the door) and only one action (opening the door) with two states (door open, door closed). One can consider the action of opening a door also as an objective, while in addition a basis heuristic is also implied (pushing the door). Thus, to open the door, I push it, and it will open or will remain closed. This is a description of the simplest mechanical model I can build. 

Of course, I can consider more factors: if I don’t apply enough force, the door might not open, so the force applied needs to be higher than the resistance. However, “force” and “resistance” are indirectly considered in our model and for the sake of simplicity can be ignored. Still, I can extend the model with one more heuristic – trying again. I can try again by applying the same force, or by pushing the door with my body’s force (maximal force) if the door didn’t open the first time. In the later case we deal thus with a third heuristic. A fourth heuristic might be to slightly increase the push force. 

When one talks about repeatability of an action a loop is implied with an exist condition - I push the door, it doesn’t open, then I try to apply more force. When just enough force is applied the door opens, and this is the exit criterion from the loop. If after several tries the door doesn’t open, then one has the choice of persisting or of breaking the option – another exit criterion. Typically a few tries are enough, though one might be tempted to try again later, though this behavior might include other drivers and probably the existence of other models. 

One can extend the model by including a door latch and with it adding more heuristics, pushing down/up the latch, if a simple push of the door doesn’t work. With it the number of states increases – latch push down, respectively up. Also here the force applied plays an important role and a loop is implied. The more objects are involved, more complex the model becomes. Of course, one doesn’t need to consider all the heuristics and states, just the more relevant or more probably one. 

On the other side, the model we are trying to build even if reflects to some degree how the mind works, is deeply submerged into our cognitive space to the degree that it become an automatism. However, we can become aware of the entire mechanism as soon we try to describe it into a functional unit, even if the process might not always be that straightforward as we would like to. In addition, the are many elements which we are forced to ignore or simplify. 

Previous Post <<||>> Next Post