3. Template theory is explained by a match of sensory information to an internal construct of the object that is held in long-term memory. A metaphor for this theory is the “lock-and-key mechanism” because there must be an exact match between sensory information and the mind’s construct of that information. There are five main problems with template theory. One is that it doesn’t work for degraded stimuli; we can still recognize objects when part of the info is missing. Another problem arises with the “straw man” argument; we can also recognize different objects that belong to the same category. Also, template theory is simply not parsimonious. We would need an infinite number of templates for every variation of anything we ever perceived. Template theory also does not allow for novel stimuli. We sometimes see things we have never seen before, but we can still identify the object. The last problem, which seems to sum up all the other problems, is that template theory is simply too restrictive for it to pertain to our whole capabilities of perception.

4. Feature analysis proposes that whole bits of sensory information are matched to its individual small units that are held in our long-term memory. For example, take the letter “A”. According the feature analysis, the letter A is first converted into sensory input. Next comes the “feature analysis” stage, where feature units are matched to parts of the incoming stimulus. (The features matched in this case are two different diagonal lines and a crossbar.) Those parts are then “activated” to make the feature more accessible. Aciviation from the feature level unties then feeds up to the “letter analysis” stage, where the letter unit most activated (in this case, the letter “A”) is then recognized. The feature analysis theory does have problems, however; one is that it doesn’t work with degraded stimuli. (Recall the WORK/WORD example from class). Also, feature analysis is not parsimonious. In reality, familiar objects are recognized more quickly than non-familiar objects. If feature analysis occurred with every object recognition process, we’d expect the recognition time for familiar objects to equal that of non-familiar objects. Also, feature analysis has a tough time with different objects that have the same features. (Ie, in THE CAT, we can tell that there’s a letter H in the first word and a letter A in the second word, but feature analysis would not be able to.) Feature analysis is an example of bottom-up processing because it focuses on the importance of the stimulus (the letter A) and information does not flow down from concept to sensory input. It therefore can also not explain context effects – the point that the context in which a letter is found influences how it is perceived.