The figure below represents the five groups of subsystems that are involved in encoding new memories.
The perceptual encoding subsystems are necessary for memory to occur. Perception determines the information that will become a memory as well as the time and place that will be tied to a memory. Perceptual subsystems are also important in that they are used to encode language, which enhances encoding of memories and defines the context for remembering and recalling stimuli. Perceptual encoding subsystems also are important in the storage of unimodal perceptual information, or information from one sense (auditory, visual, etc.).
Associative memory holds representations of information that arrives from a perceptual modality. This information is descriptive in nature; defining categories, abstract relations, structural information about objects, and other nonperceptual types of information. Associative memory thus gives detailed organization to stored information, determining the importance of information that is stored. Input from any perceptual modality can activate associative memory.
Memory formation subsystems work to store representations and associations between representations of information. Memory formation subsystems coordinate the sending of information to be stored in perceptual encoding subsystems and associative memory. Memory formation subsystems also allow associations to be formed of information stored in both of these areas. The memory formation subsystems rely on many brain structures to accomplish its tasks, particulary the hippocampus, limbic thalamus, and basal forebrain. Please see the Brain Structures page for further information.
This subsystem stores direct connections between a representation for a stimulus and a representation for a response. Studies have shown that the striatum (made up of the caudate nucleus and putamen) plays a role in storage of this type of memory. The striatum is important for skill acquisition memory. It receives perceptual information, and indirectly sends information to the premotor and supplementary motor cortex.
The decision subsystem is activated when specific, important information is stored. The decision subsystem received information from associative memory and sends commands to the memory formation subsystems to store memories and associations between memories. It thus initiates, coordinates, and oversees the storage of certain individual information.
The first step in storing new information is to determine if it is new. New information will not activate any similar stored representation. If there is no similar representation that matches the entire new input, the information is new. After information is determined to be new, two subsystems become involved in storing the new information. These two subsystems are always active and operate on all new input, automatically storing information of stimuli that is paid attention to.
This subsystem is part of the basal forebrain. When the memory formation subsystems determine that information is new, the print-now subsystem is activated. The print-now subsystem is aware of the activity that has been induced by the new stimuli in the other subsystems (perceptual encoding, associative memory, etc.) and prepares for later structural changes that will occur in connections at the neuron level so that the new stimuli will be encoded and remembered. However, this subsystem does not make these structural changes.
Weight Adjustment Subsystem
The weight adjustment subsystem changes the importance or weight of the neuron connections responsible for holding the new stimulus. It is initiated by the print-now subsystem. Thus this subsystem is responsible for storing the properties and important information of the new stimulus. This subsystem depends on processes in the hippocampus and cortex. The hippocampus oversees the changing connections of neurons- it is a "catalyst" for the neural network to change its connections in a specific way. The strength and ability of the connections overseen in this step determines the permanence of the new memory stored. This process is thus very complex and takes a long time to finish.
This process involves storing the association of familiar information with a new context. This process is seen when we "memorize" items on a shopping list (we already know the items, but we must remember them in this context). Many subsystems are involved in remembering these associations. The print-now and weight adjustment subsystems first identify the information as familiar and then associate it with previously stored, similar stimuli. This association is then stored using something like a "place holder" in the associative memory that points to the important stored representations. These representations of information are actually stored in the perceptual encoding subsystems. Later experience with the stimuli activates the place holders in associative memory, identifying the context that the stimuli was associated with.
To encode and remember examples and prototypes, abstract properties of objects must be gathered and generalizations must be drawn. Our memory subsystems are organized so that this process is automatically done. Categories or prototypes that are formed and represented in memory do not have strict boundaries, but consist of examples that are organized based on those that are most typical of the category. Thus categories consist of the basic properties of a pattern that are given weights of importance in the neural network. Incoming examples are generalized to the categories, and some are determined to fit better than others. Prototypes are formed by generalizing an average of individual examples. Thus, individual examples can be stored and remembered, and general prototypes can also be formed and remembered. Examples and prototypes are stored in seperate networks in the brain. Studies have shown that the left cerebral hemisphere plays an important role in the storage of prototypes and the right cerebral hemisphere plays an important role in the storage of examples.
Often, we intentionally pay attention to and seek to memorize and store information. This type of memory is called intentional memory. This type of memory is often more successful than automatic, incidental memory for many reasons.
1. When we intentionally pay attention to information, we retain this pattern of information in the activation subsystems and thus increase the success of storage by the weight adjustment subsystem. By intentionally paying attention to information, we also continue the signal from the print-now subsystem to the weight adjustment subsystem. Thus the information is attended to more intensively and for longer periods of time, making it more likely that the information will be successfully stored.
2. The more information and details sent of a stimulus we are trying to remember, the more likely the stimulus will be stored. The more information sent, the more opportunities there are for associations and encoding to take place. This is called the depth of processing effect.
3. The decision subsystem will take the information that one is trying to store and organize it into chunks. This chunking of information makes it easier to remember things because more time can be spent on less items.
4. In intentionally seeking to store information, retrieval cues are often used to make storage easier and more successful.
5. Intentional storage of information can make use of visual as well as verbal methods of storage. By using visual and verbal methods to store and remember information, the chances of remembering information are increased.