User:Graeme E. Smith/Collections/Model Series/Datamining/Declarative Memory

Declarative Memory

‌One way of looking at Declarative Memory, is that it is memory that you have told yourself that you have. Another way of looking at it, is if you do not have a topical listing of what you know, it is going to be very hard to find anything that you do know. To understand this, let us look at the concept of a Database. A Database is a program that stores and retrieves data. One of the first things you have to do, to store data in a database, is to link each record to a key of some type.
‌That key, is then used to develop an index to the data in the database. By having a key to each record, you can put in different sized records, and still, be able to find every record in the memory separately. Without that key, the record data, if it is not set up in equal sized records, does not have any way of being isolated from the bulk of the data.((indent|8}}To make the data really accessible you need multiple keys, and to be able to retrieve the data according to any one of the keys, and organize it into a record that makes sense. Some databases allow you to designate fields for the data, which allows you to then use the field as a key to extract the data with. To make this database efficient you need to have multiple indexes for the same data, one for each field, so you can use the individual index to extract data in a particular order.

Definition of a Meta-Index

‌Such an list of fields, must also have an index, and so we end up with an index to a list of indexes each to the same list of data records. When we apply an idea to itself we use the word Meta to describe the Top level of the application, so an index to a list of indexes, is a Meta-index.
‌Given a Meta-index that references indexes to each memory, or group of memories, your memories could be randomly spread throughout your storage device, and you could still find them. This is important, because unlike a computer where everything is put in a specific spot or location and all you need is the location, implicit memory is phenomenal, meaning that you can't predict exactly where each memory will fall within memory. Once found it is usually fairly stable, but there is evidence that suggests that sometimes the location changes without notice.
‌This means that the clump that designates a specific memory element may not be completely stable. The fact that many people report Tip of the Tongue type effects where familiar words are suddenly no longer accessible even though we know that we know them, suggests that even familiar memories can shift out of the focus of explicit memory, and therefore that the index entries can be wrong. In this case, we have a dislocation between the Feeling of Knowing that defines Declarative memory and the end-linkage from the index to the memory.

Feeling of Knowing

‌Essentially, what a feeling of knowing amounts to is a recognition that a particular topic has an entry in the meta-index complex, and that therefore it can be found if the entry is correctly indexed. There is no evidence that the accuracy of the Feeling of Knowing is greater than might be found by searching an index to an unstable location, in fact the number of false positives in experiments done on the feeling of knowing suggests that the search is more general than it is exact.

Neural Correlates of a Meta-Index

‌Evidence seems to suggest that if a Meta-Index exists to the Cerebral Cortex, it is stored in the Hippocampal area and controlled by the Subiculum. Certainly centers specializing in Episodal and Emotive Memory have been found in the Hippocampus, mapping coordinate systems seem to be generated by portions of the Entorhinal Cortex, and the Subiculum seems to service a number of other sites in the general area including the Para-hippocampus. It is thought that part of the Entorhinal Cortex might be the output of the Meta-index, and that the output, would be some symbolic representation of the CLUMP needed to find the data in the Cerebral Cortex.
‌To understand how such a symbolic representation is possible we simply need to look at the connections from all the memory systems in the brain, to the Striate Cortex. These connections seem to be multiplex and to link different types of clumps together into a database that allows us to select a particular clump in much the same way that the Meta-Index allows us to find a specific memory by topic. All we need is the ability to access the index by entry place-code, and we can essentially symbolize every explicit memory that we hold in clump storage. Essentially we have a uniform sized data element that defines all existing clumps whether in the main memory loop, or the Meta-Index loop. This has the neat effect of allowing us to reference Meta-index elements in the Meta-Index. The interesting factor about this ability to symbolize anything with a uniform code, is that we can encode the symbols into a map, and trigger each of them with the firing of a single neuron or group of neurons. This makes mapping location in episodal memory much easier, since we need to map the location not the size of the element stored, and we only need a single neuron or small groups of neurons to symbolize a much larger storage element.
‌Sure, our index becomes a map, instead of a list, but what is a map, but a way of listing things in two dimensions? or even, as may be seen in the episodal memory in 3 dimensions plus time?
‌The idea of a Meta-Index, allows our memory to retrieve the symbolic representation of the Clump index entry and thus to find the clump that addresses the implicit memory image of a particular memory, so we can rehearse the clump and retrieve the quale of the actual memory.

Antegrade Amnesia

‌A declarative memory is thus a memory that is first of all explicit and second of all listed in the meta-index complex, so that it can be found. This should be testable in that there should be an antegrade amnesia if the hippocampal area is removed or damaged in some way. Such an amnesia has been found to exist, unfortunately a companion retrograde amnesia has also been found to exist that indicates that this model is as yet incomplete. If the retrograde amnesia was total, it would fit the pattern, but since it only covers about 2.5 years before the damage to the hippocampal area, it does not fit the pattern.