Is the Multi store model of memory perfect?

“Multistore model of memory best explains the nature of memory ” evaluate the statement in theoretical perspective and empirical evidences. (20 marks) (CSE 2019)
Model answer :
Human memory system is very fascinating. All the continuity in behaviour is because of this.
How does our memory system work? This has been an intriguing question.
The information processing model of memory attempted to explain the structure and functioning of human memory using the computer analogy. It assumes that human memory system comprises at least two subsystem, STM and LTM like RAM and ROM systems of computer. This Duplex theory emphasizes the distinction of STM and LTM. Moreover, at the initial stage, a sensory register kind of memory was later added and renamed as the multi store model due to its very structure. The distinction is primarily made on the basis of duration or term of holding the information. Sensory memory holds information for very brief period ( about a second), the STM holds for more( about 20/30 second) and the LTM holds for the longest.
Functionally, sensation takes information to sensory register, attention to STM and meaningful processing to LTM.
This theory answers the basic questions of memory very well. How do we encode, store, retrieve and sometimes forget information.
There is debate over the nature of encoding, duration, capacity, functionality of each system but the debate doesn’t discard the analogy of multi stores.
But there are theories that have different perspectives on the human memory system. Craik and Lockhart criticised the distinction among sensory, short and long term memory system. They argue” why we should see these systems as distinct and different. Why not to see them as a part of single system…Rather to distinguishing them on the basis of duration of holding information, why we can’t see the difference just in terms of elaboration which causes different longevity of the information we hold “
Moreover, Tulving argues that apart from duration & elaboration, nature of material that we encode also matters.
Empirically, all models have evidences to support their points.
Without maintenance rehearsal, information is lost and with meaningful rehearsals information it lives longer . But such an evidence is indirect and could be interpreted differently. Rogers, for instance, found that distinction of maintenance and meaningful rehearsals is not very appropriate. Actual difference is in the nature of elaboration. Briefer duration of memory could be due to physical or acoustic elaboration whereas semantic and self reference based elaboration give longer retention.
Secondly, the distinction between STM and LTM on the nature of encoding is also questioned. STM doesn’t have only acoustic encoding as suggested by Atkinson model. ( Baddeley).

The hard evidence for the multi store model, however, comes from neuroscience studies of HM. HM has functional STM but dysfunctional LTM.

Critics raise the question of Atkinson and Shiffrin’s explanation of the pathway that information must pass from STM in order to reach LTM (K. F case, in this despite having dysfunctional STM , his LTM remains intact) . Likewise, procedural memory of HM improved with practice.
Moreover, incidental memory, that develops without intention, also questions the control mechanisms suggested by the multi store model.
We, therefore, can judge that STM and LTM kind of distinction suggested by the multi store model has some merits but it doesn’t answer all the questions related to its functioning. The alternate, the level of processing model by Lockhart, has good explanation for functioning but doesn’t discard the structural distinction completely.
We, thus, don’t have sufficient evidences and reasons to say that multi store model is the best theory of memory.