Memory Architecture, Working Memory, Forgetting Theories & Retrieval Dynamics

Memory is the cognitive faculty by which organisms encode, consolidate, store, and retrieve mental representations of past experience. Rather than functioning as a passive, verbatim recording device, memory operates as a dynamic, reconstructive information-processing system distributed across distinct neuroanatomical circuits.

For CUET PG Psychology (HUQP20), Unit 5 questions in this domain focus heavily on:

1. Architectural Models of Memory: Atkinson and Shiffrin's Modal Model (Sensory, Short-Term, Long-Term), Sperling's partial-report technique, and the serial position curve (primacy vs. recency).
2. Alan Baddeley's Working Memory Model: The Central Executive, Phonological Loop, Visuospatial Sketchpad, and the Episodic Buffer (Baddeley, 2000).
3. Long-Term Memory Taxonomies: Larry Squire and Endel Tulving's distinction between Explicit / Declarative (Episodic vs. Semantic) and Implicit / Non-Declarative (Procedural, Priming, Conditioning) memory.
4. Encoding & Retrieval Principles: Craik and Lockhart's Levels of Processing, Tulving's Encoding Specificity Principle (context- and state-dependent memory), and Transfer-Appropriate Processing.
5. Theories of Forgetting: Ebbinghaus's forgetting curve, Proactive versus Retroactive Interference, Delos Wickens's Release from Proactive Interference, and retrieval failure (cue dependency).
6. Reconstructive Memory & Neurobiology: Elizabeth Loftus's Misinformation Effect, Patient H.M., amnesic profiles, and Long-Term Potentiation (LTP) mediated by NMDA glutamate receptors.

1. Multi-Store Memory Architecture: The Modal Model

Formulated by Richard Atkinson and Richard Shiffrin (1968), the Modal Model conceptualizes memory as an information-processing pipeline moving through three structural stores:

1. Sensory Memory: Iconic vs. Echoic

• Iconic Memory (Visual): Duration $<500\text{ ms}$; massive capacity.
  • George Sperling's Landmark Partial-Report Paradigm (1960): Flashed $3 \times 4$ letter grids for 50 ms:
    • Whole Report: Participants recalled only ~4.5 letters before the visual trace decayed.
    • Partial Report: Sounding a high, medium, or low pitch tone immediately after grid offset to signal which row to report resulted in 100% accuracy for any cued row, proving the entire display was held in iconic storage.
• Echoic Memory (Auditory): Duration roughly 3 to 4 seconds (Darwin, Turvey, & Crowder); essential for speech comprehension and parsing sequential language.

2. Short-Term Memory (STM)

• Capacity: George Miller's (1956) classic span of $7 \pm 2$ items ("The Magical Number Seven"), enhanced via chunking (recoding discrete elements into meaningful cognitive units). Nelson Cowan (2001) revised this to $4 \pm 1$ chunks when rehearsal is blocked.
• Duration: Brown-Peterson distractor task proved that without rehearsal, STM traces decay within 15 to 20 seconds.
• Primary Coding: Auditory/acoustic coding (Conrad's acoustic confusion errors).

3. The Serial Position Curve: Primacy vs. Recency

When participants freely recall an unorganized word list, accuracy plots as a U-shaped curve:

• Primacy Effect: Superior recall of words at the beginning of the list; mediated by Long-Term Memory (early words receive dedicated elaborative rehearsal and consolidate into LTM).
• Recency Effect: Superior recall of words at the end of the list; mediated by Short-Term Memory (last words are still sitting in the active STM buffer).
• The Distractor Manipulation: Inserting a 30-second backward-counting distractor task between list presentation and recall completely abolishes the Recency Effect, while the Primacy Effect remains 100% intact.

2. Working Memory Architecture (Baddeley & Hitch)

Alan Baddeley and Graham Hitch (1974; revised 2000) replaced the concept of a passive, unitary STM with a dynamic, multi-component Working Memory System: