Chapter 7 Human Memory

Introduction

Memory is a fundamental human ability that significantly impacts our daily lives, influencing everything from recalling simple names to solving complex problems.

It allows us to retain a sense of personal identity, maintain social connections, and make decisions based on past experiences.

Given its crucial role in nearly all cognitive functions (like perception, thinking, and problem-solving), psychologists have extensively studied how information is committed to memory, how it is kept over time, why it is lost, and how memory can be improved.

Systematic study of memory dates back over a century. Pioneering work by German psychologist Hermann Ebbinghaus in the late 19th century, using himself as a subject, provided initial insights into forgetting patterns, showing that forgetting is initially rapid but slows down over time.

Later, Frederick Bartlett proposed that memory is not just a passive storage system but an active, constructive process. Using meaningful materials like stories, he demonstrated that what we remember is influenced by existing knowledge and undergoes changes during storage and retrieval, leading to qualitative differences between the original information and the recalled version.

This chapter explores these various aspects of memory, including different theoretical models and phenomena.

Nature Of Memory

Memory is essentially the capacity to retain and recall information over varying durations, depending on the cognitive task at hand.

This could involve holding information for just a few seconds (like a phone number before dialing) or retaining skills and knowledge for many years (like basic arithmetic learned in school).

Memory is understood as a process comprising three distinct but interconnected stages:

(a) Encoding: This is the initial stage where incoming information is processed and registered in a form usable by the memory system. When sensory organs receive a stimulus, it generates neural impulses. Encoding involves deriving meaning from this input and representing it in a way that allows for further processing and storage in memory.

(b) Storage: Following encoding, information needs to be retained over time. Storage refers to the process of maintaining or holding onto the encoded information within the memory system so it can be accessed later.

(c) Retrieval: To use stored information, it must be accessed and brought back into conscious awareness. Retrieval is the process of recovering stored information for use in cognitive tasks like problem-solving, decision-making, or simply recalling past events.

Memory failure can occur at any of these stages. Information might be forgotten because it was not properly encoded, inadequately stored, or cannot be successfully retrieved when needed.

Information Processing Approach : The Stage Model

Early views often saw memory as a simple storage container for all learned information. With the advent of computers, a new perspective emerged: the Information Processing Approach.

This approach views human memory as a system that processes information in a manner analogous to a computer, involving input, storage, manipulation, and output.

Both computers and humans register, store, and process information to perform tasks.

This analogy led to the development of the first influential model of memory by Atkinson and Shiffrin in 1968, known as the Stage Model.

The Stage Model proposes that information moves through a series of distinct memory stores or stages.

Memory Systems : Sensory, Short-Term And Long-Term Memories

According to the Stage Model, memory consists of three separate but interacting systems:

1. Sensory Memory:

• This is where incoming sensory information is initially registered.
• It has a large capacity, capturing a near-exact replica of the stimulus.
• However, information is held for a very brief duration, typically less than one second.
• Sensory memory is modality-specific; for example, visual sensory memory is called iconic memory, and auditory sensory memory is called echoic memory.
• Most information in sensory memory decays rapidly unless attended to.

2. Short-Term Memory (STM):

• Information that is attended to from sensory memory enters the STM.
• It has a limited capacity, typically holding about 7 ± 2 items or chunks of information (Miller's Magic Number).
• Information is held for a short duration, usually less than 30 seconds, unless actively maintained.
• According to Atkinson and Shiffrin, encoding in STM is primarily acoustical (based on sound).
• STM is fragile, as information is lost quickly without maintenance.

3. Long-Term Memory (LTM):

• Information that is processed further from STM enters the LTM.
• It has a vast, potentially unlimited capacity.
• Information can be stored for a very long duration, potentially a lifetime.
• Encoding in LTM is primarily semantic (based on meaning).
• Forgetting from LTM is often considered to be a problem with retrieval rather than a loss of stored information itself.

Control Processes: The Stage Model also includes control processes that regulate the flow and maintenance of information:

• Attention: Selects which information moves from sensory memory to STM.
• Maintenance Rehearsal: Repeating information (silently or verbally) to keep it active in STM, preventing decay.
• Elaborative Rehearsal: Linking new information in STM to existing knowledge already stored in LTM. This deep processing based on meaning facilitates transfer to LTM and long-term retention.
• Chunking: Organising smaller units of information into larger, meaningful chunks to increase the effective capacity of STM.

Challenges to the Stage Model: While influential, the Stage Model has faced criticism. Some studies suggest that encoding in STM can also be semantic, and in LTM can be acoustic. The case of KF, a patient with damaged STM but intact LTM, contradicted the idea that information must pass through STM to reach LTM.

These challenges led to alternative models of memory.

Working Memory (Box 7.1)

A modification of the STM concept, proposed by Baddeley, suggests that short-term memory is not a single passive store but a dynamic system called working memory. It functions as a mental "workbench" where information is actively held, manipulated, and processed during cognitive tasks.

Working memory consists of multiple components:

• Phonological Loop: Briefly stores auditory and verbal information (decays quickly if not refreshed).
• Visuospatial Sketchpad: Holds visual and spatial information (also has limited capacity).
• Central Executive: The control center that manages attention, allocates resources to the other components, integrates information from the loops and LTM, and plans/monitors behaviour.

Levels Of Processing

Proposed by Craik and Lockhart (1972), the Levels of Processing view suggests that memory retention is not determined by passing through distinct stores but by the depth at which information is processed.

The more deeply information is processed, the better it will be remembered.

Processing can occur at different levels:

• Shallow Processing:
  • Structural (Physical) Level: Focusing only on the physical features of the stimulus, such as the shape of letters or the appearance of a word (e.g., whether it's written in capitals).
  • Phonetic (Acoustic) Level: Processing information based on its sound or how it is pronounced (e.g., does this word rhyme with another?).

  Processing at these shallow levels typically results in fragile memories that are quickly forgotten.

• Deep Processing:
  • Semantic Level: Processing information based on its meaning. This involves understanding the significance of the information, relating it to existing knowledge, and thinking about its implications (e.g., thinking about the properties of a 'cat' when encountering the word).

  Semantic encoding is the deepest level of processing and leads to much better and longer-lasting memory retention.

This perspective has a practical implication: effective learning involves actively engaging with the meaning of the material, making connections, and reflecting on it, rather than just relying on simple repetition (rote memorisation). Deep processing through understanding and relating information enhances long-term retention.

Types Of Long-Term Memory

Just as short-term memory is seen as multi-component (working memory), long-term memory is also considered to consist of various types, holding different kinds of information.

One major classification divides LTM into:

• Declarative Memory: Memory for facts, events, and general knowledge that can be consciously recalled and verbally stated. This includes information like dates, names, definitions, and concepts. (Also called explicit memory).
• Procedural Memory: Memory for skills, procedures, and how to perform actions. This includes learned abilities like riding a bicycle, tying shoelaces, or playing a musical instrument. The knowledge is often difficult to verbalise and is demonstrated through performance. (Also called nondeclarative or implicit memory).

Within declarative memory, Tulving proposed a further distinction:

• Episodic Memory: Memory for specific events or episodes from one's own life, including their context (time, place, associated emotions). These memories are autobiographical, often emotionally charged, and are "dated" (we remember when they happened). However, we don't remember every life event, and unpleasant experiences may be recalled less vividly than pleasant ones.
• Semantic Memory: Memory for general knowledge, facts, concepts, rules, and meanings that are not tied to specific past experiences. This includes knowledge like capital cities, mathematical formulas, vocabulary, and logical principles. Semantic memory is typically affect-neutral and less susceptible to forgetting compared to episodic memory. This type of memory is "undated."

Box 7.2 and Box 7.3 provide additional details on memory types and measurement methods.

Box 7.2: Long-term Memory Classification (Additional Concepts)

• Flashbulb Memories: Vivid, detailed, and seemingly permanent memories of emotionally significant or surprising events (e.g., remembering exactly where you were when a major historical event occurred). While perceived as accurate, research shows they can still be prone to errors over time. The vividness might be due to high arousal and perhaps deeper processing/more retrieval cues.
• Autobiographical Memory: Personal memories of one's own past, similar to episodic memory. Research shows these memories are not uniformly distributed across the lifespan; there is a period of childhood amnesia (lack of memories from the first 4-5 years), a peak in memories from young adulthood (reminiscence bump), and good recall of recent events in old age.
• Implicit Memory: Memory that influences behaviour without conscious awareness or intention. It is often revealed through performance rather than verbal report (e.g., knowing how to type without being able to label keys on a keyboard). Implicit memories are retrieved automatically and can be demonstrated through techniques like priming. This type of memory is distinct from conscious, explicit memory (declarative memory) and is observed even in patients with amnesia for explicit memories.

Box 7.3: Methods of Memory Measurement

Psychologists use various methods to measure memory depending on the type of memory being studied:

• Free Recall and Recognition (for facts/episodes - Declarative Memory):
  • Free Recall: Participants learn a list of items (e.g., words) and later recall them in any order. Memory is measured by the number of items correctly recalled.
  • Recognition: Participants are shown items they learned mixed with new, unfamiliar items (distractors) and asked to identify which items they previously learned. Memory is measured by the number of correctly recognised "old" items. Recognition is generally easier than free recall.
• Sentence Verification Task (for Semantic Memory): Participants are presented with statements and asked to quickly judge whether they are true or false. The speed of response is used to infer the accessibility and organisation of semantic knowledge.
• Priming (for Implicit Memory): Participants are briefly exposed to stimuli (e.g., words). Later, their response to related stimuli is measured (e.g., completing word fragments). If prior exposure (priming) makes them faster or more likely to respond with the previously seen stimuli, it indicates implicit memory, even if they consciously don't remember seeing the initial stimuli.

Knowledge Representation And Organisation In Memory

Given the vast amount of information stored in long-term memory and our efficient retrieval of it, the way this knowledge is organised is crucial.

Research on semantic memory retrieval tasks (like verifying factual statements) provides insights into this organisation.

The basic unit of knowledge representation in LTM is often considered a concept – a mental grouping of similar objects, events, or ideas.

Concepts are organised into categories, which are themselves concepts but serve to group other concepts based on shared features (e.g., 'fruit' is a category containing concepts like 'mango', 'apple').

Knowledge can also be organised into schemas – mental frameworks representing our knowledge and assumptions about the world (e.g., a 'drawing room' schema includes expected objects and their typical arrangement).

At a higher level, knowledge in LTM is thought to be organised in a hierarchical network structure, as proposed by Collins and Quillian (1969). This network consists of nodes (representing concepts) connected by labeled relationships (indicating category membership or attributes).

In this model, properties applicable to an entire category are stored at the highest level of the hierarchy, promoting cognitive economy by avoiding redundancy. For example, "Can breathe" is stored at the 'Animal' node, not repeated for 'Bird', 'Fish', 'Canary', etc. Verifying statements takes longer if it requires traversing more links in the hierarchy (e.g., verifying "Canary is an animal" takes longer than "Canary is a bird").

Beyond verbal formats, information can also be encoded and stored as images – concrete representations capturing perceptual attributes. The dual coding hypothesis (Paivio) suggests that concrete information (nouns, objects) is encoded both verbally and visually (as images), while abstract concepts primarily use a verbal code. Information encoded both ways is generally recalled more easily.

Images can contribute to mental models – internal representations of how the environment is structured, formed from both visual and verbal information (e.g., creating a mental map from directions).

Box 7.4: Memory Making : Eyewitness and False Memories

Eyewitness Memory: Often used in legal proceedings, eyewitness testimony is considered important but can be unreliable. Research by Loftus demonstrates that eyewitness memory is prone to flaws and can be influenced by post-event information, particularly leading questions. The way questions are phrased (e.g., using "smashed" vs. "contacted" in a car accident scenario) can alter memory for details like speed. High emotional arousal during violent events can also impair detailed encoding.

False Memory: It's possible to create memories of events that never actually happened. Studies show that by having individuals vividly imagine events that didn't occur (imagination inflation), they can later come to believe these imagined events were real experiences. This highlights the constructive nature of memory and its susceptibility to suggestion and internal mental processes like imagination.

Memory As A Constructive Process

Early views of memory, influenced by Ebbinghaus, often saw it as a process of simple reproduction – retrieving stored information exactly as it was originally encoded. Any deviation was considered an error.

Bartlett challenged this, arguing that memory is an active, constructive process. What we remember is constantly being changed and modified based on our existing knowledge, beliefs, expectations, and psychological state.

Using meaningful materials (stories, fables) and the method of serial reproduction (participants repeatedly recalling the material over time), Bartlett observed systematic 'errors'. Participants often altered the material to fit their existing knowledge (schemas), omitted irrelevant details, elaborated on main points, and made the story more coherent and rational according to their understanding.

Bartlett used the term schema to describe the active organisation of past experiences and knowledge that influences how new information is interpreted, stored, and retrieved. Memory construction means information is encoded and stored not in isolation, but within the framework of a person's existing knowledge and expectations.

Nature And Causes Of Forgetting

Forgetting, the inability to recall information, is a common experience. Various theories attempt to explain why we forget.

Ebbinghaus's early studies using nonsense syllables demonstrated a typical forgetting pattern: a rapid initial drop in retention followed by a much slower decline over time (Figure 7.3).

Major theories of forgetting include:

• Trace Decay Theory (Disuse Theory): Suggests that memory leaves physical "traces" in the brain. If these traces are not used or rehearsed over time, they naturally fade or decay, making the information unavailable. This theory is considered inadequate because factors other than time elapsed (like interference) also cause forgetting. Studies show people forget less during sleep than during wakefulness, contrary to what a simple decay theory would predict.
• Interference Theory: Proposes that forgetting occurs when other stored information interferes with the retrieval of target information. This theory assumes associations are formed during learning and interference happens during recall when competing associations are activated.
• Proactive Interference: Earlier learning interferes with the recall of newly learned information. Example: Difficulty remembering new French words because you keep recalling previously learned English words.
• Retroactive Interference: Later learning interferes with the recall of previously learned information. Example: Difficulty remembering old English vocabulary after learning new French vocabulary.

Experimental designs (Table 7.1) are used to isolate and demonstrate proactive and retroactive interference.

Type of Interference	Phase 1	Phase 2	Testing Phase
Retroactive Interference	Learns A	Learns B	Recalls A
Experimental Group	Learns A	Learns B	Recalls A
Control Group	Learns A	Rests	Recalls A
Proactive Interference	Learns A	Learns B	Recalls B
Experimental Group	Learns A	Learns B	Recalls B
Control Group	Rests	Learns B	Recalls B

• Retrieval Failure Theory (Cue-Dependent Forgetting): Argues that information is still stored in memory but is inaccessible due to the absence or inappropriateness of retrieval cues – stimuli or hints that help access stored information. Tulving suggested that forgetting isn't always loss of memory traces, but rather a failure to access them without the right cues. Providing appropriate cues (like category names for a list of words) can dramatically improve recall. The context of learning can also serve as retrieval cues.

Box 7.5: Repressed Memories

Sigmund Freud suggested that highly painful, threatening, or embarrassing traumatic experiences can be unconsciously blocked from conscious memory, a process called repression. These 'repressed memories' are thought to reside in the unconscious mind, unavailable for normal retrieval.

Extreme stress or trauma can also lead to forms of psychological amnesia, such as a fugue state, where individuals may lose their personal identity and past memories, sometimes adopting a new identity, as a psychological escape mechanism.

Stress and high anxiety, like during exams, can also impair memory retrieval, causing temporary forgetfulness even of well-learned material.

Enhancing Memory

Most people want to improve their memory. Various strategies, known as mnemonics, can help enhance memory and retrieval.

Mnemonics often use images or organisation to make information more memorable.

Mnemonics using Images: These techniques involve creating vivid, interacting mental images:

• Keyword Method: Useful for learning foreign language vocabulary. Find an English word (keyword) that sounds similar to the foreign word and create a mental image of the keyword interacting with the meaning of the foreign word. Example: Spanish 'Pato' (duck) -> Keyword 'pot' -> Imagine a duck in a pot.
• Method of Loci (Memory Palace): Involves associating items to be remembered with a series of familiar physical locations (loci). Visualise a well-known route or building, then place mental images of the items at specific points along that route. To recall, mentally walk through the locations and retrieve the associated items. Effective for remembering items in a specific order.

Mnemonics using Organisation: These involve structuring the material to be learned:

• Chunking: Grouping smaller units of information into larger, meaningful chunks to increase STM capacity and facilitate LTM storage (e.g., remembering phone numbers as groups of digits).
• First Letter Technique: Creating a word or sentence from the first letters of the items you need to remember. Example: VIBGYOR for the colours of the rainbow (Violet, Indigo, Blue, Green, Yellow, Orange, Red).

Beyond specific mnemonics, a broader approach to memory improvement involves applying knowledge about memory processes:

• (a) Engage in Deep Level Processing: As suggested by the Levels of Processing view, focus on understanding the meaning of the information. Ask questions, relate it to what you already know and your life experiences. Semantic processing makes information part of your existing knowledge structure, improving retention.
• (b) Minimise Interference: Avoid studying very similar subjects consecutively, as this increases interference. Distribute your learning sessions with breaks or switch to unrelated subjects to reduce both proactive and retroactive interference.
• (c) Give Yourself Enough Retrieval Cues: Identify potential retrieval cues within the material while learning and create strong associations between these cues and the content. These cues will later help you access the stored information during recall.

The PQRST method (Preview, Question, Read, Self-recitation, Test) is another strategy for effective learning and memory. Preview the material, formulate questions, read to find answers, try to recall and recite the information yourself, and finally test your understanding.

It is important to note that no single technique guarantees perfect memory. Memory improvement depends on multiple factors, including physical health, motivation, interest in the subject, and familiarity with the material. Using a combination of strategies tailored to the specific learning task is most effective.