Processes of Human Memory

Narrative Explanatory Content

1.1 What is Memory? The Three Core Processes

Memory is the cognitive process of retaining, retrieving, and using information about stimuli, images, events, ideas, and skills after the original information is no longer present. It is fundamental to our identity, our ability to learn, and our capacity to function in the world. Psychologists conceptualise memory as involving three distinct stages:

• Encoding: The process of converting incoming sensory information into a form that can be stored in memory. This is like typing information into a computer.
• Storage: The process of holding information in memory over time. This is like saving a file on a hard drive.
• Retrieval: The process of locating and bringing stored information into conscious awareness. This is like opening a saved file.

Forgetting can occur at any of these three stages. If information is not encoded properly, it will never be stored. If it is stored but the memory trace decays, it cannot be retrieved. If it is stored, but we lack the right cues, retrieval can fail.

1.2 The Multi-Store Model (MSM): A Foundational Framework

In 1968, Richard Atkinson and Richard Shiffrin proposed one of the first and most influential models of memory, known as the Multi-Store Model (MSM). They suggested that memory is not a single entity but consists of three separate, unitary stores through which information flows in a linear sequence. This model is often called the "modal model" because it formed the foundation for so much later research.

1.3 The Three Stores of the MSM

The MSM posits three distinct memory stores, each with its own characteristics regarding coding, capacity, and duration:

1. The Sensory Register: This is the first point of contact for information from the environment. It is not a single store but has different registers for each sense (e.g., the iconic store for visual information and the echoic store for auditory information). It has a very large capacity but an extremely brief duration. Information here will decay almost instantly unless we pay attention to it.
2. Short-Term Memory (STM): Information that is attended to in the sensory register passes into the STM. This store has a very limited capacity and a short duration. Information is held here through maintenance rehearsal (repetition) and is primarily coded acoustically (by sound).
3. Long-Term Memory (LTM): If information in the STM is rehearsed for long enough (prolonged rehearsal), it is transferred to the LTM. This store is believed to have a potentially unlimited capacity and duration, holding information for a lifetime. Information here is primarily coded semantically (by meaning).

1.4 The Flow of Information: Attention and Rehearsal

The MSM is a structural model that emphasises the processes that move information from one store to the next. Attention is the critical process that selects information from the vast sensory register to be passed on to the STM. Without attention, information decays and is lost. Maintenance Rehearsal (rote repetition) is the process that keeps information active in the STM. According to Atkinson and Shiffrin, the more you rehearse information in STM, the more likely it is to be transferred to LTM. To recall information from LTM, it must be transferred back into STM, a process known as retrieval.

Narrative Explanatory Content

2.1 The Sensory Register: A Fleeting First Impression

The Sensory Register is the memory store that receives information from the environment via our senses. It has a very large capacity but information decays incredibly quickly. The key study investigating this store was conducted by Sperling (1960).

• Capacity & Duration (Sperling, 1960): Sperling flashed a grid of 12 letters onto a screen for just 50 milliseconds. When he asked participants to recall the whole grid, they could only manage about 4-5 letters. However, when he used a tone to indicate which row to recall immediately after the grid disappeared, recall was very high. This showed that the capacity of the sensory register is large (all 12 letters were available), but the duration is extremely short (less than half a second), as the information decays before it can be reported.
• Coding: Information is coded in its raw, sensory form (e.g., iconic for visual, echoic for auditory).

2.2 Short-Term Memory (STM): The Bottleneck

Information only passes from the sensory register to STM if it is paid attention to. STM is often seen as a "bottleneck" in the memory system because its capacity is so much smaller than the other stores.

2.3 Capacity of STM: The Magic Number Seven

The capacity of STM refers to how much information it can hold at one time.

• Jacobs (1887): Developed the digit span technique. He found that the average span for digits was 9.3 items, and for letters was 7.3 items.
• Miller (1956): In his famous paper "The Magical Number Seven, Plus or Minus Two," Miller reviewed existing research and concluded that the capacity of STM is between 5 and 9 items. He also noted that we can increase this capacity by chunking—grouping sets of digits or letters into meaningful units. For example, the string FB-IA-TT-CI-A is 10 items, but if chunked as FBI-ATT-CIA, it becomes three meaningful chunks that are easier to remember.

2.4 Duration of STM: The 18-Second Window

The duration of STM is how long it can hold information without rehearsal.

• Peterson & Peterson (1959): They gave participants a trigram (e.g., "YCG") to remember and a 3-digit number. To prevent rehearsal, participants had to count backwards from the number for a set period (a retention interval) of 3, 6, 9, 12, 15, or 18 seconds. They found that after 3 seconds, recall was around 80%, but after 18 seconds, it dropped to just 10%. This suggests that the duration of STM is very short (around 18-30 seconds) unless the information is actively rehearsed.

Coding in STM: Research by Baddeley (1966) found that participants had more difficulty recalling acoustically similar words (e.g., cat, cab, can) from STM than semantically similar words (e.g., great, large, big). This suggests that STM primarily codes information acoustically.

Narrative Explanatory Content

3.1 Introduction to Long-Term Memory (LTM)

Long-Term Memory (LTM) is the final store in the Multi-Store Model, and it's the repository for our lasting memories, knowledge, and skills. Unlike the fleeting Sensory Register or the bottleneck of STM, LTM is our vast archive of information. According to the MSM, information is transferred from STM to LTM through prolonged or elaborative rehearsal. Once in LTM, a memory is thought to be relatively permanent. To be consciously aware of a memory from LTM, it must be retrieved back into STM.

3.2 Duration of LTM: A Lifetime of Memories

The duration of LTM is considered to be potentially unlimited—from minutes to a lifetime. A key study that provides compelling evidence for this is from Bahrick et al. (1975).

• Bahrick et al. (1975) - High School Yearbooks: Researchers tested the memory of 392 American ex-high school students (aged 17-74) for their former classmates. They used a series of memory tests, including free recall of names and photo-recognition tests. The results were remarkable:
  • Participants tested within 15 years of graduation were about 90% accurate in photo recognition.
  • Participants tested after 48 years were still around 70% accurate.
  This study has high ecological validity as it used real-life meaningful memories, and it strongly suggests that the duration of LTM can indeed be a lifetime.

3.3 Capacity and Coding in LTM

Like its duration, the capacity of LTM is believed to be practically unlimited. We can continue to learn and store new information throughout our lives without running out of space. There is no known limit to how much information a person can store in LTM.

The primary form of coding in LTM is semantic (based on meaning). This was demonstrated by Baddeley (1966) in the same study where he tested STM.

• Baddeley (1966) - Coding Study: When participants were asked to recall word lists after a 20-minute interval (to ensure the information was in LTM), they had the most difficulty recalling words that were semantically similar (e.g., 'great', 'large', 'big'). They had less trouble with acoustically similar words. This suggests that LTM organises information based on its meaning, and confusion occurs when meanings are similar.

3.4 Comparing the Memory Stores

The MSM proposes three qualitatively different stores. The contrast between STM and LTM is particularly stark and is supported by a wealth of evidence.

Narrative Explanatory Content

4.1 Supporting Evidence: The Serial Position Effect

One of the strongest pieces of evidence for the existence of separate STM and LTM stores is the serial position effect. This is the tendency to recall the first and last items in a list better than the items in the middle.

• Glanzer & Cunitz (1966): Participants were asked to recall a list of words. The results consistently formed a U-shaped curve.
  • The Primacy Effect: Words at the beginning of the list were recalled well. The MSM explains this by suggesting these words have been rehearsed and transferred to LTM.
  • The Recency Effect: Words at the end of the list were also recalled well. The MSM explains this by suggesting these words are still in the STM.
  When Glanzer and Cunitz introduced a delay with an interference task (like counting backwards) after the list, the recency effect disappeared, but the primacy effect remained. This supports the model because the interference task displaced the items from the STM.

4.2 Supporting Evidence: The Case of Henry Molaison (HM)

Clinical case studies of patients with amnesia provide powerful evidence. The most famous is Henry Molaison (HM). To treat severe epilepsy, HM had surgery to remove parts of his brain, including the hippocampus. Afterwards, his memory was profoundly affected.

• HM's Condition: His STM was largely intact; he could hold a conversation. However, he was unable to form new long-term memories (a condition called anterograde amnesia). He could not remember what he had for breakfast or recognise doctors he had met countless times. He was, in a sense, stuck in the present.
• Support for the MSM: HM's case strongly supports the idea that STM and LTM are separate, distinct stores located in different parts of the brain. His STM worked, but the mechanism to transfer information to LTM was broken.

4.3 Limitations of the Multi-Store Model

Despite its influence, the MSM is now considered too simplistic. There are several key criticisms:

1. Oversimplification of STM and LTM: The model sees both STM and LTM as single, unitary stores. However, evidence suggests they are more complex. For example, the Working Memory Model shows STM is an active processor, not a passive store. The case of HM also challenges the unitary view of LTM; while he couldn't form new episodic memories (events), he could learn new motor skills (procedural memories), suggesting LTM has different components.
2. Oversimplification of Rehearsal: The MSM claims that the amount of maintenance rehearsal determines transfer to LTM. However, Craik & Lockhart (1972) argued that the type of rehearsal is more important. Elaborative rehearsal (linking information to existing knowledge) is more effective than simple maintenance rehearsal (rote repetition). We remember things we are interested in, even if we don't rehearse them.
3. Linear Flow: The model suggests information flows in one direction from STM to LTM. However, we often use information from LTM to help us process information in STM (e.g., chunking). This suggests a more interactive relationship between the stores.

4.4 Conclusion: The Legacy of the MSM

The Multi-Store Model was a pioneering effort that provided the first comprehensive framework for studying memory. Its distinction between STM and LTM and its emphasis on structure were hugely influential and generated a vast amount of research. While its specific claims about unitary stores and the role of rehearsal have been challenged and refined by later models, it remains a crucial starting point for understanding the architecture of human memory.

Narrative Explanatory Content

5.1 A Shift from Structure to Process

In response to the criticisms of the Multi-Store Model, particularly its emphasis on maintenance rehearsal, Fergus Craik and Robert Lockhart (1972) proposed a different approach. Instead of focusing on separate memory stores, they focused on the processes involved in memory formation. They argued that the strength and duration of a memory trace are not determined by which store it is in, but by the depth at which the information is processed.

5.2 The Levels of Processing (LOP) Theory

The LOP theory suggests that memory is a byproduct of perception and comprehension. There is a continuum of processing, from shallow to deep.

• Shallow Processing: This involves focusing on the superficial, physical features of a stimulus. It leads to a fragile memory trace that decays quickly.
  • Structural Processing: Encoding the physical appearance of something (e.g., "Is the word in capital letters?").
  • Phonemic Processing: Encoding the sound of something (e.g., "Does the word rhyme with 'cat'?").
• Deep Processing: This involves focusing on the meaning of the information and making connections to existing knowledge. This is a form of elaborative rehearsal.
  • Semantic Processing: Encoding the meaning of the information (e.g., "Would the word fit in this sentence: 'He met a ___ in the street'?").

The core idea is that deeper, more meaningful processing leads to a more durable and long-lasting memory trace.

5.3 Evidence for LOP: Craik & Tulving (1975)

The classic study supporting LOP was conducted by Craik and Tulving. They gave participants a list of words and asked them a question about each word. The questions were designed to induce different levels of processing.

After the question phase, participants were given a surprise recognition test for the words. The results showed that words processed semantically (deeply) were remembered far better than words processed structurally or phonemically (shallowly). This supports the theory that depth of processing, not just rehearsal, affects memory.

5.4 Evaluating the Levels of Processing Theory

LOP theory has been influential and has practical applications, particularly in education (e.g., encouraging students to make mind maps rather than just re-reading notes). However, it also has significant limitations.

• Strengths: It shifted the focus of memory research from structure to process and explained why some things are easier to remember than others. It highlights that memory is an active process.
• Weakness (Circularity): The major criticism is that the argument is circular. How do we know that semantic processing is deep? Because it leads to better memory. Why does it lead to better memory? Because it is deep processing. The theory fails to provide an independent way to measure 'depth' of processing.
• Weakness (Vagueness): The terms 'deep' and 'shallow' are not clearly defined and are difficult to operationalise. The theory describes rather than explains how deeper processing leads to better memory.

Narrative Explanatory Content

6.1 The Rationale for the Working Memory Model

Alan Baddeley and Graham Hitch (1974) were critical of the MSM's depiction of Short-Term Memory. They argued that STM is not a single, passive store but a complex and active system responsible for manipulating information and coordinating tasks. They pointed to several problems with the MSM's STM:

• It can't explain how we can perform two different tasks at once (dual-task performance).
• It doesn't account for evidence from brain-damaged patients like KF, who had impaired verbal STM but intact visual STM.
• It overemphasises the role of rehearsal for transfer to LTM.

To address these issues, they proposed the Working Memory Model (WMM), which is a model of STM only. It sees STM as an active 'workbench' where information is held and manipulated to perform cognitive tasks.

6.2 The Structure of the Working Memory Model

The WMM is a multi-component model. The original 1974 model consisted of three main parts:

6.3 The Central Executive: The Attentional Control System

The Central Executive (CE) is the most important but least understood component of the WMM. It is not a memory store itself but an attentional control system that manages and coordinates the other components ('slave systems').

• Analogy: Think of it as the CEO of a company. It doesn't do the work itself, but it decides what needs to be done, delegates tasks to the appropriate departments (the slave systems), and monitors their progress.
• Key Functions:
  • Directing attention and focus to particular tasks.
  • Coordinating the slave systems.
  • Switching between tasks.
  • Retrieving information from LTM.
• Capacity: The CE has a very limited processing capacity. It cannot attend to too many things at once. For example, when you are driving in a complex situation, you might have to stop talking because the CE is fully occupied with the driving task.

6.4 Evaluating the Central Executive

While the concept of an attentional controller is a major step forward from the MSM, the Central Executive is the most criticised part of the model.

• Vagueness: Critics argue that the CE is too vague and its functions are not clearly defined. Baddeley himself admitted in 2003 that it was "the most important but least understood component." Saying it's responsible for 'attention' doesn't explain how it works.
• The Homunculus Problem: The CE is sometimes criticised as being a 'homunculus'—a little man inside the head who makes all the decisions. This doesn't explain anything; it just pushes the question back a level (i.e., what controls the little man's brain?).
• Evidence: Evidence for the CE is sparse. Some comes from patients with frontal lobe damage who show impaired decision-making and attentional control, suggesting the CE might be located in the prefrontal cortex. However, it is difficult to design tasks that test the CE in isolation.

Narrative Explanatory Content

7.1 The Phonological Loop: The "Inner Ear" and "Inner Voice"

The Phonological Loop (PL) is the slave system responsible for dealing with auditory and verbal information (the sounds of language). It is subdivided into two parts:

• The Phonological Store (the "Inner Ear"): This stores the words you hear. It has a very limited duration of about two seconds.
• The Articulatory Process (the "Inner Voice"): This allows for maintenance rehearsal of information in the phonological store. It's like repeating a phone number to yourself over and over. This process has a capacity of what you can say in about two seconds.

Evidence for the Phonological Loop:

• Phonological Similarity Effect: Baddeley (1966) found that it's harder to remember a list of acoustically similar words (e.g., cat, cab, can) than dissimilar words. This is because they are stored in the phonological store and get confused.
• Word Length Effect: Baddeley et al. (1975) found that people can remember more short words (e.g., 'sum', 'harm') than long words (e.g., 'opportunity', 'university'). This is because the articulatory process can rehearse more short words in its two-second window. This effect disappears if participants are given an articulatory suppression task (e.g., repeating "the the the..."), which blocks the inner voice.

7.2 The Visuo-Spatial Sketchpad: The "Inner Eye"

The Visuo-Spatial Sketchpad (VSS) is the second slave system. It is responsible for storing and manipulating visual and spatial information. For example, you use it when you are asked to picture the route from your home to college or to count the number of windows in your house.

Logie (1995) suggested the VSS is also subdivided:

• The Visual Cache: Stores visual data (e.g., information about form and colour).
• The Inner Scribe: Records the arrangement of objects in the visual field and rehearses information in the visual cache.

7.3 Evidence for Separate Slave Systems: Dual-Task Studies

The strongest evidence for the WMM's multi-component structure comes from dual-task studies. The model predicts that if two tasks use the same slave system, performance will be poor. If they use different slave systems, performance should be good.

Baddeley & Hitch (1976) conducted a study where participants had to perform a reasoning task (using the CE) while simultaneously reciting a list of numbers (using the PL). They found that as the number of digits increased, performance on the reasoning task only slowed down slightly. This shows that the two tasks can be performed concurrently, supporting the idea of separate components.

7.4 How the WMM Improves on the MSM

The existence of these separate slave systems directly addresses the weaknesses of the MSM's unitary STM. It can explain:

• How we can perform dual tasks.
• The case of KF, who had a damaged Phonological Loop but a functioning Visuo-Spatial Sketchpad.
• The active nature of short-term memory as a 'workbench' for complex cognitive tasks.

Narrative Explanatory Content

8.1 A Problem for the Original WMM

The original 1974 model was highly successful, but Baddeley identified a key problem. The model explained how we process separate visual and auditory information, but it didn't explain how these streams are integrated. For example, when you watch a film, you effortlessly combine the visual images (from the VSS) with the sounds and dialogue (from the PL) into a single, coherent experience or 'episode'. The Central Executive was proposed to do this, but it has no storage capacity of its own. Where was this integration happening?

8.2 The Solution: The Episodic Buffer (Baddeley, 2000)

To solve this problem, Baddeley added a fourth component to the model in 2000: the Episodic Buffer (EB).

• Function: The EB is a temporary, general-purpose store that integrates information from the PL, the VSS, and LTM into a single 'episode'. It's like a mental TV screen where all the different sensory information is combined.
• Capacity: It has a limited capacity of about four 'chunks' of information.
• Link to LTM: Crucially, the EB provides the link between working memory and LTM, allowing us to incorporate our long-term knowledge into our current experience.

8.3 Overall Evaluation of the Working Memory Model

The WMM is a significant improvement on the MSM's view of STM and is widely accepted today. However, it is not without its weaknesses.

8.4 Conclusion: MSM vs. WMM

The MSM and WMM should not be seen as competing models of memory as a whole. The MSM is a general model of the entire memory system, while the WMM is a detailed model specifically of one part of that system—short-term memory. The WMM effectively replaces the MSM's concept of a unitary STM with a more complex and plausible multi-component 'workbench'. While the MSM was a vital first step, the WMM provides a more accurate and useful account of the active processes involved in short-term and working memory.

Narrative Explanatory Content

9.1 Beyond a Unitary LTM

A major criticism of the Multi-Store Model is that it views LTM as a single, unitary store. However, our experience tells us that our memory for facts (e.g., "Paris is the capital of France") feels very different from our memory for personal events (e.g., "I remember my first trip to Paris"). Endel Tulving (1972) was the first to propose that LTM is not one store, but several. He made a distinction between two types of declarative memory (memories that can be consciously recalled and 'declared').

9.2 Episodic Memory: Your Mental Diary

Episodic memory is the LTM store for personal events. It's like a mental diary or autobiography. These memories are 'time-stamped'—you remember when they happened. They also involve recalling the context (where you were, who you were with) and the emotions you felt at the time.

• Examples: Your first day of school, your last holiday, what you had for breakfast this morning.
• Characteristics: Time-stamped, complex, requires conscious effort to recall.

9.3 Semantic Memory: Your Mental Encyclopaedia

Semantic memory is the LTM store for our knowledge of the world. This includes facts, concepts, and the meaning of words. It's like a mental encyclopaedia or dictionary. These memories are not time-stamped; you know that a dog is an animal, but you probably don't remember when or where you first learned that fact.

• Examples: Knowing that London is in England, the meaning of the word 'psychology', understanding that 2+2=4.
• Characteristics: Not time-stamped, shared knowledge, less personal.

9.4 Evidence for the Distinction

The distinction between episodic and semantic memory is supported by several lines of evidence:

• Clinical Case Studies: The cases of HM and Clive Wearing are crucial. Both men suffered severe damage to their episodic memory. They could not recall personal events from their past or form new ones. However, their semantic memory was largely intact. HM knew what a dog was but couldn't remember ever stroking one. Clive Wearing, a musician, knew how to read music and play the piano, but couldn't remember a single musical lesson. This shows a clear dissociation between the two systems.
• Brain Scanning Evidence: Tulving et al. (1994) had participants perform various memory tasks while their brains were scanned using a PET scanner. They found that when recalling episodic memories, different areas of the prefrontal cortex were active compared to when recalling semantic memories. This provides biological evidence for separate systems.

Narrative Explanatory Content

10.1 Procedural Memory: The Memory for "How"

In addition to episodic and semantic memory, there is a third type of LTM: procedural memory. This is our memory for actions, skills, and how to do things. These memories are acquired through practice and repetition.

• Examples: Riding a bicycle, tying your shoelaces, playing the guitar, driving a car.
• Characteristics: These memories are non-declarative or implicit—we are not consciously aware of them. We can perform the skill, but we often find it very difficult to explain to someone else exactly how we do it. For example, try explaining how you keep your balance on a bike. The memory is in the action itself.

10.2 Evidence for Procedural Memory: HM and Mirror-Drawing

The most compelling evidence for a separate procedural memory system comes, once again, from the case of HM. While he could not form new episodic or semantic memories, he was able to learn new skills.

• The Mirror-Drawing Task: HM was asked to trace a shape (e.g., a star) while looking at his hand in a mirror. This is a difficult task that requires practice. Over several days, HM's performance on this task improved significantly. However, each day he had no conscious recollection of ever having done the task before.
• Implications: This shows a clear dissociation. His procedural memory was working (he was learning the skill), but his episodic memory was not (he couldn't remember learning it). This strongly suggests that procedural memory is a separate system, located in a different part of the brain (e.g., the cerebellum) from declarative memory.

10.3 The Big Picture: Explicit vs. Implicit Memory

The three types of LTM can be organised under two main headings: explicit and implicit memory.

• Explicit Memory (Declarative): This refers to memories that can be consciously recalled and described. It involves 'knowing that'. Both episodic and semantic memory are types of explicit memory.
• Implicit Memory (Non-Declarative): This refers to memories that we are not consciously aware of. It involves 'knowing how'. Procedural memory is the main type of implicit memory.

10.4 Evaluating the LTM Types Model

The model of LTM types is a significant improvement on the MSM's unitary store and is well-supported by evidence.

• Strengths: It has strong support from clinical case studies (HM, Clive Wearing) and brain scanning studies, which show that different types of memory are stored in different parts of the brain. This gives the model high scientific credibility. It also has real-world applications, for example in helping people with amnesia (they may not be able to remember facts, but they can still learn new skills).
• Weaknesses: The interaction between the different LTM stores is not fully explained. As discussed last session, episodic and semantic memories are often intertwined. It may be that the distinction is not as clear-cut as the model suggests, and the systems are more interdependent.

Narrative Explanatory Content

11.1 The Search for the Engram: Where is Memory?

For centuries, scientists have searched for the 'engram'—the physical trace of a memory in the brain. We now know that memory is not located in one single spot, but is distributed across a network of brain regions. However, certain structures play critical roles in specific types of memory.

• The Hippocampus: Located in the temporal lobe, this is arguably the most important structure for forming new explicit (episodic and semantic) memories. It acts as a 'gateway' or 'librarian', consolidating information from working memory into long-term storage in other parts of the cortex.
• The Prefrontal Cortex: This area at the front of the brain is heavily involved in working memory (the Central Executive) and in retrieving both episodic and semantic memories.
• The Cerebellum: Located at the back of the brain, this is crucial for procedural memory—the learning of motor skills.
• The Amygdala: This structure is vital for processing the emotional component of memories, particularly fear.

11.2 The Role of the Hippocampus: Lessons from HM

The case of Henry Molaison (HM) provided the first definitive evidence of the hippocampus's role. The surgical removal of his hippocampus left him unable to form new long-term declarative memories. This demonstrated that the hippocampus is not where memories are permanently stored, but it is essential for the consolidation process—transferring memories from a fragile state into a permanent one in the cortex. His ability to learn the mirror-drawing task showed that the hippocampus is not needed for procedural memory.

11.3 How Neurons Learn: Long-Term Potentiation (LTP)

If memory is a network, how does that network physically change when we learn something? The leading theory is Long-Term Potentiation (LTP). This is a long-lasting strengthening of the synaptic connections between neurons.

The principle was famously summarised by Donald Hebb: "Neurons that fire together, wire together."

• When you first learn something, a specific pattern of neurons fires.
• If this pattern is repeated (through rehearsal or use), the synaptic connections between these neurons become stronger and more efficient. The pre-synaptic neuron becomes more likely to release neurotransmitters, and the post-synaptic neuron becomes more sensitive to them (e.g., by growing more receptors).
• This makes it more likely that this specific pattern of neurons will fire again in the future. This strengthened pathway is the physical basis of the memory trace (the engram).

11.4 Analogy: Pathways in a Forest

Imagine trying to walk through a dense forest. The first time, it's very difficult and you have to forge a new path. The second time you walk the same route, it's a little easier. If you walk that same path every day, it will become a well-trodden, clear, and efficient walkway. LTP is the neural equivalent of creating this well-trodden path. The more you use a neural circuit, the stronger and more efficient it becomes.

Narrative Explanatory Content

12.1 Introduction to Interference Theory

Why do we forget things? Interference theory suggests that forgetting occurs because one memory blocks or disrupts another, causing one or both memories to be distorted or forgotten. It proposes that forgetting from LTM is most likely to occur when two memories are similar. This theory suggests that the memory is still stored, but it has become temporarily inaccessible.

There are two types of interference:

12.2 Proactive Interference (PI): The Past Disrupts the Future

Proactive Interference occurs when old information stored in LTM interferes with the learning and recall of new information. The prefix 'pro' means 'forward', so you can think of it as an old memory moving forward to disrupt a new one.

• Example 1: You get a new phone number, but you keep accidentally giving out your old one. The old number is proactively interfering with the new one.
• Example 2: You learn Spanish, then you try to learn Italian. You keep using Spanish words by mistake.

12.3 Retroactive Interference (RI): The Future Disrupts the Past

Retroactive Interference occurs when the learning of new information interferes with the recall of old information. The prefix 'retro' means 'backward', so you can think of it as a new memory moving backward to disrupt an old one.

• Example 1: You get a new phone number and learn it well. Now, you find it difficult to remember your old phone number. The new number has retroactively interfered with the old one.
• Example 2: A teacher learns the names of her new students this year and now has trouble remembering the names of students from last year.

12.4 Evidence and Evaluation

Much of the evidence for interference comes from artificial lab studies where participants learn lists of words (e.g., McGeoch & McDonald, 1931, found that interference is strongest when memories are similar). While these studies provide good control, they lack ecological validity.

A more naturalistic study was conducted by Baddeley & Hitch (1977):

• They asked rugby players to recall the names of the teams they had played against during a season.
• The time since the last game was the same for all players, but the number of intervening games they had played was different (due to injury, etc.).
• They found that recall was not affected by how long ago the games were, but by the number of games played in between. Players who had played more games forgot more team names. This is a real-world demonstration of retroactive interference.

Evaluation: Interference theory is a strong explanation for forgetting, especially when memories are similar. However, it is not a complete explanation. It doesn't explain why we forget things when there is no obvious interfering memory. It also struggles to explain why some memories seem to be forgotten and then spontaneously recover later.

Narrative Explanatory Content

13.1 Retrieval Failure and the Encoding Specificity Principle

Have you ever had a memory on the "tip of your tongue"? You know the information is in your LTM, but you just can't access it. This is an example of retrieval failure. This theory suggests that forgetting occurs when we lack the necessary cues to access a memory. The memory is available, but not accessible.

This is based on the Encoding Specificity Principle (ESP), proposed by Tulving (1983). The ESP states that a cue will be most effective in helping us remember if it was present when we encoded the memory. In other words, memory is most effective when information available at encoding is also available at retrieval.

There are two main types of cue-dependent forgetting:

13.2 Context-Dependent Forgetting

This occurs when the external environment at retrieval is different from the external environment at encoding. The environmental cues are missing.

• Example: You go downstairs to get something, but when you get there, you forget what you came for. You go back upstairs to the room where you had the original thought, and you suddenly remember. The room itself acted as a context cue.
• Godden & Baddeley (1975) - The Diver Study: This is the classic study. They had deep-sea divers learn a list of words either on land or underwater. They were then asked to recall the words either on land or underwater.

  Godden & Baddeley (1975) Results

  • Learn on land, recall on land = Good recall
  • Learn underwater, recall underwater = Good recall
  • Learn on land, recall underwater = Poor recall
  • Learn underwater, recall on land = Poor recall

  The results showed that recall was significantly better when the recall context matched the learning context. This is strong evidence for context-dependent forgetting.

13.3 State-Dependent Forgetting

This occurs when your internal state (e.g., your mood, or whether you are drunk or sober) at retrieval is different from your internal state at encoding. The internal physiological or psychological cues are missing.

• Example: If you revise for an exam while drinking lots of coffee, you might recall the information better if you are also caffeinated during the exam.
• Goodwin et al. (1969): They asked male volunteers to remember a list of words when they were either drunk or sober. At recall (24 hours later), some were drunk and some were sober. They found that information learned when drunk was more available when in the same state later.

13.4 Evaluation of Retrieval Failure Theory

Retrieval failure is considered one of the most important explanations for forgetting from LTM.

• Strengths: It has a huge amount of research support, from highly controlled lab studies to more realistic studies like Godden & Baddeley's. It also has significant real-world applications. For example, the Cognitive Interview uses principles of cue-dependency to improve eyewitness recall by encouraging them to 'mentally reinstate the context'.
• Weaknesses: The effects of context are not that strong in real life. Baddeley (1997) argued that the contexts have to be extremely different (like on land vs. underwater) for the effects to be seen. Simply being in a different room is unlikely to cause significant forgetting. Also, like LOP theory, the ESP can be criticised for being circular. If a cue leads to recall, it must have been encoded. If it doesn't, it must not have been. It's impossible to test this principle without assuming it's true.

Narrative Explanatory Content

14.1 Eyewitness Testimony and Reconstructive Memory

Eyewitness testimony (EWT) is the evidence given in court by someone who has witnessed a crime. For a long time, it was considered highly reliable. However, research since the 1970s, pioneered by Elizabeth Loftus, has shown that memory is not like a video recording. It is a reconstructive process. When we recall an event, we don't just play it back; we actively rebuild it. This reconstruction can be influenced by our expectations, beliefs, and information we receive after the event.

One of the most powerful factors that can distort EWT is misleading information, which can come in the form of leading questions.

14.2 Leading Questions: Loftus & Palmer (1974)

A leading question is one that, because of its phrasing, suggests a certain answer. For example, "Was the knife in his left hand?" suggests the person was holding a knife.

Loftus & Palmer (1974) - Experiment 1:

• Procedure: 45 students were shown video clips of car accidents. They were then asked a critical question: "About how fast were the cars going when they ___ each other?". The verb was changed for different groups.
• The Verbs: Smashed, collided, bumped, hit, contacted.
• Findings: The verb used had a significant effect on the speed estimate.

  Mean Speed Estimates (mph)

  • Smashed: 40.5 mph
  • Collided: 39.3 mph
  • Bumped: 38.1 mph
  • Hit: 34.0 mph
  • Contacted: 31.8 mph

14.3 Response-Bias vs. Substitution Explanation

The results of Experiment 1 could be explained in two ways:

1. Response-Bias Explanation: The verb doesn't change the memory, it just influences how the witness answers. 'Smashed' encourages a higher estimate.
2. Substitution Explanation: The verb actually changes the memory itself. The word 'smashed' causes the witness to remember the accident as being more severe than it was.

Loftus & Palmer (1974) - Experiment 2: To test which explanation was correct, they did a second experiment.

• Procedure: 150 students were shown a clip of a multi-car accident. They were split into three groups and asked the critical question with the verb 'smashed', 'hit', or no question at all (a control group). One week later, they returned and were asked "Did you see any broken glass?". There was no broken glass in the video.
• Findings: Significantly more participants in the 'smashed' condition (32%) reported seeing broken glass than in the 'hit' (14%) or control (12%) conditions.

This supports the substitution explanation. The leading question had actually changed their memory of the event.

14.4 Evaluation of Research on Misleading Information

• Strengths: This research has had a huge positive impact on the real world. It has highlighted the dangers of leading questions and has led to major changes in how police interview witnesses. The findings are reliable and have been replicated many times.
• Weaknesses (Ecological Validity): The use of students watching video clips in a lab is very different from witnessing a real crime. The emotional stress and consequences are missing. Some argue that in real life, witnesses might be more resistant to misleading information.
• Weaknesses (Demand Characteristics): Participants in a lab study may figure out what is expected of them and change their answers to please the experimenter. This could be an alternative explanation for the results.

Narrative Explanatory Content

15.1 Post-Event Discussion (PED)

When a crime occurs, it is common for witnesses to talk to each other before the police arrive. This is called post-event discussion (PED). While this may seem harmless, research shows it can be a powerful source of memory contamination. Witnesses may inadvertently incorporate information they have heard from others into their own memory of the event.

15.2 Evidence for PED: Gabbert et al. (2003)

Fiona Gabbert and her colleagues conducted a classic study to investigate the effects of PED.

• Procedure: Participants were put in pairs. Each partner watched a video of the same crime, but filmed from a different point of view. This meant that each partner could see elements in the event that the other could not. For example, only one partner could see the title of a book being carried by a woman. Both partners then discussed what they had seen before individually completing a test of recall.
• Findings: The results were striking. 71% of the participants mistakenly recalled aspects of the event that they did not see in their video but had picked up in the discussion. In a control group, where there was no discussion, the figure was 0%.
• Conclusion: This shows that witnesses will often go along with each other, either to win social approval or because they believe the other witness is right and they are wrong. Gabbert et al. called this phenomenon memory conformity.

15.3 Explaining the Effects of PED

There are two main explanations for why PED affects memory:

1. Memory Conformity: As Gabbert suggests, this is a social influence explanation. Witnesses conform to the memories of others because they want to be liked (normative social influence) or because they believe the others are more accurate (informational social influence). This may lead to a change in their public statement, but not necessarily their private memory.
2. Source Monitoring Error: This is a cognitive explanation. The witness genuinely incorporates the information from the other person into their own memory. They remember the information, but they forget the source of that information (i.e., they think they saw it, but they actually heard it). This is a genuine memory distortion.

15.4 Evaluation of Research on PED

• Strengths: Like the research on leading questions, this has important real-world applications. It demonstrates the importance of police separating witnesses as soon as possible after a crime to prevent contamination of their testimony.
• Weaknesses (Ecological Validity): The Gabbert study involved watching a staged crime on video. This is very different from a real event. Foster et al. (1994) argue that when a witness thinks they are watching a real robbery, and their testimony will have real consequences, their recall is more accurate. This suggests that the lab studies may overestimate the effect of PED in the real world.
• Individual Differences: Research has shown that age can be a factor. Older people seem to be more prone to source monitoring errors and may be more susceptible to the effects of PED than younger people.

Narrative Explanatory Content

16.1 Anxiety's Negative Effect: The Weapon Focus Effect

Witnessing a crime, especially a violent one, is a highly stressful and anxiety-provoking experience. Does this high anxiety make memory better or worse? Some research suggests it has a negative effect.

The weapon focus effect is the finding that the presence of a weapon draws attention away from other details, like the perpetrator's face, leading to poorer recall for those details.

• Johnson & Scott (1976):
  • Procedure: Participants were led to believe they were taking part in a lab study. While seated in a waiting room, they heard an argument in the next room. In the 'low-anxiety' condition, a man then walked through the waiting area carrying a pen with grease on his hands. In the 'high-anxiety' condition, they heard breaking glass and the man walked out holding a paper knife covered in blood.
  • Findings: Participants were later asked to identify the man from a set of 50 photos. In the low-anxiety (pen) condition, 49% were able to identify him. In the high-anxiety (knife) condition, only 33% could identify him.
  • Conclusion: The tunnel theory of memory suggests that the witness's attention narrows to focus on the weapon, because it is a source of anxiety.

16.2 Anxiety's Positive Effect: The Fight-or-Flight Response

In contrast, other research suggests that anxiety can actually enhance memory. The stress of witnessing a crime triggers the fight-or-flight response, which increases our alertness and may improve our memory for the event because it could be important for survival.

• Yuille & Cutshall (1986):
  • Procedure: They conducted a study of a real-life shooting in a gun shop in Canada. The shop owner shot a thief dead. There were 21 witnesses, 13 of whom agreed to participate. They were interviewed 4-5 months after the event and their accounts were compared to the original police interviews. The witnesses were also asked to rate how stressed they had felt at the time.
  • Findings: The witnesses were highly accurate in their accounts, and there was little change in their testimony after 5 months. Those who reported the highest levels of stress were the most accurate (about 88% compared to 75% for the less-stressed group).
  • Conclusion: This real-world study suggests that anxiety can have a positive effect on the accuracy of EWT.

16.3 Reconciling the Contradiction: The Yerkes-Dodson Law

So how can anxiety have both a positive and a negative effect? The Yerkes-Dodson Law (1908) provides an explanation. It suggests that the relationship between performance (in this case, memory recall) and arousal/anxiety is curvilinear, not linear. It looks like an inverted 'U'.

16.4 Evaluation of Research on Anxiety and EWT

• Ethical Issues: Creating anxiety in participants for research purposes is controversial. This is why studies like Johnson & Scott's are carefully designed to be as ethical as possible, but it raises questions about psychological harm.
• Alternative Explanations for Weapon Focus: It may not be anxiety that causes the weapon focus effect, but surprise. Pickel (1998) conducted an experiment using unusual objects (like a raw chicken) in a hairdressing salon. Eyewitness accuracy was poorer for the unusual objects. This suggests that it is the unusualness of the weapon, not the anxiety it creates, that draws attention.
• Problems with Field Studies: Real-life studies like Yuille & Cutshall's are high in ecological validity but lack control. The witnesses who experienced the highest stress were also closest to the event, which could be an extraneous variable. Also, post-event discussion could have influenced their memories.

Narrative Explanatory Content

17.1 The Problem with the Standard Police Interview

Given the research on memory fallibility, psychologists became concerned about the methods used by police to interview witnesses. The standard police interview often involved:

• Frequent interruptions.
• Closed, short-answer questions.
• Questions that were not in a logical order.
• Unintentional use of leading questions.

Fisher & Geiselman (1992) argued that these techniques were not compatible with how human memory works and could actually contaminate a witness's memory. In response, they developed the Cognitive Interview (CI), a method based on psychological principles of memory.

17.2 The Four Techniques of the Cognitive Interview

The original CI is based on four main techniques designed to increase the amount of accurate information recalled.

17.3 The Enhanced Cognitive Interview (ECI)

Fisher et al. (1987) later developed the Enhanced Cognitive Interview (ECI). This added some extra social elements to the original CI to improve communication between the interviewer and the witness. These include:

• Minimising distractions.
• Actively listening to the witness.
• Asking open-ended questions.
• Allowing the witness to pause and think.
• Avoiding interruptions.

The ECI is designed to reduce witness anxiety and build rapport, which is thought to improve the accuracy and quantity of information recalled.

17.4 Preliminary Evaluation of the CI

The CI is one of the most successful applications of cognitive psychology.

• Effectiveness: A meta-analysis by Köhnken et al. (1999) combined data from 50 studies and found that the ECI consistently provided more correct information than the standard police interview. It is now used by police forces around the world.
• Practical Issues: Despite its effectiveness, the CI is not always used properly. It is time-consuming to conduct and requires special training, which can be expensive. Some police forces may not have the resources to implement it fully.
• Increase in Inaccurate Information: The Köhnken et al. meta-analysis also found that while the CI increases the amount of correct information, it also slightly increases the amount of incorrect information recalled. However, the increase in correct information is much greater.

Narrative Explanatory Content

18.1 The Journey Through Memory

This module has taken us on a journey through the psychology of memory. We started with the foundational models of memory (MSM, WMM), which provided a theoretical architecture. We then broke down LTM into different types (episodic, semantic, procedural), showing that it is not a single store. We explored the biological basis of these systems in the brain. Next, we examined why memory fails, looking at explanations for forgetting (interference, retrieval failure). Finally, we applied this knowledge to the crucial real-world domain of eyewitness testimony, looking at factors that distort it (misleading information, anxiety) and how we can improve it (the Cognitive Interview).

18.2 A Toolkit for Critical Evaluation: GRAVE

A key skill in psychology is not just knowing the theories and studies, but being able to critically evaluate them. A useful mnemonic for evaluation is GRAVE:

18.3 The Role of Case Studies vs. Lab Experiments

Throughout this module, we have seen two main types of research methodology:

• Case Studies (e.g., HM, Clive Wearing): These provide rich, in-depth data and can study unique situations that would be unethical to create in a lab. They are vital for showing what can happen when a part of the brain is damaged. However, the findings are difficult to generalise because the individual is unique.
• Laboratory Experiments (e.g., Loftus & Palmer, Peterson & Peterson): These allow for high levels of control, establishing cause and effect and ensuring reliability. However, they often use artificial tasks and environments, which lowers their ecological validity.

The best psychological theories are those that are supported by evidence from a range of different methodologies.

18.4 Key Debates in Memory Research

As you prepare for your assignments, think about the key debates that have run through this module:

• Is memory a single entity or multiple systems? (Unitary vs. Multi-component)
• Is memory a passive recording or an active reconstruction?
• Is forgetting a failure of storage or a failure of retrieval?
• How do we balance the need for controlled research with the need for real-world relevance? (Internal vs. Ecological Validity)

Narrative Explanatory Content

19.1 Deconstructing the Assignment Question

The first step to a successful essay is to fully understand what the question is asking. Let's break down the formative assignment task.

The core of this question is comparison and evaluation. You need to show that you understand both models and, more importantly, that you understand the relationship between them—specifically, how the WMM is a more detailed and sophisticated elaboration of the MSM's concept of short-term memory.

19.2 Structuring Your Comparative Essay

A clear structure is vital for a comparative essay. Avoid simply describing one model and then the other. The best essays integrate the comparison throughout.

• Introduction (approx. 75 words): Briefly introduce memory models. State that the MSM was a pioneering model but was criticised for its oversimplified view of STM. State your thesis: that the WMM offers a more detailed and plausible account of short-term memory, addressing the key limitations of the MSM.
• Main Body Paragraph 1: The MSM's View of STM (approx. 200 words):
  • Briefly describe the MSM's view of STM as a single, unitary store that functions primarily for storage and rehearsal.
  • Introduce a key limitation: The MSM cannot explain how we can perform two different tasks at once (dual-task performance).
  • Use evidence, such as the case of KF (Shallice & Warrington, 1970), who had poor auditory STM but normal visual STM, to challenge the idea of a single STM store.
• Main Body Paragraph 2: The WMM's Solution (approx. 200 words):
  • Introduce the WMM as an alternative model of STM.
  • Describe its key components: the Central Executive, Phonological Loop, and Visuo-Spatial Sketchpad.
  • Explain how this multi-component view directly addresses the limitations of the MSM. For example, dual-task performance is possible if the tasks use different slave systems (e.g., a visual task and a verbal task). KF's case can be explained by damage to the phonological loop but an intact visuo-spatial sketchpad.
• Main Body Paragraph 3: Further Comparison and Evaluation (approx. 150 words):
  • Compare the role of rehearsal. The MSM sees it simply as maintenance, while the WMM (specifically the phonological loop) sees it as an active process.
  • Evaluate the WMM. A key strength is the research support from dual-task studies (e.g., Baddeley & Hitch, 1976). A key weakness is the vague nature of the Central Executive.
• Conclusion (approx. 75 words): Summarise the main points. Reiterate that while the MSM was a crucial first step, the WMM provides a more comprehensive and evidence-supported model of the complex processes involved in short-term memory.

19.3 How to Write an Evaluative Paragraph (PEEL)

To get high marks, your evaluation must be structured. Use the PEEL method:

Narrative Explanatory Content

20.1 Deconstructing the Summative Assignment Question

The summative assignment asks you to take a position on a fundamental debate in memory research and defend it with evidence. This requires more than just description; it requires a sustained argument.

This question has two parts: 1) Is forgetting a necessary function? 2) Is memory distortion a bigger problem than forgetting? A good essay will address both parts within a single, coherent thesis.

20.2 Developing a Thesis and Structuring Your Argument

Your thesis is your answer to the question. There is no single "right" answer; what matters is how well you support your position.

Example Thesis (Refuting the statement): "While memory distortion poses a significant challenge, particularly in legal contexts, this essay will argue that the complete and permanent loss of information through processes like interference represents a more fundamental failure of the memory system than its susceptibility to reconstruction. Forgetting is not a 'function' but a system breakdown, whereas distortion is a byproduct of an otherwise adaptive reconstructive process."

Example Thesis (Defending the statement): "This essay will defend the view that forgetting is an adaptive function that prevents cognitive overload, arguing that the real danger lies in the unreliability of memory, as demonstrated by research into eyewitness testimony, where distortion can have devastating real-world consequences."

Suggested Essay Structure:

• Introduction (approx. 200 words): Introduce the debate about the nature of forgetting and memory fallibility. State your clear thesis. Outline your line of argument.
• Argument Part 1: The Nature of Forgetting (approx. 600 words):
  • Analyse explanations for forgetting like Interference Theory (proactive/retroactive) and Retrieval Failure (cue-dependency).
  • Argue whether these represent a "necessary function" (e.g., interference clears out old, irrelevant information) or a "failure" (e.g., retrieval failure shows the information is there but inaccessible).
  • Use research like Baddeley & Hitch (retroactive interference) and Godden & Baddeley (context-dependent forgetting) as evidence.
• Argument Part 2: The Problem of Distortion (approx. 600 words):
  • Analyse factors affecting eyewitness testimony. Focus on misleading information, citing Loftus & Palmer's (1974) car crash study. Explain how leading questions can distort memory.
  • Analyse post-event discussion, citing Gabbert et al. (2003). Explain how memory conformity can occur.
  • Critically evaluate this research (e.g., artificiality, demand characteristics).
• Argument Part 3: Synthesis and Nuance (approx. 400 words):
  • Directly compare the two problems. Argue which is "bigger". For example, you could argue that while forgetting is personally frustrating, memory distortion has greater societal consequences (wrongful convictions).
  • Discuss the role of anxiety (weapon focus effect) as a factor that can cause both forgetting (of peripheral details) and distortion (of central details).
  • This section is where you bring your arguments together to support your thesis.
• Conclusion (approx. 200 words): Summarise your main arguments. Restate your thesis in light of the evidence discussed. Offer a final, thoughtful conclusion on the nature of memory's fallibility.

20.3 Achieving a Distinction: The Art of Critical Analysis

To achieve the highest marks, your essay must be consistently analytical and evaluative. Every piece of evidence should be used to serve your argument.