Tulving's Gold Memory Study

Endel Tulving is one of the most influential cognitive neuroscientists in the study of memory. His research using PET scans (Positron Emission Tomography) provided some of the first brain imaging evidence that different types of long-term memory are processed by different brain regions, supporting the distinction between episodic and semantic memory.

---

Background

Tulving had previously proposed (1985) that long-term memory consists of different types — episodic memory (personal events), semantic memory (facts and knowledge), and procedural memory (skills). He used brain imaging to test whether these different types of memory involve different brain regions.

---

Tulving's PET Scan Study (1989)

Aim

To investigate whether episodic and semantic memory are associated with activity in different brain regions.

Method

• Participants were injected with a radioactive tracer (gold-labelled radioactive material) that emits tiny amounts of radiation
• The tracer travels through the bloodstream to the brain, where active brain areas receive more blood flow and therefore more tracer
• A PET scanner detected the radiation, creating an image showing which brain areas were most active
• Participants performed tasks involving either:
  • Episodic memory — recalling personal experiences
  • Semantic memory — retrieving factual knowledge

Results

Task	Brain Region Most Active
Episodic memory encoding (forming new episodic memories)	Left prefrontal cortex
Episodic memory retrieval (recalling personal memories)	Right prefrontal cortex
Semantic memory retrieval (recalling facts)	Left prefrontal cortex

Key Finding: HERA Model

Based on his findings, Tulving proposed the HERA model (Hemispheric Encoding/Retrieval Asymmetry):

• Encoding of episodic memories primarily involves the left prefrontal cortex
• Retrieval of episodic memories primarily involves the right prefrontal cortex

This asymmetry (different hemispheres for encoding and retrieval) was an important discovery in understanding how memory works in the brain.

Tulving's HERA model

flowchart TD
    A[Memory task] --> B{Type of memory?}
    B -- Episodic --> C[Encoding]
    B -- Episodic --> D[Retrieval]
    B -- Semantic --> E[Retrieval]
    C --> F[Left prefrontal cortex]
    D --> G[Right prefrontal cortex]
    E --> H[Left prefrontal cortex]
    F --> I["HERA: Hemispheric<br/>Encoding/Retrieval Asymmetry"]
    G --> I
    H --> I

---

Significance

1. Different Types of LTM Use Different Brain Areas

Tulving's PET scan research provided biological evidence that episodic and semantic memory are distinct systems, not just theoretical constructs. This supported his earlier theoretical work on types of LTM.

2. Brain Imaging Technology

The study demonstrated the power of PET scanning as a tool for studying brain function. It allowed researchers to observe the brain in action during different cognitive tasks, rather than relying only on case studies of brain-damaged patients.

3. Linking Psychology and Neuroscience

Tulving's work was important in bridging the gap between cognitive psychology (which studies mental processes) and neuroscience (which studies the brain). By showing that psychological concepts (types of memory) correspond to distinct patterns of brain activity, he helped establish the field of cognitive neuroscience.

---

Evaluation

Strengths

• Provided objective, biological evidence for the distinction between episodic and semantic memory
• Used brain imaging technology (PET scans), which provides measurable, quantifiable data
• Supported the theoretical distinction between memory types that Tulving had proposed earlier
• The HERA model has been supported by subsequent research using fMRI

Weaknesses

• PET scans have poor temporal resolution — they show which areas are active over a period of time but cannot capture the rapid, moment-by-moment changes in brain activity
• PET scans involve exposure to radiation (from the radioactive tracer), which raises ethical concerns about participant safety
• The tasks used to test episodic and semantic memory in the lab may not perfectly represent how these memory systems are used in everyday life
• Correlation vs causation — showing that a brain area is active during a task does not prove it is responsible for that task
• Small samples — brain imaging studies often use small numbers of participants, limiting generalisability
• The distinction between episodic and semantic memory may not be as clear-cut in the brain as the study suggests — some brain regions are involved in both types

---

Key Points

• Tulving used PET scans to show that episodic and semantic memory involve different brain regions.
• Episodic memory encoding: left prefrontal cortex. Episodic retrieval: right prefrontal cortex.
• Semantic memory retrieval: left temporal lobe.
• The HERA model describes the hemispheric asymmetry in episodic memory encoding and retrieval.
• PET scans provide objective biological evidence but have limitations (radiation, poor temporal resolution).

---

Worked case study: the HM case (Henry Molaison, 1953)

The HM case is the most famous case study in the history of memory research and provides vital context for understanding why Tulving's separation of memory systems mattered.

Aim: To understand the role of the medial temporal lobe — specifically the hippocampus — in forming new long-term memories, following a surgical procedure that removed these structures.

Procedure: In 1953, 27-year-old Henry Molaison (known as "HM" until his death in 2008) underwent bilateral medial temporal lobectomy performed by neurosurgeon William Scoville to control severe epilepsy. The surgery removed most of his hippocampus and surrounding structures on both sides. Brenda Milner and Suzanne Corkin studied HM for over five decades using memory tests (digit span, mirror-drawing task, paired-associate learning) and clinical interviews.

Findings: HM's epilepsy improved, but he suffered profound anterograde amnesia — he could no longer form new long-term episodic or semantic memories. He could not remember people he met, what he had just eaten, or that his uncle had died (he re-experienced the grief each time he was told). However, two memory systems remained intact: his short-term memory (digit span normal) and his procedural memory — Milner showed he could learn the mirror-drawing task, improving with practice, yet had no conscious recollection of ever having done it before.

Conclusion: The hippocampus is essential for consolidating new explicit (episodic and semantic) long-term memories but not for short-term memory or procedural memory. Different types of memory therefore depend on different brain regions, directly supporting Tulving's multi-store view of long-term memory.

GRAVE evaluation: