Going Down Memory Lane
When we look back at early research on memory in psychology, we find that memory was once conceptualized as consisting of three separate yet unitary stores: the sensory register, short-term memory, and long-term memory. This framework, known as the Multi-Store Model of Memory, was proposed by Atkinson and Shiffrin (1968) and became one of the most influential theories in cognitive psychology (Mulyono et al., 2023).
According to the Multi-Store Model, the sensory register receives information from the sensory organs. It possesses an extremely large capacity but retains information for less than two seconds. Information that receives sufficient attention is transferred to short-term memory, where it is primarily encoded acoustically. Short-term memory has a limited capacity of approximately 5–9 items and a duration of around 18–30 seconds.
Information that is rehearsed successfully is then transferred into long-term memory, where it is primarily encoded semantically. Unlike the previous stores, long-term memory is believed to have an essentially unlimited capacity and potentially lifelong duration (Atkinson & Shiffrin, 1968).
Although highly influential, the Multi-Store Model treated long-term memory as a single, unified storage system. Given the enormous amount and diversity of information humans retain throughout life, psychologists began questioning whether long-term memory could truly function as one homogeneous system.
This question laid the foundation for a new generation of memory research.
Today, cognitive psychology recognizes that long-term memory is far more complex than originally proposed. Rather than functioning as a single storage system, long-term memory consists of several distinct yet interconnected subsystems. Among the most widely accepted distinctions are episodic, semantic, and procedural memory.
- Episodic memory stores personally experienced events, such as remembering your last birthday or your most recent visit to the dentist.
- Semantic memory contains general knowledge and factual information, such as knowing that Paris is the capital of France.
- Procedural memory is responsible for learned skills and habits, including activities such as riding a bicycle, typing, or driving a car.
These memory systems differ not only in what they store but also in how they operate.
Both episodic and semantic memory are explicit (declarative) memory systems, meaning they require conscious retrieval (Riedel & Blokland, 2015). Procedural memory, however, belongs to the implicit memory system, allowing learned skills to be performed automatically without conscious awareness.
What Was the Aim of Tulving’s Gold Study?
One of the most influential studies supporting the distinction between different types of long-term memory was Endel Tulving’s Gold Study conducted in 1989.
The primary aim of the study was to investigate whether episodic and semantic memories activate different regions of the brain, thereby providing neurobiological evidence that long-term memory is composed of multiple systems rather than a single unified store.
To examine this question, Tulving recruited six participants who were injected with a small amount of radioactive gold. As the tracer circulated through the bloodstream and reached the brain, positron emission tomography (PET) was used to measure changes in cerebral blood flow. Because increased blood flow reflects increased neural activity, more active brain regions appeared as brighter areas on the PET images.
The study compared brain activity during episodic memory retrieval, in which participants recalled personal experiences, with activity during semantic memory retrieval, in which they recalled factual or general knowledge (Tulving, 1989).
Methodology and Results
Participants completed eight experimental trials, consisting of four episodic and four semantic memory tasks presented in random order.
At the beginning of each trial, participants received a cue prompting them to retrieve either a personal memory or factual knowledge while their brain activity was simultaneously recorded using PET imaging.
The findings revealed noticeable differences in patterns of brain activation.
For three participants, episodic and semantic retrieval consistently activated different brain regions. However, the remaining three participants produced inconsistent activation patterns and were therefore excluded from the final analysis.
Among the participants whose data were retained, recalling episodic memories produced greater activation within the frontal and temporal lobes, whereas retrieving semantic memories produced stronger activation in the parietal and occipital lobes (Tulving, 1989).
Conclusion
Based on these findings, Tulving concluded that episodic and semantic memories rely on different neural systems and are likely stored and processed within distinct regions of the brain.
Although the study did not address procedural memory directly, it provided important neurobiological support for the idea that long-term memory is not a single, unitary system but instead consists of multiple specialized memory systems.
Tulving’s work represented one of the earliest demonstrations that different forms of remembering are associated with different patterns of brain activity, helping establish one of the central principles of modern cognitive neuroscience.
Criticisms
Despite its considerable influence, Tulving’s Gold Study has received several important criticisms.
The most significant limitation concerns the extremely small sample size. Only six participants took part in the study, and because half of them produced inconsistent results, the final conclusions were based on data from only three individuals. Such a small sample substantially limits the generalizability of the findings, making it difficult to determine whether similar activation patterns would be observed in larger and more diverse populations.
A second limitation involves experimental control.
Participants were instructed to think about either episodic or semantic memories while undergoing PET scanning. However, researchers had no direct way of verifying exactly what participants were thinking during the scanning period. It is therefore possible that some participants unintentionally shifted their attention, recalled unrelated memories, or became distracted.
If this occurred, the PET scans may have captured brain activity unrelated to the intended memory task, thereby reducing the internal validity of the study.
Finally, PET technology available during the late 1980s offered relatively low temporal and spatial resolution compared with modern neuroimaging techniques such as functional magnetic resonance imaging (fMRI). Contemporary research has therefore expanded upon Tulving’s findings using more advanced imaging methods, generally providing stronger evidence for the existence of multiple long-term memory systems.
Overall Evaluation
Despite its methodological limitations, Tulving’s Gold Study remains one of the landmark investigations in cognitive neuroscience. It challenged the assumption that long-term memory functions as a single storage system and provided early neurobiological evidence supporting the distinction between episodic and semantic memory.
Subsequent research employing larger samples and more sophisticated neuroimaging techniques has largely supported Tulving’s central proposition that different forms of long-term memory rely on partially distinct neural networks.
Consequently, Tulving’s work continues to play a foundational role in contemporary theories of memory and remains one of the most influential studies in the history of cognitive psychology.
References
Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In K. W. Spence & J. T. Spence (Eds.), The Psychology of Learning and Motivation (Vol. 2, pp. 89–195). Academic Press. https://doi.org/10.1016/S0079-7421(08)60422-3
Mulyono, Y. O., Cabrera, E., Sukhbaatar, U., Cabrera, L., & Cabrera, D. (2023). A systemic lens on the multi-store model/modal model. Journal of Systems Thinking, 3(1), 1–6. https://doi.org/10.54120/jost.0000035
Riedel, W. J., & Blokland, A. (2015). Declarative memory. Handbook of Experimental Pharmacology, 228, 215–236. https://doi.org/10.1007/978-3-319-16522-6_7
Tulving, E. (1989). Memory: Performance, knowledge, and experience. European Journal of Cognitive Psychology, 1(1), 3–26. https://doi.org/10.1080/09541448908403069