Working memory allows us to remember a phone number just long enough to dial it or to keep track of the steps in a recipe. This cognitive function enables us to navigate through sequences of items and events effortlessly, creating an instinctive awareness of what comes next. Despite seeming automatic, our brains engage in a delicate balancing act, maintaining a few items in mind while too tracking the timing of each occurrence. How does this intricate system function?
Traditional models of working memory largely emphasize persistent neural activity or gradual learning through repetition. While these models are effective at explaining what information is stored, they fail to clarify how we can successfully recall the order of items or the timing of events, particularly in novel sequences. Research indicates that individuals can recall new sequences, including their order and timing, without any prior practice. This suggests that working memory must be capable of storing temporal information in real-time, raising the question: where is this information stored within the brain?
Recent research by Gianluigi Mongillo and Misha Tsodyks proposes that temporal information is directly encoded in the dynamics of the synapses connecting neurons. Their findings suggest that rather than storing temporal information separately, fleeting changes in synaptic signaling may embed a sense of time within the memory itself. This innovative approach builds on their previous operate, which established a synaptic theory of working memory.
Understanding Synaptic Augmentation
Mongillo and Tsodyks expand their synaptic theory by incorporating a mechanism known as synaptic augmentation, a form of short-term plasticity that develops gradually and can persist for tens of seconds. In contrast to classical facilitation and depression—forms of short-term plasticity that occur over much shorter timescales—synaptic augmentation allows synapses to strengthen gradually. When a neuron fires repeatedly during a sequence, its outgoing synapses not only transmit signals but also increase in strength, recording how recently and how often they have been active. This leads to a time-dependent gradient across synapses, effectively encoding the timing of each item in a sequence.
For instance, consider a sequence of items represented by different neural populations. As each item is presented, the synaptic strengths in the corresponding neurons increase. Because this augmentation builds up and decays slowly, the resulting pattern retains a temporal fingerprint of the sequence, with earlier items exhibiting stronger augmentation than those presented later. This gradient allows working memory to replay the sequence in the correct order and approximate timing, whether at real speed or in a time-compressed manner.
Dynamic Memory Representations
This theory resonates with findings from experimental studies indicating that working memory activity is dynamic rather than static. Memory representations can fluctuate in strength over time, can be briefly reactivated, and may even be replayed during sleep or rest, often in a time-compressed format believed to aid learning and memory consolidation. Importantly, Mongillo and Tsodyks argue that without synaptic augmentation, the system tends to stabilize, causing all past items to be maintained equally and erasing any temporal distinctions.
Critically, their model offers testable predictions regarding how temporal information is encoded in working memory. The gradual buildup and decay of synaptic augmentation aligns with behavioral observations that working memory can sustain temporal intervals of several seconds without active rehearsal. The researchers predict that selectively disrupting mechanisms of augmentation would impair memory concerning the order and timing of items, while leaving the memory of the items themselves intact.
Neural Activity Patterns and Implications
The model proposed by Mongillo and Tsodyks, who are affiliated with prestigious institutions including the Institute for Advanced Study in Princeton and Sorbonne University, provides a framework for interpreting patterns of neural activity observed in electrophysiological studies. It accounts for ramping activity during memory delays and the so-called “activity-silent” states where latent information can be reactivated by cues. By linking these dynamic patterns to underlying synaptic changes, the model bridges the gap between observable neural activity and the synaptic processes underpinning working memory.
their work reframes a long-standing question in cognitive neuroscience: time is not merely an addition to memory but an emergent characteristic of how synapses dynamically change. The brain does not require a separate “clock” to timestamp experiences; instead, it utilizes its own plasticity mechanisms to leave a temporal imprint on memory. These insights will continue to influence our understanding of cognition as an active, evolving process shaped by brain rhythms and gradual changes in neuron connections.
As research progresses, the implications of this model could have far-reaching effects on how we understand memory, and cognition. Future studies may delve into the potential applications of these findings in addressing memory-related disorders or enhancing cognitive functions. Engaging in discussions around these advancements can foster better awareness and understanding of the intricate workings of our cognitive processes.
What are your thoughts on the relationship between time and memory? Share your views in the comments, and consider sharing this article to spark further discussion.
