Forgetting

Why would nature design such a flaw within an otherwise impressive information storage machinery? Why delete information, if, as we’ve previously described, long-term memory seems to hold virtually unlimited capacity?

The truth is, forgetting should not be viewed as a deficit, but rather as an ability. The ability to let go of information is, in many respects, as worthy as the ability to store it in the first place. Although it may not receive as much praise as memory, forgetting is an impressive process in itself, often necessary for our brains to function in a way we deem normal. ¹

In “The seven sins of memory: How the mind forgets and remembers”², Daniel Schachter divides the ways memory mechanisms may fail an otherwise normal brain in seven main categories:

1. Transience - the classical forgetting over time We forget considerably more than we remember. As far back as 1885, Herman Ebbinghaus made this a mathematical matter, describing what was further confirmed³, and we today refer to as “the forgetting curve.” It shows how we inevitably lose learned information with time. Ebbinghaus proved this result by testing himself, and his capacity to recall a set of non-sense syllables over time. He found that memory retention, starting from 100% at the time of learning, drops in a near-exponential manner down to 60% within one day, 40% in 10 days, after which the decline slows down and eventually plateaus. The results of this curve are what we’d experience if we went to an exam, having attended or read the course only once. However, another effect we are familiar with is that revising the information is significantly faster than learning it and leads to a much more attenuated curve, meaning we retain more over a longer period.⁴
2. Absent-mindedness - attention deficits that affect the way a memory formed in the first place. As we’ve described before in our article, attention is what drives the encoding of information into short and long-term memories, and the depth of processing specific information dictates how well that will be remembered.
3. Blocking - well stored information is temporarily inaccessible, on the ‘tip of the tongue’ This frustrating ‘glitch’ is often viewed as a memory engram that is intact, just out of reach at the moment. Often enough, external retrieval cues (such as the first letter of the word) may aid us to gain access and produce the stored information. Interestingly enough, the number of TOTs (tip-of-the-tongue experiences) seems to increase with age and associate with gray matter atrophy in the left insula, a relevant area for speech production.⁵ An imaging study showed simultaneous activation of both areas involved in storing a certain piece of information and areas involved in speech production during a TOT experience. This suggests that the foundation of this event may rather be a miscommunication between memory and speech, even if both processes seem to work fine individually.⁶
4. Suggestibility - altering existing memories through some form of deception and
5. Misattribution - producing memories that never took place In “Biological Time Bomb” (1968), a book that seemed to predict several biotechnological advances such as artificial insemination and organ transplants, Gordon R. Taylor mentioned that selective erasure of memories should have become possible by 1975, and transmission of memories by 2000. While we still lack the level of precision required in those claims, it appears we may have actually been altering each other’s memories for longer than we realize. The way Elizabeth Loftus, an influential figure in the field of false memories, puts it, the “misinformation effect” appears when people who witnessed an event are exposed to new, often misleading information, which leads them to alter their very memory. As we have seen before, remembering is, in itself, a constructive process, one in which the brain does not necessarily produce a perfect replica of the initial experience, but one that is coherent, fits the other memories and cues from the environment. Her research shows that, by specifically altering those cues, memories can be distorted or even created anew. The set of an accident, the identity of the perpetrator, and sometimes even the very existence of the accident can be altered in the victims’ memory just by the way an interrogation is performed, or through psychotherapeutic methods supposed to help retrieve ‘repressed’ memories. Blurring the line between repressed and false memories, putting into question the very validity of eyewitness testimonies employed in trials, she shows, sometimes through dramatic circumstances, just how susceptible to alteration the process of remembering actually is.^7,8,9 The concept of producing a non-existent memory may seem absurd at first. In a 1959 study, James Deese¹⁰ proves just how easy it actually is. He asks participants to recall somewhat related words – like, for example: red, green, brown and blue. Upon recall, many patients produce colors that were not actually there, like orange or black.¹¹ More common misattributions, that we come across daily, are misattributing the source of a memory – whether we got a certain information from the news, a friend, a documentary; misattributing a face to a wrong context; or, ultimately, misattributing an imagined event as reality. This is more than just a brain quirk, as it has been both proven experimentally and experienced by many.¹¹ In a large-scale study of people’s memories of 9/11 (‘flashbulb memories’, which seem particularly vivid due to the emotional content), people were found to not only forget or mistake more and more details as time passed, but there are several accounts of false memories among those who think they saw the crash as it was happening.^12,13
6. Bias - altering existing memories by someone’s own current beliefs. If suggestibility meant that others can alter our memories, by what we perceive they expect to hear from us, bias stands for the way in which we can alter our own memories. Our current knowledge and belief system are what we filter all information through, and memories make no exception. We may tweak and trim our older memories, slightly adjusting them to better fit what we currently think of the world and ourselves.¹⁴ Not only that, but it seems that bias works the other way around as well. Studies show that our decision-making process is highly influenced by how well and how quickly we remember an alternative from past choices, even if objectively it may not be the best in the current context. This peculiarity, a form of intrinsic trust in the efficiency and selectivity of our own memory systems, was associated with increased connectivity between the hippocampus and an area essential for decision-making within the prefrontal cortex.¹⁵
7. Persistence - things that people would want to forget, but seem to be unable to. Forgetting certain unwanted or painful memories (sometimes termed motivated forgetting) is a feat we rely on to function according to current circumstances and not dwell excessively on past ones. More than just a Dali painting, the Persistence of Memory can prove particularly distressful, as is the case for PTSD patients.

When we start questioning the mechanism behind forgetting, our instinctual understanding of it may make it feel like it should be an exclusively passive process, something that just ‘happens’ over time. For quite a while, that was the commonly accepted view in neuroscience as well. Forgetting was thought to be the result of memory engrams (neural circuits that store a piece of information) inevitably decaying over time. ¹

This is consistent with some of the most common observations about aging, such as the progressive volume loss in various areas of the brain. ¹⁶ The hippocampus, an essential center for converting short-term memories into long-term memories, is one of the most affected areas.¹⁷ Alongside it, the cerebral cortex suffers apparent alterations with age as well – it becomes thinner, and the sulci become deeper and more prominent.¹⁸ While this is the natural course of the aging brain, the process is very much accentuated in pathologies that are known to associate significant memory deficits, such as Alzheimer’s or dementia.¹⁹

At a cellular level, this shrinking translates to neurons also reducing their sizes, losing complexity in their dendritic arborizations, and suffering deterioration of the myelin sheath, ultimately reducing and altering the number of available synapses. ^16,20 Not only this, but various neurotransmitter levels also go down with age.^16,21 Overall, all the necessary structural and functional premises are there to support the fact that connectivity and synaptic plasticity are altered through aging. Since as we age, we tend to get more forgetful, it makes sense to suppose that these kinds of connectivity deficits are the main ones involved in losing engrams, and thus in the classical loss of memories over time.

However, as more and more recent evidence suggests, that is not the only way we forget. In fact, forgetting can sometimes prove to be surprisingly active in nature.

In this regard, another classical theory of forgetting mentions a rather active mechanism that may be at play – that of interference. Specifically, the concept that memories are not entirely isolated from one another, and, much like improperly separated electrical wires, they may alter one another’s conduction of information. There are two kinds of interference: retroactive interference, by which old memories may make us forget or alter newer ones, and proactive interference, by which newer memories cause us to lose old ones. This mechanism is usually thought to apply to similar memories, like confusing old and new passwords or trying to learn a new language that is very similar to another one you have previously learned – such as Spanish and Italian.²²

More recent advancement in the field of active forgetting was brought about in 2017 by Ronald Davis and Yi Zhong ²³. Intrinsic forgetting relies on a special subset of cells, the forgetting cells, to trim, erase or otherwise degrade existing engrams. These cells are supposed to release dopamine onto engram cells, mobilizing a particular signaling pathway that ultimately alters the neuron’s shape and synaptic behavior, essentially trying to bring it back to its pre-engram shape. The proposers of this theory raise the supposition that this memory-erasing process may be a default mechanism within the brain, competing with antagonistic ones such as memory consolidation to decide which memories stay with us and which don’t.

Yet another counter-intuitive cause for forgetting may be neurogenesis – the production of new neurons in the adult brain. Previous beliefs that neurogenesis stops at the adult age have been proven wrong. While the hippocampal volume may decrease with aging, studies suggest that it is not through the rate at which new neurons are generated. What seems to be altered is the regional blood flow and the neuroplasticity of this newly-generated cells.²⁴

Regardless of age, more neurogenesis may not necessarily mean better memories. In a set of experiments, Paul Frankland and his colleagues trained mice to do a certain task, after which they artificially increased the level of neurogenesis. Surprisingly enough, they found that mice with induced neurogenesis remembered how to perform the task considerably worse than their control counterparts. This led them to believe that the induced neurogenesis may have made it difficult for the recently stored engram to be recovered. Going on with an electronics analogy, Dr. Frankland explains:
“If you go in and start rewiring something, any information stored in that circuit might be degraded.” ²⁵

Ultimately, we have to ask, why go through all these complicated mechanisms for the intentional loss of information? After all, we don’t only lose bad memories, but rather all our memories seem to be affected by forgetting to some degree. In an analogy to artificial intelligence, Blake Richards, a researcher in neural circuits and machine learning, thinks of it as a way to prevent an effect called overfit – when a mathematical model becomes so familiar and good at predicting the data it has been programmed with, that it cannot properly predict new information. Similarly, if we didn’t generalize an experience and recalled its every detail, it may be harder for us to take a specific lesson away from it, or analyze and compare it with other similar experiences. ¹

It seems not only us, but mostly all species have achieved a form of equilibrium between memory consolidation and forgetting. We require the ability to forget to gain perspective, properly detach ourselves from the particular details of every passing day and only keep the very essence. There is no denying that this process fails us at times. Still, it is only through appreciating its complexity and importance that we can hope to understand, and, eventually, tackle it.

Resources:

1. Gravitz, L. The forgotten part of memory. Nature 571, S12–S14 (2019).
2. The seven sins of memory. Available at: link. (Accessed: 30th March 2020).
3. Murre, J. M. J. & Dros, J. Replication and Analysis of Ebbinghaus’ Forgetting Curve. PLoS One 10, e0120644 (2015).
4. The Forgetting Curve: The Reason to Review Your Notes. Available at: link. (Accessed: 31st March 2020).
5. Shafto, M. A., Burke, D. M., Stamatakis, E. A., Tam, P. P. & Tyler, L. K. On the tip-of-the-tongue: Neural correlates of increased word-finding failures in normal aging. J. Cogn. Neurosci. 19, 2060–2070 (2007).
6. It’s on the Tip of My Tongue: What Happens When We Can’t Remember a Name - Cognitive Neuroscience Society. Available at: link. (Accessed: 31st March 2020).
7. Creating False Memories. Available at: link. (Accessed: 31st March 2020).
8. Elizabeth Loftus: Falsifying memories | Mo Costandi | Science | The Guardian. Available at: link. (Accessed: 31st March 2020).
9. Evidence-based justice: Corrupted memory : Nature News & Comment. Available at: link. (Accessed: 31st March 2020).
10. Deese, J. On the prediction of occurrence of particular verbal intrusions in immediate recall. J. Exp. Psychol. 58, 17–22 (1959).
11. How Memories are Distorted and Invented: Misattribution - PsyBlog. Available at: link. (Accessed: 1st April 2020).
12. How Accurate Are Memories of 9/11? - Scientific American. Available at: link. (Accessed: 1st April 2020).
13. False Memories of 9/11 More Common Than You Think & Here’s Why. Available at: link. (Accessed: 1st April 2020).
14. List of memory biases - Wikipedia. Available at: link. (Accessed: 1st April 2020).
15. Mechera-Ostrovsky, T. & Gluth, S. Memory beliefs drive the memory bias on value-based decisions. Sci. Rep. 8, 1–10 (2018).
16. How the Brain Changes With Age. Available at: link. (Accessed: 31st March 2020).
17. O’Shea, A., Cohen, R. A., Porges, E. C., Nissim, N. R. & Woods, A. J. Cognitive aging and the hippocampus in older adults. Front. Aging Neurosci. 8, (2016).
18. Zheng, F. et al. Age-related changes in cortical and subcortical structures of healthy adult brains: A surface-based morphometry study. J. Magn. Reson. Imaging 49, 152–163 (2019).
19. Du, A.-T. et al. Different regional patterns of cortical thinning in Alzheimer’s disease and frontotemporal dementia. Brain 130, 1159–66 (2007).
20. Peters, R. Ageing and the brain. Postgraduate Medical Journal 82, 84–88 (2006).
21. Carlsson, A. Brain neurotransmitters in aging and dementia: Similar changes across diagnostic dementia groups. Gerontology 33, 159–167 (1987).
22. Proactive and Retroactive Interference | Simply Psychology. Available at: link. (Accessed: 31st March 2020).
23. Davis, R. L. & Zhong, Y. The Biology of Forgetting—A Perspective. Neuron 95, 490–503 (2017).
24. Boldrini, M. et al. Human Hippocampal Neurogenesis Persists throughout Aging. Cell Stem Cell 22, 589-599.e5 (2018).
25. To Remember, the Brain Must Actively Forget | Quanta Magazine. Available at: link. (Accessed: 1st April 2020)

‍