Why Do We Forget? The Science of Memory Decay and Forgetting

An hour after a lecture, you remember half the material presented. A week later, without review, you retain perhaps 10%. A year later, if the information was not used or reinforced, you have forgotten nearly all of it. This is not a reflection of poor intellect. It is a property of how memory works, documented with precision more than a century ago and confirmed repeatedly in modern neuroscience.

Forgetting is often experienced as a failure, a sign of cognitive weakness or age. But in evolutionary terms, forgetting is a feature, not a bug. A brain that retained every sensory impression, every conversation, every meal, every footstep would become increasingly difficult to navigate. Relevant information would be buried under vast quantities of irrelevant details. The ability to forget — to selectively discard information that is no longer useful — is as important to a functional memory system as the ability to remember.

This article explores the mechanisms by which we forget, the rate at which forgetting occurs, the distinction between decay and interference, and the evidence that forgetting serves important cognitive functions.

Encoding, Storage, and Retrieval: The Three Stages of Memory

Before examining forgetting, we must understand how memories are formed and maintained.

Encoding is the process by which sensory information is converted into a form that the brain can store. When you read a word, the visual system processes the letters, language areas recognize the meaning, and the prefrontal cortex engages attention to select the word for storage. Encoding involves strengthening synaptic connections — the connections between neurons. When two neurons fire together repeatedly, the synaptic connection between them strengthens, a process summarized by Donald Hebb's principle: "neurons that fire together wire together." The stronger the synaptic connection, the easier it is to reactivate the memory later.

Encoding is not passive. Effortful encoding — deliberate attention, relating new information to existing knowledge, visualizing — produces stronger, more durable memories than incidental encoding (passive exposure without effort). This is why studying with full attention produces better retention than background reading.

Storage is the persistence of encoded information over time. Memories are stored as patterns of synaptic strength distributed across networks of neurons. There is no single neuron that contains "the memory of your first day of school." Rather, millions of neurons have slightly stronger connections with their neighbors because of the patterns of firing during that experience. The pattern of synaptic strengths distributed across these millions of neurons constitutes the memory.

Synaptic strengthening involves molecular processes. Repeated activation of a synapse triggers the release of calcium ions into the postsynaptic neuron, which activates signaling cascades. These cascades activate genes that produce more neurotransmitter receptors and adhesion molecules, physically strengthening the synapse. This process is called long-term potentiation (LTP) — the long-lasting increase in synaptic strength following repeated activation. LTP is considered the cellular basis of memory formation.

Retrieval is the reactivation of a stored memory. When you recall an event, you are reactivating a subset of the neurons and synapses that were active during the original encoding. The prefrontal cortex and medial temporal lobe work together to search through memory networks and reconstruct the memory. Retrieval is not a playback of a stored recording but a reconstruction: the brain activates a pattern of firing similar to (but not identical to) the original pattern, and the phenomenal experience of "remembering" arises from this reactivation.

Importantly, each time a memory is retrieved, it becomes strengthened. Retrieval practice — consciously retrieving a memory repeatedly — produces stronger, more durable memories than the equivalent time spent encoding without retrieval. This principle underlies the effectiveness of testing and interleaved practice in learning.

Source: Dudai, Y. (2012), The Neurobiology of Memory: Concepts, Findings, Trends, Oxford University Press — comprehensive overview of encoding, storage, and retrieval mechanisms.

Source: Squire, L.R. (2004), "Memory systems of the brain: a brief history and current perspective," Neurobiology of Learning and Memory, Vol. 82 — foundational review of memory systems and their neural bases.

The Forgetting Curve: Quantifying Memory Loss

In 1885, German psychologist Hermann Ebbinghaus conducted a series of experiments on his own memory. He created lists of nonsense syllables (meaningless three-letter combinations like "ZUV" or "GIB") and tested his retention of them after varying intervals of time. He found a striking pattern: the rate of forgetting is not uniform. Immediately after learning, most information is retained. After one hour, about 50% is lost. After a day, about 70% is lost. But after several weeks, the loss plateaus — the remaining information is forgotten very slowly.

This pattern, now called the forgetting curve, has been replicated thousands of times with various types of information (faces, words, facts, procedures). The curve follows a logarithmic decay: the steepest drop occurs in the first hours and days, then the rate of forgetting slows dramatically. Mathematically, the curve is described by the equation:

Retention = 100 × (1 + t/c)^−b

where t is time since learning, c and b are constants that depend on the material and the learner. For new, arbitrary information, a typical value of b is about 1.25, producing the rapid initial loss.

The forgetting curve has profound implications for education. It suggests that studying material once produces rapid forgetting. To combat this, learners must review material repeatedly. Furthermore, the timing of review is critical: reviewing material shortly before forgetting (when retention has dropped to 50% or so) is more effective at strengthening the memory than reviewing material that is still fresh in mind. This principle underlies the effectiveness of spaced repetition — reviewing material at increasing intervals (1 day, 3 days, 1 week, 2 weeks, 1 month) — which produces much better long-term retention than massed practice (reviewing material many times over a short period).

Spaced repetition — a learning technique in which material is reviewed at increasing time intervals. The optimal spacing depends on the desired retention interval. For a student who wants to remember material for an exam one month away, reviewing on days 1, 3, 7, 14, and 21 is more efficient than reviewing multiple times on day 1. Research indicates that spaced practice can increase retention by factors of 2 to 5 compared to massed practice.

Time Since Learning	Typical Retention (%)	Implication
Immediately after learning	90–100	Most information is accessible
1 hour	50	Half is lost; first review should occur now
1 day	30	Two-thirds lost without review
1 week	10–20	Steep loss without intervening review
2 weeks	5–10	Remaining memories are more durable; slower decay
1 month	3–5	Highly stable; will persist with occasional reactivation

Source: Ebbinghaus, H. (1885), Memory: A Contribution to Experimental Psychology, Dover Publications (1964 reprint) — the original work establishing the forgetting curve.

Source: Cepeda, N.J. et al. (2006), "Distributed practice in verbal recall tasks: A review and quantitative synthesis," Psychological Bulletin, Vol. 132 — meta-analysis of spaced repetition effectiveness across hundreds of studies.

Two Mechanisms of Forgetting: Decay and Interference

Forgetting occurs through at least two distinct mechanisms: decay (passive weakening of memories over time) and interference (new information competing with or overwriting old information).

Decay theory proposes that memories naturally weaken over time through disuse. The synaptic connections underlying memories are strengthened through repeated firing during encoding, but if the connections are not reactivated, they gradually weaken. Proteins and neurotransmitter receptors that were produced during strengthening are degraded and not replaced. The structure of the synapse returns toward its baseline state. Without periodic reactivation, the synapse gradually loses the excess strength that constitutes the memory.

Decay is not purely passive. Recent research on memory reconsolidation has shown that every time a memory is retrieved, it enters a labile (unstable) state before being re-stored. During this reconsolidation window, the memory can be strengthened or weakened depending on what happens. If the retrieved memory is reinforced (the correct answer is provided, or the memory is paired with new information), it becomes stronger. If it is not reinforced, it becomes slightly weaker. This means decay is not just the absence of reinforcement but an active biological process of synaptic weakening during reconsolidation.

Interference theory proposes that new learning interferes with the retrieval of old memories. When you learn two similar pieces of information (two phone numbers, two historically similar events, two lists of words), the new information can impair retrieval of the old. This can occur in two ways:

Retroactive interference — new information impairs the retrieval of old information. After learning a new phone number, you may have trouble recalling your previous phone number, even though the old number is still encoded in your brain. The new information has interfered with your ability to access the old.

Proactive interference — old information impairs the learning or retrieval of new information. If you have known one phone number for years, learning a new number may be slower and more error-prone, because the old number keeps coming to mind and competing with the new one. The old information has interfered with the new.

Interference effects are strongest when the two pieces of information are similar. Learning two very different languages creates less interference than learning two similar languages (learning Spanish after Italian creates more confusion than learning Spanish after Mandarin). The similarity between memories determines how much they interfere with each other.

At the neural level, interference may occur because memories compete for the same neural resources or because new learning creates a pattern of synaptic strength that is similar enough to a previous memory to activate it, but not identical — the similar but incorrect pattern comes to mind instead of the target memory. This is related to the phenomenon of context-dependent forgetting — memories are easier to retrieve in the context in which they were learned, because the context provides additional cues that activate the memory.

Source: Bjork, E.L. and Bjork, R.A. (1992), "A new theory of disuse and an old theory of stimulus fluctuation," in Learning and Memory: The Ebbinghaus Centennial Conference, Lawrence Erlbaum — detailed treatment of decay and interference mechanisms.

Source: Xue, G. et al. (2010), "The neural mechanisms of forgetting: a prospective, transient weakening of pulse," Current Biology, Vol. 20 — neuroimaging evidence for the neural basis of interference and forgetting.

Motivated Forgetting: Intentionally Suppressing Unwanted Memories

Beyond passive decay and interference, people can intentionally suppress memories — a process called motivated forgetting or suppression. After a traumatic event or an embarrassing experience, you might try to push the memory out of your mind, to stop thinking about it. The question is whether this intentional suppression actually weakens the memory or merely inhibits its retrieval.

Research using functional neuroimaging shows that suppressing a memory involves activation of the prefrontal cortex, particularly the dorsolateral prefrontal cortex (dlPFC), which sends inhibitory signals to the medial temporal lobe (hippocampus and surrounding structures). The hippocampus is crucial for retrieving memories, and when the prefrontal cortex inhibits hippocampal activity, memory retrieval is suppressed.

Crucially, repeated suppression of a memory produces lasting weakening of that memory. Over time, with repeated suppression attempts, the memory becomes harder to retrieve even when the person is not actively trying to suppress it. This is genuine forgetting, not merely retrieval inhibition. The suppression-induced weakening appears to involve the same reconsolidation mechanisms that underlie decay: each suppression attempt causes the memory to enter a labile state, and if it is suppressed (not reinforced), it becomes slightly weaker.

However, motivated forgetting is imperfect. Complete suppression of a memory is difficult and often impossible. Attempting to suppress a memory often paradoxically increases its accessibility (the ironic rebound effect), perhaps because the suppression attempt itself retrieves the memory briefly, strengthening it. Additionally, cues in the environment can trigger involuntary retrieval of suppressed memories, and the suppressed memory remains in the brain, potentially accessible through different retrieval routes.

This has implications for trauma and intrusive memories. Suppression attempts alone are often ineffective for managing traumatic memories. More effective approaches include cognitive processing (understanding the trauma and its context), exposure (repeatedly retrieving the memory in a safe context until it is no longer emotionally overwhelming), and reconsolidation-based interventions (retrieving the memory and pairing it with new, non-threatening information before re-storage).

Source: Anderson, M.C. and Green, C. (2001), "Suppressing unwanted memories by executive control," Nature, Vol. 410 — seminal fMRI study of memory suppression and prefrontal inhibition of hippocampal retrieval.

Source: Levy, B.J. and Anderson, M.C. (2012), "Purging of memories from conscious awareness," Trends in Cognitive Sciences, Vol. 16 — review of mechanisms of motivated forgetting and suppression.

Tip-of-the-Tongue Phenomena: Forgetting in Action

The tip-of-the-tongue (TOT) state — the experience of knowing that you know something but being unable to retrieve it — provides insight into the distinction between loss of accessibility and loss of information.

When you experience TOT, you often have partial information: you might know the first letter of the word, the number of syllables, the meaning, or the context in which you learned it, but you cannot retrieve the word itself. This demonstrates that the word is not completely forgotten — the information is still encoded in memory. Rather, the specific retrieval path to the word's phonological form (how it sounds) is temporarily inaccessible.

The more you try to retrieve the word directly, the less likely you are to succeed. But if you stop trying and engage in other activity, the word often comes to mind spontaneously (after hours or even days). This is called the pop-up effect. The pop-up effect suggests that the retrieval failure is not due to the information being permanently lost but due to interference from competing memory traces or an insufficiently activated retrieval cue.

TOT experiences are more common for rarely used words, proper nouns, and information not used recently — exactly the cases predicted by decay and interference theory. TOT also becomes more frequent with age, as the decay of less-used memories accelerates, and as more competing memories accumulate (greater interference).

Source: Brown, R. and McNeill, D. (1966), "The 'tip of the tongue' phenomenon," Journal of Verbal Learning and Verbal Behavior, Vol. 5 — foundational study of TOT phenomena and partial retrieval.

Why Forgetting Is Adaptive

At first glance, forgetting seems wholly negative — a failure of the memory system. But forgetting is adaptive. A brain that perfectly retained every experience, every conversation, every piece of irrelevant detail would become increasingly dysfunctional.

Cognitive clutter reduction — without forgetting, the mind would accumulate vast quantities of irrelevant information. The route to your childhood home, the faces of people you met once at a conference, the specific words of a conversation from years ago, the details of dozens of books you read — all would persist equally vividly. Navigating memory for relevant information would become increasingly difficult. The ability to forget allows the mind to discard information that is no longer useful and focus on relevant, recent information.

Updating beliefs — forgetting enables the mind to update outdated beliefs and information. If you perfectly remembered that someone had wronged you a decade ago, even after they had changed and apologized, forgetting would enable forgiveness and renewed trust. Conversely, if outdated facts (a colleague's old address, a business that closed years ago) persisted as vividly as current facts, it would lead to errors and confusion. Forgetting allows the mind to progressively update its model of the world.

Reducing rumination — the ability to forget or suppress unwanted thoughts and memories is crucial for mental health. Intrusive memories and persistent negative thoughts contribute to anxiety, depression, and posttraumatic stress disorder. The capacity for motivated forgetting is a protective factor, allowing people to move past negative events. When this capacity is impaired (as in PTSD, where intrusive memories persist despite suppression attempts), quality of life is severely affected.

Prioritizing learning — forgetting recent, low-priority information while retaining older, high-priority information makes cognitive sense. Spaced repetition and the forgetting curve exploit this property: by allowing forgetting to bring a memory to the edge of loss and then reactivating it, the mind is signaled that the information is important. The reactivation during reconsolidation produces stronger memories. Memories that are frequently used will be retrieved often and will stay strong, while memories that are never retrieved will gradually fade — an efficient system for prioritizing what matters.

Recent research on inferential forgetting has shown that forgetting can actually enhance memory in some cases. When people forget specific details of an experience, they often retain and strengthen the gist — the general meaning or principle. Forgetting the specific numbers from a statistics lecture but retaining the key concepts is adaptive if the concepts are more useful than the memorized numbers.

Source: Bjork, R.A. (1978), "The updating of human memory," in Psychology of Learning and Motivation, Academic Press — theoretical paper on the adaptive functions of forgetting.

Source: Schacter, D.L. (1999), Searching for Memory: The Brain, the Mind, and the Past, Basic Books — popular treatment of the adaptive significance of forgetting.

Normal Forgetting Versus Pathological Memory Loss

It is important to distinguish normal forgetting from genuine pathological memory loss, which occurs in dementia, severe head injury, or other neurological conditions.

In normal forgetting, the information was encoded but is now inaccessible due to decay or interference. In pathological amnesia, the problem is typically in encoding or storage: the information was never properly encoded in the first place, or the neural substrates that store memories have been damaged.

Anterograde amnesia — difficulty forming new memories after brain injury. A person with anterograde amnesia may be unable to form new long-term memories despite normal short-term memory. This reflects damage to the hippocampus or other medial temporal structures necessary for encoding. In normal forgetting, encoding is successful initially; the information is simply inaccessible later.

Retrograde amnesia — loss of memories formed before brain injury. A person with severe retrograde amnesia may have lost years of memories. This is distinct from normal forgetting, which affects all time periods roughly equally (though remote memories may be more resistant to interference due to having been retrieved and reinforced many times).

Dementia, by contrast, involves both progressive loss of memories and progressive loss of the ability to form new memories. In Alzheimer's disease, the hippocampus and cortical regions degenerate, destroying the neural substrate of memory itself. This is qualitatively different from normal aging, in which memory decline is often modest and is partly offset by the benefit of larger lifetime experience.

Source: Squire, L.R. and Bayley, P.J. (2007), "The neurobiology of remote memory," Annual Review of Neuroscience, Vol. 30 — distinction between normal memory decline and amnesia syndromes.

Frequently Asked Questions

Q: Is it true that humans only use 10% of their brain, and the other 90% is unused memory?
A: No, this is a widespread myth. Humans use nearly all of their brain, and most of the brain is active almost all the time (even during sleep). The myth may arise from confusion between brain usage and memory capacity. The brain has vast memory capacity, but forgetting is a feature, not a limitation. You are not unable to access 90% of your memories because they are in an unused part of the brain; you forget information because forgetting is adaptive.

Q: Can memories be completely forgotten if you don't use them?
A: For practical purposes, yes. If information is not retrieved or used for many years and has not been reinforced through spaced repetition, it becomes increasingly inaccessible. However, in some cases, the original encoding may not be completely erased: cues or retrieval contexts might trigger recall of information thought to be completely forgotten. It is difficult to say with certainty that a memory is permanently lost versus merely inaccessible given current retrieval cues.

Q: Is it better to study all night before an exam or to space studying over many days?
A: Spaced studying is substantially more effective. Studying all night before an exam (massed practice) will result in good retention for the exam, but most of the information will be forgotten within a few weeks. Spacing the same amount of study time over many weeks (spaced practice) results in much better long-term retention. For retention beyond the exam, spaced practice is worth the investment of time.

Q: Why do embarrassing moments feel unforgettable?
A: Embarrassing or emotional experiences are encoded more strongly than neutral experiences. The amygdala (emotion-processing region) enhances memory consolidation when emotional arousal is high. So embarrassing moments often become very durable memories. However, they are not truly unforgettable, even if they feel that way. Over time, without rehearsal, even emotional memories fade — though they typically fade more slowly than neutral memories.

Q: Can you recover forgotten memories through hypnosis or memory recovery therapy?
A: Hypnosis does not reliably recover genuine lost memories. It can, however, produce false memories. A person under hypnosis is highly suggestible and may produce plausible-sounding memories based on expectation, imagination, or the hypnotist's suggestions. These "recovered memories" are often false, even if they feel genuine to the person experiencing them. This is a significant concern in the context of recovered memory therapy, in which therapists attempt to recover allegedly repressed memories of abuse. Research suggests most such recovered memories are not genuine but are confabulations created through suggestion.

Q: Is the forgetting curve the same for everyone?
A: The general shape of the forgetting curve is similar across people, but the specific rate varies with the individual, the material, and the context. Older adults typically forget faster than younger adults. People with higher intelligence or more expertise in the domain show slower forgetting (stronger retention). Meaningful, well-understood information is forgotten more slowly than arbitrary nonsense syllables. The principle of spaced repetition applies across all these groups, but the specific intervals that are optimal may differ slightly.

Q: What is the difference between forgetting and not remembering?
A: Forgetting implies that something was once remembered and is now inaccessible. Not remembering could mean the information was never encoded in the first place. For example, how many windows does your house have? You might not remember even though you have seen your house thousands of times. This is not forgetting; the information was never specifically encoded. Forgetting, by contrast, would involve once knowing something clearly and now being unable to recall it.

---

Read next:

• How Memory Works in the Brain Explained
• What Causes Phobias? The Science
• Why Do We Get Bored? The Science

For more on memory and learning, explore related explainers on the neuroscience of learning and effective study strategies.