Why a memory can feel gone but still be there
The useful distinction is between two separate things. Storage strength reflects how well a memory is established and retained; retrieval strength reflects how accessible it is at a given moment [1][5]. A memory can have high storage strength and low retrieval strength, which is why something you know well can be hard to recall on demand yet come back with the right cue.
This matters because forgetting is not a reliable index of how much was originally learned. As one account puts it, if forgetting simply tracked the strength of original learning, we could predict retention from learning strength alone, and we cannot [5]. That gap between what is stored and what is retrievable is where the idea that forgetting can help long-term retention gets its footing.
Forgetting can act as a filter that protects what matters
One way forgetting may help is by sorting weak memories from strong ones. In a study of 90 healthy adults who learned word pairs, items studied twice (weak) and three times (strong) were compared after either a night of sleep or a matching daytime waking period. Strong items were recalled better than weak ones in every group, but wakefulness disproportionately impaired the weak items, while sleep preserved weak and strong items about equally [2].
That pattern is consistent with forgetting doing useful work: when weak traces fade and strong ones survive, what remains is a leaner set of more durable memories. The study's authors describe sleep as protecting declarative memories regardless of initial encoding strength, and they found no significant link between classic sleep features such as slow oscillations or spindles and overnight retention [2]. So the filtering effect is real in the data, but its mechanism is not pinned down by these sleep measures.
A faded memory can be brought back, and sometimes that helps
Low retrieval strength does not have to be permanent. Work on forgetting distinguishes types and consequences of forgetting and notes that access to forgotten material can be reinstated at full strength, including material that was intentionally forgotten [4]. In other words, a memory that is hard to reach now may still be recoverable, which is what makes the storage-versus-retrieval split practically important.
There is also a developmental angle. In young mice, age-related differences in spatial memory retention were traced to differences in initial memory strength rather than to storage failure, and appropriate reminders led to recovery; the authors propose that infantile forgetting may support flexibility [3]. That is a proposal about a possible benefit, not a demonstrated purpose, and it comes from mice, so it should not be read as a settled account of human memory.
What this means for how you study and review
The practical upshot is that a memory feeling temporarily out of reach is not proof that learning failed. If storage strength is high, a cue or reminder can restore access [4], and the weak-versus-strong pattern suggests that the memories most at risk during an ordinary waking day are the ones that were only lightly encoded to begin with [2]. Building stronger initial encoding, and giving memories time to be recovered rather than only re-exposed, fits the evidence better than treating every lapse as loss.
The bigger theoretical picture, that forgetting and reminiscence together shape long-term retention through both storage and retrieval processes, was sketched decades ago [1]. What the newer work adds is specificity: sleep appears to protect memories across encoding strengths [2], while the mechanisms behind that protection remain unresolved. The honest summary is that forgetting can enable long-term retention by lowering retrieval strength without destroying storage, and by letting weak traces drop away, but how much this helps and through which biological route is still an open question.
Sources used in this answer
Toward a theory of the development of long-term retention
The paper examines how memory strength, including forgetting and reminiscence, and storage and retrieval processes contribute to long-term retention, with forgetting potentially facilitating retention.
Sleep Preserves, Wake Differentiates: Strength-Dependent Forgetting of Declarative Memories Across the Retention Interval.
In 90 adults learning word pairs, wakefulness selectively impaired weakly encoded items while sleep preserved weak and strong items similarly, and classic sleep oscillations showed no significant association with overnight retention.
Age-dependent changes in spatial memory retention and flexibility in mice
In mice, age-related differences in spatial memory retention were attributed to differences in initial memory strength rather than storage deficits, and appropriate reminders produced recovery, with infantile forgetting proposed to enhance flexibility.
Types and consequences of forgetting: Intended and unintended
The paper reviews types and consequences of forgetting and notes that access to forgotten material, including intentionally forgotten material, can be reinstated at full strength.
Extending the domain of memory retrieval
The paper argues that forgetting is not merely an index of the strength of original learning, since retention cannot be predicted from learning strength alone.