Samskara: Why Do Old Memories Surface in Meditation?

What a samskara actually is

The word is built from sam (together, complete) and the root kr (to make). A samskara is something made and left behind, the residue of an act of attention. In the Yoga Sutra, smriti sits among the five movements of the mind, and by the fourth chapter Patanjali makes the relationship between memory and impression explicit: they are one in form. In that account, recall is the impression rising again rather than a record being fetched from storage.

The classical image is a furrow. Water crossing the same ground enough times cuts a channel, and after that the water goes where the channel goes. Nothing about the channel is remarkable. It is the sum of every previous crossing, held in the shape of the ground. This is why the tradition files habit and memory under the same heading. Both are impressions with a slope.

It also explains a detail practitioners notice and rarely mention: the memories that surface are seldom the important ones. A corridor at a school. The particular sound a screen door made. The tradition would say these grooves were cut early and crossed often, and depth of impression has little to do with the significance we would assign the event now.

The Sanskrit vocabulary makes one more distinction that English tends to collapse. A samskara is the impression. A vasana is the leaning it produces, the tendency that outlives any memory of where it came from. Someone can lose the event entirely and keep the flinch. That gap between a forgotten cause and a live effect is most of what people are describing when they say practice brought something up.

Why old memories surface during meditation and not during the day

An ordinary day supplies the mind with unbroken external input. Something to look at, something to answer, something next. When that supply stops, the brain does not stop with it. It changes sources.

Research on the default network describes this directly. Andrews-Hanna and colleagues characterise self-generated thought by its relative independence from ongoing sensory input, and autobiographical memory is among its most reliable products. Decoupling from sensory input appears to support retrieval rather than interrupt it. Cutting the outside world hands that machinery the floor.

So the sequence is not obscure. You sit down, the input drops away, and a system that was running underneath it all along becomes audible. It explains the timing, too. The memories rarely arrive in the first minute. They arrive somewhere past the tenth, once the mind has stopped rehearsing the day and the supply of fresh material has run out. Brewer's group found the main nodes of that network relatively deactivated in experienced meditators, across concentration, loving-kindness and choiceless awareness alike. Beginners have not built that yet, which is a decent explanation for why the first weeks of practice are the loudest ones. It is also why tears show up unannounced on a morning that had nothing sad in it.

Childhood memories coming back: what the retrieval research shows

The stronger finding is that practice seems to sharpen retrieval rather than blur it. Nyhus and colleagues at Bowdoin gave one group four weeks of mindfulness training and compared them against a waitlist control. Theta power during episodic memory retrieval was higher after training than before, in right frontal and left parietal channels, and source memory improved alongside it. Source memory is the part that tracks where a memory came from rather than only what it contains.

That distinction matters more than it sounds. A memory arriving with its source attached is a memory that can be placed: this happened, it was then, it is over. A memory arriving without one gets experienced as mood, and mood without a source tends to get read as a fact about the present. Half the difficulty people describe in early practice is a sourcing problem wearing the costume of an emotional one.

There is a second body of work worth naming carefully, because the bridge to meditation is an inference rather than a finding. Lee, Nader and Schiller's review of reconsolidation describes how a retrieved memory becomes briefly unstable before it restabilises, and how that window can be used to update it. The experiments behind this are largely fear-conditioning studies in controlled settings, not studies of people on cushions. But the principle is suggestive, and it is the closest modern account of why the tradition insisted that meeting an impression is never neutral. Retrieval writes as well as reads.

Are memories during meditation normal, and when are they not

Normal, common, and documented. Lindahl, Britton and colleagues published the broadest study of meditation-related difficulty in 2017, built from more than a hundred interviews with practitioners and teachers across Theravada, Zen and Tibetan traditions. Their taxonomy includes the re-experiencing of past events, and they note it often followed a release of physical pressure or tension in practice. It appears in the research as a category, not as evidence that someone has practised badly.

The same study is the reason to name the floor. Where a memory carries trauma, unstructured silence is a poor container for it. The signals worth respecting are specific: the memory does not resolve after the sit, sleep gets worse across several nights, the body stays activated for hours, or there is a sense of watching yourself from outside. None of that means stop practising forever. It means this particular practice, alone and without structure, is the wrong instrument, and a trauma-informed teacher or therapist is the right one. Fear arriving during practice follows the same rule.

How to clear samskaras without trying to erase them

Patanjali does not propose deletion, which is the first surprise for anyone arriving with a modern instinct to process and be done. His method is displacement. When an impression pulls hard, cultivate its opposite, and cultivate it in a way the body can register: pratipaksha bhavana. The groove stays. What changes is how often the water is sent down it.

Read against the retrieval work, that is a better instruction than it first appears. Nothing in the research suggests a memory can be removed. Plenty of it suggests that the conditions under which a memory is met shape what it becomes. Meeting an old impression in a settled body, with the breath long and the room quiet, is different from meeting it at two in the afternoon with a phone in hand. The tradition arrived at that conclusion without instruments. The instruments have not contradicted it.

The practical weight lands on continuity rather than intensity. An impression that took twenty years to cut will not be met once. It will be met on an ordinary Tuesday, briefly, and then again some weeks later, and the second meeting will be quieter than the first. That is the whole mechanism. A daily seat is not a discipline imposed on the memory. It is the only condition under which the memory has somewhere ordinary to arrive.