What’s the Use of Memory? A Practice of Memory and Saṁskāra

YS I.11 – anu-bhūta-viṣaya-asaṁpramoṣaḥ smr̥tiḥ
Memory is the retention of images of sense objects that have been experienced

YS IV.9 – jāti deśa kāla vyavahitānām-apy-āntaryāṁ smr̥ti-saṁskārayoḥ ekarūpatvāt.
Because they are identical, there is an uninterrupted connection between memory and saṁskāra, even though they might be separated by birth, time, and place.

YS II.10 – tasya praśānta-vāhitā saṁskārat
The mind’s undisturbed flow occurs due to saṁskāras.


Memory, as Proust reminds us, is not always a reflection of the way things actually were. Indeed, human memory, unlike, say, computer memory, fulfills a different function. The function of memory has been analyzed by both contemporary psychology and in the literature of classical yoga, with some interesting convergences and equally interesting divergences. Here we will examine the purpose of remembering from both the contemporary psychological perspective and the perspective of classical yoga, as exemplified by Patañjali’s Yoga Sutras. The purpose of laying out this framework is to suggest what the yogic practice of memory is, and what it can contribute to the effort toward liberation.

Some contemporary memory researchers point to three purposes served by remembering (Bluck, Alea, and Habermas, 2005). The first function is to provide a sense of continuity for the self. To a large extent, our concept of our self as a being with temporal extension – i.e. someone who existed in the past, is existing now, and will continue to exist as the same personality in the future, having a set of experiences along the way – is a construct of memory. This construct exists to serve the purposes of the working self. The working self is the set of circumstances, feelings, priorities, goals, attributes, and abilities we attribute to ourselves right now, and the lens of the current working self biases how we remember ourselves in the past. A very large body of psychological research has demonstrated the interesting manner in which we reconstruct our pasts (e.g. Ross, 1989) to serve our current needs. Perhaps unsurprisingly, the way we remember ourselves being in the past is usually closer to how we are now than how we actually were. Any given memory is thus closer to a dynamically reconstructed simulation, rather than an object we had put away in a drawer in our mental deep storage, later to be retrieved and pulled into the spotlight of consciousness. But this reconstruction is not only exerted by the working self upon memory; memory also constrains the working self-concept that can manifest. The working self is constructed based on the options available in memory; that is, I can be this kind of person I am now because of the events I have selectively kept salient in memory (or that circumstances have kept salient). Memory and self, then, dynamically feed one another, and both are subject to reconstruction (Conway & Pleydell-Pearce, 2000).

In addition to maintaining the (perhaps illusory) sense of continuity for a personal self, memory also serves to maintain continuity of relationships, and cultivating social connection with others. Other people are woven into our personal memories, and sharing memories with others draws us closer to them.  Talking about shared events with other people facilitates a sense of closeness. Indeed, memory seems especially malleable when we want to create a shared reality with someone else. In conversation, people wanting to relate will converge on a shared representation of an event, and this collective memory undergirds a sense of shared identity as family members, members of a community, or a tradition (Hirst & Yamashiro, 2016).

Finally, memory is not only for the purposes of a nostalgic, backwards-looking recherche du temps perdu – it is also the means by which we can use past experiences to make decisions about activity in the future. A very interesting new line of research seems to indicate that the same brain networks involved in remembering our personal past are also co-opted to imagine ourselves in the future (Szpunar, 2010). The way we think about the future is thus strongly conditioned by the way we think about the past.

Coming, as psychologists do, from a more or less implicit set of Western cultural values, these functions have the end goal of cultivating a well-defined individual self functioning effectively in a social environment. Remembering is a practice by which we develop the narrative coherence that can undergird a strong and healthy self-concept; it is a means of nurturing social connection with the other people in our various communities; and it is used to inform future activity. These three functions of remembering coincide with the functions laid out in classical yoga; however, Patañjali’s end goal is radically different.


Remembering undoubtedly plays a dominant role in our mental activity. In the yoga literature, conscious mental activity is called citta-vr̥tti – citta being consciousness, and vr̥tti being fluctuations. Traditional commentators frequently use the metaphor of citta being the ocean, and vr̥ttis being the waves; this water metaphor is not without parallel in western psychology, where James (CIT) discussed the stream of consciousness, as the continuous, connected, but always changing flow of thoughts. Patañjali categorizes five types of fluctuations – pramāṇa(right knowledge), viparyaya (error), vikalpa (imagination), nidrā (dreamless sleep), and, finally, smr̥ti (memory). Yoga, as Patañjali defines it, is citta-vr̥tti-nirodhaḥ – restraint stilling the fluctuations of consciousness. These five vr̥ttis then are the real target of yoga practice. In yoga soteriology, the citta-vr̥tti of memory plays a vital role, both in maintaining saṃsāra, the cycle of suffering, and as a vessel carrying the yogi to the threshold of ultimate liberation.



In order to understand how yogis may use memory to carry themselves towards liberation, we must first address how memory ordinarily functions in the scheme of classical yoga. Experiences are represented within citta; we may be conscious of something we see, or feel, or infer, imagine, or are told. That conscious experience forms a trace in citta called a saṁskāra. These traces are retained over time, and accumulate within citta. Saṁskāras may become unconscious, may interact with one another dynamically, and may later become conscious again if we encounter something in the internal or external environment that triggers the saṁskāra’s reactivation. Saṁskāra may also remain unconscious – perhaps for many lifetimes. That is, to riff on the neuropsychologist Dan Schacter’s (Schacter, Wagner, & Buckner, 2000) scheme for memory, saṁskāras may be explicit (consciously remembered) or implicit (unconsciously priming behavior) to different degrees; even implicit saṁskāras, however, may influence citta-vr̥tti. When I perform an āsana, I am not consciously thinking of every time I have performed that āsana in the past; nonetheless, the saṁskāras accumulated over a long history of practice inform my current performance. In the language of psychology, I have developed an implicit procedural memory of a skilled action, such that I may execute it with expertise and without excessive conscious awareness of how I accomplish it. All sorts of unconscious memories shape habits and skills. Our store of unconscious and conscious saṁskāras, our memories and habits, can conduce towards liberation, or may further unfreedom and suffering.

Memory, in yoga, results when saṁskāras accumulated during this life become reactivated in citta. Our “full package” of saṁskāras that condition citta-vr̥ttis, our karmaśaya, however, does not only contain saṁskāras accumulated during this life. According to the traditional Hindu model, saṁskāras transmigrate with the citta from lifetime to lifetime, accumulating as they go along, and so in addition to the saṁskāras imprinted during this life, citta- vr̥ttis may be agitated by a vast number of saṁskāras from previous lives. Even if one were to look askance at the literal idea of citta and its karmaśaya being reincarnated in different bodies, it is undoubtedly true that an individual’s autobiographical events are not the only sorts of “memories” that can strongly impact her conscious experience as an embodied being; I would say the genetic heritage of billions of years of evolution, social karma like histories of colonialism, capitalism, and experiments in political science, both “liberatory” and “oppressive,” education systems, a community’s collective memory, etc., are all saṁskāras that may extend their influence from our distant, pre-birth past to impact citta-vr̥ttis. Our karmaśaya does not start with this particular self.

Saṁskāras, then, accumulate, and become expressed to various degrees in citta- vr̥ttis. When a specific context triggers activation of a saṁskāra from the personal past, it can instigate the citta-vr̥tti Patañjali classifies as smr̥ti – memory. Smr̥ti is an important class of citta-vr̥tti, because in Patañjali’s yoga, memory mediates the link between avidyā (ignorance) and karma(activity). The unskillful use of memory conduces to bondage and suffering – that is, memory often acts as a kliṣta-vr̥tti, or detrimental fluctuation of consciousness. Memory from this life and saṁskāras from previous lives activate in response to current stimuli, and lead us to judge a newly encountered experience as something desirable or undesirable based on our memory of how that experience affected us in the past. Because we retain a memory of pleasurable objects, we seek them out. If we encounter an object that was pleasurable in the past, we desire it, and attempt to attain it. If the pleasurable object is attained, the vr̥tti cognizing that attainment enters citta as a saṁskāra, to be activated at the next encounter, and the cycle repeats. If the pleasurable object is not attained, we experience duḥkha (frustration) or krodha(fury, rage) and, again, the experience imprints a saṁskāra that conditions future behavior. All this can be categorized under Bluck et al.’s (2005) second purpose of memory – using past experiences to guide future behavior. While operating in prakṛti, this seems a perfectly reasonable way to go about operating. However, we can notice here a startling lack of agency – activity is being driven by a compulsion to seek pleasure and avoid pain, and activity is at the mercy of the past and current circumstances. As long as memory is driving this fairly automatic activity, we are caught in an endless pursuit of illusory pleasures (see my article on Maya). Pleasure is a problem, in yoga, because it leads to attachment to objects that in reality are impermanent, and incapable of providing ultimate fulfillment. If the pursuit of pleasure is our sole motivation, we as an unavoidable consequence become subject to aversion, frustration, and anger if the desired object is unattainable – and, within prakṛti, all objects of desire are unattainable in any ultimate sense. The pursuit of pleasure is something that chains us to the inherently dissatisfactory world. Fixation robs us of freedom, because acting on fixations is compulsive.


In ordinary use, then, memory plays a key role in maintaining the wheel of saṃsāra. Experiences accumulate as saṁskāras, and these traces are activated as memories in response to different contexts to support affective fixation. This affective fixation drives further karma, and the deposit of further saṁskāras. However, memory does not have to play this role. Depending on the content of the memory or saṁskāra, the vr̥tti may be kliṣta or akliṣta – it may be either afflictive and detrimental, or conducive to liberation. Memory may be used in ways that drive further suffering, attachment to unfree activity, identification with an isolated, individual self, fear of loss, fear of death. Memory may also be used in ways that accumulate akliṣtic, sattvicqualities – qualities of lightness, joy, and clarity. The yogi uses memory to keep her eye on the prize, so to speak, and the yogic use of memory plays an important role in maintaining a dedicated practice and cultivating samādhi. The first stages of samādhi, samprajñāta-samādhi, are states of meditative absorption in which the yogi uses conceptual resources to deepen absorption. In YS I.20, deeper stages of Samadhi are preceded by Śraddhā – faith, virya – vigor, smr̥ti – memory, and prajñā -purvāka – discrimination.

Discrimination (prajñā -purvāka) refers specifically to discrimination between desire for illusory and ultimately unsatisfactory prakṛtic objects, on the one hand, and the desire for complete liberation on the other (YS II.15). Faith (Śraddhā) is confidence in the efficacy of the yogic endeavor, and belief that this practice will conduce to the attainment of liberation; in times of lax practice, when we allow kliṣta-vr̥ttis to accumulate, we must remember Śraddhā, and why we practice. And finally, the yogi must remember to reinvigorate his practice unremittingly, to practice virya, and engage in practice wholeheartedly, with all his energy and capacity. Intentional remembering is thus required for a sustained effort towards the goal of yoga. Just as do kliṣta-vr̥ttis, akliṣta-vr̥ttis make saṁskāric records, and, by the same process, contribute to memories, habits, and patterns of behavior that reinforce yogic goals. If the yogi has experienced the clarity and happiness of samādhi, the saṁskāras resulting from that experience will be sattvic. Faced with contexts that trigger such sattvic akliṣta-vrttis – perhaps seeing the clean corner and meditation cushion where she practices – the yogi will desire to pursue that practice of dhyāna (meditation) in the future. A liberating practice results from this positive feedback loop between akliṣta-vr̥ttis, saṁskāras, and skillful action. With continual practice, sattvic saṁskāras will accumulate, which will allow citta to become clearer and clearer, more stable, self-content, less destabilized by objects in the external world of prakṛti. When these sattvic saṁskāras trigger akliṣta-vr̥ttis, they obstruct kliṣta-vr̥ttis (YS I.50), and place constraint (nirodha) upon the ordinary saṁskāras that are compelled by prakṛti motivations (YS III.9). The proximal goal of cultivating akliṣta-vr̥ttis, and thus the saṁskāras they contribute to the karmaśaya, is to clarify the buddhi – the intelligence – sufficiently that it can reflect the Atman. In dhyāna, memories that drive activity in the external world are suppressed by memories accumulated during a practice of deepening samādhi. A long term, dedicated practice has to be sustained by intentionally cultivated memory. Again, these accumulated akliṣta saṁskāras do not have to be conscious; they may be unconscious, procedural memories that enable skilled activity – like entering samādhi – without conscious, reflective awareness of how that skill is being accomplished. If memory plays a vital role in perpetuating saṃsāra, it plays an equally vital role in freeing the yogi from saṃsāra.

The Last Saṁskāra

Habitual accumulation of saṁskāras can create a karmaśaya that is predominantly kliṣta or akliṣta; however, even if the karmaśaya becomes completely saturated with sattvic saṁskāras, producing the clearest akliṣta-vr̥ttis, these fluctuations of consciousness remain phenomena within prakṛti. As laid out in my previous article on Maya, citta as a phenomenon within prakṛti can be decomposed into several functional components. These are buddhi – the intellect, of which smr̥ti is one function, manas – the aspect of mind that categorizes sensory experience, and assigns affective valence to different stimuli from the “external” world, and ahaṃkāra – the sense of individual personality. As a jivanmukta – the personality liberated in this life – the yogi fails to construct the ahaṃkāra within citta, but can continue to use buddhi and manas to engage with the world according to his dharma. The accumulation of akliṣta saṁskāras contributes to the fading away of the ahaṃkāra; partially by definition, an aklista-vrtti is a vr̥tti that does not contribute to shoring up the ahaṃkāra. Regular practice of dhyāna suppresses the kliṣta-vr̥ttis and their afflicting saṁskāras. The jivanmukta, then, skillfully uses buddhi to clarify consciousness as much as is possible while still acting in prakṛti.

Patañjali’s yogi goes beyond this, aiming for complete liberation from prakṛti. Having cultivated within buddhi the qualities enumerated in YS I.20 – Śraddhā, virya, smr̥ti, samādhi, and prajñā -purvāka – the yogi attains the first stage of samādhi – savitarka-samādhi – where concepts, memory, and thoughts still support absorption. In the transition to the second stage of Samadhi, these memory scaffolds are dropped away – YS I.43 describes the state thus – “Absorption without conceptualization (nirvitarka–samādhi) occurs when memory has been purged and the mind is empty, as it were, of its own nature. Now only the object [of meditation] shines forth [in its own right].” In this stage, attention is constrained to the object of meditation (a mantra, Isvara, the breath, etc.), with no conceptual interpretation, memories or saṁskāras, imposed on the perception. The object of meditation, completely purged of discursive meaning, is experienced in what Mircea Eliade terms its “existential nakedness.” As samadhi progressively deepens, perceptions of form, of ahaṃkāra, of prakṛti itself, drop away. Finally, buddhi attains its most reified form in the last of the samāpattis, or bija-samādhis – samādhi with seed, or citta from which the yogi may still sprout future karma. At this point, there is only one final thought for the yogi who would be completely free. Edwin Bryant humorously dubs this virāma-pratyaya the “Terminator thought” – the thought to end all thoughts. The final akliṣta-vr̥tti, once citta has been nearly completely purified, is the determination to end all thought. Following the release even of this vr̥tti, consciousness is released from all support within prakṛti; with no representation of anything within citta, the Ātman’s attention is not drawn to anything within prakṛti, and can rest in itself.

The intentional cultivation of memory thus plays a crucial role in yoga practice. Just as the practice of āsana releases detrimental holding patterns within the body to cultivate a body that is more free and open, the yogic practice of memory aims to cultivate a karmaśaya that supports akliṣta-vr̥ttis, and a stream of consciousness that is free, and light. And, just as the yogi practices āsana for purposes that transcend physical mastery per se – that is, the purpose of cultivating physical mastery is to condition a body that can most effectively practice deep and solid dhyāna – the practice of yogic remembering also serves an ultimate purpose beyond the pleasant experience of sattvic citta-vr̥ttis per se. The yogic cultivation of memory and saṁskāras is for the purpose of developing a karmaśaya that most effectively supports the goals of yoga, namely, restraint of kliṣta-vr̥ttis, and the dominance of akliṣta-vr̥ttis to such an extent that the virāma-pratyaya can make its appearance. The final aim of this cultivation of memory is to discipline citta to the extent that eventually memory and all other phenomena within citta can be released, and Ātman rest awareness in its own pure existence.


Can practitioners who are not quite ready just yet to separate from prakṛti benefit from yogic memory practices? Of course. What the cultivation of akliṣta-vr̥ttis and the  accumulation of sattvic saṁskāras really amounts to is intentionally using memory as a means of yoking consciousness to practice. Lots of remembering is involuntary (Berntsen, 2009) and implicit; this is a normal aspect of human memory. However, just as practicing āsana involves holding physical shapes that at first feel constrained, unnatural and uncomfortable, yet after long practice yield great agility, pleasure and, ironically, greater freedom of movement, intentional cultivation of yogic remembering can also support a conscious experience that is clearer, lighter, and more pleasant.

This may look like cultivating habits of attention. We can become aware of when our attention is drawn involuntarily toward objects that support ignorance, egoism, aversion, fear of loss, clinging to a static and inflexible self-concept. We can notice which objects of attention we invest in emotionally; does attending to this object support liberation, or does it contribute to more suffering? If an object of attention is not liberating, can attention be redirected? Many modern people, under the thrall of suspect Freudian assumptions, seem allergic to the idea that suppressing an afflicting thought could possibly be salutary, but this is precisely how Patañjali suggests using akliṣta-vr̥ttis to subjugate kliṣta-vr̥ttis (YS I.50). Indeed there is convincing evidence that, given comparable degrees of trauma, more psychopathological symptoms are found in people whose personal narrative centers the trauma – i.e. who attribute extensive personal meaning to the event(s), and who frequently ruminates about the event(s) – than to people whose personal narratives do not center the traumatic event(s) (Berntsen & Rubin, 2006). Where the attention goes, there goes everything else – self-construct, emotional investment, etc. And, in the yogic scheme, so also goes the accumulation of saṁskāras, which then facilitate attention going in that direction in the future.

Pay attention to the way you are using memory; which memories do you spend a lot of time thinking about, and what is the emotional impact of thinking about that memory? How does that memory occupying space in consciousness impact your actions? Your sense of self? Your feeling of connection to others? What sort of saṁskāras are you developing by entertaining these memories? Skillful use of attention is exactly like skillful practice of āsana – the more frequently practiced, the stronger and more flexible it becomes, and each time such attentional control is practiced, we lay down saṁskāras that will support future acts of attentional control that are stronger and more skillful.

Rituals are also important memory practices by which we can remind ourselves of and recommit to our longer-term goals. Rituals may look like remembering to maintain a daily meditation practice or other sādhanā, regular training at a committed time and place, the development of procedural memory for highly skilled tasks through devoted repetition, and any number of other liberating habits that may usefully contribute to more proximal goals.

As a beloved teacher of mine repeatedly reminded me, the practice has to be real. Understanding the framework for the practice conceptually may be a starting point, but the point of remembering this entire conceptual framework is to have a systematic guide for actual practice. The goal of yogic memory is not, after all, remembering a lot of fancy Sanskrit words and philosophical systems, although these can certainly be useful tools. The goal is to be free.


The discussion presented here on memory and the saṁskāras in the Yoga Sutras drew extensively from two sources, Mircea Eliade’s Yoga: Immortality and Freedom, and Edwin Bryant’s incomparable translation of Patañjali’s Yoga Sutras, which includes commentary from the traditional interpreters all the way from Vyāsa up to colonization (and includes Bryant’s own very illuminating perspectives). Both works are highly, highly recommended for anyone interested in the deeper nuances of yoga philosophy. Any merit in this piece is largely due to those two works, and errors in interpretation are my own.


Bluck, S., Alea, N., Habermas, T., and Rubin, D. (2003). A tale of three functions: The self-reported uses of autobiographical memory. Social Cognition, 23, 91-117.

Conway, M.A., and Pleydell-Pearce, C.W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261-288.

Berntsen, D. (2009). Involuntary autobiographical memories. New York: Cambridge University Press. Chapters 2,4, and 5.

Berntsen, D. and Rubin, D. (2006). When trauma becomes a key to identity: Enhanced integration of trauma memories predicts posttraumatic stress disorder symptoms. Applied Cognitive Psychology, 21, 417-431.

Bryant, E.F. (2009). The Yoga Sutras of Patañjali: A New Edition, Translation, and Commentary. New York: North Point Press.

Eliade, M. (2009). Yoga: Immortality and Freedom. Princeton, NJ: Princeton University Press.

Schacter, D.L., Wagner, A.D., and Buckner, R.L. (2000). Memory systems of 1999. In E. Tulving & FIM Craik (Eds.), The Oxford Handbook of Memory, 627-643. New York: Oxford University Press.

Hirst, W., and Yamashiro, J. (2016). Social aspects of forgetting. In M.L. Meade, A. Barnier, P. Van Bergen, C. Harris, and J. Sutton (Eds.). Collective Remembering: How Remembering with Others Influences Memory. New York: Cambridge University Press.