• Andrew Kelly


mudrā/seal, signet ring, stamp, token, ritual gesture, internal practice

mudrā/seal appears in the haṭhayogic texts in two very different senses. These are nonetheless deployments of the same metaphor. They pick up two aspects of what a seal does, closing and asserting.

A seal closes something with a sign. It is not a lock that relies on force but an act of authority (even if that was only the authority of someone sealing a private message). The English word seal comes from Latin sigillum/little sign, a diminutive of signum: the smallness refers to the size of the image on a signet ring. Some types of (literal) mudrā were also worn as rings: convenient for use and a safe place to keep an eye on them. Having (and being seen with) such a mudrā was a mark of status, implying you have the authority to use it.

Our purely mechanical use of seal (hermetic seal, sealing a wound or a gas leak) is secondary and recent: those examples lack the visual, meaning element, the stamp of declaration, and are mere fact.

A letter might be sealed with a blot of hot wax and stamped with a seal indicating both who it comes from and that it has not not been opened. A container of cargo might be sealed with a tag that indicates it has been inspected, contains certain items in a certain quantity, and that it not been interferred with since. In ancient Mesopotamia such commercial use of seals predates writing (and may have had a role in the development of writing). In these cases the image, just like any writing involved, is a potent emblem rather than decoration. Ownership and authority is asserted, and a situation is held in a fixed state until the seal is (legitimately) broken. The illegitimately broken seal is meant to expose itself as such.

In the seal then something is blocked and held, and something is declared usually involving an assertion of some kind of power.

The wearing of such 'power-objects' (here mundane social and political power) provided the metaphor for mudrā as a site of power on the body.


mudrā+adhipa/seal master, signet-ring lord:

officer commanding a fortress

First century copper alloy signet-ring from Afghanistan, now in the British Museum. The image shows two naked women in headdresses flanking a column on top of which stands a lion.

10th century terracotta seal, about 5cm across, of the monastery/university of Nalanda, in Bihar, India; now in the British Museum. The humble material suggests this would have been used in low level transactions: monastery shopping?


In the haṭhayogic texts mudrā is either a set gesture, typically but by no means only, made with the hands (such as ‘prayer position’ añjalī mudrā); or it is some kind of internal operation. One type in the second category is the tightening of some key point in the body: all of the bandhas (such as the draw-down of the chin and the perineal grip) are mudrās. But the range of practises included under this heading is very broad.

The first use, the external, symbolic kind, is older then the second type. These are set gestures of some significance: pointing to mean ‘I’ll take that one’ or ‘three of those, please’ is instead just ordinary communication. When a god or an existential prophet such as the Buddha or Mahāvīra makes a-bhaya-mudrā, the sign of fearlessness, they are making a claim or giving an invitation for how to live and think of the world: fear not because, or fear not if you...

Mixed messages are possible: below the many-armed Kālī/Durgā holds her upper right arm in a-bhaya-mudrā while holding a severed head and an axe in her other arms. The lower right hand performs varada-mudrā, a sign of generosity and the giving of boons. (A Kalighat style image from around 1865, now in the Vioctoria and Albert Museum in London.)

In tantric practice mudrās were a form of ritual magic believed to bring on higher mental states or possession by gods; or conversely pefforming a particular (the god’s own) mudrā might be a sign that the god had in fact entered the person. No doubt that means they played a role in some kind of wider ecstatic context or performance: not just that casually positioning your fingers will flip a god into you.

Gestures (like facial expressions) tend not only to be received externally as an act of communication but to affect the mental state of the performer. These mudrās can be thought of as inward-bound gestures to condition mental states. Dhyāna-mudrā, a hand position for meditation, is not a gesture of natural or obvious significance (it’s elegant and fitting but still must be explained): but it may operate as a tool of habituation and as a kind of enactment trigger, a signal to the system that meditation is to occur, is to be entered, an anchor reminding that it is underway. This may make it a practical kind of magic. This seems easier to believe than that a series or gestures elaborated in one corner of the world have inherent power.


In the haṭhayogic contexts the boundary between āsana and mudrā was not hard and fast. Gestural mudrās were not limited only to the hands: they may instead be facial actions or require a whole hand-and-body position to launch their meaning; and this or that internal operation are in effect whole body engagements. This makes it easy to see why they might confused. But there is a kind of basic opposition: āsana was for static immobilisation and mudrās were an active intervention.

The viparīta-karaṇī for instance is an inversion (a going upside down) and in some versions involved the chin-lock: these led to it being considered not as an āsana but a mudrā.

Haṭhayoga elaborated a psycho-spiritual plumbing in which the person is like a vast and complex refinery in which the movements of literal and not-so-literal substances and fluids must be controlled and manipulated, sent here, blocked there, in pursuit of various kinds of transformation. Everything in that realm counted as mudrā. Some of it goes to very odd places: refashioning the tongue by progressively severing its tendons and giving daily yanking butter massages (do not try this at home) so finally it can reach back to seal the throat or reach out to touch the brow (warnings that speech may henceforth be impaired); and under the title vajrolī-mudrā a variety of truly distasteful interventions into the male reproductive organs, along with rather misogynist (or at least spectacularly self-centred) sexual practices.


In the context of modern yoga, among the large set of internal mudrās on offer more or less only the bandhas are in common use, both in breathing techniques and in āsana. In our prāṇāyāma sessions we’ve made use of ṣaṇ-muhkī-mudrā/the six aperture mudrā that amplifies the humming of ‘bumble-bee breathing’ in a pretty obvious, not-so-mysterious, but quite wonderful way: it amounts to a sensory filter, deleting sight and minimising external sound while instead opening up a void of internal resonance for the humming vocal chords.

As for the gestural mudrās, they have taken on a very great prominence as a form of visual display, almost as advertising: a means of marking ‘this is yoga’ on thousands of (postural) yoga pages. How significant they are as part of actual practice is another matter.

If habituation around a conventional, more or less arbitrary, signification is what makes a gestural mudrā useful (as against supposing they have some inherent magical power), then only a frequent mudrā is likely to be of much significance.

Añjalī-mudrā is part of the ceremonious, communal feel of a shared yoga session; and no harm if that’s all it is. At the same time for me it is a very import tool for bringing a particular kind of focus into āsana practice. From a tantric perspective even that would be a serious demotion.


Human physiology and our capacities at experiencing it is the (relatively) stable ground of all yoga practice. What the tantric world offered into yoga practice was a demonstration of how malleable our internal perceptual systems are: it is very striking that ordinary vocabulary pays almost no attention to this leaving us with very few words for it.

The intellectual superstructure that lies on top of the bodily practice (including meditation) is highly variable and always was. It can be changed quite rapidly: Iyengar yoga was very different from that of his teacher (who also changed in the course of his own life)—not because what they did differed physically that much but because they thought and meant different things and had different aims in differing historical circumstances. Iyengar pursued an internality in āsana practice that had perhaps never been placed there before: he looked out to a new global landscape while his teacher faced directly Indian subjection to the British Raj. It’s possible that both types of mudrā pointed in that interior direction away from the militant externality that had driven much of the innovation in yoga practice out of the 19th century.

Take what is useful and make what you can with it.


The long-ā on the end of the word marks it as a feminine noun.

Mudrā looks like a fully regular home-grown word following a pattern where a verbal root is following by a suffix -ra to create a noun or adjective. This is a very common word type (vakra bent, miśra mixed, rudra howler: the -ra doesn’t add a precise meaning). There is actually a word mudra (mud/rejoice + ra) meaning joy-bringing, and it would be very straightforward to switch this into a long-ā feminine. But hard to believe mudrā, originally a quite pedestrian and practical word meaning stamp and signature (or passport or in modern usage typeface!), can come from a word for ‘joy’.

Another view is that it is a loanword from a Mesopotamian source, perhaps from Sumerian.


A historical sourcebook on more or less every constituent of pre-modern haṭhayoga is James Mallinson and Mark Singleton’s 2017 work The Roots of Yoga, one of whose chapters is devoted to mudrā.

A Harappan/Indus Valley seal.