The Yoga Asanas before the Modern Period

December 17, 2014 – 1:16 am


The history of the specifically postural yoga that is the root of the yogas that are practiced in the modern West is not well known. We know that most forms of current Western practice can be traced back to the teachers Vivekananda, Yogananda, Sivananda, Kuvalayananda, Hariharananda, Krishnamacharya, and a few others. Beyond that, it is claimed that those forms derive from the Hatha yoga described in the Hathayogapradipika (hereinafter, HYP,) and the ideological underpinning tends to look to the Raja yoga (as described unhistorically by Vivekananda in his book of that name) and Patanjali’s Yoga Sutras. It is usually claimed that this postural yoga does indeed have roots in deep antiquity, with some claiming to find its sources in the Bhagavada Gita of the Mahabharatra, or certain of the Puranas, and to see its traces in the postures displayed in ancient statuary – perhaps as far back as the Indus Valley civilisation (where the pasupati seal is sometimes taken to be a representation of seated Siva in a yoga posture.)

These claims of ancient origins are, of course, supposed to inspire confidence as to authenticity and, for some reason, effectiveness. We would be well-advised however, not to rest too much of our faith in the effectiveness of yoga (for whatever its effect is supposed to be) on an appeal to ancient authority, for it may well be that our postural yoga is almost entirely of recent invention – perhaps taking its current form within just the last hundred or so years. It is argued, for example, that Krishnamacharya’s version, which he taught to BKS Iyenagar, K Pattabhi Jois, and TKV Desikachar, was an extraordinary mix of techniques from hatha yoga, British military calisthenics, and South-western Indian wrestling traditions (Sjoman, NE, The Yoga Tradition of the Mysore Palace, New Delhi:Abhinav:1996 ref. in White, DG (2011) “Yoga, Brief History of an Idea” (Chapter 1 of Yoga in Practice)).

We do know that Patanjali’s work, which dates to ca. 400 CE, mentions no asanas at all, and that, in fact, “prior to the end of the first millennium CE, detailed descriptions of asanas were nowhere to be found in the Indian textual record.” (White, op. cit., 17) And even then, the 15th C. HYP by Svatmarama was the first to make asanas central to Hatha Yoga. In that work are listed just 15 asanas, whose descriptions are mostly cobbled together from a large number of slightly earlier works. The source information in the following table is summarised from Mallinson’s Hatha Yoga and the descriptions of the poses are from the HYP (tr.  Pancham Sinh, 1.18-58)







Palms to ground, navel on elbows, raise body straight


Also found in in 10th C Vimanarcanakalpa, Padmasamhita, and Ahirbudhnyasamhita


Right ankle left of anus and v.v.


Padmasana, then hands through thighs and raise on palms


Sit with legs out, grasp toes, head to thighs

Sivasamhita (at par. 92 called ugrasana)


Kukkutasana, then hands cross behind back to grasp neck, then lie on back



Grasp toes with hands, draw to ears, bending like a bow.


[Right foot on left thigh, sole up. Left foot outside right knee. Twist to left. Right hand grasps right foot. Left hand behind back.]



Sit with hands under thighs, body straight



Vasisthasamhita, Yogayajnavalkya, and Sivasamhita


Right ankle to left side and v.v.


Left foot on right thigh and v.v.


Right foot to root of left thigh, right hand behind back to grasp toe and v.v.


Lie on ground like a corpse


Left heel to perineum, right above member, chin to chest


Right foot on left thigh, right hand behind back to grasp toe and v.v. Chin to chest.

OR feet on thighs, soles up, hands on thighs, palms up. Chin on chest, tongue on root of upper teeth.


Left heel to right of perineum and v.v., hands on thighs


Left heel to left of perineum and v.v., hold feet together with hands

I’m a little confused by this, however. The HYP doesn’t list the Matsyendrasana (listed by Mallinson) by that name, and if it is an alternative name for the Matsyasana, then the former shouldn’t be listed amongst the non-seated postures. (The description of the posture given is from Hewitt, J. (1977) The Complete Yoga Book. I have no idea of its accuracy; certainly its description of Matsyasana is nothing like that given in the HYP.) In fact, it’s not at all clear to me how Mallinson has determined that some are seated and others non-seated.

In any case, the textual sources that follow begin to increase the number of taught asanas that are accepted as canonical. the Yogacintamani (early 16th C) describes 35 asanas. (A manuscript of the same text dated 1660 lists 110 asanas and describes fifty-five ) The Gherandasamhita (ca. 1700 CE) teaches 32 asanas, and, eventually, the Hatharatnavali (early 18th C) taught 84 asanas – the number that the HYP had declared were taught by Siva (1.35). (Clearly this doesn’t refer to the Sivasamhita, because that only describes four postures.) The same number is also taught in the roughly contemporaneous Jogpradipika (1737) of Jayatarama.

The Gherandasamhita is now supposed to be one of the core texts of Hatha Yoga, so the list of taught postures in it is quite important. It includes all the postures listed for the HYP except the Kukkutasana (although the descriptions can sometimes be different, as for the Viasana) as well as the following:




[A version of Guptasana?]


[Virasana on one side, the other leg extended]


Padmasana lying on the belly


Similar to Padmasana but with hands covering the heels, palms facing upward


Staying on tip toe, heels not touching the ground, the buttocks rest on the heels


Similar to Gomukhasana but sitting on the heel

Mandukasana Virasana

With knees apart from each other


An inverted Mandukasana


[Stand, arms overhead, one foot to root of thigh]


Virasana with the hands on the knees


One leg in Mandukasana, the other bent on the floor with heel touching the perineum



Prone, arms extended, lift legs and arms


Lying on the stomach, legs apart


Similar to Dhanurasana but with legs crossed


[Prone, raise torso, arms bent in front.]

Yogasana Padmasana

With hands on the knees


*Note that the descriptions here are from various sources. I haven’t actually been able to gain access to a copy of the Gherandasamhita.

On the other hand, Sjoman (op. cit. p 38) notes that all these later texts seem to embody a literary tradition rather than a tradition of practice coming from the HYP. In fact, he says of the modern practitioners: “their practices have no real textual justification and there is no continuous tradition of practice that can be traced back to the texts on yoga.” And all scholars are agreed that there is no connected tradition, either philosophical or practical, joining Patanjali’s YS to the mediaeval texts and practices that really began in the 11th C. So we have at least two gaps in the tradition before we get to the modern period.


  1. 2 Responses to “The Yoga Asanas before the Modern Period”

  2. That’s a really good article. You did a good job!

    By Rajesh on Jun 23, 2015

  3. Thanks. I found it interesting. And, just as a curious side note, I’ve just watched a British documentary in which actor Dominic West and his good friend Dr Sir James Mallinson went to the Kumbh Mela at Allahabad, where Mallinson was made a mahant by his guru Babaji of the Thirteen Renouncers. I didn’t even realise who it was until I saw the name in the credits.

    By SteveGW on Jul 28, 2015

Post a Comment