This word occurs in the names of a number of poses in Iyengar’s 1966 Light on Yoga where it refers to the open angle between the thighs—or usually does if triangle pose/trikoṇāsana is an exception.
three corner pose
bound angle pose
supine bound angle pose
upaviṣṭa+koṇa+āsana (ṣ = sh)
seated angle pose
pārśva+koṇa+āsana (ś = sh)
flank angle pose
lateral angle pose
supine angle pose
Koṇa does not seem to appear in names of āsanas listed in the haṭhayogic texts. Two of the koṇa poses above are standing poses, which are mostly modern creations and so not going to be in premodern texts. The seated pose baddha-koṇāsana does occur in the texts but under another name: bhadra+āsana, auspicious pose.
Notably, in the list above all but one of the poses includes a differentiating adjective. That is, a descriptor that in effect allows a name to be split between two or more poses.
The pool of such terms in Light on Yoga includes words like:
supta supine, lying on the back
baddha bound, meaning hands or feet are joined
pārśva flank, to the side (ś = sh)
eka-pāda one leg
dvi-pāda two leg
Terms like these were deployed to make new names when the repertoire of poses was expanded in the first part of the twentieth century. A few of them (e.g. baddha, ardha) can be found earlier but are used only rarely.
In the 20th century poses newly imported from Indian wrestling and western calisthenics and gymnastics (or poses inspired by them but newly created) needed names. Variations too were being systematically listed and described in writing more than ever before. The use of numbers (1, 2, 3) was another method of multiplying names. In effect, although new in themselves, these were ways of reducing the appearance of novelty: fewer totally new names needed to be inserted into the roster.
Light on Yoga has 200 or so āsanas.
About 80 of these take their names from external items of one kind or another (things, animals, persons, gods): many of these are old. Around 30 name body parts as the sole indicator (e.g. ‘head pose’): almost none of these are old. Then that base of 120 pose names is enlarged by 27 numberings (e.g. Warrior 1, 2, 3) and by 65 poses using the modifying set given above (e.g. ‘upward bow pose’ alongisde just ‘bow pose’).
Light on Yoga was published in 1966 but exactly when and by who this type of expansion was first carried out is not entirely clear. As the work of Jason Birch and others makes clear, although the early twentieth century was a time of tremendous innovation in āsana practice, it wasn’t innovation out of a fixed tradition: a culture of development existed which then suddenly accelerated.
The 1905 Yogasopana (‘Yoga Staircase’) uses the modern Sanskrit names for headstand (śīrṣāsana, pronounced ‘sheershahsana’) and shoulderstand (sarvaṅgāsana) and not the names found for them in the hathayogic texts. It may be that the particular forms now regarded as basic for this pair of poses were influenced by European gymnastics. This may have led to the name shift. In any case some degree of reworking in the naming scheme was already under way even before the influx of new poses.
Mysore in the 1930’s and 40’s in Krishnamacharya’s “yoga hall” was a site of experiment in yoga practice. It was closed in 1950 but the work continued elsewhere. First comes innovation, then only later, perhaps much later, codification. Krishnamacharya and his students Jois and Iyengar (along with several others) all eventually produced books setting out repertoires of poses: they largely use the same pose names, but not entirely.
But the Mysore people were not the only ones who put out expanded repertoires and not all modern yoga passed through Mysore.
In 1960 Swami Vishnudevanda published The Complete Illustrated Book of Yoga (with about 150 poses) and in 1970 Swami Yogeshwaranda published First Steps to Higher Yoga (with 250 poses). The use of “Swami” (Sanskrit swām+ī ‘self-possessor’) followed by adopted Sanskrit names tells us we are a different milieu from Iyengar and Jois—who though they took on some guru trappings kept their own surnames and are not draped in orange. The Swamis may appear (and claim) to be more traditional than Jois and Iyengar, but in fact they too are a distinctly modern development, just a different one. Publishing in English is already an indication in itself.
Both of their books contain large sets of named and photographed āsanas, including modern standing poses. In general their poses are similar to those in Light on Yoga but the Sanskrit names are often completely different. There are for instance many ‘bird’ poses but often different species from those in Iyengar and/or applied to different poses.
They show however a very similar mix of simple (old) names drawn from items in the world with (new) analytic and descriptive names, along with distinctions by number. That is, though the details vary the prinicples of name production are the same.
From the 1970 First Steps to Higher Yoga:
Ivengar’s virabhadrāsana 3/warrior 3 (Virabhadra is a mythological name, not actually a word for warrior; though vira means something like hero.)
Iyengar’s marīcyāsana 1: (c = ch)
For those with troubled knees there is this terrifying proposal — going about ‘hither and thither’ on your knee caps; which may possibly provide a basis for a new form of tap dancing:
And some koṇa. The triangle illustrated below has several features that differ from the Iyengar version: the angle of the feet and the angle of the torso out of the hips. This version is older. The horizontal upper arm is an Iyengar variation: one system’s base pose is another system’s variation.
There are unresolved questions in this history: no doubt in part because only some of what happens ever gets recorded in texts; and because for cultural reasons innovators tended to play down their invention. Investigations in this area include N. E. Sjoman’s 1996 The Yoga Tradition of the Mysore Palace and Jason Birch’s The Proliferation of Āsana-s in Late Mediaeval Yoga Texts.