alan little’s weblog

yoga sceptic

7th October 2005 permanent link

One of the things in my perpetually unfinished drafts folder is an essay on “materialist yoga”, in which I talk about my belief that many of the effects and benefits traditionally claimed for yoga are real observations of real phenomena, which should in principle be explainable without resorting to some supernatural concept of “god”.

So I’m interested when I see studies that seem to be going in that direction, or entire vast bibliographies of such studies; and I’m frustrated when I see such things being done with what look to me like absolutely basic mistakes.

One thing I immediately and automatically assume is bullshit is people claiming that their religion’s sacred texts, including the particular archaic language in which they were originally composed, are the direct literal word of God. (I once saw a website whose author believed the King James Bible was what God had in mind all along. Why then did it take Him sixteen hundred years’ worth of drafts in other languages to get it right? He must manage His to-do list about as effectively as I do mine) One variant of that claim, distressingly common in yogic circles, is that the sound of the Sanskrit language in Hindu sacred texts has some kind of special effect on consciousness, irrespective of whether or not one actually understands the texts. Hmm.

So this abstract in the bibliography I just mentioned rather jumped out at me:

This study tested the prediction that reading Vedic Sanskrit texts, without knowledge of their meaning, produces a distinct physiological state. We measured EEG, breath rate, heart rate, and skin conductance during: (1) 15-min Transcendental Meditation (TM) practice; (2) 15-min reading verses of the Bhagavad Gita in Sanskrit; and (3) 15-min reading the same verses translated in German, Spanish, or French. The two reading conditions were randomly counterbalanced, and subjects filled out experience forms between each block to reduce carryover effects. Skin conductance levels significantly decreased during both reading Sanskrit and TM practice, and increased slightly during reading a modern language. Alpha power and coherence were significantly higher when reading Sanskrit and during TM practice, compared to reading modern languages. Similar physiological patterns when reading Sanskrit and during practice of the TM technique suggests that the state gained during TM practice may be integrated with active mental processes by reading Sanskrit.

Travis, Frederick, T. Olson, T. Egenes, and H. K. Gupta. Physiological patterns during practice of the Transcendental Meditation technique compared with patterns while reading Sanskrit and a modern language. International Journal of Neuroscience, Jul. 2001, 109(1-2):71-80

It’s unfair to comment on a paper based on the abstract. I would read the whole thing if I could, but the International Journal of Neuroscience doesn’t have online archives. So, thoughts based on what I can glean from the abstract:

A basic factual error in the first sentence doesn’t exactly inspire confidence: the Bhagavad Gita isn’t Vedic. It was composed hundreds of years later than the Vedas, in a language that had changed significantly from Vedic Sanskrit. (wikipedia:Sanskrit) I wouldn’t expect peer reviewers for a neuroscience journal to know this, but somebody who is sufficiently interested to want to do a study like this should.

I’ll give the authors the benefit of the doubt by assuming they know more about EEG readings and measures of statistical significance than they do about the history of Sanskrit. So assuming the effect they say they found is real – what could be causing it?

• Western yoga or meditation students might regard Sanskrit, and not German, Spanish or French, as inherently associated with things sacred or mystical – or just as something connected with their meditation or yoga practice – and respond to it accordingly.
  How could a study control for this? Not easily. It would probably be quite difficult to find a group of study subjects who had meditation experience but wouldn’t recognise Sanskrit (Japanese Zen practitioners?).
• Classical Sanskrit is not a natural language. By the time of the Bhagavad Gita it was well on the way towards being a formal language used only for scriptural and literary purposes; it wasn’t anybody’s everyday language. Perhaps it might therefore have an inherently more regular and calming-to-the-mind structure than vernaculars like German, Spanish or French?
• The Bhagavad Gita, like most ancient sacred texts, was passed down by oral tradition for centuries before it was written down (although not as long as the Vedas were). As such, it must have rhythms and structure that are particularly suited to be memorised, highly evolved to fit comfortably into slots in the human mind. Modern translations produced in post-oral cultures are not likely to have these qualities. Good verse in any language might just be inherently more pleasing to the mind than clunky prose translations, regardless of where the verse came from.
  These things might be recognisable to some degree even if one doesn’t understand the languages concerned. Lots of people listen to and appreciate singing in languages they don’t understand.
• Now I’m really clutching at straws but … Vedic Sanskrit contends with Mycenaean Greek (much less of which has survived) for the title of oldest extant Indo-European language. The early Classical Sanskrit of the Bhagavad Gita is still quite similar to it. Maybe (very tentatively) it therefore has some kind of special resonance for native speakers of Indo-European languages? The study was done in America, so I assume the subjects were native English speakers – repeat with subjects who aren’t native speakers of Indo-European languages. (Those Japanese zen practitioners are looking more and more attractive)

The claim that merely reading Sanskrit texts without understanding them has special effects is a very strong one. It’s more commonly said that reciting Sanskrit texts has a special meditative effect. It’s obvious how that could be:

• Most of the people who can chant long Sanskrit passages from memory are Brahmins who (a) understand the language fluently and (b) have been trained from childhood to memorise these texts. Reciting from memory requires a very high degree of mental focus - pratyahara, and is therefore inherently a meditative activity.
• Some western yoga students (e.g. me) also memorise and can recite a few snippets of Sanskrit scripture. But we associate the language strongly with our yoga practice and, since we don’t understand the texts like Brahmins do, possibly have to concentrate even more to remember them correctly without any clues from the meaning.

I remain sceptical about the alleged specialness of the Sanskrit language. If the language itself, irrespective of the setting and of what the text actually says, really does have the kind of special effect that some people – apparently including the authors of this study – believe, then my hypothesis would be that the effect comes from one or more of the following, in decreasing order of probability/importance:

1. people associate it with particular activities & states of mind and react to it accordingly
2. verse from oral cultures has been very highly refined and evolved over many generations to fit comfortably into the human mind
3. classical Sanskrit is an artificial language, produced by many generations of clever people taking an existing natural language and deliberately refining it in the direction of order and aesthetic pleasingness.

No divine intervention required.

related entries: Yoga

all text and images © 2003–2008