Everyone who studies Sanskrit deep enough knows the rule of reading a Sanskrit sentence: “first think through the phrase syntactic structure, then proceed to translation of every single word.” And never start interpreting from the first word unless you have comprehended the meaning of the entire sentence.
Paradoxically this rule might be also applied for dealing with the whole text, be it translation or commenting on it. You first conceptualize the entire text structure and logic, and then see into separate lines. And never start from the very beginning… In fact, this is a standard rule for reading any complex text in foreign languages. Yet it is somehow omitted when it comes to handling Sanskrit sutras.
The Hindu commentary tradition assumes that a text is thoroughly analyzed in the line by line manner from the beginning to the end. The problem of understanding text logic and inner structure, not to speak of its deconstruction, simply goes beyond the scope of established Hindu methodology. While the idea of singling out the text historic “layers” seems to be a blasphemy: a sutra is traditionally treated as a sacred text that has its inner completeness, perfection and harmony (even in case it obviously doesn’t), and all we need is to find and comprehend them with the help of various intellectual ingenuities.
It thus develops into religiously-scholastic manner of old texts interpretation that “hypnotizes” also European scholars, so that they also start reading a text from the beginning and by default treat it as an integral piece. This principle of text work has considerably predetermined the structure of my blog on Yoga Sutras. I started commenting on the text ab initio, from the first sutra on. However, the farther into the text, the more I realized how confining this approach is. I have already mentioned in the opening articles that Sutras have many discrepancies and inconsistencies, and I could not ignore them any longer. It became necessary to study the structure and inner logic of the text. And quite unexpectedly the issues under consideration have morphed into the question of Sutras authorship and dating…
So I shall give a brief outline of this study results. I guess it can take a couple of articles, but it will make a good contribution into the text analysis further progress.
What do we expect from any integral text (notwithstanding a cultural tradition), written by a single reasonable person and claimed to give a comprehensive perspective on the subject?
1. Consistency of terms and notions meaning
2. Absence of internal contradictions
3. Narrative continuity
4. Consistency of style
5. Uniqueness of definitions
6. Homogeneity of enumerations
7. Correspondence between the structure of the text and its logical structure.
I shall now explain:
1. The consistency of terms and notions meaning assumes that specific words when repeated throughout the text are used in more or less same sense. This point is essential, since one word can have different meanings in different texts. For instance, in Medieval treatises on yoga the word “siddha” may refer to “supernatural powers”, while the texts on Nyaya may use it to define something that was derived with the help of a text or an evidence. Of course, experts in both of the fields know the ins and outs of the terms usage. Also, there are some words that are so polysemantic that their usage in different meanings is possible even within one sentence. Like “artha” or “yoga”, for instance. But these are, of course, exceptions.
2. The absence of internal contradictions. One can hardly suggest that ancient sages had problems with logical thinking. Those who just tried to read Nyaya Shastra understand they were brilliant at logic. Neither they can be blamed for forgetfulness. Since we know that four thousand sutras of Panini’s Ashtadhyayi – a complex system of hyperlinks and abbreviations – were traditionally learnt by heart. As well as Vedas: for thousands of years they were passed on in oral form only. So you wouldn’t assume that by the end of a sutra one forgets the subject he was previously writing about.
3. Narrative continuity. The term “sutra” means “a thread” and thus suggests that all lines (that are, by the way, also called sutras) are logically related to each other. A category is first introduced, and then it is elaborated and elucidated. And so on.
4. The uniqueness of definitions. Sutra always gives definitions. Obviously, being defined once, a word doesn’t need a second definition.
5. The homogeneity of enumerations. I shall explain further what this point means.
6. The correspondence between the structure of the text and its logical structure, i.e. it would be logical to have a new theme introduced in each new chapter.
Now, Yoga Sutras do not meet any of these criteria.
1. The text does use same words in essentially different contexts.
2. Some terms are double-defined (like samadhī or kaivalya).
3. The text has “orphans” – the lines that are not related to the narrative thread.
4. And there are “orphan” fragments as well.
5. The text has contradictive recitals.
6. There is confusion in terms hierarchy levels.
7. Different fragments of YS describe radically different mystical experience and are aimed at solution of different problems.
All these aspects can be noticed only while working with original text, as the translators smooth out the problems unconsciously. For instance, remember my mentioning here the case of the word “kleśa”: used in a bit different context in different parts of the text, it was translated [into Russian – transl. note] by Ostrovskaya and Rudoi with the help of different words. The same happens to “kaivalya” and some other notions.
Disclaimer: hence starts a complex part of the article. Should the reader feel lazy about going deep into details go directly to the resume. Otherwise welcome to read further…
Let us analyze the examples of every listed inconsistency:
Violation of unity in the notions meaning.
The meaning of some terms is not the same in different parts of the Sutras. For instance, smṛti. The word “smṛti” is easy for analysis: it consists of the root smṛ - to remember, to recall - and the suffix -ti that builds a feminine abstract noun. [Probably we can say that] it is similar to the –ity suffix in “unity”, “equality” etc.
The term “smṛti” is first used by the author in his enlisting the types of vṛtti in the sutra 1.6. Let me remind it to be:
The [vrittis] are: pramāṇa, viparyaya, vikalpa, nidrā and smṛti.
Later, in the line 1.11, Patanjali defines the term as:
Anubhūta-viṣayа-аsampramoṣaḥ smṛtiḥ ॥1.11॥
i.e. “not-losing”, “not-forgetting” (asampramoṣaḥ) the experience (anubhūta) of the (perceived) objects (viṣayā).
Smṛti stands for not-losing the experience of perceived objects. That is, in the given context smṛti is a type of vṛtti, something that we try to get under control by means of yoga that is yogaś citta-vṛtti-nirodhaḥ. However, in the line 20 of the same chapter we see the following:
śraddhā-vīrya-smṛti -samādhi-prajñā-pūrvaka itareṣām ॥1.20॥
To attain a certain artful state, “for others” (itareṣām) it is necessary to use faith (śraddhā), valor (vīrya), samādhi and “smṛti.”
Here we may see the listed features - śraddhā, vīrya and samādhi – to be positive states of mind that are essential for attaining a new state. And smṛti is listed among them! Yet the previous lines have qualified smṛti as a vṛtti that we are trying to get control of! That is, in this line smṛti is used in a radically different context. It is given in the meaning it is used by Buddhists – “remembrance”. The way it is given in Satipatthana Sutta – a treatise that is in fact dedicated to this particular aspect (sati is a Pali version for pronouncing smṛti). That is, smṛti is a notion that can be at a stretch interpreted as “consciousness”. And it is not what one must try to get rid of. It is rather on the contrary.
Let us look further. The sutra 1.43 reads:
smṛtipariśuddhau svarūpaśūnyevārthamātranirbhāsā nirvitarkā ॥1.43॥
When memory is emptied etc. (smṛti-pariśuddhau svarūpa-śūnya) there comes “nirvitarkā samapati”.
In this case we speak about emptying smṛti as a kind of memory reservoir, for reverting it back to its genuine form (svarūpa). This is the third meaning of smṛti. And finally, in the chapter 4 smṛti is used to denote “memory”, nothing but plain memory.
To sum it up, we see that in Yoga Sutras the word “smṛti” is used in four different contexts as a minimum. It would have been normal if it was a “minor” notion. But this term was defined in the sutra 11. So in fact the author has introduced a definition that he never then used. Something went wrong...
The second example is the word “asmitā” or “I-am-ness”. Asmi stands for “am”, while –tā is a secondary suffix producing an abstract feminine noun. “I-am-ness” does not, of course, sound harmonious, but the Sanskrit form – asmitā – is nice.
And now, on the one hand we know it to be a kleśa. Remember the line 2.3:
avidya- asmitā-rāga-dveṣa -abhiniveśāḥ kleśāḥ ॥2.3॥
And in this meaning it is what we are fighting with, trying to burn its seeds down and so on. Yet it then goes differently in the line 17 of the first chapter:
vitarka-vicāra-ananda-asmitā-rūpa-anugamāt samprajñātaḥ ॥1.17॥
Consciousness (saṃprajñātaḥ) results from consistency of: opinion, analysis, admiration and experience of I-am-ness.
The asmitā mentioned in this line is obviously not a kleśa. Because vitarka, vicāra and ananda are in fact stages of attaining a specific state, and asmitā here is also used to denote a stage.
Another case to consider is the word abhyāsa – the exercise. The word breaks down into abhi-āsa, where the root “as” means “to throw”, and abhi- is “toward”; that is, “multiple throwing”. The term has in fact assumed the yogic meaning of exercise and practice in form of multiple repetition. The line 12 of the Sutras says that:
abhyāsa-vairāgyābhyāṃ tan-nirodhaḥ ॥1.12॥
Then a definition of abhyāsa is given, and it tells abhyāsa is:
tatra sthitau yatno'bhyāsaḥ ॥ 1.13॥
Here yatna means “effort”, while sthiti is a “sustained effort”. In this way, “abhyāsa is a sustained effort.” However, 19 lines away we read that he who wants to hold something needs “abhyāsa with one object” (eka-tattvа-аbhyāsaḥ). Previously abhyāsa was meant to work with vṛtti that are subject to nirodha (tan-nirodhaḥ), and here the object of abhyāsa is said to be one object (eka-tattvа) out of those listed further in the lines 1.33 – 1.40. Thus, in this situation the word “abhyāsa” is used in a different context.
Inaccurate definitions. Sutras are meant to give definitions. At the end of Yoga Sutras’ first chapter we find a definition of “samāpatti”. A very nice definition it is:
Kṣīṇa-vṛtter abhijātasyeva maṇer grahītṛ-grahaṇa-grāhyeṣu tat-stha-tad-añjanatā samāpattiḥ॥1.4॥
Samāpatti is a state in that conscious mind is coloured by cognizing, cognition and the subject of cognition, just like a gem is given an additional quality by its base it was placed on. A transparent diamond looks red when placed on a red base, and looks green when placed on a green one.
Further on two more variants of samāpatti are defined. But two lines after the text goes:
tā eva sabījaḥ samādhiḥ ॥ 1.46॥
That is, this samāpatti is the samādhi.
Consider this: we give a complex definition, where we say “this is A”. And then immediately add that “A is B”. From the position of logic this is a very poor syllogism. Why don’t we say from the beginning that “this is B»? Why introduce additional category?
This runs very much contrary not to Indian logic only, but to common-sense logic in general. Unless we assume that the lines on samāpatti were added later and have thus disrupted the initial definition of samādhi due to some polemic purposes. It’s like “what you call samāpatti is our samādhi”.
Obvious logical inconsistencies. These are a bit more difficult to understand, but nevertheless. For instance, the line 1.16 represents that the supreme state of vairāgya is when one dwells in the state of guṇa-vaitṛṣṇyam. That is, the word “tṛṣṇā”, “vaitṛṣṇā” means “disengagement”, disengagement from all guṇa(s). While the last line of the third chapter states that
sattva-puruṣayoḥ śuddhi-sāmye kaivalyam iti ॥1.55॥
When the purity of sattva becomes equal to Puruṣa [there comes] kaivalya.
But wait, wasn’t the author previously claiming it to be disengagement from all guṇa(s), the three of them? Then what made him mention sattva here? Looks like a contradictive case.
Another example. The line 2.25 reads that “Viveka [helps to] eliminate different forms of citta that are called pariṇāma(s).” But in the line 3.9 the category of “nirodha-pariṇāma”, i.e. “nirodhic parinama”, is introduced. That does not need elimination. It is rather the goal. Another contradictive case.
Violation of enumerations hierarchy. The whole of the chapter 3 – or, rather, chapter 3 and a part of chapter 2 – sūtrakāra (i.e. the creator of the sutra) explains the “aṅga (s)” of yoga, with samādhi named as one of them. On the other hand, there is an aṅga called niyāma that consists of tapas, śauca, īśvara-praṇidhāna, svādhyāya etc. That is, there is the first level of recitation hierarchy: yāma, niyāma, asana, prāṇāyāma, pratyāhāra, dhāraṇa, dhyāna, samādhi. And there is the second string of aṅga(s), i.e. sub- aṅga(s). One of them is tapas.
Nevertheless, in the line 1 of the chapter 4 they are placed side by side:
janmauṣadhi-mantra-tapaḥ-samādhi-jāḥ siddhayaḥ ॥4.1॥
i.e. siddhi (siddhayaḥ) emerge (jāḥ) as a result of emergence (janma) of medical herbs (оṣadhi), mantra(s), tapas and samādhi.
Out of sudden, tapas and samādhi have turned into proportional notions. This exemplifies the case of recitation hierarchy violation.
Designation of the same phenomena by different words. In the sutra 1.30 the author introduces a specific term to denote scattering of mind - citta-vikṣepa. As to the opposite state, the line 1.33 tells it to be citta-prasādanam, while in the sutras 3.11-3.12 the same is referred to as citta-ekāgratā. Why so? That’s unclear.
Or here is an even better example. The end of the chapter 2, praise to prāṇāyāma, says that:
dhāraṇāsu ca yogyatā manasaḥ ॥2.53॥
Then manas by virtue of pranayama becomes applicable, or suitable (yogyatā), for dhāraṇa, (or, to be more exact, “for dhāraṇa(s)” (plural form is used)).
But in a couple of lines another definition of dhāraṇa reads as follows:
deśa-bandhaś cittasya dhāraṇā ॥3.1॥
Dhāraṇais the state of keeping citta [focused] on an object.
Not manas but citta! That is, we were speaking about manas in the first case, and then did the same with citta in the second one. But in actual fact, manas and citta in Indian tradition are used to denote slightly different notions. And even more: these terms are used in scope of different systems.
Here is another example. I have already mentioned vāsana(s) and saṃskāra(s). Now, how come we work with saṃskāra(s) the whole chapter 2, but then shift to deal with vāsana(s) in the chapter 4? Given that these two are almost equal notions. And I don’t even inquire about fundamental difference between vṛtti(s) and kleśa(s). Because the part of the chapter 2 that tells about elimination of vṛtti is in fact substituted with elimination of kleśa(s). But we do have very exact lists of vṛtti and kleśa(s). And from the definitions of vṛtti(s) and kleśa(s) we see these to be different notions. Yet another puzzle riddle.
Duplication of definitions. The Sutras has four definitions of kaivalya. And they unquestionably are definitions: the word “kaivalya” is placed at the end of the line which marks here a definition. But why give four definitions of one term throughout one sutra? Moreover, two of them are placed almost in a row, with only five lines in between. Has the author forgotten his having already defined the term earlier? There are also two definitions of samādhi, and so on.
“Orphan” lines and fragments. Here is the line 27 of the chapter 2:
tasya saptadhā prāntabhūmiḥ prajñā ॥2.27॥
i.e. “this is the way to attain seven-stage wisdom”
It leaves us curious and waiting for detailed explanation of the mentioned seven stages and the wisdom. Yet the reader’s interest is ignored by the author. Moreover, this line is not linked to the previous one; as if it hangs in the air, a “seven-stage wisdom” – and that’s it. This is not typical for sutras style. If sūtrakāra writes “five vṛtti(s)”, he explains what these vṛtti(s) are. If he mentions kleśa(s), he describes five kleśa(s). If there are “seven stages”, there should be some explanation of what these stages are. But there is none.
There are even fragments that are “orphan”. For instance, the one about īśvara-praṇidhāna. The sutra 22 of the chapter 1 describes the types of students that are:
mṛdumadhyādhimātratvāt tato'pi viśeṣaḥ ॥1.22॥
[there are] also difference because of [qualities that are] poor, moderate or excellent (literally: the difference because of poorness, moderateness and excellence)
i.e. mṛdu is a flabby one, madhya is mediocre and adhimātra is active. And they differ in their tempo of attaining the state of yoga (citta-vṛtti-nirodha). Then suddenly the line 23 reads as:
Īśvarapraṇidhānād vā ॥1.23॥
i.e. “or as a result of īśvara-praṇidhāna”
How is this related to students? Did the author forget about them? Went over the next five lines in a row, and then came back to this point again?
On top of that, these lines about īśvara-praṇidhāna have no logic at all. We deal with this topic neither before, nor after them. They are a “close-loop system”, an obviously alien infusion that is followed by the main text. Moreover, we know Yoga to be originally an atheistic tradition. And though you can, of course, insert a “religionizing” fragment into the text, but it still won’t have any logical link with the rest of the lines. This is what happened in fact.
There is one greater example of the “orphan” fragment in the text of the chapter 3 that continues the narrative thread of the chapter 2 about the aṅga(s) of yoga. The author gives a nice definition of dhyāna. Then, a brilliant definition of samādhi (a second one, by the way). And then he defines saṃyāma as a unity of dhāraṇa -dhyāna- samādhi. Now, what do we expect to read next? The explanation of how this saṃyāma can be used, of course! What we read instead is a mind-blowing philosophical fragment dedicated to relations between substance and object, dharma and dharmin, confusion of categories and so on. It takes 12 lines and results in the following statement:
pariṇāma-traya-saṃyamād atītānāgata-jñānam ॥3.16॥
the saṃyāma [practiced] upon this or this results in that.
And we obviously feel this fragment to be an odd one, an insertion of some later period. This is what I call an “orphan” fragment.
Use of terms before giving them a definition. The Sutras introduces the category of “tapas” in the 2nd chapter only, though it starts using it already in the chapter 1. A detailed definition of the word “samādhi” is laid down in the chapter 3, but we first see it at the beginning of the chapter 1, which does not comply with the logic of sutra style. Sutra first introduces a term, and then uses it.
Funny thing is that there are explicit inconsistencies in the neighboring lines. Like, for instance, a truly funny and very explicit contradiction between the lines 2.15 and 2.16.
Pariṇāma-tāpa-saṃskāra-duḥkhair guṇa-vṛtti-virodhāc ca duḥkham eva sarvaṃ vivekinaḥ ॥15॥
Indeed, for a wise man everything is suffering – because of exposure to continuous change, anxiety, saṃskāra(s), as well as due to contradictory development of guṇa(s).
Well, this is a nice Buddhist idea: “everything in life is suffering.”
But we then read the next line:
heyaṃ duḥkham anāgatam॥2.16॥
The suffering (duḥkham) that has not yet come (anāgatam) is for elimination, or must be eliminated (heyaṃ - future passive participle (prescriptive gerundive) is used).
But wait, if everything is suffering, then how can suffering be eliminated? This is a contradiction between two neighboring lines. An Indian logician could not have allowed such a lapse.
So hence comes the question: why? How could this have happened? Provided that we don’t assume that people who were writing sutras were mentally incompetent. How could have they tolerated these inconsistencies? And the answer is:
Yoga Sutras is not an integral text.
It is not a text that was written by one person at one time; it is a text that was being composed piece by piece. Also, it consists of several fragments, and each one of them is absolutely consequent internally, meets all criteria of a sutra, has its own terminology, its range of problems and internal logic. In fact, all essential elements, even its own conclusion. These fragments come from different esoteric, mystic traditions that though allied, use essentially different systems of terms.
Moreover, some lines are inclusions of later periods. Most of them are borrowings from Buddhism. These inclusions can be found and their heterogeneity can be proved, since there are sutra-texts containing the same exact lines, and they [these lines] are harmoniously entwined into those texts.
And these were commentators who “sewed” them all together. The problem of commentators in hermeneutic tradition of India is that they were trying to find text’s internal logic even in case there wasn’t any. And by the way, when we read the commentaries we understand that Vyāsa and Bhoja reacted to different traditions.
I have already mentioned that each part describes specific mystic experience and corresponding mind techniques. These parts are evidently of different age, and this point complicates the problem of Yoga Sutras dating. I believe the time gap between various parts to be ca. 700 years. It was a long process of the text construction by different traditions with widely different approaches.
What are these Traditions and parts of Yoga Sutras? How can they be disjointed correctly and conclusively? This is what the following articles shall be about.
To be continued…