Devahūti has asked about bhakti yoga. Kapila's answer opens by refusing to give one description:
O beautiful one! It has been thought that there are many kinds of paths for bhakti yoga. Depending on their own nature, qualities and sentiments, men have divided it into various branches.1
And then four grades, distinguished by what the devotee is after:
If a person who engages in this has the intention of causing violence, or is insolent, jealous or angry, his insight is differentiated and he is a tamasa devotee. If a person is in pursuit of material objects, fame and prosperity, and worships me in places of worship, his insight is differentiated and he is a rajasa devotee. If a person desires to free himself of karma by dedicating them to the supreme, and if he worships the one who should be worshipped, his insight is differentiated and he is a sattvika devotee.1
Three grades, each with his insight is differentiated attached. Then the fourth, which does not have it:
I reside in everyone's heart. Just as the waters of the Ganga flow towards the ocean, merely by listening to my qualities, the mind continuously flows towards me. This is said to be the characteristic of the nirguna type of bhakti yoga. This is the devotion towards Purushottama that is without any motive.1
Causing violence, or is insolent, jealous or angry — and this is still called bhakti yoga.
That is the scheme's most surprising feature and it is easy to read past. A tāmasa devotee is not a failed devotee, an aspiring devotee, or someone outside the category. He is a kind of devotee, listed first, in a fourfold taxonomy of a single practice.
So a person who worships with the intention of harming someone is inside bhakti, at the bottom of it.
Which means the volume's category is defined by the object of attention rather than by the quality of the person. That is consistent with the claim Uddhava makes elsewhere — that the asuras are devotees because their intelligence is focused, along the path of enmity, on him.2 Here the same logic is not an inference by a grieving survivor; it is a graded scheme delivered by a descent.
His insight is differentiated appears after each of the first three and not after the fourth.
The phrase names the defect: bheda-dṛṣṭi, seeing difference — the same failure Viṣṇu threatens people over in the previous Skandha, where perceiving a difference between himself and brahmins, cows and the refugeless is called impaired vision.3
So all three motivated grades share one diagnosis. The tāmasa devotee, the rājasa devotee and the sāttvika devotee are not failing in three ways. They are failing in the same way, at three intensities, and the failure is that there is still a gap between the worshipper and what is worshipped across which something can be transacted.
Motive requires two parties. That is what makes it differentiated insight.
Look at what the nirguṇa description does not contain.
The first three each specify an action the devotee performs — engages, pursues, worships in places of worship, dedicates, worships the one who should be worshipped.
The fourth specifies no action at all. Merely by listening to my qualities, the mind continuously flows towards me.
The only thing the person does is listen, and the listening is qualified by merely. Everything after it is the mind's own behaviour — flows, in the present continuous, like a river.
The Ganges does not decide to reach the ocean. The image is chosen precisely: a river's motion toward the sea is not a choice, not an effort, and not a transaction. It is what water does on a slope.
Register: this is a reading of the grammatical difference between the four descriptions. The text supplies the four and marks only the nirguṇa one as motiveless. What can be shown is that the first three are described by what the devotee does and the fourth by what the mind does.
The scheme's names are the three guṇas plus one, and the fourth is not a fourth guṇa.
Nirguṇa means without qualities. So the taxonomy runs tāmasa, rājasa, sāttvika — and then off the scale entirely.
Which makes sāttvika devotion, the best of the graded kinds, still a differentiated insight. A person who worships in order to dissolve their karma by dedicating it upward is doing something admirable and is doing something with a purpose, and purpose is what the fourth grade lacks.
The scheme's top item is not the best version of the practice. It is the version that has stopped being a practice — no motive, no transaction, no gap.
And this is where the passage goes next, in the same breath:
Even if salokya, sarshti, samipya, sarupya and ekatva are offered, such people who are devoted to me do not accept anything other than service to me. This is said to be the extreme form of bhakti yoga.4
The motiveless devotee is offered five liberations and declines all five — which is the volume's signature claim, and it follows directly from having no motive.5
Three absences worth marking.
He does not tell Devahūti to move up the scale. The four are described and no instruction to ascend is attached. The passage that follows lists practices — image-worship, chanting, respect for the great, compassion for the poor, yama and niyama — but frames them as my dharma generally, not as a route from third grade to fourth.4
He does not condemn the lower three. The tāmasa devotee is not warned or excluded. He is classified.
And he does not say the fourth can be produced. The Ganges image works against that: rivers do not arrange to flow downhill. The only input named is merely by listening to my qualities, and the volume elsewhere says that unwavering devotion comes only through association with devotees — an environmental condition rather than an achievement.6
So the highest grade is described as a state that arrives, in a discourse that immediately goes on to prescribe a great deal of activity.
This is a four-item taxonomy of the volume's own central practice, delivered by a descent to establish a doctrine, and it is never applied to anybody.
Nobody in the four Skandhas is identified as a rājasa devotee. The asuras are not sorted into the tāmasa grade despite fitting its description exactly. Uddhava's claim about them is made independently, in a different Skandha, without reference to this scheme.2
Which places it with the volume's other stated-once classifications — the ten characteristics that order nothing, the sixteen modes of life that describe no one, the five characteristics of intelligence.7
The Bhāgavata builds taxonomies readily and operates them almost never, and this is the case where the omission is most conspicuous, because a scheme sorting devotees by motive is exactly the instrument its own narratives would need.
You are assessing participation in something — a community, a practice, a team — and the obvious variable is how much people do.
The scheme suggests a different axis and it has a specific payoff.
Sort by what they are after, not by what they perform. The tāmasa and sāttvika devotees may do identical things: attend, offer, observe. What differs is the transaction each is running, and that is what predicts behaviour when the practice stops paying.
Expect the top of your scale to still be transactional. The sāttvika devotee's motive — dissolving karma — is the best available motive and is still a motive. In most organisations the highest recognised tier is the person with the most admirable reason, and this scheme's point is that admirable reasons are on the same axis as bad ones.
And be careful what you claim about the motiveless grade. Kapila describes it and does not prescribe it. Anything you can produce by instruction is by definition motivated; a category defined by the absence of motive cannot be a target without ceasing to be itself. The honest handling is the passage's: describe it, do not require it, and keep prescribing the ordinary practices to everyone.
The failure mode to avoid is the one the scheme invites — using the fourth grade as a standard people are measured against, which converts it into an aspiration and therefore into a motive.
The strongest evidence that the sorting variable is motive is the repeated clause. His insight is differentiated attaches to all three motivated grades and to none of the fourth, which makes the diagnosis uniform across them and absent from the exception.
Tension — a person acting with intent to cause violence is classified as a devotee. The volume neither excludes nor condemns this grade. Preserved.
Tension — the fourth grade is described as arriving unbidden and the same chapter prescribes an extensive programme of practice. The relation between the programme and the grade is never stated. Preserved unresolved.
Open question. Is nirguṇa bhakti here a fourth item on the same scale or a different kind of thing that the numbering conceals? The name says it is off the guṇa scale entirely, and the passage presents it as the last of four.
This vault's Gauḍīya material holds this scheme in developed form — the Bindu and the Bhakti-rasāmṛta-sindhu tradition working nirguṇa or uttamā bhakti as the definitional target, with the motivated kinds as preliminary.
What the primary text supplies that the later systematisation does not is the flatness. Kapila lists four, marks three with one diagnosis, describes the fourth without instruction, and moves on. No ascent, no stage-theory, no method for converting one into another.
The Five Liberations Refused documents the sentence that immediately follows, where the motiveless devotee declines sālokya, sārṣṭi, sāmīpya, sārūpya and ekatva. The two together are the primary-text form of the tradition's central claim, and they arrive as description rather than as doctrine.
With Deficiency Motivation vs Growth Motivation. That page works the distinction between action driven by a lack — needing something one does not have — and action arising from sufficiency, where the behaviour is an expression rather than a repair, and the two produce different relationships to the same activity.
Kapila's first three grades are deficiency-motivated: harm to inflict, objects to acquire, karma to discharge. The fourth is neither — it is a river.
What the pairing produces: the psychological framing keeps both poles inside motivation. Growth-motivated action is still someone wanting something, just wanting it from abundance rather than from lack, and the framework's value is in explaining why the same behaviour feels different from the two positions. The Bhāgavata's fourth grade removes motivation entirely and describes the result as flow toward, with no wanting on either side. Which raises a question the deficiency/growth pair cannot pose: whether the growth pole is a genuine third thing or simply deficiency with the lack better supplied. The scheme's answer is unambiguous — it puts the best motivated grade, sāttvika, on the same axis as the worst and marks all three with the same defect. If that is right, growth motivation is a good position on a scale rather than an escape from it, and what lies off the scale is not describable in the vocabulary of wanting at all.
Sharpest implication. All three motivated grades share one diagnosis — his insight is differentiated — which means the scheme is not ranking devotees by quality but by how much gap is left. Violence, acquisition and karmic hygiene are three sizes of the same gap, and the fourth grade is not better motivated but unmotivated, described in the only grammar available for something nobody is doing: a river going downhill.
Generative questions.