An Upaniṣad describes the soul at liberation as a bird escaping a net that has been cut.
A Buddhist text describes the same moment as a bird escaping its cage.
Put those side by side and almost everyone does the same thing: nods. Same picture, near enough. Trapped creature, sudden opening, flight. Two traditions, no contact, one image — and the coincidence feels like evidence. This is how most cross-tradition reading works, and it is how a great deal of comparative religion gets written.
Bodri says they are not the same phenomenon. Not similar-but-differently-emphasised. Different attainments, described by images that happen to converge.
And then he says the thing that should stop you cold: "if no one told you this, there would be no way for you to know, because the descriptions sound so similar."1
Notice what makes that sentence dangerous rather than merely cautionary.
If two accounts of different states are describable in the same words, then matching the words gives you no information at all. Not weak information. None. The resemblance you detected is a fact about the available stock of metaphors — cages, nets, birds, light, dissolving, oceans — which is small, and which every tradition draws from because human beings have roughly the same things to compare experiences to.
Bodri gives a second case that is harder to wave away, because it is technical rather than poetic. The yin shen of the Taoist schools and the "thought-born body" of the Buddhist ones are described almost identically: a subtle body, separable, capable of travel. On his account the thought-born body arises only by purifying the sensation skandha, while the yin shen can be had without that.1 Same description, different production route, different rank — and no vocabulary in either tradition marks the difference.
So the practitioner who has read both, and who experiences one, will confidently name it the other. Not through carelessness. Through having done exactly what a careful reader is supposed to do.
The awkward part is where the warning appears: in a book whose entire method is matching traditions to each other.
Chapters one and two consist very largely of the operation Bodri here declares unreliable. Mahāmudrā's four yogas matched to five Mahāyāna stages. Sikh khaṇḍs matched to the preparatory path. Jing, chi and shen matched to the three guṇas and to sat-chit-ānanda. A twelfth-century Cistercian's "air transforming into sunshine" matched to a Taoist energetic transition.
Every one of those is a lexical match. Every one is the move the two-birds passage says cannot be trusted.
This is not a gotcha, and the page should not be written as one. It is the book's own most serious objection to itself, stated by the author, in his own voice, more precisely than a critic would put it — and then not acted on. What it means practically is that the correspondence work in this source comes in two grades, and a reader has to sort them: the phenomenon-against-phenomenon matches, which survive the warning, and the vocabulary matches, which do not.
He does supply the criterion that would sort them — compare results, not terms — and he supplies it after most of the matching is already done. See Buddhism as a Comparative Reference Grid for that criterion and how consistently it gets applied.
There is a mechanism underneath all this, and it is the part of the page with the longest reach.
Bodri argues that specialist cultivation vocabulary decays — not because languages change, but because practitioners stop reaching the states the words name. A term is coined by someone describing something they experienced. It gets transmitted. The people transmitting it have not had the experience. They still use the word correctly, in the sense of using it where the tradition uses it. But the word has quietly stopped pointing at anything they know.
Nothing announces this. The tradition's texts are unchanged. Its teachers are sincere. Its vocabulary is intact and in constant use. What has gone is the referent, and referents leave no gap behind them.
This is why he thinks the two-birds problem is getting worse over time rather than better. The more a tradition's terms float free of their states, the more freely they can be matched to anyone else's floating terms — and the matching will feel increasingly easy and increasingly convincing.
To his credit, Bodri does not merely claim this. He stages it, and the staging is the best thing in the argument.
He prints a 1973 English rendering of a dense technical passage on the parts of consciousness — the kind of translation produced by a competent scholar working from a text whose subject matter they have not experienced. Then he turns to the reader and asks a direct question: "Can you understand this translation?"2
You cannot. Nobody can. The words are all English and the sentences parse and the meaning does not arrive.
Then he prints a modern rendering of the same passage beside it, and the meaning arrives.
The demonstration makes a point that an assertion could not: the failure was not yours, and it was not the original text's. It sat in a chain of transmission where each link was faithful to the words. Fidelity to vocabulary is not the same as transmission of meaning, and the first can persist for centuries after the second has stopped.
You are reading two accounts and they seem to be describing the same thing.
Before you write that down, do this.
Strip both to phenomena. Not what the writer called it — what they said happened. Was the breathing affected? For how long? Eyes open or closed? Was there anything the following day? Most of the time this step alone dissolves the match, because one of the two accounts turns out to specify almost nothing. That is not a failure of the account. It is a finding about how much you had been supplying yourself.
Then hunt for a distinguishing observation. Ask: if these were different states that happened to be described alike, what would tell them apart? Bodri's own answer in the yin shen case is a production criterion — did it arise by itself, or was it manufactured by technique? You may not have his answer. But if you cannot name any observation that would separate the candidates, you do not have an identification. You have a resemblance.
Write the smaller claim. "Resembles" is a true sentence. "Is the same as" is usually not, and it is the sentence that gets quoted forward by everyone who reads you afterwards, stripped of your hedges.
And check the direction of the pull. The two-birds problem does not distribute evenly. It bites hardest when you want the traditions to agree — which, if you are reading comparatively at all, is most of the time.
Set against the failures, Bodri produces one cross-tradition identification that shows what the strong version looks like — and it is worth examining precisely because it is the same author doing the same operation well.
At a low stage of practice a sweet secretion is reported in the mouth. He names it across traditions: soma, ambrosia, madya, the Sufis' sweet wine, amṛta, the grail wine, the fountain of youth.3
That looks exactly like the lexical matching the warning forbids. What makes it different is what he attaches.
First, a phenomenon with a public check: it is reportedly tasteable by a third party. Whatever else is going on, this is not purely interior, and that puts it in a different evidential class from "a sense of oneness."
Second, an explanation for the convergence that does not require transmission between traditions. His argument is that the correspondence exists "not because it's been a popular symbol to pass along from school to school, but because it's such a low stage of achievement" — i.e. it is common because it is early, and early things get reached by many people independently.
Third, and most unusually, a research programme. In a footnote he proposes where to look: amylase concentrates in genital tissue, which on his scheme is where the transformed substance originates; alternatively the compound might be a tryptamine, though it produces no psychological effects; and the pineal gland forms not from brain tissue but from cells originating in the roof of the fetal mouth.4 He names what would count as finding it.
None of that establishes the claim. But compare its structure to "the bird in the net is the bird in the cage." One offers a public observation, a reason for convergence, and a way to be wrong. The other offers a shared metaphor. That is the difference the two-birds passage is pointing at, and the fact that both come from the same author in the same book is the clearest available evidence that the difference is a matter of discipline rather than of ability.
The two-birds problem is not only a comparativist's problem. It arrives first for the practitioner, who has to name their own experience — and at least one tradition built a handrail for exactly that, which Bodri quotes at length.
The Śūraṅgama runs through a catalogue of states a practitioner may encounter early on. Seeing inside your own body. The ten directions bathed in gold. Space filled with unclashing colours. Seeing in a dark room as though in daylight. The earth transformed into a pure land. Hearing your family's voices from far away at midnight.
Each entry is impressive. And each one gets the identical verdict, repeated verbatim after every single item: "This is a temporary stage and is not sainthood. If you do not confuse this with sainthood, it is an excellent level of achievement."5
Read the structure of that sentence, because it is doing something unusual. The state is not debunked and not diminished — it is called excellent. What is policed is the register of the claim made about it. The same experience is an achievement or a disaster depending only on what the person concludes.
That is the two-birds guard, installed inside a tradition, aimed at its own practitioners rather than at rival schools. It does not require the practitioner to distinguish their state from a similar one — which they cannot do — only to refrain from upgrading it. The refrain is the instrument, and its repetition is the design: you cannot read the passage without the correction arriving each time the temptation does.
Strongest support. The translation demonstration is the load-bearing evidence, because it is a demonstration and the reader participates in it. The two named cases — the birds, and yin shen versus thought-born body — are both stated by the source with the distinguishing criterion attached in the second case, which is what a good example of this problem looks like.
The tension the source does not close, and it is structural. Bodri warns that vocabulary matching is unreliable and does it constantly. He warns against making "hard and fast determinations" in the middle of a chapter of hard and fast determinations. He never reconciles this, never flags a particular match as one of the unreliable ones, and never applies his own criterion retroactively to the matches he has already made. [n≥3 PATTERN] on both sides.
A second, quieter tension. If terminology decays because practitioners stop reaching the states, then the warning applies to Bodri's own vocabulary as much as to anyone's — and his defence against that is his asserted attainment, which is the one warrant a reader cannot check, in a book whose selling point is that attainment claims are checkable.
Open questions. How would you distinguish a genuine cross-tradition convergence from the small size of the human metaphor stock — is there any test besides the phenomenon-against-phenomenon one? And if referents can drop out of a tradition without leaving a gap, what would a tradition's own early-warning system for that even look like?
This page is the reason the vault's own attribution error happened, and the two make an unusually clean pair.
The Bhūmi-Kāṇḍa Correspondence System carries the four preparatory sub-stages, which reached it through an intermediary corpus where they had been re-badged as devotional stages. What dropped out in transmission was not a citation but a doctrinal-tradition marker: a Buddhist scheme of the path of preparation, arriving somewhere else wearing different clothes, with nothing in the vocabulary to signal the change.
That is precisely the mechanism Bodri describes, observed from outside a tradition by this vault while he was describing it from inside one. Neither observation knew about the other. One mechanism, two independent sightings — and it is the strongest generative material this source produced.
Against Ālaya Equivalents Across Schools, this page is the standing caution. That page's whole business is cross-school identification of a technical term, which is exactly the operation most exposed to the two-birds problem. Bodri's own contribution there is a scope criterion rather than a lexical one — he rejects identifying the ālaya with Jungian archetypes or a collective unconscious on the grounds that it holds the seeds of the material world and of shared environments, so nothing indexed to the psyche can be it. That is the right shape of argument: a property one candidate has and the other cannot.
Words that outlive their referents — Decline Mechanisms Hub. Most accounts of institutional decline track something visible going wrong: capture, corruption, drift, a shock. This mechanism is worse because nothing goes wrong at all. The texts are preserved, the terms are used correctly, the transmission is faithful — and the thing being transmitted has quietly become a word. The insight the pairing produces: a vocabulary is a lagging indicator of a capability, and it lags without limit — which means the standard diagnostic for institutional health, is the tradition still teaching what it always taught, cannot detect this failure by construction. What would detect it is Bodri's own harder test: measure the institution by the states its members actually reach. That is a much less comfortable audit, and it is available to any field with a claimed output.
The metaphor stock is smaller than the phenomenon space — Mysticism and Numinous Experience Hub. Comparative work on mystical experience runs on reported descriptions, because that is nearly all the evidence there is. The two-birds problem says the evidence base has a systematic bias built in: human beings describing interior states reach for a small shared set of images, so convergence in description is expected even if the underlying states are various. That inverts the usual inference. Where cross-cultural agreement is normally treated as the strongest available support for a common core, this argument makes it the weakest kind of support — unless the agreement is on phenomena with public markers rather than on imagery. The practical consequence is a ranking of evidence that the field could adopt without accepting any of Bodri's metaphysics: reported bodily change and duration outrank reported imagery, always, and a convergence of imagery alone is closer to a fact about language than a fact about experience.
Sharpest implication. When two traditions describe a state in the same words, the shared words are the weakest evidence available that they mean the same thing — and they feel like the strongest. Everything else on this page follows from that inversion, including the reason the vault's own mis-attribution happened and went unnoticed.
Generative questions.