Why the commentaries don’t always line up
The Brahma Sūtras are a single, terse set of aphorisms, but each commentator inherits them a little differently. Some read a cluster of sūtras as one continuous argument; another divides that same cluster into two. Some treat a short aphorism as complete in itself; another sees it as the conclusion of a point begun a verse earlier. None of this is an error in any one tradition. It reflects centuries of independent commentarial judgment about how the text itself is structured.
Vedanta Atlas exists to make that divergence visible rather than smoothing it over. To do that, the project separates several layers that are easy to collapse into one: the original source text of each edition; the structural record extracted from it (where each sūtra begins and ends, and what commentary follows it); a shared reference numbering that gives every sūtra a stable address; and an alignment layer that says how one edition’s passages correspond to another’s. Keeping these apart means a disagreement about correspondence never has to be resolved by quietly rewriting the structure or the source. It can simply be recorded as what it is.
What the alignment labels mean
The Atlas compares its editions in pairs, sūtra by sūtra, and each pairing is given one of a small set of labels.
Exact: the passages represent the same corresponding textual unit, even when translation, emphasis, or interpretation differs. Exact does not mean identical wording; it means both editions are addressing the same unit of the text.
Split: one edition treats as a single sūtra what another divides into two or more. Merge is the reverse: material that one edition separates, another treats as one continuous unit. Reordered marks cases where the same material appears in both editions but in a different sequence, often because a commentator has relocated an argument to close out a topic rather than open it.
Tradition-specific means a passage in one edition has no corresponding unit in the other edition or editions currently being compared. It does not mean the passage is necessarily unique across all of Vedānta: as more traditions are added to the Atlas, a passage marked tradition-specific today may turn out to correspond to something in a commentary not yet included.
Unresolved: the evidence does not currently support a confident correspondence. Some unresolved cases may become clearer through additional evidence or future traditions; others may remain genuinely uncertain. Unresolved is a genuine scholarly category, not a placeholder for work still to be done, and it is treated as first-class information rather than a defect to be eliminated.
How uncertain cases were handled
Every correspondence in the Atlas begins as a proposal from an automated comparison of the two texts in a pairing. Most proposals are straightforward matches. The comparison also surfaces the cases where it is not confident (a shorter or longer passage on one side, an unusual gap, a lexical mismatch), and those cases receive direct attention from a human reviewer, who reads the surrounding passages in both editions and records a judgment alongside the evidence, rather than in place of it.
Beyond the cases specifically flagged as uncertain, a persisted, representative sample of the remaining, more routine correspondences was also individually reviewed, and its results were used to support accepting the rest of that population rather than reviewing every single one by hand. Vedanta Atlas preserves the distinction between these two kinds of judgment throughout its records: a correspondence individually reviewed by a person is marked differently from one accepted on the strength of that sample. Neither is hidden or blurred into the other.
Vedanta Atlas preserves the integrity of each tradition while making its complexity navigable. For the editions themselves, see Sources.