The Arian Witness

The Arian Witness is Krishna Mohan Banerjea’s landmark work of Indian Christian thought. Its full title — The Arian Witness, or, The Testimony of Arian Scriptures in Corroboration of Biblical History and the Rudiments of Christian Doctrine — states its thesis: that the ancient Aryan (Vedic) scriptures themselves bear witness to truths fulfilled in Christ.

Banerjea, a high-caste Brahmin convert and scholar of both Sanskrit tradition and Christian theology, argues that the sacrifices and doctrines of the Vedas corroborate biblical history and point forward to the gospel. He gives particular attention to the theme of self-sacrifice in Vedic religion, reading it as a foreshadowing of Christ’s atonement.

The book advances a ‘fulfillment’ rather than a ‘rupture’ account of the relationship between Hinduism and Christianity — presenting the Christian faith not as a foreign import but as the completion of India’s own oldest religious inheritance. It stands as a foundational document of nineteenth-century Indian Christian apologetics, written from the inside of both traditions.

The Arian Witness was written in English, so no translation is involved — but the free scan of it is badly damaged, and that is what our edition fixes. The transcription in circulation was recognized around 2007 and renders the running head as “AEIANS, CASPIANS AND KASIAPAS”; it does not contain the phrase “Ahura Mazda” anywhere, though Banerjea uses it throughout. Ours was recognized afresh from the page images at higher resolution, which cuts measured damage by about a third. Banerjea’s Sanskrit needed a different remedy: no OCR can read Devanagari set in 1875 letterpress, so those passages are line noise in the free text. Where he cites the verse, we have restored it from the Rig Veda itself — the Aufrecht edition, in Unicode Devanagari — and every restoration had to prove itself: a distinctive word of the corpus verse must appear in the English he printed beside it. Ten passages meet that test, six of them proved by svasva, the word he renders “well-horsed” and on which his second chapter turns. Citations that could not be proved were left alone rather than filled with a guess. His translations are untouched; we have added none of our own. The free scan is linked below and is the same book; what differs is that you can read this one.

The free transcription of this book is badly damaged, and this edition exists to repair it. There is only one scan of the 1875 Calcutta printing anywhere online — a Google scan, at low resolution — and HathiTrust does not hold the book. So every page of this edition was read again against the page images of that scan: the English, the footnotes and, above all, the Sanskrit.

Banerjea argues from inside the Sanskrit tradition, and he quotes it in Devanagari: the Rig Veda above all, but also the Brahmanas, the Upanishads, the Mahabharata and Ramayana, the Puranas, Sayana's commentary and the grammarians. No optical character recognition can read Devanagari set in 1875 letterpress, and in the free transcription every one of those passages is line noise — Latin letters shaped vaguely like Devanagari glyphs.

All 231 Sanskrit passages in the book are now set in Devanagari, each read off the page image of the printed page. They are set as his compositor set them: the same words and the same extent (often a single line of a verse), joined the way the 1875 page joins them, with its spellings. Wherever Banerjea names what he is quoting, the reading was then checked against that text — the Rig Veda verse by verse from Sanskrit Wikisource, the Brahmanas from GRETIL — and 119 passages were checked that way. Where his printing differs from the standard text, the edition follows his page and records the difference. Where the type is too worn to be certain of a character, the best reading is given and recorded as uncertain; there are a few dozen of these in the whole book.

His footnotes had fared almost as badly. The scan read each page on its own, so a note arrived as loose text in the middle of a paragraph, or welded into a sentence, and a paragraph that crossed a page arrived in two pieces with the notes wedged between them. All 270 notes are now separated and linked to the place in the text where he put them, every divided paragraph is whole, and dozens of passages that had dropped out of the transcription entirely — whole footnotes among them, a long passage of the Satapatha Brahmana, a line of Kalidasa, a line of Sophocles and a line of Strabo — are restored from the page.

The English was read against the page too, and more than eight hundred recognition errors are corrected, each recorded with the reading it replaced. His Greek and Hebrew are restored in their own scripts. Dated terms are modernized so a reader can follow and look them up: Muslim for Mahometan and Musalman, Hindu for Hindoo, Parsi for Parsee, Sanskrit for Sanscrit, Kashmir for Cashmere; and the Indian and Persian names he romanized with the accents of his day are given in the plain form in use now — Sayana, Narada, Mahabharata, Himalaya, Prajapati. His Sanskrit itself, in Devanagari, is exactly as he printed it.

Running heads, page numbers, printer's signature marks and the scanner's stray marks are removed. His table of contents, whose page numbers referred to the 1875 printing, is replaced by the ebook's own. Every change made in re-reading the book is logged with the page it came from.

The free scan is linked on this book's page at tsbakerbooks.com. It is the same book. What is different here is that you can read it.