The verse that was never about mantras

Some time in the 1570s, in Vārāṇasī, a poet writing in Avadhi set down four lines about the age he believed he was living in. The age was Kali, last and worst of the four. His subject was what the gods had done about it.

Śiva and Pārvatī, he wrote, seeing the age and wanting the world’s good, made a jāla of śābara mantras. The word is ordinarily Englished as a net. What the commentators do with it comes later, and it turns out not to be the interesting part.

Four hundred years later a publisher in Delhi printed those same lines on the opening page of a cheap paperback of spells.

That is this whole treatise in two facts.

The poet was not writing about mantras. He was writing about the age, and about a god’s response to it, and the mantras enter his sentence as the instrument of that response. He gives them one word. He does not explain what is in them, who may use them, or how. The line has since carried the entire weight of a genre, several publishing houses, and a definition repeated for over a century, and it was never a description of anything. It was a poet’s aside about the state of the world.

Then he says something harder, and it is the reason the line is quoted and the reason it is misread.

The letters are ill-matched. There is no proper sense. There is no repetition to perform.

And by the power of Maheśa, they work anyway.

Read it once and it sounds like an attack. Generations have read it that way, and the popular literature quotes the first half and stops. A poet does not spend three quarters of a couplet demolishing a thing he is about to credit to Śiva. He is granting the charge. Someone had evidently already said these syllables are gibberish, and Tulsīdās does not argue. He concedes it in the plainest words available and then says the power is not in the sense.

That is a claim about where power lives. It is not a claim about what the words mean.

The poem carries its own date, and the date says what he was doing when he wrote the line. He was at the beginning of a retelling of the life of Rāma in Avadhi, the spoken tongue of the eastern Gangetic plain, and he had stopped, before the story, to defend the language he was using. In the same Bālkāṇḍ invocation, at the block carrying the date, he writes संबत सोरह सै एकतीसा, sixteen hundred and thirty-one, and adds the day: नौमी भौम बार मधु मासा, the ninth tithi, a Tuesday, in the month of Caitra, at Ayodhyā. Saṃvat 1631 is 1574 of the Common Era.

Four lines of that defence are now the standard proof text for an entire category of Indian mantra. They appear at the top of the article, in the preface of the paperback, on the slide, in the video description, in the forum post, quoted loosely and with confidence: Śiva made these mantras for the Kali age, they have no meaning, and they work anyway.

The lines are real. The poem they come from is one of the most widely read books ever written in an Indian language.

Almost nobody quotes the sentence they belong to. How often is not measurable from anything in this project’s hands. That is an impression formed from the material read, and it is offered as one.

This chapter is about how a concession became a definition. It begins with the couplet itself, phrase by phrase, because the couplet is harder than its reputation.

Where the lines stand

The passage stands in the Bālkāṇḍ, the first book of Tulsīdās’s Rāmcaritmānas, inside the block that closes at dohā 15. The chaupāī is the four-part rhyming line that carries the narrative; the dohā is the two-line couplet that stops it and seals the sense.

The chaupāī number is not stable, and this treatise has to say so before it cites anything. This book cites the verse as Bāl 1.15.3, and that address is this book’s own convention. No printed edition has been shown to us carrying it. Gita-Press-derived digital texts number the chaupāī block by the dohā that precedes it, which gives 14.3 for these lines. The standard Hindī commentarial compendium, Mānas-pīyūṣa, edited by Añjanīnandan Śaraṇ and published by the Gita Press in 1968, prints the same two lines as chaupāī 5 and 6 of dohā 15, and its own subject index sends the reader to शाबरमंत्र १५,५. The editor’s own preface explains that he renumbered across the edition after collating old manuscripts, on the principle that the chaupāīs preceding a dohā belong to that dohā, which means the attachment of this block to dohā 15 is Śaraṇ’s own innovation rather than the press’s ordinary practice. So the most-quoted verse in this book carries three addresses. This chapter names all three and labels each, and does not pretend the difference away.

On the Devanāgarī itself. It agrees character for character across four digital copies, and those four are not four sources: all descend from the Gita Press line of text, so under any honest test of independence they are one witness repeated. A separate witness, a proofread transcription of a printed edition carrying a Hindī ṭīkā hosted on Hindi Wikisource, disagrees at three points: हर-गिरजा for हर गिरिजा, सिरजा for सिरिजा, and, in the following chaupāī, सो महेस where the Gita Press line reads सो उमेस. None of the three touches the sense of this verse. That transcription names no press and no year, and nothing load-bearing now rests on it.

कलि बिलोकि जग हित हर गिरिजा। साबर मंत्र जाल जिन्ह सिरिजा॥ अनमिल आखर अरथ न जापू। प्रगट प्रभाउ महेस प्रतापू॥३॥

kali biloki jaga hita hara girijā, sābara mantra jāla jinha sirijā anamila ākhara aratha na jāpū, pragaṭa prabhāu mahesa pratāpū Seeing the Kali age, for the good of the world, Hara and Girijā created the jāla of sābara mantras: ill-matched letters, no meaning and no japa, their power manifest by the majesty of Maheśa.

Rāmcaritmānas, Bālkāṇḍ, the chaupāī block closing at dohā 15 · 1.15.3 in this treatise’s own convention; Gita-Press-derived texts number the block by the preceding dohā, giving 14.3; 15.5 in Mānas-pīyūṣa #

Rendered plainly, phrase by phrase. kali biloki, seeing the Kali age. jaga hita, for the good of the world. hara girijā, Hara and Girijā. sābara mantra jāla jinha sirijā, who created the jāla of sābara mantras. Then the second half. anamila ākhara, ill-matched letters. aratha na jāpū, no meaning, no japa. pragaṭa prabhāu, their power is manifest. mahesa pratāpū, the majesty of Maheśa.

Four beats. A claim about who made them, a claim about what they are made of, a concession that they make no sense, and an assertion that they work.

That is the whole of it. Everything else attributed to this verse in the popular literature has been carried in from somewhere else.

What the line actually says

The verse says the mantras were made deliberately, in response to a condition. The condition is the Kali age. The purpose is jaga hita, the good of the world. This is a statement about intention, and it is the part the popular tradition transmits accurately.

The verse says they are a jāla. The dictionary leaves the word open between a net, a web and a tangle. The commentarial tradition is less undecided than the dictionary. Mānas-pīyūṣa, in its word-gloss, renders जाल simply as समूह, a mass, a collection, a heap of them, and a note further on speaks of an ocean of such a mass. The dictionary sense of net is available to a reader of Tulsīdās and the commentators do not take it. On their authority the line says a great body of mantras was made, and says nothing about meshes.

The verse concedes three things in a single half-line. The letters are anamila, ill-matched, badly joined. Mānas-pīyūṣa unpacks that word technically: the letters have no maitrī, no compatibility, in the sense the mantra manuals give the term. There is no artha, no meaning. There is no japa.

Then it asserts, without argument, that the power is pragaṭa, manifest, out in the open, visible. And it grounds that power in one place only: the pratāpa of Maheśa.

No mechanism is offered. No vibration, no phoneme theory, no subtle science. The verse gives a maker, a defect, and a result, and it declines to connect them with anything except the maker’s authority.

Hold that shape. Everything after this is an unpacking of it.

Hara and Girijā, together

The popular accounts give these mantras to Śiva. He is the standard attribution, in the books and the videos and the article headers: Śiva composed them, Śiva sealed them, Śiva’s power runs them.

The verse names two.

हर गिरिजा. Hara and Girijā. Śiva and Pārvatī, the second named by her mountain descent, the daughter of the giri. The compound is a dvandva, two names standing side by side as two agents, and the commentarial tradition reads it that way without hesitating. Mānas-pīyūṣa glosses the line in Hindī with a plural relative and a compound subject: it was Śiva and Pārvatī who composed the mass of śābara mantras. That gloss carries a named editor, a named press and a year.

The pronoun then corroborates it. That pronoun is जिन्ह, jinha, plural and oblique. A single subject in this construction would take jehiṁ, and Tulsīdās uses precisely that form twice inside the same invocation for a single revered agent: rāmāyana jehiṁ niramayeu of Vālmīki, bhava sāgara jehi kīnha of Brahmā. So the plural at jinha sirijā is a choice he did not have to make, in a metrical slot either form would have filled.

He wrote the plural.

It cannot be pressed further than that. Avadhi permits the honorific plural, so जिन्ह on its own could in principle mark respect for one agent rather than count two. What settles the line is the dvandva plus the commentary, and the pronoun agrees with them.

The next chaupāī keeps the couple in view rather than dropping back to Śiva alone. In the Gita Press line of text it names him सो उमेस, so umesa, the lord of Umā, an epithet carrying her name inside it. In the Wikisource printed edition the same words read सो महेस, so mahesa, and her name is not in it. This chapter is not going to pick one quietly, because the argument would then rest on an edition rather than on a poem. What can be said without choosing is that the passage does not announce a shift from a pair to a single agent. It is worth recording that one of the named Mānas commentators, quoted later in this chapter, writes सो उमेश in his own couplet on this verse, which is a small independent vote for the reading with her name in it.

A small point of morphology, with a large consequence for how the category is described. The tradition as Tulsīdās received it, or as he chose to state it, makes these mantras a joint act. The transmission dropped one of the two makers somewhere along the way, and the drop is now near-universal.

The letter that was lost is three characters long.

He spelled it sābara

The Sanskrit form is शाबर, śābara, and that is the form nearly every modern discussion uses, including the discussions that quote this verse to authorise themselves.

Tulsīdās did not write it that way. The mūla carries the dental sibilant, the way the word behaves when it stops being Sanskrit and starts being a spoken language. Avadhi does this. The palatal sibilant flattens. The word arrives in the mouth of the region and comes out slightly different from how the grammarians left it.

A printed page shows the phenomenon plainly. In Mānas-pīyūṣa, at the opening that carries this verse, the verse line prints the vernacular form and the commentary’s own Hindī paraphrase directly beneath it prints शाबरमन्त्रसमूह, with the palatal sibilant restored. The notes on the same page and the next then alternate between शाबर and शावर inside a few lines of each other. Three spellings of one word inside one printed opening of one scholarly edition.

Nobody did anything wrong on that page. A modern gloss reaches for the standard modern spelling, and the standard modern spelling is Sanskrit. The printed page performs, in miniature, exactly the tension the verse is about.

Because the word in question is a word for mantras made of vernacular, non-Sanskritic syllables. And the poet spelled it vernacularly.

The claim is enacted in the spelling of the word for the claim.

The verse is an argument about a poem

Here is where the standard reading comes apart.

The verse does not sit in a chapter on mantra. It does not sit in a ritual manual, or in a digression on the powers of the Kali age, or in the mouth of one of the poem’s speakers. It sits in the maṅgalācaraṇa, the invocation, the opening sequence of the Bālkāṇḍ where Tulsīdās speaks in his own voice about his own undertaking before the story of Rāma begins.

This is the passage where a poet salutes his teachers, names his subject, admits his inadequacy, and says what kind of book he is about to write.

And the block ends at dohā 15 like this:

सपनेहुँ साचेहुँ मोहि पर जौं हर गौरि पसाउ। तौ फुर होउ जो कहेउँ सब भाषा भनिति प्रभाउ॥१५॥

sapanehũ sācehũ mohi para jauṁ hara gauri pasāu, tau phura hou jo kaheũ saba bhāṣā bhaniti prabhāu If in dream and in waking Hara and Gaurī are gracious to me, then let all I have said of the power of a vernacular composition come true.

Rāmcaritmānas, Bālkāṇḍ, dohā 15 #

The Wikisource edition reads साँचेहु for साचेहुँ and कहउँ for कहेउँ. Neither moves the sense.

Bhāṣā, the vernacular, the spoken tongue, the language that is not Sanskrit. Bhaniti, an utterance, a composition, a poem. Prabhāu, power.

The couplet that closes the block is about the power of a vernacular composition. That is the topic. That is what “all I have said” refers back to. And the same divine couple named in the chaupāī, Hara and Girijā, are named again here as Hara and Gaurī, invoked for their grace on the poet personally.

Now read the argument as a single movement.

Śiva and Pārvatī, seeing the age, made mantras out of vernacular syllables. Those syllables are ill-matched. They carry no Sanskritic sense. They have no japa. And they work, on the strength of Śiva’s power alone.

Therefore a poem in Avadhi can work too.

That is this treatise’s reading, and it would be an inference and nothing more if the commentarial tradition did not state it. It states it. Mānas-pīyūṣa, in the note that closes its discussion of the verse, says that what Gosvāmījī is really showing here is Śaṅkara’s compassion and favour toward himself, that he cites the śābara-mantra creation as an instance of that spontaneous mercy, and that just as the śābara mantras are accomplished by Śaṅkara’s grace, so the Rāmcaritmānas, being equally his prasāda, holds the same power. The analogy this chapter draws is the analogy the named Hindī commentary draws.

Tulsīdās is not expounding śābara mantras. He is citing them. They are his precedent. He is a poet about to write the life of Rāma in a regional spoken language, in a literary culture where Sanskrit held the prestige and the authority, and he needs a warrant. He finds one in a category of mantra that everybody already accepted as effective and that everybody already knew was linguistically irregular. If the gods themselves made potent speech out of broken vernacular letters, then the objection that his own language is unfit collapses.

He is arguing for his book.

Ill-matched letters, no meaning, no japa

The concession in the second half of the chaupāī deserves separate handling, because it is routinely softened, and because the Indian commentarial tradition does not read it in one way. It reads it in four.

Mānas-pīyūṣa sets out the anvaya of anmil ākhar arath na jāpū and says outright that it can be construed in several ways. Four stand across the two pages it gives the verse, and they contradict each other. The base gloss takes the letters as ill-matched and the mantras as carrying neither proper meaning nor japa, with an alternative in the same breath under which japa alone is what is primary. One of the four, attributed to a siglum rather than to a named commentator, arrives at the opposite conclusion: the śābara mantras are not devoid of meaning, the meaning simply does not tally with the letters. Another denies that any rule of the languages touches these letters at all, and that is the one a later chapter of this book comes back to.

This treatise reads na jāpū as the absence of a prescribed japa: no counted repetition, no rule governing the manner of it, no vidhi. The commentary makes that contrast itself, distinguishing mantras that carry a japa-vidhi, a rule of recitation, and bear fruit only when it has been kept from the śābara mantras, which have no such prescription and no puraścaraṇa at all. The reading is the commentary’s before it is ours.

What the commentary also does is refuse to let us hold it alone. Four construals stand on the page, one of them saying the mantras do have meaning, one of them saying japa is the whole of it, and this chapter is following one of the four. A specialist will know the others exist, and now so does the reader.

Neither gloss anywhere says the mantras cannot be uttered. Something with no artha can still be spoken, and the tradition that uses this material has always spoken it.

So the line is a triple concession. The letters do not fit together. There is no meaning to recover. There is no japa.

And then the poet does not defend any of that.

He does not say the meaning is hidden. He does not say the sense is encoded and available to the initiated. He does not say the sound works below the level of semantics. The modern gloss on all of this, that the power lies in intent rather than in linguistic form, is a reasonable position and a genuinely later one. It is a defence.

Tulsīdās offers no defence. He grants the charge in full and asserts the result anyway, and he grounds the result in a person: mahesa pratāpū.

That is a harder position than the modern one, and it is four and a half centuries older. The modern gloss protects the mantras by explaining how they could work despite the defect. The verse does not protect them. It says the defect is real and the power is manifest and the reason is Maheśa.

Whatever else this is, it is not an argument about linguistics.

The couplet nobody had opened

There is a doctrine attached to this verse in the popular literature. It holds that śābara mantras are uniquely exempt from kīlana, the locking or nailing of a mantra, and that this is why they work at once while other mantras must first be unlocked. Search for it in English, in transliteration or in Devanāgarī and what comes back is content farms.

The standard commentary on the verse carries it outright.

Mānas-pīyūṣa, in a note on this line, reports the author of the Mānas-mayaṅka giving a couplet of his own:

सर्पादिक विषहरण कलि, साँबर रचे तुरंत। सो उमेश कलि अघ दहन मानस यश बिरचन्त॥

Those two lines are the Mānas-mayaṅka’s, quoted in Mānas-pīyūṣa.

And it glosses it. The sense of the gloss, paraphrased, is that when the Vedic and Tāntric mantras had been kīlita, locked, Śiva composed the śābara mantras for the benefit of creatures; and that because the other mantras had become locked, the śābara mantras alone were left bearing fruit. The Mānas-pīyūṣa siglum table names the Mānas-mayaṅka as the work of Paṇḍit Śivalāl Pāṭhak.

Two things follow.

The first is that the exemption doctrine has a textual basis. It has a named Hindī commentarial pedigree, it is attached to precisely this verse, and it is reported in a Gita Press scholarly compendium. The negative that stands beside it is narrower, and it is a real finding: nothing in the Sanskrit Tantric literature reached by this project supports the exemption, and the place the popular tradition got it is the Mānas commentary. Chapter nine is built on that.

The second is that the couplet is this chapter’s own argument in two lines by somebody else. Read what it says. The one who composed the sā̃bara mantras that draw out the poison of snakes is the same Umeśa who now composes the glory of the Mānas. A commentator on Tulsīdās, working inside the tradition, saw that the verse links the making of the mantras to the making of the poem, and wrote a couplet to say so.

Note also, in passing, three more spellings: साँबर in the couplet, शाबर and शावर in the notes around it. This word will not sit still, and chapters four and five are about that.

The date of the Mānas-mayaṅka is not established here. It is a printed Hindī commentary collected in a 1968 compendium, and what would fix its century is the Mānas-mayaṅka itself with its imprint read. The claim, until then, is that the doctrine belongs to the named commentarial tradition and is older than the internet, and not that it is old in any particular century.

The commentators, named

Which commentators on the Mānas actually treat this verse, and what are their names. Mānas-pīyūṣa’s siglum table and its works-consulted list answer both questions. Those whose coverage reaches Bālkāṇḍ dohā 15 include the Mānas-mayaṅka of Paṇḍit Śivalāl Pāṭhak; the Mānas-paricārikā of Bābā Jānakīdās of Ayodhyā, a ṭīkā on the first forty-three dohās of the Bālkāṇḍ; the Mānas-mārtaṇḍa of Bābā Jānakīśaraṇ, a tilak over the same span; the Mānas Abhiprāya-dīpak of Gurusahāylāl; the Mānas-dīpikā of Bābā Raghunāthdās of Kashi; and, among the printed ṭīkās collected, the work of Paṇḍit Rāmeśvar Bhaṭṭ and of Bābū Śyāmasundar Dās, who also edited the Hindī Śabdasāgara this treatise leans on in four later chapters.

That is the apparatus. It is Hindī, it is Indian, it is named, and it was there the whole time.

There is one more thing in Mānas-pīyūṣa’s discussion of the verse, and this chapter hands it forward rather than using it. The note that closes that discussion records the tradition that Śiva and Pārvatī appeared in Bhīl form in the Kali age, that Śiva composed the śābara mantras in the Bhīl language, and that Gaṇeśa wrote them down at Pārvatī’s command. Then it gives an etymology in four words: सँवर is the word for Bhīl, and the name arose from that. An Indian commentator, on this verse, deriving the name of the category from the speech of a named forest people. Chapter two is where that belongs.

What this chapter does not print

Mānas-pīyūṣa prints five specimen śābara mantras, sourced to a named ṭīkākāra. This chapter has read them and does not reproduce them, and neither will any chapter of this book.

That is a position and not an omission, and it is worth stating once, here, at the first place it arises. This treatise exhibits the form of the genre through everything that can be shown without handing over an instrument: the distributions of its formulae, the catalogue descriptions, the lexicographers’ glosses, the commentators’ analyses of what its letters do and fail to do. Its function is withheld. Chapter twenty sets out the reason at length, from the documented record of what has been done with material of this kind, and a reader who wants the argument rather than the rule will find it there.

Later in the same invocation

There is a second claim of the same shape in the same invocation, and the pair of them show what Tulsīdās was doing in these pages.

At dohā 33 of the Bālkāṇḍ, still in the frame where the poet is setting up his authority to tell this story at all, he writes that Rāma has taken avatāra in many ways, that the Rāmāyaṇas number a hundred crore and are boundless, that by the difference of kalpas the beautiful deeds of Hari differ, and that the great sages have sung them in many ways.

Read the passage beside the śābara verse and the strategy shows.

One claim licenses many Rāmāyaṇas. If the sages sang the story differently because the world-ages differed, then a new telling in a new language in a new age is not a deviation from a fixed original. There is no fixed original to deviate from.

The other claim licenses mantras with no meaning. If the highest speech in the tradition can be irregular and still be potent, then irregularity is no disqualification.

Two arguments, in one invocation, both saying that non-canonical, irregular, unlicensed forms carry real authority. One makes room for the poem’s content. The other makes room for its language.

He needed both. He was writing the story of Rāma, which Vālmīki had already written in Sanskrit, and he was writing it in Avadhi, for people who could not read Sanskrit. Both halves of that sentence were vulnerable, and he answered both before the story started.

The reading this chapter does not take

A treatise that spends four sections arguing against the popular reading owes the reader the rival to its own.

The maṅgalācaraṇa is formally a sequence of vandanā, salutations. A commentator reading the block that closes at dohā 15 has an obvious alternative available, and it needs no special pleading: the chaupāī is a praise of Hara and Girijā, and the śābara mantras are adduced as an instance of their compassion in the Kali age. On that reading the verse is about the mantras, as an item of Śiva’s glory, and it leads into the poet’s prayer that the same lord be favourable to him.

Mānas-pīyūṣa takes both at once, and says so in a single note: the mantras are mentioned as an instance of Śaṅkara’s spontaneous mercy, and the mention is there because the poem is claiming the same grace. The praise reading and the precedent reading are not rivals in the commentary. They are one move.

This chapter follows the commentary. It prefers the precedent reading as the governing one because of where the block ends, at a couplet whose subject is the power of a vernacular composition and whose verb is a petition for grace on the poet. What would overturn that preference is a Mānas commentator who reads dohā 15 as closing something other than the bhāṣā argument, or an edition in which the śābara chaupāī does not sit in that block. Neither has been found.

The title of this chapter is a compression, and compressions overstate. The verse is a doctrine of mantra in the sense that it makes four claims about mantras. What it is not is a treatise on them, and what it is doing where it stands is arguing for a poem.

What this leaves us

None of this tells you whether śābara mantras work, what they are, where they came from, or how old the category is. Those questions belong to the chapters that follow, and they need different evidence entirely.

What this establishes is narrower and worth being exact about. The single verse most often produced as authority on this subject is a poet’s argument for his own book, made in his own voice, in the opening invocation of that book, attributing the mantras to a divine couple rather than to Śiva alone, and spelling their name in the vernacular whose legitimacy is the point at issue. The verse concedes that the mantras are formless and asserts their power anyway, without offering a theory of how. The commentarial tradition on the verse says four different things about the second half of that line and one thing about its purpose, and the one thing is what this chapter has argued.

One limit is left, and it is the Mānas-mayaṅka. The commentary quoted at note 1(ga) has not been reached in its own printing by anybody here, and its imprint is not established. What would close it is that book on a table, with its title page read.

The tradition does not need this reading to continue. It has continued for a long time with the verse read the popular way, and the people who use these mantras were never waiting on a philologist.

The next chapter takes up the word inside the verse: where sābara came from, who it was said of, and what it was already carrying before anyone attached it to a mantra. A Hindī commentator has already told us it means the speech of the Bhīls. That answer has a long history behind it, and some of that history is other people’s.

The verse still says what it says.

It was written to make room for a poem. And the poem was written in the language the mantras are in.