If you have ever sat through a Sundarkāṇḍa recitation, you know the sound. The opening invocation arrives as a single rolling wave of Sanskrit — atulita-bala-dhāmaṃ hema-śailābha-dehaṃ danuja-vana-kṛśānuṃ jñāninām agragaṇyam — and it goes past at speed, magnificent and completely opaque.
Most people have never been told that each of those compounds means something specific, or that the specificity is the point.
Vedprakāś Śāstrī's essay is an attempt to slow the wave down word by word, and it rests on a single methodological claim: a poet as exact as Tulsīdās pads nothing. No word is there for sound. Every epithet is doing interpretive work, and the string of them is a compressed character study rather than a garland of praise.1
That claim is either obviously true or obviously false depending on what you think devotional poetry is for, and the essay's value is that it commits and then does the work.
The move that makes the essay more than a glossary shows up in the second epithet.
Hema-śailābha-deham — a body like a golden mountain. Read as description, it says he is gold-coloured and large, which is not interesting and is contradicted elsewhere in the same volume where he is variously red, blue and gold-toned.
Śāstrī reads it as process. Gold does not arrive gleaming. It is tempered by fire, and the gleam is what fire does to it. So the epithet is not reporting a colour; it is stating a mechanism, and the instruction folded into it is to let hardship do its work rather than avoiding it.1
That converts the whole invocation from a portrait into a curriculum. And once you have accepted it for one compound you have to run it on all of them, which is the essay's actual project.
Whether Tulsīdās meant it is a real question and the essay does not raise it. What can be said is that the reading is available — Sanskrit compounds of this type genuinely are ambiguous between attribute and process, and the tradition has always exploited that.
There is a structural oddity about where this material sits, and it changes what the epithets are doing.
The invocation opens the Sundarkāṇḍa. It is the first thing a reciter says, before a single event of the book has been narrated.
So the reader meets foremost among the wise, treasury of every virtue, fire to the forest of demons — nineteen verdicts — and only then gets the ocean crossing, Surasā, Laṅkinī, the grove, the fire. The judgments precede the evidence.
Narrative does not usually work this way and the reason it works here is that nobody is reading the Sundarkāṇḍa to find out what happens. The audience knows. What the invocation supplies is not suspense but a reading frame: here is what to notice, installed before the noticing starts.
Which means Śāstrī's project is less strange than it first appears. He is not extracting hidden meanings from decorative praise. He is unpacking a set of interpretive instructions that were placed at the front of the book on purpose, by a poet who expected the book to be recited by people who already knew the plot.
The epithets are the apparatus. The story is the demonstration.
Not everything gets the treatment, and where Śāstrī declines to find a hidden layer is informative.
Goṣpadī-kṛta-vāriśam — he who reduced the ocean to a cow's hoofprint. Maśakī-kṛta-rākṣasam — he who crushed rākṣasas like mosquitoes. Śāstrī reads these plainly, as his speed and strength restated in the most extreme available terms.1
No mechanism, no instruction. Hyperbole doing what hyperbole does.
That restraint is what makes the rest credible. A commentator who finds a teaching in every compound is running a machine, and the output tells you about the machine. A commentator who says this one is just enormous praise has demonstrated he can tell the difference, which is the constraint the naming essays in this same volume conspicuously lack.
He does the same with rāmāyaṇa-mahā-mālā-ratnam, the jewel set into the Rāmāyaṇa's garland, which he takes literally: pull Hanuman's story out and the epic is visibly poorer.1 A structural claim about the text, stated as a structural claim, with no allegory attached.
The patronymic gets the essay's most human reading and it is worth the space.
Añjanānandanam — the one who fills Añjanā with joy. A patronymic is normally a pointer: whose son. Śāstrī asks what the ānandana is doing there and answers by asking what actually makes a mother in this tradition glad. Two things, on the old lines: a son's valour, or a son's devotion to Rāma. Hanuman supplied both in full, which is what lifts the epithet above a polite naming.1
Then he pairs it with two others in the same verse and the pairing is the argument.
Vīram — a true hero — which he cashes out in four specific acts: walking into guarded Laṅkā without fear, killing Akṣakumāra, burning the golden city, and coming through the flames unmarked. And jānakī-śoka-nāśanam, the reliever of Sītā's grief.1
Read vīram alone and you have a warrior. Read it against jānakī-śoka-nāśanam and the valour acquires an object: all of it undertaken in the middle of a search for a grieving woman. His heroism was never for its own sake; it was in the direct service of ending somebody else's suffering.1
That is epithet-pairing as an interpretive tool, and it is the essay's real contribution — the meaning is not in the compounds individually but in which ones the poet chose to put next to each other.
The essay is worth reading against a version of itself that would have been much worse, because the difference is instructive.
The bad version reads each epithet in isolation, extracts a virtue, and produces a list: strength, endurance, discernment, wisdom, leadership, devotion, speed. Nineteen compounds, nineteen virtues, a complete inventory and nothing learned.
Śāstrī avoids this by attending to adjacency. Añjanānandanam next to vīram next to jānakī-śoka-nāśanam. Manojavam next to jitendriyam. Maruttulya-vegam next to śrī-rāma-dūtam. In each case his reading is that the second term is a constraint on the first, and that the pairing rather than either word carries the teaching.
That is a real technique and it generalises well beyond this material. A list of qualities is a description. A list of qualities arranged in pairs is an argument about which qualities are dangerous without which others.
And it explains why the invocation is composed the way it is. If the pairings are load-bearing, the order is not decorative and the verse cannot be reshuffled for metre without loss, which is exactly the claim Śāstrī's whole method requires and never states.
He also closes honestly. He admits he cannot gather all of it in — Rāma, Bharata, Sītā, Kṛṣṇa and Arjuna have each praised Hanuman in their own words — and ends not with more analysis but with a bow, borrowing Tulsīdās's own words.1 An essay on the sufficiency of exact language declining to add its own is a defensible way to stop.
Worth stopping on the number, because it is doing something.
Śāstrī treats nineteen distinct compounds.1 That is not a name and a couple of honorifics. It is a sustained accumulation, and accumulation on that scale is a formal choice with costs — it slows the opening down, it is difficult to memorise, and any three of the compounds would have established reverence perfectly well.
Two things follow from paying for that.
The first is coverage. Nineteen compounds can reach strength, appearance, wisdom, discernment, leadership, devotion, speed, restraint, service, lineage and effect. Three cannot. If the invocation is a compressed character study rather than a salutation, it needs the surface area, and the length is the price of the completeness.
The second is redundancy against loss. A reciter who does not know what goṣpadī-kṛta-vāriśam means still has eighteen other handles, and one of them will land. A tightly economical invocation transmits nothing to anybody who misses its single image.
That is a real design property of liturgical language and it runs opposite to every instinct in ordinary composition, where redundancy is a fault. A text built for repetition by people of wildly varying preparation wants more surface than a text built to be read once by somebody paying attention.
Which also explains why the invocation survives while longer expository material in the same tradition does not.
You are writing the paragraph that describes something you care about — a product, a person, a body of work — and it has become a list of adjectives.
Fast, reliable, thoughtful, rigorous, warm. Every one true. The whole thing inert.
Run Śāstrī's diagnostic on your own sentence. Take each adjective and ask: is this a state or a process? Golden, or tempered-in-fire-until-gold? The list version is all states, which is why it does nothing — states have no mechanism in them and a reader has nothing to do with a state but nod.
Then the harder move, which is the pairing. Go through your adjectives and find, for each one, the quality that makes it safe. Fast is a liability without judgment. Rigorous is a liability without warmth. Thoughtful is a liability without a deadline.
Now write them as pairs, and let the pairing be visible. Not fast and thoughtful — that is still a list with an and in it. Fast enough to matter, and never faster than the thinking. The constraint has to be doing work in the sentence.
Most of your list will not survive this, because most adjectives on a list of five have no dangerous version and therefore no constraint to pair with. Those are the decorative ones, and dropping them is the whole exercise.
What you will be left with is two or three pairs, which is roughly what the Sundarkāṇḍa invocation is once you strip the extreme-praise compounds Śāstrī himself reads plainly.
Nineteen epithets are treated, each named in Sanskrit and located to a source — the Sundarkāṇḍa invocation, the Rāma Rakṣā Stotra, the Rāmcaritmānas.1 For a devotional essay this is unusually systematic, and the method is stated openly rather than assumed, which makes it arguable with.
The founding claim is unargued. That a poet as exact as Tulsīdās pads nothing is asserted, not demonstrated, and it is doing all the work — every subsequent reading depends on the premise that a hidden layer must exist because the poet would not waste a word. A reader who thinks devotional poetry uses conventional epithets for metre and euphony will find nothing here persuasive, and Śāstrī does not engage that reader.
The rendering remove is total. This is a page about what specific Sanskrit compounds mean, held in English translation of a Hindi essay. [RENDERING — verify against Hindi] applies to every reading on it, and none of them should be quoted as an interpretation of Tulsīdās without checking the original.
A tension with the volume's own form material. Śāstrī reads hema-śailābha-deham as instruction rather than colour-description partly because it works better that way. Ramlal's treatise catalogues the colour question at length and finds golden, gold-toned, blue and red across sources, treating the variety as evidence that no fixed form was ever the point.2 Both are reasonable responses to the same problem and they are not the same response.
Open question. Does the pairing structure hold across the whole invocation, or only in the three cases Śāstrī highlights? If a systematic pass finds unpaired virtue-terms, the technique is a local observation rather than a compositional principle — and that is checkable by anybody with the Sanskrit.
Śāstrī and Somchaitanya Shrivastava are the volume's two close readers and they disagree about where meaning is stored.
Shrivastava works on names and his unit is the root. His claim is that a name unpacks into a biography, and that different names are different instruments — invoke one and what returns is shaped by that name's own meaning.
Śāstrī works on epithets and his unit is the compound in its position. His claim is that the poet's arrangement is the argument, and his best readings come from adjacency rather than from etymology.
The two methods have opposite failure modes, which is the useful part. Shrivastava's has no constraint — a root can be made to yield almost anything, and his essay endorses five incompatible derivations of one name without discomfort. Śāstrī's has a constraint built into it, because a claim about adjacency can be checked against the actual order of the verse and can fail. He also demonstrates the constraint working by declining to find a teaching in the hyperbolic compounds.
For this vault's purposes that makes Śāstrī the more reliable of the two and Shrivastava the more generative. Both are worth having and they should not be cited in the same register.
Gawande's Sentence Craft: Visual Action Verbs. The craft principle is that a sentence does its work through verbs that can be pictured, and that adjectival description is the register writing falls into when it has stopped knowing what it is claiming. Śāstrī's process-reading of hema-śailābha-deham is that conversion performed on somebody else's text — turning a colour into a tempering, an attribute into an action. What the pairing produces is a point neither makes alone: Sanskrit's compound structure lets a single word be simultaneously adjectival and processual, so the ambiguity Gawande's principle exists to resolve is, in this language, permanent and productive. English forces the writer to choose between golden and tempered; the compound does not. Which means the interpretive freedom Śāstrī exploits is a property of the language rather than a discovery about the poet — and it also means the tradition's commentators have a lever English-language writers do not, and that the same lever is why nothing they claim can be falsified from the compound alone.
Title as Doorway and Souvenir. The title-craft argument is that a good title does two jobs at different times — it gets a reader in, and afterwards it is what they carry away and use to refer to the thing. An epithet-string is that function multiplied and distributed: nineteen small titles, each a doorway into one aspect, and collectively the thing a reciter carries. The insight the pairing yields is about why the invocation survives as the most-recited passage in Hanuman devotion when longer and more informative material does not. It is not memorable because it is beautiful; it is memorable because it is nineteen souvenirs, and a person who has recited it holds nineteen separate handles on the figure rather than one summary. That is a genuinely different retention architecture from narrative, and it suggests the compressed-epithet form is doing a job — portable, partial, re-enterable at any point — that neither a story nor a doctrine can do.
Sharpest implication. If the pairings are load-bearing, then the Sundarkāṇḍa invocation is not praise at all — it is a compressed argument that certain capacities are dangerous unpaired, delivered in a form that gets recited by people who have no idea they are reciting an argument. Which raises a general and slightly unsettling possibility about liturgy: the material a tradition arranges for daily repetition may be carrying its most operational content, precisely because that is the content worth installing below the level of attention.
Generative questions.