A modern student of rhetoric opens a textbook on figures of speech and finds dozens — simile, metaphor, metonymy, synecdoche, hyperbole, litotes, oxymoron, paradox, alliteration, assonance, consonance, parallelism, antithesis, chiasmus, anaphora, epistrophe, polysyndeton, asyndeton, zeugma, syllepsis, anadiplosis, epanalepsis. The catalog runs long; rhetorical-figure inflation is a feature of the literary-theoretical tradition. Pick up Bhāmaha or Mammaṭa, the later Sanskrit poetics-handbooks, and the lists are even longer — hundreds of named figures spread across categories.
NS XVII.43 names four. "Four figures of speech available in drama are: Simile (upamā), Metaphor (rūpaka), Condensed Expression (dīpaka), and Yamaka."1 Just four. The framework's minimal-cardinality claim is doctrinal. Bharata is asserting that the working space of dramatic figure-of-speech is covered by four operations; everything else is sub-variation or combination.
This is sharp craft-claim, not theoretical impatience. The four are not chosen at random. Each operates on a distinct symbolic mechanism. Each produces an aesthetic effect the other three do not produce. Each is testable as deployable craft. The framework's claim is that these four exhaust what figures-of-speech do — every other named figure in the later tradition is one of the four with sub-variations.
Upamā (Simile) — NS XVII.44: "When in a poetical composition anything is compared on the basis of some similarity it is an instance of Simile. It relates to quality and form."2 The most basic comparison. A is like B. NS XVII.45-49 catalogs the number-of-objects-compared (one with one, one with many, many with one, many with many) — "your face is like the moon" (one to one), "stars shine like the moon" (many to one), "having an eye like that of a hawk, a peacock, and a vulture" (one to many), "elephants are like clouds" (many to many).3 NS XVII.50 gives five sub-kinds: simile of praise, censure, conceit, uniqueness, partial likeness.4 Simile is the primary symbolic operation — comparing one thing to another on the basis of similarity. The framework treats Upamā as the alaṃkāra-substrate; the other three operate as elaborations of the basic comparison-move.
Rūpaka (Metaphor) — NS XVII.59: "An image of slight likeness which is conceived due to indecision from objects characterised by similar limbs, is called Metaphor."5 Where Upamā says A is like B, Rūpaka says A is B. The likeness has tipped into identity. NS XVII.60 gives the example: "The pool of water and women, with their lotus-faces, Kumuda-smiles, beautiful and open Nilotpala-eyes and swans cackling around, seem to be talking to one another." The women have become the lotuses; the framework treats the image-fusion as the defining feature. The reader does not hold two separate things in mind; the metaphor produces one composite image.
Dīpaka (Condensed Expression) — NS XVII.57: "When words agreeing with different sets of words are combined into one sentence by way of illuminating them together, it is an instance of Condensed Expression."6 NS XVII.58 gives the example: "In that region fullness was always effected by swans in the lakes, by flowers in the trees, intoxicated bees in the lotuses and by friendly groups of men and women in the parks and the gardens." One verb (effected fullness) governs four parallel constructions (swans in lakes, flowers in trees, bees in lotuses, groups in parks). The framework names this as the condensed operation — one element doing the work of many, the sentence carrying multiple parallel meanings through compression. Modern rhetoric calls this zeugma when it is one word stretched across multiple constructions.
Yamaka — NS XVII.61-86 catalogs the most-developed of the four. Yamaka is phonetic-repetition with semantic-variation — the same syllables recur but the meanings differ. NS develops multiple sub-types: Pādānta (foot-end repetition), Kāñci, Āmredita (last-word reduplication), Catur-vyavasita (all-feet-similar), Mālā (single-consonant with varied-vowels), and others. The example at XVII.82: "vāraṇānām ayam eva kālo vāraṇānām ayam eva kālaḥ" — the same syllables repeated, with the meaning shifting from time of the vāraṇa flower to time when elephants are free from disease to time for enemies to come to time for battle. One sound, multiple meanings.
The framework's claim is that the four cover the entire working space of figures-of-speech because each operates on one of four primary symbolic operations:
Modern cognitive-poetics has developed similar claims independently. Lakoff and Johnson's Metaphors We Live By (1980) treats metaphor as one of a small number of primary cognitive operations. Modern symbolic-cognition research suggests that the human mind operates with a small set of primary symbolic transformations and a much larger set of cultural-specific instantiations. The framework's four-fold claim reaches what cognitive-research has confirmed — that figures-of-speech operate on a small substrate of primary operations.
The proliferation of named figures in later poetics (Bhāmaha's tens, Mammaṭa's hundreds) is sub-typing rather than new operations. Hyperbole is a kind of Upamā (comparison-to-an-exaggerated-similar); irony is a kind of Yamaka (form-meaning mismatch); chiasmus is a kind of Dīpaka (condensed-parallel structure). The framework's minimum-cardinality holds; what looks like proliferation is taxonomic-elaboration on the four primary operations.
The four-alaṃkāra framework gives modern writers a working economy that the encyclopedic figure-of-speech catalogs in modern rhetoric do not. A writer who has explicit access to all four primary operations can deploy them deliberately; a writer who has implicit access to dozens of named figures often deploys them by feel without recognizing which primary operation is being used.
As prose-craft diagnostic. When a sentence works, identify which of the four it deploys. Most strong prose-effects can be attributed to one or two of the four operating cleanly. When a sentence fails, the operation is usually muddled — the writer was reaching for a comparison but the comparison did not establish the similarity-basis (failed Upamā), or for a metaphor but the image-fusion did not occur (failed Rūpaka), or for condensation but the parallel-constructions did not align under the governing element (failed Dīpaka), or for repetition-with-variation but the variation-was-too-thin (failed Yamaka).
As writing-pedagogy economy. Teaching four primary operations is more tractable than teaching dozens of named figures. A writer who has internalized the four operations can recognize most named figures as variants and deploy them in turn.
As anti-figure-proliferation discipline. Modern rhetoric textbooks list many figures of speech as if each were a distinct operation. The framework's minimal-set claim is that most are sub-varieties of these four. A writer who deploys them deliberately is doing more with less.
Four-as-minimal vs later expansion: NS's four contrasts with the hundreds of alaṃkāras later Sanskrit poetics names. The minimal framework is operationally elegant; whether it is empirically complete is debated. The page reads the proliferation as sub-typing rather than as new operations.
Yamaka-asymmetry: NS develops Yamaka much more elaborately than the other three. The asymmetry suggests either Yamaka was the framework's specialty (most-developed in the dramatic tradition) or the most-contested rhetorical category requiring extended treatment. Modern reading typically treats Yamaka as less central than Simile and Metaphor; NS's emphasis cuts against that habit.
Alaṃkāra-lakṣaṇa boundary: the four alaṃkāras are listed separately from the thirty-six lakṣaṇas in the same chapter. The boundary is operationally precise — alaṃkāras decorate the sentence; lakṣaṇas perform discourse-acts — but later poetics sometimes blurs the two. The framework's distinction is structurally important and worth preserving in modern reading.
Bharata is the single primary source on the four alaṃkāras, and the productive tension is between the minimal-four commitment and the internal-elaboration of Yamaka. NS gives four categories at the major level; within Yamaka it gives at least five sub-types. The framework operates at minimal cardinality at the categorical level and full elaboration at the sub-type level. This is consistent with NS's broader pattern — a small number of major categories with significant internal articulation.
The framework's claim is that the cardinality is not arbitrary. Four primary operations covers the working space because four matches what the receiving mind actually does with figures-of-speech. The audience's mind operates on substitution-by-similarity, substitution-by-identity, condensation, and repetition-with-variation; these are the four primary symbolic operations available to consciousness. The framework's four are not Bharata's arbitrary count; they are the empirical-observation of what working figures-of-speech actually exploit.
What this reveals about the framework: NS is reaching what modern cognitive-poetics has independently confirmed. The mind's symbolic-cognition operates with a small set of primary transformations. Bharata observed the same set without the cognitive-science vocabulary and codified them as a working craft-tool. The later Sanskrit poetics that proliferated into hundreds of figures was doing taxonomic-elaboration; NS's four is the load-bearing structural claim.
The four-alaṃkāra framework is a creative-practice page, but its underlying claim — that figures-of-speech operate as a minimal-but-elaborate-internally catalog — runs into adjacent vault domains where the same cardinality-claim recurs.
behavioral-mechanics: Eleven Propaganda Ploys — Comprehensive Tactical Taxonomy — propaganda theory names eleven specific tactics that produce influence-effect. The structural parallel to NS's four alaṃkāras is at the minimal-cardinality level: both frameworks insist that working craft requires a small named catalog at the right resolution — neither too few categories nor too many. NS treats four as the working count for aesthetic figures; propaganda theory treats eleven as the working count for influence tactics. The frameworks differ in domain but share the methodological commitment to finite-cardinality named catalogs. The insight neither framework produces alone: figures and tactics operate at a resolution finer than generic-craft language but coarser than infinite-proliferation. NS's four and propaganda's eleven represent the working middle. The cardinality-claim is empirical: the receiving mind operates at a specific resolution for figures-of-speech (four) and at a different resolution for tactical-discourse (eleven), and effective craft matches its own cardinality to what the receiver actually distinguishes. Modern rhetoric textbooks that catalog dozens of figures are operating above the receiving-mind's working resolution; modern stand-up training that operates with three or four bit-types is operating below the working resolution for short-form comedy. The framework's resolution-matching is the methodological insight that transfers across domains.
psychology: Symbolic Thinking vs Concrete Thinking — cognitive-developmental psychology distinguishes the symbolic mode (similes, metaphors, analogies) from the concrete mode (literal description). The four alaṃkāras operate centrally in the symbolic mode — they are the structural mechanisms by which symbolic-cognition produces aesthetic effect. The framework's claim that four figures cover the working space matches developmental psychology's observation that symbolic-cognition operates on a small number of underlying transformations (substitution, displacement, condensation, repetition with variation). Bharata's four roughly map: Upamā = substitution-with-comparison; Rūpaka = displacement (the image takes the position of the original); Dīpaka = condensation (one word governs many); Yamaka = repetition-with-variation. The convergence suggests the four are not arbitrary — they correspond to the four primary symbolic operations the human mind performs. The handshake's deeper claim: NS's framework and modern cognitive-poetics are converging on the same empirical structure. The human mind has roughly four primary symbolic-cognition operations; the dramatic-tradition observed this and codified the operations as the working figures-of-speech catalog. The later proliferation into hundreds of named figures is cultural-historical elaboration on the same four primary operations.
If NS's four is the working count of irreducible figures of speech, the explosion of named alaṃkāras in later Sanskrit poetics may represent sub-typing rather than new categories. The implication for modern writing-pedagogy: rhetoric training does not need encyclopedic figure-vocabulary; four primary figures plus the thirty-six lakṣaṇas (discourse-acts) covers the working craft-territory. A writer who has internalized the four can do more with less than a writer who has memorized dozens of named figures without internalizing the primary operations they all rest on.
Generative Questions
I couldn't find any books on English rhetoric. Almost none.7 Mark Forsyth on why he wrote The Elements of Eloquence in 2013 — one nun's book from the early 20th century, one from Puttenham in the 1580s, then nothing. The Sanskrit alaṃkāra-tradition catalogued at NS XVII.43 has a structurally identical sibling in the Greek rhetorical tradition (progressio, diacope, chiasmus, anaphora, epistrophe, anadiplosis, tricolon, polyptoton, isocolon, periodic sentence, adynaton, many more) — but the Sanskrit tradition has been continuously preserved in Indian classical-arts pedagogy across two millennia while the Greek tradition was largely lost in Western writing pedagogy after the Romantics displaced rhetoric-as-calculation.
Both traditions converge on three operational claims: rhetorical figures are usable craft-tools rather than stylistic decorations; named formulas allow deliberate practice and transmission across generations; great writers use the figures intuitively whether or not they know the names. Forsyth's strongest evidence is the McCartney British Library manuscript case — McCartney went back and deliberately inserted anadiplosis into Here There and Everywhere on revision without knowing the Greek term.
Forsyth's pragmatic stance is operationally honest: the technical names do not matter at all. Indeed the ancient Greeks couldn't even agree on half of these technical names. The Sanskrit tradition has the same property — Bharata's four-figure enumeration sits alongside expanded taxonomies elsewhere; later commentators added figures; the technical-name multiplication is not load-bearing. Both traditions agree: the formulas are what matters; the names are scaffolding. The cross-tradition synthesis (alaṃkāra + Greek-rhetoric) gives writers a richer toolkit than either tradition alone. See Forsyth Rhetorical Figures Methodology.