§7.1 Madhva's distinctive move
Standard Sanskrit semantics distinguishes mukhya-vṛtti, the primary or direct function of a word, from gauṇa-vṛtti or lakṣaṇā, its secondary or indicative function. “The village on the Ganges” means, primarily, a village situated in the river; since that is absurd, the word is taken secondarily to indicate the bank. Every school accepts this machinery. What they dispute is when the shift to the secondary is licensed.
Śrī Madhvācārya's move — his most revolutionary contribution to Sanskrit semantic theory — is to argue that for Vedic words the primary function is not what the lexicographer records. He holds that all words, in their primary function, denote Viṣṇu (sarva-śabda-vācyatva), and that their application to ordinary objects is the derived, restricted usage. The technical name for this in the Dvaita literature is parama-mukhya-vṛtti, the super-primary function.
- Mukhya-vṛtti (conventional primary) — denotes finite deities or worldly objects through limited conventional usage (laukika-vyavahāra).
- Parama-mukhya-vṛtti (supreme primary) — denotes the Supreme Lord in the fullest etymological sense (nirukta / yoga), because the qualities the root words describe exist infinitely and originally only in Him.
The argument, in the tradition's terms: Viṣṇu is guṇa-paripūrṇa, complete in every perfection without remainder. A word denotes by virtue of a property. Any property a word denotes, He possesses in full and as its source; other bearers possess it partially and derivatively. Therefore the word applies to Him fully and primarily, and to them partially and secondarily.
In ours: this is an argument about the direction of derivation. Consider “president.” The word does not primarily mean the individual currently occupying an office and secondarily mean the office; it primarily names a function, and individuals bear the name by occupying it. Madhva's claim has this shape, raised to the ontological level: qualitative predicates name their perfect instance primarily, and imperfect bearers derivatively.
§7.2 Aitareya Upaniṣad Bhāṣya: sarvaghoṣāḥ
The theory does not stop at articulated names. In his commentary on the Aitareya Upaniṣad, Madhvācārya extends it to the whole acoustic field: not only human speech, but literally every sound in the cosmos is a vibration signifying the Supreme Lord.
"The surging of the ocean, the thunder of the clouds, the crashing of falling trees, the beating of drums — all sounds whatsoever are names of Him alone."
This is the maximal statement of parama-mukhya-vṛtti and the metaphysical keystone of the entire Brahma-Vaiṣṇava system. Sound is not an arbitrary human invention laid over a mute world. The universe is an acoustic reality whose every resonance terminates in the same referent. The theory of names in §7.1 is the special case of this general claim restricted to the sounds that happen to be words.
§7.3 Garuda Purāṇa & Bharadvāja Saṁhitā corroboration
Two further passages state the mechanism with unusual directness: the names of the gods are Hari's own names, delegated. They are presented here as illustration, not proof — see the grading note below, and §7.6 for what the doctrine actually rests on.
"O Devī, names such as Brahmā and the rest are in fact names of Hari, which He Himself bestowed upon the celestials. But the resolute One did not give away names such as Keśava — just as a king does not surrender his own name even on quitting his capital."
"Where the Supreme Person Puruṣottama is declared by undifferentiated words — the names of the devas and the rest, which are impure — that mode of speaking they call intermixed (vyāmiśra)."
The Bharadvāja passage supplies the term for the mechanism: vyāmiśra, the intermixed mode. A devatā name reaches Puruṣottama truly, but in a mixed register in which the delegated bearer and the source are not held apart. That is precisely what §7.4's vacuity objection needs answered, and precisely what the Garuda passage narrates as an act — delegation with reserve.
§7.4 The objections, stated at full strength
This is the most attacked position in the system, and the objections are good. A document that presented it without them would be advertising bad faith.
Circularity. You conclude that all words denote Viṣṇu from the premise that Viṣṇu is the perfect bearer of all qualities — which is what you were trying to prove.
Response, and what it costs: partially conceded. The theory of names is downstream of the theological conclusion, not independent support for it. Madhva establishes Viṣṇu-sarvottamatva from śruti by samanvaya first; the semantic theory then explains how the corpus can be read that way. Presented as independent proof it is circular, and we do not present it that way. This is why §7.6 rests the conclusion on BG 15.15 and the sūtras rather than on the semantic theory.
Unfalsifiability. If every word can mean Viṣṇu, no passage can ever count against the thesis.
Response: answered by the guard conditions of §6.2 and the licensing conditions of §9.2. Without those, the objection lands squarely. This is why we state the stopping condition S3 as prominently as the rule itself. Most occurrences of most names pass S2 and halt at S3; the platform's default display for those sites is the ordinary referent.
Derivational licence. Nirukta lets you derive nearly any meaning from nearly any root, especially with a lexicon that is not independently attested.
Response: the strongest objection, and it is not fully answerable. Madhva's Nirukti attributed to Vyāsa is not extant independently. His derivations must therefore be assessed one at a time on grammatical plausibility rather than accepted wholesale. Our editorial position: a nirukta derivation is Grade B at best unless independently supported by standard Nirukta or by a second commentator reasoning separately. Register entry D-05.
Vacuity. If every name denotes Him, the names carry no information and the corpus says nothing distinctive.
Response: rejected. The claim is not that names are synonyms but that they are aspectual — each denotes Him under a distinct quality. “Indra” reaches Him as splendour, “Agni” as the conveyor of oblation. Information is preserved; only exclusivity is denied. This is exactly why the Sahasranāma is a thousand names and not one name repeated a thousand times.
The practical objection. If every word denotes Viṣṇu, then gām ānaya, “bring the cow,” would mean “bring Viṣṇu,” paralysing ordinary life.
Response: conventional life operates on mukhya-vṛtti — relative distinctions of jīvas and matter within the conditioned realm — while philosophical inquiry (brahma-jijñāsā) operates on parama-mukhya-vṛtti. The transcendent meaning does not destroy the conventional utility; it locates its source. The two registers are addressed separately in the data model (§19.2) for exactly this reason.
§7.5 The nirukta etymological apparatus
Classical nirukta — the discipline formalised in Yāska's Nirukta, one of the six Vedāṅgas — proceeds from the position that Vedic words are transparent to their roots. A word is understood by being opened: identify the root, the affix, the sense-modification, and read the compound sense. This is not folk etymology; it is a technical procedure with rules, and it is the standard apparatus of the entire commentarial tradition, not a Mādhva invention.
- Agni → agri-nī, He who leads foremost → primarily the Supreme Lord, secondarily the fire deity.
- Indra → from the root sense of power, splendour, rule → primarily the Supreme Lord, secondarily Śakra, holder of the office.
- Śiva → from śiv, the eternally auspicious → primarily the Supreme Lord, secondarily the destroyer of Tripura.
- Rudra → ru-drāvayati, He who drives away sorrow, or rodana-kāraka → primarily the Supreme Lord, secondarily the deva Rudra.
§7.6 The doctrine stated as a conclusion
The preceding sections supply machinery. This one states what the machinery yields, at three widths, each with its own warrant and its own grade. The widths are nested: each is the previous one extended, and each carries a weaker warrant than the one before it. Stating them separately is what keeps the strongest claim from borrowing credit from the weakest.
What the conclusion does not license. Width 1 is a claim about the referent of the corpus, not a licence to read any given passage as being about Him at its ādhibhautika layer. The stopping condition S3 stands undisturbed: at the great majority of sites the ordinary referent bears all the predicates, and the platform displays the ordinary referent. Unity of ultimate referent and identity of proximate reading are different claims, and this method asserts the first while denying the second — as it also denies, at CF-03, the popular gloss that unity of referent makes all worship and all destinations equivalent (BG 9.25).