As the overview article showed, Jones concluded that the resemblance among Sanskrit, Greek, and Latin “could not possibly have been produced by accident.” But what path did he actually take to get there? The distance between “these look alike” and “these share a common ancestor” turns out to require several distinct steps. Here we trace that logic more carefully.
Three possibilities — accident, borrowing, common ancestry
Broadly speaking, there are three reasons languages can resemble each other: pure coincidence; borrowing from a neighbor; or descent from a shared parent that later split apart. Jones never lays these three out as an explicit list of his own. But the shape his argument actually takes ends up working through exactly this sequence. The skeleton of his case is a process of ruling the first two out, one at a time.
(Japanese and English renderings of quotations throughout are ToL’s own.) Ruling out “accident” is the easier step — the resemblance wasn’t confined to a word or two, but reached into verb roots and grammatical form, the system as a whole. A single coincidental match between words is easy to find if you survey enough of the world’s languages. But an entire system of inflectional patterns lining up this consistently isn’t something Jones thought mere probability could explain.
Ruling out “accident,” though, doesn’t settle the matter. Next comes “borrowing,” and that one takes a more careful argument. Words cross borders easily. So couldn’t Sanskrit vocabulary simply have flowed into Greek and Latin at some point?
A second clue, drawn from the pattern of conquest
Jones answers this question with material drawn not from linguistics but from the history of conquest. In the same section on “Languages and Letters,” discussing the relationship between Hindi and Sanskrit, he writes:
…the basis of the Hindustání, particularly the inflexions and regimen of verbs, differed as widely from both those tongues, as Arabick differs from Persian, or German from Greek. Now the general effect of conquest is to leave the current language of the conquered people unchanged, or very little altered, in its ground-work, but to blend with it a considerable number of exotick names, both for things and for actions… like the Turks in Greece, and the Saxons in Britain.
Speakers of Hindi, Jones observes, took in a great deal of vocabulary from Sanskrit and Arabic, yet held on to the “ground-work” — how verbs conjugate, how sentences are put together. This wasn’t an armchair conclusion; it was drawn from the living language Jones heard around him every day in India — the kind of raw linguistic data only a judge actually resident there could have gathered.
Turkish rule over Greece never turned Greek grammar Turkish. The Saxon conquest of Britain left the native grammatical skeleton intact. A conqueror’s language leaves vocabulary as its calling card; the grammatical foundation stays put. Having established this general rule, Jones gains his argument: if it were borrowing alone, the grammatical skeleton shouldn’t resemble the source at all. Put the other way around, resemblance reaching that deep into the skeleton is exactly what borrowing can’t explain. Because the resemblance between Sanskrit and Greek and Latin reached into precisely that layer — verb roots, grammatical form — borrowing falls away, and only common ancestry is left standing.
What makes this move clever is that Jones doesn’t try to settle the question from inside linguistics alone — he imports, so to speak, a control group from outside it: a pattern observed repeatedly across recorded history. There is, in fact, no evidence that Sanskrit and Greek and Latin ever stood in anything like a conqueror-conquered relationship to begin with. And even if they had, the Turkish-Greek and Saxon-British cases show that contact through conquest leaves vocabulary behind, not grammatical structure. So even granting some hypothetical contact, it still wouldn’t explain the depth of resemblance actually observed. Only one explanation is left. This shape of reasoning — laying out several hypotheses that could account for an observed fact (the deep resemblance among Sanskrit, Greek, and Latin), then checking each against known patterns and eliminating them one by one — is recognizably close to modern scientific inference. Jones called himself a “philologer,” but what he was actually practicing here has more in common with natural history or geology: reconstructing past events from present evidence. Like a geologist reading rock strata to reconstruct ancient geological upheavals, Jones was trying to infer a past event no longer directly observable — the branching of languages — from the shape of languages as they exist today.
The same logic, extended to writing systems
Jones doesn’t stop once he’s done with spoken language. He turns the same comparative move toward a different object entirely: writing. The idea is that if you can trace the lineage of spoken words, you should be able to trace the lineage of written symbols the same way. Having discussed India’s own script (the Nagari alphabet), he continues:
…the square Chaldiack letters, in which most Hebrew books are copied, were originally the same, or derived from the same prototype, both with the Indian and Arabian characters. That the Phenician, from which the Greek and Roman alphabets were formed by various changes and inversions, had a similar origin, there can be little doubt.
Devanagari, the square Chaldean (Hebrew) letters, Arabic script, Phoenician and the Greek and Roman alphabets descended from it — Jones is reaching for a common prototype across writing systems with entirely different lineages. Where his linguistic argument rested on verb roots and grammatical form, here the evidence is the shape of the letters themselves. Whether the object is language or script, the same posture holds throughout: look past surface features to the structural resemblance underneath — a stance later inherited by both comparative linguistics and comparative paleography.
That said, this claim about scripts is built with far less care than the claim about language. Where the linguistic argument marshaled specific evidence — verb roots, grammatical forms — the claim about the origin of scripts amounts to little more than a stated intuition: “there can be little doubt.” Modern paleography does not, in fact, support a direct link between the Nagari family of scripts and the Semitic family (Phoenician, Hebrew, Arabic) — on this point, Jones’s guess turned out to be wrong. One address holds both an insight that has held up for more than two centuries (the linguistic kinship) and a guess that didn’t survive scrutiny (the script kinship) — which is itself a reminder not to treat his argument as infallible, but to weigh the strength of each claim on its own. Within a single lecture, one claim holds up for two centuries while another is soon overturned — the discourse itself embodies how historical “discoveries” are rarely uniformly correct, but sort, in hindsight, into what gets kept and what quietly gets forgotten.
What “philologer” actually meant
Of the resemblance among Sanskrit, Greek, and Latin, Jones writes that “no philologer could examine them all three without believing them…” to share a common source. “Philologer” here doesn’t just mean someone good with languages — it means an expert capable of actually comparing several languages and examining their internal structure. Jones himself had mastered numerous languages, including Sanskrit, Greek, Latin, Persian, and Arabic, to a genuinely literate level. Part of why this single sentence was received as a testable hypothesis rather than a passing thought is that the person saying it was, himself, in a position to carry out that very comparison.
Turned around, this argument is also an appeal to verifiability: “I have compared these three myself, and anyone else who does the same should reach the same conclusion.” Jones isn’t asking his readers to take his authority on faith. He builds an open invitation — check it yourself — directly into the sentence. That stance is exactly what let later scholars pick up the claim as a hypothesis and put it to the test, rather than as a pronouncement to be accepted or rejected on faith. Presenting an authority’s assertion not as something to accept on faith but as something anyone could verify is what let this one paragraph become the starting point of a discipline, rather than just a memorable line.
Two discoveries in one address
Seen this way, the Third Anniversary Discourse isn’t simply a lecture that pointed out a resemblance between Sanskrit and Greek and Latin. It’s a working demonstration of how to turn an observation of resemblance into a testable claim. Rule out accident, rule out borrowing, back the case with a pattern drawn from outside linguistics entirely, then extend the same move to an entirely different object, writing. That whole sequence may be Jones’s real contribution. He left behind not just a conclusion — the idea of an Indo-European family — but a pattern of thinking for reaching one, which makes this discourse a double discovery. The next article looks at what the scholars who inherited that pattern actually went on to verify, and how.
Summing up, and what comes next
Jones’s logic became the prototype for what would later be called the “comparative method.” He never turned it into a formal methodology, though — he demonstrated the intuition that resemblance too deep to borrow implies common ancestry, and left untouched the work of actually identifying which sounds correspond to which. Why, for instance, does Latin pater correspond to English father, with a p becoming an f? Working out regular sound correspondences like this one — forging Jones’s intuition into a testable law — became the next generation’s task, one taken up more than a generation later by scholars in Europe.
Looking back, the strength of Jones’s case lay in each piece of evidence being independently checkable. “The verb roots and grammatical forms resemble each other” — anyone can verify that by comparing the texts. “Conquest leaves vocabulary but not grammatical structure” — anyone can verify that by surveying historical cases. The weakness was what came after: he offered no concrete procedure for how or why the shared ancestor actually diverged into its later forms. That gap, between a powerful intuition and a still-unbuilt method, is exactly what drew the next generation of scholars in.