As we saw in the overview article, nouns split into substantives and adjectives, corresponding to substance (what subsists by itself) and accident (what exists only through a substance). But reading Part II, Chapter II closely, the authors are not satisfied with that simple correspondence. They push one level deeper and ask: what exactly does a word mean, and how? “White,” “wise,” “round” — what do these adjectives actually point at? Whiteness itself, or the white thing? We rarely stop to notice the question at all — when we say “white snow,” we never pause to ask whether the adjective “means” the whiteness or the snow. But the authors take this deceptively simple question head-on, and it is where the chapter’s whole argument begins.
1. Substance vs. Accident — the starting point
Les objets de nos pensées sont ou les choses, comme la terre, le soleil… ce qu’on appelle ordinairement substance ; ou la manière des choses, comme d’être rond, d’être rouge… ce qu’on appelle accident.
(ToL working translation. Disclaimers on subsequent quotes are omitted; the same applies throughout.) The objects of our thoughts are either things — such as earth and sun — what is ordinarily called substance; or the manner of things — such as being round, being red — what is called accident.
A word that means a substance is a substantive; a word that means an accident is an adjective. “Earth” and “sun” name substances themselves, so they are substantives; “round” and “red” name the manner of a substance, so they are adjectives — a simple, straightforward correspondence so far. But the authors immediately add a caveat.
on ne s’est pas tant arrêté à la signification qu’à la manière de signifier.
What people actually attended to was not so much the content of the meaning as the manner of meaning.
2. “White” Means Two Things at Once — the Discovery of Connotation
So why, then, can an adjective not stand alone at the end of a sentence? The authors’ answer is remarkably precise.
ce qui fait qu’un nom ne peut subsister par soi-même, est quand, outre sa signification distincte, il en a encore une confuse, qu’on peut appeler connotation…
What makes a noun unable to subsist by itself is this: besides its distinct signification, it also carries a vague one, which we may call connotation.
“White” (blanc) distinctly means “whiteness” (blancheur). But at the same time it vaguely points at “something to which whiteness applies” — snow, paper, a wall, anything. It is this second function, connotation, that leaves the adjective hanging, leaning on a noun. Strip the connotation away, conversely, and the adjective turns into an independent substantive — “colored” (coloré) becomes “color” (couleur), “hard” (dur) becomes “hardness” (dureté), “prudent” (prudent) becomes “prudence” (prudence).
This double structure also explains something so obvious we never think to question it: why an adjective can modify a noun (substantive) at all. “White paper” works, on this account, because “white” is not simply an independent name for “whiteness” — it carries a hanging, unfilled part (its connotation), which lets it attach directly to a substantive like “paper” that fills that slot. The substantive “whiteness,” by contrast, is already complete in itself, so it cannot attach to “paper” directly — which is why we can only say “the whiteness of paper,” never “whiteness paper.” Joining two substantives requires some separate device (a preposition, a case ending); an adjective needs no such mediator — its connotation is a built-in “vacancy” that lets it lean directly on a substantive.
This operation is close to what modern linguistics calls “derivation” — the work of a suffix that turns an adjective into an abstract noun. But Arnauld and Lancelot’s real interest lies less in the change of form than in what that change does to the “manner of meaning.” Forming “hardness” from “hard,” on their account, is not adding anything — it is stripping away the connotation (the vague pointing) and leaving only the distinct meaning behind. This reading turns what looks like mere morphology into an analysis of a word’s semantic structure itself. And the operation runs both ways: not only adjective-to-substantive (dur→dureté), but the reverse — giving a substantive a fresh connotation to make it an adjective (couleur→coloré) — follows, by implication, from exactly the same principle turned inside out.
3. Directly but Vaguely / Indirectly but Distinctly
From here the authors draw a seemingly paradoxical conclusion. What “white” means directly is not, in fact, whiteness — it is the subject.
ils signifient le sujet directement… in recto, quoique plus confusément ; et… la forme qu’indirectement… in obliquo, quoique plus distinctement.
Adjectives signify the subject directly — in recto — though more vaguely; and signify the form (the attribute) only indirectly — in obliquo — though more distinctly.
“White” signifies its subject directly — what it actually points to in the sentence is the white thing itself. But only vaguely, without specifying which thing. Conversely, the attribute “whiteness” is signified only indirectly — it is never named outright — yet its content is perfectly distinct. This observation, that directness and distinctness need not coincide and may in fact run opposite each other, is remarkably fine-grained semantics for its time.
This paradoxical pairing — direct but vague for the subject, indirect but distinct for the attribute — can look at first like mere hair-splitting. But there is a real payoff: it explains how an adjective and a substantive divide their labor inside a phrase like “white snow.” “Snow” is the subject (substantive), forming the sentence’s skeleton; “white” leans against that snow, adding the information “whiteness” to it. If the adjective signified only whiteness directly, and never pointed at a subject at all, there would be no way to explain why “white” can combine with the particular substantive “snow” in the first place. It is precisely because the adjective signifies its subject directly — however vaguely — that it can freely combine with any substantive and add new information to it. That, one suspects, was the point the authors were driving at all along.
4. Why “King” Behaves Like a Substantive Even Though It Is Logically an Adjective
Finally, the authors take up the reverse exception. Words for professions and stations — “king” (roi), “philosopher” (philosophe), “painter” (peintre), “soldier” (soldat) — stand alone in a sentence just like substantives. Yet on reflection these words signify an accident (a role someone occupies), so logically they ought to be adjectives.
tels sont les noms de diverses professions des hommes, comme roi, philosophe, peintre, soldat.
Such are the names of the various professions of men, as king, philosopher, painter, soldier.
Why, then, are they treated as substantives? The authors’ answer is simple: the thing these words lean on (the addressee of their connotation) is almost always fixed to one single category — “a human being” — so it goes without saying and can safely be left out. Because omitting it causes no ambiguity, these words have, in practice, acquired the substantive’s privilege of standing alone.
Turn this explanation around, and it also says that a word whose connotation cannot be pinned to a single addressee in advance can never become a substantive. “White“‘s connotation can attach to snow, paper, a wall, almost anything — which is exactly why “white” never stands alone and always needs some noun. “King“‘s connotation, by contrast, is effectively pinned to just one category, “human being” — so dropping it causes no confusion for the listener. “Prudent” (prudent) is a similar but subtly different case: prudence can apply to more than humans (a “prudent decision,” for instance), so its connotation cannot be pinned to a single addressee, and it resists becoming an independent substantive. What makes profession-words special is that their addressee is confined almost entirely to one kind of substance — human beings. Whether a word can become a substantive, in other words, is not a fixed property it is born with, but a matter of degree — how tightly its connotation’s addressee can be pinned down. Here too, Arnauld and Lancelot treat individual words not as isolated entries but as points within a single semantic system running through the whole language.
Summary
| Point | Content |
|---|---|
| Starting point | Words meaning substance = substantives; words meaning accident = adjectives |
| The refining move | Not “what is meant” but “how it is meant” — the presence of connotation |
| The adjective’s double structure | It signifies the attribute distinctly and the subject vaguely, both at once |
| The paradox | Subject: direct but vague (in recto, confuse). Attribute: indirect but distinct (in obliquo, distincte) |
| The exception | Profession-words like “king,” “philosopher” stand alone like substantives because their connotation’s addressee is fixed to “a human being” |
The term “connotation” itself would go on to be inherited, in transformed guises, by 19th-century logic (Mill’s intension and extension) and 20th-century semantics (Frege’s sense and reference). The observation that a single word can, at once, “mean distinctly” and “point vaguely” began, more than 350 years ago, with one French adjective.
Why It Still Matters
The word “connotation” today is mostly used loosely, as something like “an implied nuance.” What Arnauld and Lancelot meant by it was far more precise — an observation spanning syntax and semantics about a single word carrying a distinct meaning and a vague pointing at once. What sets their usage apart from the modern one is that it functions as a tool to explain actual grammatical behavior: why an adjective cannot stand alone, why it freely combines with nouns, why profession-words alone get to stand alone as an exception. Modern semantics handles this same duality with tools like the intension/extension distinction, or variable binding in predicate logic. A predicate like “white(x)” carries one variable x — corresponding to connotation’s addressee — and depending on what fills that variable, becomes “white snow” or “white paper”: this modern notation is close to a formal rewriting of exactly what Arnauld and Lancelot were reaching for with the word “connotation.” Yet even with sharper tools in hand today, the question at the starting point — why does a single word like “white” mean both “whiteness” and “the white thing” at once, and with different degrees of clarity — remains just as fresh as it was when this 17th-century chapter first posed it. The eye that could find this much intricate semantic machinery inside the adjective, of all things — the least glamorous part of speech in the whole grammar — is exactly why this book is still read after more than 350 years. Even in treating a profession-word like “king” as a boundary case wavering between substantive and adjective, the authors were not imposing rules from above; they built their theory by patiently gathering up the subtle wavering already present in actual usage. The next article turns this same eye toward the verb — toward finding, beneath the seemingly simple act of “affirming,” just how precise a structure lies hidden.