A verb means “affirmation,” as we saw in the overview article. But an actual verb does more than affirm — it also tells you who is affirming, how many of them, and when, all through a single ending. “I see” and “we saw” are both, grammatically, just one verb each, yet that single word folds in the speaker’s person, whether it’s singular or plural, and the temporal relation in which the event occurred. Chapters XIV–XV and XVIII of Part II unpack this compression one layer at a time.
1. Where Did “I” Go? — Person as Ellipsis
Why does a verb change form for person? The authors tell it as a history of necessary shortcuts.
la diversité des personnes et des nombres dans les verbes, est venue de ce que les hommes, pour abréger, ont voulu joindre dans un même mot… le sujet de la proposition.
(ToL working translation. Disclaimers on subsequent quotes are omitted; the same applies throughout.) The diversity of persons and numbers in verbs arose because people, wanting to abbreviate, sought to join the subject of the proposition into the same word as the verb.
“When I speak,” the authors continue, “the subject of the proposition is always the first-person pronoun (ego, moi, je).” So rather than stating the pronoun every time, it sufficed to give the affirming word a certain ending marking that one is speaking of oneself — and that is what we call the first person. The same logic produced the second person and, in the plural, both again. Since the subject of a proposition is often neither speaker nor addressee, the authors add, a third person had to be invented to attach to every other possible subject.
Here the authors insert an unexpectedly sharp observation. The word “person” (personne) properly applies only to rational, intelligent substances — yet the third person is used for things too. The very name “third person,” in other words, is an imprecise borrowing.
naturellement ce qu’on appelle troisième personne devroit être le thème du verbe…
What we call the third person should, by nature, be the theme (base form) of the verb — and indeed it is, in every Oriental language.
The logical order, in other words, runs backward from what we might expect: first comes a form that simply affirms without pointing at anyone (the third person), and only afterward are special inflections layered on for the first and second persons. This is why Latin can nearly omit its first- and second-person pronouns (video, vides suffice on their own) — yet, oddly, a vernacular like French insists on restating them, je vois, tu vois. The authors guess this is because conjugated endings often collapse the distinction between persons (aimer’s first and third person are both aime).
The authors note one more use of the same person-and-number ending. Strip the pronoun away from the verb ending and use it alone, and it becomes an imperative — vois (look!), aime (love!), lis (read!). The very same material that marks “who” when paired with a pronoun marks “command” — an altogether different mood — when the pronoun is dropped. One ending, depending on context, can carry entirely different grammatical functions — an observation that resonates with the tense discussion below.
This observation anticipates, by more than 350 years, what modern language typology calls the “pro-drop” versus “obligatory subject” distinction. Languages like Latin, Spanish, or Japanese, where the ending (or context) alone identifies the person, can drop the subject pronoun; languages like French or English, where the ending’s person distinctions have worn down, cannot. The reasoning the authors offer — “conjugated endings collapse together” — is strikingly close to what modern syntacticians cite when explaining pro-drop availability. The same compression governs plural forms: nous voyons (we see) and vous voyez (you [pl.] see) fold the pronouns nos and vos into the ending, by exactly the same logic of ellipsis as the singular. The authors further note that Greek adds a dual number (used, if rarely, for exactly two things) beyond singular and plural, and that some Oriental languages mark the sex — masculine or feminine — of a given person’s subject with distinct endings, comparing across languages just how finely the categories of person and number can be subdivided.
2. Three Simple Tenses, Three Compound Tenses
Tense is another piece of information folded into the verb ending. A single letter or two can tell you when an event occurred.
Il n’y a que trois temps simples : le présent… le passé… et le futur.
There are only three simple tenses: present (amo, I love), past (amavi, I have loved), and future (amabo, I will love).
French introduces one further distinction here — whether the past is described as “just now completed” or “completed at some unspecified time.”
l’un qui marque la chose précisément faite… on nomme défini, comme j’ai écrit ; et l’autre qui la marque indéterminément faite… on nomme indéfini ou aoriste, comme j’écrivis.
One marks the thing as precisely done — called the definite preterite (j’ai écrit, I have written); the other marks it as done at an unspecified time — called the indefinite, or aorist (j’écrivis, I wrote).
The authors’ example is a fine one: “I wrote this morning” can only be j’ai écrit ce matin, never j’écrivis ce matin — the aorist form, they observe, applies only to a past at least a full day removed from the day of speaking, a rule most French speakers apply without ever consciously noticing it.
This défini/indéfini pair is, in fact, the very distinction modern French draws between the passé composé and the passé simple. Modern school grammar usually frames the difference as one of register — passé simple for writing, passé composé for speech — but as of 1660 the authors were tracking something else: sheer temporal distance from the day of utterance. They add that Spanish and Italian sometimes conflate the two preterites, already comparing, across the Romance languages of the 17th century, how unstable this défini/indéfini boundary could be.
The section on compound tenses is where the authors’ prose is at its sharpest. The imperfect (imparfait) depicts a point in the past “as though it were present.”
cum intraret coenabam, je soupois lorsqu’il est entré… l’action de souper est bien passée au regard du temps auquel je parle ; mais je la marque comme présente au regard de la chose dont je parle.
“I was dining when he entered” — the act of dining is indeed past relative to the moment I am speaking now, but relative to the other event I am describing (his entrance), I mark it as present.
The pluperfect marks “a further past within the past”; the future perfect marks “a past as seen from a future point.” The authors present this as a systematic, logically derivable set — and even note that a fourth compound form, a future relative to the present, could in principle have existed, but the vernaculars made do with the simple future instead. Latin, for instance, has only cum coenabo intrabis (you will enter while I am dining) — “my dining” is itself future, yet relative to another future event (your entrance) it is marked as present. The authors carefully point out this one unfilled square in the otherwise neat grid completed by the imperfect, pluperfect, and future perfect.
Three simple tenses and three compound tenses: six tenses in all. Where the simple tenses mark three absolute positions — present, past, future — the compound tenses always mark the relative order of two events, which way one is displaced from the other. The authors let this structural symmetry emerge across the whole chapter.
If the past can be split this finely, the future should split the same way — and the authors point out exactly the same precise/vague pair for the future. Greek has a “near future” (paulopost-future, literally “a little after”) carrying the nuance of “as good as done already,” something close to the Japanese “about to…” Against this, the simple future (ferai, faurai) states flatly that something will happen at some unspecified time. The authors line this pair up neatly against the définie/indéfinie preterite pair, showing that the whole tense system runs along a single “precise or vague” axis. They add that Oriental languages such as Hebrew have only a two-way past/future split, without the finer imperfect/pluperfect distinctions — which, they note, often produces ambiguity. The implication: tense subdivision isn’t a mark of a language’s advancement so much as a device for avoiding it. The flip side, of course, is that a language with fewer tenses isn’t inferior — it simply makes the same distinctions through context or adverbs rather than endings.
3. The Verb Inside “I Think, Therefore I Am”
In Chapter XVIII the authors bring up an unexpected case: the “merely affirming” verb (the substantive verb, estre/sum) can sometimes mean the concrete attribute of existence itself.
le verbe substantif sum, je suis, est souvent adjectif… on y joint le plus général de tous les attributs, qui est l’être ; comme lorsque je dis, Je pense, donc je suis, je suis signifie je suis un être, une chose.
Even the substantive verb sum, I am, is often adjectival, because instead of taking it as meaning affirmation alone, one joins to it the most general of all attributes — being — as when I say, I think, therefore I am, where “I am” means “I am a being, a thing.”
Citing Descartes’ Je pense, donc je suis (Discourse on Method, 1637), the authors find, in a single verb, a philosophical weight that exceeds mere grammatical affirmation — the verb chapter unexpectedly connects to Descartes’ conclusion.
This observation carries real weight. A verb that is supposed to “merely affirm” sometimes shoulders the concrete content of “existing” — a seemingly small exception that is, for their theory of the verb overall, no mere aside but an honest admission of where the theory strains. For the Port-Royal authors, a verb’s essence was supposed to be affirmation, nothing more and nothing less — as we saw in the overview article, the “is” in “God is just” was meant to be a purely logical operation binding subject to attribute. Yet the verb “to be” (estre) turns out to be special: its own content can become “existing” itself. “I think” alone is an ordinary verbal sentence, but once “therefore I am” follows, that “am” is no longer a bare “yes, indeed” — it becomes a substantive claim: “I am one being.” The same word, estre/sum, behaves transparently as a logical copula in one context and, in another, carries the heaviest attribute imaginable, existence itself. Arnauld formalized this duality precisely, in a grammar-book chapter, barely two decades after Descartes. Philosophy would later revisit this as the question “is existence a predicate?” (fully argued out by Kant and Frege) — its seed, already visible on this page, deserves more notice than it usually gets.
The same chapter also covers the traditional threefold division of verbs into active (actif), passive (passif), and neutral (neutre). The active means an action paired with its corresponding passion (battre, to strike / être battu, to be struck), and the authors further distinguish a “real action” that actually reaches its object (rompre, to break; tuer, to kill) from a merely “intentional action” that only apprehends its object in thought (aimer, to love; connoître, to know; voir, to see). This distinction is precociously sophisticated semantics: “to break” physically changes its object, but “to love,” “to know,” “to see” do not — an insight that prefigures the modern linguistic distinction between “affect verbs” and “experiencer verbs.”
The neutral covers verbs that mean no action, only a quality or state (albet, he is white; friget, he is cold; tepet, he is lukewarm). Classical languages like Latin and Greek build active and passive from the same stem through inflection alone, and, the authors note, some Oriental languages give a single verb three active forms each with its own passive, plus a reflexive that turns the action back on its own subject (roughly, s’aimer, to love oneself). Vernaculars like French, lacking a passive ending, substitute the substantive verb plus a past participle instead (je suis aimé). The active/passive/neutral division itself is inherited from ancient grammarians, but the authors’ move to ground it in a semantic criterion — does the action’s target actually change — reflects the rationalist stance that defines this grammar.
Summary
| Point | Content |
|---|---|
| Origin of person | An ellipsis that compresses the pronoun (je, tu) into the verb ending. Logically, the third person (unspecified affirmation) is the verb’s base form |
| Simple tenses | Present, past, future. French further splits the past into definite and indefinite (aorist) preterites |
| Compound tenses | Imperfect (a present within the past), pluperfect (a past within the past), future perfect (a past seen from the future) |
| The double nature of the substantive verb | “To be” (estre) can exceed mere affirmation and mean the attribute of existence itself — as in Descartes’ Je pense, donc je suis |
| The threefold verb classification | Active (action), passive (undergoing action), neutral (quality/state only). Vernaculars lack a passive ending and substitute the substantive verb plus participle |
Who is speaking, how many of them, when the event took place, and sometimes even a philosophical claim about existence — all folded into a single verb ending. What Arnauld and Lancelot saw in the verb was never a mere conjugation table, but a miniature of how human thought compresses into the finite signs of language.
Why It Still Matters
Modern linguistics groups person, number, and tense together under the terms “inflection” or “agreement,” describing cross-linguistic variation flatly as differences in parameters. But the answer to the “why” behind that description — why does a verb change form for person at all, why must present/past/future be supplemented by compound forms like the imperfect and future perfect — was already given by Arnauld and Lancelot in the 17th century, inside a single grammar book. Their answer is consistent throughout: grammatical form is not arbitrary decoration but a reasoned device for compressing, into as few words as possible, the thought the speaker wants to convey — who, when, in what relation. That the Port-Royal Grammar can connect the analysis of a single verb to the heart of Cartesian philosophy is part of why it is still read, more than 350 years later, in both the history of grammar and the philosophy of language: its habit of seeing a reason behind every form.