1. Dialectic is the science of arguing well. Now we argue, of course, with words. Words, then, are either simple or combined. Simple words are those that signify some one thing, as when we say "homo, equus, disputat, currit" ("a man, a horse, he argues, he runs"). And do not be surprised that "disputat," though it is composed of two parts, is nonetheless counted among the simple words. For the matter is made clear by the definition. We said that a word is simple if it signifies some one thing. And so "disputat" is included in this definition, by which "loquor" ("I speak") is not included when we say it. For although it is a single word, it does not have a simple signification, since it also signifies the person who is speaking. For that reason it is already subject to truth or falsity, for it can be both denied and affirmed. And so every first and second person of a verb, although it is uttered by itself, will nonetheless be counted among the combined words, because it does not have a simple signification. For whoever says "ambulo" ("I walk") makes both the walking understood and his very self who walks; and whoever says "ambulas" ("you walk") likewise signifies both the thing that is done and the one who does it. But the one who says "ambulat" ("he walks") signifies nothing other than the walking itself. For this reason the third person of a verb is always counted among the simple words, and it cannot yet be either affirmed or denied—except in the case of those verbs to which, by the custom of speaking, the signification of a person necessarily adheres, as when we say "pluit" ("it rains") or "ninguit" ("it snows"): even if it is not added who rains or snows, still, because it is understood, the word cannot be counted among the simple ones.

2. Combined words are those that, linked together, signify several things, as when we say "homo ambulat" ("a man walks") or "homo festinans in montem ambulat" ("a man hurrying up the mountain walks") and the like. But of combined words, some comprise a complete thought, like those just mentioned, while others await something to complete the thought—like those very ones we have just spoken, if you take away the word "ambulat." For although the words "homo festinans in montem" are combined, the statement still hangs unfinished. Once, therefore, we have set aside those combined words that do not complete a thought, there remain the combined words that do comprise a thought. Of these there are likewise two kinds. For a thought is either comprised in such a way as to be liable to truth or falsity, as is "omnis homo ambulat" ("every man walks") or "omnis homo non ambulat" ("not every man walks") and the like; or a thought is completed in such a way that, though it carries out the mind's intention, it still cannot be affirmed or denied—as when we command, when we wish, when we curse, and the like. For whoever says "perge ad villam" ("go to the farmhouse") or "utinam pergat ad villam" ("would that he might go to the farmhouse") or "dii illum perduint" ("may the gods destroy him") cannot be charged with lying or believed to speak the truth. For he has affirmed or denied nothing. Therefore neither do such thoughts come into question in a way that calls for a debater.

3. But those that do call for a debater are either simple or combined. Simple ones are those that are uttered without any coupling to another thought, as is the one we cited, "omnis homo ambulat" ("every man walks"). Combined ones are those about whose coupling a judgment is made, as is "si ambulat, movetur" ("if he walks, he is moved"). But when a judgment is made about the coupling of thoughts, it goes on until one arrives at the conclusion. Now the conclusion is what is produced from things granted. What I mean is this. He who says "si ambulat, movetur" ("if he walks, he is moved") wants to prove something, so that, once I have granted this to be true, it remains for him to show that the man walks and that the conclusion follows, which can no longer be denied—namely, that he is moved—or it remains for him to show that he is not moved, so that the conclusion follows, which likewise cannot but be granted—namely, that he does not walk. Again, if he should want to put it this way, "homo iste ambulat" ("this man walks"), it is a simple thought; and if I grant it, and he adds another, "quisquis autem ambulat movetur" ("but whoever walks is moved"), and I grant this too, then from this coupling of thoughts—though they were uttered and granted one by one—there follows that conclusion which must now of necessity be granted, namely, "homo iste igitur movetur" ("this man, therefore, is moved").

4. Having briefly established these things, let us consider the parts one by one. For there are two primary ones: one concerning the things that are said simply, where there is, as it were, the matter of dialectic; the other concerning the things that are said in combination, where the work, as it were, now appears. The part concerning simple words is called "on speaking" (de loquendo). But the one concerning combined words is divided into three parts. For once we have set aside the combination of words that does not complete a thought, the combination that completes a thought in such a way as not yet to raise a question or call for a debater is called "on the uttered statement" (de eloquendo); the one that completes a thought in such a way that a judgment is made about simple thoughts is called "on the proposition" (de proloquendo); the one that comprises a thought in such a way that a judgment is made even about the coupling itself, until one arrives at the conclusion, is called "on the sum of propositions" (de proloquiorum summa). Let us, then, set out each of these parts more carefully.

5. A word (verbum) is the sign of any given thing, which can be understood by the hearer when uttered by the speaker. A thing (res) is whatever is either perceived by the senses, or understood, or hidden. A sign (signum) is what presents both itself to the senses and something beyond itself to the mind. To speak (loqui) is to give a sign by an articulate utterance. And I call "articulate" that which can be captured in letters. Whether all these things that have been defined are rightly defined, and whether up to this point the words of the definition ought to have been pursued by further definitions, that passage will indicate where the discipline of defining is treated. For now, attend closely to what is at hand. Every word is a sound. For when it is in writing, it is not a word but the sign of a word; since, when the letters are examined by the reader, what bursts forth in the voice occurs to his mind. For what else do written letters present but themselves to the eyes, and beyond themselves voices to the mind? And we said a little earlier that a sign is what presents both itself to the senses and something beyond itself to the mind. The things we read, therefore, are not words but signs of words. But just as, although a letter is the smallest part of an articulate utterance, we nonetheless misuse this term so as to call it a "letter" even when we see it written—although it is wholly silent and is no part of an utterance but appears as the sign of a part of an utterance—so too it is called a "word" when it is written, although what shines forth is the sign of a word, that is, the sign of a signifying utterance, and not the word itself. Therefore, as I had begun to say, every word is a sound. But what it sounds is nothing to dialectic. For the sound of a word is at issue when one asks or observes how it is softened by the arrangement of its vowels or gapes by their clashing, and likewise how it is knotted by the interposition of consonants or roughened by their massing, and of how many and what kind of syllables it consists, where the poetic rhythm and accent lie—matters that the grammarians treat as the business of the ears alone. And yet, when these are discussed, it is not outside dialectic. For this is the science of arguing. But since words are signs of things, when they hold this office regarding things—and those are signs of words by which the latter are discussed (for since we cannot speak of words except with words, and when we speak we speak only about some things)—it occurs to the mind that words are signs of things in such a way that they do not cease to be things themselves. When, therefore, a word issues from the mouth, if it issues on its own account—that is, in order that something be asked or argued about the word itself—then it is certainly a thing subject to discussion and inquiry, but the thing itself is called a "word" (verbum). But whatever the mind, and not the ears, perceives from a word and holds enclosed within the mind itself is called the "sayable" (dicibile). And when a word issues not on its own account but to signify something else, it is called an "utterance" (dictio). The thing itself, however, which is now neither a word nor the conception of a word in the mind—whether it has a word by which it can be signified or does not—is called nothing other than a "thing" (res), now by its proper name. Let these four, then, be held distinct: word, sayable, utterance, thing. What I called "word" both is a word and signifies a word. What I called "sayable" is a word, yet does not signify a word, but what is understood in the word and is held in the mind. What I called "utterance" is a word, but signifies those two at once, that is, both the word itself and what occurs in the mind through the word. What I called "thing" is a word, which signifies whatever is left over besides those three that have been mentioned. But I see that these things must be illustrated by examples. Suppose, then, that a boy is questioned by some grammarian in this way: "arma quae pars orationis est?" ("'arma'—what part of speech is it?"). What was said, "arma," was said on its own account, that is, a word on account of the word itself. But the rest, when he says "quae pars orationis" ("what part of speech"), was perceived by the mind or uttered by the voice not on its own account, but on account of the word that was uttered, "arma." But when they are perceived by the mind, before the voice they will be sayables; and when, for the reason I have given, they have burst forth into voice, they have become utterances. But the very word "arma," which here is a word, was an utterance when it was spoken by Virgil: for it was not uttered on its own account, but in order that by it might be signified either the wars that Aeneas waged, or the shield, or the other things that Vulcan forged for the hero. But the wars themselves, or the arms, which were borne or not borne by Aeneas—those very things, I say, which, when they were being waged and existed, were seen, and which, if they were now present, we could even point to with a finger or touch, and which, even if they are not thought of, do not on that account fail to have existed—those things, then, in themselves are neither words nor sayables nor utterances, but things, which are now called by their proper name "things." It is therefore for us, in this part of dialectic, to treat of words, of sayables, of utterances, and of things. In all of which, since partly words and partly non-words are signified, there is nonetheless nothing about which it is not necessary to argue with words. And so let those things be discussed first through which we are permitted to discuss the rest.

6. Therefore any word whatever, the sound aside—about which to argue well belongs to the faculty of the dialectician, not to the discipline of dialectic, just as Cicero's speeches for the defense belong indeed to the rhetorical faculty, but rhetoric itself is not taught by them—every word, then, apart from what it sounds, of necessity calls four things into question: its origin, its force, its inflection, its ordering.
The origin of a word is asked about when one inquires why it is so called—a matter, in my judgment, too inquisitive and not very necessary. Nor have I been pleased to say this because the same seems true to Cicero as well. For who would need authority in a matter so clear? But if it were at all of much help to set out the origin of a word, it would still be foolish to undertake what is surely endless to pursue. For who could discover, of whatever has been said, why it was so said? Add to this that, like the interpretation of dreams, so too the origin of words is judged according to each man's ingenuity. For look: someone supposes that "verba" ("words") are so called because they strike the ear (verberent), as it were. No, says another, because they strike the air (aerem). But what does it matter to us? It is no great dispute, for each derives the origin of this term from striking (verberando). But see what a quarrel a third raises from the side. For, because (he says) we ought to speak the truth (verum), and a lie is by nature itself something hateful, "verbum" was named from "verum" (the true). Nor was a fourth wit lacking. For there are those who suppose that "verbum" is indeed derived from "verum," but that, the first syllable having been duly noted, the second ought not to be neglected. "Verbum," when we say it (they say), has its first syllable signifying "verum" (the true), its second a sound. For this is what they want "bum" to be—whence Ennius called the sound of feet "bombum pedum" ("the boom of feet"), and the Greeks "βοῆσαι" (boēsai) means "to shout," and Virgil says "reboant silvae" ("the woods resound"). Therefore "verbum" is so called as if from "verum boando," that is, "sounding the truth." And if this is so, the name itself prescribes that we not lie when we make a word; but I fear that those very men who say such things are lying. So it now falls to you to judge whether we should suppose "verbum" to be derived from "verberando" (striking), or from "verum" alone, or from "verum boando" (sounding the truth)—or rather not to trouble ourselves about where it comes from, since we understand what it signifies. Still, I want you to take brief note of this topic—that is, the origin of words—for a little while, lest we seem to have passed over any part of the work undertaken. The Stoics maintain (whom Cicero in this matter, being Cicero, mocks) that there is no word whose certain origin cannot be explained. And because it was easy to press them in this way—if you were to say that this is endless, since the origin of the very words by which you interpret the origin of some word may in turn be sought from you—they say that one must inquire until one arrives at the point where the thing chimes by some likeness with the sound of the word, as when we speak of the ringing of bronze, the neighing of horses, the bleating of sheep, the clangor of trumpets, the grating of chains. For you see that these words sound just as the very things do that are signified by these words. But because there are things that make no sound, in their case a likeness of touch holds good, so that, if they touch the sense gently or harshly, the gentleness or harshness of the letters, as it touches the hearing, has likewise given them their names: as "lene" ("smooth") itself, when we say it, sounds smoothly. Who, again, would not judge "asperitas" ("roughness") to be rough in its very name? It is smooth to the ears when we say "voluptas" ("pleasure"), rough when we say "crux" ("cross"). Thus the things themselves affect us as the words are perceived. Honey (mel)—as sweetly as the thing itself touches the taste, so smoothly does it touch the hearing by its name. "Acre" ("bitter, harsh") is rough in both respects. "Lana" ("wool") and "vepres" ("brambles"): as the words are heard, so are those things touched. These, they believed, are, as it were, the cradles of words, where the sense of things agreed with the sense of sounds. From here the license of naming proceeded to the likeness of things themselves to one another; so that, while (for the sake of example) "crux" was so called because the harshness of the word itself agrees with the harshness of the pain that the cross produces, "crura" ("legs") were nonetheless so named not on account of the harshness of pain but because in length and hardness they are, among the other limbs, more like the wood of the cross. Then they came to extended use (abusio), so that a name is borrowed not from a like thing but from a neighboring one, as it were. For what likeness has the signification of "parvum" ("small") and "minutum" ("diminished"), since a thing can be small that not only is not at all diminished but has even grown somewhat? Yet we say "minutum" for "parvum" on account of a certain neighboring relation. But this extended use of a term is within the speaker's power, for he has "parvum" so that he need not say "minutum." This bears more on what we now wish to show: that when a "piscina" ("fishpond," also a swimming-pool) is spoken of in the baths, in which there is no fish nor anything like fish, it nonetheless seems to be named from fish (pisces) on account of the water, where fish have their life. So the term was borrowed not by a transfer through likeness but by a certain neighboring relation. And if someone should say that men become like fish by swimming, and that from this the name "piscina" was born, it is foolish to resist, since neither account is at odds with the matter and each is hidden from us. Yet this turns out well, that by this one example we can now distinguish how the origin of a word that is seized from a neighboring relation differs from the one that is drawn from a likeness. From here advance was made even to a contrary. For "lucus" ("grove") is thought to be so called because it scarcely lets in light (luceat), and "bellum" ("war") because it is not a fair thing (res bella), and the name "foedus" ("treaty") because it is not a foul thing (res foeda). But if it is named from the foulness of a pig (porcus), as some would have it, the origin returns to that neighboring relation, since what is brought about is named from that through which it is brought about. For that neighboring relation extends very widely and is divided into many parts: either through the agent, as this very case of the treaty from the foulness of the pig, through which the "foedus" is brought about; or through the effects, as "puteus" ("well") is believed to be so called because its effect is drinking (potatio); or through that by which a thing is contained, as they would have "urbs" ("city") named from "orbis" ("circle"), because under good auspices the place is wont to be encircled by the plow—a thing Virgil too recalls, where "Aeneas marks out the city with the plow"; or through that which contains, as if someone should affirm "horreum" ("granary"), with a letter changed, to be named from "hordeum" ("barley"); or by extended use, as when we say "horreum" and wheat is stored there; or the whole from a part, as by the name "mucro" (the tip, which is the topmost part of a sword) we call the sword; or a part from the whole, as "capillus" ("hair") as if "capitis pilus" ("the hair of the head"). Why should I go on further? Whatever else can be reckoned up, you will see that the origin of a word is contained either in a likeness of things and sounds, or in a likeness of the things themselves, or in a neighboring relation, or in a contrary. This we cannot pursue beyond the likeness of sound, but even this we cannot always do. For words are innumerable whose origin, of which an account can be given, either does not exist, as I judge, or is hidden, as the Stoics contend.
Yet see for a moment how they think one arrives at those cradles of words—or rather their stock, indeed their seed—beyond which they forbid the origin to be sought, nor, if anyone should wish, can anything be found. No one denies that the syllables in which the letter "v" holds the place of a consonant—as they are at the beginning of these words, "vafer, velum, vinum, vomis, vulnus" ("crafty, sail, wine, plowshare, wound")—give off a thick and, as it were, forceful sound. The custom of speaking confirms this too, since from certain words we remove these syllables lest they burden the ear. For whence is it that we say "amasti" more readily than "amavisti," and "abiit" rather than "abivit," and innumerable things in this manner? Therefore, when we say "vim" ("force"), the sound of the word, as has been said, being forceful as it were, agrees with the thing it signifies. Now from that neighboring relation, through what they bring about—that is, because they are violent (violenta)—"vincula" ("bonds") can seem to be so called, and "vimen" ("withe") by which something is bound. From this come "vites" ("vines"), because they grasp by their windings the props on which they rest. Hence already, on account of a likeness, Terence called a bent old man "vietus" ("shrunken, wizened"). Hence the ground that is worn winding by the feet of travelers is called "via" ("way"). But if "via" is rather believed to be so called because it is worn by the force (vis) of feet, the origin returns to that neighboring relation. But let us suppose it is named from the likeness of a vine (vitis) or a withe (vimen), that is, from bending. Someone, then, asks me: why is it called "via"? I answer: from bending, because the ancients called what is bent and curved, as it were, "vietum"; whence they also call "vieti" the pieces of wood of wheels that are encircled by the rim. He goes on to ask whence "vietum" is said to mean "bent"; and here I answer: from the likeness of a vine. He presses and demands whence the name "vitis" comes thus; I say, because it binds (vinciat) the things it has grasped. He scrutinizes whence "vincire" ("to bind") itself is derived; we shall say, from "vis" (force). He will ask why "vis" is so called; the account will be returned that, by its robust and, as it were, forceful sound, the word agrees with the thing it signifies. Beyond that he has nothing to ask. But in how many ways the origin of words is varied by the corruption of their utterances, it would be foolish to pursue. For it is both long and less necessary than the things that have been said.

7. Now let us briefly consider the force of words, as far as the matter allows. The force of a word is that by which is known how much it avails. And it avails just as much as it can move the hearer. Further, it moves the hearer either according to itself, or according to what it signifies, or jointly from both. But when it moves according to itself, this pertains either to the sense alone, or to art, or to both. The sense is moved either by nature or by custom. It is moved by nature when it is offended if someone names King Artaxerxes, or is soothed when it hears "Euryalus." For who, even if he had never heard anything of these men whose names these are, would nonetheless not judge that there is the greatest harshness in the one and smoothness in the other? The sense is moved by custom when it is offended if someone is called (for example) "Motta," and is not offended when it hears "Cotta." For here it makes no difference to the sweetness or unsweetness of the sound, but the inner recesses of the ears are decisive—whether they receive the sounds passing through them as known guests or as strangers. The hearer is moved by art, however, when, on a word being uttered to him, he attends to what part of speech it is, or to anything else he has learned in those disciplines that are handed down about words. But a word is judged from both—that is, both from sense and from art—when reason notes what the ears measure and assigns the name accordingly: so that, when "optimus" ("best") is said, as soon as one long and two short syllables of this name have struck the ear, the mind by art at once recognizes a dactylic foot. But now a word moves not according to itself but according to what it signifies, when, the sign having been received through the word, the mind beholds nothing other than the very thing of which that is the sign which it received—as when "Augustine" is named, nothing other than I myself am thought of by one to whom I am known, or some other man occurs to the mind, should one who does not know me hear this name, or one who knows another who is called Augustine. But when a word moves the hearer at once both according to itself and according to what it signifies, then both the utterance itself and what is announced by it are attended to together. For whence is it that the chastity of the ears is not offended when it hears, "with his hand, his belly, his member he had wasted his good patrimony"? Yet it would be offended if the obscene part of the body were called by its filthy and vulgar name, since the thing is the same of which each term is used—except that in the one the baseness of the thing signified is veiled by the decorousness of the signifying word, while in the other the ugliness of both would strike the sense and the mind: just as a harlot is no other woman, but yet seems otherwise in that dress in which she is wont to stand before a judge, and otherwise in that in which she lies in a debauchee's bedchamber. Since, then, it appears that the force of words is so great and so manifold—which we have briefly and summarily touched on, as time allowed—a twofold consideration arises here regarding the sense: partly for the sake of setting forth the truth, partly for the sake of preserving decorum; of which the first pertains especially to the dialectician, the second to the orator. For although argument ought not to be inept, nor ought eloquence to be deceitful, still in the former the desire for learning often, indeed almost always, despises the delights of the hearing, and in the latter the more untutored multitude supposes that what is said ornately is also said truly. Therefore, since it is apparent what is proper to each, it is plain both that the debater, if he has any care for giving delight, should be sprinkled with a rhetorical coloring, and that the orator, if he wishes to persuade of the truth, should be strengthened by sinews and bones, as it were, of dialectic—which nature itself, in our bodies, could neither withhold from strength of force nor allow to lie exposed to the offense of the eyes.

8. And so now, for the sake of judging the truth, which is dialectic's profession, let us see what impediments arise from this force of words, of which we have scattered some seeds. For what impedes the hearer in seeing the truth, in the case of words, is either obscurity or ambiguity. Between the ambiguous and the obscure there is this difference: that in the ambiguous several meanings present themselves, of which it is unknown which ought rather to be taken, whereas in the obscure nothing, or little, that might be attended to appears. But where there is little that appears, the obscure is like the ambiguous: as if someone setting out on a journey were met by some place where two or three or even (so to speak) many roads meet, and there, by the thickness of a fog, none of the roads that exist were visible. He is first deterred from proceeding by the obscurity; but when the fogs have begun to thin somewhat, something is seen, of which it is uncertain whether it is a road or rather the proper and brighter color of the ground. This is the obscure that is like the ambiguous. But when the sky clears, as far as is enough for the eyes, the laying-out of all the roads is now plain, but by which one must proceed is doubtful, not from obscurity but from ambiguity. Likewise there are three kinds of obscure things. One is what lies open to the sense but is closed to the mind: as if someone should see a painted pomegranate who had neither ever seen one nor at all heard what sort of thing it was—it is not a matter of the eyes but of the mind that he does not know what thing the picture is of. The second kind is where the thing would lie open to the mind, were it not closed to the sense: as is a man painted in the dark. For when it appears to the eyes, the mind will not doubt that it is a painted man. The third kind is that in which the thing is hidden even from the sense, which nonetheless, if it were laid bare, would stand out no more to the mind—which kind is the most obscure of all: as if an untutored man were forced to recognize that painted pomegranate even in the dark.
Now turn your mind to words, of which these are likenesses. Suppose in your mind that some grammarian, having summoned his pupils and made silence, said in a lowered voice "temetum" ("strong drink"), which, being uttered by him, those who sat near enough heard well enough, those farther off too little, and those farthest off were grazed by no utterance at all. But of these, those who sat near did not know what "temetum" was; those who were farther off, by I know not what chance, partly knew what "temetum" was and partly did not; while from those who had not received the master's voice, what "temetum" was lay wholly hidden—all were impeded by obscurity. And here you now discern all those kinds of obscurities. For those who had no doubt about the hearing were suffering that first kind, to which is like the pomegranate that is painted in the light but unknown to those who do not know it. Those who knew the word but had received the utterance too little or not at all with their ears were laboring under that second kind, to which is like the image of a man, but in a place not well lit or wholly dark. But those who lacked not only the utterance but also the signification of the word were wrapped in the blindness of the third kind, which is the foulest of all. And what was said—that there is a certain obscure thing like the ambiguous—can be discerned in those to whom the word was indeed known but who had perceived an utterance neither wholly absent nor wholly certain. Therefore he will avoid all kinds of speaking obscurely who uses both a voice clear enough and not impeded by the mouth, and words that are very well known. See now, in the same example of the grammarian, how differently ambiguity impedes from the way obscurity of a word does. Suppose that those who were present had both received the master's voice well enough by sense, and that he had uttered a word that was known to all—suppose, for instance, that he had said "magnus" ("great") and then fallen silent. Attend to what uncertainties they would suffer on hearing this name. For what if he is about to say "quae pars orationis" ("what part of speech")? What if he is about to ask about meters, "qui sit pes" ("what foot it is")? What if he is going to ask about history, for instance, "magnus Pompeius quot bella gesserit" ("how many wars Pompey the Great waged")? What if, for the sake of commending poems, he is about to say "magnus et paene solus poeta Vergilius" ("the great and almost only poet, Virgil")? What if, to rebuke the negligence of his pupils, he will then burst out into these words, "magnus vos erga studia torpor invasit" ("a great sluggishness toward your studies has come over you")? Do you see how, with the fog of obscurity removed, that which was spoken of above has, as it were, stood out as a meeting of many roads? For this one thing that was said, "magnus," is both a noun and a trochaic foot and Pompey and Virgil and the sluggishness of negligence, and any others, even innumerable ones, not mentioned, which nonetheless can be understood through this utterance of the word.

9. And so it was most rightly said by the dialecticians that every word is ambiguous. Nor let it disturb you that in Cicero Hortensius cavils in this way: "They say that they dare to explain ambiguous things clearly. The same men say that every word is ambiguous. How, then, will they explain ambiguous things by ambiguous things? For this is to bring a light, extinguished, into the darkness." Wittily indeed and cleverly said—but this is what, in the same Cicero, Scaevola says to Antonius: "in short, that you may seem to speak eloquently to the wise and even truly to the foolish." For what else did Hortensius do in that passage but, by sharpness of wit and charm of speech, pour darkness over the unskilled as if with an unmixed and sweet cup? For when it was said that every word is ambiguous, it was said of single words. But ambiguous things are explained by arguing, and no one argues with single words. Therefore no one will explain ambiguous words by ambiguous words. And yet, although every word is ambiguous, no one will explain the ambiguity of words except with words—but words now combined, which will not be ambiguous. For just as, if I were to say "every soldier is two-footed," it would not follow that a cohort thereby consisted of two-footed soldiers, so, when I say that every word is ambiguous, I am not saying it of a sentence, not of an argument, although these are woven from words. Every ambiguous word, therefore, will be explained by an argument that is not ambiguous.
Now let us see the kinds of ambiguities; of which there are two primary ones: one in things that are even spoken, the other that causes doubt only in things that are written. For whether one hears "acies" or reads it, one can be uncertain, unless it is made clear by the sentence, whether a battle-line of soldiers (acies militum), or the edge of a blade (acies ferri), or the sharpness of the eyes (acies oculorum) is meant or written. But if someone finds, for example, "leporem" written, and it does not appear in what sense it is set down, he will surely be in doubt whether the penultimate syllable of this word should be lengthened, from "lepos" ("charm"), or shortened, from "lepus" ("hare")—an ambiguity he plainly would not suffer if he received the accusative case of this noun by the voice of a speaker. But if someone should say that even a speaker could pronounce it badly, then the hearer would be impeded not by ambiguity but by obscurity, yet of that kind which is like the ambiguous, because a word pronounced in bad Latin does not draw the one considering it into diverse notions but drives him toward what is apparent. Since, therefore, those two kinds differ greatly between themselves, the first kind is again divided into two. For whatever is said and can be understood in several ways—those several meanings, namely, can either be contained not only under one term but also under one definition, or they are only held by a common term but are explained by diverse expoundings. Those that one definition can include are named "univocal" (univoca); but those which, under one name, must be defined diversely have the name "equivocal" (aequivoca). First, then, let us consider the univocal, so that, since this kind has now been laid open by a definition, it may be illustrated by examples. When we say "homo" ("man, human being"), we say boy as much as young man, as much as old man, as much fool as wise, as much great as small, as much citizen as foreigner, as much townsman as countryman, as much one who once was as one who now is, as much sitting as standing, as much rich as poor, as much one doing something as one at rest, as much rejoicing as grieving, or neither. But in all these designations there is nothing that, just as it has received the name "man," is not also enclosed by the definition of "man." For the definition of man is "a rational mortal animal." Can anyone, then, say that a rational mortal animal is only a young man and not also a boy or an old man, or only a wise man and not also a fool? On the contrary, both these and the rest that have been enumerated are contained as much by the definition as by the name "man." For whether a boy, or a fool, or a poor man, or even one asleep—if he is not a rational mortal animal, neither is he a man; but he is a man; therefore he must of necessity be contained by the definition too. And about the rest, indeed, there will be no doubt. But about a small boy, or a fool or downright idiot, or one asleep or drunk or raving, it can be doubted how they can be rational animals. It can certainly be defended, but for those hastening to other matters it is long. For the matter at hand this is enough: that this is not a correct and valid definition of man unless both every man is contained by it and nothing besides man. These, then, are the univocals, which are enclosed not only by one name but also by one definition of that same name, although they can also be distinguished from one another by their own proper names and definitions. For the names "boy," "youth," "rich," and "poor," "free," and "slave," and whatever other differences there are, are diverse; therefore they have diverse proper definitions among themselves. But just as they have one common name, "man," so too "a rational mortal animal" is one common definition.

10. Now let us look at the equivocals, in which the tangle of ambiguities grows almost into an endless thicket. I will nonetheless try to distinguish them into definite kinds. But whether my ability matches my effort, you will judge. Of the ambiguities, then, that come from equivocals, there are three primary kinds: one from art, the second from usage, the third from both. By "art" I now mean the names that, in the disciplines of words, are imposed on words. For a "noun" is defined one way among the grammarians, a "dactylic foot" another way, and an "equivocal" another. And yet this one thing I say, "Tullius," is both a noun and a dactylic foot and an equivocal. And so, if someone should demand of me that I define what "Tullius" is, I answer by the explanation of whichever notion he likes. For I can rightly say: "Tullius" is a noun by which is signified a man, a certain supreme orator, who as consul crushed Catiline's conspiracy. Attend subtly: I have defined the noun itself. For if that very Tullius—who, were he living, could be pointed to with a finger—were to be defined for me, I would not say: "Tullius" is a noun by which a man is signified; but I would say: Tullius is a man, and so I would add the rest. Likewise I could answer in this way: "Tullius" is a dactylic foot consisting of these letters—for what need is there now to enumerate those letters? It is permitted also to say this: "Tullius" is a word through which all are equivocal among themselves, together with this very word, the things that were said above, and anything else that can be found. Since, therefore, this one thing I have said, "Tullius," I have been allowed to define so variously according to the terms of the arts, why do we doubt that there is a kind of ambiguity coming from equivocals which may rightly be said to arise from art? For we said that equivocals are those that, just as they cannot be held by one name, so cannot be held by one definition. See now the second kind, which we recalled comes from the usage of speaking. By "usage" I now call that very thing on account of which we come to know words. For who would seek out and collect words for the sake of words? And so now suppose that someone hears in such a way that he knows nothing is being asked about parts of speech or about meters or about any discipline of words. Still he can be impeded, when "Tullius" is said, by the ambiguity of equivocals. For by this name both the man himself who was the supreme orator, and his painted image or statue, and the volume that contains his writings, and anything of his corpse there is in his tomb, can be signified. For with diverse notions we say: "Tullius freed his fatherland from destruction," and "A gilded Tullius stands on the Capitol," and "The whole of Tullius is to be read by you," and "Tullius is buried in this place." For there is one name, but all these things are to be explained by diverse definitions. This, then, is the kind of equivocals in which no ambiguity now arises from the discipline of words but from the very things that are signified. But if either confounds the hearer or reader—whether what is said from art or what is said from the usage of speaking—will it not rightly be reckoned a third kind? Of which an example appears more openly in a sentence, as if someone should say: "many have written in dactylic meter, as is Tullius." For here it is uncertain whether "Tullius" is set down as an example of the dactylic foot or of the dactylic poet—of which the former is taken from art, the latter from the usage of speaking. But it occurs in simple words too, as if a grammarian should utter this word to his listening pupils, as we showed above.
Since, then, these three kinds differ among themselves by manifest considerations, again the first kind is divided into two. For whatever, from the art of words, makes ambiguity can partly serve as an example of itself, and partly cannot. For when I have defined what "nomen" ("noun") signifies, I can set this very thing forth by way of example. For this thing I say, "nomen," is certainly a noun; for by this rule it is inflected through the cases when we say "nomen, nominis, nomini," and so on. Likewise, when I define what "dactylus pes" ("dactylic foot") signifies, this very thing can serve as an example. For when we say "dactylus," we utter one long syllable and then two short ones. But when it is defined what "adverbium" ("adverb") signifies, you cannot give this very thing as an example. For when we say "adverbium," this very utterance is a noun. Thus by one notion "adverb" is certainly an adverb and is not a noun; but by another, "adverb" is not an adverb, because it is a noun. Likewise, when it is defined what "pes creticus" ("cretic foot") signifies, this very thing cannot serve as an example. For this utterance, when we say "creticus," consists of a first long syllable and then two short ones, whereas what it signifies is long syllable and short and long. Thus here too, by one notion "cretic" is nothing other than a cretic and is not a dactyl; but by another, "cretic" is not a cretic, because it is a dactyl.
Likewise the second kind, which was said to pertain to the usage of speaking apart from the disciplines of words, has two forms. For equivocals arise there either from the same origin or from a diverse one. I call "from the same origin" those which, although they are held by one name and not under one definition, nonetheless flow as it were from one source: as is the case that "Tullius" can be understood as both a man and a statue and a volume and a corpse. These indeed cannot be concluded under one definition, but they have one source, namely the true man himself, whose both that statue and those books and that corpse are. But when we say "nepos," from a far diverse origin it signifies a son's son (a grandson) and a spendthrift. Let us, then, hold these distinct, and see into what that kind which I call "from the same origin" is again divided. For it is divided into two, of which one comes about by transfer (translatio), the other by inflection (declinatio). I call it "transfer" when, by a likeness, one name comes to be of many things—as "Tullius" is said both of the man in whom there was great eloquence and of his statue—or when the part is named from the whole, as when his corpse can be called "Tullius"—or the whole from a part, as when we say "tecta" ("roofs") for whole houses—or the species from the genus: for "verba" ("words") are principally all those things by which we speak, but nonetheless "verbs" are properly named which we inflect through moods and through tenses—or the genus from the species: for although "scholastici" ("scholars") are properly and indeed primarily so called who are still in the schools, yet this name has been borrowed for all who live among letters—or the effect from the agent, as "Cicero" is the book of Cicero—or the agent from the effect, as "terror" is what produces terror—or what is contained from the container, as "domus" ("house"): even those who are in the house are so called—or the reverse, as "castanea" ("chestnut"): the tree too is so called—or anything else that can be found which, from the same origin, is named as it were by transferring. You see, as I think, what ambiguity this makes in words. But those things pertaining to the same origin which we said are ambiguous by the condition of inflection are of this sort. Suppose, for example, that someone has said "pluit" ("it rains," or "it has rained") <* * *> and these must certainly be defined diversely. Likewise, one who says "scribere" ("to write")—it is uncertain whether he has pronounced the active infinitive or the passive imperative. "Homo," although it is one name and one utterance, nonetheless becomes one thing from the nominative and another from the vocative, as "doctus" and "docte" ("learned," nominative, and "O learned one," vocative), where the utterance too is diverse. ["Doctius" is one thing when we say "doctius mancipium" ("a more learned slave"), another when we say "doctius illo iste disputavit" ("this man argued more learnedly than that one").] By inflection, then, the ambiguity has arisen. For I now call "inflection" whatever befalls words by bending them, whether through their utterances or through their significations. For "hic doctus" ("this learned man") and "o docte" ("O learned one") is bent even through the utterances, but "hic homo" ("this man") and "o homo" ("O man") only through the significations. But to grant and pursue this kind of ambiguities bit by bit is almost endless. And so it will be enough to have marked the topic itself this far, especially given your ability. See now those that come from a diverse origin. For these too are still divided into two primary forms, of which one is that which comes about by the diversity of languages: as when we say "tu," this one utterance signifies one thing among the Greeks and another among us. This kind had to be noted in any case; for it is not prescribed to each man how many languages he should know or in how many languages he should argue. The other form is that which makes ambiguity within one language indeed, yet by the diverse origin of the things that are signified in one term—such as the case we set down above about "nepos." This again is split into two. For either it occurs under the same kind of part of speech—for "nepos" is a noun as much when it signifies a son's son as when it signifies a spendthrift—or under a diverse one: for not only is it one thing when we say "qui" <* * *>, another, as has been said, "qui scis ergo istuc nisi periculum feceris" ("how then do you know that, unless you have made the trial?"), but also the former is a pronoun, the latter an adverb.
Now from both—that is, from both art and the usage of words, which we had set down as the third kind among the equivocals—there can exist as many forms of ambiguities as we have enumerated in the two preceding kinds.
There remains that ambiguous kind which is found in writings alone, of which there are three species. For such an ambiguity comes about either by the quantity of syllables, or by the accent, or by both: by quantity, as when "venit" is written, it is uncertain as to tense on account of the hidden quantity of the first syllable; by accent, as when "pone" is written, it is uncertain whether from "pono" ("I place"—imperative "pone," "place!") or, as has been said, "pone sequens namque hanc dederat Proserpina legem" ("following behind, for Proserpina had laid down this law"—the adverb "pone," "behind"), on account of the hidden place of the accent; but it comes about from both, as is the case we mentioned above about "lepore." For not only must the penultimate syllable of this word be lengthened but also accented, if it is derived from "lepos" and not from "lepus."