previous next
I take up now — all books being closed — a sentence of very simple structure, of which every word and every construction are familiar, say a certain passage in Livy.1 I tell the story of the context: Two assassins have got admission, on the pretext of a quarrel to be decided, into the presence of Tarquin. One of them diverts the attention of the king by telling his tale, and the other brings down an axe upon the king's head; whereupon they both rush for the door.

In order that the interpretation shall be done absolutely in the order in which a Roman would do it, without looking ahead, I write one word at a time upon the board (as I will again do upon the board before you), and ask questions as I go, as follows : —2

Tarquinium. "What did Livy mean by putting that word at the beginning of the sentence?"

"That the person mentioned in it is at this point of conspicuous importance."

"Where is Tarquinium made?"

"In the accusative singular."

"What does that fact mean to your minds?"

Here most of them are somewhat dazed, not being used to that word meaning, the very word that ought constantly to be used in dealing with syntax, or so-called "parsing." So I very probably have to say, "May it mean the duration of time of the act with which it is connected?"

They say, "No."

I ask, "Why not?"

Somebody says, "Because the name of a person cannot indicate time."

I say, "Give me some words that might indicate time." They give me dies, noctes, aetatem, etc. Then I ask, "May it mean extent of space?"

They say, "No," give me similar reasons for their answer, and, upon my asking for words that might indicate extent of space, they give me, perhaps, mille passuum, tres pedes, etc.

Then I ask, "May it indicate the extent of the action of the verb, the degree to which the action goes?" They say, "No," for a similar reason. But when I ask for words that might mean the degree of the action, they commonly cannot tell me, for the reason that, strange to say, the grammars do not recognize such a usage; though sentences like "he walks a great deal every day" (multum cottidie ambulat) are even more common than sentences like "he walks three miles every day" (cottidie tria milia passuum ambulat), and the accusatives mean essentially the same thing in both sentences.

Then I ask, "May it mean that in respect to which something is said, — as regards Tarquin, — the accusative of specification?" To a question like that, I am sorry to say that a great many always answer yes, for students get very vague notions of the real uses of the accusative of specification. Somebody, however, may be able to tell me that the name of a person is never used in the accusative of specification, and that in general the use of the accusative of specification, in the days of Cicero and Virgil, was mostly confined to poetry. "What words were used in the accusative of specification in prose?" Here I never get an answer, although the list is determinate, short, and important. So I have to say, "I must add to your working knowledge a useful item; write in your note-books as follows: partem, vicem, genus with omne or a pronoun (quod, hoc, id), sexus with virile or muliebre, hoc and id with aetatis, the relative quod and the interrogative quid, are used in Latin prose in all periods as accusatives of specification. Here, then, is a bit of definite information which may enable you, when you first meet one of these words again (you will do so quite early in your first book of Livy), to walk without stumbling through a sentence where you would otherwise trip." Then I go back to Tarquinium. "May it be," I ask, "an accusative of exclamation?"

They say, "Possibly so."

I say, "possibly yes, though in historical narration you would hardly expect such an exclamation from the historian." Next I ask, "May it be a cognate accusative?"

To that they answer, "No"; telling me, perhaps with some help, that "the name of a person cannot be in any sense a restatement of an act, — cannot mean an activity."

"Well, then, what does this accusative case mean?"

By this time a good many are ready to say: "Object of a verb, or in apposition with the object." But I ask if one thing more is possible, and some one says: "Subject of an infinitive."

"Yes," I answer; "and one thing more yet?"

"Predicate of an infinitive," someone suggests.

"Now," I ask, "what have we learned from all this? Given the name of a person or persons in the accusative with no preposition, how many and what constructions are possible?" All are ready now to answer, Object of a verb, or subject or predicate of an infinitive. "Good," I say. "Keep those possibilities always fresh in your mind, letting them flash through it the moment you see such a word; and that having been done, WAIT, and NEVER DECIDE which of these possible meanings was in the mind of the Roman speaker or writer until the rest of the sentence has made the answer to that question perfectly clear. Now tell me what constructions are possible for an accusative like hiemem."

They answer, "duration of time, apposition, object of verb, subject or predicate of an infinitive."

"For an accusative like pedes?"

They answer, "extent of space, apposition, object of verb, or subject or predicate of an infinitive."

"For an accusative like multum?"

"Extent of action, apposition, object of verb, or subject or predicate of an infinitive."

"For an accusative like vitam?"

"Cognate accusative, apposition, object of verb, or subject or predicate of an infinitive."

Now I ask, "Can any one tell me what constructions we may expect if the verb turns out to be some word like doceo or celo?" They all give the answer, and therewith I have already passed in rapid review practically the whole matter of the accusative constructions; and, what is more, — and this is vital, — I have done it from a very practical standpoint. I have not asked a student to "parse" a word after seeing its full connection in the sentence (an exercise which loses four-fifths of its virtue by this misplacement), but I have demanded anticipatory parsing, — I have put my questions in such a way that my students have learned for all accusatives what instantaneous suggestions of the possible parts a word is playing in the sentence they may get, at first sight of the word, from the very nature of the word.

Then I pass on. "We have our King Tarquin before our eyes, as the person on whom the interest of the sentence centres, and we know that he is the object of an action, or the subject or predicate of an infinitive action; or, possibly, in apposition with such an object, subject, or predicate. To proceed, the next word, moribundum, is what and where made?"

"Adjective, nom. sing. neut., or acc. sing. masc. or neut."

Don't smile at all this. The habit of getting a young student to think all these things out, even where he could not go astray if they were not asked of him, saves many a getting lost in difficult places. "What is probable about moribundum, as we have it in this particular sentence?"

"That it belongs to Tarquinium."

"Right. Now keep that picture in mind: Tarquinium moribundum, the King, breathing his last, acted upon or acting. Now for the next word: Tarquinium moribundum cum. What is cum?" Some say, with perfect readiness, "preposition," some say "conjunction."3

"But," I answer, "if you are used to the right spelling, you know with an instant's thought that no Roman that ever lived could tell at this point whether it was preposition or conjunction. In order to tell, you must wait for — what?"

"Ablative or verb," they answer.

Then we go on, "Tarquinium moribundum cum qui. What does qui at once tell us about cum?"


"Right. What do we know now, with almost absolute certainty, about Tarquinium? What part of the sentence does it belong to?"

Here, I grieve to say, a chorus of voices always answers, "Main verb"; for, in some mysterious way, students arrive at the universities without having learned that the Romans delighted to take out the most important word, or combination of words, from a subordinate introductory sentence, and put it at the very start, before the connective, — a bit of information worth a great deal for practical reading. That habit of expression I now tell them, and then ask, "Given a sentence beginning with mors si, what do you know?"

"That mors is the subject or predicate of the verb introduced by si."

"Given a sentence introduced by Hannibali victori cum ceteri?"

"That Hannibali depends on something in the cum-sentence."

Now we go back to our sentence, and the word qui. "What part of speech is it?"

"Relative," they say.

"Or what else?" I ask.


"Where is it made?"

"Nom. sing. or plur., masc."

"If it is a relative, where in the sentence as a whole does its antecedent lie?" They should answer, "Inside the cum-clause." The cum serves as the first of two brackets to include the qui-clause.

"If, on the other hand, it is an interrogative, what kind of a question is alone here possible?"

"Indirect, and in the subjunctive," they answer.

"In that case, what kind of a meaning, speaking generally, must the verb introduced by cum have?"

"It must be able to imply asking of some kind."

"Rightly said; perhaps we may have such a sentence as, “"When everybody inquired who these men were"” — Cum qui essent omnes quaererent; or perhaps we shall find that qui is relative. The next word is circa, — Tarquinium moribundum cum qui circa. What part of speech is it?"


"What then may it do?"

"It may modify a verb, an adjective, or another adverb."

We proceed: Tarquinium moribundum cum qui circa erunt.

"What, now, about circa?"

"It modifies erant."

"What was the number of qui?"


"Was it relative or interrogative?"


"How do you know?"

"Because erant is not subjunctive."

"Right. Now qui circa erant is as good as a noun or a pronoun, — an indeclinable noun or pronoun, in the plural. Think of it in that way, as we go on. Tarquinium moribundum cum qui circa erant excepissent. I don't ask to-day the meaning of the mode of excepissent, because the world is in so much doubt about the question of the history and force of the cum-constructions. But what was Livy's meaning in writing the accusative Tarquinium?"

"Object of excepissent."

"Yes, and what was the subject of excepissent?"

"The antecedent of qui."

"Yes; or, looking at the matter more generally, the subject was qui circa erant."

"Before going on, what picture have we before us? What has the sentence thus far said? This:

See Tarquin, dying! See the bystanders! See them pick him up!

Our curiosity is stimulated by the very order. The next word is illos, — “Tarquinium moribundum cum qui circa erant excepissent” ... What does the position of illos, first in the main sentence proper, tell us?"

"That the people meant by it are of special prominence at this point."

"Who do you suppose these illos are, these more distant persons, thus set in emphatic balance against Tarquinium, each leading its clause?

"The assassins," the whole class say.

"What do we know about Livy's meaning from the case?"

Now they all answer in fine chorus and completeness, "Apposition, object of main verb, or subject or predicate of an infinitive."

We proceed: “Tarquinium moribundum cum qui circa erant excepissent, illos fugientes” ... "What part of speech is fugientes?"


"Which one?"

"Present active."

"Then you see a running-away going on before your eyes. What gender?"

"Masc. or fem."

"What number?"


"Then you see some two or more men or women running away. What case?"

"Nom. or acc."

"On the whole, do you feel sure you know the case?"

"Yes; accusative."

"Belonging to what?"



"Because of course the assassins, the illos, would run away."

"Yes," I say; "but it cannot possibly mislead you to wait until there isn't a shadow of a doubt. We will go on: “Tarquinium moribundum cum qui circa erant excepissent, illos fugientes lictores” ... Here you have another set of people, the king's body-guard. In what case?"

"Nom. or acc. plural."

"Which?" They do not know. "Well, then, can illos agree with lictores, if you consider forms alone?"


"In that case, fugientes would have to go with illos lictores, wouldn't it?"


"But would the lictors run away?"


"Would the assassins?"


"Certainly. Then fugientes does not belong with lictores, and does belong with illos; and illos seems to be, just as we suspected at first sight of it, the assassins. However, we must ask ourselves one more question, Is apposition possible between illos and lictores?"

"No; for they are very different people."

"Is any relation of a predicate possible between them? Can the one be the predicate of an infinitive of which the other is the subject?"

"No; because, as before, they are very different people."

"Still it is possible that lictores is accusative. If it is, it may be object, in which case illos is necessarily subject, for, as we have seen, they cannot be in apposition; or, it may be subject, in which case, for the same reason, illos must be object. In either case, they must be in direct opposition to each other, one of them (we don't yet know which) being subject, the other, object; while, if lictores is nom., you still have the same relation, only you know which is subject and which is object. In any event, you see they are set over against each other, together making subject and object. Now keep the results of this reasoning ready for the countless cases in which such combinations occur. Given two nouns like bellum Saguntum: what are the constructions?"

"One is the subject of a verb, and the other the object, and we can't yet tell which."

"Right. Now I will give you a still more involved combination, but of a very commonly occurring kind, — quae nos materiem. What do you make out of that?"

Some clever boy will say, "Nos must be the subject of a verb, either finite or infinitive, and quae and materiem are object and predicate-object."

"Good. Then what kind of meaning does the verb probably have?"

"One of calling."

"Right. The words are from Lucretius, and the verb he used was vocamus. Treasure up that combination and the meaning of it."

"Now we go back to the assassins who are running away, and the king's body-guard. I will inform you that there is just one more word in the sentence. What part of speech is it?"


"Active or passive?"


"Right. What does it tell?"

"Tells what the lictors do to the assassins."

"What mode, then?"


"What two tenses are possible?"

"The perfect and the historical present."

"Right. Now the situation is a pretty dramatic one. Which of these two tenses should you accordingly choose, if you were writing the story?"

"The present."

"So did Livy. Now tell me what you think the verb is."

"Interficiunt," somebody says. "Capiunt," says another, hitting the idea but not the right word, which is comprehendunt, get hold of them well, — "nab 'em"; or, as our tamer English phrase might put it, "secure them."

"Now let us render into English the sentence as a whole, translating not merely Livy's words, but the actual development of the thought in his mind.

Tarquinium, there's Tarquin; moribundum, he's a dying man; cum qui circa erant, you see the bystanders about to do something; excepissent, they have caught and supported the king; illos, you turn and look at the assassins; fugientes, they are off on the run; lictores, there are the king's body-guard; we hold our breath in suspense; — comprehendunt, THEY'VE GOT 'EM!

So, then, that Latin order, which looks so perverted to one who is trained to pick the sentence to pieces and then patch it together again, gives us the very succession in which one would see the actual events; weaves all the occurrences together into a compact whole, yet keeping everywhere the natural order; while any order that we may be able to invent for a corresponding single sentence in English will twist and warp the natural order into a shape that would greatly astonish a Roman."

"Finally, with the understanding and sense of the dramatic in the situation, which we have got by working the sentence out as Livy wrote it, compare the perversion of it, which we get by working it out correctly on the first-find-your-subject-of-the-main-sentence-and-then-your-predicate, etc., method:

the lictors secure the assassins as they run away, when those who were standing by had caught and supported the dying Tarquin.

The facts are all there, but the style, the soul, is gone."

1 1.41, Tarquinium moribundum cum qui circa erant excepissent, illos fugientes lictores comprehendunt.

2 The sentence grows upon the board by the addition of one word after another.  To obtain the same result in print, with each new word the whole of the sentence thus far given will be repeated.  And, for the sake of greater clearness, answers will be distinguished from questions by the use of italics. [Nowadays, some teachers find it convenient to write the sentence on a transparency and uncover one word at a time.]

3 The fact that it is possible for students, without a moment's reflection, to plunge at things in this sadly well-known way shows how thoroughly ineffective the prevailing method of teaching beginners is in developing a sharp and self-suspicious observation.  That charge, it will be seen, cannot be brought against the method advocated in this paper.

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 United States License.

An XML version of this text is available for download, with the additional restriction that you offer Perseus any modifications you make. Perseus provides credit for all accepted changes, storing new additions in a versioning system.

hide References (1 total)
  • Cross-references in notes from this page (1):
    • Livy, The History of Rome, Book 1, 41
hide Display Preferences
Greek Display:
Arabic Display:
View by Default:
Browse Bar: