CHAPTER 4

Direct Questions, Irregular Verbs, Compound Verbs

The singular imperative form fer is paralleled by irregularities in the singular imperatives of dico, duco and facio: dic, duc, fac (which survives in the English words “facsimile” [= “fax”] and “factotum”, literally “make a similar one” and “do everything”. Unlike ferte, the corresponding plural imperatives of those other verbs are regular: dicite, ducite, facite. dice, duce and face, but not fere, are all attested in early Latin; it is not clear why the irregularity arose.

Latin, like English and the Romance languages, but unlike classical Greek or German, is not, in general, a free-compounding language, employing nouns and adjectives as well as prepositions and verbs to form complex words. (A remarkable exception is suovetaurilia, the sacrifice of a pig [sus], a sheep [ovis] and a bull [taurus].) The poet Horace was scornful and suspicious about sesquipedalia verba “words a foot and a half long”. The longest word in Latin till well after the classical period is the comic coinage subductisupercilicarptor “a person who criticizes, drawing his eyebrows from below”, a mere 24 letters. (In his Rudimenta Grammatices, the most widely used Latin grammar in Italy and many other parts of Europe at the end of the Quattrocento, Niccolò Perotti parades Dioclitianopolitanissimorum [27 letters], a rather improbable superlative form of “the citizens of Dioclitianopolis”, while both Dante and Shakespeare cite forms of the honorificabilitudinitas, which had appeared in grammar books as early as the Carolingian period. In its longest form, the dative/ablative plural, this word can be anagrammatized into hi ludi F. Baconis nati tuiti orbi “these plays, F. Bacon's offspring, are preserved for the world”, a rather clumsy phrase lending putative support to the theory that Francis Bacon wrote Shakespeare's plays. [The word also has the curious distinction of being composed of alternating consonants and vowels, none of which is an e, even though that is the most frequent letter in Latin.] Dante also offers as an Italian word sovramagnificentissimamente [27 letters].

German compounds such as

Saucissenkartoffelbreisauerkrautkranzwürste

or

Oberammergaueralpenkräuterdelikatessenfrühstückskäse

sound rather indigestible.

The longest word in Greek, which has much greater facility than Latin for compounding words, is an Aristophanic concoction, describing an extravagant gallimaufry of various types of food in 171 letters:

The 4th century BC comedian Anaxandrides lists 100 delicacies in an impressively remorseless catalog quoted at Athenaeus Wise Men at Dinner 131d. (It is often hard to detect an intellectual stratum in Greek Comedy.) An anonymous six-line epigram quoted by Athenaeus at 162a, consists of a mere 14 words, all except two (both καί [kai (‘and’)] being derogatory terms for philosophers; the first means ‘sons of drawers up of eyebrows’ and is only 15 letters long, not a match for subductisupercilicarptor, but the length of the 12 compounds ranges from 14 to 20 letters, a sustained feat that Latin never comes close to equalling.

The prize in English probably goes to James Joyce for his 100-letter “thunderwords” in Finnegans Wake, words such as

(with the not very apparent meaning of “cough”), but the longest admitted to the second edition of the OED is pneumonoultramicro-scopicsilicovolcanoconiosis, “a factitious word alleged to mean ‘a lung disease caused by the inhalation of very fine silica dust’ but occurring chiefly as an instance of a very long word”.

Not only is Latin not a free-compounding language, but the Romans’ attitude towards the creation of new words was very conservative:

Although Latin is not a free-compounding language, it managed to create compound words for the droppings of mice, pigs and sheep (respectively muscerdae, sucerdae, and ovicerdae [this last is of slightly doubtful manuscript authority]), and perhaps other specific types that happen not to have been transmitted.

There are no generally applicable principles for the changes made in assimilation; ab, ob and sub, for example, produce three quite different variations in abstuli, obtuli and sustuli. In making this same point, that prepositions may change when forming compound words, Quintilian cites abstulit, aufugit, amisit, all using the one preposition, ab (Education of the Orator 1.6).

Sometimes assimilation was not applied at all; whereas, for example, cum is always altered in compounds, to co- or cog- or col- or com- or con- or cor-, and the same writer might alternate inconsistently between, for example, affero and the unassimilated adfero. (It would usually be more accurate to say that the medieval scribes to whom we largely owe our texts of classical writers concerned themselves very little about these matters, and varied the forms with minimal thought for consistency.) Since the assimilated forms require more attention, they will be used in this course.

Here is a list of prefixes used in compounding verbs, but not found separately. Ancient grammarians refer to such prefixes by the rather charming term loquellares, “little bits of speech”:

perire is very rarely used in its literal sense. Why “go through” should mean “die” is unclear; it presumably alludes to the notion of passing from life to death. Similarly, the standard word for “kill” is interficere, a compound of inter “between” and facere, but the implication of the preposition is again unclear.

interficere is not used in the Romance languages. Some of the standard Romance words for “kill” are also obscure. The Span./Port. matar [cf. matador] may be derived from the Latin mactare “slaughter”, but it has been suggested that it is related to the Persian shah mat “the king is dead” [i.e. checkmate], reflecting the long occupation of the Iberian peninsula by the Arabs. The French tuer seems to be a grim euphemism, derived from tutari [for this form of the present infinitive, see Chapter 15] “take care of”.

Since the Italian uccidere comes from occidere (a compound of ob and cadere), that term also must have been current in spoken Latin. St. Augustine (Heptateuch Questions 7.56) interestingly notes that Christians preferred the gentler terms occurrere “run into” and compendiare “abridge” (with which he compares the military slang allevare, literally “lighten”).

Good English style resists the placing of prepositions at the end of a clause (as in e.g. “This is the sort of word-order I will not put up with”, to say nothing of “It was snowing hard, but not hard enough to come back in out of from”). Latin does not separate prepositions from verbs in this way.

Latin does not generally favor the compounding of a verb with more than one prefix: “The wolf goes from the house to the cave” is best translated as lupus ad speluncam a casa abit or adit. (abadit would look bizarre to a Roman.) Other than with re-, which significantly is one of the few verbal prefixes which are not prepositions (see above), there are only occasional exceptions.

For example, recolligo (re + cum + lego) “gather together again”, recompono (re + cum + pono) “place together again” [“decompose” does not go back to classical Latin], reconduco (re + cum + duco) “lead together again”, repercutio (re + per +cutio), literally “strike back thoroughly”.

Such double compounds occur mostly in poetry, and many are very striking. Vergil twice uses transadigere, compounded of trans “across” + ad “to” + ago, agere, egi, actum 3 “drive”. At Aeneid 12.270ff., hasta .../... iuvenem .../transadigit costas (“the spear pierced the young man in the ribs”), the weapon (sc. hasta “spear”) is the subject and the prepositions govern two accusatives, iuvenem and costas. At lines 505ff. of the same book, Aeneas ...//transadigit costas ... ensem “Aeneas drove his sword (at and) through his ribs”, the construction and sense are different, the weapon (sc. ensem “sword”) being the direct object of the verb, leaving only costas to be governed by the prepositions.

Only direct questions (e.g. “Do you like pigs?”, “Where are the pigs?”) will be considered in this chapter. Indirect questions (e.g. “The farmer asked the pirate whether he liked pigs”, “I know where the pigs are”) will be introduced in Chapter 25.

The Romans themselves did not have a universally accepted system of punctuation in the classical period. The question-mark is thought to be derived from qo, an abbreviation for quaestio, “question”, written occasionally by scribes at the end of interrogative sentences. The decision to punctuate as a question or not is one which a modern editor of a classical text routinely makes with little or no guidance from the manuscripts. The problem is particularly acute in the dialogues of Roman drama. It would make sense to inform readers at the start of a sentence that what they are reading is not a statement, but a question, would it not? ¿Why is Spanish the only language in the world to have introduced the logical refinement of indicating questions and exclamations at the beginning of such sentences?

It may seem a bizarre fact that Latin has no simple equivalent to “Yes”. A periphrasis, such as ita vero “so indeed” or minime “(not) in the least”, may be used, or the verb may be repeated from the question, as e.g. audisne, agricola, piratam? audio. Likewise, there is no word for “No”. A negative response to audisne, agricola, piratam? might be non audio. Much the same principles of affirmation and negation are used in many other languages; e.g. classical Greek, Chinese, Irish and other Celtic languages.

The phonetic variations between the two main branches of medieval French were characterized by their different ways of saying “yes”: the langue d’oïl (derived from the Latin hoc ille, literally “that man (does/says) this (thing)”, and the origin of the modern oui) was spoken north of the Loire, and developed into standard modern French; the langue d’oc (simply hoc “this”) developed into Occitan, still spoken by more than two million people in the Languedoc and other parts of southern France.

The use of si in Italian, la lingua del si, and other Romance languages to mean “yes” is derived from sic “thus”. French uses si specifically to contradict a negative statement, as in “Vergil is not a good poet. Yes, he is”. German uses doch for this purpose, but modern English lacks such a refinement: in Old and Middle English, simple affirmation was expressed by “yea”, whereas “yes” was the equivalent to French si and German doch.

The standard French negation ne ... pas, as in Je ne sais pas “I do not know”, is derived, by fine semantic bleaching, from non “not” and passus, a 4th decl. Latin noun meaning “step”, i.e. as if it were “I have made no progress towards knowing”.

Although nonne is by far the most regular way to mark a question expecting a positive response, some writers preferred to introduce such questions as if they were open questions (i.e. with or without the marker -ne), and then insert non later in the sentence: agricola(ne) nautam non videt?

It is also possible to use markers other than nonne or num in asking leading questions; for example, when he wished to save a defendant from the ghastly punishment for parricide, Augustus is said to have asked him certe patrem tuum non occidisti? “Of course, you didn’t kill your father?” (Suetonius, Life of Augustus 33).

“The traditional punishment for parricide is as follows: the condemned person is beaten with blood-colored sticks, then sewn up in a sack with a dog, a rooster, a viper and a monkey, and thrown into the deep sea, if the sea is nearby; otherwise, in accordance with the law passed by the deified Hadrian, he is thrown to wild beasts” (Justinian’s Digest 48.9.9). Monkeys are not indigenous to Italy; their use in punishing parricides is not mentioned till the end of the 1st century AD. In 100 BC, Publicius Malleus, who had murdered his mother, was the first to be sewn in a sack and thrown into the sea (Livy, History of Rome Summary of Book 68). Seneca claims that this punishment was inflicted more frequently by Claudius in a five-year period than throughout all earlier time (On Mercy 1.23).

The double question “Do you love the sailor or the farmer?” can be asked in four different ways:

Such sentences are formulated, however, much more frequently with an than with anne.

For questions introduced by interrogative pronouns and pronominal adjectives (i.e. “Who … ?”, “Which sailor … ?” etc.), see Chapter 18.

Latin has very few irregular verbs. Apart from sum, possum, eo, and fero, the only other ones you will meet in this course are the closely related volo, nolo, and malo (all in Chapter 10), and fio (Chapter 15). One or more of the tenses of these verbs, usually only the present, must be learned individually.

These verbs are conventionally called “irregular”, even though it might be more accurate to describe them as “unpredictable”. The form sum, for example, although it seems so remote from equivalent verb-forms in -o, is quite explicable as a development of a regular IndoEuropean form. It is, however, neither helpful nor reassuring to the beginner to know that sum, es and est follow the same “regular” IndoEuropean system as do am, are and is; we have to regard them as irregular forms. As an example of what might be called the “hidden regularity” of esse and posse, the use of -se rather than -re in the present infinitive active appears also in other “irregular” verbs: ferre (= ferse), velle and its compounds nolle and malle (= velse, nolse, malse).


Spanish has two verbs meaning “to be”, ser (derived from esse) and estar (derived from stare “stand”), which distinguish between permanent and temporary conditions: soy espagñol “I am Spanish”, but estoy cansado “I am tired”; Latin does not make such a distinction. In the sentence Acquí está la llave (“Here is the key”), está (“is”) is a form of estar, and is used in denoting the location of the key. In the sentence ’A key’ es una llave (“’A key’ is una llave”), es (“is”) is a form of the verb ser, derived from the Latin esse (“be”), and states that llave is now and always is the word for “key”. Latin has only the one verb, esse, for both such functions.

A cap(p)ella singing has nothing to do with goats (capellae). It is music “in the manner of the chapel” [an abbreviation of Ital. alla cap(p)ella]. On the other hand, one the most favored suggestions for the origin of the term “tragedy” is τράγου ᾠδή (tragou ode “goat-song”), a goat (τράγος [tragos) being given as a prize for the best song (ᾠδή [ode]) in honor of Dionysus, the god of drama; goats were an appropriate sacrifice to Dionysus because they damaged his sacred vines.

The constellation and astrological sign Capricorn is often associated with Amalthea, the she-goat which suckled the infant Jupiter on Crete; according to that version of the legend, Capellaecornu would be a more accurate name. (For caper, capri masc. 2 “he-goat”, see Chapter 5). The Romans often seem, however, to take no account of grammatical gender: despite the obvious fact that only the cock-bird has a magnificent tail, at Amores 2.6.55 explicat ipsa suas ales Iunonia pinnas “The bird of Juno [i.e. the peacock] spreads out its wings of its own accord”, Ovid uses the third declension feminine noun ales, alitis “bird”, its gender being reinforced by the feminine form of the adjective Iunonia.