Accidence and syntax

Nouns: gender, number, case and the five declensions#

StandardOCR H443 Appendix 5d Accidence - nouns of all standard types, together with bos, domus, Iuppiter and visOCR H443 Appendix 5d Syntax - standard patterns of case usageDfE GCE AS and A level subject content for ancient languages

The five declensions

Fig. 1The genitive singular is the key that identifies the declension, which is why it is the emphasised cell. Read down each column once, then read across each row: the horizontal patterns - identical dative and ablative plural, identical nominative and accusative in every neuter - are what actually save time in an unseen.

A Latin noun carries three pieces of information at once, and reading Latin is largely the habit of extracting all three from a single ending. Gender is a property of the word rather than of the thing it denotes: it is learned with the word and it never changes. Number is singular or plural. Case is the noun's grammatical job in its own clause - subject, direct object, indirect object, possessor, instrument - and because Latin marks that job on the word itself rather than by position, the word order of a Latin sentence is free to do other work: emphasis, contrast, suspense, the sound of the line. That freedom is exactly what makes Latin hard to read quickly and rewarding to read well. The candidate who translates by guessing from word order is reading English with Latin words in it; the candidate who reads the endings is reading Latin.

There are five declensions, and the genitive singular identifies which one you are in - which is why dictionaries and vocabulary lists print it. Genitive in -ae is first declension (puella, puellae, almost all feminine); in -ī, second (dominus, dominī, mostly masculine; bellum, bellī, neuter); in -is, third (rēx, rēgis - the largest and most varied declension, with all three genders); in -ūs, fourth (manus, manūs, mostly feminine in practice for the common nouns, though the declension is predominantly masculine); in -eī or -ēī, fifth (rēs, reī, feminine, apart from diēs, which is usually masculine). Learn the noun as a pair, not as a single form: writing corpus in a vocabulary list tells you nothing, whereas corpus, corporis, n. tells you that the stem is corpor-, that the declension is the third, and that the neuter rule will apply.

Three rules cut across the whole system and are worth stating separately. First, in every neuter noun of every declension the nominative and the accusative are identical, in both numbers, and the nominative and accusative plural always end in short -a: bellum, bella; corpus, corpora; mare, maria. That single rule resolves a large proportion of the ambiguities in an unseen. Second, the dative and ablative plural are always identical within a noun: puellīs, dominīs, rēgibus, manibus, rēbus - so context, not form, decides which one you are looking at. Third, the third declension splits into consonant stems and i-stems, and the difference shows up in exactly three places: genitive plural -ium rather than -um (urbium, not urbum), neuter nominative and accusative plural -ia rather than -a (maria), and neuter ablative singular -ī rather than -e (marī). The specification's literature papers add a printing convention here that catches candidates out: an accusative plural in -is is printed as -es, so a form you might have expected as urbīs appears on the paper as urbēs.

The specification singles out four nouns by name - bōs, domus, Iuppiter and vīs - because each is irregular enough to defeat the ordinary rules. Bōs, bovis (ox, cow) has the unexpected stem bov- and a genitive plural boum. Domus (house, home) declines partly as fourth and partly as second declension and, more importantly, behaves like a place name: domī means at home (a locative), domum means homewards, domō means from home, all without a preposition. Iuppiter has the oblique stem Iov-, so Iovis, Iovī, Iovem, Iove. Vīs means force in the singular (vīs, vim, vī, with almost no genitive or dative in use) and strength or troops in the plural, where it declines as an i-stem third-declension noun: vīrēs, vīrium, vīribus. None of these is difficult, but each is a trap in an unseen precisely because it looks like something else.

Finally, know what the cases actually do, because the tree of case usage matters more in translation than the ability to recite endings. The nominative marks the subject and also the complement after sum, fīō and the passive of verbs of naming and making - rēx factus est, he was made king, with a nominative after the passive verb. The accusative marks the direct object, the goal of motion with ad and in, and extent of time or space (multōs annōs, for many years; decem pedēs, ten feet). The genitive marks possession, description, part of a whole (pars mīlitum, part of the soldiers) and the object of certain verbs of remembering and forgetting. The dative marks the indirect object, the person interested or affected, possession with sum (mihi est liber, I have a book) and the agent with a gerundive. The ablative is the workhorse: means, manner, accompaniment, cause, comparison, time when, separation, description, and the absolute construction. Vocative, for direct address, differs from the nominative only in the singular of second-declension nouns in -us.

Worked example

Reading a sentence by its endings alone

Constructed illustrative sentence (not a quotation): eō tempore rēgis fīlia mīlitibus urbis magnam spem salūtis dedit. Identify the declension, case, number and function of every noun and adjective, then translate.

Find the verb and its person
dedit is third person singular perfect active of dō, dare, dedī, datum. The subject is therefore singular, and dō characteristically takes an accusative of the thing given and a dative of the person given to - so before reading a single noun you already know the shape of the sentence you are looking for.
eō tempore
tempus, temporis, n. is third declension; tempore is ablative singular. eō is the ablative singular neuter of is, ea, id, agreeing with it. An ablative of time when, without a preposition: at that time. Time when is one of the ablative uses that never takes in.
rēgis fīlia
fīlia is first declension, nominative singular, feminine - the subject. rēgis is third declension, genitive singular of rēx, rēgis, m.: a genitive of possession dependent on fīlia. The king's daughter. Note that the genitive stands before its noun here; it could equally have stood after it, and nothing in the meaning would change.
mīlitibus
mīles, mīlitis, m., third declension. The ending -ibus is dative or ablative plural and the form alone cannot decide between them. The verb decides: dō governs a dative of the person, so this is dative plural, the indirect object. To the soldiers.
urbis
urbs, urbis, f. is an i-stem third-declension noun; urbis here is genitive singular, dependent on mīlitibus - the soldiers of the city. It would be an easy misreading to take urbis as a nominative or accusative plural, but a plural in that declension would be urbēs, and in any case the sentence already has its subject.
magnam spem salūtis
spēs, speī, f. is fifth declension; spem is accusative singular, the direct object. magnam is the accusative singular feminine of magnus, -a, -um, agreeing with spem in case, number and gender. salūtis is genitive singular of salūs, salūtis, f., an objective genitive: the hope is directed at safety, so hope of safety, not safety's hope.
Render it as English
Assemble the pieces in the order English needs, not the order Latin used: subject, verb, indirect object, direct object. Nothing in the Latin word order is lost, because the emphasis Latin achieved by putting eō tempore first can be reproduced in English simply by keeping it first.

Result: At that time the king's daughter gave the soldiers of the city great hope of safety. Nouns: tempore (3rd, abl. sg., time when); rēgis (3rd, gen. sg., possession); fīlia (1st, nom. sg., subject); mīlitibus (3rd, dat. pl., indirect object); urbis (3rd i-stem, gen. sg., possession); spem (5th, acc. sg., direct object); salūtis (3rd, gen. sg., objective genitive). Adjectives: eō (abl. sg. n., agreeing with tempore); magnam (acc. sg. f., agreeing with spem).

Typical mistakes

Treating -a as a reliable feminine nominative: it is equally the neuter plural (bella, corpora) and the first-declension ablative singular (puellā).
Assuming the noun nearest the verb is its subject; in Latin the subject may be the last word of the sentence, or be carried only by the verb ending.

Active revision

Take twenty lines of continuous Latin prose from any A-Level-level author. For every noun, write down the declension, the case, the number and the single word in the sentence that governs the case. Where a form is genuinely ambiguous - a first-declension -a, a third-declension -is, a neuter -a - write both possibilities, then state which one the sentence forces and why. You should end with a list of governing words, not a list of endings.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level in Latin (H443) specification (Cambridge OCR) · GCE AS and A level subject content for ancient languages (Department for Education)

Adjectives, pronouns, comparison, numerals and prepositions#

StandardOCR H443 Appendix 5d Accidence - adjectives of all standard types from all three declensions, including dives, pauper and vetusOCR H443 Appendix 5d Accidence - adverbs; comparison of adjectives and adverbs; pronouns and pronominal adjectives; cardinal numbers 1-1000 and ordinals 1st-10th; uses of prepositionsDfE GCE AS and A level subject content for ancient languages

Adjective endings: the 2-1-2 type and the third-declension type

Fig. 2Two patterns cover almost every adjective in the language. The emphasised cell is the ablative singular of the third-declension type: -ī, not -e, is the mark of an i-stem, and it is the form that distinguishes fortis from the consonant-stem exceptions dives, pauper and vetus.

A Latin adjective agrees with its noun in case, number and gender - three agreements, all of which must be right, and none of which requires the adjective to stand anywhere near the noun. This is the single most useful fact for reading verse, where an adjective can be separated from its noun by half a line, and the agreement is the only thread that connects them. Adjectives come in two families. The first, and much the largest, is the 2-1-2 type: bonus, bona, bonum, using second-declension endings for the masculine and neuter and first-declension endings for the feminine. The second is the third-declension type, which has three sub-patterns distinguished only by how many distinct forms they show in the nominative singular: three terminations (ācer, ācris, ācre), two (fortis, forte), or one (ingēns, gen. ingentis; audāx, gen. audācis). Whichever sub-pattern it belongs to, a third-declension adjective is an i-stem: ablative singular in -ī, genitive plural in -ium, neuter plural in -ia.

Three third-declension adjectives are named in the specification because they refuse to be i-stems: dīves (gen. dīvitis), pauper (gen. pauperis) and vetus (gen. veteris) all take the consonant-stem endings, so ablative singular dīvite, paupere, vetere and genitive plural dīvitum, pauperum, veterum. They are single-termination adjectives, so the nominative singular is the same in all three genders, and the neuter plural is -a, not -ia. Learn the genitive singular of every third-declension adjective as you learn the adjective itself, exactly as you do with a noun: it is the only form that reveals the stem, and vetus, veteris is no more predictable from vetus than corpus, corporis is from corpus.

Comparison is regular and productive. From the stem of the adjective, the comparative adds -ior (masculine and feminine) and -ius (neuter) and then declines as a consonant-stem third-declension adjective: fortior, fortius, gen. fortiōris, abl. fortiōre. The superlative adds -issimus, -a, -um and declines as a 2-1-2 adjective: fortissimus. Adjectives in -er form the superlative in -errimus (ācer, ācerrimus; miser, miserrimus), and six adjectives in -ilis form it in -illimus (facilis, facillimus; difficilis, difficillimus; similis, simillimus, and their relatives). Adverbs behave in parallel: the positive adverb from a 2-1-2 adjective ends in -ē (lātē, from lātus) and from a third-declension adjective in -iter (fortiter); the comparative adverb is simply the neuter singular of the comparative adjective (fortius, lātius); the superlative adverb ends in -ē (fortissimē, lātissimē). The irregular sets must be memorised outright: bonus, melior, optimus (bene, melius, optimē); malus, peior, pessimus (male, peius, pessimē); magnus, maior, maximus; parvus, minor, minimus; multus, plūs, plūrimus. Two constructions attach to the comparative: quam plus a matching case (fortior quam frāter, braver than his brother), or the bare ablative of comparison (fortior frātre), which the specification lists explicitly - and which is much the commoner of the two in verse.

Pronouns and pronominal adjectives are a small, closed, high-frequency set, and they repay outright memorisation because they appear in every sentence you will ever read. The workhorse is is, ea, id (he, she, it, that), with the irregular genitive eius, dative eī, and plural eōrum, eārum, eōrum, eīs. Alongside it stand hic, haec, hoc (this, near me: gen. huius, dat. huic, acc. hunc, hanc, hoc), ille, illa, illud (that, over there: gen. illīus, dat. illī), ipse, ipsa, ipsum (himself, the very), and īdem, eadem, idem (the same). The relative quī, quae, quod has the same irregular pattern as the interrogative: gen. cuius, dat. cui, acc. quem, quam, quod, abl. quō, quā, quō, and plural quī, quae, quae with gen. quōrum, quārum, quōrum and dat./abl. quibus. The governing rule for relatives is the one candidates most often break: a relative pronoun takes its number and gender from its antecedent but its case from its own clause. In mīles quem vīdī, the pronoun is masculine singular because mīles is, but accusative because it is the object of vīdī. The reflexive sē (acc. and abl. sē or sēsē, gen. suī, dat. sibi) has no nominative, refers back to the subject of its own clause, and inside indirect speech refers back to the speaker - a distinction that decides the sense of whole paragraphs of oratio obliqua.

Two smaller systems complete the list. Numerals: the cardinals from one to a thousand are needed, but only three decline - ūnus, -a, -um (with the genitive ūnīus and dative ūnī, the pronominal pattern), duo, duae, duo, and trēs, tria; the rest, from quattuor to centum, are indeclinable, while mīlle is an indeclinable adjective in the singular and a neuter noun in the plural (mīlia) taking a partitive genitive, so tria mīlia mīlitum, three thousand soldiers. The ordinals first to tenth (prīmus, secundus, tertius, quārtus, quīntus, sextus, septimus, octāvus, nōnus, decimus) are ordinary 2-1-2 adjectives. Prepositions: those governing the accusative include ad, ante, apud, circum, contrā, inter, per, post, propter, trāns; those governing the ablative include ā or ab, cum, dē, ē or ex, prae, prō, sine. Two are dual: in and sub take the accusative when they express motion towards (in urbem, into the city) and the ablative when they express position (in urbe, in the city). That one distinction is worth more marks in an unseen than any other single point of prepositional usage.

Worked example

Making every agreement explicit

Constructed illustrative sentence (not a quotation): cōnsul, vir fortissimus, mīlitēs ācrēs quōs in Galliā collēgerat trāns flūmen dūxit. Justify the form of every adjective and pronoun, then translate.

cōnsul ... dūxit
cōnsul is nominative singular, the subject; dūxit is third singular perfect active of dūcō. Everything between them is either in apposition to the subject or belongs to the object phrase, and the length of that gap is entirely normal in prose.
vir fortissimus
vir is nominative singular masculine, in apposition to cōnsul - a second naming of the same person, so it takes the same case. fortissimus is the superlative of fortis and agrees with vir: nominative, singular, masculine. Note that the superlative here is best rendered in English not by 'bravest' but by 'a very brave man', which is one of its two standard senses.
mīlitēs ācrēs
mīlitēs is accusative plural (the direct object of dūxit); the nominative and accusative plural of a third-declension masculine noun are identical, so the case is settled by the fact that the sentence already has a subject. ācrēs is the accusative plural masculine of ācer, ācris, ācre, agreeing with mīlitēs. In the plural the three-termination adjective shows no separate masculine and feminine forms.
quōs
The relative pronoun takes its gender and number from its antecedent mīlitēs, so masculine plural; it takes its case from its own clause, where it is the object of collēgerat, so accusative. Hence quōs. A candidate who wrote quī here would be giving it the case of the antecedent, which is the standard error.
in Galliā
Ablative singular, because in here expresses position, not motion: he had gathered them in Gaul. Had the Latin been in Galliam the sense would be motion into Gaul, and the whole clause would describe a different event.
collēgerat and trāns flūmen
collēgerat is pluperfect - the gathering is prior to the leading, and English must mark that with 'had gathered'. trāns takes the accusative, and flūmen, flūminis, n. is a neuter third-declension noun whose nominative and accusative are identical; the preposition settles the case.

Result: The consul, a very brave man, led across the river the keen soldiers whom he had gathered in Gaul. Agreements: fortissimus (nom. sg. m., with vir); ācrēs (acc. pl. m., with mīlitēs); quōs (m. pl. from mīlitēs, acc. from its own clause); Galliā (abl. after in of position); flūmen (acc. n. after trāns).

Typical mistakes

Reading in plus the ablative as motion: in urbe means in the city, in urbem means into the city, and the two produce quite different narratives.
Giving a relative pronoun the case of its antecedent rather than the case its own clause requires.

Active revision

Write out the full declension of ācer, ācris, ācre and of vetus, veteris side by side, then say in one sentence what the difference between them is and why the specification names vetus. Then take a page of verse and, for every adjective, draw a line to the noun it agrees with; count how many of your lines cross another word, and note the largest gap you found.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level in Latin (H443) specification (Cambridge OCR)

The verb: four conjugations, six tenses, three moods, two voices#

StandardOCR H443 Appendix 5d Accidence - verbs of all standard types from all conjugations in all moods, voices and tenses, together with deponent, semi-deponent, defective, irregular and impersonal verbs including the impersonal passive of intransitive verbsOCR H443 Appendix 5d Accidence - compound verbs of regular formation using all the common prefixesDfE GCE AS and A level subject content for ancient languages

The six indicative tenses across the conjugations

Fig. 3Read the future row first. The split between the -bit future of the first two conjugations and the -et future of the last three is the single most common source of misread tenses in an unseen, and the emphasised cell - reget, a future that looks like a present - is where it bites.

Everything a Latin verb can do is generated from four principal parts, and the discipline of learning all four is what separates a candidate who can read from one who can only guess. Take regō, regere, rēxī, rēctum. The first part gives the person endings of the present system and identifies the conjugation; the second, the infinitive, gives the present stem (reg-) from which the present, future and imperfect are built in both voices; the third gives the perfect stem (rēx-) from which the perfect, pluperfect and future perfect active are built; the fourth, the supine, gives the participial stem (rēct-) from which the perfect passive participle, the future active participle and the whole passive perfect system are built. A verb whose third and fourth parts you do not know is a verb whose perfect tenses you cannot recognise, and the perfect is the ordinary tense of Latin narrative.

The four conjugations are identified by the vowel before the -re of the infinitive: -āre (first, amō, amāre), -ēre with a long e (second, moneō, monēre), -ere with a short e (third, regō, regere) and -īre (fourth, audiō, audīre). A fifth pattern, the mixed or third -iō conjugation (capiō, capere), takes third-declension endings in some forms and fourth in others, and it is worth learning as a separate paradigm because capiō, faciō, iaciō, fugiō and their compounds are extremely common. The distinction between the second conjugation's long ē and the third's short e is not a nicety: it is the difference between monēre and regere, and it is the reason the future tense is formed one way in the first two conjugations (-bō, -bis, -bit) and quite another way in the last two (-am, -ēs, -et). A candidate who has not internalised that split will read reget as an imperfect or a subjunctive rather than as a future.

The six indicative tenses divide cleanly into a present system and a perfect system. Present, future and imperfect are built on the present stem and describe an action as ongoing, or about to happen, or repeated and continuous in the past - imperfect regēbat is properly 'he was ruling' or 'he used to rule', not 'he ruled'. Perfect, future perfect and pluperfect are built on the perfect stem and describe an action as completed: rēxit is 'he ruled' or 'he has ruled', rēxerat is 'he had ruled', rēxerit is 'he will have ruled'. In the passive, the present system uses a single word with passive endings (regitur, regēbātur, regētur) while the perfect system uses two words - the perfect passive participle plus a part of sum (rēctus est, rēctus erat, rēctus erit). Note carefully what that means for translation: rēctus est is a perfect, 'he was ruled', not a present.

Beyond the indicative sit the subjunctive, the imperative and the infinitive. The subjunctive has four tenses only - present, imperfect, perfect and pluperfect - and its uses divide into main-clause uses (jussive, hortatory, deliberative, optative, potential, and the generic and potential uses the specification names) and the very large family of subordinate uses treated later in this topic. Its formation is best remembered by markers rather than by paradigms: the present subjunctive replaces the stem vowel (regam, regās; but amem, amēs, with the tell-tale e in the first conjugation), the imperfect subjunctive is the present infinitive plus personal endings (regerem, regerēs - which is why it is the easiest subjunctive to spot), the perfect active subjunctive is the perfect stem plus -erim, and the pluperfect active subjunctive is the perfect stem plus -issem. The imperative is regular (amā, amāte; regē is not a form - the second-declension-style singular imperative of the third conjugation is rege), and four verbs have irregular short imperatives: dīc, dūc, fac and fer.

Finally, the specification demands the awkward verbs by name, and they are worth a page in your own notes. Deponents are passive in form and active in meaning: sequor (I follow), loquor (I speak), patior (I suffer), conor (I try), morior (I die), ūtor (I use, with the ablative), proficīscor (I set out). Semi-deponents are active in the present system and deponent in the perfect: audeō, ausus sum; gaudeō, gāvīsus sum; soleō, solitus sum; fīdō, fīsus sum. Defective verbs have only some forms, notably the perfect-form presents meminī (I remember), ōdī (I hate) and coepī (I began). Impersonals occur only in the third person singular: licet (it is permitted), oportet (it is right), placet (it pleases), and the group with an accusative of the person and a genitive of the cause - mē paenitet, mē pudet, mē taedet. The impersonal passive of intransitive verbs, which the specification lists explicitly, is the one that most often defeats a candidate in an unseen: pugnātum est does not mean 'it was fought' in any English sense but 'there was fighting', 'a battle took place'; ventum est means 'they came'. Add to this the irregular verbs sum, possum, eō, ferō, volō, nōlō, mālō and fīō, and the compound verbs built from every common prefix - ad-, con-, dē-, ē- or ex-, in-, ob-, per-, prae-, prō-, re-, sub-, trāns- - with the consonant and vowel changes they cause: capiō becomes accipiō, faciō becomes efficiō, teneō becomes retineō.

Worked example

Parsing four verbs that look like something else

Constructed illustrative clauses (not quotations). Parse the verb in each and translate: (a) hostēs ad flūmen secūtī sunt. (b) diū ācriter pugnātum est. (c) sī hoc fēcerit, omnia āmittet. (d) cōnsul veritus est nē mīlitēs fugerent.

(a) secūtī sunt
Third person plural perfect indicative of the deponent sequor, sequī, secūtus sum. It is passive in form and active in meaning, so 'they followed', not 'they were followed'. The tell-tale is the dictionary form: a verb whose first principal part ends in -or has no active forms at all. sequor takes a direct object in the accusative, and hostēs here is that object; the subject is unexpressed and must be supplied from context.
(b) pugnātum est
Third person singular perfect indicative passive of pugnō, an intransitive verb - so this is the impersonal passive that the specification lists by name. There is no subject and no direct object to be found, because there is none: the Latin says that fighting happened. Render it as 'there was fierce fighting for a long time' or 'they fought fiercely for a long time'. A candidate who hunts for a subject here will lose the whole clause.
(c) fēcerit
Formally ambiguous between future perfect indicative and perfect subjunctive of faciō. The conditional clause with sī and the future āmittet in the main clause settle it as a future perfect: Latin is being precise about sequence where English is lazy. Translate 'if he does this, he will lose everything' - English uses a present where Latin insists on a future perfect, because the doing precedes the losing.
(d) veritus est
Third person singular perfect of vereor, verērī, veritus sum - a deponent, so active in meaning: 'the consul feared'. Verbs of fearing take nē for a positive fear and ut or nē nōn for a negative one, and the subordinate verb is subjunctive; the sense of nē is inverted relative to its usual 'lest'.
(d) fugerent
Third person plural imperfect subjunctive of fugiō, formed as always by adding personal endings to the present infinitive fugere. The main verb is a perfect with a past sense, so the sequence is historic and the imperfect subjunctive expresses an action contemporaneous with it. The clause means 'that the soldiers would run away' - not 'lest they should run', which is archaic English and blurs the sense.

Result: (a) They followed the enemy to the river. (b) There was fierce fighting for a long time. (c) If he does this, he will lose everything. (d) The consul was afraid that the soldiers would run away. The four traps in order: a deponent, an impersonal passive, a future perfect masquerading as a subjunctive, and a fearing clause with its inverted negative.

Typical mistakes

Translating a perfect passive such as captus est as a present ('he is captured') instead of a perfect ('he was captured').
Reading a deponent as a passive: secūtus est means 'he followed', not 'he was followed'.

Active revision

Choose ten verbs you met for the first time in your last piece of reading. For each, write out all four principal parts from memory, then form and translate the third person singular of every one of the six indicative tenses in both voices, plus the four subjunctive tenses active. Mark the forms you had to look up: those are the verbs to revisit next week, not the ones you got right.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level in Latin (H443) specification (PDF) (Cambridge OCR)

The non-finite system: participles, infinitives, gerund and gerundive#

AdvancedOCR H443 Appendix 5d Syntax - uses of the infinitive (as subject, as complement, prolative, historic)OCR H443 Appendix 5d Syntax - uses of the participle (including ablative absolute); uses of the gerund and gerundive (including the gerundive of obligation)DfE GCE AS and A level subject content for ancient languages

The non-finite system of rego, regere, rexi, rectum

Fig. 4The two dashes are the important cells: Latin has no present passive participle and no perfect active participle for an ordinary verb. Every English 'having done' must therefore be rebuilt - by a deponent, by an ablative absolute, or by a cum-clause.

The non-finite forms are where Latin achieves its compression, and they are the reason a single Latin period can carry what English needs three sentences to say. Latin has three participles for an ordinary transitive verb and not one more: a present active (regēns, regentis - ruling, while ruling), a perfect passive (rēctus, -a, -um - having been ruled), and a future active (rēctūrus, -a, -um - about to rule, intending to rule). Two gaps in that pattern cause almost all the trouble. There is no present passive participle, so 'being ruled' must be rendered by a clause; and there is no perfect active participle, so 'having ruled' must be rendered by cum plus the pluperfect subjunctive or by an ablative absolute. The one exception is precious: a deponent verb, being passive in form and active in meaning, supplies a perfect participle that is active in sense - secūtus means 'having followed', profectus 'having set out'. When you meet one, you have found the compact past-active participle that Latin otherwise lacks.

The ablative absolute is the construction to master above all others, because it is everywhere in historical prose and it is the standard way Latin subordinates one event to another. It consists of a noun or pronoun in the ablative together with a participle in the ablative agreeing with it, and it is grammatically detached from the rest of the sentence - hence 'absolute', from absolūtus, loosed. urbe captā means literally 'the city having been captured', and the point of translating it is to choose the English connective that the context actually implies: 'when the city had been captured', 'after the city was captured', 'since the city had been captured', 'although the city had been captured'. Latin leaves that logical relation unstated; a good translation states it. A variety without any participle at all is equally common because sum has no present participle: Caesare cōnsule means 'when Caesar was consul', mē duce 'under my leadership'.

The infinitive is a verbal noun and does four jobs that the specification lists explicitly. As subject: errāre hūmānum est, to err is human. As complement, after sum and similar verbs. Prolative, completing the sense of another verb - volō īre, I want to go; possum vidēre, I am able to see - which is the commonest use of all. And historic, a narrative present infinitive used with a nominative subject to convey rapid, breathless action: mīlitēs currere, clāmāre, arma capere, 'the soldiers began running, shouting, seizing weapons'. That last is a genuine A-Level construction: a nominative subject with an infinitive and no finite verb in sight is not a mistake in the text but a deliberate device, and it is a marker of Livy's battle narrative in particular. Its English rendering should carry the same speed - a string of past tenses, not a laborious 'proceeded to'.

The gerund is a verbal noun, neuter, singular, and it exists only in the oblique cases: genitive regendī, dative regendō, accusative (after ad) regendum, ablative regendō. It means the act of doing something - ars regendī, the art of ruling; ad urbem capiendam is not a gerund but its close relative. The gerundive, regendus, -a, -um, is a verbal adjective meaning 'needing to be done', and it agrees with a noun like any other adjective. Latin strongly prefers the gerundive-plus-noun construction to a gerund with a direct object, so where English says 'for capturing the city' Latin says ad urbem capiendam - literally 'towards the city to-be-captured'. Recognising that phrase as a purpose expression, rather than trying to construe capiendam as a passive participle, is a standard A-Level hurdle.

The gerundive's most important use is the gerundive of obligation, also called the passive periphrastic: the gerundive plus a part of sum expresses necessity. urbs dēfendenda est means 'the city must be defended'. The person on whom the obligation falls goes into the dative, not into a phrase with ā or ab - urbs cōnsulī dēfendenda est, 'the city must be defended by the consul'. That dative of agent is peculiar to this construction and is examinable in exactly the form just given. Intransitive verbs form the same construction impersonally: eundum est, 'one must go'; mihi eundum est, 'I must go'. Finally, note the future infinitive passive, rēctum īrī, which uses the supine plus the passive infinitive of eō and is invariable: it appears almost only in indirect statement, and it is worth learning as a fixed shape rather than analysing it every time.

Worked example

Unpicking a participle-heavy period

Constructed illustrative sentence (not a quotation): urbe captā, dux, mīlitēs ad flūmen secūtus, nūntium Rōmam mīsit quī cōnsulibus omnia dīceret; nam pāx magnō studiō petenda erat. Translate, naming every non-finite form.

urbe captā
Ablative absolute: urbe is ablative singular of urbs, captā is the perfect passive participle of capiō agreeing with it in case, number and gender. The construction is grammatically independent of the rest of the sentence, and it precedes the main action in time. Choose the connective the narrative supports - here, plainly temporal: 'after the city had been captured'.
dux ... mīsit
The main clause is short and sits at the two ends of the sentence: dux (nominative singular, subject) and mīsit (third singular perfect active of mittō). Everything between is subordinate. Finding the main verb first, and only then reading inwards, is the method for every long Latin sentence.
mīlitēs ... secūtus
secūtus is the perfect participle of the deponent sequor, so it is active in meaning: 'having followed'. It is nominative singular masculine, agreeing with dux, and it takes its own direct object mīlitēs. English will not tolerate the participle phrase in the same position, so unpack it: 'the leader, who had followed his soldiers to the river' or 'having followed his soldiers to the river'.
nūntium ... quī ... dīceret
nūntium is the accusative object of mīsit. quī introduces a relative clause whose verb dīceret is imperfect subjunctive - a relative clause of purpose, one of the constructions listed under Description in the specification. Translate 'to tell the consuls everything', not 'who was telling', because the subjunctive here marks intention, not description.
cōnsulibus omnia
dīcō takes a dative of the person told and an accusative of the thing told: cōnsulibus is dative plural, omnia the neuter accusative plural of omnis. omnia is another instance of the neuter rule - nominative and accusative alike - and the verb resolves it.
pāx ... petenda erat
The gerundive of obligation in a past tense: petenda is the gerundive of petō agreeing with pāx (nominative singular feminine), and erat supplies the tense. 'Peace had to be sought.' magnō studiō is an ablative of manner: 'with great enthusiasm'. Note that no agent is expressed here; had the Latin added cōnsulī it would have been a dative of agent, not a phrase with ab.

Result: After the city had been captured, the leader, having followed his soldiers to the river, sent a messenger to Rome to tell the consuls everything; for peace had to be sought with great enthusiasm. Non-finite forms: captā (perfect passive participle in an ablative absolute); secūtus (perfect participle of a deponent, active in sense); dīceret (finite, but the subjunctive of a purpose relative); petenda (gerundive of obligation with erat).

Typical mistakes

Translating a deponent perfect participle passively: profectus means 'having set out', not 'having been set out'.
Trying to construe ad plus a gerundive as a noun phrase rather than as an expression of purpose.

Active revision

Find five ablative absolutes in a page of Latin narrative prose. For each, write out four different English renderings using four different connectives - when, after, since, although - then decide which the context supports and write one sentence justifying the choice from the surrounding narrative. This is the exercise that turns a mechanical construction into a reading skill.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level in Latin (H443) specification (PDF) (Cambridge OCR)

Syntax drill: patterns of case usage and the subordinate clauses#

AdvancedOCR H443 Appendix 5d Syntax - standard patterns of case usage; negation; direct statement, question and command; prohibitions, exhortations and wishesOCR H443 Appendix 5d Syntax - subordinate clauses: indirect statement (including extended oratio obliqua), question, command and prohibition; description; purpose; result; conditional; causal; temporal; subordinate clauses within indirect speech; fearing, prevention and precaution; concessive; comparativeOCR H443 Appendix 5d Syntax - constructions using quominus and quin; use of dum and dummodo to mean 'provided that'

Standard patterns of case usage

Fig. 5The ablative branch is the one to learn first and revise last: it carries more distinct functions than the other five cases together, and almost every grammar question about an ablative is asking which of these branches you are on.

Accidence tells you what a form is; syntax tells you what it is doing, and the second is where marks are won and lost. The efficient way to hold case usage in the head is as a small tree: each case has one central idea and a handful of specialised branches that all descend from it. The accusative is fundamentally the case of the goal - the goal of a verb's action (direct object), the goal of motion (ad urbem), and the extent covered (multōs annōs, decem pedēs). The genitive is fundamentally the case of the defining relation - whose, of what kind, of what part, of what value. The dative is the case of the person concerned - to whom, for whom, in whose interest, and the special cases of possession with sum and agency with the gerundive. The ablative is the widest and therefore the one to break down: separation (from), instrument and means (by, with), and accompanying circumstance (time when, manner, description, and the absolute construction). Almost every ablative you meet belongs to one of those three families, and naming the family is what a grammar question asks for.

Subordination in Latin is signalled in two ways: by the conjunction, and by the mood of the verb. The indicative in a subordinate clause states a fact; the subjunctive marks the clause as something other than a plain statement - purposed, resultant, reported, feared, conceived. This is why the same conjunction can introduce two quite different clauses. cum with the indicative in a present or future context is simply 'when'; cum with the imperfect or pluperfect subjunctive is the circumstantial cum-clause, translated 'when', 'since' or 'although' according to context. ut with the indicative means 'as' or 'when'; ut with the subjunctive introduces purpose or result. quod with the indicative gives a fact ('because', in the writer's own voice); quod with the subjunctive gives an alleged reason ('on the grounds that'). Reading Latin fast means noticing the mood as soon as you notice the conjunction.

Indirect statement is the construction that dominates Latin prose and the one that most repays drilling. After any verb of saying, thinking, knowing, perceiving or hoping, Latin does not use a 'that' clause: it puts the subject of the reported statement into the accusative and its verb into the infinitive. dīcit rēgem venīre, 'he says that the king is coming'; dīxit rēgem vēnisse, 'he said that the king had come'; dīcit rēgem ventūrum esse, 'he says that the king will come'. The tense of the infinitive is relative, not absolute: the present infinitive means contemporaneous with the main verb, the perfect infinitive prior to it, the future infinitive subsequent to it. That relativity is the whole difficulty, and it is why rēgem vēnisse becomes 'had come' after a past main verb but 'has come' after a present one. Where the reported subject is the same person as the speaker, Latin uses the reflexive sē, and getting sē right in a long stretch of oratio obliqua is one of the marks of a strong A-Level candidate: sē refers to the speaker, eum to somebody else.

Extended oratio obliqua, which the specification names, is indirect statement sustained over several sentences - a whole speech reported. Inside it, three things happen at once. Main statements become accusative-and-infinitive. Main-clause questions and commands become subjunctive. And every subordinate clause inside the reported speech has its verb in the subjunctive too, whatever mood it would have had in direct speech. That last rule is what the specification means by 'subordinate clauses within indirect speech', and it explains a page of Livy that would otherwise look like an unmotivated pile of subjunctives. When you meet such a passage in an unseen, the reading strategy is to find the verb of saying, mark where the reported speech begins, and then read every subjunctive inside it as an ordinary indicative until you reach the end.

The remaining clause types are best learned as a checklist with their triggers, and the table sets them out. Two families are worth extra attention because they involve inverted or unfamiliar logic. Verbs of fearing use nē for what you fear will happen and ut or nē nōn for what you fear will not: timeō nē veniat, 'I am afraid that he will come'; timeō ut veniat, 'I am afraid that he will not come'. Verbs of preventing, hindering and refusing use nē or quōminus after a positive main verb and quīn after a negative one: nōn dubitō quīn vēnerit, 'I do not doubt that he has come'. Finally, dum with the present indicative means 'while', dum with the subjunctive means 'until' or, with modo and dummodo, 'provided that' - three senses from one word, decided by the mood and by whether a condition is being laid down. Negation follows its own small rules too: nōn negates a statement, nē a command or wish, and two negatives in Latin make an emphatic positive, so nōn nūllī means 'some' and nēmō nōn means 'everybody'.

The subordinate clauses of Appendix 5d

Fig. 6A checklist rather than a theory: in an unseen, the trigger comes first and tells you what to expect, so train the recognition of the trigger words themselves. The emphasised row is the one that carries most of the prose you will read.

Worked example

Naming and translating six constructions in one period

Constructed illustrative passage (not a quotation): cōnsul, cum hostēs iam appropinquāre cognōvisset, mīlitibus imperāvit ut pontem statim dēlērent, nē urbs caperētur; tam ācriter autem pugnātum est ut nēmō ex hostibus flūmen trānsīre posset. dux tamen nōn dubitāvit quīn Rōmānī vīcissent. Name each construction and translate.

cum ... cognōvisset
cum with the pluperfect subjunctive: a circumstantial cum-clause. The pluperfect places the knowing before the ordering. English needs a connective that reflects the causal colour here - 'when the consul had learned' or, better given what follows, 'since the consul had learned'.
hostēs iam appropinquāre
Indirect statement dependent on cognōvisset: hostēs is the accusative subject, appropinquāre the present infinitive. Present infinitive means contemporaneous with the main verb, and the main verb is past, so English gives 'that the enemy were now approaching'.
imperāvit ut ... dēlērent
Indirect command. imperō takes a dative of the person commanded (mīlitibus) and ut with the subjunctive for the command itself. Contrast iubeō, which takes an accusative and an infinitive - a distinction worth a mark in a grammar question. dēlērent is imperfect subjunctive, historic sequence.
nē urbs caperētur
A negative purpose clause: nē with the imperfect subjunctive passive of capiō. 'So that the city should not be captured.' Note that nē here is purpose, not fearing; the trigger is the absence of any verb of fearing.
tam ācriter ... ut ... posset
A result clause, announced in advance by tam. The signposting word is the point: tam, tantus, adeō, ita and sīc all warn you that an ut-clause of result is coming, and that ut will therefore not mean 'in order that'. pugnātum est is again the impersonal passive - 'there was such fierce fighting that ...'.
nōn dubitāvit quīn ... vīcissent
Prevention and doubting: a negatived verb of doubting takes quīn with the subjunctive. vīcissent is pluperfect subjunctive, prior to the main verb in historic sequence, so 'that the Romans had won'. English inverts the negative into a positive assertion: 'the leader had no doubt that the Romans had won'.

Result: Since the consul had learned that the enemy were now approaching, he ordered his soldiers to destroy the bridge at once, so that the city should not be captured; and there was such fierce fighting that none of the enemy could cross the river. The leader, however, had no doubt that the Romans had won. Constructions: circumstantial cum-clause; indirect statement; indirect command with imperō; negative purpose; result with tam ... ut; quīn after a negatived verb of doubting.

Typical mistakes

Translating cum plus the subjunctive as a bare 'with', or as 'when' where the context clearly demands 'although'.
Reading the tense of an infinitive in indirect statement as absolute rather than relative to the main verb.

Active revision

Take a paragraph of reported speech from a Latin prose author and rewrite it as direct speech: turn each accusative and infinitive back into a nominative and a finite indicative, each subordinate subjunctive back into whatever mood it would have had, and each se back into a first-person pronoun. Then translate both versions. The differences between them are exactly the constructions the examiner will test.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level in Latin (H443) specification (PDF) (Cambridge OCR) · GCE AS and A level subject content for ancient languages (Department for Education)

Exam technique: answering accidence and syntax questions#

StandardOCR H443 Component 02 Section A - translation, comprehension and questions on syntax and accidence on an unseen prose passage (50 marks)OCR H443 Assessment objective AO1 - demonstrate knowledge and understanding of the languageOCR H443 Appendix 5d - accidence and syntax required for Components 01 and 02

Section A of Component 02 asks three kinds of question about a passage of unseen prose, and the accidence-and-syntax questions are the ones that reward precise vocabulary rather than good instincts. They are short-answer questions worth one or two marks each, and they are answered in English. What the examiner wants is a technical description that could only fit the form in front of you: the answer to 'what part of the verb is dīxisset?' is 'third person singular pluperfect subjunctive active of dīcō', not 'a pluperfect' and not 'he had said'. Every element of that description carries information, and a partial description is a partial answer. Get into the habit of producing all of person, number, tense, mood and voice for a finite verb, and case, number and gender for a noun, adjective or pronoun - even when you suspect only one of them is being asked about.

The second kind of question asks you to explain a case or a mood rather than merely to name it. 'Why is cōnsulī in the dative?' is not answered by 'because it is dative'. It is answered by naming the function and the governing word: 'it is a dative of agent with the gerundive dēfendenda' or 'it is the indirect object of imperāvit' or 'it is governed by persuāsit, which takes the dative'. Three parts, then: the technical name of the usage, the word in the passage that produces it, and - if the usage is unusual - a two-word gloss of the sense. The same applies to moods: 'why is the verb subjunctive?' is answered by 'it is the verb of a purpose clause introduced by ut', not by 'because it follows ut', which is not true of result clauses, indirect commands or ut meaning 'as'.

The third kind asks you to name a construction as a whole. Here the marks are for the technical term, and the terms are a small closed list that you should be able to recite: indirect statement, indirect question, indirect command, purpose clause, result clause, ablative absolute, gerundive of obligation, cum-clause (circumstantial, causal or concessive), conditional clause, relative clause of purpose or characteristic, historic infinitive, impersonal passive. Learn them in English, because that is the language of your answer, and learn them precisely: 'accusative and infinitive' and 'indirect statement' are both acceptable names for the same thing, but 'reported speech' is vaguer and 'a that-clause' is wrong, since Latin does not use one.

Timing is the practical constraint. Component 02 lasts one hour fifteen minutes for fifty marks, and if you have chosen Section A those fifty marks are split between a translation of part of the passage, comprehension questions and the grammar questions. The grammar questions are the fastest marks in the paper for a prepared candidate and the slowest for an unprepared one, because they are pure recall. Do not write essays: a one-line answer that contains the technical term, the governing word and, where asked, the sense, is complete. Conversely, do not abbreviate to the point of ambiguity - 'pluperf. subj.' is fine, 'plup.' is not, and a bare 'ablative' where the question asked you to explain the ablative earns nothing.

One discipline underpins all of this: answer from the form, not from the translation. It is perfectly possible to translate a sentence correctly while misparsing a word in it, and the grammar questions are designed to catch exactly that. If you are asked about vēnisset and you have translated 'he had come', check the ending before you answer: the -isset marks a pluperfect subjunctive, and if you write 'pluperfect indicative' because your English used a plain past perfect, you have described your translation rather than the Latin. Read the word, identify the stem, identify the ending, then name the form - in that order, every time.

Worked example

Ten grammar questions with model answers

Using the constructed illustrative passage from the previous section - cōnsul, cum hostēs iam appropinquāre cognōvisset, mīlitibus imperāvit ut pontem statim dēlērent, nē urbs caperētur; tam ācriter autem pugnātum est ut nēmō ex hostibus flūmen trānsīre posset - answer the following in the style of Component 02 Section A.

Parse cognōvisset
Third person singular pluperfect subjunctive active of cognōscō, cognōscere, cognōvī, cognitum. Two marks: one for the person, number and tense, one for mood and voice. A candidate who stops at 'pluperfect' has given half an answer.
Why is cognōvisset subjunctive?
Because it is the verb of a circumstantial cum-clause. The trigger is cum, and cum with the imperfect or pluperfect subjunctive in a past narrative regularly gives the circumstances of the main action. Naming cum without naming the clause type would not be enough.
In what case is hostēs, and why?
Accusative plural. It is the subject of the infinitive appropinquāre in an indirect statement dependent on cognōvisset. Note the two-part shape of the answer: the case, then the construction that produces it.
What part of speech is appropinquāre, and what is its function?
It is the present active infinitive of appropinquō. It is the verb of the indirect statement, and its present tense marks the approach as contemporaneous with the act of learning - which is why English translates it as 'were approaching' after a past main verb.
In what case is mīlitibus, and why?
Dative plural. imperō governs the dative of the person commanded, unlike iubeō, which takes an accusative with an infinitive. The governing word is the mark-bearing part of the answer.
Name the construction introduced by ut ... dēlērent.
An indirect command. dēlērent is imperfect subjunctive in historic sequence after the perfect imperāvit. 'Purpose clause' is wrong here: the ut-clause reports what was commanded, not why the commanding happened.
Name the construction introduced by nē ... caperētur, and distinguish it from the previous clause.
A negative purpose clause. It states the object the consul hoped to achieve, whereas the preceding ut-clause states the content of his order. Both use the subjunctive; only the context distinguishes them, which is why the examiner has set them side by side.
Parse pugnātum est and explain the construction.
Third person singular perfect indicative passive of pugnō. Because pugnō is intransitive, this is an impersonal passive: there is no subject, and the sense is 'there was fighting'. This is one of the forms the specification names explicitly in Appendix 5d.
Why is posset subjunctive?
It is the verb of a result clause introduced by ut, signalled in advance by tam with the adverb ācriter. The signpost word is worth naming: the examiner is testing whether you can distinguish an ut of result from an ut of purpose, and tam is the evidence.
In what case is hostibus in the phrase ex hostibus, and what does the phrase express?
Ablative plural, governed by the preposition ex. The phrase is partitive in sense - 'none of the enemy' - Latin regularly preferring ex or dē with the ablative to a partitive genitive after words such as nēmō, ūnus and quisque.

Result: Ten answers, each in one sentence, each containing the technical term and the governing word. Total time under examination conditions should be about eight minutes; if it takes longer, the problem is recall of the terminology rather than understanding of the passage - and terminology is the part that can be learned by heart.

Typical mistakes

Answering a parsing question with a translation ('it means he had said') instead of a grammatical description.
Naming the conjunction rather than the construction: 'because of ut' is not an explanation of a subjunctive.

Active revision

Take any twelve-line passage of Latin prose and write your own grammar paper on it: five parsing questions, three 'explain the case' questions and two 'name the construction' questions. Leave it for a week, then answer your own paper under timed conditions and mark it strictly, awarding nothing for an answer that omits the governing word. Setting the questions teaches you what an examiner can see in a passage.

Active recall

Recall the key points — then reveal.

Sources: OCR A Level Latin (H443) - specification at a glance (Cambridge OCR)

Key points

Reading a sentence by its endings alone

Key points

Making every agreement explicit

Key points

Parsing four verbs that look like something else

Key points

Unpicking a participle-heavy period

Key points

Naming and translating six constructions in one period

Key points

Ten grammar questions with model answers

Sources

Accidence and syntax

Key points

Reading a sentence by its endings alone

Key points

Making every agreement explicit

Key points

Parsing four verbs that look like something else

Key points

Unpicking a participle-heavy period

Key points

Naming and translating six constructions in one period

Key points

Ten grammar questions with model answers

Sources