Notes · LatinUK · A-Levels

Vocabulary and linguistic range

There is no Defined Vocabulary List for A-Level Latin. OCR publishes one for AS Level (H043) and expects A-Level candidates to treat it as the floor of their vocabulary rather than its ceiling, extending it by wide reading. That single decision changes how the whole qualification must be prepared: nobody can hand you a finite list to learn, so vocabulary has to be built the way a reader builds it - from texts, in families, with the derivational machinery of the language doing most of the work. These notes set out what that means in practice. First, what the specification actually requires and how to run a vocabulary notebook that repays the time. Second, word-building: the prefixes and suffixes that turn a hundred simple verbs into a thousand compounds, and the sound changes they cause. Third, semantic range - the fact that a Latin word covers a territory rather than translating to a single English word, and that translating it well means choosing inside that territory. Fourth, a drill on principal parts, because the perfect stem is a vocabulary problem as much as a morphological one. And finally the examinable skill this all serves: meeting a word you have never seen, in an unseen passage, with no dictionary, and producing a translation you can defend.

5 sections~43 min reading time3 competenciesLevel Standard 3 · Advanced 2

T·0222 / 10

Exam profile

AO1 · Demonstrate knowledge and understanding of the Latin language: a vocabulary and range of structures sufficient to read unadapted Latin at speedAO2 · Demonstrate knowledge and understanding of literature by reading the set texts in the original language without constant recourse to a glossaryAO3 · Critically analyse and evaluate literature, using an author's choice of word as evidence for a literary claim

Operators:translatededucegive the meaningidentify the rootexplain the sensedistinguish

basic level

AS Level Latin (H043) has a Defined Vocabulary List, and everything in the AS papers is written within it or glossed. An AS candidate's task is therefore finite: learn the list, and learn it in both directions, so that a form on the page produces a meaning and an English meaning produces a Latin form. Word-building matters even at this level, because the list contains simple verbs whose compounds appear freely.

higher level

The A-Level removes the list. Vocabulary is expected to grow by reading, which means the A-Level candidate must be able to deduce as well as recall, must know the derivational system well enough to unpack an unfamiliar compound on sight, and must be alert to the wider semantic range that real authors exploit - the metaphorical use of a military term, the technical use of an everyday one, the moral colour a word carries for a Roman reader that its English equivalent does not carry for us.

Depth

Text

Contents · 5 sections

Vocabulary and linguistic range

Vocabulary without a list: what the specification requires#

StandardOCR H443 - there is no Defined Vocabulary List at A-Level; learners build on the AS Level Latin (H043) Defined Vocabulary List through wider readingOCR H443 Component 01 - unseen translation from Livy and Ovid, with no defined vocabularyDfE GCE AS and A level subject content for ancient languages

The absence of an A-Level vocabulary list is the most consequential administrative fact in the specification, and it is routinely underestimated. At GCSE and at AS Level a candidate can, in principle, learn every word that will appear. At A-Level that guarantee is withdrawn: the unseen passages come from Livy and Ovid as those authors wrote them, and the set texts are real literary works. OCR's expectation is stated plainly - learners build on the Defined Vocabulary List published for AS Level Latin (H043) and extend it through wider reading. The practical consequence is that vocabulary preparation stops being an exercise in memorising a finite document and becomes a two-year habit: read a great deal, record what you meet in a form you will actually revise, and learn the machinery of the language well enough that a word you have never seen is often not really unknown at all.

Start by making the AS list secure, because it is not a small thing to have. It contains the high-frequency core of the language - the words that make up the majority of any continuous passage - and a candidate who is fluent in it is already reading most of the running text of a Livian narrative without hesitation. Fluency here means something stricter than recognition. It means that dux produces 'leader' without a pause, that dux, ducis, m. produces the third-declension stem duc- without a pause, and that dūcō, dūcere, dūxī, ductum produces the perfect dūxit without a pause. Slow recognition is as expensive as no recognition in a timed translation, because the time is spent at the level of the individual word and there are nine hundred of them in a paper.

Above that floor, vocabulary grows from reading, and the reading has to be of the right kind. Two habits do most of the work. The first is extensive reading: covering ground, with a facing translation or a good commentary at your elbow, at a pace that keeps the narrative moving. This is where you meet words in their natural frequencies, and frequency is what fixes a word in the memory far more reliably than a list does. The second is intensive reading: taking twenty lines and working every word of them until nothing is unexplained. Extensive reading builds range; intensive reading builds precision. A course that does only the first produces candidates who half-know a great many words; a course that does only the second produces candidates who know four hundred lines perfectly and stall on the four hundred and first.

The vocabulary notebook is worth designing rather than improvising, because most of them fail in the same way: they become undifferentiated lists of word-plus-gloss that nobody rereads. A useful entry has four parts. The full dictionary form, with all the information the form carries - not 'corpus: body' but corpus, corporis, n., and not 'fero: I bear' but ferō, ferre, tulī, lātum. The sense in the passage where you met it, in the words the passage needed. One other sense the word commonly carries, so that the entry records a range and not a single point. And the phrase it came in, three or four words long, because words are recalled in company. An entry built that way takes twenty seconds to write and can be revised at a glance in six months.

Finally, be honest about the size of the target. Nobody can compute how many words an A-Level candidate needs, and any figure offered is a guess. What can be said with confidence is where effort pays: the commonest few hundred words repay outright memorisation, the next band repays repeated encounter in reading, and the long tail of rare words - the ones an author uses once - is not worth memorising at all, because the skill that handles them is deduction, not recall. Plan accordingly. Time spent making the core automatic and the deduction method reliable is worth several times the same time spent extending a list into territory where the words are too rare to reappear.

Worked example

Building four notebook entries from one sentence

Constructed illustrative sentence (not a quotation): lēgātī, quī ab hostibus missī erant, cōnsulem ōrāvērunt ut obsidibus acceptīs bellum fīnīret. Write the notebook entry you would make for lēgātus, obses, ōrō and fīniō.

lēgātī
Full form: lēgātus, lēgātī, m. Sense here: 'envoys', men sent to negotiate - the word is a perfect passive participle of lēgō, lēgāre (to send as a deputy) used as a noun, which is why it means 'one who has been sent'. Other common sense: a legate, that is, a senior officer serving as a commander's deputy, and by extension the commander of a legion. Phrase to record: lēgātī ab hostibus missī. Recording both senses matters because the military sense and the diplomatic sense are equally common in historical prose and only context separates them.
obsidibus
Full form: obses, obsidis, m. and f. Sense here: 'hostages', people handed over as a guarantee. Note the third-declension stem obsid-, which the nominative obses conceals - this is exactly why the entry must carry the genitive. Related word worth recording alongside it: obsidiō, obsidiōnis, f., a siege, and obsideō, obsidēre, to besiege; all three descend from sedeō, to sit, with the prefix ob-, so the underlying image is of sitting in someone's way. Phrase: obsidibus acceptīs, an ablative absolute.
ōrāvērunt
Full form: ōrō, ōrāre, ōrāvī, ōrātum. Sense here: 'begged', 'entreated' - and note the construction, because for this verb the construction is half the entry: ōrō takes an accusative of the person entreated and ut with the subjunctive for the request. Other sense: to speak formally, to plead a case, which is the sense that produces ōrātiō (a speech) and ōrātor. A word whose noun you already know is half-learned before you meet it.
fīnīret
Full form: fīniō, fīnīre, fīnīvī, fīnītum. Sense here: 'should end', transitive - to bring something to an end. Root: fīnis, fīnis, m., an end or boundary, and in the plural fīnēs, territory, the land inside the boundaries. Recording the root noun with the verb gives you two words for the price of one and explains the plural sense, which otherwise looks arbitrary. Phrase: bellum fīnīret.
What the four entries have in common
Each entry carries a form, a sense in context, a second sense or a governed construction, and a phrase. Three of the four also carry a related word - lēgātus from lēgō, obses from sedeō, fīniō from fīnis. That last column is the one that compounds: after a term of entries like these, a large proportion of new words arrive already attached to something in the notebook, and the notebook stops growing linearly.

Result: Translation of the sentence: The envoys, who had been sent by the enemy, begged the consul to accept hostages and end the war. Four entries, each in four parts - dictionary form, sense in context, second sense or construction, phrase - and three of them cross-referenced to a root that was already known. The entry for obses also carries obsidiō and obsideō, which is three words recorded in the time one takes.

Typical mistakes

Learning a word in one direction only: recognition without production is enough for translation but not for the composition option, where the English word has to produce the Latin one.
Recording a single English gloss and then defending it in translation even when the sentence plainly wants a different sense of the same Latin word.

Active revision

Take the last twenty lines of Latin you read and mark every word you had to look up. Sort them into three piles: words that are in the AS Defined Vocabulary List and should not have needed looking up; words that are transparent compounds or derivatives of something you know; and words that are genuinely new. Only the third pile belongs in your notebook. The first pile is a revision problem and the second is a word-building problem, and treating all three as the same problem is why most vocabulary notebooks stop working by the second term.

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)

Word-building: prefixes, suffixes and the shape of a Latin word#

StandardOCR H443 Appendix 5d Accidence - compound verbs of regular formation using all the common prefixesOCR H443 - vocabulary extended by wider reading rather than by a defined listDfE GCE AS and A level subject content for ancient languages

Common verbal prefixes and the changes they cause

Fig. 1Read the third and fifth columns together. Almost every compound in the language is the fourth column with the first column stuck on the front, and the only reason it does not look like it is the assimilation in column three and the vowel weakening visible in column five. The emphasised cell shows both changes at once: ad plus capiō gives accipiō, with the d assimilated to c and the root a weakened to i.

Latin builds words with a small and highly regular kit, and learning the kit is the single highest-yield piece of vocabulary work available to an A-Level candidate. The verb system is where the return is greatest. Take a simple verb you already know - capiō, take - and prefix it: accipiō (receive), excipiō (take out, receive, make an exception of), incipiō (begin), recipiō (take back, recover), suscipiō (undertake), dēcipiō (deceive, literally take away), praecipiō (take beforehand, instruct), percipiō (take thoroughly, perceive), concipiō (take together, conceive). Nine common verbs from one root and eight prefixes, and the same operation applies to dūcō, mittō, ferō, faciō, veniō, cēdō, pōnō, tollō, teneō, agō and a dozen more. A candidate who has done this consciously for twenty core verbs has effectively added several hundred words to their reading vocabulary, and has added them in a form that will not be forgotten because it is a system rather than a list.

Two sound changes accompany the process and must be recognised on sight, because they disguise the root. The first is assimilation: the final consonant of the prefix accommodates itself to the first consonant of the verb. ad- becomes ac- before c (accipiō), af- before f (afferō), ag- before g, ap- before p (appellō), at- before t (attingō), ar- before r (arripiō). in- becomes il- before l (illūstrō), im- before m and p (immittō, impōnō), ir- before r (irrumpō). con- becomes com- before b, m and p (compōnō, commoveō) and col- or cor- before l and r; before a vowel it reduces to co- (coeō, coerceō). ob- gives oc-, of-, op- (occurrō, offerō, oppōnō); sub- gives suc-, suf-, sup-, sus- (succēdō, sufferō, suppōnō, suscipiō); ex- gives ef- before f (efferō, efficiō). None of this is arbitrary - every change makes the word easier to say - and once you expect it you stop being surprised by it.

The second change is vowel weakening in the root. A short a or e in the first syllable of the simple verb regularly weakens to i in the compound, or to e before r: capiō gives accipiō, faciō gives efficiō and perficiō, iaciō gives coniciō and abiciō, cadō gives incidō, teneō gives retineō, premō gives opprimō, habeō gives prohibeō and adhibeō, salīō gives īnsiliō, damnō gives condemnō, and legō gives colligō. Where the root vowel is long or a diphthong it usually survives: dūcō gives dēdūcō, claudō gives inclūdō with the diphthong reduced to ū, caedō gives occīdō. That last pair is worth memorising as a pair, because occidō with a short i (from cadō) means to fall, to set, to die, while occīdō with a long ī (from caedō) means to kill - and the difference in an unseen is the difference between a general who died and a general who slaughtered somebody.

Prefixes carry meaning as well as sound, and their meanings are worth learning as a short list: ad- towards, ab- away, ante- before, circum- around, con- together or thoroughly, contrā- against, dē- down or away or utterly, ē or ex- out of, in- into or against, inter- between, ob- in the way of, per- through or thoroughly, prae- in front or beforehand, prō- forward or on behalf of, re- back or again, sub- under or secretly, super- over, trāns- across. Two of them do double duty and must be distinguished by context. in- prefixed to a verb means into or against (invādō, incurrō), but prefixed to an adjective or a participle it is the negative (incertus, indoctus, invictus, ignōtus - where in- has assimilated before g). And per- and prae- are both intensifiers as well as directional prefixes: perterritus is thoroughly frightened, praeclārus is outstandingly distinguished.

The suffix system builds nouns, adjectives and adverbs with the same regularity, and knowing it lets you read a derived word without ever having met it. From verbs: -tor and -trīx give the agent (victor, victrīx; imperātor; ōrātor), -tiō and -siō give the action or its result (ōrātiō, obsidiō, occāsiō), -men and -mentum give the result or the instrument (flūmen from fluō, agmen from agō, monumentum from moneō, impedīmentum from impediō). From adjectives: -tās, -tūdō and -ia give the abstract quality (lībertās from līber, celeritās from celer, magnitūdō from magnus, fortitūdō from fortis, audācia from audāx, prūdentia from prūdēns). From nouns: -ōsus means full of (perīculōsus, glōriōsus), -eus means made of (aureus, ferreus), -ānus and -īnus mean belonging to (Rōmānus, urbānus, dīvīnus), and -ulus, -culus and -ellus make diminutives (homunculus from homō, nāvicula from nāvis, libellus from liber, ocellus from oculus). Two verb suffixes complete the picture: -tō or -itō makes a frequentative, a verb of repeated action (dictō from dīcō, agitō from agō, clāmitō from clāmō, cursō from currō), and -scō makes an inceptive, a verb of beginning (crēscō, senēscō, pallēscō, obdormīscō). A frequentative in an unseen tells you an action was done again and again, which is a piece of narrative information most candidates simply discard.

Worked example

Unpacking eight compounds of one verb

Take the simple verb pōnō, pōnere, posuī, positum (to place, to put). Work out the form and the meaning of the compounds with ad-, com-, dē-, dis-, ex-, im- (in-), op- (ob-) and prae-, and say in each case which part of the meaning comes from the prefix.

ad- and com-
ad plus pōnō gives appōnō (the d assimilates to p): to place beside, to serve at table. com plus pōnō gives compōnō: to place together, hence to arrange, to compose, to settle a quarrel. Notice that con- has taken its com- form before the p, exactly as the table predicts, and that the metaphorical senses - composing a poem, settling a dispute - are all reachable from 'placing things together'.
dē- and dis-
dēpōnō: to place down, hence to lay aside, to give up, to deposit. dispōnō: to place apart, hence to distribute, to arrange at intervals, to draw troops up in position. dis- is the prefix of separation and appears also as dī- before certain consonants (dīmittō, dīvidō); it always adds the idea of apart, in different directions, or thoroughly.
ex- and im-
expōnō: to place out, hence to set out, to expose, to explain - the last of these being a metaphor English shares in 'expound'. in plus pōnō gives impōnō (the n assimilates to m before p): to place upon, to impose, and with the dative of the thing placed upon. The dative is part of the word's meaning and belongs in the notebook entry with it.
op- and prae-
ob plus pōnō gives oppōnō: to place against, to set in opposition, to oppose - again with a dative. praepōnō: to place in front, and idiomatically to put someone in charge of something, again with the dative: exercituī praepositus, put in command of the army. praepositus then functions as a noun meaning a commander, which is how a participle becomes a piece of vocabulary.
The principal parts do not change
Every one of these compounds keeps the principal parts of the simple verb, adjusted for the prefix: appōnō, appōnere, apposuī, appositum; expōnō, expōnere, exposuī, expositum. This is the reason the exercise pays: one set of principal parts, learned once, serves nine verbs. When a compound's principal parts do surprise you - as with faciō, whose compounds are efficiō, efficere, effēcī, effectum with the weakened vowel throughout - the surprise is systematic and applies to the whole family.
Reading test
In a passage you might meet cōnsul mīlitēs in colle disposuit or urbe captā, dux arma dēposuit. In the first, dis- plus the idea of placing gives 'the consul drew up his soldiers at intervals on the hill'; in the second, dē- gives 'the leader laid down his arms', that is, surrendered. Neither sentence needs a dictionary if pōnō and the prefix list are secure.

Result: Eight compounds - appōnō, compōnō, dēpōnō, dispōnō, expōnō, impōnō, oppōnō, praepōnō - all built from one set of principal parts, all with meanings predictable from the prefix, and three of them (impōnō, oppōnō, praepōnō) governing a dative that belongs in the vocabulary entry. The same exercise on ferō, mittō, veniō, cēdō and teneō yields another forty verbs.

Typical mistakes

Failing to see the root through an assimilated prefix: suscipiō is sub plus capiō, and a candidate who does not undo the assimilation treats it as a new word.
Confusing the two verbs that are both spelled occido without macrons: occidō, from cadō, means to fall, to set or to die, while occīdō, from caedō, means to kill.

Active revision

Choose five simple verbs from the core vocabulary - suggested: ferō, mittō, veniō, pōnō and cēdō. For each, write out every compound you can form with the prefixes ad-, con-, dē-, ex-, in-, ob-, prae-, prō-, re-, sub- and trāns-, giving the assimilated spelling where assimilation occurs and the English meaning. Then find, in whatever you are currently reading, three of the compounds you wrote down. If none of your fifty-five forms appears in a page of prose, you have chosen the wrong five verbs.

Active recall

Recall the key points — then reveal.

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

Semantic range: one Latin word, several English words#

AdvancedOCR H443 Component 01 - translation into English of unseen prose and verse, marked for accuracy and for senseOCR H443 Components 03 and 04 - translation of the set text into English as part of Sections A and BDfE GCE AS and A level subject content for ancient languages

A Latin word does not translate to an English word; it covers a territory, and the English words available cover different territories that overlap it imperfectly. This is the single most important thing to understand about vocabulary at A-Level, and it is what separates a translation that scores from one that merely decodes. Consider rēs, reī, f., the most elastic noun in the language. In different sentences it is a thing, a matter, an affair, a business, a fact, a circumstance, a situation, property, or an event; rēs pūblica is the state, rēs gestae are achievements, rēs mīlitāris is warfare, rē vērā means in fact, and quā dē rē means for which reason. Nobody learns those as ten separate words. What you learn is the territory - rēs is the maximally general noun, the placeholder for whatever the sentence is about - and then you choose the English word that the particular sentence needs. A translation that renders every rēs as 'thing' is legible but wooden, and in a marked translation woodenness costs.

The same is true of the great abstract nouns, and here the stakes are literary as well as linguistic. virtūs is built on vir, a man, and its first sense is manliness; from there it runs through courage, then excellence of any kind, then moral virtue. When a Roman historian praises a general's virtūs he is not calling him virtuous in the modern sense, and translating it so imports a Christian colouring the Latin does not have. fidēs is trust, and it points in both directions at once - the trust one places and the trustworthiness one has - so it means good faith, loyalty, credibility, a promise, and also protection: in fidem venīre is to come under someone's protection. pietās is dutifulness towards gods, country and family, and 'piety' catches only the first third. imperium runs from a single order, through the legal power of a magistrate to give orders, to a sphere of command, to the empire itself. grātia is favour given, favour received, influence, and thanks; grātiās agere is to thank, and grātiā placed after a genitive means for the sake of.

Verbs behave the same way, and the commonest verbs have the widest ranges. petō is to seek, and therefore to ask for (auxilium petere), to make for (urbem petere), and to attack (hostem petere) - three quite different narrative events from one verb, distinguished only by object and context. ferō is to carry, to bear, to endure, and to report or say (fertur, it is said). agō is to drive, to do, to spend time (vītam agere), to plead a case, to discuss. tollō is to lift, and therefore to raise, to remove, and to destroy. cēdō is to go, and therefore to yield, to withdraw, to grant. In each case the English words are far apart and the Latin word is one thing: the trick is to hold the underlying image and let the sentence select. If you find yourself deciding between 'attack' and 'ask for' when you meet petō, look at the object.

Then there are the words that look like English words and are not, and these cost marks in every examination series. hostis is the public enemy, the enemy in war; a personal enemy is inimīcus, and rendering hostis as 'a host' or inimīcus as 'a hostile' is both wrong and expensive. familia is the household, including its slaves and property, not the modern family; the family in the sense of a bloodline is gēns or domus. līberī, from līber, means children - specifically the free-born children of the household. pius is dutiful, not pious. patior is to suffer or allow, not to be patient. dēleō is to destroy. cōnfīdō is to trust, and it governs the dative or the ablative rather than an object. imperātor is a commander, and only in imperial contexts does it approach 'emperor'; princeps is the leading man, the first citizen. magistrātus is a public office, or the man holding it, not a magistrate in the modern judicial sense. Keep a running list of these in the back of the notebook; they are a finite and predictable set.

Idiom is the last layer, and it is learned by phrase, not by word. Some of the commonest patterns in narrative prose have no word-for-word English: rēs ita sē habet, the matter stands thus; quid tibi vīs?, what do you mean?; nihil est quod timeās, there is no reason for you to be afraid; certiōrem facere aliquem dē aliquā rē, to inform someone about something (literally to make him more certain); grātiam habēre, to feel gratitude, as against grātiās agere, to express it; iter facere, to march; castra pōnere, to pitch camp; proelium committere, to join battle; poenās dare, to pay the penalty, where the subject is the guilty party and not the punisher - a reversal that catches candidates every year. These phrases are high-frequency in exactly the historical prose Component 01 draws on, and each one learned is a whole clause that will not need to be reasoned out under time pressure.

Worked example

One verb, five sentences, five English verbs

Constructed illustrative sentences (not quotations). Translate petō in each, and justify the choice from the sentence: (a) lēgātī ā cōnsule auxilium petīvērunt. (b) hostēs, castrīs relictīs, montēs petēbant. (c) mīles gladiō ducem petīvit. (d) puer patrem lacrimīs petēbat. (e) haec silva saepe ā mercātōribus petitur.

(a) auxilium petīvērunt
The object is an abstract noun, auxilium, and the source is marked with ā plus the ablative, ā cōnsule. Neither 'made for' nor 'attacked' can take that shape: you cannot make for help, and attacking help is meaningless. The sense is 'asked for': 'The envoys asked the consul for help.' The construction is part of the evidence - petō in this sense regularly has ā or ab with the person asked.
(b) montēs petēbant
The object is a place, the subject is an army that has just abandoned its camp, and the tense is imperfect, describing a movement in progress. The sense is 'were making for': 'The enemy, having abandoned their camp, were making for the mountains.' Note that 'were attacking the mountains' is grammatically possible and narratively absurd, which is exactly the kind of judgement the mark scheme is testing.
(c) gladiō ducem petīvit
The object is a person, the instrument is a sword in the ablative, and the subject is a soldier. The sense is 'attacked' or, more precisely here, 'made for with intent to strike': 'The soldier attacked the leader with his sword.' The ablative of instrument is the decisive evidence; it belongs to a physical action, not a request.
(d) patrem lacrimīs petēbat
The object is a person, as in (c), and the instrument is in the ablative, as in (c) - but the instrument is tears, and the subject is a child. The physical sense will not carry tears. The sense returns to entreaty: 'The boy was entreating his father with tears.' This pair, (c) and (d), is the point of the exercise: the same grammatical shape, and the meaning decided by the lexical content of the ablative.
(e) ā mercātōribus petitur
A passive, with the agent in ā plus the ablative and a wood as the subject. The sense is 'is visited', 'is resorted to', 'is sought out': 'This wood is often frequented by merchants.' English has no single verb that covers the Latin here, so the translator chooses one that fits the sentence and accepts that a different sentence would need a different one.
The underlying territory
All five senses descend from one image: directed movement towards a target. Ask for, make for, attack, entreat, resort to - each is that image with a different kind of target and a different degree of literalness. Holding the image is what lets you choose quickly; holding five separate glosses means testing each against the sentence in turn, which is slow.

Result: (a) The envoys asked the consul for help. (b) The enemy, having abandoned their camp, were making for the mountains. (c) The soldier attacked the leader with his sword. (d) The boy was entreating his father with tears. (e) This wood is often frequented by merchants. One verb, five English verbs, and in each case the deciding evidence is the object, the construction or the lexical content of the ablative - never the dictionary order.

Typical mistakes

Rendering every occurrence of rēs as 'thing', which produces English that no examiner can read as a natural rendering of the Latin.
Reading poenās dare as 'to inflict punishment': the subject of the phrase is the person being punished.

Active revision

Take the five words rēs, virtūs, fidēs, petō and ferō. For each, find five occurrences across your reading - not made-up sentences, but places in a real text - and write out the phrase plus the English rendering the context requires. Then write one sentence explaining what all five renderings have in common, which is the underlying territory. Do the same again in a month with five different words; twenty words treated this way is worth two hundred treated as glosses.

Active recall

Recall the key points — then reveal.

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

Morphology drill: principal parts and the perfect stem#

AdvancedOCR H443 Appendix 5d Accidence - verbs of all standard types from all conjugations in all moods, voices and tensesOCR H443 Component 01 - unseen translation, in which the perfect is the ordinary tense of narrativeDfE GCE AS and A level subject content for ancient languages

A verb's principal parts are a vocabulary problem disguised as a morphological one. The present stem is usually predictable from the infinitive, and the fourth part is usually predictable from the third; what is not predictable is the perfect stem, and the perfect is the tense in which Latin narrative is written. If you know veniō, venīre but not vēnī, then vēnērunt - one of the commonest words in historical prose - is opaque to you. The drill in this section is therefore not decorative: the whole of Component 01 Section A is a narrative in the perfect and pluperfect, and every unrecognised perfect stem is a clause lost. Learn all four parts, always, from the first encounter. The cost of doing so is a few seconds per verb; the cost of not doing so is paid at speed in the examination hall.

Perfect stems are not random, and grouping them by formation makes several hundred verbs memorable. There are five families. The -v- perfect belongs to the first and fourth conjugations and to some others: amāvī, audīvī, dēlēvī, petīvī, cupīvī. The -u- perfect belongs mostly to the second conjugation: monuī, habuī, tenuī, aluī, voluī, and note posuī from pōnō, which looks second-conjugation and is not. The -s- or sigmatic perfect belongs mostly to the third: rēxī from regō, dūxī from dūcō, scrīpsī from scrībō, mīsī from mittō, gessī from gerō, mānsī from maneō, iūssī from iubeō, sēnsī from sentiō, clausī from claudō. The lengthened-vowel perfect changes the root vowel and adds nothing: vēnī from veniō, lēgī from legō, ēgī from agō, fēcī from faciō, cēpī from capiō, fūgī from fugiō, vīdī from videō, sēdī from sedeō, ēmī from emō. And the reduplicated perfect repeats the initial consonant with a vowel: cecinī from canō, cecidī from cadō, tetigī from tangō, dedī from dō, stetī from stō, poposcī from poscō, fefellī from fallō, cucurrī from currō, momordī from mordeō. A sixth, smaller group changes nothing at all: vertī, solvī, dēfendī, ostendī, contendī.

Two consequences of that grouping are directly useful when reading. First, an unfamiliar perfect can often be traced back to its present: if you meet trāxit and do not recognise it, the -x- says sigmatic perfect, -x- comes from a root ending in -c- or -g-, and trahō is the candidate - which it is. If you meet pepercit, the reduplication says the present begins with p and the root is parc-, giving parcō, to spare. Second, the reverse operation is what the composition option needs: given the English 'he spared', you must produce pepercit, and given 'they had drawn up', dūxerant. Recognition and production are separate skills built on the same knowledge, and a candidate preparing Section B of Component 02 needs both.

The perfect stem also carries two contractions that Livy and Ovid use constantly and that defeat unprepared candidates. The first is syncope: where the perfect stem ends in -v-, that -v- and its following vowel may drop out before s or r. amāvistī becomes amāstī, amāvisse becomes amāsse, audīvistis becomes audīstis, nōvisse becomes nōsse, dēlēvērunt becomes dēlērunt, petīvērunt becomes petiērunt. The word on the page looks shorter and stranger than the form in your paradigm, and the only defence is to have met the phenomenon in advance. The second is the alternative third-person plural perfect ending -ēre for -ērunt: dīxēre for dīxērunt, fuēre for fuērunt, vīdēre for vīdērunt. This is not an infinitive, though it looks like one, and it is extremely common in verse and in the historians who write in a poetic register. A candidate who reads dīxēre as 'to speak' has lost the main verb of the sentence.

Finally, keep the fourth principal part in view, because it is not a curiosity. The supine stem generates the perfect passive participle, which generates the entire passive perfect system and every ablative absolute in the language; it generates the future active participle in -ūrus, which generates the future infinitive and therefore every future-referring indirect statement; and it generates a great deal of vocabulary - the -tor agent nouns, the -tiō action nouns, and the -tūrus and -tus adjectives that behave as ordinary words. From regō, rēctum you get rēctus, rēctūrus, rēctor and, at one remove, dīrēctiō and corrēctor. From videō, vīsum you get vīsus, vīsūrus, and the noun vīsiō. Learning the fourth part is therefore not learning one form; it is learning the door into a whole quarter of the vocabulary.

Worked example

Reading five disguised perfects

Constructed illustrative clauses (not quotations). Identify the verb, give its principal parts, parse the form and translate: (a) haec cum dīxēre, discessērunt. (b) rēx urbem oppugnāre nōn ausus est. (c) mīlitēs nostrī hostēs fūdēre. (d) nōnne haec omnia nōstī? (e) pepercit victīs et superbōs domuit.

(a) dīxēre
The -ēre ending is the alternative third person plural perfect indicative active, equivalent to dīxērunt. The verb is dīcō, dīcere, dīxī, dictum - a sigmatic perfect, with the -c- of the root combining with the -s- to give -x-. Parse: third person plural perfect indicative active. Translate 'when they had said these things, they departed'. The trap is real: dīxēre is one letter away from looking like an infinitive, and the sentence has no other finite verb to fall back on.
(b) ausus est
audeō, audēre, ausus sum - a semi-deponent, active in the present system and deponent in the perfect. So ausus est is third person singular perfect indicative, active in meaning: 'he did not dare'. Two errors are available here and both are common: reading it as a passive ('he was dared', which is meaningless) and reading it as a present. oppugnāre is a prolative infinitive completing ausus est.
(c) fūdēre
Again the -ēre third plural, this time of fundō, fundere, fūdī, fūsum - a lengthened-vowel perfect, with the present-stem nasal infix dropped. The military sense of fundō is to rout: 'Our soldiers routed the enemy.' Note how much is packed into the recognition: the ending gives the person and tense, the vowel length gives the stem, and the object gives the sense.
(d) nōstī
A syncopated form of nōvistī, second person singular perfect of nōscō, nōscere, nōvī, nōtum. The perfect of nōscō means 'I have got to know' and therefore, in effect, 'I know' - a perfect with present force, like meminī and ōdī. So nōnne haec omnia nōstī? is 'surely you know all this?' Two things had to be undone at once: the syncope and the present force of the perfect.
(e) pepercit and domuit
pepercit is a reduplicated perfect: parcō, parcere, pepercī, parsum, to spare, and it governs the dative - hence victīs, dative plural of the perfect passive participle of vincō used as a noun, 'the conquered'. domuit is a -u- perfect: domō, domāre, domuī, domitum, to subdue, and it governs the accusative - superbōs, 'the proud'. Translate: 'He spared the conquered and subdued the proud.' The two participles used as nouns are a normal Latin economy and should be read as such.

Result: (a) When they had said these things, they departed. (b) The king did not dare to attack the city. (c) Our soldiers routed the enemy. (d) Surely you know all this? (e) He spared the conquered and subdued the proud. The five obstacles: the -ēre third plural twice, a semi-deponent perfect, a syncopated perfect with present force, and a reduplicated perfect governing the dative. Not one of them is an unknown word; every one of them is a known word in an unexpected coat.

Typical mistakes

Reading a third plural perfect in -ēre (dīxēre, fuēre) as a present infinitive, which removes the sentence's main verb.
Treating a syncopated form (amāsse, nōsse, petiērunt) as an unknown word rather than as a contraction of a paradigm you already know.

Active revision

Write out, from memory, all four principal parts of these twenty verbs: agō, capiō, cadō, caedō, canō, cognōscō, currō, dō, dūcō, ferō, faciō, fīō, gerō, iubeō, mittō, pōnō, stō, tangō, tollō, volō. Then classify each perfect by its formation - v, u, s, lengthened vowel, reduplicated, unchanged - and write the third person plural perfect of each in both the -ērunt and the -ēre form. Mark every one you could not do without looking it up, and repeat the exercise a week later with only those.

Active recall

Recall the key points — then reveal.

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

Exam technique: deducing an unfamiliar word#

StandardOCR H443 Component 01 - unseen translation with no defined vocabulary listOCR H443 Component 02 Section A - translation and comprehension on an unseen prose passageOCR H443 Assessment objective AO1 - demonstrate knowledge and understanding of the language

The four questions to ask of an unknown word

Fig. 2The order matters as much as the content. Candidates who guess well are running this procedure quickly; candidates who guess badly are starting at question four. The emphasised node is the step that is skipped most often and pays best: the syntax around the word constrains its meaning more tightly than the story does.

You will meet words you do not know. That is not a failure of preparation; it is the design of the paper, since there is no A-Level vocabulary list and the passages are real. What separates candidates is not how few unknown words they meet but what they do when they meet one. The wrong response is to stop, stare, and eventually guess from the general drift of the paragraph - which produces a plausible-sounding sentence that scores nothing because the mark scheme rewards the sense of the Latin, not the sense of the story. The right response is a fixed procedure, run in a few seconds, that extracts everything the word itself and its surroundings will give before any guessing happens. The four questions in the diagram are that procedure, and they are worth rehearsing until they are automatic, because under examination pressure an unrehearsed method is no method.

Question one: what is it, grammatically? The ending usually answers this before the meaning is in play, and knowing the part of speech is already a large constraint. A word in -ibus is a dative or ablative plural of a third- or fourth-declension noun; a word in -ērunt is a third person plural perfect; a word in -endus is a gerundive; a word in -ntem is an accusative singular present participle. Establish that first and write it in the margin if it helps. A candidate who knows they are looking at an accusative singular noun governed by a particular verb has already narrowed the search enormously, and has also learned something that will earn credit even if the meaning never fully arrives: a translation that puts an unknown word in the right grammatical slot is far more likely to keep the marks attached to the rest of the clause.

Question two: what is the stem, once the prefix and the suffix are stripped off? This is where the word-building work pays. praetermittere is prefix praeter- plus mittere: to send past, hence to omit. incrēdibilis is negative in- plus crēd- plus -ibilis: not able to be believed. inexpugnābilis is in- plus ex- plus pugn- plus -ābilis: not able to be stormed. Strip mechanically and in order - prefix first, then suffix, then look at what is left - and be alert to assimilation, which hides the prefix, and to vowel weakening, which hides the root. A word that survives this treatment and is still unrecognisable is genuinely new, and you now know exactly what you do not know, which is useful information.

Question three: what does the syntax require? The sentence around the word constrains its meaning more tightly than most candidates realise. If the word is the object of a verb of motion, it is a place or a person. If it is the subject of a verb of speaking, it is animate. If it stands in the ablative after a preposition of separation, it is a thing being left. If a verb governs a dative in the sentence, the class of verbs that govern the dative - helping, favouring, trusting, obeying, sparing, commanding, persuading - is a short list, and that list is now the candidate set for your unknown verb. This is a genuinely powerful step, and it is the one candidates skip, going straight from the shape of the word to a guess about its meaning.

Question four: what does the narrative require - and then, having answered it, write something. The context tells you what kind of event would make sense here, and the combination of a grammatical slot, a stripped stem and a narrative expectation will very often produce a translation that is right, or right enough to keep the surrounding marks. Two disciplines govern this last step. Write a real English word, not a blank and not a transliteration: an examiner cannot award anything for a gap, and an untranslated Latin word in an English sentence is a gap with extra steps. And do not let one unknown word damage the clause around it - if you cannot get the participle, translate the main verb, the subject and the object accurately and accept the loss of one mark rather than abandoning five. Move on, and if there is time at the end, come back.

Worked example

Five unknown words, deduced in order

Constructed illustrative passage (not a quotation): cōnsul, cōpiīs ante urbem īnstrūctīs, oppugnātiōnem parābat. mūrus tamen inexpugnābilis vidēbātur, et mīlitēs, longō itinere dēfatīgātī, imperia ducis praetermittere coepērunt. tandem obsidiōnem solvit et exercitum incolumem redūxit. Deduce inexpugnābilis, dēfatīgātī, praetermittere, obsidiōnem and incolumem.

inexpugnābilis
Question one: the ending -is on a word standing after mūrus, with vidēbātur as the verb, makes it a third-declension adjective in the nominative singular, agreeing with mūrus - a complement after the passive of videor, 'seemed'. Question two: strip in- (negative, because this is an adjective), strip -ābilis (able to be), and the stem is expugn-, from expugnō, to storm or take by assault, itself ex- plus pugnō. Question three and four: a wall that seemed not-able-to-be-stormed. Translate 'the wall, however, seemed impregnable'.
dēfatīgātī
Question one: -ātī after a masculine plural noun mīlitēs, set off by commas, is a perfect passive participle, nominative plural masculine, agreeing with mīlitēs. Question two: strip dē- (an intensifier here, 'utterly'), and fatīgātus from fatīgō, to tire out, is left; even if fatīgō is unknown, the phrase longō itinere in the ablative of cause supplies the rest. Question three: a participle agreeing with the soldiers and governed by an ablative of cause. Question four: soldiers, after a long march, are exhausted. Translate 'the soldiers, worn out by the long march'.
praetermittere
Question one: an infinitive, and coepērunt is a verb that takes a prolative infinitive, so this is what the soldiers began to do. Question two: praeter- (past, beyond) plus mittere (to send, to let go): to let go past, to pass over, to neglect. Question three: it governs imperia ducis, the accusative object - so whatever it means, it is done to the leader's orders. Question four: exhausted soldiers begin to do something to their commander's orders. 'They began to disregard the leader's orders.' The stem work alone nearly gets there; the object confirms it.
obsidiōnem
Question one: accusative singular, third declension, object of solvit. Question two: the suffix -iō, -iōnis makes an action noun; the stem obsid- is ob- plus sed-, the weakened form of sedeō, so the underlying image is sitting in someone's way. Question three: the object of solvō, to loosen or release - so something that can be loosened. Question four: after a failed assault, a commander who withdraws his army. A siege. Translate 'at last he raised the siege'. Note that solvere obsidiōnem is an idiom and 'loosened the siege' is not English; the mark scheme wants 'raised' or 'lifted'.
incolumem
Question one: -em on an adjective agreeing with exercitum, accusative singular. Question two: in- negative plus colum-, related to a verb of harming that you may not know. This is the case where step two fails, and that is worth practising too. Question three: an adjective describing the army that the commander led back. Question four: the narrative is a withdrawal after an unsuccessful siege, and the sentence is plainly saying something favourable about the state of the army. 'Safe', 'unharmed', 'intact'. Translate 'and led his army back unharmed'. The guess is constrained enough by steps one, three and four to be worth a mark even though step two produced nothing.

Result: Full translation: The consul, having drawn up his forces before the city, was preparing an assault. The wall, however, seemed impregnable, and the soldiers, worn out by the long march, began to disregard the leader's orders. At last he raised the siege and led his army back unharmed. Four of the five words were fully recoverable from prefix, suffix and stem; the fifth was recovered from its grammatical slot and the narrative alone - which is the case the four-question order is designed for.

Typical mistakes

Leaving a gap or copying the Latin word into the English translation, which can never earn a mark.
Guessing from the general drift of the paragraph before extracting what the ending and the stem will give.

Active revision

Ask someone to take a passage of Latin prose at your level and blank out six words at random with a marker. Translate the passage anyway, and for each blanked word write down, in order, the four answers: what its ending must have been, what the sentence needs it to mean, what part of speech it is, and your best English word. Then uncover the six words and score yourself. You will find that the grammatical predictions are almost always right and the lexical ones are right about half the time - which is exactly the ratio that makes the method worth using.

Active recall

Recall the key points — then reveal.

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

References & sources

Sources

Cambridge OCR

Department for Education

GCE AS and A level subject content for ancient languages

Vocabulary and linguistic range

5 sections~43 min reading time3 competenciesLevel Standard 3 · Advanced 2

Vocabulary without a list: what the specification requires#

Standard

OCR H443 - there is no Defined Vocabulary List at A-Level; learners build on the AS Level Latin (H043) Defined Vocabulary List through wider reading

OCR H443 Component 01 - unseen translation from Livy and Ovid, with no defined vocabulary

DfE GCE AS and A level subject content for ancient languages

Vocabulary and linguistic range

Key points

Building four notebook entries from one sentence

Key points

Unpacking eight compounds of one verb

Key points

One verb, five sentences, five English verbs

Key points

Reading five disguised perfects

Key points

Five unknown words, deduced in order

Sources

Vocabulary and linguistic range

Key points

Building four notebook entries from one sentence

Key points

Unpacking eight compounds of one verb

Key points

One verb, five sentences, five English verbs

Key points

Reading five disguised perfects

Key points

Five unknown words, deduced in order

Sources