Unseen translation

The shape of Paper 1: five tasks, 150 marks, three hours#

FoundationAQA 7677 section 2.2 — Paper 1: written exam, 3 hours, 150 marks, 50 per cent of A-level; three sections, all answered; no access to a dictionaryAQA 7677 section 4.4 — duration 3 hours, total raw mark 150; there is no Defined Vocabulary List for this componentAQA 7677 section 4.4.1 to 4.4.3 — the five tasks and their conditions

Paper 1: the five tasks and where the marks are

Fig. 1The three sections carry exactly 50 marks each. The suggested times are arithmetic, not the board's: 150 marks in 180 minutes is 1.2 minutes per mark, and the allocation shown uses the full three hours, so a working plan shaves a little from each block to hold ten minutes back for checking. The emphasised cell is the composition, the largest single question in the whole qualification.

Paper 1 is the largest object in the qualification: three hours, 150 marks, half the A-level, and every mark on it belonging to AO1. It has three sections and all of them are compulsory. Section A works on unseen prose and is worth 50 marks in three parts — translate a passage of unseen prose into English for 20, answer comprehension and grammar questions on a passage of unseen prose for 25, and point a passage of unseen prose for 5. Section B works on unseen poetry and is worth 50 marks in two parts — translate a passage of unseen poetry into English for 25, and answer comprehension and grammar questions on a passage of unseen poetry for 25. Section C is a single task worth 50: translate a prose passage of approximately 100 words from English into pointed Biblical Hebrew. Five tasks, three equal fifty-mark blocks, and no dictionary in any of them.

Section 4.4 adds detail the section headings do not, and each detail is worth planning around. Within Section A the translation passage and the comprehension passage are two different unseen passages, not one passage used twice; the same is true in Section B. One or two sentences of background information are given in English as an introduction to each Hebrew text. Names of people and places are glossed, and so are words that are very rare or used unusually, if the meaning cannot be derived from the available text. The split between comprehension questions and grammar questions is approximately even, so of the 25 marks in each of those two questions, roughly half go to understanding the passage and roughly half to naming its forms. And the pointing task is described in two places with two measures that agree: approximately 25 words, and approximately two lines of narrative prose.

Because the marks are so evenly distributed, the paper punishes an unbalanced candidate severely. Prose and poetry carry 50 marks each; translation across the two sections carries 45; comprehension and grammar across the two sections carries 50; composition alone carries 50. A candidate who is strong on prose narrative and weak on verse has left a third of the paper in doubt before starting. A candidate who translates well but cannot name a form in the specification's terminology has left another third in doubt. There is no optionality anywhere in the paper to hide a weakness behind, which is unusual: the ancient-language A-levels in other languages often allow a choice between composition and comprehension, and 7677 does not.

Time is the other resource to plan. One hundred and fifty marks in one hundred and eighty minutes is a rate of 1.2 minutes per mark, which gives 24 minutes for the prose translation, 30 for the prose comprehension and grammar, 6 for the pointing, 30 each for the two poetry questions and 60 for the composition. That allocation uses the whole three hours exactly, which is why the sensible plan is to shave a little from each block and hold ten minutes back. Where should the ten minutes go? Almost certainly to the composition, because it is the only task where an error you make is an error you wrote, and because its pointing column rewards a systematic check. The pointing question, at 6 minutes for 5 marks, is the one place where running over time is never worth it.

Two features of the paper repay a deliberate strategy. First, the English introductions are free intelligence and should be read before anything else: they name the participants, situate the scene and often give away the shape of the action, which converts several guesses into inferences. Second, the two comprehension and grammar questions between them are worth 50 marks, exactly as much as the composition, and they are the most answerable part of the paper, because grammar questions have determinate answers that can be learned from section 3.3.2. A candidate who has drilled the terminology can bank a substantial part of those 50 marks with much greater certainty than any translation block offers.

Worked example

Building a defensible time plan for the three hours

You have three hours and 150 marks. Construct a plan that reads the whole paper first, holds ten minutes in reserve, and protects the composition. Show the arithmetic and say what you would sacrifice if you fell twenty minutes behind.

Start from the rate, not from the feel
One hundred and eighty minutes divided by 150 marks is 1.2 minutes per mark. Any block that takes more than 1.2 minutes a mark is being subsidised by another block. Write the proportional figures out once: 24, 30, 6, 30, 30, 60. They sum to 180, which tells you immediately that a proportional plan leaves nothing for reading or checking, and therefore that the real plan must be sub-proportional everywhere.
Take the reserve off the top
Reserve five minutes at the start to read all three sections, including the English introductions and the composition passage, and ten minutes at the end. That leaves 165 minutes of working time, or 1.1 minutes per mark. Scaling the proportional figures gives roughly 22, 27, 5, 27, 27 and 55. Round to 22, 28, 5, 28, 27 and 55, which sums to 165. Reading the composition passage at the start costs nothing and does something useful: it lets the back of your mind work on it while you translate.
Decide the order
There is no rule that you must answer in the printed order, and there are two defensible strategies. The first is to do the composition second, after Section A, on the grounds that it is the biggest block and should not meet you tired. The second is to do it last but with a hard start time written on the page. What is not defensible is leaving the start time to chance, because the composition is the only task where being short of time changes the mark by twenty rather than by two.
Protect the pointing question with a hard stop
The pointing question is 5 marks and about 25 words. Five minutes is the right allocation and it should be enforced absolutely, because the marginal return collapses: the grid gives 5 marks for one to five errors and 4 marks for six to ten, so the difference between a careful five minutes and an agonised fifteen is at most one mark. Point the words you are sure of, make a reasoned attempt at the rest, and stop.
The triage if you are twenty minutes behind
Sacrifice from the block with the lowest marginal return per minute, which is almost never the composition. In practice that means finishing the comprehension and grammar questions in note form rather than in sentences, since those answers are marked for content rather than for prose; abandoning the last-pass check on a translation you have already completed; and, if it comes to it, writing the last two blocks of the composition as unpointed Hebrew, which the grid still credits at up to 2 marks each, rather than leaving them blank at 0.

Result: Five minutes to read everything, then 22 minutes for the prose translation, 28 for the prose comprehension and grammar, 5 for the pointing, 28 and 27 for the two poetry questions and 55 for the composition, with ten minutes held at the end. If twenty minutes are lost, take them from the prose of the comprehension answers and from the final check, never from the composition, and always write something pointed in every composition block.

Typical mistakes

Skipping the English introduction and starting on the Hebrew, which throws away the free context the specification promises.
Treating the pointing question as a puzzle worth solving properly, and spending fifteen minutes on 5 marks while the 50-mark composition waits.
Planning the course as though Section B were an appendix to Section A, when the two carry identical marks and the poetry translation is worth more than the prose one.

Active revision

Take a full past-style paper structure and write out your own time plan on the front of it, block by block, in minutes, with a running clock time beside each. Then sit the paper to that plan with an alarm at each boundary, and record where you overran. Do this three times. The point is not to be fast but to discover which block silently eats your time, because it will be the same block every time and it can only be fixed once you have named it.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 2.2 and 4.4 (AQA) · GCE AS and A level subject content for ancient languages (Department for Education)

Reading biblical narrative prose: the chain and its interruptions#

StandardAQA 7677 section 4.4.1 — students will translate an unseen passage of Biblical Hebrew prose into English; one or two sentences of background information will be given in EnglishAQA 7677 section 3.3.1.3 — vav conjunctive; vav conversive/consecutiveAQA 7677 section 3.3.1.5 — word order; relative, oath and interrogative clauses

Biblical narrative is built out of one construction repeated: the vav consecutive with the imperfect. A paragraph consists of a chain of clauses each beginning וַ plus an imperfect, each carrying the action one step forward, and the English equivalent is a sequence of past-tense main clauses. Reading such a chain is easy once you accept that it is a chain: do not look for subordination that is not there, and do not translate every vav as and, because English tolerates far fewer of them than Hebrew requires. The chain also fixes the word order. Each clause begins with its verb, so the subject arrives second and the object third, which means you cannot know who is acting until you have read past the verb. The discipline that follows is simple and unnatural: read to the end of the clause before you commit to anything.

What is interesting is where the chain breaks, because a break is always meaningful. If a clause begins with something other than a vav consecutive — a subject, an object, a prepositional phrase, or a vav followed by a noun rather than a verb — the narrative has stopped advancing and is doing something else. It may be giving background information, in which case the natural English is a pluperfect or a now-clause; it may be marking a contrast between two participants, in which case English wants a but or an as for; it may be describing a state rather than an event, in which case it will often be a verbless clause or a participle. Recognising the break is worth more than any single item of vocabulary, because it tells you what kind of sentence you are in before you translate a word of it.

Direct speech is the second great feature of the genre and it has its own conventions. Hebrew has no quotation marks, so speech begins immediately after the verb of saying, and it very often begins after the infinitive construct לֵאמֹר, literally to say, which functions as a spoken colon and should not be translated at all in modern English. Speech may be nested — a messenger reports what he was told to say — and the pronouns inside it refer to the speaker and hearer of the inner speech, not the outer. A common source of confusion is that a chain of speech verbs continues inside the reported words, so that a candidate reading mechanically can end up attributing a speech to the wrong character. The safeguard is to mark, in pencil, where each speech opens and closes before translating any of it.

Three grammatical habits of the genre should be automatic by the examination. First, the formula וַיְהִי, and it came to pass, introduces a time setting and is usually followed by a temporal expression and then by the main clause; English can drop it entirely or render it as and it happened that. Second, the object marker אֵת, usually written אֶת with a maqaf, precedes a definite direct object and is not translated: seeing it tells you that what follows is the object and not the subject, which resolves the word-order ambiguity Hebrew otherwise leaves open. Third, the definite article and the relative particle אֲשֶׁר do the work that English does with the, who, which and that, and אֲשֶׁר is invariable, so the English relative pronoun has to be chosen from the sense of the clause rather than read off the Hebrew.

Finally, prose narrative gives you a great deal of redundancy and you should use it. Names are repeated where English would use a pronoun; the same verb often recurs at the start of consecutive clauses; the narrative frequently states an intention and then reports its fulfilment in almost the same words. That redundancy is a gift to a reader working without a dictionary, because a word you could not parse in one clause may reappear parsed for you in the next. It also has a literary function that Papers 2 and 3 will ask about, since the repetition of a command in its execution, or the pointed failure to repeat it, is one of the ways biblical narrative makes a judgement without stating one. In the language paper you use the redundancy to read; in the literature papers you use the same observation to argue.

Worked example

Mapping a narrative period before translating it

Constructed teaching Hebrew, not a quotation. Map the structure of the following before translating: mark the vav consecutive chain, the breaks in it, the direct speech and its boundaries. וַיְהִי כַּאֲשֶׁר שָׁמַע הַכֹּהֵן אֶת־הַדְּבָרִים הָאֵלֶּה וַיִּקְרָא אֶל־הַנַּעַר לֵאמֹר לֵךְ וּקְרָא אֶת־כָּל־הָעָם אֲשֶׁר בָּעִיר׃ וַיֵּלֶךְ הַנַּעַר וַיַּעַשׂ כַּאֲשֶׁר צִוָּה אֹתוֹ הַכֹּהֵן׃

The temporal frame
וַיְהִי is the frame-setting formula and is not itself the main clause. What follows it, כַּאֲשֶׁר שָׁמַע הַכֹּהֵן אֶת־הַדְּבָרִים הָאֵלֶּה, is a temporal clause introduced by כַּאֲשֶׁר, when: the verb שָׁמַע is a qal perfect third masculine singular, the subject הַכֹּהֵן, the priest, follows it, and the object is marked by אֶת and is definite. So the shape is: it came to pass, when the priest heard these words. English will normally drop the formula and begin at when.
The main clause and the start of speech
וַיִּקְרָא אֶל־הַנַּעַר is the first link of the chain proper: vav consecutive with the imperfect of קרא, and he called, with the addressee marked by אֶל. Then לֵאמֹר, the infinitive construct of אמר, opens the direct speech. Do not translate it; mark it as a colon and open your quotation there. Everything from לֵךְ to הָעִיר is inside the speech, and the imperative לֵךְ, go, is the proof, since an imperative can only be spoken.
Inside the speech
לֵךְ וּקְרָא is two imperatives joined by a vav conjunctive, go and call. The object is אֶת־כָּל־הָעָם, marked as definite by both the object marker and the article, and then אֲשֶׁר בָּעִיר is a relative clause with no verb: literally who in the city, which English must complete as who are in the city. Two of the specification's listed constructions therefore sit inside a five-word command: a relative clause and a verbless clause. The speech ends at הָעִיר and the sof pasuq closes the verse.
The second sentence: execution reported
וַיֵּלֶךְ הַנַּעַר וַיַּעַשׂ resumes the chain with two more vav consecutive clauses, and he went and he did. Note the redundancy the genre offers: the same verb הלך that was commanded is now reported, which confirms your reading of the imperative. Then כַּאֲשֶׁר צִוָּה אֹתוֹ הַכֹּהֵן is a second temporal or comparative clause, as the priest had commanded him, with צִוָּה a piel perfect and אֹתוֹ the object marker with a third masculine singular suffix.
Turning the map into English
The map now dictates the translation. Two sentences, four narrative clauses, one embedded speech containing two imperatives, and two subordinate temporal clauses. English wants: a subordinate clause, a main clause, a quotation, then a sentence with two verbs and a subordinate clause. The vavs at the head of the narrative clauses become full stops or ands as English taste requires, not one and each. That is what the specification means by idiomatic English which accurately transfers the meaning.

Result: And it came to pass, when the priest heard these words, that he called to the young man and said, Go and call all the people who are in the city. And the young man went and did as the priest had commanded him.

Typical mistakes

Translating every vav as and, producing an English paragraph of thirty coordinated clauses that reads as a bad translation and can lose the idiomatic-English credit.
Committing to a subject before reading past the verb, and then having to rebuild the clause when the real subject arrives.
Missing the boundary of a piece of direct speech, so that the narrator's words are attributed to a character or the reverse.

Active revision

Take a chapter of set text 1 and mark up its structure before translating any of it. Rule a vertical line down the margin and mark every clause that begins with a vav consecutive; then mark differently every clause that begins with anything else, and beside each write what you think it is doing — background, contrast, description, speech. Bracket every stretch of direct speech and label the speaker. Only then translate, and see how many decisions the mark-up had already made for you.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, section 4.4.1 (AQA)

Translating unseen prose: the method and the mark scheme#

StandardAQA 7677 section 4.4.1 — students should be able to translate unprepared passages into idiomatic English which accurately transfers the meaning of the Biblical Hebrew, while showing an awareness of its morphology and syntaxAQA 7677 section 4.7.1 — translation passages will be divided into sections, each worth 5 marks; the general principle in assessing each section will be the proportion of sense achieved

The mark scheme decides the method, so read it first. Translation passages are divided into sections each worth 5 marks, and each section is marked on the proportion of sense achieved out of 5: 5 for an accurate translation with one slight error allowed, 4 for mostly correct, 3 for more than half right, 2 for less than half right, 1 for little recognisable relation or meaning to the Hebrew, and 0 for no response or none worthy of credit. Three consequences follow immediately. The unit of scoring is the block, not the word, so a block in which the syntax is right and one noun is wrong scores 4 while a block in which every word is glossed but the clause structure is wrong may score 2. Nothing rewards a blank. And there is no separate credit for elegance, so time spent polishing English that is already accurate is time taken from a block that is not.

The method that fits that scheme has four passes and takes less time than it sounds. The first pass is structural and touches no vocabulary: read the English introduction, then read the Hebrew through marking clause boundaries, the vav consecutive chain, the breaks in it, and the boundaries of any direct speech. The second pass is grammatical: for each clause identify the verb and parse it, then find the subject and the object, using the object marker where it appears. The third pass is lexical and is the only one where you should be looking up anything in your memory: gloss the words you know, and deduce the ones you do not by the method of stripping affixes and finding radicals. The fourth pass is the English one: write the whole thing out as connected English prose, not as a gloss.

Idiomatic English is explicitly required and it has a small number of recurring moves. Hebrew coordinates where English subordinates, so a chain of vav consecutive clauses becomes a mixture of main clauses and subordinate ones. Hebrew repeats a proper name where English uses a pronoun, so the second and third mentions should usually become he or the king. Hebrew states an infinitive construct with a preposition where English uses a when-clause or a by-clause. Hebrew uses a construct chain where English often prefers an adjective, so a phrase of the form the man of God is better rendered by the standard English equivalent than by a literal genitive. And Hebrew has a much higher tolerance of the word and, so most of them should disappear.

Set against that is the requirement, in the same sentence of the specification, that the translation show an awareness of the Hebrew's morphology and syntax. Idiomatic does not mean loose. If the Hebrew has a hiphil, the English should show a causative; if the Hebrew has a niphal, the English should show a passive or a reflexive; if the Hebrew has an infinitive absolute reinforcing its verb, the English should show the emphasis with a word like surely or indeed. The line to hold is that every grammatical feature of the Hebrew should be visible somewhere in the English, and that no feature of the English should be invented that has no source in the Hebrew. Adding an explanatory phrase because the sentence seems to need one is the commonest way of drifting from translation into paraphrase, and it costs marks in a scheme that measures the proportion of sense achieved against the original.

Two rules govern what to do when you are stuck. First, never write two alternatives separated by an oblique stroke. An examiner marking the proportion of sense achieved must decide what you have said, and a hedged answer reads as the weaker of the two options. Choose, write, and move on. Second, if a whole clause defeats you, translate what you can identify and leave the structure visible: writing and the king something the people preserves the subject, the object and the clause type, and under this grid a clause with a missing verb may still be more than half right. Both rules are applications of the same principle, which is that the scheme rewards sense conveyed and never punishes an honest attempt more than it punishes silence.

Worked example

A prose translation worked clause by clause and then marked

Constructed teaching Hebrew, not a quotation. Translate, working clause by clause and justifying each decision, then divide the result into 5-mark blocks and mark it. וַיְהִי כַּאֲשֶׁר שָׁמַע הַכֹּהֵן אֶת־הַדְּבָרִים הָאֵלֶּה וַיִּקְרָא אֶל־הַנַּעַר לֵאמֹר לֵךְ וּקְרָא אֶת־כָּל־הָעָם אֲשֶׁר בָּעִיר׃

Clause one: the temporal frame
וַיְהִי is the frame formula and carries no content of its own. כַּאֲשֶׁר שָׁמַע הַכֹּהֵן is a temporal clause: כַּאֲשֶׁר, when; שָׁמַע, qal perfect third masculine singular of שמע, he heard; הַכֹּהֵן, the priest, definite, subject. The perfect after כַּאֲשֶׁר is past. English: when the priest heard. Note that the subject follows the verb, so a reader who committed to a subject at שָׁמַע would have had to rebuild.
The object of the temporal clause
אֶת־הַדְּבָרִים הָאֵלֶּה. The object marker אֶת tells you at once that what follows is the direct object, which removes any ambiguity about whether the priest is hearing or being heard. הַדְּבָרִים is the masculine plural of דָּבָר with the definite article, and הָאֵלֶּה is the demonstrative agreeing with it in number and definiteness. דָּבָר means word, matter or thing; the plural with a demonstrative after a verb of hearing points firmly to words. English: these words.
Clause two: the main narrative clause
וַיִּקְרָא אֶל־הַנַּעַר לֵאמֹר. The vav consecutive resumes the chain and makes this the main clause of the sentence: and he called. The preposition אֶל marks the addressee, הַנַּעַר, the young man or the lad. לֵאמֹר is the infinitive construct of אמר used as a speech marker; it is not translated, and it opens the quotation. English: he called to the young man and said, or simply he called to the young man, followed by a quotation. Both are accurate; the second is more idiomatic.
Clause three: the speech itself
לֵךְ וּקְרָא is two masculine singular imperatives, go and call, the first from הלך and the second from קרא, joined by a vav conjunctive written as shuruk because the qof carries a vocal sheva. The object is אֶת־כָּל־הָעָם, all the people, definite twice over. אֲשֶׁר בָּעִיר is a relative clause with no verb: who are in the city, where the preposition has absorbed the article. English: Go and call all the people who are in the city.
Dividing and marking
The passage is about twenty-two words and would be divided into perhaps two 5-mark blocks. Block one runs to הָאֵלֶּה and block two to the end. Suppose a candidate wrote: And it happened when the priest heard these words, and he called to the boy saying, Walk and read all the people who are in the city. Block one is accurate with one slight infelicity and scores 5. Block two has two lexical errors, walk for go and read for call, both from correct roots with the wrong sense chosen, but the syntax is entirely right; it is mostly correct and scores 4. Nine out of ten, and the two errors are of exactly the kind the semicolon convention in the vocabulary list warns about.

Result: And it came to pass, when the priest heard these words, that he called to the young man and said, Go and call all the people who are in the city. Marked in two 5-mark blocks, the structure carries the marks: a candidate who chose the wrong sense of a known root in two places still scores 9 out of 10, because the clause architecture is right.

Typical mistakes

Translating word by word from the left of the line rather than clause by clause, which in a verb-first language guarantees a rebuild halfway through every sentence.
Producing a literal English that keeps every Hebrew and, which the specification's requirement of idiomatic English explicitly counts against.
Padding a translation with explanatory phrases that have no counterpart in the Hebrew, which turns translation into paraphrase and reduces the proportion of sense credited.

Active revision

Take a passage of unseen prose of about eighty words and translate it under timed conditions using the four-pass method, writing nothing in English until the fourth pass. Then divide your own translation into 5-mark blocks of roughly ten words each, exactly as the mark scheme would, and mark it yourself against the six descriptors. Count how many blocks lost marks for a wrong word and how many for a wrong structure. If the second number is larger, your problem is the first pass, not your vocabulary.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 4.4.1 and 4.7.1 (AQA)

The pointing question: 5 marks, about twenty-five words#

AdvancedAQA 7677 section 2.2 — in Section A students will point a passage of unseen prose (5 marks)AQA 7677 section 3.1 — students will be required to point approximately two lines of narrative proseAQA 7677 section 4.4.1 — students will also be required to point approximately 25 words of Biblical Hebrew proseAQA 7677 section 4.7.2 — the number of pointing errors will be added up and marks awarded according to the grid

How the pointing question is marked

Fig. 2Errors are counted first and converted afterwards, which is why the question rewards partial knowledge so well: five errors still scores full marks and ten still scores four. The third column expands the specification's four rules into the five ways a single point can go wrong. The emphasised cell is the band that matters most in practice, because the gap between five errors and six is the only place on the grid where one point costs a whole mark.

Nothing else in any ancient-language A-level in England resembles this question, so it is worth stating exactly what it is. You are given approximately twenty-five words of unpointed Biblical Hebrew narrative prose — the specification also describes it as approximately two lines — and you must supply the pointing. It is worth 5 marks out of 150, which is 3.3 per cent of Paper 1 and 1.7 per cent of the A-level. That is a small number, and it should govern how much time the question gets; but the knowledge it tests is not small, because the same knowledge underwrites the pointing column of the 50-mark composition, and there it is worth a great deal.

The marking is unusual and generous, and knowing how it works changes how you attempt the question. Errors are counted and then converted by a grid: one to five errors scores 5, six to ten scores 4, eleven to twenty scores 3, twenty-one to thirty scores 2, thirty-one to forty-five scores 1, and more than forty-five scores 0. Note what that means. You are allowed five errors for full marks. A passage of twenty-five words contains somewhere between eighty and a hundred and twenty separate points, so full marks require roughly ninety-five per cent accuracy, but four marks require only about ninety per cent and three marks about eighty-five. The grid widens steeply, which means that a candidate who knows the system reasonably well but not perfectly still scores well, and a candidate who leaves words unpointed does not.

The specification defines what counts as one error and there are four rules. An incorrect vowel is one error. A vowel on a letter which should be unpointed is one error. A missing vowel or dagesh is one error. And a dagesh in a letter which should not have one is one error. Two things follow. First, every point is counted separately, so a word with four vowels and a dagesh has five opportunities to be wrong; over-pointing is punished exactly as heavily as under-pointing. Second, there is no rule about consonants, because the consonants are given: this is not a spelling test. A candidate who spends the five minutes worrying about the letters has misread the question.

The method that fits the marking is to parse before pointing. You cannot point a word until you know what it is, because the vowels are the morphology: a form you have identified as a qal imperfect has one vowel pattern and a niphal perfect has another, and guessing the pattern without the identification is guessing five points at once. So work word by word in three moves. First, identify the word class and, for a verb, the binyan, aspect, person, number and gender. Second, write the vowels the pattern requires. Third, apply the environmental rules — a guttural takes a hataf vowel where a plain sheva would be expected, a begadkephat letter at the start of a word or after a silent sheva takes a weak dagesh, the definite article doubles what follows it, a word at a major pause takes its pausal form.

Two pieces of strategy follow from the grid. The first is that partial pointing is worth having: a word you point correctly costs nothing even if the next word defeats you, and a word left unpointed contributes an error for every vowel it should have had, which is typically two or three. Point everything, even where you are unsure. The second is that the highest-value words to get right are the long ones and the ones with articles and prefixes, because those carry the most points: a four-syllable verb with a vav consecutive and a suffix may carry six or seven separate marks of pointing. Working through the passage from the words you are most confident about, rather than strictly left to right, is a legitimate and efficient strategy in a question marked by counting.

Worked example

Estimating your own mark before the examiner does

You have pointed a twenty-five-word passage and you have four minutes left. Rather than re-reading it hopefully, audit it systematically for the error classes the specification names. What are the five checks, in order of return, and what mark would each recover?

Check one: gutturals and resh with a dagesh
Scan the passage for alef, he, het, ayin and resh, and confirm that none of them carries a dagesh. This is the cheapest check on the paper because it is purely mechanical, requires no parsing, and every hit is a guaranteed error under rule four. It also often reveals a second error, because a candidate who put a dagesh in a guttural usually also failed to lengthen the preceding vowel, and so has two errors in one word.
Check two: the definite article
Find every definite article and confirm three things: that it carries the right vowel for the letter that follows — patah with a dagesh normally, qamats before alef, ayin and resh, patah without a dagesh before he and het, segol before an unaccented guttural with qamats — and that the doubling is present where it should be and absent where it cannot be. Articles are frequent, each carries two points, and the rules are short enough to be checked exhaustively in under a minute.
Check three: shevas under gutturals
Scan for any guttural carrying a plain sheva. There should be none: a guttural takes a hataf vowel where a plain vocal sheva would be expected. This is the single most frequent error class in candidates' pointing, and it usually recurs, so finding one instance is worth re-scanning for the rest. Each instance is one error under rule one, and in some words it drags a second error after it in the preceding vowel.
Check four: begadkephat letters
Find every bet, gimel, dalet, kaf, pe and tav and apply the two-question test: is the letter preceded by a vowel? If yes, any dagesh in it must be strong and must be justified by a pattern, an article, a vav consecutive or an assimilated nun. If no — that is, at the start of a word or after a silent sheva — it should carry a weak dagesh. Missing weak dageshes at the start of words are common and each is one error under rule three.
Check five: the pause
Identify the word at the end of the passage and, if the passage spans a verse division, the word at the main internal pause. Those words should carry pausal forms, which typically means a lengthened stressed vowel. This check is last because it is the hardest to be sure of and because it usually affects only one or two words; but each pausal form is worth one or two points and, unlike the other checks, it is the kind of detail that separates a five-error attempt from a three-error one.

Result: Five mechanical checks, in order of certainty and speed: no dagesh in a guttural or resh; every article correctly pointed for its following letter; no plain sheva under a guttural; every begadkephat letter tested for weak against strong; pausal forms at the pauses. In a twenty-five-word passage these five checks routinely find between three and eight errors, which is exactly the range that moves a mark from three to four or from four to five.

Typical mistakes

Pointing before parsing, which means guessing a whole vowel pattern rather than deriving it from an identified form.
Leaving a hard word unpointed, which converts an uncertain two or three errors into a certain two or three errors with no chance of being right.
Adding a dagesh to a guttural or to resh, which is always an error and is one of the few that can be avoided by a mechanical check.

Active revision

Take a passage of set text 1 that you have already translated, copy out its consonants without any points, put it away for a week, and then point it from cold. Mark your own attempt against the printed text using the specification's four error rules and count. Convert the count to a mark on the 4.7.2 grid. Repeat weekly with a new passage and plot the error count: what you are looking for is not perfection but a downward trend, because the grid rewards improvement long before it requires mastery.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 4.4.1 and 4.7.2 (AQA)

Pointing a passage, decision by decision#

AdvancedAQA 7677 section 3.3.1.4 — syllables and vowels including sheva and hataf vowels; weak and strong dagesh; the effect of the guttural and labial letters on pointing; the effect of the accents on pointing; pausal formAQA 7677 section 4.7.2 — the four rules by which pointing errors are counted

The pointing question is a writing task disguised as a reading task, and the only way to learn it is to do it slowly enough to say why each mark goes where it does. The order of work is fixed: identify, then point the pattern, then apply the environment. Identification comes first because the vowels are the morphology and cannot be guessed independently of it. Pointing the pattern means writing the vowels that the identified form requires in a strong verb or a regular noun. Applying the environment means adjusting for the letters actually present — gutturals, resh, the begadkephat six, the labials — and for the word's position in the clause, which is where pausal forms come in.

The environmental rules are few enough to list and should be committed to memory as a checklist. Gutturals and resh never take a dagesh, and where a pattern requires doubling the preceding vowel usually lengthens to compensate. Gutturals prefer a-class vowels and take hataf vowels in place of a plain vocal sheva. The six begadkephat letters take a weak dagesh at the beginning of a word and after a silent sheva, and no dagesh after a vowel. The definite article carries patah with a following dagesh, except that it is qamats before alef, ayin and resh, patah without a dagesh before he and het, and segol before an unaccented guttural pointed with qamats. The inseparable prepositions take hiriq before a vocal sheva and the matching short vowel before a hataf. The vav conjunctive becomes shuruk before the labials bet, mem and pe and before any vocal sheva.

Segolate nouns deserve their own note because they are frequent and because their pointing looks irregular until you see the rule. A segolate is a noun stressed on the first syllable with a helping vowel in the second, and its two commonest patterns are two segols, as in מֶלֶךְ, king, and a segol with a preceding tsere, as in סֵפֶר, book. The class also includes nouns whose helping vowel is patah because the second or third letter is a guttural. In the plural the stress moves to the ending and the first vowel reduces to a vocal sheva, so מֶלֶךְ becomes מְלָכִים and סֵפֶר becomes סְפָרִים. Recognising a segolate is therefore worth two points of pointing in the singular and three in the plural, and there are several in any twenty-five words of narrative.

Verbs in a vav consecutive chain are the other high-value target, because a chain of them is what narrative prose is made of and because each carries a predictable cluster of points. The vav takes patah and the following prefix letter takes a strong dagesh, except where that letter is a yod carrying a vocal sheva, in which case no dagesh is possible and the classic form is וַיְהִי. The prefix vowel is then hiriq in the qal and patah in the hiphil; the stem vowel is holam in the qal imperfect of a strong verb; and in many forms the stress is drawn back towards the beginning of the word, which shortens the final vowel. That last effect is why the vav consecutive form of a hollow verb is וַיָּקָם with qamats rather than the shuruk of the plain imperfect יָקוּם.

Two habits keep the error count down. The first is to point every word, however uncertain, because an unpointed word contributes an error for every point it is missing and cannot possibly contribute a correct one. The second is to reserve the last minute for the mechanical checks that need no parsing at all: no dagesh in a guttural or resh, no plain sheva under a guttural, a weak dagesh on every word-initial begadkephat letter, and an article correctly pointed for whatever follows it. Those four scans take under a minute in a twenty-five-word passage and routinely find three or four errors, which on the specification's grid is often the difference between four marks and five.

Worked example

Pointing nine words and justifying every mark

Constructed teaching Hebrew, not a quotation. The consonantal text is given without points, as the examination would give it. Point it, saying for each word what you identified it as and which environmental rules applied. ויאמר המלך אל עבדיו לכו ושמרו את הדרך אשר בין הערים

Word 1: the narrative verb
Four consonants after the vav, and the sequence is the commonest opening in the language: a vav consecutive with the qal imperfect of אמר, third masculine singular. Point the vav with patah and double the yod with a strong dagesh; the yod then takes holam as the stem vowel of this irregular pe-alef verb, the alef is quiescent and takes no vowel at all, the mem takes segol and the resh closes the word. Result: וַיֹּאמֶר. Five points on one word, and note that the alef being unpointed is itself a decision: a vowel on it would be an error under rule two.
Word 2: article plus segolate
Identified as the definite article plus מלך, king, a segolate noun in the singular absolute. The article takes patah and doubles the mem with a strong dagesh; the segolate pattern gives segol under the mem and segol under the lamed, and the final kaf takes a silent sheva. Result: הַמֶּלֶךְ. Note that the kaf takes no dagesh: it follows a vowel, so the weak dagesh that a word-initial kaf would take is excluded here.
Words 3 and 4: preposition and suffixed plural
אל before a maqaf is the preposition to, pointed with segol rather than the tsere it takes when it stands alone, because a word joined by maqaf loses its own stress. עבדיו is the masculine plural of עֶבֶד, servant, with the third masculine singular suffix, his servants. The ayin is a guttural and takes hataf patah where a plain vocal sheva would otherwise stand; the bet takes qamats, the dalet qamats, and the yod and vav form the suffix. Result: אֶל־עֲבָדָיו.
Words 5 and 6: two imperatives
לכו is the masculine plural imperative of הלך, go, which behaves as a pe-yod verb and loses its initial he: lamed with vocal sheva, kaf with shuruk. Result: לְכוּ. ושמרו is the vav conjunctive plus the masculine plural imperative of שמר: the imperative is שִׁמְרוּ with hiriq, and the vav takes a plain sheva because shin is neither a labial nor a letter carrying a vocal sheva. Result: וְשִׁמְרוּ. A candidate who wrote shuruk on this vav would incur one error under rule one.
Words 7 to 9: object, relative and the guttural article
את before a maqaf is the object marker with segol. הדרך is the article plus the segolate דֶּרֶךְ, way: patah on the he, strong dagesh in the dalet — strong, though dalet is a begadkephat letter, because it follows the article's vowel — then segol, segol and a silent sheva. אשר is the relative particle: the alef is a guttural and takes hataf patah, then segol under the shin. בין is the preposition between: the bet begins the word and therefore takes a weak dagesh, then tsere, then yod and a final nun. הערים is the article plus the plural of עִיר, city; because the ayin is a guttural pointed with qamats and unaccented, the article takes segol and there is no dagesh. Result: אֶת־הַדֶּרֶךְ אֲשֶׁר בֵּין הֶעָרִים.

Result: וַיֹּאמֶר הַמֶּלֶךְ אֶל־עֲבָדָיו לְכוּ וְשִׁמְרוּ אֶת־הַדֶּרֶךְ אֲשֶׁר בֵּין הֶעָרִים — and the king said to his servants, Go and guard the road which is between the cities. Nine words, about thirty separate points, three strong dageshes, one weak dagesh, two hataf vowels required by gutturals, and one article that takes segol rather than patah because of the guttural after it. A real examination passage is about twenty-five words, so roughly three times this.

Typical mistakes

Pointing the first syllable of a segolate as though the stress were on the second, which produces the plural pattern in a singular noun.
Putting a strong dagesh in the yod of the vav consecutive form of the verb to be, where the vocal sheva makes doubling impossible.
Running out of time and leaving the last words unpointed, which is the single most expensive thing that can happen in this question.

Active revision

Copy out fifteen words of narrative prose without points and point them under a strict five-minute limit, writing beside each word a two-word note of what you identified it as. Then check against the printed text and, for every error, decide whether it was an identification error or an environment error. Identification errors are cured by grammar practice; environment errors are cured by the checklist. Most candidates find that after a month the identification errors have gone and the environment errors have not, which tells you where the remaining marks are.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 3.3.1.4 and 4.7.2 (AQA)

Reading biblical poetry: parallelism, terseness and word order#

AdvancedAQA 7677 section 4.4.2 — in Section B students will translate an unseen passage of Biblical Hebrew poetry into English and respond to comprehension and grammar questions on a second unseen poetry passageAQA 7677 section 3.3.2.5 — parallelism, pausal form, subject, object, ellipsis

What the two halves of Paper 1 share and what each demands alone

Fig. 3The two sections carry identical marks and rest on the same grammar list, but they require different reading habits: narrative is organised by a chain of clauses and is generous with signposts, while verse is organised by pairs of lines and omits them. Only Section A carries the pointing question. The regions are deliberately short, because the figure engine truncates a region label past about twenty-six characters.

Section B is worth 50 marks, exactly as much as Section A, and its translation question is worth 25 against Section A's 20. Yet it is the half of the paper candidates most often under-prepare, because poetry cannot be read with the habits that work on narrative. Three things change at once. The vav consecutive chain that organises a prose paragraph is largely absent, so the clauses do not tell you their order. The prose signposts — the definite article, the relative particle and the object marker — appear far less often, so the boundaries of phrases are less explicit. And the line is short, so there is less context around any given word. Against these losses stands one great gain, and the whole method of reading verse is built on it: the lines come in pairs, and the second line tells you about the first.

Parallelism, which the specification lists as a term you must be able to use, is the organising principle of biblical verse. A verse typically consists of two short lines, or cola, in which the second stands in a definite relation to the first. Where the second restates the first in different words, the relation is usually called synonymous; where it states the opposite, antithetical; where it carries the thought forward rather than repeating it, the older textbooks call it synthetic, though many modern scholars prefer to describe the second line as sharpening, specifying or intensifying the first rather than merely completing it. The taxonomy is worth knowing and worth attributing rather than asserting, because it is a description scholars have proposed and refined rather than a feature the text labels. What is not in dispute is the pairing itself, and the pairing is what a reader uses.

Terseness is the second feature and it is what makes verse hard. A poetic line uses fewer words than a prose clause would to say the same thing, partly by omitting the particles and partly by ellipsis — a word stated in the first colon and not repeated in the second, which the reader must carry over. Ellipsis is a listed term and is examinable: a grammar question may show you a colon with no verb and ask what supplies it. The reader's discipline is therefore to count the elements in each colon and, when the second has fewer, to look back rather than to invent. The commonest ellipsis is of the verb; the next commonest is of a preposition, and the next of the subject.

Word order is freer in verse than in prose, and the freedom is used for effect. A poet may put the object first, or the adverbial phrase first, or split a construct chain across a line break; a common pattern places the elements of the second colon in the reverse order of the first, so that the pair forms a chiasmus. Because the verb-first default is not reliable, a reader must identify the verb by its form rather than by its position, which is another reason the accidence is the foundation of everything. Poetry also uses short and archaic forms — pronouns and verb endings that a prose narrative would spell out in full — and prefers pronominal suffixes to separate words, so a poetic line packs more grammar into fewer letters.

The last feature to expect is the pausal form, which is far commoner in verse than in prose because verse lines end at pauses by definition. A word in pause typically lengthens its stressed vowel and may move its stress, so that a form which looks unfamiliar at the end of a line may be a word you know perfectly well in its contextual dress. The specification lists pausal form as a term, with in pause accepted. Two practical uses follow. In reading, a pausal form marks where the line divides, which is free structural information in a text printed without line breaks. In answering a grammar question, naming a pausal form is a precise and rewardable observation of the kind the comprehension and grammar questions on Section B are designed to elicit.

Worked example

Taking a couplet apart before translating it

Constructed teaching Hebrew in the shape of a biblical verse line, not a quotation. Identify the parallelism, find the ellipsis, name any pausal forms, and only then translate. רָחֲקוּ מִמֶּנִּי אֹהֲבָי · וּמִבֵּיתִי רֵעָי

Count the elements in each colon
The first colon has three: a verb, a prepositional phrase and a subject. The second has two: a prepositional phrase and a subject. The mismatch is the signal. Since the second colon has no verb, and since a verb cannot simply be absent from a clause, the verb of the first colon is elided and must be carried over. That is ellipsis, and it is the specification's own term.
Name the parallelism
The two subjects, אֹהֲבָי and רֵעָי, are near-synonyms: those who love me and my companions. The two prepositional phrases, מִמֶּנִּי and וּמִבֵּיתִי, are not synonyms but are of the same kind, both expressing separation from something. So the relation is synonymous, with the second colon specifying rather than merely repeating: not only from me but from my house. Naming it synonymous is defensible and is the answer a grammar question would want; adding that the second colon narrows the first is the observation that would earn credit in a literary question.
Parse the verb and the pausal forms
רָחֲקוּ is the qal perfect third common plural of רחק, to be far or to go far away; the hataf patah under the het is the guttural rule at work, since a plain vocal sheva is impossible there. Both nouns are at line ends and both are in pause: the ordinary contextual forms would carry patah before the final yod of the first person singular suffix, and in pause that patah lengthens to qamats. Recognising that is what stops a reader from treating either word as unfamiliar.
Parse the prepositional phrases
מִמֶּנִּי is the preposition מִן with the first person singular suffix, from me, and its doubled nun is the assimilation the preposition always produces. וּמִבֵּיתִי is three morphemes: the vav conjunctive, written as shuruk because it stands before a mem; the preposition מִן, whose nun has assimilated into the bet and left a strong dagesh; and בַּיִת, house, in its suffixed form with the first person singular suffix. And from my house.
Translate as a couplet, not as two sentences
English should preserve the pairing, which means keeping the two lines and letting the elided verb do its work rather than repeating it woodenly. Those who love me have gone far from me, and my companions from my house. A translation that supplies the verb twice is not wrong but is less faithful to the shape; a translation that runs the two lines into one sentence loses the parallelism the comprehension question will ask about.

Result: A synonymous couplet with the verb elided in the second colon and both line-final nouns in pausal form. Those who love me have gone far from me, and my companions from my house. The ellipsis, the parallelism and the two pausal forms are each separately nameable in the terminology of section 3.3.2.5.

Typical mistakes

Reading a poetic line as though it were a prose clause and looking for an object marker or a relative particle that the genre routinely omits.
Inventing a verb to fill a colon that is elliptical, instead of carrying over the verb from the parallel line.
Treating a pausal form as an unknown word, which converts a familiar item into an obstacle at the very point where the line ends.

Active revision

Take twenty verses of set text 3 and twenty of set text 4 and write each colon on its own line, so the parallelism is visible. For each pair, name the relation between the cola in one word and say what evidence you used. Then mark every ellipsis and write in the carried-over word in brackets. Finally, list every word that appears in pausal form and write its contextual form beside it. Do this often enough and the shape of a verse becomes visible before you have read a word of it.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 3.3.2.5 and 4.4.2 (AQA)

Translating unseen poetry: 25 marks, and what the English must show#

AdvancedAQA 7677 section 2.2 — in Section B students will translate a passage of unseen poetry into English (25 marks)AQA 7677 section 4.4.2 — students should be able to translate unprepared passages into idiomatic English which accurately transfers the meaning, while showing an awareness of morphology and syntaxAQA 7677 section 4.7.1 — translation passages are divided into sections each worth 5 marks, marked on the proportion of sense achieved

The poetry translation is worth 25 marks, five more than the prose translation, and it is marked on exactly the same grid: blocks of 5 marks, scored on the proportion of sense achieved. The blocks in a poetry passage will usually correspond to verses or to pairs of cola, which means that the unit of scoring is the same as the unit of composition, and that a candidate who reads couplet by couplet is already working in the right units. That alignment is worth exploiting. Translate a whole couplet before writing any of it, because the second colon may change your reading of the first; and write the couplet out as a couplet, on two lines, because doing so makes the parallelism visible to the examiner as well as to you.

The English should show the poetry, but showing it does not mean decorating it. Four moves cover almost everything. Preserve the line division, because it is a feature of the original and costs nothing. Preserve the parallelism by using different English words for the different Hebrew words, rather than collapsing a synonymous pair into one term repeated. Supply an elided word where English requires it, but supply the word the Hebrew elided rather than a livelier substitute. And keep the order of ideas within the line where English allows, because a fronted object or a chiastic arrangement is part of what the poet did. What you should not do is add adjectives, intensifiers or connectives that have no source in the Hebrew, however much they improve the rhythm: the mark scheme measures sense conveyed against the original.

Certain features of poetic grammar need a decision rather than a rendering, and it is better to make the decision consciously. Poetry uses the perfect and imperfect with a much looser relation to time than narrative does, so a perfect may describe a general truth, a completed event, or an event so certain that it is spoken of as past. Where the context does not settle it, choose the reading that fits the argument of the passage as a whole and stay consistent. Poetry also uses short and archaic forms whose person or number can be ambiguous, and it uses pronominal suffixes heavily; when a suffix could refer to two different antecedents, choose the one the parallelism supports. In each case, a consistent reading of the whole passage will score better than a series of locally plausible guesses that contradict one another.

Vocabulary is the practical difficulty and the method for it is the one from Section A applied harder. Strip affixes, find radicals, name the binyan, and then let the parallel colon narrow the sense. Because verse is terse, the parallel is often the only context available, and using it is not a workaround but the intended route: a poet who states something twice has given the reader two chances at it. Where both cola contain unknown words, fall back on structural translation — write the clause with its syntax intact and its unknown elements rendered by the most general word that fits, since a clause that is right in shape with one vague noun is more than half right, while a blank is nothing.

The final consideration is presentation, and it matters more here than in prose. Write the translation as verse, one colon per line, aligned so that the parallel members sit under each other where they can. Leave a blank line between verses. Number the verses if the passage is numbered. None of this earns a mark directly, but all of it makes the structure of your reading legible to an examiner working through a proportion-of-sense grid block by block, and legibility is worth something real when a marker has to decide whether a block is mostly correct or more than half right. It also helps you: a translation laid out as verse shows its own gaps, and a colon with nothing under it is a visible reminder that a line has been skipped.

Worked example

A poetry translation and the grid applied to it

Constructed teaching Hebrew in the shape of a biblical couplet, not a quotation. Translate, then mark two candidate answers against the 5-mark proportion-of-sense grid. רָחֲקוּ מִמֶּנִּי אֹהֲבָי · וּמִבֵּיתִי רֵעָי

Read the whole couplet before writing
The second colon has no verb, so the couplet must be read as a unit before a word of English is written. The verb רָחֲקוּ governs both halves. Deciding that first prevents the commonest failure, which is to write a complete English sentence for the first colon and then discover that the second has nothing to attach to. Reading the couplet whole is not a refinement; it is what makes the second line translatable at all.
Choose English words that keep the pair distinct
אֹהֲבָי and רֵעָי are different words for overlapping ideas, and English has the resources to keep them apart: those who love me and my companions, or my friends and my neighbours. What loses the feature is rendering both as friends, which flattens a parallel pair into a repetition and reads to an examiner as a candidate who did not distinguish them. The specification lists parallelism as a term you must be able to use, and using it in the translation is the cheapest way of showing that you have seen it.
Supply the elided verb honestly
English cannot leave the second line verbless, so the verb must be supplied. Supply the one that was elided — have gone far, or are far — and not a livelier substitute such as have fled or have deserted me, which imports a sense the Hebrew does not carry. If you want to signal that the verb is supplied rather than present, square brackets are acceptable in a translation and cost nothing.
Mark candidate A
Candidate A writes: My friends have gone far from me, and my friends from my house. The verb is right, both prepositional phrases are right, both subjects are rendered, the ellipsis has been recognised. The single loss is that the two distinct nouns have been collapsed into one word. Under the grid this is an accurate translation with one slight error, or at worst mostly correct: 5 or 4 out of 5. The structure is what carries it.
Mark candidate B
Candidate B writes: They were far from me, my lovers, and my house, my neighbours. Here the verb is rendered but its subject has been misread as an unexpressed they with the noun in apposition; more seriously, the preposition on בֵּיתִי has been dropped, so from my house has become and my house, and the second colon no longer parallels the first. Two structural errors and one dropped morpheme: less than half right, so 2 out of 5. The vocabulary was better than candidate A's; the reading was worse, and the grid rewards the reading.

Result: Those who love me have gone far from me, and my companions from my house. Candidate A, whose only fault was to collapse a synonymous pair, scores 4 or 5; candidate B, who knew more words but lost the preposition and the parallel structure, scores 2. Under a proportion-of-sense grid the architecture of the couplet is worth more than any individual word in it.

Typical mistakes

Translating line by line from the top without reading the couplet whole, so that the second colon arrives too late to correct the first.
Adding connectives and intensifiers to make the English flow, which introduces sense that is not in the Hebrew and reduces the proportion credited.
Writing the translation as continuous prose, which hides the parallelism from the examiner and hides your own omissions from you.

Active revision

Take a passage of twelve verses from set text 4 and translate it under timed conditions, laid out as verse with one colon per line. Then divide your translation into 5-mark blocks of one verse each and mark it against the six descriptors of section 4.7.1. For every block below 4, write one sentence saying whether the loss was lexical, grammatical or structural. Repeat with a passage from set text 3, which is prophetic poetry and behaves differently, and compare the two profiles.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 4.4.2 and 4.7.1 (AQA)

Key points

Building a defensible time plan for the three hours

Key points

Mapping a narrative period before translating it

Key points

A prose translation worked clause by clause and then marked

Key points

Estimating your own mark before the examiner does

Key points

Pointing nine words and justifying every mark

Key points

Taking a couplet apart before translating it

Key points

A poetry translation and the grid applied to it

Sources

Unseen translation

Key points

Building a defensible time plan for the three hours

Key points

Mapping a narrative period before translating it

Key points

A prose translation worked clause by clause and then marked

Key points

Estimating your own mark before the examiner does

Key points

Pointing nine words and justifying every mark

Key points

Taking a couplet apart before translating it

Key points

A poetry translation and the grid applied to it

Sources