Vocabulary and linguistic range

There is no defined vocabulary list, and what follows from that#

FoundationAQA 7677 section 3.5 Vocabulary — there is no defined vocabulary list for the A-level in Biblical Hebrew; students know and build upon words and regular compounds of the words listed in the Core Vocabulary List for Biblical HebrewAQA 7677 section 3.1 — there is no Defined Vocabulary List for this component; students should build on and develop their range of vocabulary through wider reading of texts during the A-level courseAQA 7677 section 4.4.1 — names of people and places in the text will be glossed, as well as words that are either very rare or used unusually, if the meaning cannot be derived from the available text

How the vocabulary list is notated and how the exam glosses

Fig. 1The left-hand block is the Core Vocabulary List's own notation as section 3.5 describes it; the right-hand block is what section 4.4.1 promises about an examination passage. The emphasised cell is the qualifier that decides how much help you will actually get: a rare word is glossed only if its meaning cannot be derived from the available text.

Section 3.5 states the position flatly and states it twice, once there and once in 3.1: there is no defined vocabulary list for the A-level in Biblical Hebrew. What exists instead is the Core Vocabulary List for Biblical Hebrew, and the specification requires students to know and build upon the words, and regular compounds of the words, that it contains. Two sentences later comes the qualification that decides how the list should be used: this Core Vocabulary List is primarily intended as a guide for teachers to assist in the planning of schemes of work. It is a floor and a planning tool, not a syllabus you can complete. Students should be introduced to a greater range of vocabulary through wider reading of original material, and are expected to study a range of genres in order to develop a wider vocabulary and a more complex understanding of syntax and accidence. Vocabulary is thus the only genuinely unbounded demand in a qualification that is otherwise specified to the level of individual grammatical terms.

Two categories count as core without being printed. The specification says that in addition to the words in the list, verbs formed from the listed roots are part of the core vocabulary, and so are cardinal and ordinal numbers. The first of those is the more useful, because it converts the list from an inventory of words into an inventory of roots: if a root appears with a root sign, every verb built on it in every binyan is fair game and is expected. The list's own notation is worth learning, because it tells you what to memorise. Verb roots are marked with a root sign and are not pointed, which is why these notes also write roots unpointed. Every other part of speech is pointed with its regular pointing, which is a strong hint that the pointing of a listed noun is itself examinable material. In the English column a comma marks a synonymous or closely related alternative and a semicolon an unrelated meaning; in the Hebrew column a comma marks an alternative form of the same word and a semicolon an alternative form with a distinct meaning, marked by a semicolon in the English column too.

The list also has a cognate column, and it is the most efficient page in the document, because it shows the root system doing its work. Beside the root אבה, meaning to accept or to be willing, it sets the noun אֶבְיוֹן, poor or oppressed; beside אָדוֹן, lord or master, it sets a Divine name. Reading down that column trains exactly the habit the unseen requires: seeing a word you have not learned and recognising the word you have. A student who learns the list as a flat vocabulary of individual items has bought perhaps two thousand words; a student who learns it as a set of roots with their cognates has bought several times that number, and has bought the ability to guess the rest.

The examination compensates for the open vocabulary in three specific ways, and each is worth counting on. One or two sentences of background information are given in English as an introduction to each Hebrew text, which tells you who the participants are and often what is about to happen. Names of people and places in the text will be glossed, so proper nouns are free and no candidate should lose time on them. And words that are either very rare or used unusually will be glossed, but — and this is the sentence to read carefully — only if the meaning cannot be derived from the available text. The board's working assumption is that you can derive. Against those three helps stands one absolute: no access to a dictionary during any of the three assessments, and in Papers 2 and 3 no access to the set texts either, except for the extracts printed on the question paper.

The practical consequence is that the reading you do is the vocabulary syllabus. The specification says as much when it notes that over the course of the A-level, students will have studied four different books of the Bible in preparation for the language and literature exams. Those four books are the ones you must read closely anyway, and their vocabulary is the vocabulary an unseen is drawn from, because an unseen prose passage is biblical narrative and an unseen poetry passage is biblical verse. Time spent on set text 1 is therefore not time taken from Paper 1: the two books of narrative prose you must know for Paper 2 are the best possible preparation for the unseen prose of Section A, and the prophetic and psalmic vocabulary of Papers 3 is the best preparation for the unseen poetry of Section B. Planning the course as though language and literature were separate wastes the largest single efficiency available.

Worked example

Reading four entries as the list intends them to be read

Using the conventions of section 3.5, say exactly what each of these entries commits you to learning: (a) a root printed as אבה with a root sign and glossed accept, be willing, with the cognate אֶבְיוֹן glossed poor, oppressed; (b) an entry glossed destroy; lose; (c) an entry glossed lord, master; (d) the absence of any entry for the numeral seven.

(a) a root, its cognate, and everything between them
The root sign tells you this is a verb root and that it is not pointed, so you learn the three consonants and not a vowel pattern. The gloss commits you to the qal sense. Because the specification says verbs formed from the listed roots are part of the core vocabulary, you are also committed to the other binyanim of this root wherever they occur. The cognate column then adds a noun, אֶבְיוֹן, that is not obviously connected in English — willing and poor look unrelated — but which the list places there so that you learn the pair together. That pairing is the method in miniature: learn the family, not the item.
(b) a semicolon in the English column
A semicolon marks an unrelated meaning, so destroy; lose is not one idea loosely rendered twice but two distinct senses of the same root, and only the context will decide which is meant. In an unseen this is the commonest reason a translation goes wrong at the level of sense rather than of form: the candidate recognises the word, chooses the first gloss they learned, and produces a clause that is grammatically correct and semantically absurd. The habit to build is to learn both senses at once and to make the choice consciously.
(c) a comma, and a cognate that is a Divine name
The comma in lord, master marks a synonymous or closely related alternative, so here there is one idea and two English words for it, and either will translate the noun. The cognate column beside it carries a Divine name, which the specification writes with a dash between the first and second letters. Two things follow. First, the vocabulary of Divine names is on the list and is therefore examinable. Second, the dash is a presentational convention and not a grammatical one: the word is pointed and construed exactly as it would be without it.
(d) what is not printed but is still required
There is no need for the list to print every numeral, because the specification states separately that cardinal and ordinal numbers form part of the core vocabulary. So an entry you cannot find may still be required. The same applies to a verb you meet in the hiphil whose qal is on the list: it is not a new word, it is a listed root in another binyan, and the specification has already told you that you are responsible for it.
What the four entries add up to
Between them these four cases cover almost everything the notation can do: a root with its family, a semicolon separating senses, a comma joining them, and a category that is required without being printed. Read the list once with this in hand and it stops being two thousand items and becomes a few hundred roots with their attendants, which is a memorisable object.

Result: (a) Learn the three consonants unpointed, the qal sense, every binyan of the root, and the cognate noun as part of the same family. (b) Learn both senses, because a semicolon marks unrelated meanings and only context chooses. (c) Learn one sense with two English words, and note that the Divine name in the cognate column is pointed and inflected normally despite the specification's dash. (d) Learn the numerals anyway: the specification makes cardinals and ordinals core without printing them.

Typical mistakes

Treating the Core Vocabulary List as a syllabus to be completed and then stopping, when the specification calls it a teacher's planning guide and a floor.
Learning listed nouns without their pointing, when the list prints them pointed and the composition will require them pointed.
Preparing the unseens and the set texts as separate projects, so that the same vocabulary is learned twice and the reading that would have made both easier is never done.

Active revision

Open the Core Vocabulary List at any page and cover the English column. Work down the Hebrew column writing your own gloss for each entry, and mark with a cross every root whose derived nouns you cannot produce. Then take a chapter of set text 1 you have already translated, list every word in it that is not on the list, and sort that list into three piles: words you could have deduced from a root you know, words the exam would have glossed as rare or as a proper noun, and words you would simply have had to know. The third pile is the only one that needs learning by heart, and it will be smaller than you expect.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, section 3.5 Vocabulary (AQA) · GCE AS and A level subject content for ancient languages (Department for Education)

The root: how a few hundred roots become a reading vocabulary#

StandardAQA 7677 section 3.3.2.2 — root, with shoresh as an acceptable alternative; first, second and third root letterAQA 7677 section 3.3.1.1 — nouns formed with preformative letters mem, alef and tav, and afformative letter nunAQA 7677 section 3.2 — derivations of nouns which are derived from verbal forms, and of proper nouns

Biblical Hebrew builds almost its whole vocabulary from roots of three consonants, and a lexicon is organised by root rather than by word. The specification's term is root, with shoresh accepted as an alternative and the Hebrew word marked unacceptable; the individual consonants are the first, second and third root letters, for which pe-, ayin- and lamed-, the words radical, and the Roman numerals I, II and III are all accepted. Learning a root is therefore not learning a word but buying an option on a family of them, and the families are large. From מלך come מָלַךְ, he reigned, מֶלֶךְ, king, מַלְכָּה, queen, מַלְכוּת, kingdom or reign, and הִמְלִיךְ, he made someone king — five items and one act of memory. From שׁפט come שָׁפַט, he judged, שֹׁפֵט, a judge, and מִשְׁפָּט, judgement or justice or custom, a word that runs through the prophets and the psalms alike.

The formation rules are regular enough to be predictive, and section 3.3.1.1 names the letters involved: nouns are formed with preformative mem, alef and tav and with afformative nun. The preformative mem is the workhorse. Put a mem in front of a root and you very often get the place where the action happens, the instrument with which it happens, or the abstract of it: from a root meaning to slaughter for sacrifice comes מִזְבֵּחַ, an altar; from שׁפט comes מִשְׁפָּט; from a root meaning to count comes מִסְפָּר, a number. The specification's term is preformative, with noun prefix accepted and introductory rejected. The same predictive power runs through the binyan participles: a piel participle begins with a mem, so a word beginning with mem may be a noun with a preformative or a participle, and telling them apart is a matter of the vowel pattern.

One root family repays being learned in full because it demonstrates every mechanism at once. The root ספר gives the qal סָפַר, he counted; the piel סִפֵּר, he recounted or told, where the intensive binyan has moved the sense from counting to narrating; the participle סֹפֵר, a scribe or secretary, a word that names a court office in the books of Kings; the noun סֵפֶר, a book, scroll or letter; and with a preformative mem the noun מִסְפָּר, a number. Five forms, one root, three different parts of speech and a shift of meaning that would look arbitrary in a bilingual list but is transparent once the family is seen whole. This is the unit in which vocabulary should be learned at A-level, and it is also the unit in which grammar questions are often set, since asking a candidate to relate a noun to its verbal root is exactly what section 3.2 has in mind.

Section 3.2 makes derivations examinable in two directions, and the second is easy to overlook: derivations of nouns which are derived from verbal forms, and of proper nouns. Proper nouns matter because the biblical narrative frequently explains a name at the moment it is given, and set text 1 does it three times. At I Samuel 1:20 the narrative attaches Samuel's name to the act of asking that has run through the chapter; at I Samuel 4:21 the child born as the ark is captured is named in a phrase the narrative itself glosses by the departure of the glory from Israel; at I Samuel 7:12 Samuel sets up a stone and gives it a name that means stone of help. In each case the etymology is not a scholarly reconstruction but a claim the text makes about itself, which is why it can be quoted in an answer and why the specification asks for it.

The root method has limits and an honest course teaches them. Not every set of three letters that looks like a root is one: some words are borrowed, some are primitive nouns like אָב, father, and אֵם, mother, that have no verbal family, and some roots are homonyms, so that identical consonants carry unrelated meanings and only context separates them. Nor is the meaning of a binyan always predictable from the qal; many piel verbs are factitive or denominative rather than intensive, and many hiphil verbs have drifted from a strictly causative sense. The root is therefore an index, not a definition. Used as an index it is the single most powerful tool a candidate has in an unseen; used as a definition it produces confident nonsense, which under the proportion-of-sense mark scheme costs more than an honest gap.

Worked example

Building one root family and testing it against a text

Constructed teaching exercise. From the root ספר, produce the members of the family, gloss each, and then explain how each of the two nouns is formed. Finally, say what a grammar question could reasonably ask about the word סֹפֵר.

The verb in two binyanim
The qal perfect third masculine singular is סָפַר, he counted. The piel is סִפֵּר, he recounted or told, with hiriq under the samekh and a strong dagesh doubling the pe, which is the visible mark of the binyan. The semantic step from counting to telling is a real one and not a coincidence of English: to recount is to count over. Naming the binyan is half the answer to any question on either form; naming the shift of sense is the other half, and section 3.3.2.2 gives you the vocabulary to do it, since intensive is one of the forces the terminology list expects.
The participle that became a job
סֹפֵר is the qal active participle, with holam on the first radical and tsere on the second, which is the participle's own vowel pattern. As a participle it means counting or one who counts; as a noun it means a scribe or a royal secretary, and it is used as a title in the historical books. That is the specification's category of participles used as verbs, nouns or adjectives doing its work: the same form is a verb in one clause and an office-holder in the next, and only the syntax decides.
The two nouns and how they are made
סֵפֶר, a book, scroll or written document, is a segolate noun formed directly from the root with no added letter; its stress falls on the first syllable and the second vowel is a helping vowel, which is why the plural is סְפָרִים with the stress moved and the first vowel reduced. מִסְפָּר, a number, is formed with a preformative mem, the letter section 3.3.1.1 names first. The term to write is preformative, or noun prefix; introductory is listed as unacceptable.
What an examiner could ask about the participle
Three questions are available and each has a different answer. What form is it: the qal active participle, masculine singular. What is its root: ספר. Why is it translated as a noun here: because it stands as the subject or in apposition to a name rather than governing an object, and the specification recognises participles used as nouns. A fourth question, from section 3.2, would ask for the derivation, and the answer is that the office is named from the act of writing or counting that defines it.
The general lesson
Learning this family costs one act of memory and yields five words, two of which are common in the historical books, plus a ready-made answer to three kinds of grammar question. Doing the same for a hundred roots yields something in the region of five hundred words, which is a far better return than five hundred separate memorisations, and it produces the derivational knowledge that section 3.2 asks for as a by-product.

Result: סָפַר qal, he counted; סִפֵּר piel, he recounted or told; סֹפֵר qal active participle, counting, and as a noun a scribe or secretary; סֵפֶר, a segolate noun, a book or scroll or letter; מִסְפָּר, a noun with preformative mem, a number. One root, five forms, three parts of speech.

Typical mistakes

Assuming that every noun beginning with mem has a preformative, when it may be a participle or simply a word whose first radical is mem.
Deriving a meaning from a root and then defending it against the context. The context wins; the root only suggested.
Reciting a name etymology that the biblical text does not itself give, which turns a rewardable observation into an unsupported claim.

Active revision

Choose ten high-frequency roots from set text 1 and build a family sheet for each: the qal perfect and imperfect, any other binyan that occurs in the set texts, the participle, and every noun formed from the root, all pointed, with a one-word gloss each. Then take a page of unseen prose you have not translated and, for every word you do not know, write the three root letters you think it has before you write anything else. Check afterwards how often the root alone would have given you a usable sense, and how often it would have misled you.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 3.2 and 3.3.1.1 (AQA)

Deducing an unglossed word under time pressure#

AdvancedAQA 7677 section 4.4.1 — words that are either very rare or used unusually will be glossed, if the meaning cannot be derived from the available textAQA 7677 section 4.7.1 — the general principle in assessing each section will be the proportion of sense achieved out of 5AQA 7677 section 3.1 — extend and demonstrate knowledge of vocabulary and linguistic structures through wider reading

The order of operations on an unglossed word

Fig. 2Every step here is an accidence rule run backwards, which is why deduction improves as the grammar does rather than as vocabulary does. The emphasised node is the pivot: nothing before it uses knowledge of the word, and nothing after it is possible without it. The dashed return edge is the only loop, and the specification's proportion-of-sense mark scheme means it should be taken at most once before you commit.

The examination will gloss a rare word only if its meaning cannot be derived from the available text, which is a statement about what the board expects you to be able to do. Deduction is therefore a taught skill and not a last resort, and it has an order of operations that works far better than staring. Strip the prefixes first, in the fixed order vav, preposition, article, interrogative he; strip any pronominal suffix from the end; strip the affixes that belong to the inflection, the plural ending, the feminine ending, the imperfect prefix. What remains is a stem. Only then look for three consonants in it, and only then consider what the vowel pattern is telling you. Doing these steps in the wrong order is the commonest reason a candidate fails to recognise a word they actually know: a familiar root becomes unrecognisable under two prefixes and a suffix.

Finding three radicals in a stem often requires reconstruction, because the weak classes hide letters. If the stem has only two visible consonants, ask whether a nun has assimilated and left a strong dagesh, whether a middle vav or yod has been vocalised, or whether a final he has dropped. If a hataf vowel appears where a plain sheva would be expected, a guttural is present and it is probably the first radical. If a letter carries a strong dagesh with no article and no vav consecutive in sight, suspect either an assimilated nun or a doubled second radical, which would make the form a piel, pual or hithpael. Each of these is a rule you already know from the accidence; deduction is simply running those rules backwards, and it is worth practising in that direction explicitly.

The binyan is the next piece of evidence and it constrains meaning even when the root is unknown. A niphal is passive or reflexive, so whatever the root means, the subject is having it done to them. A hiphil is causative, so the subject is making something happen. A piel is often intensive or factitive. A participle has no time of its own and takes it from the clause. A stative verb — one whose qal perfect is pointed with tsere or holam rather than patah — describes a condition rather than an action, so its best English is often an adjective. These constraints let you write a clause that is right in shape even when one word in it is a guess, and shape is what the mark scheme is actually measuring.

Context is the final and strongest filter, and it comes in three layers that the examination has deliberately provided. The English introduction of one or two sentences tells you the situation. The narrative logic tells you what kind of thing must be happening: a verb between a subject who is a king and an object that is a city is doing one of a small number of things. And parallelism, in a poetry passage, gives you a second line saying something close to the first, which is why an unknown word in a poetic line is often decodable from its partner in the parallel colon. Use the layers in that order and the space of possible meanings collapses quickly.

The last skill is knowing when to stop, and the mark scheme makes the decision for you. Translation passages are divided into sections each worth 5 marks, and the principle is the proportion of sense achieved: 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. Nothing in that grid rewards a blank, and nothing punishes a reasonable guess more than an omission. So a clause in which the syntax is right, the verb is right, the subject and object are right and one noun is a plausible guess will usually score 4; the same clause left unfinished will score 2 or less. Guess in the direction the context points, write the clause out in full, and move on. Time spent agonising over one noun is time taken from a whole block of marks elsewhere.

Worked example

Deducing three words that were not glossed

Constructed teaching Hebrew, not a quotation. Assume the English introduction has told you that a king is besieging a city. Work out each of the following without a lexicon, showing every step: (a) וּבְמַלְכוּתוֹ (b) וַיַּמְלִיכוּ (c) הַנִּשְׁמָרִים

(a) four morphemes and a familiar root
Strip in order. The initial shuruk is the vav conjunctive, written that way because it stands before a bet. The bet is the inseparable preposition, in. At the other end, the final vav with holam is the third person masculine singular pronominal suffix, his. What remains is מַלְכוּת. The ending ־וּת is an abstract noun ending, and the three radicals are mem, lamed, kaf. The root is מלך, to reign, which you know. So the word is and in his kingdom, or and in his reign — and note that the abstract can be either the territory or the period, which the context must choose.
(b) a prefix vowel that names a binyan
Strip the vav consecutive and you have יַמְלִיכוּ, an imperfect third person masculine plural. Now read the vowels rather than the letters: patah under the prefix and a long hiriq written with yod in the stem is the hiphil imperfect pattern, in which the characteristic he of the perfect has vanished and left its vowel behind. The radicals are again mem, lamed, kaf. A hiphil is causative, so this is not they reigned but they caused someone to reign, that is, they made him king. Translate: and they made him king. Compare וַיִּמְלְכוּ, and they reigned, where the hiriq under the prefix and the absence of the long vowel mark a qal.
(c) an article, a binyan and a plural
Strip the article הַ, noting the strong dagesh in the nun. What remains is נִשְׁמָרִים, and the plural ending ־ִים marks it as masculine plural. The prefixed nun with hiriq before a stem is the niphal, and with the participle vowel pattern this is a niphal participle: being kept, or those who are kept or guarded. The article makes it a substantive: the ones who are guarded, or the garrison, or those kept safe, depending on the situation. With a siege in the English introduction, the guarded ones inside the city is the natural sense. Radicals shin, mem, resh; root שמר; binyan niphal; form participle, masculine plural, with the definite article.
Writing the clause even where the sense is uncertain
Suppose you had never met מַלְכוּת and could get no further than an abstract noun from the root to reign. Writing and in his royal something still preserves the syntax, the possessor and the preposition, and under the proportion-of-sense grid that is a mostly-correct clause. Writing nothing preserves none of them. The decision to commit is worth two marks per block, which across a 20-mark translation is a grade.
What made each one solvable
In (a) it was segmentation: the word was long but every piece was familiar. In (b) it was the vowel pattern, which named the binyan before the root was identified and told you the meaning would be causative. In (c) it was the combination of a binyan marker and an article, which together turn a verb into a noun. None of the three required a word you had not learned; all three required grammar you already had, applied in reverse.

Result: (a) וּבְמַלְכוּתוֹ: vav conjunctive plus preposition bet plus the abstract noun מַלְכוּת from the root מלך plus the third masculine singular suffix — and in his kingdom, or in his reign. (b) וַיַּמְלִיכוּ: vav consecutive plus hiphil imperfect third masculine plural of מלך — and they made him king. (c) הַנִּשְׁמָרִים: definite article plus niphal participle masculine plural of שמר — those who are guarded.

Typical mistakes

Looking for the root before stripping the prefixes, so that a known word behind a vav, a preposition and an article is never recognised.
Forgetting that a strong dagesh may be an assimilated nun, and so failing to reconstruct pe-nun roots, which are among the commonest verbs in narrative.
Leaving a gap rather than committing to a reasoned guess, which under the proportion-of-sense grid converts a likely 4 into a certain 2.

Active revision

Take an unseen prose passage and, before translating, write out every word of more than four letters as a segmentation with plus signs between the morphemes. Then, for every word you cannot immediately gloss, write in this order: the three radicals you think it has, the weak class if any, the binyan, and a one-word guess. Only then translate. Afterwards, check your guesses against a lexicon and count how many were right, how many were wrong but produced a sensible clause, and how many were wrong in a way that broke the sentence. The third figure is the one to reduce.

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)

Four books, four vocabularies#

StandardAQA 7677 section 3 — over the course of the A-level, students will have studied four different books of the Bible in preparation for the language and literature examsAQA 7677 section 3.4.1 — the four set texts, valid for the lifetime of the specificationAQA 7677 section 3.5 — students are expected to study a range of genres in order to develop a wider vocabulary

The specification is unusual in telling you exactly which four books you will have read by the end: I Samuel, II Kings, Jeremiah and Psalms. Since it also tells you that the unseen passages are prose in Section A and poetry in Section B, the overlap is nearly total. Two of your set texts are the same genre as the unseen prose and two are the same genre as the unseen poetry, and the vocabulary you learn for the literature papers is the vocabulary the language paper draws on. Planning as though the two were separate projects is the largest single waste available in this course. The practical form of the insight is a working glossary organised by set text but audited against the unseens: for every word you learn in II Kings 9 to 13, ask whether it is court and administrative vocabulary that would appear in any narrative, or whether it belongs only to that story.

Narrative prose has a small, dense core. Verbs of motion and position — הָלַךְ to go, בּוֹא to come or go in, יָצָא to go out, שׁוּב to return, יָשַׁב to sit or dwell, קוּם to arise, נָפַל to fall — carry the plot, and half of them are weak roots, which is why the weak classes matter so much for reading. Verbs of speech and perception — אָמַר to say, דִּבֶּר in the piel to speak, שָׁמַע to hear, רָאָה to see, יָדַע to know — carry the dialogue. Verbs of transaction — נָתַן to give, לָקַח to take, שָׁלַח to send — carry the action. Around them stands a nominal core: מֶלֶךְ king, עַם people, עִיר city, בַּיִת house, אֶרֶץ land, יָד hand, דָּבָר word or matter, יוֹם day. Perhaps sixty words carry the skeleton of any narrative passage, and they are worth knowing to the point of instant recognition in every inflected form.

Set text 1, I Samuel 1 to 7, adds the vocabulary of the sanctuary, because its first chapters are set at Shiloh and its last four are the story of the ark. Expect the words for vow and offering, for priest and for prophet, for the ark itself, and for the Philistine cities and their god; expect also the recurrent vocabulary of asking and giving that the narrative uses when it explains Samuel's name at I Samuel 1:20. Set text 2, II Kings 9 to 13, adds the vocabulary of the court and of the coup: anointing, conspiracy, chariot and horseman, the words for a royal house and for the dynasty, the technical vocabulary of the temple repairs in chapter 12, and the recurring regnal formula that opens and closes each king's reign. Those formulae repeat so often across Kings that learning one gives you many.

Set text 3, Jeremiah 13 to 18, is prophetic, and its vocabulary is a third kind again. It is dominated by the verbs of turning and returning, of building and planting and their opposites of plucking up and pulling down, and by the imagery the chapters are built on: a linen waistcloth in chapter 13, drought and famine in 14, the potter and his clay in 18. Prophetic Hebrew also uses a set of recurring formulae — the messenger formula that introduces an oracle, and the phrase that attributes the utterance to the deity — which appear so often that they should be recognised as units rather than parsed word by word. Set text 4, Psalms 50 to 59, is the vocabulary of prayer and complaint: the words for enemies and for refuge, for steadfast love and for faithfulness, for the soul and for the heart, and the verbs of crying out, waiting and trusting.

Two habits make this manageable. The first is to keep the glossary by root and not by chapter, so that a word met in Jeremiah joins the family you already built from Kings. The second is to mark each entry for the direction in which you need it: R for recognition, which is enough for reading, and P for production, which is what the composition requires and which means knowing the pointing. Most of the specialised vocabulary of the four books needs only R. The narrative core of sixty verbs and sixty nouns needs P, because those are the words a 100-word English passage of third-person narrative prose with some direct speech will actually ask you to write.

Worked example

Sorting one page of vocabulary into what you need and how deeply

Take a page of narrative from set text 2 containing, among others, the following kinds of word, and sort them: a verb of motion; the noun for chariot; a technical term for a part of the temple; a proper name; a recurring regnal formula. For each, decide whether it needs recognition only or production as well, and justify the decision from the assessment.

The verb of motion
Production, without question. A 100-word English narrative in the third person will contain several verbs of going, coming, sending and returning, and Section C requires them pointed. That means knowing not only the qal perfect but the vav consecutive form, since a narrative composition will be built on vav consecutive chains: from רכב, to ride, you need וַיִּרְכַּב, and from שׁוב, to return, you need וַיָּשָׁב, which is a hollow verb and therefore worth drilling separately. Mark these P and drill the pointing.
The noun for chariot
Recognition, with production optional. The word will appear in II Kings and may appear in an unseen with a military setting, so you must read it instantly; but an English passage set for composition is unlikely to require it, and if it does, the examiner will more probably supply a proper name than an item of military hardware. Mark it R, and if you have time later, promote it. This kind of triage is what makes a glossary of a thousand entries usable rather than paralysing.
The technical temple term
Recognition only, and even then with a note of where it came from. Chapter 12 of II Kings contains vocabulary about the repair of the house and the handling of silver that is specific to that passage. You need it to translate a printed extract in Paper 2, which is precisely the situation the specification describes when it says there is no access to the set texts in the exam. You do not need it for anything else. Mark it R and attach the chapter number, so that revision before Paper 2 can be targeted.
The proper name
Neither, in one sense, and both in another. Names of people and places are glossed in Paper 1, so you never need to recognise one cold in an unseen, and in Section C the specification says names will be provided in pointed Hebrew in a glossary, so you do not need to produce them either. But for Papers 2 and 3 you need to know who the person is and where the place is, which is a matter of the storyline and the map rather than of vocabulary — and section 3.2 asks for exactly that, including the approximate locations of important places on maps of the region.
The regnal formula
Recognition as a unit, and a very high return. The formulae that open and close a reign in Kings recur with small variations dozens of times, so learning the pattern once gives you a large amount of text for free and, better, gives you a structural marker: when the formula appears, a reign is beginning or ending, and the narrative is changing gear. In a Paper 2 answer, being able to say that a passage sits inside the formulaic frame is a literary observation as well as a linguistic one.

Result: Verb of motion: P, with the vav consecutive form drilled. Chariot: R, promote later if time allows. Temple technical term: R, tagged to its chapter for targeted revision before Paper 2. Proper name: neither for Paper 1, since names are glossed in the unseens and supplied pointed in the composition, but the person and place must be known for Papers 2 and 3. Regnal formula: R as a whole unit, and doubly valuable because it also marks the structure of the book.

Typical mistakes

Learning set-text vocabulary chapter by chapter and never merging it, so the same root is learned three times under three different words.
Treating prophetic and psalmic vocabulary as an optional extra, when half of Paper 1 is poetry and its vocabulary is thinner on context than prose.
Learning everything to recognition standard only, and then discovering in Section C that recognition does not tell you how to point a word.

Active revision

Build one glossary for the whole course, organised by root, with four columns: the root, the forms you have met, the set text or texts they came from, and R or P for the depth you need. Enter set text 1 first, then set text 2, and note how much of the second is already there. Then take an unseen prose passage and mark every word that is already in your glossary: the proportion should rise steadily through the course, and if it does not, your reading is too narrow rather than your memory too weak.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 3.4.1 and 3.5 (AQA)

The poetic lexicon: what changes when the passage is verse#

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, ellipsisAQA 7677 section 3.4.1.2 — poetry set texts: Jeremiah 13-18 and Psalms 50-59

Half of Paper 1 is poetry, and poetry uses a partly different vocabulary from prose. The difference is not a matter of register alone: a set of words appears predominantly or exclusively in verse, and a set of prose words is largely avoided there. The definite article, the relative particle אֲשֶׁר and the object marker אֵת all appear far less often in poetry than in prose, which removes three of the signposts a reader of narrative relies on. In their place come rarer synonyms for common ideas, archaic and short forms of pronouns and verbs, and a heavier use of pronominal suffixes where prose would use a separate word. The practical effect is that a poetry passage is denser, shorter and less explicit, and that guessing from context is harder because there is less context.

The compensation is parallelism, which the specification lists as a syntax term you must be able to use. Biblical verse characteristically states something in one line and then restates, sharpens or reverses it in the next, so that the second line is a commentary on the first. For a reader, that means a difficult word in one line very often has a partner in the other: if you know one and not the other, the one you know constrains the meaning of the one you do not. Scholars describe the recurring partners as fixed word pairs, and a working knowledge of the commonest is worth real marks — the pair of words for poor and needy, the pair for way and path, the two words for heart, the pair for earth and world. Learning them in pairs rather than singly is learning them in the form the poetry actually uses.

Jeremiah 13 to 18 and Psalms 50 to 59 give you two different poetic vocabularies and both are examinable, so both are worth building. Prophetic poetry is argumentative: it accuses, it summons, it threatens and it pleads, and its vocabulary is heavy with the verbs of turning and returning, of listening and refusing to listen, of building and planting and their violent opposites. Its imagery is concrete and is developed at length — the spoiled waistcloth of chapter 13, the drought of chapter 14, the potter reworking a marred vessel in chapter 18 — so its nouns cluster around whichever image is in play. Psalmic poetry is relational: it addresses, complains, trusts and praises, and its vocabulary is heavy with the nouns of relationship and refuge and with the verbs of crying out, waiting, taking shelter and being delivered.

Two grammatical features of poetry cause more trouble than any vocabulary. The first is ellipsis, which the specification names as a term: a verb stated in the first line is not repeated in the second, and the reader is expected to carry it over. A candidate who does not recognise the gap will look for a verb that is not there and mis-parse a noun to fill the space. The second is the pausal form, also a listed term: because verse lines end at strong pauses, pausal forms are far commoner in poetry than in prose, and a word in pause may look different from the same word in its contextual form, with a lengthened vowel and sometimes a shifted stress. Reading a pausal form as an unfamiliar word is a real and frequent error, and the cure is to know the commonest pausal alternations.

The honest conclusion is that unseen poetry has to be prepared deliberately and cannot be prepared by doing more prose. Read the set poetry aloud until the shape of a line is familiar; keep a separate glossary of poetic synonyms beside the prose word you already know; and practise the specific move of reading the second colon first when the first is opaque. It is also worth remembering how the marks fall. Section B carries 50 of Paper 1's 150 marks, exactly as much as Section A, and its translation question carries 25 marks against Section A's 20. A candidate who treats poetry as the hard part to be attempted last is choosing to compete for two thirds of the largest paper in the qualification.

Worked example

Using the parallel line to decode an unfamiliar word

Constructed teaching Hebrew in the shape of a biblical verse line, not a quotation. Suppose Section B gives you a couplet whose first colon you can translate as they have forsaken the way of righteousness, and whose second colon contains a noun you do not know standing where you would expect a word for path, with a verb you do not know standing where you would expect a verb of walking. Work out both.

Establish the parallelism first
Before touching either word, decide what kind of parallelism you are looking at. If the second colon repeats the sense of the first with different words, it is synonymous and the unknown items will be close in meaning to their partners. If it states the opposite, it is antithetical and the unknown items will be opposites. If it carries the thought forward, the relationship is looser and the constraint is weaker. The specification lists parallelism as a term you must be able to use, and naming the type is often itself a mark in a grammar question.
Constrain the noun by its position
In a synonymous couplet, a noun standing in the second colon where way stands in the first is very probably a synonym for way, and Biblical Hebrew has a small stock of them. That is enough to translate the line accurately even if you cannot name the precise shade: writing path or track or road preserves the whole sense of the couplet and, under the proportion-of-sense grid, will not cost you the block. Writing nothing costs the whole clause.
Constrain the verb by its morphology and its partner
Now apply the accidence. If the verb is a qal perfect third masculine plural, it matches the tense and number of the verb in the first colon, which supports the reading that the two cola are doing the same work. If it is a niphal, the subject is undergoing something rather than doing it, and the parallelism is not straightforwardly synonymous after all. Read the vowels for the binyan before you guess the sense: the binyan is visible even when the root is not, and it is the stronger constraint of the two.
Check for ellipsis before you finish
Count the elements in each colon. If the second colon has one fewer than the first, a word has probably been left out and must be carried over — most often the verb, sometimes the subject, sometimes a preposition. Supplying it in English is not padding: it is what the Hebrew expects, and naming it as ellipsis in a grammar question is exactly what section 3.3.2.5 asks for. If the second colon has one more element, look for a term that intensifies rather than repeats, which is the mark of the kind of parallelism that sharpens instead of restating.
Commit
Write the couplet out as two English lines that stand in the same relation to each other as the two Hebrew cola. Where you have guessed, guess in the direction the parallelism points, and do not hedge with two alternatives in the answer: a translation containing an oblique stroke and two options reads as an admission and is marked as the weaker of the two. Choose, write, and move on.

Result: The unknown noun is decoded as a synonym for way or path, constrained by its position opposite way in a synonymous couplet; the unknown verb is constrained first by its binyan, read from the vowels, and then by its partner verb. Both are committed to in the translation without alternatives, and any element missing from the second colon is carried over from the first and named as ellipsis if a grammar question asks.

Typical mistakes

Looking for the definite article, the relative particle and the object marker as clause signposts in poetry, where they are frequently absent by convention.
Failing to carry over an elided verb into the second colon, and then mis-parsing a noun to supply one.
Mistaking a pausal form for a different word, which turns a known vocabulary item into an unknown one at the very end of a line.

Active revision

Take ten verses of set text 4 and write each line of verse on its own line of your page, so that the parallelism is visible on the page. For each pair of lines, mark which words correspond, and where a word in one line has no partner in the other, say whether that is because the line is elliptical or because the parallelism is not synonymous. Then cover the first colon of each pair and see how much of it you could reconstruct from the second. Finally, list every word in the ten verses that you would not expect to meet in narrative prose: that list is the beginning of a poetic glossary.

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)

Active vocabulary: what the 50-mark composition actually needs#

AdvancedAQA 7677 section 4.4.3 — in Section C students will translate a prose passage of approximately 100 words from English into Biblical Hebrew; the passage will be written primarily in third person narrative prose with some direct speech; names of people and places will be provided in pointed Hebrew in a glossaryAQA 7677 section 4.7.3 — the composition passage will be divided into nine sections each worth 5 marks; additional marks to a maximum of 5 will be awarded for individual instances of idiomatic and stylish Biblical Hebrew writing

Section C is worth 50 marks, a third of Paper 1 and a sixth of the whole A-level, and it is the only place in the qualification where vocabulary must be produced rather than recognised. That changes what learning a word means. To read מֶלֶךְ you need to recognise five letters; to write it you need to know that the segol appears twice, that the noun is a segolate stressed on the first syllable, and that its plural is מְלָכִים with the first vowel reduced. Every noun you intend to use in the composition therefore has to be learned in at least two forms, singular and plural, with pointing, and every verb in at least the vav consecutive third person, since a third-person narrative is built from vav consecutive chains. The correct unit of preparation is not a word list but a small, deeply known production kit.

The specification tells you a great deal about what the kit must cover. The passage is approximately 100 words, based on a suitably ancient text and possibly adapted; it is written primarily in third person narrative prose with some direct speech; and names of people and places will be provided in pointed Hebrew in a glossary. From that you can predict the grammar with some confidence: vav consecutive narrative, verbs of motion, speech, giving and taking, a good deal of direct speech introduced by a verb of saying, prepositional phrases of place and direction, some construct chains, and probably at least one relative clause and one temporal clause. You can also predict what you will not need: an extensive abstract vocabulary, technical terms, and proper names, which are supplied.

A realistic kit is therefore small. Around fifty verbs, each known in the qal perfect, the qal imperfect, the vav consecutive third person masculine singular and plural, the imperative and the infinitive construct, all pointed, covers the great majority of what a narrative needs; add the handful you know in piel and hiphil for the causative and intensive senses you will occasionally want. Around eighty nouns, each known in singular, plural and construct, pointed, covers the furniture of a biblical scene: people, kinship, parts of the body, buildings, land, animals, time. Then the closed classes, which are small and absolutely essential: the pronouns and pronominal suffixes, the numerals to ten, the prepositions with their suffixed forms, the demonstratives, and the particles כִּי, אֲשֶׁר, אִם, הִנֵּה, לֹא, אַל, גַּם and אַף.

Where the kit fails, paraphrase. The composition is marked in nine blocks of 5 marks on the proportion of sense achieved, so a block that conveys the sense in words other than the obvious ones can still score highly, while a block left blank scores nothing and a block with a fabricated word scores very little. If you cannot write the word for a specific object, name the class it belongs to. If you cannot produce a hiphil, use two clauses: he spoke to the people and they went out, rather than he sent the people out. If you cannot manage an abstract noun, use a verb. This is not a trick; it is what the mark scheme is designed to allow, and it is what a real translator does.

Finally, the style marks. Up to 5 additional marks are awarded for individual instances of idiomatic and stylish Biblical Hebrew writing, which is a tenth of the section for a handful of well-placed constructions. A small stock of genuine biblical idioms is therefore worth more than a large stock of nouns. Learn the narrative opening וַיְהִי, the speech-introducing infinitive לֵאמֹר, the particle הִנֵּה used to present something to the hearer's attention, the temporal formula built on וַיְהִי כַּאֲשֶׁר, the infinitive absolute used to reinforce a finite verb of the same root in the pattern מוֹת יָמוּת, and the use of the construct chain where English would use an adjective. Six constructions, each of which can be dropped into almost any narrative, and each of which reads as Hebrew rather than as English in Hebrew clothing.

Worked example

Paraphrasing three sentences you cannot translate directly

Constructed teaching exercise in the shape of Section C. For each English sentence, first identify the item you probably cannot produce, then write a Hebrew rendering that preserves the sense using only the core kit. (a) The king appointed officials over the province. (b) He was extremely afraid. (c) The people rebelled against him and the city was besieged.

(a) when you cannot produce the noun
The likely gap is the word for an official or for a province. Both are avoidable. Officials over the province can be rebuilt as men over the people, or as servants over the land, using words that are certainly in the kit: אֲנָשִׁים or עֲבָדִים, the preposition עַל, and עָם or אֶרֶץ. The verb to appoint can be avoided by using a construction of giving or setting: he gave them over the people, or he set them over the land. The sense reaching the examiner is the same, the syntax is correct, and every word is pointed with confidence. A block like this can score 4 or 5 without containing a single word the candidate had to guess at.
(b) when you cannot produce the intensifier
Extremely is the trap, because the obvious adverbial equivalents are easy to get wrong. Biblical Hebrew has an idiomatic answer that is also a style mark: the infinitive absolute reinforcing a finite verb of the same root. He feared a fearing, in the pattern of מוֹת יָמוּת, produces exactly the emphasis English gets from extremely, and it reads as Hebrew rather than as translated English. If that construction is not secure, the ordinary intensifying adverb מְאֹד after the verb is safe and correct. Either way the block is complete; only the second forgoes the style mark.
(c) when you cannot produce the passive
The city was besieged is a passive with a technical verb. Two escape routes exist, and both are better Hebrew than a wrong niphal. The first is to make the enemy the subject: the enemies came against the city, or they encamped against the city, which uses ordinary verbs of motion and a preposition. The second is to use the impersonal third person plural, which Hebrew uses freely where English uses a passive. Rebelled likewise has a common verb, but if it will not come, they did not listen to his voice or they went out from under his hand both carry the sense and are idiomatic.
Building the chain
Having solved each block, join them the way biblical narrative joins them. The first finite verb sets the tense; every subsequent clause takes a vav consecutive with the imperfect. So the sequence runs and the king set men over the people, and the people did not listen to his voice, and the enemies came against the city. Three clauses, one connective, no subordination — and that is not a simplification but the actual style of the historical books, which is what the examiner is asking you to imitate.
The scoring logic
Each of the nine blocks is worth 5 marks on the proportion of sense achieved, with a separate column for spelling and pointing and the rule that a block with no attempt at pointing cannot exceed 2. A paraphrase that conveys the whole sense in secure, pointed Hebrew will therefore score at or near the top of its block, while a literal attempt containing an invented word and uncertain pointing will not. Over nine blocks the difference between those two strategies is easily fifteen marks, which is five per cent of the entire A-level.

Result: (a) Rebuild appointed officials over the province as set men over the people, using only core nouns and the preposition al. (b) Use the infinitive absolute construction for the emphasis, or the adverb meod if it is not secure. (c) Turn the passive into an active with the enemy as subject, or use the impersonal third plural. Then join all three with vav consecutive chains, which is the genuine style of biblical narrative and the one the composition is imitating.

Typical mistakes

Preparing vocabulary only to recognition standard, and discovering in the exam that you know what a word means but not how it is pointed.
Attempting a complex English sentence word for word instead of rebuilding it as two simple Hebrew clauses joined by a vav consecutive.
Inventing a plausible-looking word for something you cannot say, which under a proportion-of-sense grid scores less than an honest paraphrase.

Active revision

Write out your production kit in full, pointed, on four sides of paper: fifty verbs in six forms each, eighty nouns in three forms each, and the closed classes complete. Then take any hundred words of English narrative prose — a passage from a translation of one of your own set texts is ideal, since it is by definition suitably ancient in style — and translate it using only what is on those four sides. Mark every place where you had to paraphrase, and add the missing item to the kit only if it is a word a narrative would plausibly need again.

Active recall

Recall the key points — then reveal.

Sources: AQA A-level Biblical Hebrew 7677 specification, sections 4.4.3 and 4.7.3 (AQA)

Key points

Reading four entries as the list intends them to be read

Key points

Building one root family and testing it against a text

Key points

Deducing three words that were not glossed

Key points

Sorting one page of vocabulary into what you need and how deeply

Key points

Using the parallel line to decode an unfamiliar word

Key points

Paraphrasing three sentences you cannot translate directly

Sources

Vocabulary and linguistic range

Key points

Reading four entries as the list intends them to be read

Key points

Building one root family and testing it against a text

Key points

Deducing three words that were not glossed

Key points

Sorting one page of vocabulary into what you need and how deeply

Key points

Using the parallel line to decode an unfamiliar word

Key points

Paraphrasing three sentences you cannot translate directly

Sources