Using a Lexicon and Reading Aids

Sooner or later you will meet a Hebrew word this course never taught you. Probably in the first verse you read on your own. The skill that matters then is getting from the form on the page, say וַיִּתֵּן, to the right entry in a dictionary, נָתַן "give", and then to the right sense of that entry. In Greek you mostly strip endings and an augment. In Hebrew you strip prefixes and endings too, and then you often have to put back a root letter that has vanished. This lesson teaches that skill, by hand first. The online tools come second. They are excellent, and they make lazy readers out of people who never learned to work without them. (Lesson 20 of Intro to Classical Greek makes the same argument for Greek.)

Everything in lessons 1 to 19 goes into this. You need the prefixes from lesson 7, the suffixes from lesson 9, the idea of the root from lesson 10, the stem markers from lesson 15 and, above all, the weak verbs from lesson 16.

Which dictionary

A dictionary of an ancient language is called a lexicon. For classical Hebrew in English, a handful of works matter. Here they are, with what you can and cannot get for free.

This course describes HALOT and DCH but does not quote them: they are modern, copyrighted works, and you would need a library or a purchase to use them. Most serious readers end up owning one of them. For learning to read, BDB is enough, it is free, and its arrangement by root teaches you Hebrew while you look things up. Its English is old (it says "Gn" for Genesis and "ψ" for Psalms, and its etymologies from Akkadian and Arabic are sometimes out of date), but its sense divisions and its lists of passages are still the place most work begins.

How BDB is built

BDB follows the order of the alphabet, but the alphabet of roots, not of words. Under each root you find the verb (if there is one), then the nouns, adjectives and names made from that root. Roots are written without vowel points, and this lesson writes them that way too: ה-ל-ך, written הלך. When a verb is attested in the Qal, BDB names it by the Qal perfect third masculine singular, pointed: הָלַךְ "he went" (Gesenius, section 30, on stems and roots).

A real entry heading

Here is how the entry for "go" begins in the Open Scriptures digital edition of BDB, which records the printed page as 229. The digital file gives the heading and an outline of the senses; the printed book fills in every sense with passages, nine pages of them.

הָלַךְ 1546 vb. go, come, walk
Qal Impf. usually (629 t.) as if from ילך

1. הָלַךְ: the headword, the Qal perfect 3ms.
2. 1546: the number of times the entry counts the verb in the Hebrew Bible.
3. vb.: verb. Nouns are "n.m." or "n.f." (noun masculine, noun feminine), "n.pr.m." is a man's name, "adj." an adjective.
4. go, come, walk: the basic English senses.
5. Qal Impf. usually (629 t.) as if from ילך: "in the Qal the imperfect is usually formed (629 times) as if the root were ילך." That single line will save you when you meet וַיֵּלֶךְ below. Gesenius treats the same fact under the verbs that begin with yod (Gesenius, section 69).

After the Qal, this entry runs through the other stems the verb uses: Pi. (Piel), Hithp. (Hithpael), Hiph. (Hiphil). BDB's usual order is Qal, Niph., Pi., Pu., Hiph., Hoph., Hithp., and it skips any stem the verb does not use; a minority of entries, this one among them, put Hithp. before Hiph. Inside each stem, senses are numbered with Roman numerals, Arabic numerals and letters. For הָלַךְ the Qal is split into I. literal uses (of persons, of animals, of things) and II. figurative uses (pass away, die; live; of moral and religious life). The Hiphil means "lead, bring, carry".

Then come the derived words, still under the same root. After the verb הָלַךְ BDB lists הֵלֶךְ "traveller", הֲלִיכָה "a going, way", מַהֲלָךְ "walk, journey" and תַּהֲלוּכָה "procession". Notice the last two: they begin with מ and ת, but you find them under ה, because those letters are not part of the root. They are noun prefixes (Gesenius, section 85).

Square brackets, Roman numerals and "v."

Look at the cross-reference for רֹעִי. It sits in the ר section right after the verb, because BDB's editors knew learners would meet רֹעִי "my shepherd" in Psalm 23:1 and wonder what it was. On the same page there is a trap: רְעִי, with shewa instead of holem, is a different word, a noun meaning "pasture" that occurs once (1 Kings 5:3). One vowel separates them. That is why you always copy a form with every vowel before you look it up.

The page numbers in this lesson come from the Open Scriptures digital BDB, which records the page of the 1906 printing where each entry starts. Most later reprints keep those page numbers, which is why scholars cite "BDB 229" for הָלַךְ.

From the page to the root

The form on the page is almost never the headword. Here is a method, in the order that usually works. It is the Hebrew version of the method lesson 20 of the Greek course gives.

1. Copy the form with all its points. Dagesh, shewa and the exact vowels are clues. Leave the cantillation marks aside.
2. Take off what is attached in front. The conjunction וְ or וּ; the article הַ (with a dagesh after it); the prepositions בְּ, כְּ, לְ and מִן (often just מִ with a dagesh in the next letter); the question marker הֲ; the relative שֶׁ. The wayyiqtol וַ with a dagesh in the next letter goes too (lesson 12).
3. Take off what is attached behind. Pronominal suffixes such as ךָ, וֹ, נִי, הֶם; plural and feminine endings ִים, וֹת, ָה; verb endings such as תִּי, תָּ, וּ, נוּ.
4. Read the stem and conjugation markers, then set them aside. Imperfect prefixes א ת י נ; Niphal נִ or a dagesh in the first root letter; Piel and Pual dagesh in the middle letter; Hiphil הִ, יַ and a hireq-yod before the last letter; Hithpael הִתְ; participle מְ; noun prefixes מ and ת (lesson 15).
5. Count the letters that are left. Three: look them up. Two: a root letter has gone missing, and the vowels tell you which one (next section).
6. Find the root in BDB. If you cannot find it, try the other candidates, and look for a cross-reference entry under the form itself.
7. Find the stem inside the entry, then the sense that fits the sentence. Check the passages BDB cites under that sense. If one is your verse, you are done.

Restoring a missing root letter

Lesson 16 taught the weak verbs from the root out. Here is the same knowledge turned around, as clues you see in a form and what they point to.

Every form in the second column occurs in the books this course quotes: יִתֵּן in Genesis 43:14, יִקַּח in Exodus 21:10, יֵשֵׁב in Numbers 35:28, יָקוּם in Exodus 21:19, יִבְנֶה in Deuteronomy 25:9, יֹאמַר in Genesis 31:8 and לְהוֹשִׁיב in 1 Samuel 2:8.

Strong's numbers

In 1890 James Strong published an exhaustive concordance of the English King James Version. At the back he added a short Hebrew and Aramaic ("Chaldee") dictionary and gave every headword a number, from 1 (אָב, father) to 8674. The concordance let a reader with no Hebrew find every place an English word stood for a given Hebrew word. The numbers outlived the book. Today almost every free tool tags each Hebrew word with a Strong's number, and so does the Open Scriptures data this course uses: in its tags for Genesis 1:1, בָּרָא carries 1254 a (Strong's 1254 with a letter added, explained below) and אֱלֹהִים carries 430.

A Strong's number is a handy address. Type H1254 into a search box and you get every verse with that verb, whatever its form. That is its real value: it gets you from a form to a list of passages, and to the BDB entry, quickly.

The limits

• The glosses are short and old. Strong gives a line or two, often shaped by how the King James Version translated the word. A number is not a meaning. Read the lexicon entry and the context.
• One verb, two numbers. Strong split some verbs by the shape of the form. The verb "go" has 1980 (הָלַךְ) and 3212 (יָלַךְ). In this course's readings, וַיֵּלֶךְ in Ruth 1:1 is tagged 3212, while הֹלֵךְ in Ecclesiastes 1:4 is tagged 1980. BDB treats both as one verb. A search on one number finds only half the verb.
• Different words, one number. Homonyms that BDB separates (I. and II.) can share a number. The Open Scriptures project adds letters to split some of them: טוֹב is tagged 2896 a in Genesis 1:4 (the adjective "good") and 2896 b in Psalm 23:6 (the noun "a good thing, goodness").
• The numbers do not parse. 3212 tells you nothing about whether the form is an imperative or a wayyiqtol. You still have to read the form, or read a morphology code.
• Hebrew and Aramaic are mixed. Aramaic words from Daniel and Ezra sit in the same number series as the Hebrew ones, and names are numbered separately from the common nouns they come from.

Use the numbers to find, never to define.

The free tools

All of these were checked in October 2026. Websites change; if a description here does not match what you see, trust the site's own help pages.

Two older free books belong on this list too. Davidson's Analytical Hebrew and Chaldee Lexicon (1848) lists every form that occurs in alphabetical order, so you can look up וַיֵּלֶךְ as it stands and be told its parsing and root. Generations of students have used it to check their own work. And Gesenius' Hebrew Grammar, which this course cites by section, has a full set of paradigms at the back.

Reading morphology codes

The Open Scriptures tags are compact. Each word gets a code like HVqp3ms. Learn to read them and every free tool that uses them opens up. The first letter is the language: H for Hebrew, A for Aramaic. Then comes the part of speech and its details. When a word has prefixes or suffixes, each piece gets its own code, separated by a slash, in the order the pieces are written: HC/Vqw3ms is a conjunction plus a verb.

Upper and lower case matter: p is Piel and P is Pual; h is Hiphil and H is Hophal. Here is בָּרָא from Genesis 1:1, code HVqp3ms: H Hebrew, V verb, q Qal, p perfect, 3 third person, m masculine, s singular. "He created."

The Open Scriptures tags for a real verse

Here is Jonah 1:3, then what the Open Scriptures data give for it, word by word: each word, its lemma (Strong's number with any prefix letters), and its morphology code. The right-hand column decodes each code.

וַיָּקָם יוֹנָה לִבְרֹחַ תַּרְשִׁישָׁה מִלִּפְנֵי יְהוָה וַיֵּרֶד יָפוֹ וַיִּמְצָא אָנִיָּה בָּאָה תַרְשִׁישׁ וַיִּתֵּן שְׂכָרָהּ וַיֵּרֶד בָּהּ לָבוֹא עִמָּהֶם תַּרְשִׁישָׁה מִלִּפְנֵי יְהוָה׃

Jonah 1:3, Westminster Leningrad Codex

Notes on the output. The lemma column uses letters for prefixes: c for the conjunction, l for לְ, b for בְּ, m for מִן, d for the article, k for כְּ. A bare letter with no number, like b on בָּהּ, means the word is only a preposition plus a suffix. The added letter after a number (6965 b, 5973 a) is the Open Scriptures way of splitting a Strong's number that covers more than one word. In Ruth 1:1 you will see 1035+ on בֵּית; the plus marks a word that belongs to a name of more than one word, בֵּית לֶחֶם Bethlehem.

Three words in Jonah 1:3 are wayyiqtols of weak verbs, and the codes do not tell you which letter is weak. וַיָּקָם is hollow, וַיֵּרֶד is I-yod, וַיִּתֵּן is I-nun. Notice the vowels. In וַיָּקָם the stress moves back to the prefix, as it does in many short wayyiqtols, and the last vowel, a qamets in a closed unstressed syllable, is qamets hatuf: [va-YA-kom]. Likewise וַיֵּרֶד [va-YE-red], with segol where the plain imperfect יֵרֵד has tsere.

Reader's editions and interlinears

Two kinds of printed help sit between you and a lexicon.

A reader's edition prints the Hebrew text with footnotes that gloss the less common words and parse the hard verbs. A Reader's Hebrew Bible (Zondervan, 2008), edited by A. Philip Brown II and Bryan W. Smith, glosses the words that occur 100 times or fewer. The German Bible Society and Hendrickson publish a reader's edition of the standard scholarly text, Biblia Hebraica Stuttgartensia, with a similar apparatus. The idea is good: the frequent words you should know by heart, and the footnotes keep your eye on the Hebrew line instead of on a dictionary. Use one once you know the few hundred most common words.

An interlinear prints an English word under each Hebrew word. It is useful for one thing, checking which Hebrew word stands behind a translation. It is harmful for learning, because your eye goes to the English line every time and the Hebrew becomes decoration. The glosses also follow Hebrew word order, which makes poor English, and one gloss per word hides the range of senses a lexicon shows you.

The Masoretic notes: ketiv and qere

The Masoretes were Jewish scholars of the seventh to tenth centuries AD who added the vowel points and accents to the consonantal text and guarded it with notes, which together are called the Masorah. In the Leningrad Codex the small notes in the side margins (the Masorah parva) count words and flag unusual spellings, and the larger notes at the top and bottom of the page (the Masorah magna) give lists. The photograph below shows both.

The Leningrad Codex was copied in Cairo in 1008 or 1009 AD and is now in the National Library of Russia in Saint Petersburg. It is the oldest complete manuscript of the Hebrew Bible, and the standard scholarly editions and the Westminster and UXLC digital texts all follow it.

The most important notes for a reader are the ketiv and qere (Aramaic: "written" and "read"). The Masoretes would not change a consonant of the text they received. Where they held that a word should be read differently from how it was written, they left the written consonants (ketiv) in the line and put the reading (qere) in the margin. Printed editions mark the word with a small circle and give the qere in a note. The ketiv in the line usually carries the vowels of the qere, which can make it look like nonsense until you see the note (Gesenius, section 17). Counts differ with the manuscript and the method of counting, from roughly 850 to more than 1,500.

Readings that are never written: the perpetual qere

A few words are so common that the Masoretes did not note them each time. You simply know. These are called perpetual qere.

• יְהוָה: the divine name. As lessons 2 and 3 explained, by long Jewish practice it is read Adonai [a-do-NAI], "my Lord", and the vowels written with it point to that reading.
• הִוא: in the five books of Moses the pronoun "she" is usually written הוא, the same consonants as "he", and pointed with hireq: הִוא (Genesis 3:12). Read it הִיא [hi]. Outside the Pentateuch it is written הִיא, as in Ruth 1:3.
• יְרוּשָׁלִַם: Jerusalem is usually written without a yod before the final mem, but the Masoretes put a hireq there as well as the patah or qamets under the lamed, so that it is read Yerushalayim [ye-ru-sha-LA-yim] (Psalm 79:1). The two vowels under one letter look odd; the hireq stands for the yod that is read but not written.

Printed editions mark every ketiv and qere with a note, and so do tanach.us and the Open Scriptures data. When you read such a verse, read the qere aloud, and when you quote it, say so in your notes.

Updated October 28, 2026.