How Hebrew Builds Meaning
This is the second of three parts. Part 1 introduced a modern classroom chart that lays Hebrew’s seven verb patterns on a menorah and considered what is known of the chart’s development. Part 2 moves from that visual arrangement to the grammar it represents: how Hebrew builds words from roots and patterns.
One Word Family, in English First
Start with three English sentences:
- A student learns.
- A teacher teaches.
- A lesson is taught to someone — they are on the receiving end of it.
Those are obviously related. English marks the relationships with separate words and helping verbs: learn, teach, be taught. They describe the same activity from three positions: the learner, the teacher, and the lesson received.
Hebrew marks the same relationship differently. Rather than treating learn, teach, and be taught as three unrelated verbs, it builds them around a shared consonantal root and changes the pattern around it.
The shared root is למד—three consonants, read right to left, conventionally written in English as L–M–D. Different vowel patterns and added elements shape that root into different words and grammatical roles. The dictionary form is לָמַד (lāmad, pronounced roughly lah-mad).
Here are three real examples, each taken from the Hebrew Bible rather than invented for the occasion:
| Hebrew, as it appears | Roughly how it sounds | Where | King James wording | |
|---|---|---|---|---|
| learn | אֶלְמַ֥ד | elmad1 | Psalm 119:71 | “that I might learn thy statutes” |
| teach | מְלַמֵּ֣ד | melammed | Psalm 18:34 | “He teacheth my hands to war” |
| be taught | מְלֻמָּדָֽה | melummadah2 | Isaiah 29:13 | “their fear toward me is taught by the precept of men” |
If the Hebrew script is unfamiliar, use the “Roughly how it sounds” column. Notice that in elmad, melammed, and melummadah, the same consonants—L, M, D—remain in the same order. The surrounding letters and vowel patterns change, marking different grammatical roles and shaping how each word is understood.3
(The forms above are quoted as they stand in the Hebrew text, accent marks and all.)
What the Pattern Is Doing
On the chart, the seven patterns are traditionally named with the model root פ–ע–ל—peh–ayin–lamed (P–ʿ–L), meaning “do” or “act.” The names do more than label the patterns. As readers hear the root in Paal (often called Qal), Niphal, Piel, Pual, Hiphil, Hophal, and Hitpael, they can begin to notice recurring vowel shapes, prefixes such as ni-, hi-, and hit-, and other features that distinguish particular patterns—for example, a doubled middle consonant, marked by the small dot in the central letter. In the language, these are recurring ways a root can take shape, each performing a different grammatical role.
Return to the two teaching verbs, melammed and melummadah. Both contain L–M–D, a participial m prefix, and a doubled middle consonant. Those shared features show that the forms are related. Their vowel patterns distinguish one who teaches from something being taught: active Piel and passive Pual. Here the root keeps the words in the same learning family, while the pattern changes the role.
The ending does another job: the -ah on melummadah is grammatical agreement, not what makes the word passive. That is the distinction between a verb pattern and the additional forms built within it, and the linked notes give the details.
A root does not have to appear in all seven patterns. Across the Hebrew Bible, L–M–D appears in only three of them: Qal (or Paal) for learn, Piel for teach, and Pual for be taught. The chart shows the available structures; actual usage tells us which ones this root uses.
Hebrew roots do not form verbs alone. The same L–M–D family also gives us words for the learner and for learning itself. These are nouns, not additional verb patterns:
| Meaning | Hebrew | Roughly how it sounds | Where to see it |
|---|---|---|---|
| a pupil or disciple | תַּלְמִיד | talmid | 1 Chronicles 25:8, where the KJV calls the pupil a “scholar” in a passage about temple musicians |
| pupils or disciples | תַּלְמִידִים | talmidim | Pirkei Avot 1:1, a later rabbinic teaching that calls for raising many disciples |
| study or learning | תַּלְמוּד | talmud | Mishnah Peah 1:1, in the Mishnah, the earliest written collection of rabbinic teaching: talmud Torah means the study of Torah |
The pupil is a talmid; several pupils are talmidim, not talmudim. Talmud names learning or study and is also the name of the great collections of rabbinic discussion. Notice l–m–d inside each word. One root connects an action, the person learning, and the study itself. The seven binyanim describe the verb patterns; the nouns extend the word family without adding branches to that seven-pattern chart.
Learning and Measuring: Two Roots That Can Look Alike
Learning and measuring are not the same Hebrew root, yet their sounds, forms, and meanings invite examination. Their resemblance does not establish a historical relationship by itself, but it raises a question worth following. The L–M–D family has just shown how Hebrew connects learn, teach, and be taught. A second root, M–D–D, gives us madad, “to measure.”
There is a genuine overlap in spelling. In Zechariah 2:2, lamod, לָמֹד, means “to measure.” Its first letter is the preposition to, attached to a form of M–D–D. This is not the same root as lamad, “learn,” yet it suggests a fitting conceptual comparison: learning often proceeds through attentive measuring and comparison. Without vowel marks, both can be written למד. Even vowel-marked lamod can mean “to learn” elsewhere in scripture, so context, sentence, and grammar—not the letters alone—identify the word. BDB: learn; BDB: measure.
The comparison also belongs within a larger history of contact. Greeks adapted a North Semitic—probably Phoenician—alphabet for their own language, reshaping it to represent Greek vowels. Later, Greek translators of the Septuagint used forms of manthanō, “to learn,” when rendering Hebrew lamad.
It is worth placing these words beside one another. Hebrew madad—and lamod, “to measure,” formed with the preposition to—belongs to M–D–D. The L–M–D family gives us lamad, “to learn,” and talmid, “pupil.” Greek mathētēs, μαθητής, likewise means “pupil” or “disciple,” and belongs with manthanō and mathēma, “a thing learned” or “lesson,” the source of later mathematical vocabulary.
The comparison is not merely conceptual; it is also audible. These word families repeatedly bring learning, measuring, and instruction into close relation through similar consonantal shapes and neighboring dental sounds. Across languages, sounds such as d, t, and the two “th” sounds can shift through voicing, aspiration, and softening.4
These connected sounds, meanings, and histories do not conclusively demonstrate that the Hebrew and Greek roots share an earlier ancestor. They do show why the comparison invites further inquiry. Hebrew and Greek remain distinct language families, yet their worlds met repeatedly through writing, translation, education, travel, and trade. The echo therefore deserves serious consideration while remaining an open question rather than a proven etymology.
The full interactive model has its own tab in this series. It begins with this L–M–D comparison and then lets you explore the broader pattern system.
Roots: A Real Grammatical Tradition, With Real Trees In It
The connection begins with the vocabulary itself. Here are five terms that help us distinguish the patterns and forms of Hebrew words:
- Binyan; plural binyanim (בניין; בניינים): literally a building, used for a verb pattern, as in the learn/teach/be-taught examples above. This metaphor is architectural, not botanical. Ben-Yehuda: binyan.
- Shoresh; plural shorashim (שורש; שורשים): a root, both a plant’s root and the consonants shared by a family of words, such as L–M–D in our learning examples. Academy: root.
- Geza; plural geza’im (גזע; גזעים): a tree’s trunk or stock, and, in grammatical terminology, a stem. A stem is the base form used to make further word forms, rather than just the underlying root consonants. One caution about the English word: most English-language grammars of biblical Hebrew use stem for the binyan itself, so you will find the seven patterns called the seven stems. That is a different use of the word from geza here, and the two are worth keeping apart. The Academy of the Hebrew Language — Israel’s official body for Hebrew terminology — illustrates that base-form idea with bayit, “house,” and beit-, “house of.” Academy: stem.5
- Gizrah; plural gzarot (גִּזְרָה; גְּזָרוֹת): a root class. A binyan identifies the verb pattern; a gizrah identifies features of the root—such as a weak letter in a particular position or a repeated consonant—that make that pattern conjugate in a distinctive way. Academy: verb rules by root type.
- Mishkal; plural mishkalim (משקל; משקלים): literally weight or measure, used for a recurring noun pattern. Noun patterns belong to the wider root-and-pattern system, but they are not represented on the seven-binyan chart. Academy: noun pattern.
The root-and-stem picture helps us see how related words can share something beneath their different surfaces. It does not mean every similar-looking word is related, or that every root produces every form.
David Kimhi (c. 1160–1235) was a medieval Hebrew grammarian whose works preserved and organized earlier Jewish grammatical learning. In his Sefer HaShorashim—the Book of Roots—the same vocabulary can describe a tree’s root and a word’s root. Kimhi keeps the analogy clear, however: the consonantal root and the verb pattern built from it are distinct parts of the language, even as botanical imagery gives its grammar a memorable vocabulary.6
Kimhi is therefore an important witness to an inherited grammar vocabulary. His work does not provide evidence for the modern menorah diagram itself, but it helps us see the grammatical tradition on which that diagram draws. The chart remains a teaching arrangement: it makes recurring relationships easier to see and remember, but it is not a definitive map of how Hebrew developed. It offers a useful first framework for recurring verbal structures, yet it cannot account for every root or every irregular form. Nor do these patterns and structures prove an ancient menorah-based system. We cannot yet trace the root-and-pattern system, or its terminology, to a menorah arrangement; the historical record has too many gaps. What Kimhi gives us is a glimpse of the medieval scholarly tradition within which a later diagram like this became intelligible—and that is worth contemplating.7
Part 3 turns to the older tradition of reading the lampstand as a figure for the word, and asks how much of it can be connected to the modern chart.
Sources
Works cited.
- Academy of the Hebrew Language, terminology records.
- Biblical lexicon entries: root (H8328) and trunk/stock (H1503).
- David Kimhi (Radak), Sefer HaShorashim.
- Ben-Yehuda dictionary, entries for binyan and shoresh.
- Strong’s Hebrew lexicon, entry H3925.
- Open Scriptures Hebrew Bible, morphologically tagged text.
On the worked examples. The occurrence counts are measured rather than estimated: across the tagged text named above, H3925 occurs 86 times — 57 Piel, 24 Qal, 5 Pual. English wordings follow the King James Version.
The three forms in the table. Elmad is Qal imperfect, first-person singular: “I may learn.” Melammed is a masculine singular Piel active participle: “one teaching.” Melummadah is a feminine singular Pual passive participle: “being taught.” Psalm 18 is numbered 18:34 in English Bibles and 18:35 in the Hebrew text. Brown–Driver–Briggs lists Qal “learn” at Psalm 119:71, Piel “teach” at Psalm 18:35, and Pual “taught, trained” at Isaiah 29:13 and 1 Chronicles 25:7. L–M–D entry. ↩︎
Why melummadah means “being taught.” In Isaiah 29:13, מְלֻמָּדָה (melummadah) is a feminine singular Pual participle from L–M–D: the passive counterpart to the active Piel form illustrated by melammed, “one teaching.” It agrees in gender and number with mitsvat, “commandment” or “precept,” so the -ah ending is agreement with the precept, not the feature that makes the word passive. The Pual pattern is what identifies it as something taught or learned; the verse describes the people’s reverence as a human precept that has been taught. That grammatical observation supports the reading, but it does not by itself settle every theological or rhetorical question in Isaiah’s warning. Isaiah 29:13, Hebrew analysis; Brown–Driver–Briggs, לָמַד. ↩︎
What shifts between the forms. In melammed and melummadah, the vowel after the first root consonant changes from a to u, and the vowel after the second changes from e to a. The doubled middle consonant remains. It is the pattern as a whole—not one vowel alone—that marks the change from the active Piel role to the passive Pual role. ↩︎
Why the phonetic comparison is worth pursuing, but remains open. The early Greek alphabet retained separate letter-lines from Semitic dalet, tet, and tav as delta, theta, and tau. The Greeks adapted those letters to their own sound system; they did not merge the three into one. In the Biblical Hebrew reading tradition, the begadkefat letters preserve stop and fricative variants, including dalet /d/ and /ð/ and tav /t/ and /θ/. Greek delta later moved toward /ð/ in some Koine and later contexts, while theta followed its own path from aspirated /tʰ/ toward /θ/. These are recurrent possibilities for sounds made in neighboring dental positions, but they do not by themselves establish a shared word origin. Establishing cognacy requires systematic correspondences, chronology, and evidence of contact beyond a single pair of words. The standard Greek lexicon treats manthanō within an Indo-European word family; the historical connection proposed here therefore remains a question rather than a conclusion. Greek alphabet; Biblical Hebrew pronunciation; Greek lexicon: manthanō. ↩︎
Isaiah’s stump and Daniel’s tree. Geza is Isaiah’s word for the stump or stock of Jesse in Isaiah 11:1; it also appears in Job 14:8 and Isaiah 40:24. Daniel’s tree vision uses a related image, but Daniel 4 is in Aramaic and says ʿiqqar sharshowhi—“the stump of its roots”—with a different word for stump or stock. The images are thematically close without being one repeated technical term. Isaiah 11:1 Hebrew analysis; Daniel 4:15 Aramaic analysis. ↩︎
Kimhi’s evidence and its limit. Kimhi’s entry for gezaʿ describes a cut tree whose root remains in the ground and gathers passages including Isaiah 11:1. In his entry for shoresh, he analyzes taking root and being uprooted across verb patterns, explicitly distinguishing a grammatical root from a binyan: words can share a pattern without sharing a root. These entries show tree vocabulary and pattern analysis in the same medieval work; they contain no menorah diagram. Kimhi, entry גזע; Kimhi, entry שרש. A related caution belongs here: Kimhi’s introduction speaks of correcting the distortions of students’ mouths and lips in reading. That wording echoes Proverbs 4:24, not Isaiah 6, and concerns accurate pointing and pronunciation from reliable manuscripts rather than Isaiah’s cleansing vision. Introduction, paragraph 6. Kimhi wrote about a century and a half after the Leningrad Codex (1008–09), the oldest complete dated Hebrew Bible now surviving. He was not a Masorete or a contemporary of the formative Tiberian Masoretic circles, but he belonged to the same broad medieval culture of preserving, reading, and explaining the Hebrew text. His writings became an important bridge of preservation: Mikhlol and Sefer HaShorashim remained widely read, carrying forward earlier grammatical work whose original treatises became less accessible or were forgotten. Leningrad Codex record; history of Hebrew grammar. ↩︎
Six, eight, and seven. The Ben-Yehuda dictionary reports Kimhi arranging verb forms in eight columns. William Chomsky’s edition of the Mikhlol gives the medieval sequence as Kal, Nif’al, Pi’el, Pu’al, Hif’il, Hof’al, Po’el, Hithpa’el; it credits Moses Kimhi, David’s older brother and teacher, and describes the arrangement as followed with modifications into the present. Ben-Yehuda, binyan; Chomsky, David Kimhi’s Hebrew Grammar (Mikhlol). The extra form is Po’el, commonly treated in later teaching as a variant associated with particular root shapes rather than a separate pattern. Earlier grammarians such as Hayyuj, Ibn Janah, and Ibn Ezra counted only six, while Joseph Kimhi gave Pu’al and Hof’al distinct standing. The changing counts show that the familiar seven is a teaching and classificatory choice, not a sevenfold arrangement demonstrated by the medieval sources. ↩︎