The Debate Over Hebrew Roots

Interactive Learning Module

Machberet Menachem

• This module will explore a long-standing debate about the nature of Hebrew "שרשים" or roots, the most basic building blocks of Hebrew words.
• Today, it is widely accepted that all Biblical Hebrew verbs have tri-consonantal roots which encapsulate the basic meaning of the verb, and to which prefixes and suffixes are added to mark person, gender, voice, tense etc.
• Thus, for example, "כתב" lies at the base of all the following verbs: כָּתְבוּ, אֶכְתְּבֶנָּה, כָּתַבְתִּי, תִּכְתְּבוּ, all related to writing.It also relates to the nouns כְּתָב, כְתֹבֶת and מִכְתָּב.
• However, there is also a school of thought which disagrees and maintains that roots can be as small as one letter.
• What lies at the heart of the argument?
• In the coming slides we will trace how various grammarians and exegetes from Menachem ben Saruk to Malbim have understood the workings of Biblical Hebrew roots, attempting to answer the following questions:
• Which commentators assume that roots must be tri-consonantal and which think they are bi-literal?
• What is motivating each side to say as they do?
• Finally, and perhaps most importantly, why does it matter? How do the different theories affect one's understanding of individual Biblical words and exegesis as a whole?

• This slide will briefly introduce some concepts and terminology which might be necessary to fully appreciate the following discussion.
• The terms are those used in modern grammars, and even though medieval grammarians might not have used the same terminology and might have thought about roots differently, they will provide a framework (and language) for comparing earlier and more modern root theories.For a more thorough discussion of the concepts introduced here, see Verbs.
• Strong and weak verbs
• Strong verbs are those whose root letters are preserved in all forms of the verb (i.e they do not get lost when the root is inflected). They are also referred to as "פעלים שלמים" (complete verbs) and include roots such as שמר, כתב and גדל.
• Weak verbs are those which contain at least one root letter which will either get dropped or be unvocalized in certain inflections; these will usually be one of the אהו"י letters, a doubled root letter, or an introductory "נ". Such verbs might also be called פעלים נחים or חסרים (silent or absent verbs).
• Root position and verb class (גזרות)
  • In modern Hebrew, roots are often classified based on the position of the letter within the root. One can thus refer to a letter as being in the Root I, II or III position (from right to left), or by assigning each consonant a place based on the word "פעל" ("he acted"). The first letter of the root is assigned to "פ", the second to "ע" and the third to "ל". For example, a פ"א verb is one which opens with an "א", while an ע"ו verb is one with a "ו" as the second radical.
  • The weak verbs are thus classified based on their root position as being either: פ"א, פ"ו/י, פ"נ, ע"ו/י, ע"ע, ל"ה/י, ל"א verbs. All roots of the same class (גזרה) have similar inflection patterns.

• Early grammarians posited that roots could consist of one, two, or three (or more) letters, with the most well-known exponent of this school being Menachem ben Saruk (10th century, Spain).
• Machberet Menachem – Menachem is credited with being the first to produce a Hebrew-Hebrew lexicon, today known as "Machberet Menachem".His choice to write in Hebrew rather than Arabic made his work accessible also to non-Arabic speaking audiences. As such, it was used by many commentators, even those outside of Spain, including Rashi and his school. The work was organized by roots, or in his terminology, by a word's "יסוד" (foundation),He also uses the terms שרש and עיקר. which might contain from 1-5 letters. Each entry opened with the root and was followed by a list of verses containing that root, subdivided into semantic categories (מחלוקות) which exemplified the various possible "definitions" of the root.Unlike modern dictionaries, the lexicon does not usually contain explicit definitions; the reader was meant to understand the word from the context of the various verses cited in each subcategory. Only when a word is rare or hard to define did Menachem add a short explanation.
• See a sample entry here and here.
• What constitutes a root? – Though today most students of Hebrew take the concept of roots for granted, in the tenth century, the idea had not yet been applied scientifically to the study of the Hebrew language. Menachem, thus, had to determine on his own by which criteria a letter is considered to be part of the root.Though medieval Arabic grammarians were already studying roots as they related to their own language, Menachem was loathe to make use of cognate languages to understand Tanakh, and thus did not actively apply Arabic grammar to Hebrew.
• To highlight the difficulty, look at the word רַצּוֹתִי in Shemuel I 12:3. Assuming that there is no precedent to know how many letters a root must have, and there is of yet no system categorizing verbs into פ"י, ל"ה, ע"ו etc. classes, how is one to determine the root? How would one choose between רץ, רוץ, רצה, רצץ or even רצו or רצות‎?Commentators actually debate what is the root of this word, with most suggesting it is רצצ (to crush) and others suggesting it is רצה (to desire). Compare Rashi and R"Y Kara with Abarbanel and Malbim here.
• Menachem's system – Menachem suggested that in order for a letter to be considered a root consonant, the consonant has to exist in all forms derived from the root. If any letter is missing in any individual inflection, it must not be an integral part of the root for the basic meaning of the stem can be expressed without it.It is important to note, however, that Menachem's main goal was probably not to develop a grammatical theory of roots, but rather to formulate a principle through which he could organize his dictionary. As such, it is possible that had his goal been different, or if had been focused purely on grammatical considerations, he might have reached other conclusions.
• Uni-literal or bi-literal roots – According to him, then, a significant number of verbs are assumed to have just two (or even one) root letters.
• For example, he posits that the root meaning "to strike" (today assumed to be "נכה"), is simply "כ", for the other two consonants get lost when the verb is inflected.
• Similarly, he suggests that the root meaning "to rise" (today, assumed to be "קום") is bi-literal, "קם", for only those two letters are present in all conjugations.
• Semantics – In Menachem's system, as a root is determined by technical considerations, it often carries no independent semantic meaning. As such, a single root could often take a wide variety of unconnected definitions.
• For example, see the entry "שב" in Machberet Menachem, where he points to seven different usages of the root, each in its own "מחלקה" (semantic division). Included in these are what today would be considered the unrelated roots שבה, נשב, ישב, שוב, each with its own meaning (capture, blow, sit, return).
• Similarly, see the entry "כל" which contains fourteen (!) subcategories, including what today would be considered the roots כלה, כלא, כול, נכל, יכל, (to finish, restrain, contain, plot and be able) alongside nouns such as כלה, כלי and כל (bride, vessel and entirety).

• Dunash ben Labrat, a fellow Spanish grammarian of the 10th century and sharp critic of Menachem, questioned some of Menachem's roots and classifications, but he did not fundamentally disagree with his theory or systematically develop an alternative.
• Bi-literal roots – In most cases, Dunash does not actively argue against Menachem's suggested roots, but rather agrees that roots might be bi-literal.
• See, for instance, Dunash's analysis of the relationship between the noun "יגון" and what is today considered the verbal root "יגה", where he notes that: "ויגה מן האחדים" (that the word "יגה" has a uni-literal root, "ג"), clearly agreeing that sometimes there can be one and two letter roots.Similarly in the realm of nouns, though he attacks Menachem for assuming that the nouns "מסוה" and "סתה" are connected in meaning and both relate to the same one letter root "ס", Dunash himself does not suggest a tri-literal root for either, but two bi-literal roots, claiming that "מסוה" is related to "סו" and "סתה" to "סת".
• Disagreement – Elsewhere, though, Dunash contests Menachem, arguing that a number of the bi-consonantal roots suggested by him are actually tri-literal.
• For example, see his criticism of Menachem's suggestion that the root meaning to be tired (today assumed to be יגע) is simply "גע". Dunash counters that the "י" is actually part of the root, which has three, not two, letters. He explains that even though the "י" falls in some inflections, when it does so, it is replaced by a "ו" (as in: הוֹגָעְנוּ) and thus should not be considered absent.Elsewhere, too, he recognizes that the "י" of what today we would call פ"י verbs is part of the root. Thus, he categorizes יבש also as a tri-literal root. Here he explains that even though in the third person imperfect (יֵבוֹשׁ) the yud seems to be lost (for it is spelled with just one yud while one would have expected two, one for the prefix and one for the root), this is only because the yud plays double duty.
• The disagreement, then, is not fundamental, for it relates not to the essence of Menachem's theory but only to its application, i.e. how one defines whether or not a letter has been lost in inflection.
• Grammatical awareness – Despite the overall similarity, though, it does seem that Dunash might have been more attuned to some of the unique verbal conjugations of "weak roots" than Menachem. For example, he recognized that what we call a ל"ה verb is inflected differently than an ע"ע verb, and that therefore, even if they share the same bi-literal root, the two should not be placed in the same subclass. This recognition is significant, as it might impact meaning in addition to form.See H. Englander, "Rashi's View of the Weak ע"ע and פ"נ Roots" who notes that this awareness might be what led Ibn Ezra to suggest that Dunash "was the first to begin to awaken from the deep sleep of ignorance touching the nature of weak roots." It should be noted however, that Menachem, too, usually put weak verbs of the different classes in different divisions of the root. In addition, sometimes when Dunash disputes Menachem's divisions, it is not clear if he is indeed motivated by knowledge of different inflection patterns or simply because he thinks context does not support a certain understanding.
• See, for example, Dunash's discussion of the verb "רַצּוֹתִי" in Shemuel I 12:3. Though Dunash agrees with Menachem that its root is רץ, he questions Menachem's grouping of the verb in the same division as those verbs which today would classified under the root "רצה" (meaning to desire). He suggests that it should instead be in the same division as verbs which today we would classify as having the root רצץ (to oppress). He notes not only that the context does not support Menachem's understanding but also that the conjugation does not match that of other verbs in the class.
• That said, the vast majority of Dunash's disagreements with Menachem relate not to the number of root letters or even their subdivision but to specific translations which he thought could lead to misinterpretations of Jewish law or heresy.
• See, for example, his attack on Menacheem's suggestion that the root "מלק" means "to slaughter", with Dunash claiming that Menachem preferred the Karaite interpretation over the Rabbinical one.
• Somewhat ironically, then, it was not Dunash, but Menachem's own student, Yehuda ben David Hayyuj, known as the "father of scientific Hebrew grammar", who pushes back against his teacher and is credited with developing the alternative tri-consonantal theory.

• Criticism of Menachem – Yehuda Hayyuj was a disciple of Menachem and is possibly the "Yehuda ben David" included in the threesome of students who defended Menachem against Dunash's criticisms.As we saw, Dunash was a sharp critic of Menachem; these attacks prompted several students to defend their teacher in a polemical composition known as "תשובות תלמידי מנחם על דונש בן לברט". It was authored by Yitzchak Ibn Kapron, Yitzchak Ibn Chiquitilla, and Yehuda ben David, but it is debated if the latter is identical to Yehuda Hayyuj. For discussion, see A. Gaash, "תשובות תלמידי מנחם על דונש: אילו תשובות כתב כל תלמיד, והאם היה חיוג’ אחד מהתלמידים",Lĕšonénu: A Journal for the Study of the Hebrew Language and Cognate Subjects (2019): 297–318, and the literature cited there. He, however, was also aware that Menachem's system had certain deficiencies, related to both semantics (meaning) and morphology (word form):
• First, as we saw, in Menachem's system, not all roots contained independent meaning, and thus a single root could have many vastly different interpretations.
• Second, each root had a number of different, unrelated, inflection patterns, with, for example, "שב" (meaning ישב) and שב (meaning שוב) being conjugated totally differently from each other, and yet still considered one root.
• A new theory – Yehuda Hayyuj, thus, set out to develop an alternative, more systematic theory of roots. His knowledge of Arabic grammar allowed him to adopt and apply its grammatical guidelines to Hebrew, though he rarely explicitly compared the two. He recognized several phenomena which transformed the study of Biblical Hebrew roots:
• Weak consonants and the "נח נסתר" – When comparing the inflections of words with the weak vowel letters אהו"י , Hayyj noted that sometimes the letters appear and are pronounced, sometimes they appear but are unvocalized (a נח נגלה), but sometimes, too, they can be hidden and not appear at all (נח נסתר). Compare, for example, the word דַּיִשׁ where the yud is consonantal, the verbs "לָדוּשׁ" and "בְּדִישׁוֹ", where it is found but not vocalized (a נח נגלה), and the verb דַשְׁתִּי, where it is hidden (נח נסתר). Sometimes, when the letter does not appear, it will be replaced by a long vowel, as in "אֵשֵׁב" (from ישב). This, Hayyuj taught, does not preclude their being one of a root's radicals.
• Assimilation – This refers to the process by which certain letters like "נ" are at times absorbed into the sound of the following consonant. In contrast to weak letters which might "disappear" totally, these are marked by a strong dagesh.Thus the imperfect of נפל is "יִפֹּל"; the assimilated "נ" does not appear in the word, but is marked by a dagesh in the "פ". The same phenomenon occurs with roots with doubled root letters, such as סבב. When inflected, the first "ב" might be absorbed by the second and marked by a dagesh as in: סַבּוֹתִי.
• Metathesis – This refers to the reversing of consonantal place, as seen in many hitpa'el forms such as "נִּצְטַדָּק" where the "ט" and "צ" chaמge places, making it more difficult to discover the root.
• Three letter roots – These concepts transformed the way one looked at roots, for now it became obvious that what often appeared in Hebrew to be two or one letter roots were actually three. Just because a letter does not appear in all inflections of a word, does not mean it is not central to its core.For further discussion of Hayyug's contributions to the study of Hebrew grammar, see D. Sivan, “Biblical Hebrew Roots and Quiescents According to Judah Ḥayyuj’s Grammatical Works.” Hebrew Union College Annual 60 (1989): 115–27 and A. Wattad and D. Sivan, " הבחנות ייחודיות של חיוג’ בחיבוריו בתחומים: פונולוגיה, מורפולוגיה ותחביר ותרומתן לחקר דקדוק לשון המקרא", Beit Mikra: Journal for the Study of the Bible and Its World, (2021): 144–77.
• Inflection patterns – Hayyuj further noted that verbs could be classified by their inflection patterns, for families of verbs tend to act similarly when conjugated. He thus recognized the morphological classes (גזרות) familiar to all today as פ”ן, פ“א, פ”י (פו"י), עו“י, ע”ע, ל“ה (לו"י) verbs. It is these patterns that determine a word's root.
• To demonstrate and substantiate his theories, he wrote two works, "The Book of Verbs Containing Weak Letters." and "The Book of Verbs Containing Double Letters". These contain lists of verbal roots with weak or doubled letters and the principles that their govern their inflections.
• His work was further developed by his students R. Yonah ibn Janach and R. Shemuel haNaggid. The former composed Sefer HaRikmah, a grammatical work, and Sefer HaShorashim, the first lexicon written with the three-letter root concept. The works of Shemuel HaNaggid have not survived and are known mostly through citations of Ibn Janach, Ibn Ezra and Radak.

• Hayyuj's work spread quickly, and within one generation, virtually all Spanish commentators adopted his tri-consonantal system. However, as Hayyuj's books were written in Arabic, they were initially inaccessible to those from non-Arabic speaking lands.
• Thus, in France, for example, commentators such as Rashi (11th century) continued to follow Menachem's theory that words could have one or two letter roots.
• See Rashi on Bereshit 15:17, Bereshit 49:19 and Shemot 1:20 where, in each verse, he points out that the verbs under discussion are those "שיסודם שתי אותיות" (whose roots are bi-literal), clearly adopting the two-letter system. [The comments refer to roots that we would today call ע"ו, ע"ע and ל"ה verbs, all of which Rashi categorizes here as being bi-consonantal.]
• Nonetheless, like Dunash before him, Rashi disagrees with Menachem on some points, and, for instance, maintains that what we would today call פ"י roots are tri-consonantal.See above, that Dunash noted this with regards to these verbs as well.
• See, for example, the end of his comments on Shemot 1:20, above, where, unlike Menachem, he recognizes that the "י" in words like ירד, יצא and ישב are part of the root.
• As such, though Rashi is clearly influenced by Menachem, the degree of influence is debated, with some suggesting that differences are minor and others claiming that Rashi actually diverged significantly, coming much closer to a tri-consonatal theory (even if unaware of Hayyuj).See H. Englander ibid, I. Eldar, "תפישת השורש של חכמי צרפת הראשונים", Leshoneinu 72 (2011):487-502 and Ch. Mirsky, "שלוש סוגיות בחכמת הלשון מבית מדרשו של מנחם בן סרוק" who all conclude that Rashi did not make distinctive contributions beyond his predecessors. Cf. Ch. Gamliel, "תפישת השורש של רש"י", Leshoneinu 71 (2009):105-129, who disagrees.
• Part of the dispute revolves around the term "יסוד נופל" (or: עיקר נופל) used by Rashi to describe certain letters which appear in some but not all of a word's inflections.
• See Rashi Bereshit 49:10 (ד"ה ולו יקהת עמים) where he discusses the concept.In this comment he discusses the idea in reference to nouns, but the principle is the same when applied to verbs as well. Rashi notes that in the word יקהת, the yud "is a fundamental in the root, but sometimes it drops from it" (עיקר היא ביסוד... ופעמים שנופלת הימנו), pointing to several other letters, like "א" and "נ" which often act similarly.
• Menachem states a similar principle with regards to the letters תימנה, noting that "הן נראות כאותיות יסוד אך אינן כן משום שיש היקרויות שבהן האותיות הללו נופלות מהמילה". However, while Menachem is explicit that such letters do not actually comprise part of a word's root, Rashi is ambiguous, with some of his comments implying that he does count them and other implying the opposite.
• See, for example, Rashi Shemot 3:22 regarding the word "וְנִצַּלְתֶּם" where he explicitly writes, "והנו"ן בו יסוד", though he categorizes it as a letter that sometimes drops ("שהיא באה בתיבה לפרקים ונופלת ממנה"), comparing it to other similar words such as ונשאתם, ונתתם, and ונמלתם. [See similarly Bereshit 33:12.]
• Compare, however, Rashi Bereshit 17:1 where he discusses the word וּנְמַלְתֶּם, writing, "והנו"ן בו יתירה ליסוד, הנופל לפרקים", saying that the "נ" is as addition to the root!
• Elsewhere, too, Rashi's stance is debated, as some of his positions appear contradictory.
• See, for example, his comments on Shemot 1:20 mentioned above, where he exemplifies two-letter roots by pointing to what today would be called ל"ה verbs (פנה, גלה), implying that all such roots are bi-literal. Compare, though, his comments on the word "מְשִׁיתִהוּ" in Shemot 2:10 where it seems as if he assumes that its root is the tri-consonatal "משה", writing: "מגיזרת תיבה שפעל שלה מיוסד בה״י בסוף התיבה". Here, too, he compares it to other so-called ל"ה verbs such as בנה and עשה, implying, in contrast to the above, that they are all tri-literal.
• Given that many of Rashi's comments are open to interpretation, it is difficult to claim conclusively that Rashi moved significantly beyond Menachem and Dunash in recognizing three-letter roots, but as seen throughout the above examples, he does exhibit insight into the various classes of weak roots and how their inflection patterns differed from one another.

• ספר הדייקות – Rashbam's understanding of Hebrew's root system can be gleaned from his ספר הדייקות, the first grammatical work written by a Jew of Asheknazic lands.The work contains two parts: 1) a grammatical treatise of eight chapters discussing various families of roots, the construct state (סמיכות), masculine and feminine forms, vocalization and more 2) a grammatical commentary to Tanakh. In the heading to the second section, Rashbam expresses his intention to cover all 24 books of Tanakh but only his comments until Bereshit 7:5 have survived.
• Relationship to Hayyuj – From both the work and Rashbam's Torah commentary it is clear that, in contrast to Rashi, he firmly adopts a tri-consonantal theory. However, his application of the concept differs from that of Hayyuj and it is not clear whether and to what extent he was influenced by the Spaniard's works.On the relationship between the two works, potential influence and where Rashbam is both similar to and different from Hayyuj, see R. Merdler, "רבי שמואל בן מאיר והדקדוק העברי", PhD dissertation (Jerusalem, 2005): 289-303 and I. Eldar, ibid. Rashbam never mentions Hayyuj by name and it is unknown if he had his books. Nonetheless, it is very possible that even if he never read them, he knew of the principles laid forth therein, for by the time Rashbam wrote ספר הדייקות, R. Moshe ibn Chiquitilla (11th century) had already translated Hayyuj's work into Hebrew. Moreover, Ibn Ezra also wrote a translation in Rashbam's lifetime and spread Hayyuj's theories wherever he traveled, including France. However, as the date of composition of ספר הדייקות is debated, one might question whether it was written before or after Ibn Ezra's translation and travels. Merdler assumes that it is specifically Ibn Ezra's arrival in France and his spreading of Hayyuj's theories that prompted the composition. For an alternative view of the dating, see י. עופר, "מתי נכתב "דייקות", ספר הדקדוק של רשב"ם?", שנתון לחקר המקרא .והמזרח הקדום י"ז (תשס"ז):233-251.
• Uniqueness – One of the major differences between Rashbam and Hayyuj relates to the methodology by which each determined a word's root. While Hayyuj focused on the unique inflection patters of each class of roots, it seems that Rashbam looks at just one concrete form, the verb's perfect, third person, masculine singular form, seeing in it the verbal root.See Merdler and Eldar, ibid. As such, he divides roots into two main categories: those of two letters (the minority) and those of three letters (the vast majority).
• Bi-literal roots – These includes only one class of roots, what we know of as ע"ו or ע"י roots which take two letters in third person (קם, בא).See Merdler, p.181, that Rashbam does not view these as "weak" verbs since they maintain all their root letters throughout all their inflections, and so they might be seen as parallel to the tri-consonatal "שלמים" which function similarly; both are complete or "strong" verbs, the only difference being the number of original root letters.
• Tri-literal roots – These includes all other roots: the tri-consonatal "שלמים" whose root letters appear across all conjugations, and what Rashbam refers to as "הפעלים החטופים", those tri-literal roots which contain a letter that sometimes is lost in inflection. These include what we would call today פ"נ, פ"י, ל"ה, ע"ע verbs and the גזרות המורכבות (those like נטה, with two letters which might be lost in inflection). Rashbam notes that some of these disappear totally, while others are marked by a dagesh (as when נפל becomes "יִפֹּל") or through a variant letter (as when the "י" of "ישב" becomes a "ו" in "הושיב").
• Examples – See the following comments of Rashbam where he discusses various two or three letter roots. The examples show how he was aware of the verbal classes that we would refer to today as: ע"ו,and פ"י, ל"ה, ע"ע, פ"נ roots, recognizing their individual inflection patterns and how correct categorization impacts meaning.
• See Shemot 2:1, where Rashbam questions whether the root of "וַתְּנִיקֵהוּ" is the two letter "נק" (what we would call an ע"ו verb) or the three-letter "ינק".
• See Shemot 12:4 where he discusses the relationship between the noun "מִכְסָה" and the verb "תָּכֹסּוּ", suggesting that while the former's root is "מכס" the latter is from the tri-literal root "כסס" (noting that the word has the same inflection pattern as "סבב").In other words, he suggests that despite the similarity in meaning and overlap in letters, the noun and verb stem from different roots. Here, recognition of the different roots does not impact meaning, but one can see how it might have. [He also raises the possibility that "מכסת" could stem from the root "כסה", among "חטופי למ"ד פעל של ה"י" (what we would call a ל"ה verb), but without sharing what the word would then mean. He supports the possibility by noting that when ל"ה verbs are inflected to become nouns they take the paradigm of "מכסת". Hollow verbs and doubled root verbs have a different noun paradigm, which is further support for his claim that the root of the noun is not כסס.]
• See Dayyakot LeRashbam Bereshit 3:15 where Rashbam asserts that the root of יְשׁוּפְךָ/ תְּשׁוּפֶנּוּ is "שף" (to crush or bruise) and not "נשפ" (to blow) or "שפה" (to lay bear), questioning both Menachem and Rashi's categorizations.Menachem places the verb under the two letter root "שף", but includes it in the same subcategory of roots which we would refer to as ל"ה verbs, suggesting that it takes the meaning of today's "שפה" (lay bare). According to Rashi, the words "יְשׁוּפְךָ" and "תְּשׁוּפֶנּוּ" have different meanings, with the text intentionally playing between the two; the first means to bruise and the second to blow (or hiss). Rashbam disagrees, noting that if it were categorized to mean "נשף", there should not have been a "ו" in "תְּשׁוּפֶנּוּ".
• Rashbam's theory of roots clearly moves beyond Menachem, coming much closer to that of Hayyuj, even if he differs in details, terminology, and methodology. Interestingly, in some ways he is most similar to Ibn Ezra, who fundamentally follows Hayyuj, but like Rashbam, held that ע"ו verbs are two-letter roots.
• Rashbam's variation, however, was overshadowed by Hayyuj's system which, despite minor tweaks and disputed points, was adopted and became the standard used by all, including Ibn Ezra, Radak and others until today. For centuries, hardly anyone reverted to the bi-literal theory until it was revisited in the nineteenth century.

• In the nineteenth century, a variation of the two letter root theory surfaced, becoming popular among German linguists and commentators.
• Bi-literal theory – The pioneer of the revisited theory was R. Shelomo Pappenheim (רש"פ), who criticized the notion that roots need to be three letters, proposing instead that at the core of all three letter morphemes (what he calls "שמות")This could be a noun or adjective, or simply a sequence of three letters which itself acts much like a root and from which other verbal forms or nouns etc. can be derived. are really roots of two (or one) letters which encapsulate some fundamental meaning. The extra letters added to the root allow for variation and nuances of meaning, but all words which share a core root also share the same semantic field.
• The fundamentals – He lays out the fundamentals of his theory in the introductions to his works יריעות שלמה, devoted to differentiating between synonyms, and חשק שלמה, a book of roots,For discussion of the two works and an overview of Pappenheims's methodology, influences and impact, see Y. Wormser, "The study of Hebrew synonyms in the early Haskalah period", Journal of Jewish Studies 74 (2023): 74-91. where he describes how new words are derived from original bi-literal roots:
• Often, one of the auxiliary letters, האמנתי"ו, will be added either in the beginning, middle or end of a certain root to create a new semantically related word, but nonetheless one which has its own unique meaning.He notes that the אהו"י letters might be added in any position, while the letters מנ"ת will be added only before or after the biliteral root.
• For instance, the root "בל" whose essential meaning relates to confusion and loss of form can become נבל (to decay or languish), בלה (to wear out), הבל (referring to something which is evanescent), בהל (to be dismayed to the point of confusion), בלי (without), and more.
• Elsewhere, a three letter word is really comprised of a combination of two roots, each contributing to the word's ultimate meaning.
• For example, בגד (garment), is a compound of בא (come) and גד (bundle, as in: אגד). When the two are joined, they mean: to come into a bundle [of clothing]. The verb "ברח" combines בא (come) and רח (related to רוחה, expanse), for someone who flees goes from straits to expanse. The word "חרף", meaning to taunt, belittle or curse another out of anger, stems from the two roots "חר" and "אף" (to be kindled with anger)
• When, instead, a word has a doubled letter (e.g. סבב, פלל), the doubling is meant to show intensity of some sort.
• For example, R. Pappenheim suggests that the root "אף" relates to the word "face" while the doubled "אפף" has an intensified quality, meaning "to be surrounded" (to be faced on all sides).See the entry "אף" in חשק שלמה.
• As seen, R. Pappeneheim's approach differs greatly from that of his predecessors. While they focused on morphology (looking at word structure and verbal inflections), he focused on etymology and semantics (how meanings of words evolve).
• R. Pappeneheim was likely influenced by the German lexicography of his day, especially by the "Stammwörter principle" which suggests that many German words are derived from some basic element to which affixesThis refers to the various grammatical elements that might be added to a word to modify its meaning or function, including prefixes, suffixes and infixes or other words are joined, an approach quite similar to his own.See Y. Wormser, ibid, p. 86-91.
• Another motivator might have been R. Pappeneheim's belief in the omnisignificance of Torah and his conception that every Hebrew word must have distinct meaning.This is what promted him to write יריעות שלמה which differentiates between seeming synonyms. He rejected the idea that any two words can be truly synonymous, for why would "the Creator designate two different roots to convey the same idea"? Positing that the same root lies at the heart of many related words, then, obviated the problem. All similarity stems from the fact that the two words share a root, not because Hashem "redundantly" made two verbs with the same meaning.
• Somewhat ironically, perhaps, the method often forced R. Pappenheimer to posit connections and note similarities also between words which otherwise seem quite unrelated. And, at times, this need to show relationships between words with shared roots necessitated quite a bit of creativity.
• Thus, for example, in his discussion of the root פל (what he suggests relates to falling), he connects the words נפל (to fall), פלא (wonder, something which "falls out of the natural realm"), אפילה (darkess, which comes due to falling vapors), and פלל (to make a judgment fall upon something).

• Several commentators, including R. Mecklenburg (HaKetav VeHakabbalah) and Malbim, were heavily influenced by R. Pappenheim<,fn>HaKetav VeHakabbalah mentions him by name 180 times! both summarizing and implemented his method in several places in their commentaries.
• See, for example, HaKetav VeHakabbalah Vayikra 5:4 where he briefly reviews the essence of R. Pappenheim's approach, giving several examples where words share a bi-literal root and thus have similar meaning. He compares קצה and קצע, both derived from קץ and related to cutting,One might similarly connect קצר, קצב, and קצץ, all also related to cutting off. גלה (to reveal) and גלע (to lay bare), both relating to the root גל, something circular that can be rolled [away], and גבעה (hill) and גבה (to grow tall) from the root גב, referring to something which extends upwards.
• Malbim, too, explicitly summarizes the approach in several places in his work. See, for example, Vayikra 2:14 and 7:13 where, among other things, he points to many examples of compound roots, writing: "חלף חרף טרף שלף חנף מורכבים מן חל אף, חר אף, טר אף, של אף, חן אף". In each word, the root "אף" is joined to another root to make a compound word.
• Elsewhere, HaKetav VeHakabbalah utilizes the theory to understand a root in a less conventional way so as to answer a theological or textual difficulty.
• For example, he is uncomfortable with the suggestion that Sarah's words "לֹא יִירַשׁ בֶּן הָאָמָה הַזֹּאת עִם בְּנִי" in Bereshit 21:10 comprise a request that Yishmael be disowned from all property (as might be understood according to the standard definition of "יירש", to inherit). Thus, he posits that really Sarah is instead asking that Yishmael not have authority over or act as superior to Yitzchak, noting that "יירש" stems from the root "רש", which actually means "head" (related to ראש).
• Malbim often uses the approach not just to understand the meaning behind and relationship between words, but to understand Rabinnic Midrashim, for he assumes that this bi-literal approach to roots was shared by the Sages.See the sources above and the opening paragraphs of his work Ayelet HaShachar where he lays out the grammatical principles which he thinks underlies Rabbinic exegesis.
• • For example, he explains that Bereshit Rabbah's statement "וַיִּפְצַר בָּם מְאֹד – הִכְנִיס בָּם אַף וְצָרָה" can be understood if one recognizes that the sages are assuming that פצר is a compound word made of "אף " and "צרה". Similarly, he points to the dispute over the meaning of the word "פרך" in Shemot 1:13, noting that the opinion that suggests that it refers to "פה רך" (a "soft mouth", i.e. via enticement) is understanding it to be comprised of two individual roots.

• This module has attempted to trace the development of the various theories of Biblical Hebrew roots from Menachem b. Saruk until today.
• Menachem ben Saruk and his school assumed that roots might be anywhere from 1-5 letters, determining that a root includes only those letters which appear in all of a verb's inflections.
• Dunash and Rashi, to varying degrees,See the discussion above regarding to what extent Rashi moved beyond Menachem. followed in his path, though both exhibit greater recognition of certain inflection patterns.
• Menachem's system had two main drawbacks: it created a situation where the same root might have multiple unconnected meanings and multiple different inflection patterns. These two deficiencies were addressed by the alternative root theories offered by Hayyuj and R. Pappenheim.
• Hayyuj, focusing on morphology, recognized how weak letters might lose vocalization, be hidden, or assimilate into the following letter. This led him to conclude that all roots really comprise three letters, and that these roots can be classified according to shared inflection patterns (today's various גזרות), for roots with weak letters in the same positions act similarly. As such, looking at a word's inflection pattern can help one determine its root.
• His theory gained wide acceptance, with generations of grammarians from Ibn Janach, Ibn Ezra and others until today adopting his system. Rashbam's similar theory might also have been influenced by him.
• R. Pappenheim focused instead on semantic etymology, looking to find the shared meaning in all roots. He concluded that most words are comprised of small building blocks of meaning of one or two letters (the root) to which one of the letters האמנתי"ו might be affixed, another root might be conjoined, or a root letter be doubled, to create other morphemes in the same semantic field, but with individual meaning.
• His theory, too, attracted many followers, with two of the most well-known being Malbim and R. Mecklenburg.
• The various theories affect how one analyzes and thus how one understands various words, impacting Biblical exegesis.

• Return to the beginning of this module.
• Click here to proceed to the first module on verbs, which introduces rott classes, binyanim and various conjugation patterns.
• Access additional grammar topics and modules.
• Access additional Interactive Learning Modules.