Proto-Semitic Language and Culture

John Huehnergard

The Appendix of Semitic Roots (Appendix II) that follows this essay is designed to allow the reader to trace English words derived from Semitic languages back to their fundamental components in Proto-Semitic, the parent language of all ancient and modern Semitic languages. This introduction to the Appendix provides some basic information about the structure and grammar of Semitic languages as an aid to understanding the etymologies of these words. In the text below, terms in boldface are Semitic roots that appear as entries in Appendix II. Words in small capitals are Modern English derivatives of Semitic roots. An asterisk (*) is used to signal a word or form that is not preserved in any written document but that can be reconstructed on the basis of other evidence.

   1

The Semitic Language Family

The Semitic language family has the longest recorded history of any linguistic group. The Akkadian language is first attested in cuneiform writing on clay tablets from ancient Mesopotamia (modern Iraq) from the mid-third millennium B.C., and Semitic languages continue to be spoken in the Middle East and in northeastern Africa today.

   2

      Modern Semitic languages include Arabic, spoken in a wide variety of dialects by nearly 200 million people as the official language of over a dozen nations, and in many other countries as well; Amharic, the official language of Ethiopia; Hebrew, one of the official languages of Israel; Tigrinya, the official language of Eritrea; Aramaic, the language of the Jewish Talmud and of Jesus, first attested in inscriptions written three thousand years ago and still spoken by several hundred thousand people in the Middle East and elsewhere.

   3

      Ancient Semitic languages include Akkadian, the language of the ancient Babylonians and Assyrians; Phoenician and its descendant Punic, the language of Carthage, the ancient enemy of Rome; the classical form of Hebrew as recorded in the Hebrew Scriptures and later Jewish writings; the languages of the neighbors of the ancient Israelites, such as the Ammonites and Moabites; many early dialects of Aramaic; the classical Arabic of the Koran and other Muslim writings; Old Ethiopic texts of the Ethiopian Christian church; and South Arabian languages attested in inscriptions found in modern Yemen, such as Sabaean, the language of the ancient Sheba of the Bible.

   4

      In the same way that English is a member of the sub-family of Germanic languages within Indo-European, the Semitic languages constitute a sub-family of a larger linguistic stock, formerly called Hamito-Semitic but now more often called Afro-Asiatic. Other branches of Afro-Asiatic include ancient Egyptian (and its descendant, Coptic), the Berber languages of north Africa, the Cushitic languages of northern East Africa (such as Somali and Oromo), and the Chadic languages of western Africa (such as Hausa in Nigeria).

   5

      Various significant linguistic features allow us to classify the many Semitic languages in a way that shows the historical branching off of sub-groups. The ancient ancestor of all the Semitic languages, like Proto-Indo-European a prehistoric, unwritten language, is called Proto-Semitic or Common Semitic. The earliest branching, which includes most of the known Semitic languages, is called West Semitic; the part that remained after this branching, East Semitic, essentially includes only Akkadian. West Semitic comprises three branches: the modern South Arabian languages; the ancient and modern languages of Ethiopia; and Central Semitic. Central Semitic is further subdivided into the South Arabian inscriptional languages; classical, medieval, and modern forms of Arabic; and the Northwest Semitic languages, which include Hebrew and Aramaic. See the “Chart of the Semitic Family Tree”.

   6

Semitic Words, Roots, and Patterns

A distinctive characteristic of the Semitic languages is the formation of words by the combination of a “root” of consonants in a fixed order, usually three, and a “pattern” of vowels and, sometimes, affixes before and after the root. The root indicates a semantic field, while the pattern both narrows meaning and provides grammatical information. For example, if we represent the three root consonants abstractly as X’s, in Arabic the pattern XaXaXa produces a verb form, called the perfect, in the third person masculine singular. Applying this pattern to the root -r-m, indicating the notion of “banning, prohibiting” (see rm), Arabic forms the perfect third person masculine singular arama, “he prohibited.” Another pattern, XaXX, yields a derived noun, in this case the word arm, “forbidden place,” the source of English HAREM, while the pattern iXXX yields a verbal noun, irm, “prohibition,” the source of English IHRAM. The pattern muXaXXaX (with doubling of the middle root consonant) yields a passive participle, muarram, English MUHARRAM. This last pattern is also found, for example, in the personal name MUHAMMAD, Arabic muammad, from the root -m-d, “to praise” (see md).

   7

      In most Semitic languages, sound changes have obscured some of the underlying patterns. For example, Arabic kf, the origin of English KIF, is a dialectal variant of classical Arabic kayf, a form of the Arabic root k-y-f with the pattern XaXX. Hebrew tôrâ (TORAH) is historically an example of the pattern taXXaXat; the earlier form was *tawrawat-, from the root that was originally w-r-w in Semitic (“to guide”; see wrw), and regular sound shifts in the history of the language changed *tawrawat- to tôrâ.

   8

      The prominence of the root-and-pattern system makes it relatively easy to determine both constituents of most Semitic words. This in turn allows the comparison of individual roots across languages. Thus, for example, Arabic salm, “peace, well-being” (English SALAAM), from the Arabic root s-l-m, is clearly cognate with Hebrew lôm, which has the same meaning (English SHALOM), from the Hebrew root -l-m; both reflect the same Proto-Semitic root, lm. The patterns, too, in this case are cognate; the Proto-Semitic pattern *XaXX, still seen in the Arabic form, regularly develops into XXôX in Hebrew. For most words associated with verbal roots, however, the distribution and semantic function of the various possible patterns are specific to individual languages. The original patterns of specific words very often shifted to other patterns during the separate histories of the various languages after they branched off from their ancestral subgroups. For example, Arabic and Hebrew share a common root, -k-m, “to be wise”; but the attested form of the adjective meaning “wise” in Arabic has the pattern XaXX, akm (English HAKIM¹; see km), while in Hebrew it has the pattern XXX (a Hebrew development of Proto-Semitic *XaXaX), km.

   9

      Because of these pattern shifts, it is usually not possible to reconstruct individual words back to Proto-Semitic, only individual roots. The Appendix that follows is therefore a list of Semitic roots rather than of individual words. An important group of exceptions to this generalization includes words that denote physical objects, such as “hand,” “rock,” and “house.” While such words may be associated with derived roots of verbs (as in English to house), the substantives are clearly primary, and it is often possible to reconstruct them back to Proto-Semitic, or at least to intermediate stages, such as Proto-Central Semitic or Proto-West Semitic. Such reconstructed forms are given in the Appendix where appropriate; to facilitate the arrangement of the Appendix, they have been listed under the consonantal root that can be extracted from the reconstruction, rather than as entries unto themselves. Thus Proto-Semitic *bayt-, “house,” is listed under byt. Some of these words have only two consonants, or rarely only one, rather than the usual three consonants that make up verbal roots; thus, Proto-Semitic *il-, “god,” is listed under l, Proto-Semitic *yad-, “hand,” under yd, and Proto-Semitic *pi- or *pa-, “mouth,” under p.

   10

Proto-Semitic Sounds and Their Development in the Languages

The Proto-Semitic sound system had three short vowels, a, i, u, and three corresponding long vowels, , , ; these vowels are preserved essentially unchanged in classical Arabic but have undergone numerous developments in most of the other Semitic languages, both ancient and modern.

   11

      Proto-Semitic had 29 consonants. These are shown as the first row of sounds in the Table of Proto-Semitic Sound Correspondences. There were five triads of homorganic consonants (pronounced in the same area of the mouth); each triad consisted of a voiced, voiceless, and emphatic consonant. The emphatic consonants are characteristic of Semitic; in Proto-Semitic they were probably glottalized, that is, produced with a simultaneous closing of the glottis in the throat; this is how they are still pronounced in the Ethiopian Semitic languages. (In Arabic, however, emphatics is the name given to pharyngealized consonants, that is, those pronounced with a constriction of the pharynx and a raising of the back of the tongue.) The five triads were: (1) the interdental fricatives , , and (with and pronounced as th in English then and thin, respectively); (2) the dental stops t, d, and (with t and d as in English, and , a glottalized t, represented phonetically as [t’]); (3) the alveolar affricates s, z, , which were pronounced (ts), (dz), and glottalized (ts’), respectively; (4) the laterals l, , and (with l pronounced as in English light and as a voiceless l like the Welsh sound written ll); and (5) the velar stops g, k, q (with g as in English go, k as in kiss, and the q as an emphatic k).

   12

      In addition to these triads, there were a number of pairs of consonants that lacked an emphatic counterpart: two labial stops, voiced b and unvoiced p (the latter becoming f in Arabic, Ethiopian Semitic languages, and sometimes in Hebrew and Aramaic); two velar fricatives, voiced , pronounced like a French r, and voiceless , pronounced like the ch in Scottish loch or German Bach; two distinctively Semitic pharyngeals, voiced “ayin,” indicated by the raised symbol , and unvoiced , both somewhat like h but formed by constricting the pharynx; and two glottal consonants, the glottal stop (like the catch in the voice in the middle of English uh-oh), and glottal fricative, h. Finally, there was a sibilant, transcribed and pronounced sh in Hebrew but as a simple s in Arabic and in Proto-Semitic; and five additional resonants besides l: m, n, r, w, and y.

   13

      All of the 29 Proto-Semitic consonants are preserved as distinct sounds in the Old South Arabian languages (such as Sabaean), but in the other Semitic languages various mergers of the original consonants have occurred. Thus Akkadian, the earliest-attested Semitic language, has only 18 consonants. The outcomes of the Proto-Semitic consonants in Akkadian, Ethiopic, Arabic, Hebrew, and Aramaic are illustrated in the table "Proto-Semitic Sound Correspondences".

   14

Grammatical Forms and Syntax

Semitic nouns exhibit two grammatical genders, masculine and feminine. Masculine nouns have no special marker, whereas the majority of feminine nouns have an ending after the masculine stem, usually either -at or -t, as illustrated by the pairs *bal-, “owner, lord” (as in BAAL and HANNIBAL) versus *balat-, “owner, lady” (see bl), and *bn-, “son” (as in BENJAMIN) versus *bint- (from *bnt-), “daughter” (as in BAT MITZVAH; see bn). A few feminine nouns have no such marker, however, such as *imm-, “mother,” and *ayn-, “eye” (see yn).

   15

      The declension of the noun in early Semitic is relatively uncomplicated. There were three cases, a nominative (for subjects of sentences and for predicates of verbless sentences), genitive (for possession and after all prepositions), and accusative (for the direct object of the verb and for sundry adverbial forms). A characteristic feature of Semitic nouns is the so-called bound form or construct form, an endingless form taken by a noun when it is followed directly by a possessor noun or by a possessive pronoun suffix. For example, the Arabic word anabun means “tail,” but the ending -n is dropped in the possessive phrases anabu asadin, “tail of a lion,” and anabu-hu, “his tail.” Several English words derived from Semitic phrases, such as the star names DENEBOLA and FOMALHAUT, come from a word in the bound form.

   16

      Both Arabic and Hebrew have a definite article (but no indefinite article); in both, the article is prefixed to its noun. In Hebrew the form of the article is ha-, usually with doubling of the first consonant of the noun, as in ha-nâ, “the year” (ROSH HASHANAH) from nâ, “year” (see n). In Arabic the form of the article is al-; the a of al- is omitted when a preceding word ends in a vowel, and the l assimilates to many of the consonants it precedes. The article also causes the final n of forms such as anabun, “tail,” and asadun, “lion,” to be omitted: a-anabu, “the tail,” al-asadu, “the lion.” When a construct phrase, such as anabu asadin, “the lion’s tail,” is made definite, the article appears only on the second member of the phrase: anabu l-asadi, “the lion’s tail.” Many Arabic nouns were borrowed together with the article into European languages, especially into Spanish; this is the source of the al- in a number of English words of Arabic origin, such as ALCOHOL, ALEMBIC, and ALGEBRA, as well as other words where the article has been altered, such as ARTICHOKE and AUBERGINE.

   17

      Most Semitic languages exhibit two types of finite verbs. One type, which is usually called the perfect and is used for completed action, has a set of endings to indicate the person, gender, and number of the subject, as in Arabic daras-a, “he studied,“ daras-at, “she studied,” daras-tu, “I studied,“ and Hebrew dra, “he studied” (with no ending), drâ, “she studied,” dra-tî, “I studied” (see dr). In the other type, the subject is indicated by prefixes (and, for some forms, endings as well), and the verbal root has a different pattern of vowels from the perfect, as in Arabic ya-drus-u, “he studies,” ta-drus-u, “she studies,” a-drus-u, “I study,” and Hebrew yi-dr, “he studies,” ti-dr, “she studies,” e-dr, “I study.” The third person masculine singular form of the perfect is customarily used as the citation form of a verb; traditionally, however, its English translation is given as an infinitive, and this practice is followed in Appendix I. Thus under hgr, the Arabic verb hajara is glossed as “to depart,” although the actual meaning of that form is “he departed.”

   18

      In addition to the two forms just noted, Semitic verbs also have a rich variety of derived stems that variously modify the basic meaning of the verbal root. Thus the Arabic root k-t-b, expressing the notion of “writing,” forms a verb whose basic (perfect) form is kataba, “to write”; with a long vowel in the first syllable, ktaba, it means “to correspond (with someone)”; with in- prefixed, it is a passive, inkataba, “it was written”; and with a- prefixed, it is causative, aktaba, “to cause to write, to dictate.” For simplicity’s sake, such derived forms of the verbal root are labeled in the Appendix as “derived stems.”

   19

Lexicon and Culture

As in the case of Indo-European, the reconstruction of Proto-Semitic words and roots offers us a glimpse of the world and the culture of its speakers.

   20

      Several kinship terms can be reconstructed, a number of which suggest that Proto-Semitic society was patriarchal. Although the words for “father,” *ab- (b), and “mother,” *imm- (mm), are distinct, the word for “daughter,” *bint-, is the grammatical feminine of “son,” *bn- (bn), and “sister,” *at-, is likewise a feminine of the word for “brother,” *a- (). Separate words for “husband’s father,” *am-, and “father’s kinsman, clan,” *amm- (mm), are found, but the feminine equivalents are simply derived from these. Interestingly, the words for “son-in-law, bridegroom,” *atan-, and “daughter-in-law,” *kallat-, are unrelated to each other. A word for “widow,” *almanat-, can be reconstructed, but not one for “widower.”

   21

      Other Proto-Semitic words provide more glimpses into the social structure. That it was stratified is shown by the existence of words for “king” or “prince” (two are found,*arr- and *malk-, the latter of which is associated with the verbal root mlk, “to rule”), “lord, owner, master,” *bal- (and the feminine *balat- “lady”; see bl), and “female slave,” *amat-. (No masculine counterpart is reconstructible; slaves were perhaps acquired as prisoners of war, the males being killed.) Communities had judges who adjudicated (dyn) over local disputes.

   22

      There is no Proto-Semitic word for “religion,” but several religious terms can be reconstructed, such as *b, “to sacrifice”; m, “to anoint”; rm, “to ban, prohibit”; qd, “to be holy, sacred” (as well as ll, “to be clean, pure, holy”); and *alm-, “(cult) statue.” There is a Proto-Semtic word for “god,” *il- (l); the names of the earliest Semitic gods for the most part denoted natural elements or forces, such as the sun, the moon, the morning and evening stars, thunder, and the like (see under tr, m, wr).

   23

      There are many Proto-Semitic terms referring to agriculture, which was a significant source of livelihood. Words for basic farming activities are well represented: fields (*aql-) were plowed (*r), sown (*r), and reaped (d); grain was trampled or threshed (*dy) and winnowed (rw) on a threshing floor (*gurn-), and ground (n) into flour (*qam-). Words for several specific grains can be reconstructed, including wheat (*in-), emmer (*kun-), barley (*ir-; West Semitic only, related to Proto-Semitic *ar- “hair”), and millet (*dun-). The words for many other agricultural products may provide clues as to the original homeland of the Semites, though this is a matter of conjecture and dispute: they were acquainted with figs (*tin-), garlic (*m-), onion (*baal-, replaced in Akkadian by a Sumerian word), palm trees (*tamr- or *tamar-; see tmr), date honey (*dib-), pistachios (*bun-), almonds (*aqid-), cumin (*kammn-; see kmn), and groats or malt (*baql-), as well as oil or fat (*amn-; see mn). The early Semites cultivated grapes (*inab-) growing on vines (*gapn-) in vineyards (*karm- or *karn-), from which they produced wine (*wayn-, akin to Indo-European words for wine and probably a loanword in Proto-Semitic as well). Another alcoholic beverage, *ikar- (kr), was also known; it was stronger than *wayn-, perhaps fermented or distilled.

   24

      Also of Proto-Semitic antiquity are the names of a number of domesticated animals and several words denoting products and activities associated with them. We can reconstruct separate words for “sheep” (*immar-), “ewe” (*lair-, see lr), and “she-goat” (*inz-), as well as separate words for “flock of sheep” (*aw-) and “mixed flock of sheep and goats” (*an-). Sheep were shorn (*gzz) and the flocks “tended” or “herded” (ry, with the participle *riy-, “herder”) and given to drink (qy, a root also meaning “to irrigate”). A general word for “bovine” was *li- (feminine *liat-), in addition to which come *alp-, “ox” (lp) and *awr- “bull” (the latter perhaps a borrowing of Indo-European *tauro-, just as Proto-Semitic *qarn-, “horn” may be of Indo-European origin; see tauro- and ker-¹ in Appendix I). The pig (*zr or *nzr) and the dog (*kalb-) were domesticated, as were donkeys, for which separate words for the male (*imr-; see mr) and the female (*atn-) can be reconstructed. Dairy production is shown by *laad-, “cream,” and *imat-, “curds, butter.”

   25

      The level of technology that the reconstructed Proto-Semitic vocabulary points to is that of the late Neolithic or early Chalcolithic. The early Semites, or at least some of them, lived in houses (*bayt-; see byt) with doors (*dalt-; see dl), containing at least chairs (*kussi-) and beds (*ar-) for furniture. They dug (*kry) wells (*bir-), lit (*rp) fires (*i-), and roasted (qly) food (*lam-; see lm). A number of words dealing with mining are found: the Semites had learned to smelt (rp) ores with coal (*paam-) to obtain metals (only “silver,” *kasp-, is Proto-Semitic; words for “gold,” “copper,” “bronze,” and “iron” are not reconstructible). Bitumen (*kupr-) was used for waterproofing. They also used antimony (*kul-; see kl) and naphtha (*nap-; see np), and manufactured rope (*abl-). The early Semites made use of bows (*qat-) and arrows (*aw-). In transactions, they weighed (ql), measured (*mdd), and otherwise counted (mnw) things, and sometimes, at least, found time to play music (zmr).

   26

      Of particular interest in the reconstruction of the non-material culture of the Proto-Semites is the structure of personal names. Personal names in most Semitic languages have traditionally consisted of meaningful phrases or sentences that express a religious sentiment, usually with reference to a deity. Some names are phrases of the type “X of god,” as in the Hebrew name ydîdyh (Jedidiah), “beloved of Yah” (Yah being a shortened form of the name of the god of Israel, Yahweh; see dwd, hwy) and the Arabic name abdullh (Abdullah), “servant of Allah.” In other Arabic names, an epithet of Allah appears instead, as in abduljabbr (Abdul-Jabbar), “servant of the Almighty,” from jabbr, “powerful, almighty.” Many Semitic names constitute a complete sentence. Some of these contain a verb form, as in Hebrew zkaryh (Zechariah), “Yahweh has remembered” (that is, has remembered the parents; see zkr) and yônn (John), “Yahweh has been gracious” (see nn), and in Akkadian Sîn-a-erba (Sennacherib), “(the god) Sin has replaced the (lost) brothers for me” (see rb) and Aur-aa-iddin (Esarhaddon), “(the god) Ashur has given a brother” (see ntn). Other sentence-names are simply two words juxtaposed to form a nominal sentence with understood verb “to be,” as in Hebrew lîyh (Elijah), “Yahweh (is) my god” (see l, hwy) and abrhm (Abraham), “the (divine) father (is) exalted” (see b, rhm); Akkadian tukult-apil-earra (Tiglathpileser), “my trust (is) the heir of Esharra” (see wkl); and Amorite ammu-rpi (Hammurabi), “the (divine) kinsman is a healer” (see rp).

   27

      One-word names also occur, as in Arabic amad (Ahmed), muammad (Muhammad), and mamd (Mahmoud), which reflect different forms of a root meaning “to praise” (see md), asan (Hasan) and usayn (Hussein), both meaning “handsome, excellent,” and asad (Asad), “lion” (see d); and in Hebrew dwd (David), “beloved” (see dwd) and yônâ (Jonah), “dove” (see ywn). Most women’s names are of this type; for example, Arabic fima (Fatima), “she who weans,” ia (Aisha), “living,” Hebrew râ (Sarah), “princess” (see rr), and rl (Rachel), “ewe” (see lr).

   28

Semitic Words in English

Since English is an Indo-European language and therefore not genetically related to the Semitic family, all words of Semitic origin in English are loanwords. Roughly 700 of the words listed in this Dictionary have come into English by way of a Semitic language. For over 90 percent of these, the Semitic language is Arabic (over 400) or Hebrew (over 250). Not all such words originated in a Semitic language, however. Some of them are loanwords into Semitic from another source. In the case of several words that have come through Arabic, for example, the Arabic word is originally Persian, as in the case of JULEP, borrowed into Middle English from Old French, into Old French from Late Latin, and into Late Latin from Arabic julb; but Arabic julb itself is from Persian gulb, “rosewater.” Here, the Indo-European English, French, and Latin have borrowed from the Semitic Arabic, which in turn has borrowed from another Indo-European language, Persian. Such words will not be found in this Appendix, but if they are derived from an Indo-European root, may be found in the Appendix of Indo-European roots preceding. A number of scientific and technical terms borrowed from Arabic were first borrowed by Arabic from Greek, such as ALEMBIC, from Old French, from Medieval Latin, from Arabic al-anbq, “still” (with the Arabic article al-), from Greek ambix, “cup.” Still others have a more remote or ancient source; the common English word ADOBE, which came into English from Spanish, came into Spanish from Arabic a-ba, “the brick” (with the article al-, here assimilated to the - of the noun), but the Arabic word is of ancient Egyptian origin, Coptic tbe and classical Egyptian bt, “brick.” The word TUNIC, from Latin, entered Latin from a Semitic language akin to Hebrew (perhaps Punic-Phoenician), which in turn had borrowed it from another Semitic language, Akkadian, which in turn had borrowed it from Sumerian (gada, linen), a non-Semitic language of ancient Mesopotamia.

   29

      Most of the words that have come into English from a Semitic language, however, are also Semitic in origin. The following Appendix of Semitic roots lists over 550 such words. Again, the most common language of origin is Arabic, followed by Hebrew. There are also a few dozen words that originate in other Semitic languages, especially Aramaic and Akkadian. In most cases, an Aramaic or Akkadian word has first entered Arabic or Hebrew, whence it then found its way to English; for example, SOUK, a Middle Eastern market, is borrowed from Arabic sq, which was borrowed from Aramaic q, which in turn was borrowed from Akkadian squ, which meant “street,” from a Semitic root meaning “to be narrow or tight” (see yq).

   30

      Many of the Semitic words that have come into English fall into a few important semantic categories. The names of the letters of the Hebrew alphabet are of course Semitic in origin, though for the most part not Hebrew but Phoenician, a close relative of Hebrew; also of Phoenician origin are the names of Greek letters, such as ALPHA and BETA, and the word ALPHABET that is derived from the latter (see lp, byt). Many of the Semitic words are star names, such as ALTAIR, BETELGEUSE, and VEGA, which derive originally from Arabic words or phrases (see yr, gwz and yd, wq). Other large semantic groups are formed by words having to do with religious customs and practices, such as ABBOT, AMEN, ARMAGEDDON, AYATOLLAH, BAR/BAT MITZVAH, HALLELUJAH, JIHAD, KOSHER, MOSQUE, MUEZZIN, RABBI, SATAN, TORAH; business terms, such as ABACUS, ARSENAL, AVERAGE, MINA, SILVER, TARIFF; words for trade goods and similar items, including plant names such as BALSAM, COTTON, CUMIN, GALBANUM, HASHISH, HUMMUS, HYSSOP, MYRRH, SAFFLOWER, SESAME, but also many others, such as AMBER, CANE, CIDER, COFFEE, GYPSUM, HOOKAH, JACKET, JAR, MAT, MATTRESS, MOHAIR, NAPHTHA, RACKET, REAM, SACK, SEMOLINA, SEQUIN; names of animals, including ALBACORE, ALBATROSS, BEHEMOTH, CAMEL, GIRAFFE; and, in addition to the star names mentioned earlier, other scientific terms, such as ALCOHOL, ALGEBRA, ALIDADE, ALKALI, CIPHER, NADIR, SODA, ZENITH, and ZERO. The names of the months of the Jewish calendar and the months of the Muslim calendar have naturally entered English from Hebrew and Arabic, respectively, but it is interesting to note that most of the Jewish month names were originally at home in ancient Mesopotamia and were borrowed into Hebrew from Akkadian. The Semitic languages are also the origin of many proper names in English, such as the names of many of the books of the Bible, as well as given names such as ABRAHAM, ADAM, ANN, JACOB, JACK, and RACHEL. The name MICHAEL, which comes from Hebrew mîkl, meaning “who is like God?” (see the Appendix under my¹, l), may be humanity’s oldest continuously used name, for it is found not only in the Hebrew of the Bible but also in Eblaite, a Semitic language closely related to Akkadian, from about 2300 B.C.

   31

      In spite of the fact that the Semitic languages have been known and studied by scholars for many hundreds of years, the comparative reconstruction of Proto-Semitic is in many ways still in its infancy. The historical linguistics of the Semitic languages has not traditionally focused as much on reconstruction as Indo-European historical linguistics has, and the philological study of the individual languages has remained rather insular. This Appendix of Semitic Roots, while by no means the first comparative Semitic glossary, is the first such work to attempt systematically to give reconstructed forms and meanings for such a wide variety of roots and words. Time and further discoveries will no doubt result in the modification of some of the material here; new information on the ancient Semitic languages is constantly coming to light through archaeological finds, and ongoing linguistic study of the ancient and modern languages is sure to advance our knowledge as well.

   32

The American Heritage^® Dictionary of the English Language, Fourth Edition. Copyright © 2000 by Houghton Mifflin Company. Published by the Houghton Mifflin Company. All rights reserved.

Guide to Appendix II

The layout of Appendix II follows essentially that of Appendix I. An entry is headed by a reconstructed Semitic root in boldface followed by its meaning. The meaning, like the reconstructed root itself, is hypothetical and often difficult to pinpoint exactly; hence meanings are sometimes stated vaguely, and occasionally no meaning at all can be given:

   1

      If the root is possibly but not assuredly of Common Semitic ancestry, the subbranch of Semitic for which it is reconstructible is given first:

   2

      In general, only roots and not actual words are reconstructible for Semitic. If a reconstructed word can be given, it may take the place of a definition of the root if the root is only known from the word in question:

   3

      Following this information is a list of English words, in SMALL CAPITALS, whose histories can be traced back to the root. Only those English words that are entries in the Dictionary are given. If both uncompounded and compounded words or phrases occur, the uncompounded words are given first and separated from the compounds by a semicolon. The Semitic word from which the English words are derived is given next in italics:

   4

      If the exact Semitic source is not known, the oldest attested form to which the English word(s) can be traced is given first:

   5

      If intermediate reconstructed forms are given, as is sometimes necessary, they are preceded by an asterisk (*) to indicate that they are hypothetical, but assumed to have existed. Such a reconstructed form, if ancient enough, is given at the outset of a section before the list of English derivatives:

   6

      Intermediate stages that are attested (such as the stages coming between Arabic arba and English SHERBET) are given in the etymologies in the main vocabulary of the Dictionary, and not in the Appendix, as a general rule.

   7

      Entries are broken up into sections if the English derivatives come from different reconstructed or attested Semitic forms:

   8

      Sections of an entry are themselves divided into subsections if all members of the section ultimately come from the same word. This word is then given after the various subsections:

   9

      English derivatives of the same Semitic word may be listed separately from each other if expository clarity is thereby gained. Thus, in the following excerpt, schlemiel in fact ultimately goes back to the same Hebrew word lm given in 1a as the source of SHALOM and SHALOM ALEICHEM, but is kept separate from it in order to give a gloss to the personal name lmîl:

   10

      Numerous proper nouns in English have been borrowed from Semitic languages, like the geographical name Bethlehem and the personal name Joshua. These are treated in one of two ways. A few, such as the names of selected major biblical figures (especially those after whom books of the Bible are named), are given etymologies in the main vocabulary of the Dictionary, with references to the Appendix as necessary. Others are only given etymologies in the Appendix, and have been put in it as a special feature of this work. They are treated about the same as other words but are given fuller etymological information (since none is given in the main vocabulary of the Dictionary). When necessary, cross-references to other roots are given in boldface.

   11

      In this example, the subentry for Carthage here shows that the word comes from a Phoenician compound; the element -adat is fully etymologized here, but the reader is referred to the entry qr for fuller information on the other compound member, qart-.

   12

Alphabetical Order

The following alphabetical order is observed in this Appendix:

   13

      , , b, d, , g, , h, , , k, l, m, n, p, q, r, s, , , , , t, , , , w, y, z.

   14