^«<. i 0' ii ( =!:; cu : LIBRARY OF THE THEOLOGICAL SEMINARY PRINCETON. N. J. PRESENTED BY Ws l!RS. DONALD JOHTISON Division rldA^SA Section-^ L\. _ J^l _ _ JUN 14 1941 ' L.. GKAMMAK OF THE HEBREW LANGUAGE. BY WILLIAM HENRY GREEN, PROFESSOR IN THE THEOLOGICAL SEMINARY AT PRINCETON, N. J. SECOND EDITION, WITH CORRECTIONS AND COPIOUS INDEXES. NEW YORK: J O H isr WILEY. 1861. Entered, according to Act of Congress, in the year of 1861, by JOHN WILEY, the Clerk's Office of the District Court for the Southern District of New Yorlc. JOHN P. TROW PRINTER, STEREOTYPER, AND ELECTROTTPER, 46, 48 & 50 Greene Street, New York. PREFACE This work was begun at the instance of my friend, preceptor, and colleague. Dr. J. Addison Alexander. The aid of his counsels and suggestions was freely promised in the undertaking ; and he was to give to it the sanction of his name before the public. It appears shorn of these ad- vantages. A few consultations respecting the general plan of the book and the method to be observed in its prepara- tion, were all that could be had before this greatest of American orientalists and scholars was taken from us. De- prived thus early of his invaluable assistance, I have yet found a melancholy satisfaction in the prosecution of a task begun mider such auspices, and which seemed still to link me to one with whom I count it one of the greatest blessings of my life to have been associated. The grammatical system of Gesenius has, from causes which can readily be explained, had a predominance in this country to w^hich it is not justly entitled. The grammar of Prof. Stuart, for a long time the text-book in most common use, was substantially a reproduction of that of Gesenius. Nordheinier was an adherent of the same system in its essen- tial features, though he illustrated it with wonderful clearness and philosophical tact. And finally, the smaller grammar of Gesenius became current in the excellent translation of Prof. Conant. Now, while Gesenius is unquestionably the prince of Hebrew lexicographers, Ewald is as certainly entitled to IV PREFACE. the precedence among grammarians ; and the latter cannot be ignored by him who would appreciate correctly the exist- ing state of oriental learning. The present work is mainly based upon the three leading grammars of Gesenius, Ewald, and Nordheimer, and the at- tempt has been made to combine whatever is valuable in each. For the sake of a more complete survey of the history of opinion, the grammars of R. Chayug, R. Kimchi, Reuch- lin, Buxtorf, Schultens, Simonis, Robertson, Lee, Stier, Hupfeld, Freytag, Nagelsbach, and Stuart, besides others of less consequence from Jewish or Christian sources, have also been consulted to a greater or less extent. The author has not, however, contented himself with an indolent com- pilation; but, while availing himself freely of the labours of his predecessors, he has sought to maintain an independ- ent position by investigating the whole subject freshly and thoroughly for himself. His design in the following pages has been to reflect the phenomena of the language precisely as they are exhibited in the Hebrew Bible ; and it is be- lieved that this is more exactly accomphshed than it has been in any preceding grammar. The rule was adopted at the outset, and rigorously adhered to, that no supposititious forms should be admitted, that no example should be al- leged which is not found in actual use, that no statement should be made and no rule given the evidence of which had not personally been subjected to careful scrutiny. Thus, for example, before treating of any class of verbs, perfect or im- perfect, every verb of that description in the language was separately traced through all its forms as shown by a con- cordance; the facts were thus absolutely ascertained in the first mstance before a single paradigm was prepared or a word of explanation written. Some may be disposed, at first, to look suspiciously upon the triple division of the Hebrew vowels, adopted PRErACE. from Evvalcl, as an innovation : further reflection, however, will show that it is the only division consistent with ac- curacy, and it is really more ancient than the one which commonly prevails. The importance of the accent, especially to the proper understanding of the vowels of a word and the laws of vowel-changes, is such that the example of Ewald has been followed in constantly marking its position by an appropriate sign. He uses a Methegh for this purpose, which is objec- tionable, on account of the liability to error and confusion when the same sign is used for distinct purposes. The use of any one of the many Hebrew accents would also be liable to objection, since they not only indicate the tone- syllable, but have besides a conjunctive or disjunctive force, which it would be out of place to suggest. Accordingly, a special symbol has been employed, analogous to that which is in use in our own and other languages, thus ^i?)? Mtal' . The remarks upon the consecution of poetic accents were in type before the appearance of the able discussion of that subject by Baer, in an appendix to the Commentary of Delitzsch upon the Psalms. The rules of Baer, however, depend for their justification upon the assumption of the accurate accentuation of his own recent edition of the He- brew Psalter, which departs in numerous instances from the current editions as they do in fact from one another. Inas- much as this is a question Avhich can only be settled by manuscripts that are not accessible in this country, it seems best to w^ait until it has been tested and pronounced upon by those who are capable of doing so. What has here been written on that subject, has accordingly been suffered to re- main, imperfect and unsatisfactory as it is. The laws which regulate the formation of nouns have been derived from Ewald, with a few modifications chiefly tending to simplify them. Vi PREFACE. The declensions of nouns, as made out by Gesenius, are purely artificial. Cumbrous as tliey are, they are not ex- haustive, and the student often finds no little difficulty in deciding to which declension certain nouns of frequent oc- currence are to be referred. For these reasons they were abandoned by Nordheimer, who substituted a different sys- tem, which is itself, however, more perplexing than service- able. The fact is, that there are no declensions, properly speaking, in Hebrew; and the attempt to foist upon the lano-uao-e what is alien to its nature, embarrasses the subject instead of relieving it. A few general rules respecting the vowel-changes, which are liable to occur in different kinds of syllables, solve the whole mystery, and are all that the case requires or even admits. In the syntax the aim has been to develop not so much what is common to the Hebrew with other languages, as what is characteristic and distinctive of the former, those points being particularly dwelt upon which are of chief im- portance to the interpreter. In the entire work special reference has been had to the wants of theolooical students. The author has endeavoured to make it at once elementary and thorough, so that it might both serve as a manual for beginners and yet possess all that completeness which is demanded by riper scholars. The parts of most immediate importance to those commencing the study of the language are distinguished by being printed in large type. Peinceton, Au^itst 2'2c?, 1861. OO^sTTEXTS. PART I.— ORTIKJGRAPHY. Divisions of Grammar, §1. ORXnOGEAPniC SYMBOLS. The Letters.— Alphabet, §2; Sounds, §3; Double forms, §4; Names, § 5 ; Order, § ; Classification, § 7 ; Words never divided, § 8 ; Abbreviations and Signs of ISTumber, § 9. The Vowels.— Masoretic Points, §10; Vowel Letters, §11; Signs for the Vowels, §12; Mutual Relation of this twofold dotation, §§18, U; Pure and Diphthongal Vowels, § 15. Sh'va, silent and vocal, simple and compound, §16. Pattahh Furtive, §17. Syllables, § 18. Amhiguous Signs.— Hhirik, Shurek, and Kibbnts, §19.1; Kamets and KametS'Hhatuph, § 19. 2 ; Silent and Vocal Sb'va, § 20. Points affecting Coxsoxakts: — Daghesh-lene, §§21, 22. Daghesh-forte, §23; ditferent kinds, §24; omission of, §25. Mappik, §26. Raphe, §27. Points attached to Words.— Accents, their design, §28; forms and classes, §29; like forms distinguished, §30; poetic accents, §31; position as determined by the character of the syllables, §32. 1; in uninflected words, §32. 2. 3 ; with affixes, suffixes and prefixes, §33 ; use in distinguishing words, § 34 ; shifted in special cases. § 35. " Consecution of the Accents in Prose.— ClsMsea and their subdivisions. § 36 ; tabular view, § 37 ; explnnation of the table, § 38 ; adaptation of the trains of accents to sentences, § 39. viii CONTENTS. Poetic Consecution.— QlnnsQS &n(}i their subdivisions, §40; tabular view and explanation, §41 ; adaptation of tlie trains of accents to sen- tences, §42. Makkeph, §43. Methegh, its form and position, §44; special rules, §45; K'ri and K'thibh, meaning of the terms, §4G; constant K'ris not noted in the margin, § 47 ; their design and value, § 48. Accuracy of the points, § 49. ORTHO GRAPHIC CUANGES. Significant mutations belong to the domain of the lexicon, §§50, 51 ; eu- phonic mutations to the domain of grammar, § 52. MiTATioNS OF CoxsoxANTS at the beginning of syllables, § 53 ; at the close of syllables, §54; at the end of words, §55 ; special rules, §56. Changes of Consonants to Vowels in reduplicated syllables and letters and in quiescents, § 57. Mutations of Vowels, significant and euphonic, §58; due to syllabic changes, §59; to contiguous gutturals, § 60 ; to concurrent conso- nants, § 61 ; concurring vowels, § 62 ; proximity of vowels, § 63 ; the accent, §64; pause accents, §65; shortening or lengthening of words, § 66. PART II.— ETYMOLOGY. Roots of Words. — Design of Etymology, three stages in the growth of words, §67; pronominal and verbal roots, §68; formation and inflection of words by external and internal changes, § 69 ; parts of speech, §70. Pronouns personal, § 71 ; pronominal sufiixes, § 72 ; demonstrative, § 73 ; relative, § 74 ; interrogative and indefinite, § 75. Verbs, the species and their signification, §§ 76-80. Perfect Verbs, § 81 ; formation of the species, §§ 82, 83 ; tlieir, inflection, §§ 84, 85. 1 ; paradigm of bi|r, § 85. 2. Bemarlcs on the Perfect Veris.—Kal preterite, §-86; Infinitive, §87; Future, §88: Imperative, § 89 ; Participles, § 90 ; iSTiphal, § 91 ; Piel, §92; Pual, §93; Iliphil, §94; Hophal, §95; Hithpael, §96. Paragogic and Apocopated Future, § 97 ; and Imperative, § 98. Vav Conversive with the Future, §99 ; with the Preterite, § 100. Verbs with suffixes, §§ 101, 102 ; paradigm, § 103 ; Remarks on the Per- fect Verbs with sufiixes. Preterite, § 104 ; Future, §105; Infinitive and Imperative, § 106. Imperfect Verbs, classified, § 107. Pe Guttural Verbs, their peculiarities, §§108, 109; paradigm, §110; Remarks, §§111-115. CONTENTS. IX Ayin Guttural Verbs, their peculiarities, §110; paradigm, §117; Re- marks, §§118-122. Lamedh Guttural Verbs, their peculiarities, §123; parudigiii, §124; Remarks, §§ 125-128." Pe Nun Verbs, tlieir peculiarities, § 129 ; jjaradigm, § 130 ; Remarks, §§ 131, 132. Ayin Doubled Verbs, their peculiarities, §§133-137; paradigm, §138; Remarks, §§ 139-142. Pc Yodh Verbs, their peculiarities, §§ 143-145 ; paradigm, § 146 ; Re- marks, §§ 147-151. Ayin Vav and Ayin Yodh Verbs, their peculiarities, §§ 152-154 ; para- digm, §155; Remarks, §§ 156-161. Lamedh Aleph Verbs, their pecidiarities, § 1G2 ; paradigm, § 103 ; Re- marks, §§164-167. Lamedh He Verbs, their peculiarities, §§168, 169; paradigm, §170; shortened future and imperative, § 171 ; Remarks, §§172-177. Doubly Imperfect Verbs, §178. Defective Verbs, § 179. Quadriliteral Verbs, § 180. Nouxs, their formation, §181; Class L §§182-186; Class IL §§187, 188; Class m. §§189-192; Class IV. §§193, 194; Multiliterals, §195. Gender and Kumhcr. — Feminine endings, § 196 ; anomalies in the use of, §197; employment in the formation of words, §198; plural end- ings, § 199 ; anomalies, § 200 ; nouns confined to one number, § 201 ; Dual ending, §202; usage of the dual, §203; changes consequent upon aflixing the endings for gender and number, §§206-211, The Construct State^ its meaning and formation, §§ 212-216. Declension of Nouns, paradigm, §217. Paragogic Vowels added to Nouns, §§218, 219. Nouns Avith suffixes, §§ 220, 221 ; paradigm, § 222. Numerals.— Cardinal numbers, §§223-220; Ordinals, etc., §227. Prefixed Particles, §228; the Article, §229: the Interrogative, § 230 ; Inseparable prepositions, §§ 231-233 ; Vav Conjunctive, § 234. Separate Particles. — Adverbs, §285 ; with suffixes, §236; Prepositions, § 237 ; with suffixes, § 238 ; Conjunctions, § 239 ; Interjections, § 240. PART III.— SYNTAX. Office of Syntax, §241. 1; Elements of the sentence, 241. 2. The Subject, a noun or pronoun, § 242 ; when omitted, § 243 ; its exten- sion, §244. The Article, Avhen used, §245; nouns definite without it. §240; omitted in poetry, § 247 ; indefinite nouns, § 248. Adjectives and Demonstratives qualifying a noun. § 240. X CONTENTS. Xin/icrnh.—Ciivdmi] numbers. §§250, 251 ; Ordinals, etc., §252. Apitusitiun, §25:1 The Construct state and Sutlixes, §§25-i-256; resolved by the preposition b, ? 257. TiiK PiiEDicATE, Copula, §258; Nouns, adjectives, and demonstratives, §259. Comparison of adjectives, §260. Vo-hs. — Hebrew conception of time, §261 ; the primary tenses: use of the preterite, § 202 ; the future, § 263 ; paragogic and apocopated future, §264; the secondary tenses, §265 ; participles, §266; Infinitive, §§267-269. Oh/eef of T^rJ.v.— The direct object of transitive verbs, §270 ; transitive con- struction of intransitive verbs, §271 ; indirect object of verbs, §272 ; verbs with more than one object, § 273. Ailverbs and adverbial expressions, § 274. Neglect of agreement, § 275 ; compound subject, § 276 ; nouns in the con- struct, § 277; dual nouns, §278; changes of person, §279. Kei>etition of nouns, § 280 ; pronouns, § 281 ; verbs, § 282. IxTEEROGATivE Sextexces, §§ 283, 284. CoMPorxD Sentences. — Relative pronoun, § 285 ; poetic use of the de- monstrative, § 286 ; conjunctions, § 287. Grammatical Analysis, page 815. Index I. Subjects, ...... page 323 Index II. Texts of Scripture, . . ' . . "331 Index III. Hebrew Words, . . ..." 343 Index IV. Hebrew Grammatical Terms, . . . " 399 PART FIRST. ORTHOGRAPHY. §1 . Language is the communication of thouglit by means of spoken or written sounds. The utterance of a single thought constitutes a sentence. Each sentence is composed of words expressing individual conceptions or their relations. And words are made up of sounds produced by the organs of speech and represented by written signs. It is the province of grammar as the science of language to investigate these several elements. It hence consists of three parts. First, Orthography, which treats of the sounds employed and the mode of representing them. Second, Etymology, which treats of the different kinds of words, their formation, and the changes which they undergo. Third, Syntax, which treats of sentences, or the manner in which words are joined together to express ideas. The task of the Hebrew grammarian is to furnish a complete exhibition of the phenomena of this partic- ular language, carefully digested and referred as far as practi- cable to their appropriate causes in the organs of speech and the operations of the mind. The Letters. §2. The Hebrew being no longer a spoken tongue, is only known as the language of books, and particularly of the Old Testament, which is the most interesting and important as well as the only pure monument of it. The first step 2 ORTHOGRAPHY. §2 towards its investigation must accordingly be to ascertain the meaning of the symbols in which it is recorded. Then having learned its sounds, as they are thus represented, it will be possible to advance one step further, and inquire into the laws by which these are governed in their employment and mutations. The symbols used in writing Hebrew are of two sorts, viz. letters (ni'^nisi) and points (Q"^"!^]??). The number of the letters is twenty -two ; these are written from right to left, and are exclusively consonants. The following alphabetical table exhibits their forms, English equivalents, names, and numeri- cal values, together with the corresponding forms of the Rab- binical character employed to a considerable extent in the commentaries and other writings of the modern Jews. §3 LETTERS. Order. Forms and Equivalents. Names. Eabbinical Alpliabet. Numerical values. 1 S* qb^ Aleph i> 1 2 n Bh, B n^3 Beth 3 2 3 a Gh, G ^^^k Gi'-mel J 3 4 "I Dh, D V T Da'-leth 7 4 O n H N?;} He t) 5 G 1 V ^) Vav 1 6 7 T Z VI Zajin » 7 8 n Hh n'ln Hheth P 8 9 13 T n^tD Teth V 9 10 1 Y lii Yoclli , 10 11 3 1 Kh, K !^? Kaph 1 = 20 12 b L ^■ab' La'-medh i 30 13 •a D M Q^ Mem r 40 14 2 1 N r= Niln P 50 15 S ^^9 Sa'-mekh P 60 16 y r^ Ayin :: 70 17 s q Ph, P «a Pe S^ 80 18 ^ r Ts I'li Tsa'-dhe P 90 19 p K S^ip Kopli ? 100 20 n R iijin Resh ■5 200 21 lU Sh, S r^' Shin C 300 22 n Th, T in Tav P 400 §3. There is always more or less difficulty in represent- ing the sounds of one language by those of another. But this is in the case of the Hebrew greatly aggravated by its having been for ages a dead language, so that some of its 4 ORTHOGRAPHY. §3 sounds cannot now be accurately determined, and also by its belonging to a different family or group of tongues from our own, possessing sounds entirely foreign to the English, for which it consequently affords no equivalent, and which are in fact incapable of being pronounced by our organs. The equivalents of the foregoing table are not therefore to be re- garded as in every instance exact representations of the proper powers of the letters. They are simply approximations suffi- ciently near the truth for every practical purpose, the best which can now be proposed, and sanctioned by tradition and the conventional usa2:e of the best Hebraists. 1. It will be observed that a double pronunciation has been assigned to seven of the letters. A native Hebrew would readily decide without assistance which of these was to be adopted in any given case, just as we are sensible of no in- convenience from the various sounds of the English letters which are so embarrassing to foreigners learning our language. The ambiguity is in every case removed, however, by the ad- dition of a dot or point indicating which sound they are to receive. Thus a with a point in its bosom has the sound of h, n unpointed that of the corresponding v, or as it is com- monly represented for the sake of uniformity in notation, hh ; a is pronounced as //, ^ unpointed had an aspirated sound which may accordingly be represented hj ^/i, but as it is difficult to produce it, or even to determine with exactness what it was, and as there is no corresponding sound in English, tlie aspira- tion is mostly neglected, and the letter, whether pointed or not, sounded indifferently as y ; '^ is d, 1 unpointed is the aspirate d/i, equivalent to t/i in the ; 3 is k, 3 unpointed its aspirate kh, perhaps resembling the German ch in ich, though its aspira- tion, like that of IJ, is commonly neglected in modern reading ; £! is/*, & unpointed is jy/^ or/; n is ^'j n unpointed th in ihin. The letter "^ with a dot over its right arm is pronounced like sh, and called SIfm ; is with a dot over its left arm is called JS2?i, and pronounced like s, no attempt being made in modern §3 . LETTERS. 5 usage to discriminate between its sound and that of D Samckh. Althougli there may anciently have been a distinc- tion between them, this can no longer be defined nor even positively asserted ; it has therefore been thought unneces- sary to preserve the individuality of these letters in the notation, and both of them will accordingly be represented by s. a. The double sound of the first six of the letters just named is purely euphonic, and has no effect whatever upon the meaning of the words in which they stand. The case of O is different. Its primary sound was that of sh, as is evident from the contrast in Judg. 12 : 6 of rb3':J shibboleth with nb>p sibbolelh. In certain words, however, and sometimes for the sake of creating a distinction between different words of like orthography, it received the sound of 5, thus almost assuming the character of a distinct letter, e. g. ^r:J to break, ^ib to hope. That Sin and Samekh were dis- tinguishable to the ear, appears probable from the fact that there are words of separate significations which differ only in the use of one or the other of these letters, and in which they are never interchanged, e. g. bydi to be bereaved, bib to be xoise, bio to be foolish; •^ittj to be dnmken.'^-^zv to hire, ■n=0 to shut lip; ^rr to look, nr:) to ride, ^^o'to turn back; r^^-a a lip, nsD to destroy. The close affinity between the sounds which the'y repre- sent is, however, shown by the fact that is in a tew instances written for O, e. g. no3 Ps. 4 : 7 from ^cz , nii^qb Eccles. 1 ! 17 for wi^DD . The original identity of aj and b is apparent from the etymological connection between -^Nb leacen and nnxttJa a vessel in which bread is leaveried; ni'b to shudder, nin^b horrible, causing a shudder. In Arabic the division of' single letters into two distinguished by diacritical points is carried to a mucTi greater length, the alphabet of that language being by this means enlarged from twenty-two to twenty-eight letters. 2. In their original power u t differed from n f, and S Jk from p X-, for these letters are not confused nor liable to inter- change, and the distinction is preserved to this day in the cognate Arabic; yet it is not easy to state intelligibly where- in the difference consisted. They are currently pronounced precisely alike. 3. The letter n has a stronger sound than n the simple /t, and is accordingly represented by M; n is represented by r, although it had some peculiarity of sound which we can- not at this day attempt to reproduce, by which it was allied to the guttm-als. 6 ORTHOGRAPHY. . ^ 4. Per two letters, i5 and V, no equivalent has been given in the table, and they are commonly altogether neglected in pronunciation. iJ is the weakest of the letters, and was prob- ably always mandible. It stands for the slight and involun- tary emission of breath necessary to the utterance of a vowel unattended by a more distinct consonant sound. It there- fore merely serves to mark the beginning or the close of the syllable of which it is a part, while to the ear it is entirely lost in the accompanying or preceding vowel. Its power has been likened to that of the smooth breathing (') of the Greeks or the English silent h in hour. On the other hand 5' had a deep guttural sound which was always heard, but like that of the corresponding letter among the Arabs is very difficult of utterance by occidental organs ; consequently no attempt is made to reproduce it. In the Septuagint it is some- times represented by y, sometimes by the rough and some- times by the smooth breathing ; thus "^y^? Fof^oQ^a, "^^2? 'Hit, "pV^t, ^Jfj^alrj^. Some of the modern Jews give it the sound of )!(/ or of the Prench y;^ in campagne, either wherever it occurs or only at the end of words, e. g. 3!''ato Slimang, "l^!? (jimmbdli. §4. The forms of the letters exhibited in the preceding table, though found without important variation in all existing manuscripts, are not the original ones. An older character is preserved upon the Jewish coins struck in the age of the Maccabees, which bears a considerable resemblance to the Samaritan and still more to the Phenician. Some of the steps in the transition from one to the other can still be traced upon extant monuments. There was first a cm'sive tendency, disposing to unite the different letters of the same word, which is the estabhshed practice in Syriac and Arabic. This was followed by a predominance of the calligraphic principle, which again separated the letters and reduced them to their present rectangular forms and nearly uniform size. The cursive stage has, however, left its traces upon the five letters §5 LETTERS. 7 which appear in the table Avith double forms ; D "a 3 s 2 when standing at the beginning or in the middle of words termi- nate in a bottom horizontal stroke, which is the remnant of the connecting link with the following letter ; at the end of words no such link was needed, and the letter was continued vertically downward in a sort of terminal flourish thus, "1 1 ?] f , or closed up by joining its last with its initial stroke, thus D. a. The {"cw instances in which final letters are fonnd in the middle of words, as t^3"^n^ Isa. 9: 6, or their ordinary forms at the end, as "cn Neh. 2: 13, 5n Job 38: 1, are probably due to the inadvertence of early tran- scribers which has been faithfully perpetuated since, or if intentional they may have had a connection now unknown with the enumeration of letters or the signification of words. The same may be said of letters larger than usual, as nsSi Ps. SO : 16, or smaller, as □X"ia'?a Gen. 2:4, or above the line, as "i?^^ Ps. 80 : 14, or inverted, as >b:a Num. 10: 35, (in manuscripts and the older editions, e. g. thatof Stephanus in 1541), or with extraordinary points, as ^nj^ii^i Gen. 33 : 4, iCi}^ Ps. 27 : 13, in all which the Rabbins find concealed meanings of the most fanciful and absurd character. Thus in theif opinion the suspended 3 in H'b^Ta Judg. IS : 30 suggests that the idola- ters described were descended from Moses but had the character of Ma- nasseh. In ■|'^^^ Lev. 11 : 42 the Vav, which is of unusual size, is the middle letter of the Pentateuch; ^^i""?!! Gen. 16: 5 with an extraordinary point over the second Yodh, is the only instance in which the word is written with that letter; the large letters in Deut. 6: 4 emphasize the capital article of the Jewish faitli. All such anomalous forms or marks, with the conceits of the Rabbins respecting them, are reviewed in detail in Buxtorf's Tiberias, pp. 152 etc. §5. All the names of the letters were probably significant at first, although the meanings of some of them are now doubt- ful or obscure. It is commonly supposed that these describe the objects to which their forms originally bore a rude resem- blance. If this be so, however, the mutations which they have since undergone are such, that the relation is no longer traceable, unless it be faintly in a few. The power of the letter is in every instance the initial sound of its name. a. The opinion advocated by Schultens, Fundamenta Ling. Heb. p. 10, that the invention of the letters was long anterior to that of their names, and that the latter was a pedagogical expedient to facilitate the learning of the letters by associating their forms and sounds with familiar objects, has met vvith little favour and possesses little intrinsic probability. An interest- 8 ORTHOGRAPHY. §6 ing corroboration of the antiquity of these names is found in their preserva- tio'n in the Greek alphabet, though destitute of meaning in that language, the Greeits having borrowed their letters at an early period from the Phe- nicians,and lience the appended a of MX^a, etc.. which points to the Ara- maeic form xs^x . h. The Semitic derivation of the names proves incontestably that the alphabet had its origin among a people speaking a language kindred to the Hebrew. Their most probable meanings, so far as they are still ex- plicable, are as follows, viz: Aleph, an ox; Beth, a Aouse ; Gimel, a camel; Daleth, a door; He, doubtful, possibly a window; Vav, a hook; Zayin, a weapon; Hheth, probably a fence; Teth, probably a snake ; Yodh, a hand; Kaph, the palm of the hand; Lamedh, an ox-goad; Mem, water; Nun, a fish; Same kh, a prop; Ay\\\aneye; Fe,aviouih; Tsadhe, afsh-hook or a humerus dart; Koph, perhaps the back of the head; Resh, a head; Shin, a tooth ; Tav, a cross mark. §6. The order of the letters appears to be entirely arbi- trary, though it has been remarked that the three middle mutes n :^ 1 succeed each other, as in like manner the three hquids '= 'o 5 • The juxtaposition of a few of the letters may perhaps be owing to the kindred signification of their names, e. g. Yodh and Kaph the hand, Mem wafer and Nun a fish, Resh the head and Shin a tooth. The antiquity of the existing arrangement of the alphabet is shown, 1. by psalms and other portions of the Old Testament in which successive clauses or verses begin with the letters disposed in regular order, viz. Ps. 25 (p omitted), 34, 37 (alternate verses, 2> omitted). 111 (every clause), 112 (every clause), 119 (each letter eight times), 145 (3 omitted), Prov. 31 : 10-31, Lam. ch. 1, 2, 3 (each letter three times), 4. In the first chapter of Lamenta- tions the order is exactly preserved, but in the remaining three chapters 2? and S are transposed. 2. By the corres- pondence of the Greek and Roman alphabets, which have sprung from the same origin with the Hebrew. a. The most ingenious attempt to discover a regular structure in the Hebrew alphabet is that of I.epsius, in an essay upon this subject published in 1836. Omitting the sibilants and Resh, he finds the following triple correspondence of a breathing succeeded by the same three mutes carried through each of the three orders, the second rank being enlarged by the addition of the liquids. §7 LETTERS. Breathiiigs. Mutes. ■ 1 Liquids. ■ 1 Middle Smootli Rough n 1 n 12 a p n (^=) b525 Curious as this result certainly is. it must be confessed that the alleged correspondence is in part imaginary, and the method by which it is reached is too arbitrary to warrant the conclusion that this scheme was really in the mind of the author of the alphabet, much less to sustain the further speculations built upon it, reducing the original number and modifying the powers of the letters. h. It is curious to see how, in the adaptation of tlie alphabet to different languages, the sounds of the letters have been modified, needless ones dropped, and others found necessary added at the end, without disturbing the arrangement of the original stock. Thus the Greeks dropped 1 and p, only retaining them as numerical signs, while the Roman alphabet has F and Q,; on the other hand the Romans found 'O and superfluous, while the Greeks made of them r^ and ^; 5 and t, in Greek 'J b ^ W Dentals T D 3S IIJ Labials n ^ ^ s n has been differently classed, but as its peculiarities are those of the gutturals, it is usually reckoned with them. 2. Secondly, according to their respective strength, into three classes, which may be denominated weak, medium, and strong. The strong consonants offer the greatest resistance to change, and (ire capable of entering into any combinations which the formation or inflection of words may require. The weak have not this capacity, but when analogy woidd bring them into combinations foreign to their nature, they are either 10 ORTHOGRAPHT. ^7 liable to mutation tliemselves or occasion changes in the rest of the word. Those of medium strength have neither the absolute stability of the former nor the feeble and fluctuating character of the latter. AJ^ V ( i5 n 1 'I Vowel-Letters, ^^^^' 1 x n n 2^ Gutturals. Medium, b •» 3 "I Liquids, T D 22 ID Sibilants. Strong, -I^ ^ ^ p )- Aspirates and Mutes. The special characteristics of these several classes and the influence which they exert upon the constitution of words will be considered hereafter. It is sufficient to remark here that the vowel-letters are so called because they sometimes represent not consonant but vowel-sounds. a. It will be observed that while the p, k, and if-mutes agree in having smooth S 3 n and middle forms 3 a "i, which may be either aspirated or nnaspirated, the two last have each an additional representative p a which is lacking to the first. This, coupled with the fact that two of the alpha- betic Psalms, Ps. 25, 34, repeat S as the initial of the closing verse, has given rise to the conjecture that the missing p mute was supplied by this letter, having a double sound and a double place in the alphabet. In curi- ous coincidence with this ingenious but unsustained hypothesis, the Ethio- pic alphabet has an additional p, and the Greek and Roman alphabets agree one step and only one beyond the letter T, viz. in adding next a labial, which in Greek is divided into v and (j), and in Latin into U and V. as "i into I and J. 3. Thirdly, The letters may be divided,- with respect to their function in the formation of words, into radicals and serviles. The former, which comprise just one half of the alphabet, are never employed except in the roots or radical portions of words. The latter may also enter into the con- stitution of roots, but they are likewise put to the less inde- pendent use of the formation of derivatives and inflections, of prefixes and suffixes. The serviles are embraced in the ^8 LETTERS. 11 memorial words nbDi rna^ 'jn'^x (Ethan Moses and Caleb) ; of these, besides other uses, "jfT^s are prefixed to form the future of verbs, and the remainder are prefixed as particles to nouns. The letters T^np'asin are used in the formation of nouns from their roots. The only exception to the division now stated is the substitution of t: for servile n in a certain class of cases, as explained § 54. 4. a. Kimchi in his Mikhlol (bibsio) fol. 46, gives several additional ana- grams of the serviles made out by different grammarians as aids to the memory, e. g. n2"'a IPDxb^lU ybr his work is understmiding ; n^biy 13s ams / Solomon am wriling ; nisn -^x inibo only build thou my peace ; "p^n 3X ^TOS like a branch of the father of mult ihide ; i3-'b.x sno noB Moses has icritlen to us. To which Nordheimer has added "'ins^ '^^r^ bs*a consult the riches of my book. § 8. In Hebrew writing and printing, words are never divided. Hence various expedients are resorted to upon occasion, in manuscripts and old printed editions, to fill out the lines, such as giving a broad form to certain letters, ^i Ti H 1=3 t-i , occupying the vacant space with some letter, as p, repeated as often as may be necessary, or with the first letters of the next word, which were not, however, accounted part of the text, as they were left without vowels, and the word was written in full at the beginning of the following line. The same end is accomplished more neatly in modern print- ing by judicious spacing. §9.1. The later Jews make frequent use of abbreviations. There are none, however, in the text of the Hebrew Bible ; such as are found in the margin are explained in a special lexicon at the back of the editions in most common use, e. g. ^y} for "I'oilJ'i ef comple.tio = etc. 2. The numerical employment of the letters, common to the Hebrews with the Greeks, is indicated in the table of the alphabet. The hundreds from 500 to 900 are represented either by the five final letters or by the combination of M with the letters immediately preceding; thus 1 or pn 500, Q or "in 600, ■} »n or pnn 700, q or nn 800, T or pnn 900. Thou- 12 ORTHOGRAPHY. ^ 10 sands are represented by units with two dots placed over tliem, thus X 1000, etc. Compound numbers are formed by joining the appropriate units to the tens and hundreds, thus i^sn 421, Fifteen is, however, made not by Si'' , which are the initial letters of the divine name Jehovah, tiini, but by It: 9-|-6. This use of the letters is found in the accessories of the Hebrew text, e. g. in the numeration of the chapters and verses, and in the Masoretic notes, but not in the text itself. Whether these or any other signs of number were ever employed by the original writers of Scripture, or by the scribes in copying it, may be a doubtful matter. It has been ingeniously con- jectured, and with a show of plausibihty, that some of the discrepancies of numbers in the Old Testament may be accounted for by assuming the existence of such a system of symbols, in which errors might more easily arise than in fully written words. The Vowels. § 10. The letters now explained constitute the body of the Hebrew text. These are all that belonged to it in its original form, and so long as the language was a living one nothing more was necessary, for the reader could mentally supply the deficiencies of the notation from his famiharity with his native tongue. But when Hebrew ceased to be spoken the case was different ; the knowledge of the true proniuiciation could no longer be presumed, and difficulties would arise from the ambiguity of individual words and their doubtful relation to one another. It is the design of the Masoretic points ( nnic^ traditioii) to remedy or obviate these inconveniences by supplying what was lacking in this mode of writing. The authors of this system did not venture to make any change in the letters of the sacred text. The signs which they introduced were entirely supplementary, consist- ing of dots and marks about the text fixing its true pronun- §11 VOWELS. 13 ciatioii and auxiliary to its proper interpretation. This has been clone with the utmost nicety and minuteness, and with, sucli evident accuracy and care as to make them rehable and efficient if not indispensable helps. These points or signs are of three kinds, 1. those representing the vowels, 2. those affecting the consonants, 3. those attached to words. a. As illustrations of the ambiguity both as to sound and sense of indi- vidual words, when written by the letters only, it may be stated that ^z~i is in Gen. 12: 4 n'^'n he spake, in Ex. 6: 29 "lii'n speak and ni'-n speak- ing, in Prov. 25: 11 13^ spoken, in Gen. 37 : 14 li'n word, in 1 Kin. 6: 16 li^: (he oracle or most holy place of the temple, in Ex. 9 : 3 nrri pestilence. So pffi-iT is in Gen. 29: 10 pa^l and he watered, and in the next verse pik'^.l andhe kissed; xn-^l occurs twice in Gen. 29 : 23, the first time it is xi^;;, and he brought, the second xiljl and he came; dTinam is in Jer. 32: 37 first c-irh^rn;; ajid I will bring them again, and then D-^ni'rn-, and I will cause (hem to dwell; W'-n^^i is in Gen. 14 : 19 D-^ts^ heaven, and in Isa. 5 : 20 D^bu) putting. This ambiguity is, however, in most cases removed by the con- nection in which the words are found, so that there is little practical diffi- culty for one who is well acquainted with the language. Modern Hebrew is commonly written and read without the points: and the same is true of its kindred tongues the Syriac and Arabic, though each of these has a system of points additional to the letters. ^11. 1. The alphabet, as has been seen, consisted exclu- sively of consonants, since these were regarded as a sufficiently exact representation of the syllables into which in Hebrew they invariably enter. And the omission of the vowels occa- sioned less embarrassment, because in the Semitic family of languages generally, unlike the Indo-European, they form no part, properly speaking, of the radical structure of the word, and consequently do not aid in expressing its essential mean- ing, but only its nicer shades and modifications. Still some notation of vowels was always necessary, and this was furnish- ed in a scanty measure by the vowel-letters, or, as they are also called, quiescents, or w^rt/rc5 lectionis (guides in reading). The weakest of the palatals "^ was taken as the representative of the vowels t and B of the same organ to which in sound it bears a close affinity ; the weakest of the labials 1 was in hke manner made to represent its cognates ii and o ; and the 14 ORTHOGRAPHY. §11 two weak gutturals 5? and n were written for the guttural vowel a, as well as for the compound vowels B and cf which a is one of the elements. Letters were more rarely employed to represent short vowels ; n or "^ for c is the most frequent case ; others are exceptional. a. Medial a when written at all, as it very rarely is, is denoted by N , e. g. ■j.-tb lal Judg. 4 : 21, 5X1 dag- Neh. 13: 16 K'thibh, CSp kam Hos. 10: 14, b^nvj "zazel Lev. 16: 8, iTXi ro-s/i Prov. 10: 4 and in a few other passages, mrx-i sometimes for ramolh, ^Ni:J isavvar, "iNDXisx Hos. 4 : 6 if not an error in the text perhaps for emasak ; final a, which is much more frequent- ly written, is denoted by n, e. g. nba gala, nsbn vialka, nnx a/ta, rarely and only as an Aramasism by N, e. g. itm hhogga Isa. 19 : 17, xn-ip korhhfl Ezek. 27 : 31 K'thibh, »riz> gahh'ha Ezek. 31 : 5 K'thibh. The writing of e and e, and u is optional in the middle of words but necessary at (he end, e. g. nnii^ or cnilif tsivmhim, "^niiil tsivvUhl ; ■-!:: or 1310 shub/iii. In the former position "^ stands for the first pair of vowels, and 1 for the second, e. g. mpi3"'0 meifikuth, injioa ti'sugholhi; X for e and 6 so situated is rare and exceptional, e. g. irxn resh Prov. 6 : 11, 30 : 8, and perhaps yXJi yanels Eccles. 12:5 ; nxT zOlh, nxis poroth Ezek. 31 : 8, tPXsa bilstsOthav Ezek. 47: 11. At the end of words e is commonly expressed by "i, and o by i, though n is frequently and X rarely employed for the same purpose, e. g. ''Zih-a malkhe, I3b^ malko ; n^^ A-'i/e, nsns parO ; 8 is not sounded. 2. Sh'va may be either silent (n3 quiescens), or vocal (2!'3 mobile). At the close of syllables it is silent. But at the beginning of a syllable the Ilebrev/s always facilitated the pronunciation of concurrent consonants by the introduction of a hiatus or slight breathing between them ; a Sh'va so situated is consequently said to be vocal, and has a sound approaching that of a hastily uttered e, as in given. This will be represented by an apostrophe, thus, "iS"!^^ Vmidhhar, Dr)'7i?7' 2jf^'<-((^Meni. a. According to Kimchi (Mikhlol fol. ISO) Sh'va was pronounced in three different ways, according to circumstances. (1.) Before a guttural it inclined to the sound of the following vowel, e. g. lix';' 7/abbedh, HNb s'eth, ^"'n d"u, and if accompanied by Methegh, §44, it had the full sound of that vowel, e. g. ^sb auu, "^nn nhhi, C^is'b loolam. (2.) Before Yodh it inclined to ?', e. g. Sp?.''^ b'lja kubh, Dii3 k'yoin, and with Methegh was sounded as Hhirik, e. g. T^^ biijadh. (3.) Before any other letter it in- clined to a, e. g. i^^'^3 b"rakha, D^B"i^a g^Ill'im, and with Methegh was pronounced as Pattahh nibnp^s bamakheloth. 3. Sh'va may, again, be simple or compound. Some- times, particularly when the first consonant is a guttural, which from its weakness is in danger of not being distinctly heard, the hiatus becomes still more audible, and is assimi- lated in sound to the short guttural vowel a, or the diph- thongal t or 0, into which it entei-s. This assimilation is rep- ^17 VOWELS. 21 resented by combining the sign for Sh'va with those for the short vowels, thus forming what are called the compound Sh'vas in distinction from the simple Sh'va previously ex- plained. These are, Hhatcph-Pattahh — ; thus, "lb? ""mbdh. Hhatcph-Seghol "^; thus, ^bJ!|! "mbr. lihateph-Kamets tt; thus, "'bn hWU. a. Hhateph (''^'^,f} snatching) denotes the rapidity of utterance or the hurried character of the sounds represented hy tliese symbols. b. The compound Sh'vas, though for the most part restricted to llic gutturals, are occasionally written under other consonants in place of sim- ple Sh'va, to indicate more distinctly that it is vocal: thus, Hhateph- Pattahh ::nn Gen. 2:12, fib'^^v! ^^n. 27:38; Hhateph-Kamets nnpb Gen. 2 : 23, iisiriDX Jer. 31 : 33 ; but never Hhateph-Seghol except cy^^:!: 2 Sam. 6: 5 in some editions, e. g. that of Stephanus. This is done with so little uniformity that the same word is diflerently written in this respect, e. g. 'T'j^'S? 2 Kin. 2 : 1, ir'^^Da ver. 11. Pattahii Furtive. §17. A similar hiatus or slight transition sound was used at the end of words in connection with the gutturals. When y , n , or the consonantal ^ at the end of words is pre- ceded by a long heterogeneous vowel (i. e. another than a), or is followed by another vowelless consonant, it receives a Pattahh furtive — , which resembles in sound an extremely short a, and is pronounced before the letter under which it is written, e. g. n^n riiVt//, ?)2Ty sMmlf , 'r^^%''^^ ma/jkhVIi, vmb shuma"t, ^T\^ yi%hd. a. Some grammarians deny that Pattahh furtive can be found under a penultimate guttural, contending that the vowel-sign is in such cases a proper Pattahh. and that nr^d should accordingly be read shamaat, and ^ni yihhad. But both the Sh'va under the final letter, §1G, and the Daghesh-lene in it, § 21, show that the guttural is not followed by a vowel. The sign beneath it must consequently be Pattahh furtive, and represent an antecedent vowel-sound. In some manuscripts Pattahh furtive is writ- ten as Hhateph-Pattahh, or even as simple Sh'va ; thus, ?"^P7 or ^p*! for S"'pl. 22 orthography. ^18 Syllables. §18. 1. Syllables are formed by the combination of consonants and vowels. As two vowels never come together in the same word in Hebrew without an intervening conso- nant, there can never be more than one vowel in the same syllable ; and with the single exception of ^ occurring at the beginning of words, no syllable ever consists of a vowel alone. Every syllable, with the exception just stated, must begin with a consonant, and may begin with two, but never with more than two. Syllables ending with a vowel, whether represented by a quiescent letter or not, are called simple, e. g. "^Vl^Cij ribis? u-lCi. (The first syllable of this second example begins, it will be perceived, with the consonant y , though this disappears in the notation given of its sound.) Syllables ending with a consonant, or, as is possible at the close of a Avord, with two consonants, are said to be mixed : thus Dn^J? ham-tem, ^^^T} lia-lahM. As the vocal ShVas, whether simple or compound, are not vowels properly speak- ing, but simply involuntary transition sounds, they, with the consonants under which they stand, cannot form distinct syllables, but are attached to that of the following vowel, Pattahh furtive in like manner belongs to the syllable formed by the preceding vowel. Thus ?i"iT zrb'^, ""IS! '^nl are mono- syllables. 2. Long vowels always stand in simple syllables, and short vowels in mixed syllables, unless they, be accented. But accented syllables, whether simple or- mixed, may con- tain indifferently a long or a short vowel. a. The following may serve as a specimen of the division of Hebrew w^ords into their proper syllables ; thus, u-dha'm ''lo-hi'm b'ro' b'yo'm a-dha'm to-1'dho'th se'-pher ze' Gen. 5 : 1. irx nps C!'^ih"'bj< rwvs, O-tho' cVsa' 'lo-hl'm bidh-mu'th §19 SYLLABLES. 23 h. The reason of tne ruie for the quantity of syllables appears to be this. In consequence of their brevity, the short vowels required the ad- dition of a following consonant to make the utterance full and complete, unless the want of this was compensated by the greater energy of pronun- ciation due to the accent. The long vowels were sufficiently complete without'any such addition, though they were capable of receiving it under the new energy imparted by the accent. This pervading regularity, which is so striking a feature of the Hebrew language, was the foundation of the systema morarum advocated by some of the older grammarians of Holland and Germany. The idea of this was, that each syllable was equal to three morae, that is, three rests, or a bar of three beats ; a long vowel being equivalent to two morae, or two beats, a short vowel to one, and the initial or final consonant or consonants also to one : thus Pi^^rj k (1) + a (2) = 3. t (1) + a (1) -f U (1) = 3. An accented syllable might have one mora or beat either more or less than the normal quan- tity. This system was not only proposed by way of grammatical explana- tion, but also made the basis ol'a peculiar theory ot Hebrew prosody. See Gesenius, Geschichte d. Heb. Sprache, p. 123. c. The cases in which short vowels occur in unaccented simple sylla- bles, are all due to the disturbing influence exerted by the weak letters upon the normal forms of words ; thus, n^rn hd-clh is for r^rn, and NJinn ha-hu for hdh-hu : such words as X'^'n , X"iE, N^e, n:.n, ti^j? are formed after the analogy of ~^^. A long vowel in an unaccented mixed syllable is found in but one word, and that of foreign origin, "^Xil'i^^a bel-f shuts- tsar ; though here, as in the majority of instances falling under the previ- ous remark, the syllable receives, if not the primary, yet the secondary accent, e. g. '^niT'^n, Qinnti , iia?.n. The same is the case when a long vowel is retained before Makkeph, e. g. •'5~rd. In the Arabic, Avhich is exceedingly rich in vowels, there are comparatively few mixed syllables; nearly every consonant has its own vowel, and this more ii'equently short than long. The Chaldee, which is more sparing in its use of vowels than the Hebreiv, observes in general the same rule with respect to the quan- tity of syllables, though not with the same inflexible consistency. Ambiguous Signs. §19. It will now be possible, by aid of the principles already recited, to determine the quantity of the doubtful vowels, and to remove the ambiguity which appears to exist in certain vowel-signs. 1. Hhirik, Shurek, and Kibbuts, in unaccented simple syllables, must be long, and in unaccented mixed syllables, short, e. g. '!^T'! or tJni p-rash, ^b"' yihh-nu, i'i^n^ or i53f ghhii-lb, ^i^_ or "I^T" yid-ladli, 3^3 or "UfO hd-lCim, ^ry^ 24 ORTHOGRAPHY. §19 or ''•T^*''? mauzzi. In accented syllables, whether simple or mixed, they are always long, e. g. nn-iUJ or D^niilJ si-him, "^^ ll, via or b^ia ghkul, ^riirn^: or inw^'i cVrCi-sltu-liU, the only ex- ception being that Hhmk is short in the monosyllabic parti- cles Di? , i^"^ , S3> , ]'□ , and in some abbreviated verbal forms of the class called Lamedh-He, e. g. T^:^, aiu'^n, nn';*. The only cases of remaining doubt are those in which these vowels are followed by a letter with Sh'va, either sim- ple or compound. If the former, it might be a question whether it was silent or vocal, and consequently whether the syllable was simple or mixed. If the latter, though the syl- lable is of course simple, the weak letter which follows may interfere with the operation of the law. Here the etymology must decide. The vowel is long or short as the grammatical form may require ; thus in n'jri;!, ^^"J? , 'i^"'^^'?r Gen. 22 : 8, which follow the analogy of Viijp'^ , and in ''ino I^^- 1^ : 34, ii^^ the first vowel is short ; in D?ri?^ » ^^^- ^^^^ first vowel is long. In a few instances the grammatical form in which Hhirik is employed is itself doubtful ; the distinction is then made by means of Methegh, §44, which is added to the vowel- sign if it is long, but not if it is short ; thus, ^fi«'i,'? yi-rii, from ^'b^ ^0 facir, and ^bi^'? yi-sltnu from ]tp^ to sleep ; but ^S'^'? yir-ii from Tk") to see, and ^biB';' yisli-nu from r.biu to do a second time. 2. Kamets d and Kamets-Hhatuph o are both repre- sented by the same sign ( t ), but may be distinguished by rules similar to those just given. In an unaccented simple syllable it is Kamets ; in an unaccented mixed syllable it is Kamets-Hhatuph ; in an accented syllable, whether simple or mixed it is Kamets, e. g. "li"^ dd-bhdr, ^icsn hhojjJt-sJti, T\yq md-veth, fiTsS ldm-7nd, d'^Pa hot-tim. Before a letter with simple Sh'va, the distinction is mostly made by Methegh, §44 ; without Methegh it is always Kamets-Hhatuph, with it commonly Kamets, e. g. 5T^^n hhokJi-md, n^DH Jihd-M'md. Before a guttural with Hhateph-Kamets or Kamets-Hhatuph §19 AMBIGUOUS SIGNS, 25 it is frequently 6, though standing in a simple syllable and accompanied by Methegh, e. g. '•'inin ho-h¥n, oi^l^n to- obhdhcm. The surest criterion, however, and in many cases the only decisive one, is found in the etymology. If the vowel be derived from Hholem, or the grammatical form re- quires an or a short vowel, it is Kamets-Hhatuph ; but if it be derived from Pattahh, or the form requires an a or a long vowel, it is Kamets : thus f^i'?^|i with the prefixed con- junction vo°niyybtU, "^tP^v? with the article h[h°niyya ; ^h'^.^„ in the Hophal yo%mdh, ^n>s:n^ Isa. 44 : 13 in the Piel yfha"- rP/if(. The first vowel is u in D!^'bn^ fro^^i "'"^j ^^'"^1?^ fro^^ t-jp, nvirrij from TiJ-)iiJ, ''';^~)?ir^ Isa. 38 : 14, '^^-nn]^ Num. 22 : 11, ■'IJ-n'ix Num. 23 : 7 and the like, and the first two vow- els in such words as D^^^'S from b:^s, D^DiJ'o Isa. 30 -. 12 from CS573, nb:iy^ Deut. 20 : 2, ^ro;^ Hos. 13 : 14, ^t:];: 2 Chron. 10:10, t:y"b3)^ 2 Kin. 15:10, because they are shortened from Hholem. On the other hand the first vowel is a in ^■Hriil) Job 1G:19 from ihi^, Q-'irnn from TiJnn, \nm from n^DS , and in ^^k'^ , P^^^I? and the like, because it is originally and properly Kamets. The word •"''^^^ is in Ps. 86 : 2 the imperative sltomra, in Job 10:12 the preterite shCmrCi. a. In a very few instances Kamets-Hhatuph is found in a syllable bearing a conjunctive accent, viz.: "^E")"! Ps. 38:21, bs Ps. 35:10, also Prov. 19: 7 (in some copies), and in the judgment of Ewald 'iJO Judg. 19 : 5, comp. ver. 8 and DS Ezck. 41 : 25 ; in Dan. 11 : 12 Dln"i the points belong to the marginal reading DTi, and the vowel is consequently Ka- mets. There are also a few cases in which Kamets remains in a mixed syllable, deprived of its accent by Makkeph, §43, without receiving Methegh, viz. : -P3^ Ps. 16 : 5, "D-np Ps. 55: 19, 22, "^5 Ps. 74:5; and a final unaccented Kamets is not affected by the insertion of Daghesh-forte conjunctive, § 24, in the initial letter of the following word, e. g. c\l" Pina^ Gen. 31 : 13. When an accent takes the place of Methegh. it serves equally to distinguish*? from o, e. g. ^125:1 Ex. 21:22 v'ndglrphu. ''""sr^t Ex. 21 : 35 umakJi'ru. §45. 5. h. Inasmuch as J^"^"^ is derived from "iniD uinhhar, its first vowel might be suspected to be a; but as it is so constantly written with Hhateph-Kamets, the preceding vowel is probably conformed 1o it. It is consequently regarded and pronounced as 6. Kimchi (Mikhlol, fol. 188) declares that the first vowel in •^h'-i'^ 1 Sam. 13: 21, ni:h-i^ Eccles. 12: 11 26 ORTHOGRAPHY. § 20 and vb^^ Num. 24 : 7 was universally held to be Kamets, and that with the exception of Rabbi Jonah ben Gannach, who was of a contrary mind, the same unanimity prevailed in regard to the first vowel of '{^'i'^ Ezek. 40: 43. As, however, this last word is in every other place written without the Methegh, and there is no analogy for such words as those mentioned above having a in their initial syllable, the best authorities are now agreed that the vowel is 6. and the words are accordingly read dorbhan, etc. In T^'£^:i^ jasper, and rj^in emerald, Ezek. 28:13. which are mentioned by Kimchi in the same connection, the first vowel is Kamets. c. In some manuscripts and a few of the older printed books, e. g. Ste- phanus' Hebrew Bible and Reuchlin's Rudimenta Hebraica, Kamets- Hhatuph is denoted by ( t: ). It then differs from Kamets, but is liable to be confounded with Hhateph-Kamets. It can, however, be distinguished from it by the circumstance that Kamets-Hhatuph is always followed either by simple Sh'va, Daghesh-forte, or Methegh ; none of which ever immediately succeed Hhateph-Kamets. Such a Ibrm as i^2p Ezek. 26: 9 in the editions of Michaelis and Van der Hooght is an impossible one if(T: ) have its ordinary meaning. d. It is surprising that in so minute and careful a system of orthogra- phy as that of the Masorites, there should be no symbol for 6 distinct from that for a; and some have felt constrained in consequence to suppose that the signs for these two vowels were originally different, but became assimilated in the course of transcription. This seems unlikely, however. The probability is that a and 6. whose resemblance even we can perceive, were so closely allied in the genuine Hebrew pronunciation, that one sign was thought sufficient to represent them, especially as the Masorites were intent simply on indicating sounds without concerning themselves with grammatical relations. § 20. 1. As simple Sh'va is vocal at the beginning of a syllable and silent at its close, there can be no doubt as to its character when it stands under initial or final letters. Pre- ceding the first vowel of a word it must of course be vocal, and following the last vowel it must be silent, DP.'idt zliliartCim, n"ibT zukhart. In the middle of a word, the -question whether it belongs to the syllable of the precedirrg or the following vowel must be determined by the circumstances. If a com- plete syllable precedes, that is, either an unaccented long vowel or a vowelless consonant serving as the complement of a previous short vowel, it is vocal. If it be preceded by a short vowel which caimot make a complete syllable with- out the aid of a following consonant, or by a long accented § 21 DAGHESH-LENE. 27 vowel, it is silent : ^^^) zb-lilir^, ^"^STn tiz-k'ra, ^"^^T zikh-ru, riDbiifpn tiktoUia. Sli'va under a letter doubled by Daghesh- forte, ^ 23, is vocal, such a letter being equivalent to two, the first of which completes the previous syllable, and the second begins the syllable which follows : Q'^'^S'^ = D'^'^^n'^ haz-z kJidrlm. 2. In addition to this it is to be remarked that Sh'va is vocal after what may be called intermediate syllables ; that is to say, when the consonant under which it stands per- forms, as it occasionally does, the double office of completing one syllable and beginning the next. Thus, when it follows a consonant from which Daghesh-forte has been omitted, ^iapn;';) vayhhalcshii for vay-yhltak-k'shu, or the first of two similar letters, in order that the reduplication may be made more distinct, iSbn haVlu, n^bp kiVlatli, ibb^i tsiVh, ^bbs aVlay, "'j^pn hkiUkt, and in several other cases, which will be more particularly described in § 22. a. The same double office is performed by gutturals beginning one syllable and yet inclining to complete the one before it. § 18. 2. c. In n^n, ibr example. 5 belongs in a measure to both syllables. It properly begins the second, and yet it is preceded by a short vowel just as if il ended the first, which is accordingly to be reckoned an intermediate sylla- ble, being in strictness neither simple nor mixed, but partaking of tho nature of both. Daghesh-Lene. § 21. The second class of signs added to the Hebrew text are those which are designed to guide in the pronuncia- tion of the consonants. These are the diacritical point over Shin, Daghesh-lene, Daghesh-forte, Mappik, and Raphe. The use of the first of these has already been sufficiently explained, § 3. 1. 1. Daghesh-lene (bp t^rn) is a point inserted in the six letters n s D "i ^ n (technically called B'yhidh K'^phath), to indicate the loss of their aspiration, e. g. 3 bh, 3 b, etc. 28 ORTHOGRAPHY. ^ 21 As these letters are always aspirated after a vowel-soimtl, however shght, and never as an initial ntteranee or when fol- lowing a consonant, they invariably require Daghesli-lene whenever they are not immediately preceded by a vowel or a vocal Sh'va. It is consequently inserted in the initial aspirate of a word which begins a verse, n'^T2Ji<"ia Gen. 1:1, or which follows a word bearing a disjunctive accent (inas- much as this represents a pause of longer or shorter dura- tion), \V5 in-in^ Ex. 1 : 1, "js i nr^y Gen. 3 : 22, or ending in a consonant, "?3-^i?, ^^T- H"^!?^^ Gen. 24 : 42 ; but not if it fol- lows a word ending in a vowel and having a conjunctive accent, Qinri "i^s, ^nn nn-^n Gen. 1:2. The sacred name nin^ is followed by Daghesli-lene, even though it may have a conjunctive accent. Num. 10 : 29, Deut. 3 : 26, Josh. 10 : 30, 11 : 8, Ps. 18 : 21, because in reading the Jews always sub- stitute for it the word "'i'l^i-, which ends in a consonant. In a very few cases, however, e. g. Din ''3'ii* Ps. 68 : 18, ^nn-l]? Isa. 34:11, nn ibio Ezek. 23:42, Daghesh-lene is not in- serted after a vowel-letter, which retains its consonant sound. 2. Daghesh-lene is inserted in a medial or final aspirate preceded by a vowelless consonant, whether this be accom- panied by silent Sh'va or Pattahh furtive, c. g. n'inDD, n?iaT»; but not if it be preceded by a vowel or vocal Sh'va, whether simple or compound, e, g. fT^4^^, ^'^7-?'5? • a. Tlic primary signification of the name Daghesh is commonly ex- plained from the Syriac v-jJ^. ? (l^W), to which Castcllusin his lexicon gives the sense of piercing: This is by some applied to tlic puncture or point which is its written sign, by others to its power of sharpening the sound of letters by removing their aspiration or doubling them. Buxtorf, however, in his Chaldee Lexicon, disputes the existence of such a root in either Syriac or Chaldee, alleging that in Prov. 12: 18, the passage quoted to prove the Avord, the true reading is ].M.y} (XDJi). The six letters which receive Daghesh-lene in Hebrew have the same twofold pronunciation ni Sj^riac, a red dot called Riikhokh {^sob softness), being written beneath them when they were to be aspirated, and another called Kushoi (^ -"^-'i," hardness), being written above them when they were not. § 23 DAGHESH-LENE. 29 b. Grammarians are not agreed whether tlie aspirated or unaspirated Kound of these consonants was tlie original one. There being no data for the settlement of the question, each decides it by his own theory of pho- netic changes. Tiie correctness of the Masoretic punctuation has some- times been questioned in regard to this matter, on the ground of the im- probability of such fluctuation in the sound of these letters in the same word. But besides the Syriac analogy just referred to, the Sanskrit lan- guage shows the almost unlimited extent to which euphonic changes may be carried by a people possessing a sensitive and discriminating ear. The Sanskrit aspirates, besides being subjected to other mutations which can- not here be detailed, regularly lose their aspiration when finals, and under certain conditions when medials, throwing it back, where this is possible, upon a previous letter. Bopp Kritische Grammatik, pp. 30. 42. Similar laws prevail to some extent in Greek, c. g. 6f)L^, Tpt;(o?; rplcpo), 6[ji\po); 6v(ji, (.Tvdrjv ; ovK 1-^m, ovx e^w ; /xc^*' v/xiv. § 22. The absence of Daghesh-lene in an aspirate some- times shows a preceding simple Sh'va to be vocal when this would not otherwise have been kno^vn. In most of the cases referred to, a letter originally belonging to the succeeding syllable is by the prefixing of a short vowel drawn back to complete the syllable before it ; instead, however, of giving up its previous connection altogether, it forms an interme- diate syllabic, § 20. 2, the Sh'va remaining vocal though the antecedent vowel is short ; thus, ^nb Vhliabh with the prefix 3 becomes nnba UVhhahU, not lab^ Ul-hahh. a. The particular instances in which this may occur arc the following, viz. : (1) The Kal imperative of verbs and the Kal infinitive with suflixes, e. g. 'ns?, n23, cn::J, ^-"ins from *12?. ; yet with occasional exceptions, as CSEDxa Lev.' 23 : 39.' (2) Those forms of Pe Guttural verbs in which the first radical assumes a short vowel in place of the silent Sh'va in the reg- ular inflexion, e. g. ^"inr^, on^^n for ^"is^";, Ciarpi. (3) The construct plural of nouns iT^S from c^nzs, risrs li-om r.isr3, r.'.dnn from n'ii~ri. though with occasional exceptions, as ■'E':3"i Cant. 8 : G. bui "2UJ-! Ps. 76 : 4. •1^7:^ Isa. 5 : 10, r-,snn Ps; 69 : 10, ^is*;3 Gen. 50: 23, but cn-'Sna Judg. 7:6; ■'"53; "iija from n;2 are peculiar in omitting Daghesh in the singular with sufiixes. (4) Three feminine nouns ending in ri, Msb^ from Ti'^c, r^nb^ from l^.'l, r,"5 (only occurring with suffixes) from 'izs, but not r'':i"i'0 Also a few other nouns of diff^erent forms, viz. : O^ri^"!^ . but "^rs"}^ , J^^S'^V' : ■'nns, a^rn';3, cirii-iTS, ■,'n35<, bxrp'; Josh. 15:38^ 'cynp^ JoVh. IsVoG. (5) After prefixes, as He interrogative, e.g. cri^T'^n Gen. 29:5 from ^ri^'l?. and inseparable prepositions, e. g. "i"'^"}' from ~"'~'n . rii"i3 from r;"n, "iS'is from ~3':i. Usage is not unilbrm in the case of Kal infinitives follow- 30 ORTHOGRAPHY. §23 ing inseparable prepositions, e. g. S'^ri^t') -'^'^?3; Vs33, ^333, bbsb; i<32!p Isa. 31 :4, snrib Num. 4: 23, 8: 24; nHsb, ^3]3. (6) Tlie'suffixes of the second person Tj, C3 , '|3 never receive Daghesh-lene, T]53, Di~iiD3. Tliese rules are sometimes of importance in etymology; thus. Ti'^3'i3T^ Ezek. 27:12 must have as its ground form 'i'i3-t3J, not "iisi^ ; and C3~i5< Hos. 7 ;6, cifiH Ps. 90: 10 cannot be infinitives with suffixes, but must be from the seghoiates 3"iX, 3n'"i. 6. The omission of Daghesh-lene in the final letter of ribin Prov. 30 : 6, abbreviated from rj^oin or r|Din, is exceptional. The Daghesh occasion- ally occurring in initial aspirates after words ending in a vowel and having a conjunctive accent, is best explained not as an exception to the ordinary rules, but as Daghesh forte conjunctive, §24, e. g. "(^33 '^^'^';l Gen. 11 :31 and elsewhere, V^a nrinx Gen. 46: 28, ni whose last letter is changed to n to conform with that which follows ; *^aD from nio. 2. When the reduplication is indicative of a particular grammatical form the Daghesh- forte is called characteristic, e. g. in the Piel, Pual, and Hithpaehof verbs ; as, tjin, ^fr?^"!, and certain forms of nouns, as, ^isa . 3. When it has arisen from the necessity of con- verting a previous simple syllable into a mixed one in order to preserve the quantity of a short vowel which it contains, it is Daghesh-forte conservative; e. g. 15;^ for so;'. 4. When the initial letter of a word is doubled under the influence of the final vowel of the word preceding, it is Daghesh-forte conjunctive; e. g. ^^~'^, ^^""^n;;', ^,s2i ^mp. 5. When the last letter of an intermediate syllable is doubled in order to make the following hiatus or vocal Sh'va more distinct, it is Daghesh-forte dirimens or separative, because the letter which receives it is thus separated in part from the syllable to which it belongs ; e. g. "'isy innhhv, for "^b.TJ iubJiE. 6. When the first letter of a final syllable is doubled under the influence of a previous vowel bearing the accent (mostly a pause ac- cent, § 30. 2. ^.) 5 f<^i' the sake of increased fuflness and force of pronunciation, it is Daghesh-forte emphatic ; e. g. '^^in for '^~X} . In the first three uses named above Daghesh-forte is said to be essential, in the last three it is euphonic. a. Dan^hesh-forte conjunctive occurs regularly after the pronoun fi^, e. g. D"'2.':"i"n2^ ir^ nn Ps. 133: 1, and in a multitude of cases after final Ka- mets or Seghol in words accented on the penult or followed by Makkeph, § 43, e. g. rxrnn;?b Gen.2 : 23, cik^-nlprs Deut.27 : 7, ii nn";nT Num. 25 : 13, ^2-nr2;^ Gen. 30:"33; n^.n^'^A Num.' 34:6, 7. 9, n^ix^'nlj-a Ex. 13:1 (wliere the accent is on the ultimate), T|'^"'"'??^ Prov. 15: 1 (in some edi- tions), more rarely after other vowels, e.g. ^XJJ V^>ip Gen. 19:14, sb ^T^X^i 1 Sam. 8: 19, once after the liquid n, e. g. N^ i^.N'^ 1 Kin. 11: 22. See also § 22. 6. In a few instances words thus united are written as one, e. g. ri;TT9 Ex. 4:2 for n|f no, so cd^t? Isa. 3:15, nx^n^ Mai. 1:13, nxsxoa §25 DAGHESH-FORTE. 33 Isa. 27:8, and possibl}'- C^x-iN Isa. 33:7. See Dr. Alexanclcr'.s Com- mentary upon this passage. b. Daghesli-forte separative occurs only in the following examples : n-iSS Hos. 3: 2. T]ini'-i;3"'3 Ps. 45: 10. nini-Tn Am. 5 : 25. PinirSJ^rn Gen. 18:21. ' rphsn Gen. 37 : 32. 'i=^n Gen. 17:17. Dn"'N':in 1 Sam. 10: 24, 17 : 25, 2 Kin. 6: 32. cniixsr; .Tob 17:2. is^Esn Ex. 2:3. sinsi'i'rin (?) Judg.20: 43. nT:5.'':iri 1 Sam. 1 : 6. -i|5^n Isa. 57:6. nn;?"^ Gen. 49:10. rnts^b Prov.30:17. 1 Sam. 28 : 10 (?) Ezek. 13:20. Isa. 33:1. orinnii'oPs. 89 :' 45. Joel 1:17. Job 9:18. Nah. 3 : 17. Ex. 15: 17. Deut. 23:11. Job 30 : 8. Ps. 141:3. Prov. 4: 13. Judg. 20 : 32. Jer. 4 : 7. ibao Isa. 9:3, 10: t: ■. ' 27. "^•ZIV Deut. 32:32. •inhs:? 2 Sam. 23: ' 27, Jer. 29: 27. C2'^32? Isa. 58 : 3. C3-in-iiy Am. 5 :21. '■' -^'^isiU?) Cant. 1:8. ni::|3a Ps. 89:52. rpnir^ja Ps. 77 : 20. T^iZ'2i'J Prov. 27 : 25. iDnp52:s Ps. 119; 139. •':!inn5::i Ps. 88: 17. tninVp. (?) Ps. 37: 15.1sa.5:28. -^bnOJ Zech. 4 : 12. bnb2ti Ps. 58:9. This list is corrected and enlarged from Gesen. Lehrg. pp. 86 it Those words which are followed by a note of interrogation (?) are found in some editions but not in others. Daghesh separative may be found after He interrogative in some instances not included in the above list. c. Daghesh-forte emphatic occurs only in ^Hn Judg. 5 : 7, 1 Sam. 2:5; Jl^nil Job 29 : 21 ; ^nn^ or ^nn;; Job 21 : 13 ; WSj'-; Isar 33: 12, Jer. 51 : 58 ; na'ib Ezek. 21:15; ^laipj Ezek. 6:9; nrjOJ Jer. 51:30; ^-ins Ezek. 27: loV^^p (?) Isa. 19:6; and probably il^nn'nJob 13 : 9 (not in pause). § 25. In order to the distinct utterance of a reduplicated consonant, it must be followed as well as preceded by a vowel-sound. Dagliesh-forte is consequently never written in a final vowelless letter, with the exception of the two words PS , prip , both of which end in aspirates whose pronuncia- tion would be changed by the removal of the Daghesh. In every other instance the doubling is neglected, even though the letter be an aspirate, which will for this reason resume its aspiration ; e. g.^p., V?jb; 3D, "^^6; T\h'^^ abridged from nsn^'i; qi;»n from ns^^^ In a medial letter with Sli'va Daghesh may be written, because the Sh'va being thus ren- dered vocal the reduplication can be made audible by means 3 34 ORTHOGRAPHY. § 2G, 27 of the hiatus which it represents ; it is, however, quite as frequently omitted, the Sh'va commonly remaining vocal as if it were inserted, and compound Sh'va being occasionally substituted for simple to indicate this fact, § 16. 3. h. ; e. g. D'ln'iy for D"''7!i2? , ixc? for iiS'BS , particularly after prefixes, as Vav conversive, the article and preposition "a, so ^V>T^, T2J|n'cn. It is seldom omitted from a medial aspirate on account of the change in its sound involved : yet even this is done occasion- ally, e. g. n^:2n^ Judg. 8 : 2 for n'^itn^, ^^nn Isa. 22 : 10 for ^snn, "jiiDT from 'ji">3T. In a few rare instances it is dropped from a letter followed by a vowel, when the laws of syllables will permit and the pronunciation will not be materially affected ; e. g. nsiyn Ruth 1:13 for nsiyn . Mappik. § 2G. Mappik (p''|'52 bringing out or uttering), is a point in one of the letters i5 n 1 "i, showing that it represents a consonant and not a vowel, or in other words that it does not quiesce in the preceding vow^el-sign. It is unnecessary, however, to employ any notation for this purpose in the case of « 1 and '' , for their quiescence can be readily determined in all cases by the rules already given, ^13. Although it is much more extensively used in manuscripts, therefore, ]\Iap- pilv is in modern editions of the Hebrew Bible only inserted in final n when it retains its consonantal power ; c. g. Siini? arfsah, ns^is* artsd, rnnpb rh'ihhCth, nri)pb laJc'hJia. The point four times found in i5, ^S5^n^^ Gen. 43 : 2G, Ezra 8:18, •^s^nn Lev. 23 : 17, ^55"^ Job 33 : 21, though called aDaghesh in the Masoretic notes in the margin, is probably to be re- garded as Mappik. Raphe. § 27. Raphe (nsn weak), is a small horizontal stroke placed over a letter, and denotes the opposite of Daghesh- ^ 28 ACCENTS. 35 lene, Daghesh-forte, or IMappik, as the case may be. As no mconveriicnce can arise from its omission, it is only occa- sionally used in modern Bibles, and not with entire uni- formity in the different editions. It is chiefly found where a Mappik has been omitted in n , which according to analogy might be expected to be inserted, e.g. n'lC^n Ex. 9:18, rn?i?^ Lev. 13:4, f-'ijifna Num. 15:28, nb Num. 32:42, nrDiETS Job 31 : 22 in some copies. In 'T\}-r^z^r\ Ex. 20:4, Deut. 5:8, it is the opposite of Daghesh-forte, and shows that b may either be doubled agreeably to the point in its bosom or not. In nknr, tfb Ex. 20 : 13, Deut. 5:17, it is the opposite of Daghesh-lene, and shows that the r may either have its unaspirated sound, as the Daghcsh indicates, or may be aspu-ated. It is often referred to in the marginal Masoretic notes even where it is no longer found in the text, e.g. Judg. 10:10, 28. Accents § 28. The third class of :Masoretic additions to the text are those which relate to the words. These are the accents, Makkeph, Methegh, and the K'ri. An accent (nyb) is writ- ten upon every Avord with a twofold design, 1st, of marking its tone-.syllable, and 2dly, of indicating its relation to other words in the sentence. The great number of the accents has respect entirely to this second function, there being no difference in the quality of the stress laid upon particular syllables, such for example as is marked by the Greek acute, grave, and circumflex, but only that difference in its amount which arises from the unequal emphasis naturally laid upon the different members of a clause or period. Tlie punc- tuators have attempted not only to indicate the pauses to be made in reading, as is done by the stops in use in other lan- guages, but to represent to the eye the precise position held 86 ORTHOGRAPHY. § 29 by each word in tlie structure of the sentence, and the various grades of attraction or repulsion arising from the re- lations whether co-ordinate or subordinate which subsist among them. Every sentence is fancifully regarded as a ter- ritory, which, partitioned into its several clauses, forms em- pires, kingdoms, and principalities, ruled by their respective sovereigns, each of whom has his own train of inferiors and dependants. The accents are accordingly divided into Dis- junctives or Rulers (a^pb^), and Conjunctives or Servants (D^'in?^). The former indicate that the word upon which they are placed is more or less sej)arated from those that follow ; they mark thus the end of a clause or of the section of a clause over which they exert control. The latter indi- cate that the w^ord over or under which they are written is connected with what follows and belongs to the clause or section ruled by the next succeeding Disjunctive. a. The stress of voice denoted by the accent must not be confounded with quantity. An accented syllable may nevertheless be short, the energy with which it is pronounced not necessarily affecting its length. h. The Jews made use of the accents as musical notes in the cantilla- tion of the synagogue, whence they arc also called ni3''3i3. In the judg- ment of some this is a part, and perhaps a leading part, of their original design. Their great variety, the frequent occurrence of accents of oppo- site powers upon the same word, and the distinct system of poetical accents, favor this opinion. Such as are curious to know the details may find the mode of their employment for this purpose explained at length in Bartoloccii Bibliothoca Magna Rabbinica, vol. iv. pp. 427-444. § 29. The Disjunctive accents may be divided into four classes of various rank or power, as follows, viz : Class I. Emperors. *1. Silluk (.) :p!i^D *2. Athnahh L) nsnst §29 ACCENTS. Class II. Kings. 3. S'gholta (■••) T : : 2)ostp, 4. Zakeph Katon C) '&^P, tll?.I 0. Zakeph < Gadhol (') Piia ?);?t *6. Tiplihha (J Class III. Lulces. xnso *7. R'bhr (■) T'^i"} *8. Shalshel eth (') '^^}^'!^P. *9. Zarka {-) Xj^'lT postp. 10. Pashta (~) T : - postp. 11. Y'thibh (<) 131 n-i^^ prep. 12. T'bhir (,) •n-inn 37 Class IV. Counts. ns. Pazer / ^ \ 1TQ 14. Karne Phara I '^''\ n^ )£ "ij-i;^ 15. T'lisha Gh'dhola I ^\ nbinS xsij-'bri prep. 16. Geresli /^ \ ^l? 17. G'rashayim / "\ n-iirj-iri *18. P'sik (I) ip-^DQ The Conjunctive accents, or Servants, are the following, VIZ. *19. Merka \i / "'^^'^^ *20. Munahh \ J y 21. Merka Kh'phula ^ ti ' !ibl53 ss-i^a •22. Malipiikh V< / '^p.fl'? 23. Darga \j / ^^1'^ *24. Kudhma / ^ \ NB'lp *25. Yerahh ben Yomo \ v/ ■irr-ja nn;; 26. T'lisha K'taiina r"^ . nsijp N'^ibn ;jos/:p. 38 ORTHOGRAPHY. § 30 a. Merka Kh'phula has sometimes been reckoned among the Disjunc- tives, as by Gesenius in his Lehrgebaude ; but the absence of Daghesh- lene in the word following that on which it stands in Ex. 5:15, Ezek. 14:4, proves that it is a Conjunctive. b. According to their most probable significations, the names of the accents appear to be in part borrowed from their forms and in part from their uses. Thus the Disjunctives: Silluk, end; Athnahh, rest ; Segholta, bunch of grapes ; Zakeph, small and great, causing suspension ; Tiphhha, palm of the hand ; Whhi", square or reposing; Shalsheleth, chain; Zarka, dispersion; Pashta, c.vpansio7i or letting dawn (the voice); Y'thibh, s?'/- ting still; T'bhir, interruption; Fazer, sepajatur ; Karnc Phara, a heifer^s horns; T'lisha, great and small, shield ; Geresh, expidsioji ; G'rashayim, double Geresh; P'sik, cut off. Conjunctives: Merka, prolonging; Mu- nahh, (a trumpet) at rest, i. e., in its proper position ; Merka Kh'phula, double Merka ; Mahpakh, (a trumpet) ^^fe/VecZ; Darga, pro^-ress ; Kadh- ma, beginning ; Yerahh ben-Yomo, moon a day old. Other names are given to some of these accents, particularly where they occur in certain situations or combinations; thus Tiphhha is also called Tarhha ( ^r\y^_ ), Munahh with P'sik is called L'gharmeh ( i^ansb ), etc. c. The classification of the Disjunctives, according to their respective powers and the laws of their consecution, has been the work of Christian writers, from whom all accurate investigations of the accentual system have proceeded. In fact, this whole subject is treated by the Jewish grammarians in the crudest and most perplexed manner. Buxtorf says, in his Thesaurus Grammaticus, p. 45 : Accentuum ratio hactenus nee a quo- quam nostrorum nee ab ipsis etiam Hebraeis sufficienter explicata est. The division exhibited above is the one now commonly adopted. The current names, Imperatores, Reges, Duces, Comites, are those used by Wasmuth in his Institutio Accent. Heb. 1664. Others have divided them difTerently. The learned PfeifTer, author of the Dubia Vexata, distin- guishes one Emperor, one Archduke, four Dukes, seven Counts, and five Barons. Boston, the well-known author of the Fourfold State, in an elab- orate Latin treatise upon this subject left by him in manuscript and pub- lished shortly after his death, distributes them into three classes of superior and one of inferior rank. Mention is made, in a commendatory prefl\ce by Mill, the distinguished critic of the New Testament, of another manuscript in English, in which Boston applied his views practically in a twofold translation of the first twenty-three chapters of Genesis, with copious notes, both philological and theological. This, it is believed, has never been published. A curious little book upon the Canon by Ferdinand Parkhurst, London, 1660, makes six Regal and ten Principial Disjunctives, Y'thibh and P'sik being omitted altogether. §30.1. fourteen of the accents are written over, and eleven under, the words to which they are attached. P'sik, whose only use is to modify the power of other accents, is written after the word to which it belongs, and in the same line ^ 30 ACCENTS. 39 with it. The place of the accents is either over or under the letter preceding the tone-vowel, with the exception of the prepositives Y'thibh and T'lisha Gh'dhola, which always ac- company the initial letter of the word, and the postpositives S'gholta, Zarka, Pashta, and T'lisha K'tanna, which stand upon the final letter. Y'thibh is only used when the first is the tone-syllable. Pashta is repeated if the word on which it stands is accented on the penult, e. g. ^nn Gen. 1 : 2, or ends with two vowclless letters, e. g. nil^^l Huth 3 : 7, or if the last letter has Pattahh furtive, e. g. ^1'' Gen. 33 : 13, and in some manuscripts and editions there is a like repetition of S'gholta and Zarka. When a word bears the other preposi- tive or postpositives, there is nothing to mark its tone-syllable unless this may chance to be the one upon which the nature of the accent in question requires it to be placed. 2. Silluk has the same form as Methegh, ^44; but the former invariably stands on the tone-syllable of the last word in the verse, while Methegh is never written under a tone-syllable. Pashta is likewise distinguished from Kadhma only by its position upon the last letter of the word, and after the superscribed vowel, if there be one, e. g. '^TS^f Gen. 1 : 7, i^ij^^ Gen. 24 : 7, while Kadhma is placed upon the letter preceding the tone-vowel, e. g. "lies Gen. 2 : 19 : where this chances to be a final letter the laws of consecution only can decide ; thus, in 'n?'?! Gen. 2G : 4, ^nxs Deut. 16:3, the accent is Pashta, but in ?i?-iTb;i Gen. 17 : 8, ?inx2 1 Sam. 29 : G, it is Kadhma. Y'thibh is distinguished from Mahpakh by being written under the first letter of the word and taking precedence of its vowel if this be subscribed, e. g. ^ii?y Gen. 1 : 11, •>? Gen. 31 : 6, Deut. 10 : lY ; Mahpakh belongs under the consonant which precedes the tone-vowel, and after its vowel-sign if this be subscribed, e. g. ^'r}2'r\ Gen. 2:14, "'I Gen. 32 : 33, Deut. 4 : 7. When the initial syllable bears the tone and there is no subscribed vowel, the laws of consecu- tion must decide ; thus, in S^n the accent is Y'thibh in Gen. 40 ORTHOGRAPHY. § 31 3:15, 44:17; Deut. 10:17; but Malipakli in Josh. 17:1. § 31. The accents abeady explamed are called the prosaic accents, and are found in all the books of the Old Testament with the exception of the Psalms (n^!jrin), Proverbs C^l?!^^), and the poetic portion of Job (^'i'^i?), whose initials form the technical word ti^DS? . Here a different system of accentua- tion prevails. Thirteen of the prosaic accents, one-half of the whole number, nowhere occur in the books just named, viz, : S'gholta, Zakeph-Katon, and Zakeph-Gadhol of the Kings, Pashta, Y'thibh, and T'bhir of the Dukes, Karne Phara, T'lisha Gh'dhola, Geresh, and G'rashayim, of the Counts, Merka Kh'phula, Darga, and T'lisha K'tanna of the Con- junctives. Such as are common to both systems are in the previous table distinguished by an asterisk. The powers of some of these, however, are altered, so that a new arrange- ment of them is necessary ; and they are supplemented by additional signs formed by combining the prosaic accents or assigning them unusual positions. The scale of the poetical or metrical accents thus constituted is as follows, viz. : Disjunctive Accents. Class I. 1. Silluk (',) : li^Sfl 2. Athnahh (J Ti=sirj 3. Merka-Mahpakh C) Tiasn 4. R-'bhi" o ninsn 5. Pazer C) li^sn 6. R'bhi'* Geresh (•') liisri T. Tiphhha initial (.) ni35ri prep. 8. Zarka D •ninsn posip. 9. P'sik (0 1 "lissn postp. ^32 POSITION 1 or THE ACCENT, CoXJtIK CTivE Accents. 10. Merka ^ J ^ ninsri 11. Merka-Zarka ■""irsn 12. Mahpakh \< / "niasn 13. Mahpakli-Zarka *^'^t?^! 14. Munahh ^ J / ^i3^r^ 15. Muriahh superior / -J V liiifi 16. Yerahh ben Yomo V v/ Ti^sn 17. Kadhma / > \ liisn IS. Tiphhha V V. / -ii::3n 19. Shalsheleth ( ' \ ninsn 41 a. It will be perceived that there are fewer Disjunctives but more Conjunctives than are exhibited by the prosaic accents. Merka-Malipakh answers substantially to S'gholta; RMibi^-Geresh to Tiphhha before Silluk, and Tiphhha initial to Tiphhha before Athnahh. Tiphhha and Shalshe- leth are transferred from the list of Disjunctives to that of the Conjunc- tives, whence it comes to pass that if a word bearing either of these accents terminates in a vowel, Daghesh-lene will not be inserted in a fol- lowing initial aspirate, e. g. Oli'ia '^^^'^'^ Ps. 31 : 10, n"inrs N-':^ Prov. 8 : 3, m'53T523 Vasn';! Ps. 10:2. b. P'sik, in the poetic as in the prosaic accents, is never used alone but always in conjunction with another accent. It serves to strengthen Dis- junctives and to reduce the power of Conjunctives without disturbing the order of their consecution. It is thus used with Merka-Mahpakh Ps. 5 : 13, Fazer Ps. 10: 14, Tiphhha initial Ps. 31:4, Mahpakh Ps. 5 : 9, Munahh Prov. 1:22. Merka Ps. 10:13. Kadhma Ps 10:5. Shalsheleth Ps. 7:6. Position of the Accent § 32. The accent in Hebrew may fall either upon the ultimate or the penultimate syllable, but never at a greater remove from the end of the word. In the former case words are technically termed Milra (^i'nb^ from below), and in the latter Milel ("J^^bt! from above). 42 ORTHOGRAPHY. § 33 1. Tlie position of the accent may be considered in rela- tion either to the syllabic or to the etymological strncture of a word, that is to say, as aflected by the nature of its syl- lables on the one hand or of the elements ot which it is com- posed as a signiticant part of speech on the other. It is so far determined by the syllabic structure of words, that a long mixed syllable or a short simple syllable, whether in the ultimate or the penultimate, nuist receive the accent, §18. 2. thus: pn2\ np"r!?PT, nra, n-i': . 2. Considered in reference to their etymological structure, words exist in two conditions, (1.) their priniary uninflected state, by wdiich their essential and proper meaning is con- veyed; (2.) with added atfixes and prefixes, by which that meaning is varionsly moditied. In their nude or primary state all words, whether primitives or derivatives, are ac- cented upon the ultimate, and so continue to whatever flexion, involving no terminational appendages, they may be sub- jected. Thus,ii:2, nps, npb, ^j??, n{?s, ^ps^ ^psnn; i%, 3. The only exception is a class of words called Se- ffholates, in Avhicli the last vowel does not belonc; orio-inallv or essentially to the form, but is introduced for the sake of softening the pronunciation, §G1. 2; these are accented on the penultimate, as 1\)^, ^?D, m\, n^2, rnri, r:n3, b.)i , b^), a. 'T^r^j'. I^- 50: S is said to bo tlie only instance of a word ac^'cnted on the antepenult. The proper tone-syllable of this word is the ultimate, but upon the recession of the accent by § 35, the vowel next preceding, which has arisen from Sh'va and is unessential to the form, cannot receive it, so that it necessarily falls upon the one still fnrthtr back. § 33. The additions which words may receive at the be- ginning or end aftect the accent in proportion to the respect- ive weight accorded to them. Additions to the end of words are of two sorts, which may be distinguished as affixes and suffixes. Affixes are so welded to the word or mer2:ed in it ^33 POSITION OF THE ACCENT. 43 that in the popular consciousness they have become an in- tegral part of it, and their independent existence or separate origin is no longer thought of; such are the personal inflec- tions of verbs and the terminations indicating gender and number in nouns and adjectives. Suffixes are not so inti- mately blended with the word to which they are attached as to have lost their individual identity and independent charac- ter, and consequently are of greater weight as respects the accent; such are the fragmentary pronouns appended to verbs, nouns, and prepositions. 1. If the appendage consists of a vowel (as n^, ri, ^, i, \, r.), or begins with one (as l^,, \, ^\, Q\, ^''^, ^., '^-, °,» )^, u\, ^\), and can consequently only be pronounced by the aid of the final consonant of the word to which it is at- tached, it will attract the accent to itself or to its initial vowel from a noun, adjective, participle, or preposition, as ^p^^ , i-b'^ , D-'nn^ , V.^"^ , V2^1 > ^""^^^ from W ; Q^ffi'JI? , '''41'i^ , from TS^p . Such an appendage to a verb, if a suffix, will so far accord with the rule just given as to carry the accent forward one syllabic ; but the accent will remain in its origi- nal position if it be an affix, unless it is either dissyllabic or causes the rejection of the vowel previously accented ; D^^nn with a suffix ^^^"l^T}: > ^^^t with an affix ^^^'^nn j ^iy with a suflix r-qt , but with an affix '^rq'P , -rt^y ; C]b , r.rj? , nrj? ; bj?, n^p , '.^p , T''^p ; ^b^ , ^?0^ ' ^C'^"^ • It is to be obsen^d, however, that a paragogic n^ or n. , §G1. 6, attached to nouns, pronouns, and adverbs, and occasionally a paragogic r does not disturb the position of the accent, e. g. H'^ , nsii?; n:,r.n:; so nrn, n^i?,n7ii?, T^:) Lam. 1:1, but -^nsbi: Isa. 1 : 21 ; neither does the feminine ending r... , which is a Segholate formation, e. g. 'k'TQ , ro^'^'9 • a. Paragogic M^ receives the secondary accent Methegh in n^s, fil^s Gen. 28 : 2^b%. 7. ' 2. The appending of a simple syllable, such as the 44 ORTHOGRAPHY. § 33 suffixes '^P , ^3 , ^n , n , i^ , or the verbal affixes n , "^n , ^5 , n: , will not alter the position of the accent provided it originally stood upon the ultimate ; if, however, its original place was the penult, or if the syllable in question be attached to the word by a union vowel, the accent must be carried forward one syllable to prevent its standing on the antepenult, which is never admissible: nos , ^ns3, i^&s, '^?^5?; i^^^, Ppsia, ^:nDS'59 ; !:)b, rii'jp, ^ri'i'p . Suffixes appended to a word ending with a consonant mostly require a connecting vowel, and con- sequently shift the position of the accent. Affixes, by reason of the less weight accorded to them, commonly do not. The suffix ^ follows the general rule when preceded by a union vowel, but draws the accent upon itself when it is not, e. g. "'t , ^"2^ , ^1^ , T%^ . A consonantal appendage to a long un- accented vowel, inasmuch as it converts the ultimate into a mixed syllable, necessarily draws the accent upon it from the penult, §32. 1, e.g. ^noi?^, ^^nox^, d^npx^; ^in^h^ ^ )^ri^'n^ . 3. A mixed syllabic, whether an affix as DJ?, in, or suffix asDD, ]D, en, )r^, will attract the accent to itself, DPDbn from ^^n ; D52573 , 'ab^^b'n, from ^b-ia ; anb;in from D^nri .In the unusual form D^bs 2 Sam. 23 : 6, the accent stands upon the union vowel. 4. The only prefixes which exercise any influence upon the position of the accent, are the Vav conversive of the futm-e, which draws back the accent from a mixed ultimate to a simple penult, •yh^\ n^s^i, nfc.'i, rna^i ; and the Vav con- versive of the preterite, which throws it forward from the penult to a simple ultimate, n'^iax , J?i^5<'i ,- S?3i^^ , ^^Ti^^l , a. Some languages invariably accent the same part of the word ; thus, Bohemian and Lettish the initial syllable, Polish and Lazian, one of the Caucasian tongues, the penult of all polysyllables. Others, in which more freedom is allowed, have no respect to the etymological structure of words, but are guided entirely by the character of their syllables. Thus, in Arabic and Latin words are accented according to the quantity of the penult; the accent is given to the penult if it is long, to the antepenult § 34, 35 POSITION OF THE ACCENT. 45 if the penult is short. In others still the etymological principle is the prevailing one, and this often has a wider scope than in Hebrew. Thus, in Greek the accent has the range of the last three syllables. In San- scrit it may stand upon any syllable whatever even of the longest words. In English it is almost equally free, e. g. peremptorily, inconsideralion, its removal from its primary position upon the radical portion of the word being conditioned by the respective weight- of the formative syllables ap- pended, e. g. person, personate, personally, personify, personality, per- sonijicdtion. §34. The location of the accent being thus influenced by the etymological structure of words, it may serve to dis- tinguish words of like appearance but different formation. Thus, nn'a Gen. 30:1, nka Gen. 29 : 6, are participles, but T\tfi2 Gen. 35 : 18, "ijli Gen. 29 : 9, are preterites, the femi- nine affix receiving the accent in one case but not in the other, § 33. 1. So ^53 thei/ built from nba, but ^:3 in us; ^ii? thei/ carried captive from !"air, but laia tliey returned from i^iJJ; Tn55 he has seized, but Tri5< Job 23 : 9 I shall see from J"tTn ; V^': it shall be evil from yi?n , 5?:)^ he shall feed from n^"n ; nnt! he ivas rebellious, 'riya it icas bitter from "Ta ; "112^]? arise thou (fern.), ''ti'ip my rising uj). § 35. The position of the accent may be shifted from the following causes, viz. : 1. A Conjunctive is frequently removed from the ultimate to the penult if a Disjunctive immediately follows, whether upon a monosyllable or a dissyllable accented on the penult, in order to prevent the unpleasant concurrence of two ac- cented syllables in closely connected words, e. g. •i^'^^^ i^'^J^ Gen. li's, n^:^ nDia Gen. 4:17, t ^^T^' Dent. 32 : 3G, X)'! Ta^^) Ps. 2 :12, tiV ^^^ Isa- 36:8. In a few excep- tional cases the secondary accent Methegh remains to mark the original tone-syllable, after the principal accent has been thrown back, TZ "^^S"? Num. 24 : 22, r^ ^?.3 Isa. 40 : 7, a:^ 5?^.tjn Dent. 4 : 33. 2. The special emphasis given to the last word of a clause or section, and represented by what are called the ^^ ORTHOGRAPHY. & gg pause accents, § 30, 2, a, is sometimes rendered more distinct by a change of the accented syllable from the ultimate to the penult, e. g. -ipbx ^ .^jij. nnx, nr^. r,^;^^ j^pj^. ^3^^ ^^^, ^^ from the penult to the ultimate, particularlyln the' case of forms with Vav conversive of the future ^b^l , ^br.i ^ so ^^'^ , °r^r^ , '^'9.^'^ . The accent is in a few inVtanceV at- tracted to a short final syllable ending in a weak letter, which either loses its sound entirely, converting the syllable into a srniple one, or requires considerable effort and energy of voice to make it distinctly heard, e. g. xn^ Gen. 41 : 33 for «-i^ • so &{nn Zech. 9 : 5, Mic, 7:10, S^irn Ps. 39 : 14 for ^th^' ' \ Consecution of Accents in Prose. §36. 1. The second use of the accents is to point out the relation of words to one another. The Disjunctives in- dicate a greater or less separation between the word on which they stand and the following one ; the Conjunctives indicate a connection. The greatest separation of all is effected by Silluk, which is written under the last word of every verse, and is followed invariably by two dots vertically placed (:), called Soph Pasiik (p^DS ^d end of the verse). The next in power are Athnahh and S'gholta. When a verse was to be divided into two clauses, Athnahh was placed under the last word of the first clause, Silluk maintaining its position at the end of the verse. If it was to be divided into three clauses, which is the greatest number that any verse can have, the last word of the first clause receives S'gholta, the last word of the second Athnahh, and the last of all Silluk. Verses of one clause range from Gen. 26 : G, containing three words, to such as Jer. 13: 13 and 1 Chron. 28 : 1, con- taining more than twenty : the most common division is into two clauses, e. g. Gen. 1:1; -j^nxn . . . Q^n'bN ; three clauses § 37 CONSECUTION OF ACCENTS IN PROSE. 47 are mucli less frequent, Gen. 1:7 ! 1? • ^"^l?"?!? • • • ^TI^O* 23 : 16, 24 : 30, 2G : 28. a. In Job 1 : 8 S'gliolta occurs in a verse of two clauses without Ath- nahh, probably because tiie accentuation is conformed to that of Job 2 : 3. 2. Eacli of these clauses is capable of subdivision to whatever extent its length or character may seem to demand by the Disjunctives Zakeph Katon, Zakeph Gadhol, Il'bhi^ Pazer, and T'lisha Gh'dhola, according to the number of sec- tions to be made and the various degrees of their completeness. Thus, ill Josh. 1 : 8 the clause of Athnahh is divided into five sections, ^3 . nii^?b . . r^b^'j . , . rps^ . . . i2Jra;i , in 2 Kin. 1 : 6 into six, ^T^^V. . . . l^^i? n!?izJ . ^ky^'J'2. . . ^)'^^ . . n^^s? . The choice of the accent to govern a particular section de- pends not only upon its power, but likewise upon its rank, the more exalted officer standing in ordinary cases nearer the sovereign. Accordingly toward the beginning of a clause an inferior Disjunctive will be used, even though the separa- tion is such as would require an accent of much higher power to indicate it in a more advanced portion of the same clause. These accents, moreover, have not a fixed value like the stops in other languages ; their power is not absolute but relative, and varies endlessly with the circumstances of the case. Athnahh in Gen. 1 : 1 marks the greatest division in the verse, but that is not sufficient to require a comma. In the next verse Zakeph Katon is equal to a semicolon in the first clause and less than a comma in the second. In Gen. 27 : 16 the separation indicated by R'bhi'' is wholly rhythmical. a. Those accents which, as above described, mark the limits of clauses and sections, are denominated pause accents. ^37. In the sections thus created the accents are dis- posed relatively to the Disjunctive which marks its close. Each ruler has his servant and subordinate officer, whose 48 ORTHOGRAPHY. ^37 function it is to wait upon him. In other words, each Dis- junctive is regularly preceded by a particular Conjunctive and inferior Disjunctive ; and the train of accents in each section is formed by arranging the Disjunctives in their fixed order of succession Avith or without their regular Conjunctives until all its words are supplied. The trains proper to the different sections are shown in the foUowino- table : Primart Sections. 1 ft o O O ft k D e >^ c H „ ft O D d |2 m > •-a S5 O O • 1 J (■ ^ .(..) V .c)b' 1 '0 ^o\j. A J .A.) - .(.)>.)' Secondary Skctions. • J J jSiS U • ..0,X) H -1 -1 J J > ;> Unusual Sections. 1 1? V J J J J J § 38 CONSECUTION OF ACCENTS IN PROSE. 49 a. Accents of like forms are readily distinguishable in the table by the column in which they stand. Where perspicuity requires it the distinction will hereafter be made by appending tiieir initial letters, thus : Kadhma ''', Pashta '^, Mahpakh "", Y'thibh'^. § 38. Exjilanation of the Table.- — The trains preceding the three principal accents are exhibited in the horizontal lines of the uppermost division ; those of the ordinary de- pendent sections in the middle division, and those of rare occurrence at the bottom. 1 . Train of Silluk. — If Silluk be immediately preceded by a Conjunctive, it will be Merka ; if a Disjunctive precede it in the same section, with or without an intervening Merka, it will be Tiphhha, Gen. 1:1. If there be a Conjunctive before Tiphhha, it will be Merka, Gen. 1:1; if two Con- junctives, which occurs but fourteen times, they will be Merka Kh'plmla and Darga, Gen. 27:25, Lev. 10:1, 2 Chron. 20 : 30. The next Disjunctive before Tiphhha, in the same section, will be T'bhir, Gen. 1:4. If T'bhir be pre- ceded by one Conjunctive, it will be Darga, Gen. 1 : 12, or Merka, Gen. 1 : 26 ; if by two, the second will be Kadhma, 1 Sam. 15 : 33, or Munahh, Gen. 2:4; and if by three, the third wall be T'lisha K'tanna, Gen. 2:19. The next Disjunctive before T'bhir, in the same section, will be Geresb, Gen. 26 : 11, 27 : 4, or G'rashayim, Ex. 23 : 4. If Geresh be preceded by one Conjunctive, it wiU be Kadhma, Gen. 24 : 7, or Munahh, Isa. 60 : 17 ; if by a second, it will be T'lisha K'tanna, Gen. 2 : 5, or Munahh with P'sik, Gen. 28 : 9 ; if by a third, it will be Munahh, 1 Sam. 14 : 34 ; if by a fourth, it will also be Munahh, Deut. 1:19. a. The parentheses of the table contain alternate accents. Thus, Merka is substituted for Darga and for Mahpakh (before Pashta in the clause of Zakeph Katon) if no more than one vowel intervenes between the Conjunctive and the king which it precedes, e. g. Gen. 1 : 22, Gen. 1 : 24, 26 ; Gen. 5:17, Deut. 1 : 2, 35. Munahh is also regularly substi- tuted for Kadhma, whenever the accent stands on the initial letter of the word. Gen. 25:8, Gen. 19:35; 1 Kin. 19:7, Deut. 1:28; Gen. 19:12 ; 4 50 ORTHOGRAPHY. § 38 Eccl. 5 : 7. G'rashayim takes the place of Geresh provided the accent is on the ultimate and it is not preceded by Kadhma either on the same or the previous v^^ord, Ex. 16 : 23, 36 : 3. When two accents are included in a parenthesis the meaning is that if an additional accent is required, these two will take the place of the one before the parenthesis. P'sik has no separate place in the consecution, but is joined with the other accents to modify their power. It is constantly associated with the Disjunctive Shalsheleth to add to its strength, and occasionally with the different Conjunctives to reduce their strength, but without disturbing the order of their consecution, e.g. with Merka Ex. 16:5, Munahh Gen. 46:2, Mahpakh Ex. 30 : 34, Kadhma Lev. 11:32, Darga Gen. 42:13, T'lisha K'tanna 1 Sam. 12:3. 2. Train of Atlmalih. — If Athnahh be preceded by a Conjunctive, it will be Munahh, Gen. 1:1; if by a Disjunc- tive in its own section, it will be Tiphhha, Gen. 1:1. The accents which precede Tiphhha have already been mentioned in explaining the train of Silluk. 3. Train of B'yholta. — The first Conjunctive before S'gholta will be Munahh, Gen. 3:3; if there be two, the second will be Munahh, Lev. 8 : 31, or Merka, Gen. 3 : 14. The first Disjunctive in its section will be Zarka, Gen. 1 : 28; and if this be preceded by one Conjunctive, it will be Mu- nahh, Gen. 1 : 7, or Merka, 1 Chron. 5:18; if by two, the second will be Kadhma, Gen. 30 : 16, 31 : 32 ; if by three, the second will be Munahh and the third Kadhma, Lev. 4 : 35. The next Disjunctive before Zarka will be Geresh, Gen. 24 : 7, or G'rashayim, Ex. 39 : 3. The accents pre- ceding these have been explained in 1. 4. Train of Zalceph Katon. — The first Conjunctive before Zakeph Katon will be Munahh, Gen. 1 : 2-, the second like- wise Munahh, Gen. 27 : 45. The first Disjunctive will be Pashta, Gen. 1:2; or, if the proper place of the accent be the first letter of the word, Y'thibh, Gen. 1 :11, 2:11. The first Conjunctive before Pashta will be Mahpakh, Gen. 1:9, or Merka, Gen. 1:2; the second, Kadhma, Gen. 39 : 19, or Munahh, Gen. 1:12; the third will be T'lisha K'tanna, Ezr. 3:11. The Disjunctive before Pashta will be § 38 CONSECUTION OF ACCENTS IN PROSE. 51 Geresh, Gen. 1 : 24, or G'rashayiin, Gen. 1:11; the further consecution is explained in 1. a. In some instances Pashta is found not in the train of Zakeph Katon, but seeming to govern an independent section, e.g. Ex. 29:20, Deut. 9:6, Josh. 10:11, 2 Sam. 14 : 7, 2 Chron. 18:23. 5. Zakeph Gadhol is mostly used instead of Zakeph Katon when no other accent precedes it in its own section, whether upon the same word or one before it : "I'm ^x Gen. 9 : 4 (in some editions), in which it is preceded by Munahh, is exceptional. 6. Trai?i of JR'd/ii". — The first Conjunctive before R'bhia* will be Munahh, Gen. 1:9; the second, Munahh commonly with P'sik, Gen. 2:5, or Darga, Gen. 6:15; the third, Munahh with P'sik, Gen. 7:23, 31:29, or Merka, Ex. 14 : 10. The Disjunctive before R'bhi=^ will be Geresh, Ex. 16 : 3, or G'rashayim, Deut. 1:11, which are preceded as in 1. 7. Train of Fazer. — Pazer may be preceded by one Munahh, 1 Sam. 14 : 34, by two, Ezek. 9 : 2, by three, 1 Sam. 14 : 34, or by four, Isa. 6Q : 20. 8. Trai?i of Tlisha 6'/^V/^o/«.— T'lisha Gh'dhola is the weakest of the Disjunctives which are ever set to rule inde- pendent sections. Its weakness is in fact such, that it is sometimes drawn into the section of a stronger Disjunctive ; thus, in Gen. 1 : 12, Lev. 4 : 7, 1 Sam. 17 : 51, Isa. 9 : 5, Neh. 5 : 18, it takes the place of T'lisha K'tanna among the antecedents of Pashta, standing between it and Geresh or G'rashayim ; in Gen. 13 : 1, 21 : 14, Deut. 26 : 12, it stands similarly between T'bhir and Geresh or G'rashayim. And in many cases, perhaps in most, when it rules a section of its own, this is a mere subsection, not so much a division of one of the principal clauses as a fragment broken off from one of the larger sections at a point where T'lisha K'tanna would have stood had the connection been sufficiently close 52 ORTlIOGRAniY. § 39 to require a Conjunctive, c. g. Gen. 19 : 3, 1 Kin. 20 : 28. That tliis is not always so nppcars, lioAvever, from examples like 2 Sam. 14 : 32, Gen. 7 : 7, Isa. 66 : 19, Jer. 39 : 5, and particularly Gen. 31 : 52, where nn^TDj)!;! corresponds to the preceding- -^px-QSV T'lisha Gh'dhola may be preceded by one Mmiahh, Gen. 27 : 4G, by two, Josh. 2:1, by three, or by four, 1 Kin. 2:5. 9. Shalsheleth occurs but seven times, viz.. Gen. 19:10, 24 : 12, 39 : 8, Lev. 8 : 23, Isa. 13:8, Am. 1 : 2, Ezr. 5:12, and in every histance stands npon the initial \vord of the verse, and is accompanied by P'sik. It has consequently no antecedents. 10. Karne Phara is only used sixteen times. Its section never contains less than three Avords : its immediate prede- cessor is always Yerahh ben Yomo, to Avliich may be added one Munahh, Num. 35:5, Neh. 5:13, 13:5, 2 Chron. 24 : 5 ; two, 2 Kin. 10 : 5, Jer. 38 : 25, Est. 7 : 9, Neh. 1 : G, 2 Chron. 35:7; three. Josh. 19:51, 2 Sam. 4:2, Jer. ,13 : 13 ; four, 1 Chron. 28 : 1 ; or five, Ezek. 48:21. §39. 1. The complete trains of the several accents con- tain one Disjunctive from each of the inferior orders, dis- posed in due succession of rank, with one Conjunctive immediately preceding the first class of Disjunctives, two Conjunctives preceding the second class, three the third class, four or more the fourth class. These trains are adapted to sections of difterent length and character by omitting such of the Conjunctives, and more rarely by repeating such of the Disjunctives, as the mutual relations of the words may seem to require, and breaking off the series as soon as every w^ord in the section is supplied. Thus, while the general order of consecution is fixed and invariable, there is the utmost liberty and variety in particular cases. a. In a very few instances the Conjunctives go beyond the number here assigned. Thus, Athnahh is preceded by two Munahhs in Ex, 3: 4, § 39 CONSECUTION OF ACCENTS IN PROSE. 53 and. according to some editions, in Isa. 48: 11. T'bhir ifi preceded by (bur Conjunctives, Josh. 10: 11, 2 Chron. 22: 11, Isa. tJG:20; Pashta by four, Ex. 5 : 8. 2 Kin. 5:1, and even by five. Joeli. 10 : 51. 2. If a section consists of but a single word, this will re- ceive the appropriate Disjunctive, the entire antecedent series of the table being then omitted as unnecessary; thus, Siiliik :nb;'l Gen. 5:5; Athnahh '\ic^^'\ Gen. 24:34; Zakcpli Katon nhsn Isa. 1 : 30 ; R'bhi'^ n-^isni Gen. 7:19; Pazcr ni3si^] Gen. 22 : 2 ; T'lLsha Gh'dhola p5 Gen. 19 : 8. This, as has been already said, is the regular length of the sections of Zakeph Gadhol and Shal.sheleth ; but those of S'gholta are never composed of less than two Avords, and those of Kame Phara never of less than three. 3. In sections of greater length there is a disposition towards a regular alternation of Disjunctives and Conjunc- tives upon successive words, e. g. Gen. 23 : 11 s , , ^ , , , , Gen. 24 : 7 _, \ ' '^ , and consequently thou^^rh two or more Conjunctives may be allowed before a particular Disjunctive, only the first of these is in the majority of ca.ses employed. The actual relations of words may, however, so interfere with this regularity as on the one hand to cause the intervening Conjunctives to be dropped entirely, e. g. Gen. 1:22 ...,,, 1 Chron. 15 : 18 ' ' ',/'', or, on the other, to introduce as many Conjunctives as the table will admit, e. g. Gen. 3:14 '\, , ",^ ''^\ But if either of the three primary sec- tions consist of but two words, the first must have a Dis- junctive accent, however close its relation may be to the second, e. g. : n-^l T^^-i Gen. 9:20, S2\r? •'7jr?=^ Gen. 3 : 5, bsc^ n-j Gen. 19:4. a. In Gen. 24 : 15. where, however, editions difTer, Silluii is in a section of two words immediately preceded by Merka. 6. Sometimes an excluded term of the series will take the place of the secondary accent Methegh, §44. Tiphhha is thus five times written upon the same word with Silluk. e.g. S'urn. 15:21. and eleven times v/itb 54 ORTHOGRAPHY. ^ 40 Athnahh, e. g. Num. 28 : 26. Munahh, Gen. 21 : 17, for which Kadhma is sometimes substituted, Gen. IS: 21, ol'len stands upon the same word with Zakeph Katon. Kadhma is also joined in this manner with Munahh, Lev. 10:12, Merka, Judg. 21:21, Neh. 12:44, Mahpakh, Lev. 25:46, and Geresh, Ex. 16 : 15, 21 : 22, 35. Mahpakh with Munahh, Lam. 4 : 9. 4. Occasionally a subordinate Disjunctive or its alternate is repeated in the same section with, or without its ante- cedents. Thus, T'bhir, Deut. 26:2 ,_ , ._ '^^ , '\ so Deut. 30 : 20, 1 Sam. 20 : 21, 2 Kin. 17 : 36. Zarka, 2 Kin. 1:16 ■//""/, so ver. 6, Gen. 42:21, Jer. 21:4, Neh. 2:12. Pashta, Gen. 24:14, 42, 48, 65; 1 Kin. 20:9. Pashta, Pashta and Y'thibh, 2 Kin. 10:30, Ezr. 7:25. Geresh and G'rashayim, Gen. 28 : 9. a. There is a double accentuation of part of Gen. 35:22, and of the entire decalogue, both in Ex. 20: 2-17, and Deut. 5 : 6-21, which involves a double vocalization in certain words, e.g. :^ba Ex. 20:3, i.e. either :13a or *';s. Single Avords also occur with alternative accents, e.g. with G'rashayim or Geresh and T'lisha Gh'dhola nV Gen. 5:29. ^ii-if? Lev. 10 : 4, ^D'ij' 2 Kin. 17 : 13, n^xb? Ezek. 48 : 10, nxY Zeph. 2:15.' Poetic Consecution. §40. 1. The principle of the consecution is the same in the poetic as in the prosaic accents, although there is consid- erable diversity in the details. There is a like division of verses into clauses and sections ruled by a Disjunctive at the end, which imposes upon them its own special train of accents. The sections are fewer, however, and the trains shorter than in prose, on account of the greater brevity of the sentences in poetry for the most part. But this reduc- tion is more than compensated by the new complexity arising from the latitude allowed in the choice of Conjunctives, which it seems impossible to reduce to fixed rules, and is probably to be referred to their use as musical notes for the ^ 40 POETIC CONSECUTION. 55 cantillation of the synagogue. It should be added, that the embarrassment arising from this inherent complexity of the subject is seriously aggravated by the numerous discrepancies in the different editions of the Bible, by which the true ac- centuation in the three poetical books is often involved in doubt and uncertainty. a. In atlditioii to availing himself of the researches of others, particu- larly of Nordheimer and Ewald in their discussions of this subject, the author has examined verse by verse the entire book of Proverbs and the first division of the Psalms (Ps. 1-41), as well as other selected Psalms and portions of Job. As the result, he confesses himself quite unable to disentangle the mystery; and as the only contribution he can make towards its solution he has concluded to present in detail, and in as con- venient a form as possible, the facts observed, hoping that some future ex- ploration may discover the principle of order, if any such principle there be, in this apparently inextricable confusion. 3. Verses may consist of one, two, or three clauses, dis- tinguished by the three Disjunctives of the first class. If the verse contain but one clause, Silluk will be written upon the last word, Ps. 4:1; if it contain two clauses, the divi- sion will be made by Athnahh, Ps. 1:4, or by Merka- Mahpakh, Ps. I : 2. 3 : 3, upon the last word of the first clause ; if it contain three, the last word of the first will have Merka-Mahpakh, the last word of the second Athnahh, and the last word of the third Silluk, Ps. 1 : 1. Clauses may consist of a single section when no subdivision of them is necessary ; or they may consist of two or more sections, when the subdivision is eff'ected by U'bhi'' or Pazer, e. g. : ■ Ps. 18: 51, :, . ' Prov. 1 : 10, '. ' Ps. 41:7, • / Ps. 7:'6,, : ■ .."Ps. 17:14. 56 OUTHOGRAl'llY. §41 §41- Tlie order of the accents in the various sections is exhibited in the following table : Principal Sections. > t-H o m izi -\,0\. or -'s : ■ \' A 1 : 1 1 : ! • :.0 ^ in. ■■m\ "r(:):^;\)] J ■n .., »-Xy,Oi. 'rSh. Subordinate Sections. • '\MM\ V ±H'r" Explanation of the Table. a. Train of Silluk. — If Silluk is preceded by a single Conjunctive, it will be Munahh, Prov. 1 : 4, or Merka either alone as Prov. 1:2, or com- pounded with Zarka, Ps. 10 : 5, and P'sik, Ps. 10 : 3. If it be preceded by two Conjunctives, they will be ^ Ps. 5:5, ^ Ps. 10 : 6, Prov. 12 : 1 (in some editions), ^ ^ Prov, 25 : 26, "^ ^^ Ps, 18 : 7, ■* ■* Ps. 36 : 1, or ■" ' Prov. ^41 POETIC CONSECUTION. 57 8: 13. Ifit be preceded by three Conjunctives, they will be ^ , ^ Ps. 24:6, 'Ps. 10:2ror 'Ps.7:6). Prov. 26:25 (or ^ "Ps. 2S:8or Prov. 29 : 13), , . ^ Ps. 4:8,^^ ' Prov. 3 : 27, '' "" . If it be preceded by four Conjunctives, they will be ^ ^ , \ Ps. 89:2, ^ ^ / ^' Ps. 32:5, or ■* ■* *^ Ps. 3 : 3 (in some editions '' '' * J. If it be preceded by five Con- junctives, they will be ^ ^ \ ^ Job 32:6, 37:12 (in this latter example some editions substitute a Makkeph for Merka). If Silluk be preceded by a Disjunctive in its own section, it will be R'bhi''-Geresh, Ps. 1:1, 5:3, 10. R'bhi^-Geresh may be preceded by one Conjunctive,, Ps. 5: 4 ; by two,, , Ps. 8:2, or, ^ Ps. 31:10, 19; by three, or Ps. 73:4. There are occasional deviations from the Conjunctives of the table; thus, R'bhi^-Geresh is in Ps. 34:8 preceded by ^ " ^. In some of these cases, however, editions differ in their notation of the accents. Thus, in Ps. 5:7 some editions have ^ " before Silluk, others , "; in Ps. 18:36, Prov. 30: 17, some have , , others , ; in Ps. 20:2 some have ^ ^ , others . ; in Prov. 24 : 8 some have , , , others , , the two words being joined by Makkeph. So, again, some editions have in Ps. 9: 11 ^ before R'bhi"- Geresh, in Ps. 18:44 , in Prov. 27:19 . in Prov. 21:17 ; while other editions do not depart in these passages from tlie order given in the table. Similar discrepancies exist in the other sections likewise. 6. Train of Athnahh. — Athnahh may be preceded by one Conjunctive. ^ Ps. 5:8 (or ^^ Prov. 8:30, 34),, Ps. 5 : 3 (or ^, Ps. 35:21, , " Ps. 69:2), .Prov. 23:3,^ Ps. 14:3, Prov. 6:3 (or ^ ^ Prov. 16:10); by two, ^ ^ Ps. 6:8 (or^ ^^ Ps. 7:17), ^ ^ Prov. 28:25, Ps. 5:2 (in some editions the latter example has ^ _, ^), _, ,^ Ps. 14:5, , ^ Prov. 11:12, 14 •■21,, / Ps. 37 : 1. , ^ Prov. 8 : 21, ._ ^ Ps. 25 : 16 ; by three, ^ ^ , Prov. 24 : 21, ^ ^ ^ Ps. 6:6 (or^ ^ . " Ps. 9:10, or ^ ^ _^ Ps. 16:10), , ^ ' Ps. 10:17, ^ ' '/ Prov. 8:13, __^ ,, Ps. 18:50, ,, ._ _, Ps. 10:13, ' 'fprov. 6:27,, " Vps.72:3; by four, , ^ ^ Prov. 3: 12, ^ ^ ^ ^ Prov. 24: 16, y^ ^^ Ps. 34:7, ^ ^ \ Ps. 32:2 (in some editions), "■ ^ ^^ Prov. 1 : 19, ," '"'^Ps. 65:2. If Athnahh be preceded by a Disjunctive in its own section, it will be Tiphhha initial, Ps. 1 : 6, 26 : 4. Tiphhha initial may be preceded by one Conjunctive,^ Ps. 5:6; by two, ^ _^ Ps. 9:19(or " Ps. 14:1 or Ps. 16:9),^ 'Ps. 32:11,^ ^ Ps. 35:14,15,^ / Prov' 25 : 20 ; by three, ^'^ , ^ Ps. 23:6,, , ,' Ps. 27:1,^ / ^ Ps. 12:5 (or ^ / / Prov. 27:14),\';"^ Ps. 9:14. c. Train of Merka-Mahpakh.—Mevlia-Mahpakh maybe preceded by one Conjunctive, which is almost always Yerahh ben Yomo, Ps, 1:1 58 OBTHOGRAPHY. §42 though occasionally it is, in some editions at least, Merka, Ps.l5:5, 35: 10, or Mahpakh, Ps. 24:8, 31:10. If it be preceded by a Disjunctive in ita own section, Zarka will be employed, Ps. 1 : 1, Prov. 1:11. Zarka may be preceded by one Conjunctive, Ps. 12:7 (or ^ Prov. 1 : 22), ^ Ps. 6 : 3, ^ ^ Ps. 12 : 3, ^ * Ps. 31 : 12 ; by two, ^ ^ ' Prov. 30 : 15 (in some editions '), Ps. 24 : 10 (or Ps. 13:6), Ps. 21 : 10, ' Ps. 27 : 2 (or / ^ " Ps. 35 : 26), _ ' '' Ps. 7 ; 10 ; by three, ^ / ^ Ps. 29 : 9, ' Ps. 31:14, ' " Ps. 10: 14; or by four, ' Ps. 40 : 6. d. Train of R^bhi". — R'bhi" may be preceded by one Conjunctive, Ps. 5 : 1, ^ Ps, 8 : 2 (or ^ " Ps. 23 : 4, or _ ^ Ps. 6 : 7),^ Prov. 28 : 22, ^ Ps. 22 : 25, 'Ps. 11:2 (or ' Ps. 5 : 11) ; by two, Prov. 8 : 33. Ps. 28 : 7 (or Ps. 18 : 3), ^ / Ps. 9 : 7, ^ " ^ ^ Ps. 1 1 : 4, ^ / Ps. 26 : 1, ""^ Ps. 27 : 6 (or ^ ^^ Ps. 5:9). 'Prov. 6:22, ' Ps. 18:1 (or ' ~ Ps. 7 : 7, or ' " Ps. J i \ 'l< \|< 'IK 39 : 5), / ' Job 16:10; or by three, ^ / ^ Ps. 40 : 7, ^ , \ Ps. 41 : 7 (or " ' Ps. 39:6, or " ' " Ps. 3 : 8, or ' " Ps. 41:14), ' ' Ps. al< '^oi^"DT^ , ns?'n^"']'^K . 60 ORTHOGRAPHY. § 44 a. Tsere remains before Makkeph in 13, "^3, "IV, yiy^ ; it sometimes re- mains and is sometimes shortened in B'^, lad six, rs e. g. Gen. 16: 13 rnn^i'dU, but ver. 15 i33"CiIJ . It once remains according to some editions in TIN Job 41 : 26, a word which is three times written rx without Makkeph, Ps. 47:5, 60:2, Prov. 3:12. Comp. § 19. 2, a. b. Makkeph is occasionally found in the middle of a long word, which has been erroneously divided into two, e. g. ii^E"iia'^ Jer. 46 : 20, and perhaps tiipTipS Isa. 61 : 1. Sometimes words are thus divided without a Makkeph to unite the sundered parts, e.g. Q"';? ']? Lam. 4: 3, tfiinh "^na 2 Chron. 34: 6, and probably ^lin dliij Hos. 4: 18, ni-iQ -iQnb Isa. 2 :'20'. (See Dr. Alexander's Commentary on this passage.) The last two ex- amples are plainly intended by the punctuators to be read as separate words. This might likewise be done in the preceding examples if they were pointed CliJ "is and ch"'n2 "in3 . Methegh. § 44. Methegh (.^ri'a bridle), a small perpendicular stroke under the initial letter of the syllable to which it belongs, is a secondary accent denoting a stress of voice inferior to the main accent. As this latter always has its place in Hebrew either upon the ultimate or the penult, distinctness was promoted and monotony relieved, especially in long words, by giving prominence to one or more of the antece- dent syllables. There is a natural tendency to heighten the force of the accent by passing lightly over the immediately preceding syllable, this diminished force creating in its turn a new stress upon that next beyond it, and so on in alternate elevations and depressions to the beginning of the word. Agreeably to the principle just stated, Methegh regularly stands in polysjdlables upon the second syllable before the accent, and again upon the fourth if the w^ord have so many, e. g. ^ii^\}, vl5s?;2, ''?!J?^^'i?, Dn^p,a, Drpni'iiip^^, niiiDipri^i . And so upon two or more words connected by Makkeph, which are pronounced as one, e. g. i^J'rrx'i';' Gen. 22 : 8, nn^-aii-^s 1 Sam. 21 : 7. a. Sometimes, however, particularly when the nature of the syllables requires it, §32. 1, Methegh takes the place of the principal accent before §45 METHEGH. 61 Makkeph irrespectiv^e of the position of the accent upon the following word, i?-1"^xaJn Num. 21:35, liran-Ti^.^ Num. 21: 33, ^=i3"r|>.^. Jer. 34:1, C!in-nb-b3i Gen. 30: 32, nrsibn/l Sam. 21 : 12, '^B-'S Ex. lb": 5. b. It is to be observed that the position of Methegh is determined by that of the tone-syllable, not by that of the accentual sign when these are not coincident, as frequently happens with prepositives and postpositive?, e.g. '■'ni"'"^] Dent. 4 : 2G. sirnini'n Josh. 22:27, where the tone fells on the penult, c^p^irr; Jer. 26:21, where the tone is upon the ultimate. §45. The secondary accent is liable to be shifted from its normal position for the following reasons, viz. : 1. If the syllable which should receive it is mixed, it may be given in preference to an antecedent simple syllable, e. g. nn^nTTX) 2 Sam. 22 : 24, ^?nnna.i Job 1 : 7, niinnpnia Ezek. 42 : 5, TiJ'^iitn-bKilJ. Gen. 43:7; or if none such precede, it may be omitted altogether, e. g. DDS'a^] Jer. 33 : 24, ''?«?'?r^^ 1 Kin- 31 : 1, ni:£^n-b5-ni5 Dent. G : 25. 2. It is always given to simple syllables when followed by a vocal Sh'va, wdiether simple or compound, or a vowel which has arisen from Sh'va, the slight pronunciation proper to the Sh'va or its derivative giving new prominence to the preceding vowel, ^:ynik\ ^^.T}"^. , ^^.^}., riinj^irn Gen. 30 : 38, ^'I'lin'^ ; sometimes to intermediate syllables, § 20. 2, e. g. ^inp Isa. 9:17, 10:34, tjW Obad. ver. 11, particularly after He interrogative or when Daghesh-forte has been omitted as after the article, Vav conversive, and the prepo- sition p, e. ff. bti)2T\ , n^j-iJnn , ^^bn , ninis , D^y^nsisn , nk'j-ch ''n^ll ; rarely and only as an exception to a mixed syllable standing in the first place before the principal accent, e. g. «TC^n Gen. 1:11, n^mn Ex. 12 : 7, Zech. 14 : 2. a. It hence appears how Methegh comes to be of use in distinguishing the doubtful vowels, § 19, and to what extent it can be relied upon for this end. As it invariably accompanies the vowel of a simple syllable when followed by vocal Sh'va, it must always be found with a, I, and u preced- ing Sh'va. inasmuch as this will necessarily be vocal. Initial ^ u, the un- emphatic conjunction, is an exception, Avith which it is commonly not written, e. g. ni;^:!i Gen. 6: 19, iixbbii Gen. 31 : 4, though it is sometimes, 63 ORTHOGRAPHY. § 45 e. g. i'-^^nn^l Gen. 1 : 18, nh.t^ Judg, 5 : 12. The absence of Methegh; except in the case just mentioned, is consequently conclusive evidence of tlie shortness of tiie vowel. As, however, short v6we!s in intermediate syllables, and in a few rare Instances even in mixed syllables, may receive Methegh, the presence of this sign does not of itself determine the vowel to be long; the ultimate decision must in this case depend ou other con- siderations. 3. When by the operation of the preceding rule Me- thegh comes to stand in the first place before the accent, another Methegh is nevertheless occasionally found in the second place, the two thus standing in immediate succession, e. g. nin^a Gen. 32 : 22, ^S?.'!^ Gen. 45 : 25 ; and even three occur upon successive syllables, e.g. ?]'i)a?'D'a^ Isa. 22 :19. But commonly where there is more than one Methegh, their position relatively to each other is governed by the same rules as the position of Methegh generally with relation to the principal accent, e. g. 'nnsjrdjib , jHhirnn , 'nf;i^;?ns5'\ , 4. Methegh is sometimes written under a letter with ' Sh'va, e. g. x?""^^ Job 1:11,2:5, iBS-i5>^ Job 19 : G, nj:n:3 Ps. 2 : 3, nicn^as Jer. 49 : 18, '^i^.'a^ Ruth 1:11. a. A Methegh so situated is called Gaya (J^'^^sra bellowing) by Jewish grammarians, and, according to Elias Levita, it occurs eighty-four times, the number yielded by its name arithmetically reckoned. Methegh upon a short vowel before a compound Sh'va was called Ma"rikh (Tf~N]? pro- longing), with a short Hhirik it was called Hhiruk (p^iTJ gnashing). 5. The place of Methegh is frequently supplied by an accent chosen agreeably to the laws of consecution, § 39. 3. 6., e. g. aipnss: Isa. 66 : 13, oH^n-^sb Deut. 12 : 31, Dn?Di5) Zech. 7:14, xni^-b^l Num. 10:23, ^':r|;?^^ Josh. 22 : 12. a. The want of consistency or of uniformity, which may be occasion- ally observed, in regard to the insertion or omission of Methegh, e. g. nnnsa Cant. 1 : 7. ninsiy Cant. 3:1; tna Cant. 6 : 5. and Lam. 4 : 9 ; § 46 k'ri and k'thibh. 68 nidns Num. 31:12, ri^iS Josh. 4:13, and llie discrepancies between different manuscripts and editions, e. g. «^3^i< or nsbx Gen. 45 : 28, rinTai-^ixb or nn^f^isiy Ps. 81 : 3, if not arising in the first instance from clerical errors, are probably to be attributed to the inferior importance of the sign itself, whose place might be presumed to be sufiiciently determined even if not written. K'ri and K'thibh. §46. Various notes extracted from the Masora ('^'^"ic'a traditioii), a collection of remarks upon the text, are found in the margin of the Hebrew Bible, which are explained in the glossary at the end of most editions. The most im- portant of these are the various readings known as the K'ri ("'"ij? read), and K'thibh (^'^r^i' icritten). If in any instance traditional usage sanctioned a reading different from that which was written in the text or the K'thibh, the punctuators did not venture to alter the text itself for the sake of making the correction ; they went no further than to connect with the letters of the text the vowels of the w^ord to be substi- tuted for it in reading or the K'ri, with a reference to the margin where the letters of the substitute might be found. Thus, with the Avord I'^'Q!^''!! Josh. G : 7 is connected the marginal note "^ip "n^xin. The vowels here attached to the K'thibh belong not to it but to the unpointed word in the margin, which is accordingly "i^x^i . The proper vowels for the pronunciation of the K'thibh are not written, but must be supphed from a knowledge of the form indicated by the letters, Avhich in this case is ^n'Ci^'^'i . Again, in ver. 9, M!yT\ in the text refers to p '•ypn in the margin ; the K'ri is here ■^^ph , and the K'thibh, whose vowels are left to be deter- mined by the reader, ^5pjn. Jer. 42:6 has ^ibz^ where the marginal note is ^"p i:n:s< ; the K'ri is accordingly ^^H'?'!' , and the K'thibh ''ii? . In order to indicate that a given word was to be omitted in reading, it was left unpointed, and the 64 ORTHOGRAPHY. § 47 note "^"ip Sib*! n^nD , iDritten but not read, placed in the margin, e.g. Ti)'/:n Ezek. 48:1G, «3 2 Kin. 5: IS, inT Jer. 51 : 3. If, on the other hand, a word was to be suppHed, its vowels were inserted in the text and its letters placed in the margin, with the note S^rD sbi iip, read hut not toritten, e. g. Judg. 20:13 in the text ^ and in the marmn '':3, to be read :.:a ; so Jer. 31 : 38 n\s:a . In 1 Kin. 21 : 8 the first letter of D''"i£9^ is l^ft unpointed as superfluous, and in Job 2 : 7 "?, is explained by the margin to stand for 'i?^ : so Jer. 18 : 23 ^'Ti,.'] for vm . a. The number of these marginal readings differs in different editions. Elias Levita states that there are 848. Others have computed them to be 1,000; others still, 1,200. §47. Sometimes a different reading from that of the text is suggested by the points alone without a marginal note being added in explanation, as when a particular word or orthography is regularly substituted for another of frequent occurrence. These cases are presumed to be so familiar to the reader as to require no other index of their existence than the presence of the appropriate vowels. Thus, the divine name Mini , which the Jews had a superstitious dread of pronouncing, was and still is read by them as if it were ■'ns Lord, whose points it accordingly receives, T\-\r^^ , unless these two names stand in immediate connection, when, to avoid repetition, it is read D'^nbs and pointed Mnn.^ Gen. 15:2, Hab. 3:19. The antiquity of this superstition is attested by the Kvpwi of the Septuagint, -followed in the English as well as in other modern versions by the rendering Lord. The true sound of the name never having been noted, is noAV lost ; the only clue that is left being its ety- mology and the form which it assumes in composition, §62. 1, from which the conclusion has been variously drawn that it was nin^, ""On., or ^^"^T!*^.. The common pronunciation Jehovah is manifestly founded upon the error of combining §48 k'ri and k'tiiibii. 65 the consonants of this word with the vowels of another and an entirely different one. There is, however, especially as it is uncertain whether Yahve or Yah"va, or either of these, was its original sound, no good reason for abandoning the pro- nunciation familiar to the Christian world and hallowed by the association of constant usage for the sake of adopting another which is, or is supposed to be, phonetically more exact, any more than we need be guilty of the pedantry of preferring YesJi,a>/aIm to Isaiah because it approaches more nearly to the original pronunciation of the prophet's name. Other standing K'ris, unnoted in the margin, are s^'in , the form of the pronoun of the third person feminine which is used throughout the Pentateuch ; this is designed to be read X''n , though the sound indicated by the letters is in all proba- bility N^n . So ^bim-j read ^iis:' , and ni%ry^_ read Q:'^Tri^^ . §48. In the absence of definite information respecting the origin and sources of these various readings, it is difficult to determine with absolute precision the weight to which they are respectively entitled. The current opinion of the ablest Hebraists, based upon a carefid scrutiny of their in- ternal character and the relation which ordinarily appears to subsist between them, is that Avhile the K'ri may perhaps, in a few cases, correct errors in the K'thibh, and so restore the original reading, it is in the great majority of instances an explanatory gloss rather than an emendation. With the rare exceptions already suggested, the K'thibh is esteemed the true reading, the object of the K'ri being to remove ortho- graphical anomalies, secure grammatical uniformity, substi- tute usual for unusual, prevailing for obsolete words and forms, and occasionally to introduce euphemistic expressions. While the K'ri is probably not to be esteemed the original reading, therefore, it deserves attention as the grammatical or exegetical comment of a steadfast tradition. 66 ORTHOGRAPHY. ^ 49 Accuracy of the Points. §49. 1. All the Masoretic additions to the text designed to facilitate its reading have now been considered. The cor- rectness of the pronunciation, which they yield, is vouched for not only by the esteem in which they are universally held by the Jews, but by the scrupulous minuteness of the system, its consistency with itself and with the vowel-letters of the text, its affinity with and yet independence of the vocalization of the kindred languages the Arabic and Syriac, and the veneration for the already established text which evidently characterized its authors, since they did not venture to change the text even in the slightest particular. 2. The only additional information which has come down to us respecting the true sound of Hebrew words, is furnished by the mode of writing proper names in the Septuagint version, and the few Hebrew words preserved by ancient authors, particularly Origen and Jerome. These have been subjected to an elaborate comparison with the Masoretic punctuation, and the result has been to establish their sub- stantial agreement in the main, with, however, not a few remarkable points of divergence. In relation to this subject it should be observed, that the Hebrew pronunciation of the Seventy is inferred entirely from their mode of spelling proper names, not from words in living use in the language. The chances of inaccuracy, on the part of the translators, are here peculiarly great. Many names were -not familiar and were of rare occurrence ; and as no system of vowel notation then existed, they were left entirely to their independent knowdedge of the sound of each individual word. These words were written by them in a foreign alphabet, whose sounds did not coincide precisely Avith those of the Hebrew,, and in which the proper equivalents varied somewhat accord- ing to their combinations. The true sound w^as also de- ^49 ACCURACY OF THE POINTS. 07 parted from sometimes because the laws of Greek euphony forbade its exact reproduction. The neghgence with which they are chargeable elsewhere was also probably aggravated here, and in fact there are many instances in which they not merely deviate from the vowels but transpose or change the letters. Leaving out of view, therefore, such incidental dis- crepancies as are to be accounted for in the ways now sug- gested, a thorough and extended examination of the subject reveals, with all the general agreement, a number of regular and systematic deviations. a. These are thus stated by Ewald, Lehrbuch, p. 1 16. (1.) An e or l de- rived from a is written d, as rrnn ©apa, nrbs BaXaa/A. "i"'"^? TafSaotv, CJ")^ Maptajj. : and on the other hand, a is sometimes written e, n:on"^bfij< 0At/3f/xa, f:p Keve^; rj Ted^, especially before n, as nnp Kope, nnt Zape. (2.) e is written for l and 6 for u, C'lns Xerratot. n':n^5 Teevva, ')"''7? TeSeouv, D':"!'-^? Mecrpatyx, rmii.__ Oxol,a&, n^-U' O^ta. (3.) for the diph- thongal e and o their constituents ai and an are substituted, ""^l^ Katvav, ira Na/^aw. (4.) The vowel letters are softened into their homogeneous vowels X'^P.*^ oviKpa, "^S"]"^!] ovihaf^rjp. (5.) Vocal Sh'va is written as a full vowel, commonly a, or if an follow, o, rixn:: 2a/?aw^, bs^""i Payouv^A, C^^iS XepovfSifJi, cho SoSojtta ; the final vowel of Segholates is also written 6 if o precedes, T)^b MoXo;)(, I'c.i" yop.op. 3. The regularity of these deviations seems to be best accounted for by the assumption that the pronunciation represented in the Septuagint is that which prevailed among the Jews in Egypt, Avhich would naturally be less pure than that of Palestine represented in the vowel points, and which, moreover, betrays in the particulars recited above a strong leaning to Aramaean forms and sounds. Accordingly the view now commonly entertained is that the vowel notation of the Masorites is correct, at least in all essential particulars, and that it is properly to be put at the basis of all investiga- tions into the phenomena of the language. 68 ORTHOGRAPHY. § 50 Orthographic Changes. § 5 0. The signs thus far described represent all the sounds of the Hebrew language. Its stock of words is formed by combining these in various significant ways. The laws of such combinations, and especially the mutations to which they are subject, or which they occasion, next demand attention. When a particular idea has been attached to a certain combination of sounds, its different modifications may naturally be expressed by slightly varying those sounds. This may take place, 1. By the substitution of one letter for another of like character, and for the most part of the same organ, e. g. : n;^rt to be. exist, iTjri to live ; S'23 to pour forth, K^3 the same idea ap- plied to words, to prophesy; pJlS to encircle the neck with an ornament, pDH to strangle, p;N applied to sounds uttered in strangulation, to groan; ^:0 to go about as a spy, 'b'y-\ to go about as a merchant; G33 to collect, D^na treasures; ^'"'25 a cup, i'liis or ^"zip a helmet (of similar shape); tp tender, delicate, pT thin ; "jpti to make straight, 'SPi to straighten the beam of the balance, to weigh ; "1132 Jirst born. "^32 Jirst ripe, "ip3 the first portion of the day, the morning ; nbn to suspend, nb'n applied to a bucket, to let down; "its to cut. "iS;? to reap; -tij gold, :rh^ yellow; "(ox^ to con- ceal, "jS'sU and "(S^ to hide away as treasures, "£& to cover with boards ; I'nj to destroy by tearing down, cnj to destroy by uprooting ; ns:: to slay, nST to sacrifice; bDrj to bind, b3J to bound; n"]3 to break up, flte, tT^Q to break out, blossom, p^Q to break in pieces ; 3^p to cut off, ~:in to heto stone, a^in to cut wood, ; "inS to surround, -iZi'J to encircle the head with a crown; T\^^ to pour oid, "03 to pour in libation or in casting metals ; in^ to shine, "iHiJ to be pure; nnn to engrave, dnn to plough; iria to prove, iris to approve, choose ; nnt;: ?o cZn'?i/c, its causative HiTiirn ; "inn to break through, ipn to investigate ; arij to place, \is reflexive iS^rili. 2. By the transposition of letters, e. g. : "j^nS to deal violently, nrja to urge ; "i2p to cid with the sickle, reap, y^p^ to cut with the teelli, bite; ti'lij to blow, TliS? breath; 053 to collect, 0D3 riches, ni330^ storehouses. 3. By the addition of a letter : Thus, from the letters "i::, in which inheres the idea of compression, are formed "I* ^ to bind, "iVJ. to press together, ^:ii< to heap up, ^^^ to be § 51 ORTHOGRAPHIC CHANGES. G9 straitened, ^iJ3 to guard, besiege, ^'^'S to restrain, istn an enclosure ; from T5 are formed "in to cid, tna /o n«< off, TIJ ^o cztf /oose, go away, tn ?o s/iear. bta ?o plunder, nita /iei<;/i stone; T'^S ^o unfold, make distinct, VC^yi to spread out ; Q"i3 a vineyard,, b'C'iS a garden. § 51. Such literal cliaiiges as those just recited not only serve to express new shades of meaning, but even where the meaning remains precisely the same, they may represent diversities of other sorts. Thus, the distinction may be, 1. In point of currency or style : One form of the word being in more common and familiar use, the other more rare and savoring, perhaps, of the elevated or poetic style, e. g. : "i:jj to guard; "iiiS poetic; ri"i3 cypress, ni^a once in poetry; "iJO to shut, rarely "i?0; iT^:?p storm, ^1^^'^. rare and poetic; "?5 to cover, once tjsi^ : Tj?^ to he quenched, once '7\^"_] ; -^'P to abhor, once sxn ; bro to be foolish, once bos ; T\h']'J iniquity, once ti'^? . 2. Of antiquity : The pronunciation of a word or its form may undergo changes in the lapse of time. Of the few instances of this sort, which our imperfect data enable us to fix upon with some measure of confidence, the follow- ing may be taken as specimens, e. g. : To laugh in the Pentateucli pf^^, in other books (Judg. 16:25 ex- cepted) pnb ; to cry out in the Pentateuch P"^ , only once (Ex. 2 : 23) pSI which is the more irequent form in other books ; -"rs , nar? a. lamb, occur in the Pentateuch interchangeably with bas , nbas, which are the only forms found in other books ; a sceptre ^5'^ , but in the book of Esther li'^ano ; Damascus p'::isy\ , in Chronicles pb^s'^'n ; how 1 Chron. 13:12. Dan. 10 : 17 T^^f} , in earlier books Tpx . 3. Of Dialect : The same word may come to be pro- nounced differently by those who speak distinct though re- lated languages. Thus, the Aramsean dialects, the Chaldee and Syriac, in very many words regularly substitute i{ for the Hebrew final n , and the corresponding Unguals for the He- brew sibilants, 2 being sometimes still further weakened by the loss even of the lingual sound to that of the guttural y , e. s:. : 70 ORTHOGRAPHY. § 52 Heb. ni'Sn to wander, Chajd. N^-J , Syr. j.^^ ; Heb. rnj gold, Chald. -r,^, Syr. Isffi?; Heb. ^ri a rock, Chald. nrj , Syr. ]lLi ; Heb. U?d three, Chald. n^n , Syr. l:<.Z , Arab. e>^" J Heb. '}^-iN the carlh, Arab. ,^!f , Chald. "TX, Syr. jll?) . Other consonant changes: Heb. ',2 a son, Arab. ^Jl , Chald. ^3 , Syr. j-I: ; Heb. ba;? ^o A:///, Arab. JJci; Heb. bbp?';', Syr. '\a.f.aJ; Heb. NS3 a ^/t?'o»(?, Chald. "D-i^iS, Syr. }.aJC?aa. Arab. "Is/; Heb. r^ph^ a Jidd, Chald. Nb|?r! , Syr. jf-al , Eth. jTl^A. ." . 4. Of simple eiipliony : An alternate form of a word may be produced to facilitate its pronunciation or make its sound more pleasing, e. g. : "i?3J"ix , "i'"iX purple; tub, )'q':i to hate; ^'z^h , nr;i;: chamber, 'p5), '■!=>' Achan; -^IXl^^^l^J , -i5JX-;"]3!i33 Nebuchadnezzar; 5n'^, a.'jin Doeg ; cj^bx, n"'52^r-bi< alnmg or algiim trees ; nis'nb^o, ri"^ri?3 teeth. a. Mere varieties of orthography must not be mistaken for consonantal changes, e. g. xb occasionally for ib and vice versA, probably r^bsb for n^lb^C), and such permutations of gutturals as abound in the manuscripts of the Samaritans, who, making no distinction in the sounds of these letters, perpetually confounded them in writing, Gesen. Sam. Pent. p. 52. A like fiTulty pronunciation has been attributed to the Galileans, to which there is a probable allusion in Matt. 26:73. Buxtorf Lex. Chald. p. 434. § 52. The changes thus far described result in the pro- duction of distinct words, and belong to the domain of the lexicon rather than of the grammar. The lexicographer re- gards such words as cognate, and traces them back to their common source ; but, in the view of the grammarian, they are totally distinct. The mutations with which the latter concerns himself are such as take place in the direct deriva- tion and inflection of words. These are altogether euphonic, are more restricted in their character, and take place within far narrower limits, than those heretofore considered. When words are subjected to grammatical changes their sounds are brought into new connections, attended, it may be, with § 53 CONSONANT CHANGES. 71 a difficulty of utterance whicli demands some measure of relief, or they pass readily and naturally into other sounds, which are easier of pronunciation or more agreeable to the ear. The mutations thus induced are of three sorts, viz. : Consonant Changes, the Conversion of Consonants into Vowels, and Vowel Changes. These will require to be con- sidered separately. Consonant Changes. § 53. The first class of changes embraces those which affect the consonants. These mostly arise from the concur- rence of two consonants, creating a difficulty in the pronun- ciation or yielding a sound displeasing to the ear. This may take place either at the beginning or the close of a syllable. Syllables in Hebrew may, and often do, begin with two con- sonants, §18. 1; but the necessity of this is avoided in certain cases by the following expedients : 1. In the beginning of words the weak letter in is some- times prefixed with a short vowel, thus creating a new initial syllable to which the first consonant may be transferred. a. The only instances of this are afforded by the second and seventh conjugations of verbs, the Niphal and Hithpael, e.g. bapti = bib]:?:!! for Vi-j;:??; b-Jprri probably for ?bl?n §82. 5. h. In llS'nx Ezek. 14 : 3 X is prefixed instead of i"i . Prosthesis is more common in the domain of the lexicon, where S is always the letter used, e.g. ?i"iT , ISi^TX arm. ; bi^fn, biianx yesterday. A prefixed N. is even occasionally employed to soften the pronunciation Avithout the necessity stated above, e. g. nin:i3N , caaN:. , D'lis'i'iX, , ^■'fstx^. So in Chaldee Cix blood, Yi&h. Cn ; 'i^X garden.YL&h. "ja. In Arabic the concurrence of two consonants at the beginning of a word is regularly obviated by prefixing I . Comp. Greek x^ii, ^X^^'^- 2. The first of the concurrent consonants, if it has a comparatively feeble sound, is sometimes dropped. 72 ORTHOGRAPHY. § 53 a. This occurs regularlv in verbs whose first radical is "^ or : , and in nouns derived from such verbs, e.g. -^ ior r;^?'; , ny'n ibr i^y'7.'!, ?^3 for b^di , 'in for )h: . "'n Ezek. 2 : 10 for ^h: , brcn EzeU. 1 : 4 for b^'rnj , and perhaps ^i<3 Am. 8:8 for IN^? . X is thus dropped in ^;n: for ^irnrs, ui for "illix ; also in a few instances from the beginning of the second syllable of words, e.g. ijl^Xp^ Ezek. 28:16 for ^13NN'; ; "pTX Job 32: 11 for "("^'iNX ; D-^n^an Eccl. 4:14 for ci^.^Oxri; nirinn'2 Chron. 22:5 for Q-^rnxn ; r-iD52 Ezek. 20:37 for niON ;? ; nVsa 1 Kin. 5:25 with Daghesh-lorte conservative for nbi:x?3 ; "(in Prov. 8:17 for -rjXN; 'pip Prov. 17:4 for V'iX^.; Tib^'f! 1 Sam. "1:17 for 7|nbNi:3 . These examples likewise admit ol" a different explanation; X may give up its consonantal power, losing its sound in that of the pre- ceding vowel, agreeably to §57. 2 (2), after which it may readily be dropped altogether. J2 is occasionally dropped from the participles of the Pual or fourth conjugation, as n^sb for nf^^TD ; h in P.|D for nf?b ; n in nib Ex. 3:2 for ^^r?.'?; ^n^'^^ Ex. 7:22 for'cn'^-jnb Ex. 7:11; and perhaps 3 in nri^iD Gen. 49 : 11, which appears to be for nhsiDS . b. The rejection of a consonant from the beginning of a syllable, when not immediately followed by another consonant, is exceptional ; as "it Judg. 9:11 for 1^.;; ; nnn 2 Sam. 22:41 for nnn: ; in Ezek. 33:30 for nns?; nisBH Neh. 3:13 for niscxn ; '^nb'inn Judg. 9:9 for "n^tjnnn, and perhaps -id Jer. 42: 10, which seems to be ibr 3i"c:^. 3. The second consonant is sometimes dropped, if it is a letter of feeble somid. a. This is regularly the case with M of the article and of verbal pre- fixes, and '' as the final radical of verbs, e. g. n';'2b for n'^snb; bbp"^ for bbisn'i ; iiBj for ii'i'ba . It occurs besides in a few sporadic examples with these same letters, and more rarely still with N , 1 , and 35 , e. g. "iT for ItnT , ''i Ezek. 2 : 10 for bh^ Isa, 13 : 20 for bnii'] , ^2nTn;i 2 Sam. 22 :'40 for "'3'^i xni • ^3 Isa. 3 : 24 for "^'3, ''V for i^?, , C'^h'^ for Q^h^^; ; 13 as a particle of' entreaty, probably for ■'i-'a, np'l"3 Am. 8: 8 (K'thibh) for M^T^'? 5 ^? the name of a Baby- lonish deity for brs is a foreign contraction. The conjecture that 133 Mic. 1 : 10 is for i3>"3 in Accho is ingenious and favoured by the occurrence of n;a mi Gath in the parallel clause; but it is at variance with the points, which, upon this hypothesis, should be 123. b. In rare cases this rejection occurs even after a mixed syllable, whose final consonant is thus drawn forward, e. g. f\''!Zii. for nnrx , iihn Job 29:6 for tik^n, 3ann Ex. 2:4 for 3il^nn and'probably pex Ps" 139 : 8 with Daghesh-forte conservative for PrO'* • §54 CONSONANT CHANGES. 73 § 54. When the concurrence takes place at the close of a syllable, Avhether the second consonant belongs to the sanie syllable with the first as at the end of words, or to a differ- ent syllable as in the middle of words, the following changes may be produced. 1. An aspirate following another consonant loses its aspiration, §21 ; or if it be brought into juxtaposition with its like so as to form a doubled letter, the aspiration of both will be removed, §23. 2, unless the combination occurs at the end of a word, where the reduplication is not expressed, § 25. Thus, n)2 for nn^ , niann for Di2nrin , "jrinb for inn::b , but nn from nnn, trivia 1 Kin. 1 : 15 for nn'iffi)2 or f^nnir^, nin^. Ezek. 4 : 3 for t^^^ri-a , rhm Mai. 1 : 14 for innnir^ . 2. The first of two concurring consonants is in certain cases assimilated to the second, the doubling thus occasioned being expressed as in the case of letters originally alike by Daghesh-forte, except at the end of words, §25, where Daghesh disappears or is only virtually present, being re- sumed upon the addition of a fresh vowel or syllable. This is most frequently the case with the liquid 5 , rarely wdth ^ and ■> and only in particular w^ords ; so n of the Hithpael of verbs before "7 and "J , and in a few instances before sibi- lants and other letters, and 1 at the end of a few words before n . Thus, ^k"! for )ki^ , rn^a for n:pD^ ; n^:* for nj^)^ , nb Ezek. 27 : 23 for nibs Am. 6:2; ^^ for ^^b "ncs? ; ^ss^'!' for ^ii^^ro, ^^^''. for si2t:n\ ^ijn for ^itrn, DTsiiBn for D'biirnn , ^sasn for ^i<35nn , nssn for nssnn ; nb for ri'ib , nni? for n^ns? . a. So perhaps 3 in HE'D according to Gesenius for f^h'zxi and D^ for p~^. Compare Greek avyycv/j-; for o-uvyei/vys, TeVu/Xjaat for reVuTr/xat, and Eng. il-logical, ir-religioti, im-matnre formed by the negative prefix in. 3. A few isolated cases occur of the reverse process more common in Chaldee and Syriac, by which a doubled letter is resolved into two different consonants by the change of the 74 ORTHOGKAPIIY. § 55 first or the second member of the rediiphcation to a hqnid n or : , e. s^. '2^1?^ for bsD^a , pi?^^^ for piSTS^ , OiR^ Isa. 23 : 11 for ""'•iyri, ^h'p Job IS : 2 in the judgment of some for ^ssp cuds, though others make the 2 a radical, and give the word the sense of snares. The conjecture that ^5^r} Ps. 64 : 7, Lam. 3 : 22 is for ^^sri is unnecessary and unwar- ranted. 4. When n of the Hithpael of verbs would stand before a sibilant, it is transposed with D and 12 , and with 2 it is in addition changed to 12 . Thus, nnnc^ for nncn^ , i^ni^j^ for "listen;' , '^^.t^'iO'j for i^iwf^O , PT^^^ for p'^^fp? . a. Ill nrj^iir^n Jer. 49 : 3 the transposition does not take place in con- sequence of the number of similar letters which would thus be brought into proximity. In the cognate languages n is likewise transposed with t and changed to T. thus, Chald, "S^in for )h]r^.^; so, also, in Syriac and Arabic. The only example of a Hebrew verb whose first letter is T ap- pearing in this conjugation is !l3-;tn Isa. 1:16, where Ti is assimilated agreeably to 2. Compare with these transpositions the frequent Doric change of ^ (= 8cr) into crS, as crvpio-Soj for crvpt^w. §55. The occurrence of a consonant at the end of a word may, inasmuch as the succeeding word must necessarily begin with one, be reojarded as an additional case of the con- currence of consonants. As the contact is less close, how- ever, than when they meet in the same word, it is less fruitful of changes than in the cases already considered. 1. There are three instances in which it has been doubt- fully conjectured that a final ) has been assimilated to a fol- lowing initial 'Q; viz. nilb'ti-; Isa. 35:1 presumed to be for )im^; Di-^^s Num. 3 :49 for li^^s Ex. 21 : 30, Ps. 49 : 9 ; D^oGen. 28:12. a. Final consonants are in Sanskrit perpetually modified by the initial letter of the following word. But it is by no means clear that this is so in Hebrew, even in the examples alleged, as the forms admit of a difi'erent explanation. See in regard to the first passage, Dr. Alexander's Com- mentary. ^56 CONSONANT CHANGES. 75 2. A few cases occur of the rejection of a letter, chiefly ) and "a , from the end of a word. a. 'i of the verbal endings ",1 and "p . is almost always dropped, being only retained as an archaeic or emphatic form, and chiefly at the end of a clause, e. g. )^^-}1 Deut. 8 : 16, but mostly V^iy, "^"i^lO Gen. 32:20, com- monly ^"iann; i'^O^'ri Ruth 3:4, commonly ""ia^n • So, too, in some proper nouns', 'lina^ Zech. 12: 11, iij'S Josh. 12 : 21 ; ^'^^^, , whose original ■| is shown in the derivative "^i'^'^^ ^"^ ''^ perpetuated in the modern name Seiiun. h. In like manner a is rejected from the dual and plural terminations of nouns upon their entering into the close connection of the construct state with the following word, ''is].^ from C^ilX, ''03 from t3"^n3 . c. If the feminine endings n^ and n^ have, as is probable, a common origin, this may be best explained by the assumption that n is in many cases rejected li-om the termination, leaving only the vowel, though it is always retained when any addition is made to the word: thus, the con- struct state r^=n, absolute nisn , but with a suffix •'n^san ; n^^il^ (comp. nliTS Deut. 32 : 36), ''rP^::^;^. It is to be observed here, that this phenomenon does not establish the possibility of an interchange between the consonants ti and n, because n in this case represents not h but the vowel a. § 5G. A few other changes remain to be mentioned which are due to special causes. 1. Nun is often inserted in certain forms of verbal suffixes to prevent the hiatus between two vowels, "i"?';!^?;! Je.r. 5 : 22, or § 53. 3. a. ^nn?^. Isa. 33 : 21 for ^nnn?;: , ^ni'a'ai^.s: Ex. 15:2 for ^n'b'aini!: . Comp. Gr. ayoo-to? and Enolish indefinite article an. 2. Vav at the beginning of words is changed to "^ , e. g. ^b'^ for "i?i , "i^,^ for "iy^ , ^'i^ip? for bbjpi . The only exceptions are the four words 'il , nrn Prov. 21 : 8, ^^) Gen. 11 : 30, ^b'j 2 Sam. 6 : 23 (K'ri), and the prefixes Vav Conjunctive and Vav Conversive. 3. Vav, though capable of being reduplicated, e. g. ^?;X is in most instances relieved from this necessity by the sub- stitution of \ or by doubling the following letter in its stead, e. g. Drp^i! or n^aipx for D^pi? . 76 OHTllOCKAniY. §57 I!. Ill one iiis(anc(> aliiM' surli a clian^ri" of 1 lo ■^, a lullowinir i siillcrs the contrary chaiigo to 1 to jurvont \\\c lri])le renirroiico ol' tlio t?anie letter, T|ltl!^. 1«'^- 0:9 for T(::^":x . 4. Yodli before tlie plural termination D"" . is in a few cases eliang-ed to N to prevent the eonjnnction of like sonnds, Q>'ibn llos. 11 : 7 for D^^'^n Josh. 10 : iCO ; D>'int llos. 1! : S for a^hs Cum. 10 : IV) ; n"'s?'i^n from ^'^ri; n^^ni: (also tiisDi^) for u-^'':^^ ; \s:i5a Jcr. o8 : 1.0 for "^iba (or as some read, r'^ib3) vcr. 1 1. inauiUM- T is chan-roil to a. before ni in llic worJ riss lor M>3 from nip ; it is ooiisoqiieiitly iiiiiioeossary to assume, as deseuius ilocs!, a siiia-iilur riX3 which never occurs. Change of Consonants to Voavf.t.s. §57. The second class of changes is the conversion of consonants into vowels, or the substitution of the latter for the former. Tliis occurs, 1. Occasionally in reduplicated syllables or letters, 2ii3 for n^ns ; nit-ji-j for nir^rj; 5na for baba Gen. 11 :9 ; ninbar 2 Chron. 35 : 13 from nn?2? Prov. 19 : U. 2. ]\[uch more frequently with the quiescents. (1) A prelixed ) is softened to its homogeneous vowel u before other labials or vowelless letters, e. g. rr^i^, "^i'}^; the softening of an initial "^ to J only occurs in '^ir'^x 1 Chron. 2 : 13 for ^r"^ ver. 1 :2, t'X 2 Sam. 14:19, Mic. 0:10 for c."? . {^) ]\Icdial or final quiescents Avithout voAvels of their own often lose their sound in that of a preceding vowel. This is invariably the case with i ami "^ following their homo- geneous vowels, e. g. ^"H^n for ^"iin § 59, ni^n'^a for n'l^r;';'^, unless they are doubled, as "^'r.'^;^^ , n^s, and occasionally even ^ 57 CHANGE OF CONSONANTS TO VOWELS. 77 then, e. g. "'t?"'^ for '^t??''? . I'inal x always, and medial i< fre- quently, gives up its consonant sound after any vowel what- ever, e. g. ViT/2 , in'i-Q , rxib for nsjib . a. Medial X regularly loses its consonantal power in the future Kal of Pe Alepii verbs, e.g. bix'i ; in "itx prceeded by b, thus 12xb ; in C'^n'^X and certain ffjrms of ITiX preceded by the prefixes 3 b 3 T , thus, C'^n''!5xb , •iri'bxb but ni^xb; "^pxb', -ppxb , -pxb but r'"'^:^^ ^Pi*!^ , ^:^^^.^ • The following examples are ol'a more individual character, e.g. n-XD ibr n^X3, nb'X] l^Kin. 11:39 for ns^'X; , r,DEOxn Num. 11:4, t;->f5lX3 Jer. 40:1. iT'r^X'Jx:: Isa. 14: 23. In a lew cases this has led to a change of ortho- graphy, the X which is no longer heard being dropped, or another vowel letter substituted for it. c. g. ib^i"* Ezek. 42:5, and h'^z'^.n Hos. 11:4 from b=X, •,r!a--i Job 8:8 Ibr yiix-i,' '::i-i Deut. 32:32 for TTXi, and the exam- ples cited {'.03. 2, a. b. The consonant in never loses its sound in that of a preceding vowel like the rent of the quiescents. The letter n is often used to denote a vowel, but if in any word it properly expres.ses a consonant this is never converted into a vowel, or vice versd. The exceptions are apparent not real, as in the frequent abbreviation of the ending W^ in proper names to n"i, thus Jin'p'in, nj;r*n. The change here does not consist in the rejec- tion of the vowel 1 and the soltening of the consonant n. but the syllable in is dropped, whereupon final Kamets is written by its appropriate vowel letter, § 11. 1, a, just as ^'"'^-■'a after the rejection of 'n^ becomes ti;"'5a , So in those rare cases in which n is substituted for the suflix n, e.g. nnrb Lev. 13:4 for "^''J'^- The proper name bxnns Num. 34:28 is de- rived not from iris but ni^, a root of kindred meaning, of whose exist- ence, though otherwise unattested, this word is itself a sufficient voucher. (3) Medial N often gives its vowel to a preceding vowel- less letter and rests in its sound; "^ occasionally does the same with a homogeneous vowel, when preceded by a vowel- less prefix, a. Thus, X: c-^axn for cifcx-i , rxcn for rx'jn; tj'jx'j Ezek. 25:6 from -JX'r ver. 15; xvij Ps. 1.39*: 20 for ix-yUJ . so xii:';' Jer.''l0:5; ""J^lx-, from "pix-i ; cx-nia Neh. 6: 8 for cxni3 ; xin Isa. 51 : 20, ixn Deut. 14:5; c^><-J^I 1 Sam. ^14:33 for c\V^'n; 'n-'ixs Lsa. 10:13 for vaxs; -\'^_::^H'<. Zech. 11:5 for ~i"X'; ; this even occurs after mixed syllables, e. g. i^^xba for nixb-s ; yx:'; for 7"x:^; rixnjrb for rx-ipb, particularly in proper nouns' bxi^r-] for bx:;^^^ ,'bxi;-n for bxr-n. So, i:"nn-3 Eccles. 2:13 for '("i-in'i; i"'?'?"',^ Jer. 25:36 for rpb-;! ; r.nis-b Prov. 30:17 for rhp'b. There is no instance of this with 1. on the contrary, nia^ip Cant. 5:2, 12. (4) At the end of words ^ and ■•, when without a vowel of their own and preceded by a vowellcss letter, invariably 78 OETHOGRAPHY. § 58 quiesce in their homogeneous vowels, "i in an unaccented u, "> in 1, which draws the accent upon itself and frequently causes the dissolution of a previous syllable and the rejection of its vowel, ^nia for ina , ^nniy'? for ^nniij^ ; ^n": for '^r-o , ^-isi for ;^n3 , ''^5 for ^^^ . (5) When preceded or accompanied hy heterogeneous vowels, 1 and "^ are sometimes dropped, or if the vowel be cf, they not unfrequently combine with it, forming the diph- thongal and e, § G2. 1, e. g. p2^n for p^^r. , n'Sa for ■'"'i?, nSa for ^%, D)? for Dtip, a-ipn for D^ipn, n^ for nn)2; n^isin for ni"iLhn, ^'^^2 for TJJ!i'7P , ^I'i''^ construct state of tyj-Q , n^i const, of n^a , bi^^n for b^^;'n , r3ro for ^"7^^, . a. Vav rarely remains with a heterogeneous vowel unless accompanied by weak letters, by contrast with which it becomes comparatively strong, e.g. nin, •^h, ni^. Vowel Changes. § 58. 1. The third class of changes embraces those which take place in the vowels. The primary office of the vowels is to aid in pronouncing the consonants, to which conse- quently they are quite subordinate, merely occupying, so to speak, the interstices between them. Their number and variety being greater, howxver, than is demanded for this single purpose, they have besides to a certain extent an in- dependent value and meaning of their own in the constitu- tion of words. (1) Changes of vowels, wliile they cannot like a difference of consonants create distinct verbal roots, are yet frnitful of those minor modifications of which etymology takes cognizance, such as the formation of derivatives and grammatical inflexions, e. g. '5"i5 to he great, b^ia greatness, bina great; bt2]p he kiUed, biti]^ to kill, Vib]? kill tJiou, bbp killing, brjp killed; C^D a horse, HD'C a mare. (2) They may indicate differences in the forms of words which have ^58 VOWEL CHANGES. 79 arisen in the lapse of time; i?i in the Pentateuch means in- differently (jid or boij, in later books (jlrl is Thv^ ; s^n in the Pentateuch Ue or she, in other books she is always s''n ; the form of the demonstrative nr'^n is found only in Genesis, T^n in writers after the time of Moses, 'nT"?n in Ezekiel; the plural of the demonstratiye in the Pentateuch '^s or n'^ii , elsewhere, with a single exception, n!ps . The imperfect no- tation of the vowels in the original mode of writing by letters alone has, however, left us without the means of ascertaining to what extent such changes may have taken place. (3) They may indicate diversity of dialect, e. g. ^t3)5 to kill, Chald. Vjp, Syr. ^>-.tV^ , Arab. JCxS' , Ethiop. ^-tA : . 2. The vowel changes with which orthography is con- cerned, on the other hand, are pnrely euphonic, being in themselyes void of significance, and springing solely from the natural preference for what is easier of utterance or more agreeable to the ear. Orthographically considered, vowels are either mutable or immutable, the latter behig unaffected by those circumstances which occasion changes in the former. A yoAvel may be immutable by nature, or made so by posi- tion. A short yowel in a mixed syllable before the ac- cent is ordinarily immutable by position, being beyond the reach of the common causes of mutation, e. g. is"''^ , rinsTrip . Long vowels are immutable by nature in certain words or classes of words ; but they are only distinguishable as such by a knowledge of the etymological forms which require them. It may, however, be obseryed, as a general though not an invariable rule, that the vowels of such words and forms as are prevailingly written with the vowel letters are less liable to mutation than those which are prevailingly written without them. IMutable vowels are liable to changes both of quantity, from long to short, and the reverse, and of quality from pure to mixed {ii to o, i to e, a to r) and the reverse, these changes being confined, except in rare in- 80 ORTHOGRAPHY. § 59, 60 stances, to the cognate forms ; thus, i never passes into u or 0, nor these into a. Only as C stands in relation to both i and «, it serves to mediate the interval between them, and thus accounts for the occasional changes of i to a or the re- verse, e. g. "s^Ppn , J^^r'^i?*^ j ^^ fo^ '^■r^ ' ^^^ ' ^^ ' °?^^ comp. a. The exceptional change from it or o to e occurs only in the pro- nouns, e. g. cnbijp, belbre suffixes ^nb::p ; and in the particle nx , before suffixes r X . There are also a few examples of the change of short vowels in mixed syllables before the accent, e.g. nss^p. , construct nnsn^a, plural n"::3i'2. § 59. The mutations of vov^^els are due to one or other of the following causes, viz. : 1. Syllabic changes. 2. The influence of consonants. 3. The influence of vowels. 4. The accent. 5. The shortening or lengthening of words. As the vowel of unaccented mixed syllables is always short, and that of simple syllables long, §18. 2, it is evident that a change in the character of a syllable will involve a corresponding change in its vowel, unless the accent interfere to prevent. Accordingly, when for any cause a mixed syllable becomes simple, its short vowel will be converted into a long one ; and when a simple syllable becomes mixed, the reverse change will take place, e. g. "in , B"'"in ; n^p , P'cp . In the case of the vowels i and tt there is frequently an addi- tional change of quality, viz., of t to B and u to o, e. g. D''pn for n^-i)pri; ^;i3 for 'jiiD in place of ^i? § 56. 3. a. Dagliesli-lbrte is thus resolved by tlie prolongation of the previous vowel in 'ci-ap^ . 'C^^■^p■, -^5|q , ^?!?^3; "''i;''?'!'. , ""uiiTon ; Dij-ii'n, Di;i-n:3 ; ipn, ipin; u^^c-pr^ Eccies. 9-12 for c-'i;;?;':^ §33. 2. a; s-iporn for !i-ij?finr; ; nn^3 Lam. 1:8, if this is for !Tn3 see ver. 17; and if the conjecture of Gesenius (Thesaurus, p. 4S3) be correct as to the true reading in 1 Chron. 23:6, 24:3 cpsm for np!rn;i. § 60. Contiguous consonants may give rise to vowel changes by their individual pecuharities, as is the case with § 60 VOWEL CHANGES. 81 the gutturals, or by their concurrence. The pecuUaritles of the gutturals are fourfold, viz. : 1. A preference for the vowel Pattahh of the same organ, into which, consequently, a preceding or accompanying vowel is frequently converted, e. g. nbir for ribw ; wjh for D3?S ; mS.^^ for f^inro ; s^'Istj: for y^TiJ ; ^^^its from N'iiu . a. The instances in which tliis permutation occurs cannot easily be ennbraced under any general rules. In some cases it was optional; in others, usage decides for it or against it without, however, being absolutely uniform. The following statements embrace what is of most importance. (1) The stability of the vowel often depends upon the weight attached to it in the etymological form ; thus, IS^'d in the imperative but not in the in- finitive for s^d; rrir-i for t-h^-^ , but riia not ria ihr i'iir. (2) The vowel preceding the guttural is more liable to change than that which succeeds it, e. g. yht"^^ always, but hs^"^ and hsxi'n; n:m but Gnni; flpST but Jll'OS . (3) An accented vowel is sometimes retained where one un- accented would suffer change, e. g. siin^i but "jn*;!; '■irM; cn^. (4) O and ic are less subject to alteration than t' and e, e.g. b?3 for brQ ; a which is already cognate with the gutturals is mostly retained, though it occasion- ally becomes a before n, e.g. finx from nx, "^n-jn^ Job 31 : 24 (in most copies) from nb::^, n^^ from nn^av (5) ft in many cnses prefers the diphthongal vowels e and o, tJius itipx, "^pxiri, ^'|Si^n, "ibx"; but tiTnoxi ; ^^"^i ^tX"". (6)^ n partakes of this preference for a to a limited extent. e-g. ^?^] for "ip;''T or "lO^i; t<-i^i from nN-17 . 2. The reception of Pattahh furtive, § 17, at the end of a word after a long heterogeneous vowel (i. e. any other than a), or before a vowelless final consonant, e. g. ^_'i , i^^ir^, nS, a. This is necessary when the vowel preceding a final guttural cannot be converted into Pattahh. Sometimes the form wiUi Pattahh and that with Pattahh furtive occur interchangeably, e. g. rj|x"b and n?db , or with a slight distinction, as nfCwX, in pause :n^'::N; nsTTS, construct' nata . In a few instances a guttural preceding a final vowelless letter takes' simple Sh'va instead of Pattahh furtive, e. g. nnpb 1 Kin. 14 : 3, and in most editions S^in^'j Jer. 13 : 25. As final x is always either quiescent or otiant, it never receives Pattahh furtive. The letter 1 never takes it unless it be in a single instance, and that in a penultimate syllable Cj^i" Ps. 7:6, which is probably to be read yfrdoph ; though it might be pronounced yiraddoph, which some conceive to be an anomalous form for C^i--. , after the analogy of rn:i;: Gen. 21:6, the compound Sh'va being lengthened mto a vowel followed by euphonic Daghesh, as in the related words 6 82 ORTHOGRAPHY. § 60 n^^isn Isa. 1:6, and n"^2n Isa. 53:5, while others adopt the explanation of the old Jewish Grammarians, that it is a peculiar combination of the Kal ^i^T. and the Piel ^^'^'J . 3. A preference for compound rather than simple Sli'va, ^16. 3, whether silent or vocal, inasmuch as the gutturals are more readily made audible at the beginning than at the close of a syllable, and the hiatus accompanying them as- sumes more of the complexion of a vowel than is usual with stronger consonants. a. The gutturals occasionally retain simple Sh'va when silent. This is regularly done by a final radical n, n or 5, followed by a servile letter, e.g. Pjnij, ^^'-^il, crij;"!^, cri:;j?:u372 , innadri, with few exceptions as r,!ii3)"i';i 'Hos. 8:2, ^^h,i? Gen. 26:29, cw^0n 2 Sam. 21:6. Other cases have more of a casual or sporadic character, and occur chiefly with the stronger gutturals n and n. nrjn";, "sn:, f^^'l^ , ^^np, ^uin^nri, nirrno but mi^rn^, \riQ]i_ but ^irin^, n^n.D a possession, but f^^ni from bni a brook; niore rarely with X and V, c^s'S Lev. 4 : 13, NOi-a 1 Kin. 15:16, C^iys Isa. 11 : 15, f^'^V.k Deut. 25 : 7 but in pause '. tr^^fj Isa. 28 : 6, •'"I'lX.? Ex. 15 : 6; 1 has for the most part simple Sh'va n'i-i, cbnS';', though in a few instances it has compound "is"!?.]?, ^"^^ti.J?!!. b. (l) Among the compound ShVas the preference, unless there is some reason for choosing another, is ordinarily given to Hhateph Pattahh, as the simplest and most in accordance with the nature of the gutturals, and to this an antecedent Hhirik, when unessential to the form, is commonly made to correspond, e. g. iri? , 1b?.^_ for 'i^^'? . Sometimes, particularly with N (see 1. a. 5.) Hhateph Seghol is taken n6^X . f^i^'2^c , nnTN , ri^ii , ^rix, cinx, rri-iis^, W^_, n^h, r\^-i-J Joel 2 : 5, r|'TnNl> Jer. 13:21, which not infrequently becomes Hhateph Pattahh upon the prolongation of the word ?;"}^x. , """Ji*. Prov. 25:7, inriwS:. , "lains*^ , iin^'in^X';' Judg. 10:2, or the carrying forward of its accent "'n'lixn, ''P}13i<:V^'] , ''J\i'?'!!r!vl j "'^v't'T'TIv!!! • (2) If, however, z or 6, characteristic of the form, precede, this commonly determines the Sh'va to be selected, e. g. ^T^a^n for T'53^'n, "^k^.^ for "li^^, i^i'Q for ''t'^S; though sometimes Hhateph Pattahh is retained and the intermediate syllable, §20. 2, resolved into a simple one by prolonging the vowels, e. g. 7\-}iit\}. Josh. 7 : 7, n^?:H, l^.^'Q Isa. 1 : 31. Hhirik may, how- ever, remain short, e. g. Tj^'ln , sij?.^, ^'in^ Job 6: 22, particularly if a Daghesh-forte has been omitted from the guttural, e. g. «^£5?,J?'? and §61 VOWEL CHANGES. 83 in-ixri'^ Isa. 44: 13; l^'^^^n Job 16: 16 (K'ri in some copies), 1^^'i^n Lam. 1 : 2U,' inj<;ri Isa. 52 : 14', 'inxn 1 Satn. 28 : 14. c. Before another guttural the compound Sh'va is frequently re- placed by the corresponding sliort vowel, e. g. Ti.^Nrt for TiJ^*?!!!, "'fiT^'Sn for ■^n'T'^Jn , CD"'nNn for C2"'nNi'i ; aiui occasionally under N by a long vowel before other letters as well as gutturals, or by a short vowel with Daghesh, e. g. C-'pnis: for c^phn, l^rrnk, C^i.N for Olis , nilX for "liiN , "iSX for ^OX. This disposition to render the gutturals more audible by the aid of a vowel is further shown by their attracting to themselves the vowel of another letter, particularly in trilileral monosyllables, e. g. S'lT for S-1T (sJnt), yar\, s6i, -^i^-q, rnp 2 Kin. 12:9, csa for 1^x3, ^xa", also' 'lii-ip Ex. 2 : 20 for' JX'np' Ruth'l : 20, ^linJtPi Prov.'l : 22 for 'irixr!, !in^2xn'job20:26 for iinHDNn, o':?Di<'! Zech. 7 : 14 for nn?DNi , and by their sometimes causing an antecedent or accompanying vowel to be re- tained where analogy would require its rejection, e. g. "^xril- for ■'Jt^ia from siia, "^X'L^r! , ''i'^^ , ^P.^':^ f^'id ''P.^l^, c-^-iln^ from n^na comp'. 1. a. (4), -in^iiiTS-; Deut. 32: 10; nSr-a, n^rn .' 4. An incapacity for being doubled, whence they never receive Daghesh-forte, and the previous syllable thus becom- ing a simple one, its vowel is generally lengthened, § 59, a to a, I to t, a to o, e. g. "ii^ia for l^-o , y^-a for fxia , 1]nh;i for a. Sometimes an intermediate syllable, §20. 2. is formed, and tlie vowel remains short. (1) This is commonly the case before n, frequently be- fore n, less often before -J. rarely before X, never before "i , e.g. ens, iri"^, "iH"?, -^;n , j'NJ . (2) It is more likely to occur in the body of a word than after a prefix, e.g. '^nbn';' Ps. 119:43 from hn^, but p^n;; Job 38:24 from pHn . (3) When the guttural comes to stand at the end of the word the short vowel is often resumed, e. g. ""^nn Prov. 22 : 24 from rtinnpij I'sn Ps. 141:8 from nnsn but lirn Deut. 2:9. There are a very few in- stances in which Daghesh-forte is found in 1 , e. g. T]'^'^ n':i3 Ezek. 16 : 4, rrn-a Prov. 14: 10, r^ti"."!?.?^^ Prov. 15: 1 (in some editions), •^ix'n-d Cant. 5:2, see also §24. b. ' §61. The concurrence of consonants gives rise to the following vowel changes, viz. : 1. When two vowelless letters come together at the be- ginning of a syllable in contravention of the law in § 18, the impossible combination is relieved by giving to the first of them a short vowel. This, if there be no reason for prefer- ring another, will be the briefest of the vowels, Hhiiik, e. g. 84 ORTHOGRAPHY. §61 ■inn^ for ^^s^ , li^a for 1^73 , ipm for ^pTn . If a vowel lias been omitted from the word, the corresponding short vowel is frequently employed, e. g. '^i'p'a for '^D'p^a from tjbia ('?|^'a); iDb^ from ^■'5^ ; '^Ipnn from bnn , ^ripn;" for ^2n^ from fi-; . Or if one of the consonants be a guttural, the vowel mostly con- forms to the compound Sh'va, which it has or might have, e. g. VT^. for ^l^V., , ^pTH^ for ^pTn^^ , ''^nb for "^bnb , nbnb for 'iDPh , ^bys for ^b2?D . a. Vav before a guttural follows the rule just given; before "^j and sometimes before n or n followed by "^ , it takes Hhirik ; before other vowelless letters it gives up its consonant sound and quiesces in its homo- geneous vowel Shurek, §57. 2. (1), thus "innJi , ■'in^i , n'^'ni and rr^n;; , si^i, b. In Iriliteral monosyllables or final syllables with the vowel Pattahh, the first letter sometimes receives an accented Seghol, to which the fol- lowing Pattahh is then assimilated, e.g. filns for tjns construct of Ci^S, td]?^'? for nabriTa ^ the Seghols being liable to be changed to Pattahhs by the presence of a guttural rnstJa for riiQda . c. In sisisb Gen. 32 : 20 for CDi:|:3:3 the vowelless letters belong to different syllables, and the introduction of the new vowel makes it neces- sary to lengthen the one before it. 2. Although two vowelless letters are admissible at the end of a word, §18, the harshness of the combination is commonly relieved by the insertion of Seghol, e. g. n^"? fo^ ^'-\'} , n^b for P^b . If either letter is a guttural, Pattahh is mostly used instead, e. g. ras , bys, )n^_ . If either letter is 'i , its homogeneous vowel Hhirik is used ; if the second letter is 1, it wiU rest in Shurek, §57. 2. (4.), e. g. n^i, ^bs, ^nn, but n;;,^ . a. When the penultimate letter is n or n, it in -a few instances takes Seghol, as bnx, ',ri3 , nnB, trS. When the final letter is X, it either remains otiant, §16, or requires Seghol, ^)^, ^I'^'l , ^"l-j '^'^5 5 '^ penulti- mate t< either quiesces in the antecedent vowel or attracts it to itself, §60. 3. c, ri>l?5, nxb or nxb, t:Ni. The alternate mode of facilitating the pronunciation of gutturals before a vowelless letter at the end of a word by means of Pattahh furtive, has been explained §60. 2. 3. When the same letter is repeated with or without a mutable vowel intervening, there is often a contraction into § 61 VOWEL CHANGES. 85 one doubled letter, and the vowel is rejected or thrown back upon the precedmg consonant, e. g. ^ID'!' for ^^^S'' , 3b^ for nine;! (Daghesh-forte disappearing at the end of the word), 2b for 22^, ^22^2-: Job 31 : 15 for ^sssio;* (see 4. below); if another consonant immediately follow the contracted letters, a diphthongal vowel ''.. or i may be inserted to render the re- duplication more audible and prevent the concurrence of three consonants, "^iniio, ni'^ipn. 4. In accented syllables the diphthongal vowels e and o are employed before two consonants or a doubled consonant in preference to the pure i and ii, e. g. n^iijn , npni^n ; D^p , n-Qp ■ b^tbpn , nsb-jpn , so '^son , t:©p , ^n^i . This is still the case when at the end of a word an auxiliary Seghol or Pattahh has been inserted between the letters (according to 2.), e. g. ^b. , ^20 , b?3, np^)3 from p^i?^x? , or the reduplication of the doubled letter is no lonojer heard arjd the Dadiesh-forte does not appear, § 25, e. g. 2Dn comp. '5"'t2pn . a. The vowel e is in like circumstances often reduced to one of its con- stituents a, e. g. •'nVjpn from ^"^ii^n, ^T\\hp_ from brjp5 , nj^^n from "^n, ■^P?^! '"'?"?;P}, ^^cn, and occasionally to its other constituent r, e. g. tR"^""i^^n Irom C-i;?rri , tnc3-i^ from t;n;. The only example of Shurek in a Segholate form is nrvjn Lev. 5 : 21. 5. In unaccented syllables i and u are preferred to c and ij before doubled letters, "nx , '^nx ; -nn , ^nn ; nacn from 2cn; noH, ''2|^;<; trq comp. bh^-q , -pn, ^pn, though such forms as ^"A^n , ^jy , tr\2 , n^ri likewise occur. 6. A voAvel is occasionally given to a final consonant to soften the termination of the word, and make the transition easier to the initial consonant of that which follows ; thus, ^:^^, ^)'i^; J"in53 for n?^ ; on, n^r,; bwV, n?X; \x, n^>?; ^Eh, --isn; to preserve the brevity of the antecedent vowel, § 24. 3, becomes ■'f ., e. g. ""nns? with Q^ . becomes D^nny or Q-'^inS' . So, \ be- fore !". ^ forms n^ _ , and before i forms i^ . , e.g. Sihsy , ni^ns^? ; in like manner ^ is changed before i into uv, form- ing ii. , which, by § 56. 3, becomes i\ , e. g. ln^D>^ , by the substitution of rii for in , tii^3^^ . / \ followed by u ^ forms iv, ^M^n'pt:;^, ™'5t?;p; in-^s, l^S; ^^h^ii for ^^p^n Josh. 14 : S. U "^ .. before I ■> . or u ^ is resolved into a?/, which, joined with the appropriate semi-vowels, becomes ^ . and 1^^, the virtual ^ 63 YOWEL CHANGES. 87 reduplication of the final consonant in the one case preserv- ing the short vowel, which is lengthened in the other ; thus ''O^D with "^ . becomes ''D'D , and with ^n , rD^o . The same resolution of "' .. occurs before final ^, forming ^^ ., and by § CI. .3 ^^ ., thus ^^rj: with tj becomes V^^^^ . a. Grammarians have disputed whether in sucli Avords ns C"'^~3yj ri'2b?3 ihe point in "^ is Daghesli-forte or Mappik, §2G, and accordingly whether they are to be read ib/in'i/ylni., vtalklmyijolh, or ihliilijlm, malkhuyoth. If the exphmation giv'en above be correct, it is Daghesh- forte Conservative. Comp. C^p, c;^;^ . b. Such Ibrnis as '^■]"i3, 'i"'"}?, c^"iQ from "'"iS are only apparent excep- tions to the above rules. The word is properly ';'"}Q, and to this the addi- tions are made, the auxiliary Hhirik being dropped with the cessation of the cause from which it originated, §57. 2.(4). In D"'Ni3'i:y 2 Chron. 17: 11 from "'2"}2 and Q"' _ the vowels are kept separate by an interposed X. c. In words of n^ formation, such as nib:?, "iibi", Di'CJ:' from niL"3> and ri^, i , c"! , it might appear as though one vowel were rejected before another. But the correct explanation is that "^ is the true final radical, and the forms above given are for M^^Js", i'^cJ-'', ni';ii:;:J (like C"'p::'p) from which "1 is rejected by §53. 3. In the same way ^iL">\ ~-7', etc., from fiirs are lor ^''b^' , ^?^'>'- In such alternate Ibrms as n;j:'}b I'rom i^Va. the radi- cal "^ is retained by preserving the antecedent vowel, which, before Daghesh-forte Conservative, becomes Hhirik, §61. 5. §03. The following euphonic changes are attributable to the proximity of vowels, viz. : 1. Pattahh before a guttural is often changed to Seghol if another a follows, and the same change sometimes occiu's after a guttiu'al if another a precedes. The particular cases axe the following : a. When (_) stands before a guttural with (^) always before n, e.g. inn for ;nn, nn-j^'Q Prov. 21:22. c=nn, ^nrnsn (also when n has Hhateph Kamets, e.g. Ci"'i!:3nnn , ""nbnnn Judg. 9:9), often before n and 'J . particularly if it receives the secondary accent, e.g. Ci"inn for c^'^nn. ninb but tzWd. ^'^'rv::!P: ; f^^^"", t^-'-q-J n-2, rarely before i< and "i, 1^-i}^ Gen.' 14 : 10, riiiss Neh. 9 : IS, 26 but ""^ni^xi Ezek. 35 : 12. b. When ( ) before a guttural is followed by another consonant with (-) Of (J ^'iin.?. J ^''„^n?. but ^''^n^, xinp but nxina, N-jn;;, once before the liquid ^, e.g. '^i^sx E.\. 33:3 (or ~|rt-i!) &"d once before 2. e.g. n::rt for nijnb. c. In nxnpxj 1 Sam. 2S:15 and the combination 'i"^ cbi3 a similar change takes place after a guttural to prevent the repetition of the vowel a; so in nilS'i'i Ps. 20:4, and n:xi n:x after the liquid 3. 88 ORTHOGRAPHY. ^ 64, 65. 2. Pattalih is sometimes assimilated to a following Seghol, or to a preceding Kamets or Tsere. a. The assimilation to (..) takes place regularly in what are called Segholate forms, in which an auxiliary Seghol lias by §61. 2 been intro- duced between two vowelless letters, Ti^a forTj^i, 'Z'\^\ for n"";, y'ns for "/"IN, but ST^a, crQ J only before 1, which can combine with a and not with e, a is retained and lengthened to (J by §59, "X, Tip\- Rarely in other cases cd'i;; for ci'iV where the change is facilitated by the pre- ceding "1. b. The assimilation to (^) occurs in a few cases after a guttural with rj prefixed, e. g. ci-fn for cyn, inn for inn, ^ixn for ■]''7.x^! • c. The assimilation to (_ ) occurs in the Kal future of Pe Yodh verbs where the alternate forms are -!i|] and y 15''? • § 64. The folio v/ing vowel changes are due to tlie accent, viz. : 1. If a long vowel in a mixed syllable be deprived of its accent, it will be shortened, §18, e.g. it?T?n, '^^^^■, ^^\ nc^i; Dt\\ Db^'i; nt?;., "li??,^. a. If a vowel preceding Makkeph is incapable of being shortened, it will receive the secondary accent Methegh, agreeably to §43. 2. The accent prefers to be immediately preceded by a simple syllable and a long vowel. Accordingly an antece- dent vowelless letter often receives what may be called a pre- tonic vowel. This is commonly the simplest of the long vowels u, e. g. bt:;^ , nc;) , nnpb , l^^n;; , occasionally e, e. g. ^i?.'' , s^^'ii^ , V^^^. , V'^'^'^^ , rarely u, e. g. T\^'^^^ . Such a vowel is sometimes inserted, even though a pre-existing mixed syllable is thereby destroyed, e. g. in the plurals of Segholates and of feminine nouns derived from them, D'^i^^ from ^b)2, niib)2 from nib^ . §65. The special emphasis, with which the last word of a clause is dwelt upon, gives rise to certain vowel changes in connection with the pause accents, §36. 2. gj. These are (1) lengthening short vowels, viz., (.) and not infrequently (J which has arisen from (.) to (J, e. g. ^ti?, Tax ; P:?J^3, Mns ; rnx , rnii • 'ini? , nns? , and brino;in<2: back Kamets § G6 VOWEL CHANGES. 89 rihatupli shortened from liliolem to its original lengtli froH, ^'^r^ • (2) Restoring vowels wliicli have been dropped in the conrse of inflection, e. g. ^in2) , i^ns' ; t\2^ , T\yi ; Ti)2^ , ^1^2? . (3) Changing simple Sh'va in triliicral sylla- bles and before the snffix '^ to Seghol, e. g. ^r^'^3, ^jn^a ; ^n^^ "^^2 ; D^T^ , 031^ . (4) Changing compound Sh'va to the cor- responding long vowel, e. g. "'IK, ''rx; "^in (i^^rn), ^^n ; ^^n^ a. Pattahh sometimes remains without change, e.g. 'i^ Ps. 132:12, nnsjn 2 Sam. 2:27, siD^S? Jer. 7 : 10, ^n=i:.ri Prov. ^30:9, "'nr^Ji:: Job 34:5, i-^rtrx Neh. 5:14. Seghol more frequently, Ti^^., P"^., n-i]^ , -n^ and T|":'n . Long vowels are mostly unaltered ; only Tsere is in mixed syllables occasionally changed to Pattahh, e.g. JTPrr Isa. lS:5ibr trin, so ;2irn Isa. 42:22, !-sn Gen. 17:14, %r>^1 Gen. 21:8, 7^}.^, Gen. 25: 34, which, in one word of Segholate formation, is converted to Seghol, e. g. S.'d^, Slav Where the same word has alternate forms, one is sometimes se- lected as the ordinary and the other as the pausal form, thus Vbn;^ , "I'Bn;; ; dnn'^, ^'21!?; ^"^^^j "^^'^5 ■'fi^==^\ !''ri"^-?'^ Gen. 43:14; p");i Ec'cl. 12: il, :"ii"iv,l Sam. 13:21; T^, ny GenV49:3, rarn , na^in Lev. 26:34, 35; ^^"3?, ^^'iJ3"' . Sometimes, instead of changing the Sh'va be- fore ^ to Seghol, its vowel is shifted, thus T^3, Tj^ ; T|^, Tj^", "t^i* , and in Ex. 29: 35 "^tlia^- The position of the pause accent, so far as it differs from that of the ordinary accent, has been explained §35. 2. h. Of the pause accents, or those which mark the limits of clauses and sections, the first class, viz., Silluk, Athnahh, and Merka with Mahpakh, almost always give rise to the vowel changes Avhioh have been described; the second and third classes, S'gholta. Zakeph Katon, Zakepii Gadhol, R'bhi'' and Shalsheleth, e. g. ^-n^Il Isa. 13:8. do so frequently ; the fourth class, Pazer, e.g. 2 Kin. 3: 25,' Prov. 30:4, and T'lisha Gh'dliola, e.g. Ezek. 20 : 21, but seldom. Pausal forms are occasionally found with other Disjunctives, thus, Tiphhha irbn Deut. 13 : 5, Pashta ^n^'rtn ibid., Geresh vh"j Ezek. 40 : 4, and even with Conjunctives, e. g. "'JS Isa. 49 : 18 , nbii^n Ezek. 17 : 15, TJJ3 2 Chron. 29 : 31. \(SQ. 1. The shortening and lengthening of words has an effect upon their vowels. The shortening may take place (1) At the end of a word by the rejection of a vowel. This occurs only with (..) or (..) in certain forms of tib verbs, e. g.br>ri from nlijri, ri'^i for ^l'^''.'], in'^l 1 Sam. 21 : 14 for tiw-^l, ^'c\\ for nniii'^l . In the last two examples the short vowel is lengthened upon its receiving the accent, comp. §64.1. If the rejected vowel was preceded by two con- sonants, these will now stand together at the end of the word, and be lia- ble to the changes described §61. 2, e. g. v)"irt for i^2~n . 90 ORTHOGRAniY. ^ 6Q (2) In the body of a word by shortening a long vowel in a mixed syllable, Avhicli mnst, of conrse, be the one bearing the accent, § 32. 1, or rejecting a long vowel in a simple syl- lable before the accent (the pretonic vowel, § C4. 2), ^i'l , nn^; If^-g, 1fb)2; IC^p^; TiJ^)?^. a. This is in general the only reduction possible. The vowel of a mixed syllable, if short already, is capable of no further abbreviation ; and it cannot be rejected, or there would be a concurrence of vowelless con- sonants which the language seeks fo avoid (i^ip'in Prov. 30: 6 is an excep- tion). And the vowel of a simple syllable, if short, must have the accent, §32. 1, which preserves it from rejection. The changes above recited are confined to the last two. or, in case the accent is upon the penult, the last three syllables of the word; for the antecedent portions of polysyllables are already abbreviated to the utmost. Contractions due to the peculiari- ties of certain letters, as the gutturals and quiescents, which have been before explained, are not here taken into the account, e.g. ^'c'iJ , ^izxa; nia, nio; n'^n;', inv b. Where the last vowel cannot be shortened, it sometimes experiences a change of quality from pure to diphthongal, such as is produced by the pressure of two following consonants, §61.4, e.g. ^'^'n:;^, ^'^2;^;i ; ::"''c;in, rnir-in; nit;;', rlr; ; ivc;v tc"^; c^c^o, ''O^b. 2. If a word be lengthened by additions at the end, its vowels are liable to changes in consequence. (1) Such additions create a tendency to shorten the pre- vious part of the word in the, manner just described. Tor the normal length of words in Hebrew being dissyllabic, the genius of the language is opposed to transcending this limit any further than is absolutely necessary. If the addition is not of sufficient weight to affect the position of the accent, no abbreviation results. But if it is of weight enough to remove the accent, an abbreviation follows if it is possible for one to be made., e.g. nn^, Q''^^^, oi^nn^ for Di^")!! by §G1.1. (2) They produce changes in an ultimate mixed syllable. If the appendage begin with a consonant, the antecedent vowel will now be succeeded by two consonants and be liable to the changes consequent upon such a position, § 61. 4, e. g. rob-jjpn from b^t:;:^; n:)9p from D^p; ^pVipn from ^'''^'^^; ^GQ VOWEL CHANGES. 91 ir^btap from ^Dp . If the appendage begin with a vowel, it will attach itself to the final consonant, which will in conse- quence be drawn away from its own syllable to begin the new one. This may occasion the following changes : («) If the preceding vowel is an auxiliary Seghol or Pattahh, introduced to facilitate the pronunciation of the second of two vowelless consonants, §61.2, it wHl be rejected, inasmuch as it is no longer required for this pm-pose, e. g. isb^ from i^t , innp from nno . {b) If it be a short vowel, it must either be lengthened to adapt it to the simple syllable in which it now stands, or rejected on account of the disposition to abbreviate words upon their receiving accessions at the end, e. g. f^^'^JP and n'pif j5 from ^12)^ . The cases are very rare in wdiich a short vowel remains unchanged in consequence of its having the accent, §18. 2, e. g. rna^iq 1 Kin. 19 : 15 from ^ky^, r.b^t^n Ezek. 8 : 2 from b^TJ^n . (c) If it be a long vowel, it may be rejected, as '^'^'^'} from ^'Ci^"} , ^%tD from Dii? , or retained either unaltered, as rn'i2''f:n from D"^[?r}, ''i.l^'^ from'jiT^'a, or with a change of quality from pure to diphthongal or the reverse, n;[:in^ from pint) , ^niiDS from y\^} , ^;^pn from uyr\ , D'^bbs from t2''Sa . TABLE SHOWING THE CHARACTER AND AFFINITIES OF THE VOWELS AND THE ORDINARY LIMITS OF EUPHONIC CHANGES. Lu>is Short. Guttural, . Palatal, Labial, pure i dijjlithongal pure dijjJitJionf^al 6 pure u PART SECOND. ETYMOLOGY. Roots of Words. § 67. Etymology treats of the various kinds of words, their formation and inflections. Three successive stages are here to be distinguished. The first is the root or radical portion of words. This embraces those fundamental sounds, in Avhich the essential idea originally inheres. Roots do not enter, in their nude or primitive form, into the current use of language, but they constitute the basis upon which all actually occurring words, with the exception of the inorganic interjections, are constructed. The second stage is the word itself in its simple uninflected state ; this is formed, if a prim- itive, directly from the root, if a derivative, from a pre-existing primitive, by certain changes or additions, which serve to con- vert the radical idea into the precise conception intended, which is as yet, however, expressed absolutely. The third and only remaining stage is the word as it appears in the ac- tual utterances of speech, so modified by inflections as to suggest the definite qualifications of the idea, such as the tense of verbs, the gender and number of nouns, and the de- gree of adjectives, or its relations whether of agreement or subordination, such as the persons and modes of verbs and the cases of nouns. § 68. There are in Hebrew, as in most languages, two classes of roots, which may be denominated respectively pro- § 68 ROOTS OF WORDS. 93 nominal and verbal. Pronominal roots form tlie basis of sncli words as express the relations of things to the speaker or to one another, viz., pronouns and certain prepositions, adverbs, and otlier particles. From verbal roots, which arc by far the more numerous, spring words expressive of ideas, viz., verbs, nouns, and such particles as are derived from them. Verbal roots consist exclusively of consonants, and are almost invariably triliteral. The introduction of a vowel or vowels, even for the sake of pronouncing them, destroys their abstract radical character, and converts them into specific words of this or that description. Nevertheless, for reasons of conve- nience, the letters of the root are usually pronounced by the aid of the vowels belonging to them in the simplest form of the corresponding verb, which is mostly the third person sin- gular of the preterite, e. g. ^t?)^ , tjH'a . This must not be suffered, however, to lead to the confusion of identifying that particular verbal form with the proper radical, nor of supposing the verb to be the radical part of speech from which nouns in all cases are derived : verbs and nouns are rather to be regarded as co-ordinate branches springing from a common root. a. The ^g.v7 quadriliterals and quinqueliterals which occur are mostly- formed from pre-existing triliterals by the addition of a weak letter, or a letter similar to one of the original radicals, e. g. 00*13 to lay waste comp. CDS; r|S"bT to hum comp. t{S^•, Mssno a branch comp. ^iS?p ; C^ernb thoughts comp. D'^QSJb ; a^ind a sceptre comp. tJlntJ ; lixbir tranquil comp. 'i;^?,^; twJ'iQ to spread comp. ^"^S ; or by blending two different roots, e.g. T^£i?"i to be fresh composed of -I3"n and tis:: ; "^bbbs a certain one = ^i^^^ "^i?? ; ?"n"i2:J a frog from "i^ to leap rnn (in Arabic) a marsh. Some, which are not thus reducible, may perhaps be of foreign origin. b. Many of the triliteral roots appear to be based upon pre-existing biliterals. Thus, the cognates ^ijs, iib, tu, nn, na, T^ia, have in com- mon the two letters tJ with the associated idea of cutting, §50.3. The frequent examples of this description, together with the fact of the exist- ence of a few biliterals, e. g. 25< father, rix brother, CX mother, have suggested the thought that the ultimate roots may in all cases have been biliterals, and that the triliterals were a secondary formation. Various in- genious but unsuccessful attempts have been made to demonstrate this 94 ETYMOLOGY. § 69 position by an actual analysis, and to effect the reduction of all roots to two primitive letters. Still more extravagant and fanciful is the endea- vour, which has actually been made, to explain the origin of roots from the individual letters of which they are composed, and to deduce their mean- ings from the names, the shapes, or other peculiarities of those letters. The existence of roots and the meanings attached to them must be ac- cepted as ultimate facts. Some have arisen, no doubt, from the imitation of sounds in nature; but in most cases no satisfactory reason can be given why a given combination of sounds has that particular sense, which is in fact connected with it. § 69. The formation of words and their inflection are ac- comphshed partly by internal changes and partly by external additions. The internal changes are the insertion of vowels and the reduplication of consonants in various significant ways, e. g. bb^, b-jp , bh'p , bujb . The external additions are significant syllables welded to the root or to the word, either at the beginning or the end, e. g. btsj?, ri'piijp, Vibfp^, ^ibtoj^nn. a. The triliteral and exclusively consonantal character of Semitic roots is their most remarkable peculiarity in distinction from those of the Indo-European languages which are as prevailingly monosyllabic, the vowel being an essential constituent, while the number of consonants is variable. The fact of the vowel being an integral part of the root in these languages interferes wiih their employment of internal changes for purposes of derivation and inflection, and confines them almost entirely to external additions, e. g. voco, vocabam, vocalio, vocabulum, vocito, etc. The composition of words of which such large use is made in tiie Indo- European tongues, 6. g. ad-voco, in-voco, etc., is almost unknown in He- brew except in the formation of proper names. b. Different languages differ greatly in their flexibility, that is to say, in the variety of words which may spring from a common root, and the number of forms which the same word may assume to express the various relations into which it enters. Relations, which in some languages are expressed by flection, as the cases of nouns, tenses of verbs, concord of adjectives, are in others indicated by additional -words, as prepo- sitions, auxiliary verbs, etc., or suggested by the order of words in the sentence. c. Formative syllables, added either at the beginning or the end of words for the sake of inflection, are, in the ordinary consciousness of those who use the language, completely amalgamated with them, so that their separate origin and signification is never thought of They are thus to be distinguished from those words which, by reason of their dependent character, are attached to others as prefixes or suffixes, but yet preserve their separate identity as prefixed conjunctions and prepositions and suf- fixed pronouns. §70,71 PRONOUNS. 95 §70. The parts of speech in Hebrew arc either decUna- ble as pronouns, verbs, and nouns (including adjectives) ; or indechnable, as the article, adverbs, prepositions, conjunc- tions, and interjections. As most if not all of the syllables employed in the formation and inflection of verbs and nouns are of pronominal origin, it will be necessary to consider the pronouns first. a. The classification usual with tlie Jewish grammarians is into verbs (ni^yQ actions), nouns ( niTsd names), and particles (o"'j:72 words). Pronouns. PERSONAL PEOXOTJNS. § 71. The Hebrew pronouns are personal, demonstrative, relative, and interrogative or indefinite. The personal pro- nouns are the following, viz. -. SINGULAR. PLURAL. 1. I ^i:i5, 13X We ^y j Thou m. nPi? Ye m CP,S ( Thou/, r-x , ^lyii? Ye/. ]r)i? , ri:px o j He x^n They w?. on, msn '^•(She N^n They/, in, n^n There are, it will be perceived, distinct forms for singular and plural in the three persons, and for masculine and fem- inine in the second and third. There is no form for the neuter, as that gender is not recognized in Hebrew. a. (l) The alternate forms of the first person singular "'ili^ (in pause "i^bx with the accent on the penult except Job 33: 9), and "^rs (in pause "'SS) are used interchangeably and with perhaps equal frequency. It has been observed, however, that while the former is the more common in the Pentateuch, it never occurs in the books of Chronicles, and but once in Ezekiel, viz., 36 : 28, a passage borrowed from the Pentateuch. The usual plural of this person is ^:r!t^.") "H? occurs but six times, viz., Gen. 96 ETYMOLOGY. § 71 42 : 11, Ex. IG : 7. S, Num. 32 : 32, 2 Sam. 17 : 12, Lam. 3 : 42 ; ^3N though common in later Hebrew, occurs but once in the Old Testament, viz., Jer. 42 : 6 K'thibh, where the K'ri substitutes the usual form. (2) The second person masc. sing. nnN (in pause occasionally npis Ps. 2 : 7, 25 : 27, 40 : 18, 70 : 6, but mostly nnx ) is in five instances written S^N Avithout the final He, which is however restored in the K'ri, viz., 1 Sam. 24 ; 19, Ps. 6 : 4, Job 1 : 10, Eccles. 7 : 22, Neh. 9 : 6. and in three instances nx without the final vowel Num. 11 : 15, Deut. 5 : 24, Ezek. 28 : 14. The feminine nx is occasionally written "'nx Judg. 17:2, 1 Kin. 14:2,2 Kin. 4 : 16. 23, 8 : 1, Jer. 4 : 30, Ezek. 36 : 13 ; the K'ri invariably retrenches the superfluous "i, though it is probable that the original pronunciation proper to this orthography was "inx . The feminine plural 'nx occurs only Ezek. 34: 31, where a few manuscripts read 'ifiX ; the alternate form t^:ni< oc- curs Gen. 31 : 6. Ezek. 13 : 11. 34 : 17 ; in Ezek. 13 : 20 most editions have (3) The third person fern. sing. XT) occurs but eleven times in the books of Moses, viz.. Gen. 14:2, 20:5, 38:25, Lev. 2: 15 (in some editions), 11:39, 13: 10. 21. 16: 31, 21:9, Num. 5: 13, 14. In its stead is found Xiri a combination of the letters of the masculine with the vowel of the fem- inine. The explanation of this is that Nin hu was at that early period of common gender and used indifferently for both masculine and feminine. As this primitive usage subsequently became obsolete, the word, when used for the feminine, was read X"'n hi according to the uniform practice of the later books, and the punctuators have suggested this by giving it the corresponding vowel, §47. According to Kimchi "in Ruth 1:13 and nsn 2 Sam. 4 : 6, Jer. 50:5, stand for the masculine plural; this assump- tion is unnecessary, however, as in the first passage the feminine may have the sense of the neuter "these Ihingsy and in the last two it is an adverb of place, meaning here. 6. (1) The pronoun "liibx unites the palatal found in the nominative singular of the first person in Indo-European languages, Gr. iyw, Lat. ego, Goth, ik, with the nasal of its other parts Gr. fxi, vuii, Lat. me, nos, Goth. mik. The same combination is found in the Coptic and the Phcenician. The Arabic and Syriac have retained only the abbreviated form in the singular and the prolonged form in the plural. The second person npiX is based upon the lingua! n as the Doric tv, Lat. tu, Ger. du, Eng. ihuu ; and the third person NW upon the guttural n as the Zend ho, Gr. 6, Lat. hie, Eng. he. (2) Words in such constant and flimiliar use as the pronouns are sub- ject to more or less irregularity in all languages. The original plural termination, as will be shown more fully hereafter in the case of verbs and nouns, is c^ . In the first person n is omitted to prevent the concur- rence of nasals in the same syllable, ''iX , s:x ; the plural of the prolonged ibrm seems to be best explained by supposing it to have been originally is::x , which was in the lingular softened to ''bbx by §57.1, and in the plural by a transposition and weakening of the palatal to a guttural (comp. Gr. eycj, Sans, aham), became irn'sx^ or by §53. 2, ^lina . The plurals of the second and third persons were originally DWS, CW, which are still § 72 PRONOUNS. 97 preserved in the Arabic an