Symbols denoting vocal stresses on particular syllables in pronouncing words or sentences. 1. In every word we utter, one syllable is spoken with greater emphasis and clearer enunciation than the rest. About it, as the strongly stressed or accented element, the other unaccented, or rather less strongly accented, syllables are grouped. Thus, in the word "contradict" the last syllable is the bearer of the main accent; a weaker, secondary accent rests on the first, while the italicized intermediate syllable is unaccented. Similarly, in a sentence, some words are pronounced with marked distinctness, while others are spoken hastily, almost without a stop, and made to lean forward or backward, as the italicized words in "he is a man of the world"; "I knew it." Both the accent which belongs to every word in itself ("word-accent") and the one which indicates its rank in a sentence ("sentence-accent") are to be regarded as the vital force which welds disjointed speech-elements into harmonious sense-units. The stops become particularly noticeable when, in a larger complex of clauses, they serve to mark the limits of each clause and its relation to the others. Some pauses are bound to be made, on physical grounds, to take breath; it is nearly always so arranged that the logical pauses shall coincide with those intervals. In an ordinary page of English the word-accent is never indicated (as it is in Greek), nor do the signs of punctuation (. : ; ,) show all the stops which careful reading in accordance with sense (especially oratorical delivery or the forceful recitàl of a literary masterpiece) requires. In the Hebrew text of the Bible, on the contrary, is found an elaborate system of signs (notations of stresses, or Accents) by which the stronger as well as the weaker stresses belonging to syllables and words are marked, so that a reader who is acquainted with the use of the symbols may recite the sacred texts correctly and, in appearance at least, intelligently, without considering grammar or sense.

Name.

2. The Hebrew (Aramaic) word V01p149001.jpg), plural V01p149002.jpg), which is used in the Masorah in the sense of "accent," "accents," denotes, in the first place, "taste" (in the literal sense, as in Ex. xvi. 31); then, "judgment," "good sense" (see I Sam. xxv. 33); in Talmudic Hebrew, "sense" (V01p149003.jpg "words of sense"; V01p149004.jpg "admitting of more than one sense"). This is the oldest term which thus conclusively proves that the Biblical system of accentuation was primarily designed to mark the various degrees of logical, or sense, pausation. This method of punctilious distribution of great and small pauses led, however, to a peculiar intonation in a half-singing style which is called Cantillation; this may still be heard in (orthodox) Jewish synagogues. The Accents have the secondary function of marking this intonation, each symbol being equal to several musical notes. Hence their appellation in Arabic, laḥn, plural alḥan, as early as Ibn Koreish, and the Hebrew term V01p149005.jpg "melody," plural V01p149006.jpg.

On the term "trop" (the same as the English "trope," in the sense of a musical cadence) used by the Jews in their vernaculars, see Berliner, "Beiträgezur hebräischen Grammatik in Talmud und Midrasch," p. 29, note 4, Berlin, 1879.

Sentence-Accent.

3. All of the Hebrew Accents are properly "sentence-accents." Hence they vary in form (V01p150001.jpg etc.) in accordance with their varying pausal functions. The sign once chosen, the "word-accent" is indicated by its place in the accented syllable, above or below the initial consonant in the center; when there is a vowel sign below, the latter occupies the center, while the accent sign is placed farther to the left: V01p150002.jpg. Some of the accents are placed, without regard to the accented syllable, invariably at the beginning or at the end of the word (hence termed prepositives and postpositives: V01p150003.jpg; in the editions of Baer, the notation is repeated on the accented syllable: V01p150004.jpg). A secondary accent (V01p150005.jpg "bridle," that is, check) is indicated by V01p150006.jpg. A word may lose its accent; then it is joined by means of a hyphen (V01p150007.jpg "coupler") to the next following word: V01p150008.jpg, V01p150009.jpg; the words thus united are regarded, for purposes of accentuation, as one word: V01p150010.jpg.

Place of Word-Accent.

Hebrew words have their main accent either on the last syllable (V01p150011.jpg "below") or on the penult (next to the last syllable) (V01p150012.jpg "above"). The accent is never found farther back (for a seeming exception see below). In the majority of words (word-types) the accent falls upon the last syllable: V01p150013.jpg, etc. Penultimate accentuation is found in the pronouns V01p150014.jpg (and the shorter form V01p150015.jpg (similarly in the dissyllabic suffixes V01p150016.jpg); in verbal forms of all stems (conjugations) ending in V01p150017.jpg), V01p150018.jpg; in the causative stem (hif'il), additionally in the forms ending in V01p150019.jpg and V01p150020.jpg; the latter rule applies also to verbs V01p150021.jpg and V01p150022.jpg in all stems (except those which follow the analogy of triliteral verbs), hence V01p150023.jpg, etc.; in the noun in forms with a helping vowel like V01p150024.jpg (compare V01p150025.jpg), as in verbal forms like V01p150026.jpg; similarly in the dual ending V01p150027.jpg; with the so-called locative ending V01p150028.jpg (with a few exceptions); in verb and noun before the suffixes V01p150029.jpg) (when preceded by [V01p150030.jpg and in forms of the type V01p150031.jpg and V01p150032.jpg, V01p150033.jpg, hence V01p150034.jpg; similarly V01p150035.jpg (in V01p150036.jpg, V01p150037.jpg) and V01p150038.jpg, in V01p150039.jpg, before V01p150040.jpg (in V01p150041.jpg), V01p150042.jpg and V01p150043.jpg; in the adverbs V01p150044.jpg (also V01p150045.jpg) and V01p150046.jpg and those with the locative ending V01p150047.jpg like V01p150048.jpg and V01p150049.jpg (although not uniformly); in V01p150050.jpg forms (not uniformly, although with more regularity in verbs V01p150051.jpg and V01p150052.jpg) when the last syllable is closed and the next to the last is open, hence V01p150053.jpg, etc.; in forms of the type V01p150054.jpg the accent remains on the penultimate before V01p150055.jpg and (less uniformly) in all forms with an open penultimate.

Penultimate accentuation may also be due to recession (V01p150056.jpg "stepping back"), as in V01p150057.jpg, that is, when a non-pausal accent (see § 4) due on the ultimate precedes a pausal accent (ibid.) due on the penultimate; the non-pausal then recedes to the penultimate (and even farther back in V01p150058.jpg) on the same conditions as the secondary accent if the two words were hyphenated (see below); in point of fact, the non-pausal is intended as a substitute for the secondary accent (see § 4); the rule, however, is not followed consistently (see Jos. Wijnkoop, "Darke ha-Nesiga, sive Leges de Accentus Hebraicæ Linguæ Ascensione," Leyden, 1881; also in Hebrew, V01p150059.jpg, Amsterdam). Finally, penultimate accentuation is due to recession in pause, that is, when the accent is a pausal one, V01p150060.jpg, less often V01p150061.jpg, etc.; in V01p150062.jpg; in verbal forms ending in V01p150063.jpg and V01p150064.jpg, hence V01p150065.jpg, etc.; also in V01p150066.jpg; in forms like V01p150067.jpg (for the non-pausal forms V01p150068.jpg); before the suffix V01p150069.jpg, hence V01p150070.jpg, etc.; in adverbs and participles, for example, V01p150071.jpg. Conversely, the pausal accent may bring about ultimate accentuation as in V01p150072.jpg.

Secondary Accent (V01p150073.jpg).

Properly, the secondary accent is due upon the second syllable from the main accent, provided the intervening syllable is long, that is, open with a long vowel, closed with a short vowel, or opened, that is, originally closed, with a short vowel: V01p150074.jpg. The syllable receiving the secondary accent must also be long (open with a long vowel, opened with a short vowel: V01p150075.jpg; with a closed syllable the sign is implied, but never expressed: V01p150076.jpg). When the syllable preceding the main accent is overlong, that is, closed with a long vowel, the secondary accent will be placed there: V01p150077.jpg (imperative) and similar instances, owing to a retarded pronunciation of ō which is thus raised to å Similarly, the secondary accent will fall upon the syllable preceding the main accent when it is long (open with a long vowel, opened with a short vowel) and the syllable bearing the main accent is a compound one, that is, consists of an ordinary (simple) syllable preceded by a consonant and an incompletely reduced vowel (a V01p150078.jpg), or by a consonant and a completely reduced vowel (a vocal V01p150079.jpg) at the beginning of a word; neither combination is capable of forming a syllable by itself, nor may it be joined in speech to the preceding syllable: V01p150080.jpg. When a word is long enough, another subsidiary accent may become necessary; it is placed at the same distance from the secondary accent as the latter from the main accent, and upon the same conditions (the one to the right being the stronger): V01p150081.jpg, V01p150082.jpg. When the second syllable from the main accent is closed (with a short vowel) and the syllable next preceding is open, the secondary accent is placed upon the latter, the interval between the two accents thus exceeding the limit of one syllable: V01p150083.jpg (observe that V01p150084.jpg pre-fixed never takes a secondary accent).

Distinct from the V01p150085.jpg in the cases just mentioned (also in all forms of the verbs V01p150086.jpg and V01p150087.jpg in which the guttural closes a syllable with a short vowel, for instance, V01p150088.jpg), which the Hebrew grammarians term "light V01p150089.jpg," is the so-called "heavy V01p150090.jpg" which is found, on certain conditions, with closed syllables containing a short vowel (V01p150091.jpg, V01p150092.jpg, etc.), or (in Psalms, Proverbs, Job) with reduced vowels (vocal V01p150093.jpg, V01p150094.jpg, and so on). A third kind which does not concern us here at all is the so-called "euphonicV01p151001.jpg." See Gesenius-Kautzsch, "Hebrew Grammar" (Clarendon Press edition), § 16, 2; Stade, "Hebräische Grammatik," §§ 53-56; both rest upon S. Baer, "Die Methegsetzung," in Merx's "Archiv für Wissenschaftliche Erforschung des A. T.," 1867, pp. 56 et seq.; 1868, pp. 194 et seq., also in Latin in his edition of Proverbs. The accent is often an aid to sense, especially in words similar in sound, but different in meaning, as V01p151002.jpg "he drank," V01p151003.jpg "she put"; V01p151004.jpg "Rachel is coming," V01p151005.jpg "Rachel came." Similarly, the V01p151006.jpg; compare V01p151007.jpg, "they will fear" and V01p151008.jpg "they will see," etc.

Use of Hyphen.

Small words of frequent occurrence, as the mono-syllabic prepositions and conjunctions (V01p151009.jpg, V01p151010.jpg), the words V01p151011.jpg, V01p151012.jpg, also V01p151013.jpg, are, as a rule, joined to the following (long) word. Not only two, but three, and even four, words may be hyphenated thus: V01p151014.jpg, V01p151015.jpg. On the other hand, a long word will occasionally be joined to a following small word: V01p151016.jpg. There is always a close syntactical relation between the hyphenated words. Indeed, in every union of words, sense and rhythm are equal determining factors.

Place of Sentence-Accent.

4. The verse (V01p151018.jpg) is adopted as sense-unit. It is certainly the natural unit in the poetical portions of the Bible in accordance with the Rhythm of Parallelism. It is there equally natural to divide the verse into two halves. Accordingly, in a part of the recently discovered fragments of the Hebrew text of Ecclesiasticus, each verse occupies two short lines (stichs) running across the page; for example:

V01p151019.jpg

The Song of Moses (Deut. xxxii.) is still arranged in this fashion in the Scrolls and in the ordinary editions of the Hebrew Bible.

Determination of Accent.

Elsewhere the Verse-Division is an arbitrary, though convenient, innovation which was not permitted to penetrate into the Scrolls (the sign, marking the end of a verse, must be kept out of them; see Soferim, iii. 7). The whole of the Bible was to be read according to a rhythmical swing which even in the poetical compositions is largely determined by sense. The traditional verse, as a glance at the English Bible will show, does not always coincide with our period; nor is it always of the same length. For purposes of accentuation each verse must be dealt with separately. The problem is invariably: given a verse, determine the accentuation. The leading principle of the system is halving (extended from the poetical portions to the rest of the Bible). Each verse is divided into two parts not necessarily equal; these parts are each divided into two other parts; this process is continued until an indivisible complex of words is reached. The greater pauses are regulated by sense. Frequently, however, the logical pause is sacrificed to rhetorical effect. A characteristic deviation from the accepted method of punctuation consists in passing over introductory clauses or phrases which are treated as a subordinate part of what follows; for example, "And God said, Let there be a firmament in the midst of the waters, and let it divide the waters from the waters" (Gen. i. 6). The lesser pauses obey the laws of syntactical construction, which are obviously various in different languages. The English sentence "And the earth was waste and void" properly reads in Hebrew: "And the earth—it was waste and void"; hence there will be a pause in Hebrew after "and the earth." The order of words differs also. Compare the very opening of the Bible in Hebrew and in English. Rhetorical effect makes itself felt in connection with the smaller no less than in the case of the greater pauses. Thus, for the sake of emphasis, the pause may be shifted from one place to another; or it may be introduced within a group of words which is properly indivisible. In general, greater latitude is permissible in dealing with the slighter pauses. Individual taste will there play an important part. Rhythm is another factor. A group must consist of more than two words to admit of a marked pause within it. When thus the stops have been properly distributed in a verse, our next task is to indicate both the presence and the absence of a pause by the corresponding signs (accents). The accents are either pausal (V01p151020.jpg "stops") or non-pausal (V01p151021.jpg "servants," servi). The notation differs in Job, Proverbs, Psalms (V01p151022.jpg; hence, V01p151023.jpg) from that employed in the other (twenty-one) books (V01p151024.jpg). The two systems must be treated separately, that of the three books first.V01p151017.jpg

A. V01p151025.jpg List of Accents.

Pausal: V01p151026.jpg "cessation"), V01p151027.jpg "ascending and descending"), V01p151028.jpg "rest"), V01p151029.jpg "rhomb"), V01p151030.jpg gereshatum [see below]"), V01p151031.jpg "water-channel"), postpositive, V01p151032.jpg "thrust back"), prepositive, V01p151033.jpg "shake" or "trill"), V01p151034.jpg (great V01p151035.jpg "chain"), V01p151036.jpg and V01p151037.jpg V01p151038.jpg and V01p151039.jpg "by itself"), that is, pausal (for the meaning of V01p151040.jpg and V01p151041.jpg see below).

Non-pausal: V01p151042.jpg "lengthening"), V01p151043.jpg (V01p151044.jpg "laboring, heavy, slow"), V01p151045.jpg "going on," that is, not pausing), V01p151046.jpg "settled," that is, unvarying in its tone), V01p151047.jpg "placed above"), V01p151048.jpg "turned round"; the older form was V01p151049.jpg "wheel"; the older form was V01p151050.jpg (little V01p151051.jpg, V01p151052.jpg), pretonic. The names, it will be observed, are derived from the musical value or from the form of the accents. Other names are met with; but those given are the most common. The diagram printed above will be employed to illustrate the use of the various signs.

Explanatory Notes.

• 1. The proper measure of a poetical verse is two short lines (a distich or couplet). Such is the form of an ordinary verse in Proverbs. The main cesura is then marked by V01p152001.jpg. But frequently, as in Psalms, a verse will contain three short lines (a tristich or triplet; that is, rhythm is sacrificed to sense); or a verse may contain four short lines (a tetrastich or quatrain; that is, two rhythmical verses making one sense-verse); or a verse, not necessarily long, may be trisected purely for reasons of sense or for the sake of oratorical emphasis. The main cesura will then be marked by V01p152002.jpg (a stronger V01p152003.jpg), while V01p152004.jpg will be reserved for the secondary cesura (that is, the one between V01p152005.jpg and V01p152006.jpg). In the diagram the three by no means coordinate sections of the verse are designated by the letters A, B, and C. In a short verse, therefore, drop A, and retain B and C. In a still shorter verse (one consisting of but one short line), drop A and B, and retain C. This principle applies equally to the smaller sections on the diagram (that is, those limited by a pausal accent), the beginning of which may be lopped off to suit varying lengths.There will be found V01p152007.jpg in the sixth word from V01p152008.jpg and farther; it will be replaced by V01p152009.jpg occasionally in the fifth, and almost always in the fourth word; V01p152010.jpg is never used farther to the left V01p152011.jpg is replaced by V01p152012.jpg always in the first, occasionally also in the second, word (see note 19).
• 2. V01p152013.jpg (for V01p152014.jpg, V01p152015.jpg.
• 3. The main cesura in section A is marked by V01p152016.jpg; when a second cesura becomes necessary, V01p152017.jpg is repeated. Observe, in general, that whenever an accent is repeated, the one farther to the left is the weaker. Between V01p152018.jpg and V01p152019.jpg there must be no word (in which case V01p152020.jpg is called little V01p152021.jpg) or at least two words (then we have great V01p152022.jpg). Two V01p152023.jpgs must equally be separated by at least two words. When V01p152024.jpg becomes impossible, V01p152025.jpg takes its place. The shortest measure of section A is two words; a cesura is always required.
• 4. The servus of V01p152026.jpg is V01p152027.jpg (or V01p152028.jpg, when properly the hyphen should be employed; or V01p152029.jpg, that is, with a V01p152030.jpg). This V01p152031.jpg may occur in the same word with V01p152032.jpg (in place of the light V01p152033.jpg).Here V01p152034.jpg, "cutting off," "separating," is a line similar to the one used with V01p152035.jpg and V01p152036.jpg. It occurs (a) before or after the divine name "to prevent its being joined, in the reading, to a word which—in the opinion of the accentuators—it was not seemly to bring into contact with it"; (b) between two words of which the first ends in the same letter with which the second begins; (c) elsewhere, to mark an emphatic intonation. In all these cases, V01p152037.jpg introduces a slight pause after a non-pausal accent.
• 5. In the section limited by great V01p152038.jpg (great V01p152039.jpg's section), the main cesura is marked by V01p152040.jpg (rarely by V01p152041.jpg) and the secondary cesura by V01p152042.jpg. When only one cesura is required, it is marked by V01p152043.jpg (that is, the V01p152044.jpg section is cut off); but V01p152045.jpg is found in exceptional cases, and necessarily, when two servi are introduced (see note 12). Sections of two words may and may not have a cesura. If required, it will, of course, be marked by V01p152046.jpg. The shortest measure is one word.
• 6. Great V01p152047.jpg never has more than one servus, which is V01p152048.jpg (exceptionally V01p152049.jpg; particularly when another V01p152050.jpg precedes). When a pausal accent (V01p152051.jpg or V01p152052.jpg) precedes, it is V01p152053.jpg, but V01p152054.jpg when an open syllable directly (that is, no V01p152055.jpg intervening) precedes the tone-syllable; these accents may appear in the same word with great V01p152056.jpg taking the place of light V01p152057.jpg).
• 7. Sections of two words will occasionally have a cesura; it is omitted in the case of small words standing at the beginning of the section and accented on the first syllable, unless emphasis is desired. The cesura in little V01p152058.jpg's section is marked by V01p152059.jpg. The shortest measure of little V01p152060.jpg's section is one word.
• 8. Little V01p152061.jpg may have two servi, V01p152062.jpg (or V01p152063.jpg); or one servus, V01p152064.jpg. The two servi (V01p152065.jpg) appear occasionally in the same word (when the syllable immediately preceding the tone is open); but this rule is not always obeyed.
• 9. The cesura in V01p152066.jpg's section is marked by the same accent, and is dependent upon the same conditions as the cesura in little V01p152067.jpg's section (see note 7). A secondary cesura is seldom required; the accent marking the main cesura will then be repeated. The shortest measure of V01p152068.jpg's section is one word.
• 10. V01p152069.jpg may have two servi, V01p152070.jpg (i.e., V01p152071.jpg when the tone falls on the second letter and farther; V01p152072.jpg when on the first); or one servus, V01p152073.jpg (it may appear instead of light V01p152074.jpg in the same word with V01p152075.jpg) (or V01p152076.jpg). In a few instances three servi are found: V01p152077.jpg.
• 11. V01p152078.jpg, when a servus precedes; or when the tone falls on the third syllable or farther; in all other cases, V01p152079.jpg (the latter always between V01p152080.jpg and V01p152081.jpg).
• 12. There is no cesura in V01p152082.jpg's section. Its shortest measure is one word. Except in two instances, V01p152083.jpg has never more than one servus, V01p152084.jpg, when the tone is on the first syllable (but in two instances in the place of the hyphen); or on the second when it is simple and the first syllable is a simple closed one without heavy V01p152085.jpg when the condition mentioned in note 6 is fulfilled; V01p152086.jpg in all other cases (but V01p152087.jpg in a few instances where the V01p152088.jpg or V01p152089.jpg preceding the tone-syllable is abnormal). Two servi: V01p152090.jpg in the place of a hyphen.
• 13. The rules for the division of V01p152091.jpg's section are the same as those laid down for great V01p152092.jpg (see note 5).
• 14. V01p152093.jpg has properly only one servus, V01p152094.jpg, when the tone is on the first syllable; V01p152095.jpg when on any other syllable (but V01p152096.jpg; also V01p152097.jpg exceptionally in two places; in one of them two consecutive V01p152098.jpg's are found); always V01p152099.jpg when under a dageshed letter, except in three places, where V01p152100.jpg is found again. Exceptionally two servi are found: V01p152101.jpg; the first is properly in the place of a hyphen; once we find V01p152102.jpg, where again the first is in the place of a hyphen.
• 15. The main cesura in section B is marked by V01p152103.jpg; for a second cesura, V01p152104.jpg will be repeated; and so on. The V01p152105.jpgs may follow each other closely. Properly, between V01p152106.jpg and V01p152107.jpg at least two words should intervene. This must always be the case when V01p152108.jpg marks a subordinate cesura; otherwise an interval of one word is frequently sufficient. When V01p152109.jpg becomes impossible or undesirable, V01p152110.jpg takes its place. The interval between V01p152111.jpg and V01p152112.jpg must never exceed one word. V01p152113.jpg is frequently found in the second word from V01p152114.jpg. It is found in the first only when V01p152115.jpg's word is long; that is, when the tone-syllable therein is preceded by at least two syllables, or by one syllable, provided it is the bearer of a secondary accent (see § 3); otherwise V01p152116.jpg gives way to a servus. The shortest measure of B is two words (except after V01p152117.jpg,when one word is sufficient). Sections of two words may and may not have a cesura.
• 16. V01p153001.jpg should properly never have more than one servus. In all cases where two or more servi are found the servus immediately preceding V01p153002.jpg is a substitute for V01p153003.jpg (see note 15). Three servi: V01p153004.jpg (but V01p153005.jpg, i.e., V01p153006.jpg when the tone is on the third syllable; or on the second syllable when the first syllable is overlong; V01p153007.jpg when the condition mentioned in note 6 is fulfilled; V01p153008.jpg in all other cases). In three instances V01p153009.jpg takes the place of the middle servus; it is preceded by V01p153010.jpg and followed by V01p153011.jpg (when the tone is due on the first syllable) or by V01p153012.jpg (when the usual condition is fulfilled). Two servi: V01p153013.jpg (but V01p153014.jpg). One servus: V01p153015.jpg after V01p153016.jpg (but V01p153017.jpg); V01p153018.jpg in all other cases. More than three servi are found in three instances: in one V01p153019.jpg occupies the second place before V01p153020.jpg; in the others the multiplication of servi is due to the resolution of hyphenated words.
• 17. Theoretically, V01p153022.jpg marks the main, and V01p153023.jpg the secondary cesura in V01p153024.jpg's section; but V01p153025.jpg's section is usually too short to require two cesuras. One expects V01p153026.jpg to be the accent where only one cesura is required. Such is frequently the case. But V01p153027.jpg is employed when the section in front of the cesura must itself be bisected, or when the pausal accent requires two servi before it (in either case V01p153028.jpg is out of the question; see note 12); sometimes (in three instances) for no apparent reason (V01p153029.jpg and V01p153030.jpg are so nearly alike in pausal force that occasionally one is placed for the other). Between V01p153031.jpg and V01p153032.jpg there must be at least one word. Otherwise V01p153033.jpg gives way to a servus. The shortest measure of V01p153034.jpg's section is one word. Sections of two words, of course, have no cesura. The cesura fails likewise in the case of small words standing at the beginning of the section and accented on the first syllable, unless emphasis is desired. The foregoing rules remain in force, even when V01p153035.jpg gives way to a servus (see note 15).
• 18. V01p153036.jpg should properly never have more than one servus: V01p153037.jpg (it may be found, instead of light V01p153038.jpg, in the same word with V01p153039.jpg only when the syllable preceding the tone-syllable is overlong and has α or ō for its vowel). When two servi appear, the one adjoining V01p153040.jpg) is a substitute for V01p153041.jpg (see note 17), while the one farther to the left is V01p153042.jpg's servus (see note 12): V01p153043.jpg. Once three servi are found: V01p153044.jpg; V01p153045.jpg takes the place of a hyphen.
• 19. The main cesura in section C is marked by V01p153046.jpg; the secondary cesura by V01p153047.jpg. When only one cesura is required, V01p153048.jpg should properly mark it. However, V01p153049.jpg is employed (the two accents are presumably regarded as of equal force; see, for a similar substitution, note 17). Between V01p153050.jpg and V01p153051.jpg there must be at least one word. When V01p153052.jpg is due in the word immediately preceding V01p153053.jpg, it is replaced by a servus, V01p153054.jpg. Another servus, V01p153055.jpg, may be placed in the next preceding word. This necessitates a further change: V01p153056.jpg (marking the main cesura), which does not permit V01p153057.jpg immediately after it, and is transformed into V01p153058.jpg. V01p153059.jpg may be found in the word adjoining V01p153060.jpg only when V01p153061.jpg's word is long; otherwise V01p153062.jpg gives way to a servus. This may necessitate a further change: when the word adjoining V01p153063.jpg is itself short (that is, with only one syllable, which is not the bearer of a secondary accent, before the accented syllable); V01p153064.jpg, when due on the next preceding word, is replaced by V01p153065.jpg. The shortest measure of C is one word. But V01p153066.jpg does not permit V01p153067.jpg immediately before it; the latter accent will then be replaced by V01p153068.jpg, the other accents remaining the same as before V01p153069.jpg. Sections of two words have a cesura, provided the last word is of sufficient length to permit V01p153070.jpg before it (see above).
• 20. V01p153071.jpg should properly never have more than one servus. In all cases where two or more servi are found, the servus immediately preceding V01p153072.jpg is a substitute for V01p153073.jpg (see note 19). Three servi: V01p153074.jpg, that is, V01p153075.jpg and V01p153076.jpg upon the same conditions as before V01p153077.jpg (see note 16); where V01p153078.jpg is used before V01p153079.jpg will be employed here. Two servi: V01p153080.jpg; but V01p153081.jpg may take the place of light V01p153082.jpg in the same word with V01p153083.jpg (provided that V01p153084.jpg does not precede; see note 19); in a few places the servi are altogether irregular. One servus: V01p153085.jpg (when the tone is on the first syllable; but V01p153086.jpg), V01p153087.jpg (when on any other syllable), or V01p153088.jpg (after V01p153089.jpg). In a few instances four servi are found.

V01p153021.jpgFirst Four Verses of Pslam cx.

• 21. There is no cesura in V01p153090.jpg's section. Its shortest measure is one word. Except in a few instances, V01p153091.jpg has never more than two servi. Three servi: V01p153092.jpg or V01p153093.jpg according to the usual rule); in three passages: V01p153094.jpg and V01p153095.jpg. Two servi: V01p153096.jpg. One servus: V01p153097.jpg. The servi of V01p153098.jpg are the same as those of V01p153099.jpg.As an illustration of the application of the above diagram and rules to concrete cases, the first four verses of Psalm cx. are given above. The cesuras are indicated as in the diagram; the figures refer to the notes.

The verse from Ecclesiasticus quoted above would be accented as follows:

V01p154001.jpgB. V01p154002.jpg List of Accents.

Pausal: V01p154003.jpg "cluster of grapes"), post-positive, V01p154004.jpg and V01p154005.jpg (great and little V01p154006.jpg "upright"), V01p154007.jpg "handbreadth," or V01p154008.jpg (V01p154009.jpg, "scatterer"), postpositive, V01p154010.jpg "stretching out"), postpositive, V01p154011.jpg (V01p154012.jpg "resting"), prepositive, V01p154013.jpg "broken"), V01p154014.jpg "expulsion"), V01p154015.jpg "double V01p154016.jpg (great V01p154017.jpg (great V01p154018.jpg "drawing out"), V01p154019.jpg—that is V01p154020.jpg).—Non-pausal: V01p154021.jpg, V01p154022.jpg (double V01p154023.jpg "trill"), V01p154024.jpg (little V01p154025.jpg

Explanatory Notes.

• 1. The verse may be of varying length. In a long verse V01p154026.jpg marks the main cesura. The two sections are designated in the diagram by the letters A and B. In a short verse drop A and retain B. The shortest measure of a verse is two words. The cesura never fails.V01p154027.jpg's proper place is in the fifth word from V01p154028.jpg and farther; when due in the fourth and farther to the left, it may be replaced by V01p154030.jpg or V01p154031.jpg (in accordance with the rules laid down for the position of those accents in note 2); the substitution is common in short verses; it necessarily takes place in verses of three or two words; in the first word before V01p154032.jpg (even in a long verse), V01p154033.jpg is rarely used (except in cases of a marked logical pause).
• 2. The main cesura in A is marked by V01p154034.jpg; the second by V01p154035.jpg; for every following cesura V01p154036.jpg is repeated until the last is reached, which is marked by V01p154037.jpg Between V01p154038.jpg and V01p154039.jpg at least three words must intervene; but the proper place is at a considerable distance from V01p154040.jpg. Between V01p154041.jpg and V01p154042.jpg there must be at least one word. When V01p154043.jpg's word and the one adjoining it are both short the distance between V01p154044.jpg and V01p154045.jpg must amount to two words. When V01p154046.jpg becomes impossible V01p154047.jpg takes its place. In a few instances where the two words immediately preceding V01p154048.jpg are very short, that is, mono-syllables, and properly subject to hyphenation, V01p154049.jpg is found in the third word; this is its utmost limit to the right. The shortest measure of A is one word. Sections of two words always have a cesura.
• 3. V01p154050.jpg's servus is V01p154051.jpg (which is repeated in the few instances which call for a second servus; see note 2). In a few instances V01p154052.jpg is found in the same word with V01p154053.jpg; grammarians incorrectly call it a servus (V01p154054.jpg "inclined").
• 4. The main cesura in V01p154055.jpg's section is marked by V01p154056.jpg, which is repeated for every following cesura until the last is reached, which is marked by V01p154057.jpg. Between V01p154058.jpg and V01p154059.jpg there must be at least one word. When V01p154060.jpg's word or the one adjoining is long, V01p154061.jpg is admissible in the second word, but is not necessary. When V01p154062.jpg becomes impossible or undesirable, V01p154063.jpg takes its place. V01p154064.jpg is comparatively rare in the third word; this is its utmost limit to the right. The shortest measure of V01p154065.jpg's section is two words. The cesura never fails. When only one word is available V01p154066.jpg takes the place of V01p154067.jpg.
• 5. Between two V01p154068.jpg's there must be at least three words. When the interval is shorter the one to the left is transformed into V01p154069.jpg; the change does not affect the V01p154070.jpg next to the left, which always maintains its position, there being a sufficiently long interval between it and the V01p154071.jpg preceding it. Between V01p154072.jpg and V01p154073.jpg or V01p154074.jpg there must be at least two words; otherwise V01p154075.jpg is transformed into V01p154076.jpg or V01p154077.jpg. But V01p154078.jpg may precede another V01p154079.jpg; this is the only case in which two V01p154080.jpg's may come together.
• 6. V01p154081.jpg may have one or two servi, both V01p154082.jpg's.
• 7. The main cesura in V01p154083.jpg's section is marked by V01p154084.jpg, which is repeated for every following cesura until a point is reached when V01p154085.jpg is inadmissible or undesirable (see below); then it gives way to V01p154086.jpg; the next cesura is marked by V01p154087.jpg; then comes V01p154088.jpg which may be repeated. Between V01p154089.jpg and V01p154090.jpg there must be at least four words. It is rarely found in the fourth word. It necessarily replaces there V01p154091.jpg when the next cesura is due immediately before V01p154092.jpg then becomes unavailable (see below), and V01p154093.jpg takes its place (that is, V01p154094.jpg's section is obliterated); the interval between V01p154095.jpg and V01p154096.jpg must never exceed one word; otherwise V01p154097.jpg and V01p154098.jpg (the servus due in the second word before V01p154099.jpg; see note 13) would come together, and, on musical grounds, the two accents can not come together without a pausal accent between them. In a few instances V01p154100.jpg takes the place of V01p154101.jpg in the fourth or third word for no apparent reason. Between V01p154102.jpg and V01p154103.jpg there must be at least two words; it is found in the second only when the two next following words are both long; its utmost limit appears to be the fifth or sixth word (where it replaces V01p154104.jpg for the main cesura). When V01p154105.jpg becomes unavailable it gives way to V01p154106.jpg. Between the latter and V01p154107.jpg there need be no interval; its utmost limit is the fourth word. Between V01p154108.jpg and V01p154109.jpg there must be at least one word; it is found in the first only in the place of (V01p154110.jpg (that is, V01p154111.jpg) when the latter sign is due before V01p154112.jpg (strangely enough, the notation remains the same); its utmost limit appears to be the third word. In a section consisting of only three words V01p154113.jpg may take the place of V01p154114.jpg in the second word. The shortest measure of V01p154115.jpg's section is one word. Sections of two words may or may not have a cesura; the cesura is likely to occur when the last word is long, but it is not necessary even then. The cesura may be left out also in sections of three words provided it is due immediately before V01p154116.jpg.V01p154029.jpgIn the twenty-one books V01p154117.jpg is especially employed to mark a stop in long sections limited by V01p154118.jpg, or V01p154119.jpg, for the subdivision of which by means of pausal accents there exists no provision in the accentual system; or to mark a stop immediately before V01p154120.jpg, or V01p154121.jpg neither V01p154122.jpg nor V01p154123.jpg being available (see note 15)
• 8. (V01p155001.jpg may have one or two servi. Two servi: V01p155002.jpg. One servus: V01p155003.jpg. The latter is occasionally found in the same word with V01p155004.jpg, especially in order to indicate a compound word (V01p155005.jpg, Eccl. iv. 10, for example).
• 9. There is no cesura in V01p155006.jpg's section. Its shortest measure is one word. V01p155007.jpg may have from one to six servi, all V01p155008.jpg's. V01p155009.jpg is found in sixteen instances; in every instance V01p155010.jpg might have been used. V01p155011.jpg never stands alone; it may have as many as six servi: V01p155012.jpg etc.
• 10. There is no cesura in V01p155013.jpg's section. Its shortest measure is one word. V01p155014.jpg may have from one to five servi, all V01p155015.jpg's. V01p155016.jpg and V01p155017.jpg are constantly interchanged, particularly where the former is subordinated to V01p155018.jpg (see note 11) or to the servus that takes the place of V01p155019.jpg (see note 15).
• 11. V01p155020.jpg's section should properly be indivisible. But very often a division is introduced. The main cesura is then marked by V01p155021.jpg, and the second by V01p155022.jpg. Between V01p155023.jpg and V01p155024.jpg at least two words should properly intervene; the former is rarely found in the second word. Sometimes, when there are only two words in V01p155025.jpg's section, a cesura is introduced. Similarly, in a few very rare instances, V01p155026.jpg's section is bisected; V01p155027.jpg then marks the cesura. The reason for the phenomena just mentioned is apparently the slight and almost imperceptible difference in pausal force between the three accents: V01p155028.jpg, and V01p155029.jpg. The shortest measure of V01p155030.jpg's section is one word.
• 12. V01p155031.jpg when the accent is on the penultimate, or when V01p155032.jpg precedes; V01p155033.jpg when the accent is on the ultimate, and V01p155034.jpg does not precede.
• 13. V01p155035.jpg may have from one to five servi, but V01p155036.jpg can have only one. Three or more servi: V01p155037.jpg, etc. Two servi: V01p155038.jpg. One servus: V01p155039.jpg (when the accent is on the first letter of the word, this is the only servus V01p155040.jpg can take), or V01p155041.jpg (when on any other letter). V01p155042.jpg may take the place of light V01p155043.jpg in the same word with V01p155044.jpg when no other servus precedes (except when the V01p155045.jpg divides V01p155046.jpg or V01p155047.jpg, or when V01p155048.jpg follows, unless at the same time V01p155049.jpg precedes).
• 14. There is no cesura in V01p155050.jpg's section. Its shortest measure is one word. V01p155051.jpg may have one or two servi: V01p155052.jpg.
• 15. The rules for the division of V01p155053.jpg's, V01p155054.jpg's, and V01p155055.jpg's sections are nearly the same as those governing the division of V01p155056.jpg's section (see note 7). The following differences should be noted: V01p155057.jpg's section is seldom available (only three instances are recorded). V01p155058.jpg may be found in the second word before V01p155059.jpg, etc., though not frequently, even when the two words next following are both short; its utmost limit appears to be the fifth word (where it replaces V01p155060.jpg for the main cesura). In five passages V01p155061.jpg and V01p155062.jpg are found in the same word (second from V01p155063.jpg, etc.); there was evidently a difference of opinion among the accentuators; both accents are now chanted, V01p155064.jpg first. Between V01p155065.jpg and V01p155066.jpg there must be at least one word (but see below); its regular utmost limit is the third word; it is found in the fourth only when the next following cesura is marked by V01p155067.jpg (see above), or when it and V01p155068.jpg change places, as in Gen. i. 12; only in the latter case V01p155069.jpg may be found in the fifth word (see Deut. xvii. 5); V01p155070.jpg and V01p155071.jpg may also change places when the latter accent is due in the third word. When V01p155072.jpg becomes unavailable it gives way to a servus, its own servi remaining; V01p155073.jpg may remain when the last word is long. The section limited by V01p155074.jpg, etc., may contain no more than one word. Sections of two words may and may not have a cesura; a cesura is admissible when the latter of the two words is long and the interval between the tone-syllables considerable; but even then it is rarely introduced; the accent marking the cesura is V01p155075.jpg. The cesura may be left out occasionally also in sections of three words even when it is due at a sufficiently long distance (that is, after the first word of the section) to make V01p155076.jpg available.
• 16. When V01p155077.jpg is due on the first letter of the word and no servus precedes, it is replaced by V01p155078.jpg.
• 17. V01p155079.jpg, etc., may have as many as six servi. Four or more servi: V01p155080.jpg. Three servi: V01p155081.jpg. Two servi: V01p155082.jpg—that is, V01p155083.jpg when on the first letter, and V01p155084.jpg when elsewhere; the two servi may occasionally appear in the same word, the first replacing the light V01p155085.jpg or indicating the end of the first part in a compound word; V01p155086.jpg may take the place of V01p155087.jpg between V01p155088.jpg and V01p155089.jpg when V01p155090.jpg occurs in the latter's word, or when V01p155091.jpg precedes. One servus: before V01p155092.jpg, that is, V01p155093.jpg, when two or more syllables intervene between the servus and V01p155094.jpg at the beginning of a word and furtive V01p155095.jpg counting as syllables; V01p155096.jpg when only one syllable (even an overlong syllable) or none at all intervenes; V01p155097.jpg always remains before V01p155098.jpg, provided no other servus precedes, may replace V01p155099.jpg (in the same word with V01p155100.jpg) when the latter is due in an overlong syllable (immediately before V01p155101.jpg); but not in an open syllable separated from V01p155102.jpg by another open syllable or by an incompletely reduced vowel (V01p155103.jpg); before V01p155104.jpg that is, V01p155105.jpg, when one or more syllables intervene between the servus and the tone-syllable of V01p155106.jpg's word, V01p155107.jpg at the beginning of a word and furtive V01p155108.jpg counting as above; in a few compound words V01p155109.jpg appears in the same word with V01p155110.jpg when no syllable intervenes; V01p155111.jpg always remains before V01p155112.jpg; before V01p155113.jpg.
• 18. V01p155114.jpg when a servus precedes; otherwise V01p155115.jpg is used.
• 19. The rules for the division of V01p155116.jpg's section are the same as those governing the division of V01p155117.jpg's section except that for V01p155118.jpg there is used here V01p155119.jpg. The shortest measure of V01p155120.jpg's section is two words. Sections of two words may or may not have a cesura. The cesura always fails when the second word is short; when it is long a cesura must be introduced, unless the first word is very short, or is a word of frequent occurrence.
• 20. V01p155121.jpg may have one or two servi, both V01p155122.jpg s. V01p155123.jpg may appear in the same word with V01p155124.jpg, provided that no second V01p155125.jpg precedes, in place of light V01p155126.jpg (it must not divide V01p155127.jpg or V01p155128.jpg; see note 13), but not on the first letter; when V01p155129.jpg is inadmissible and the pausal accent preceding is not V01p155130.jpg (called here V01p155131.jpg, or a kind of V01p155132.jpg) is introduced in the place of the heavy V01p155133.jpg; when neither V01p155134.jpg nor V01p155135.jpg is admissible V01p155136.jpg is necessarily employed.
• 21. The rules for the division of V01p155137.jpg's section are the same as those governing the division of V01p155138.jpg's section except that for V01p155139.jpg there is used here V01p155140.jpg. The shortest measure of V01p155141.jpg's section is one word. Sections of two words may or may not have a cesura (a cesura may be introduced only when V01p155142.jpg's word is long).
• 22. V01p155143.jpg has usually only one servus: V01p155144.jpg. It occasionally appears in the place of light V01p155145.jpg, or in compound words, in the same place with V01p155146.jpg. In fourteen instances V01p155147.jpg is preceded by two servi: V01p155148.jpg (V01p155149.jpg is properly a weakened V01p155150.jpg is V01p155151.jpg's servus).
• 23. The rules for the division of section B are the same as those governing A except that V01p155152.jpg is not available here. The shortest measure of B is one word. Sections of two words always have a cesura.
• 24. V01p155153.jpg's servus is V01p155154.jpg. In a few instances V01p155155.jpg is found in the same word with V01p155156.jpg; see note 3).

For the sake of illustration the Second Commandment(Ex. xx 3-6) is here subjoined (according to the V01p156001.jpg; see below):

V01p156002.jpg

The use of a separate system for the three books requires an explanation. Luzzatto (in his "Prolegomeni ad Una Grammatica Ragionata della Lingua Ebraica," pp. 177 et seq.; letter to Baer appended to the latter's treatise, V01p156003.jpg, p. 55) writes that the different method of chanting in vogue for those books called for a different notation. Baer (V01p156004.jpg V01p156005.jpg, p. 3), and before him Elias Levita, believed that the shorter measure of the poetical verses is responsible for the change of the accentual system. Wickes ("Poetical Accentuation," pp. 7 et seq.) seems to combine both views when he says that the system of accentuation found in V01p156006.jpg involves "a refinement of a purely musical character," and that "the idea seems to have been to compensate for the shortness of the verses by a finer and fuller, more artificial and impressive melody." It would seem that Baer's opinion needs but a slight modification to be accepted as an adequate explanation. The accentuation of the three books may be said to be designedly adjusted to the stichic form of the poetical texts (see beginning of this section; also V01p156007.jpg, note 1). In the majority of cases the distich was found to cover the sense-verse. V01p156008.jpg was the natural sign; it is the sign of bisection in a verse in the other books of the Bible. But occasionally the sense required a sense-verse of three stichs. Had V01p156009.jpg been used to mark the main cesura, the rhythmical trisection would have been entirely obliterated. With the introduction of V01p156010.jpg was kept in its place and the rhythmical division left recognizable. Monostichs were not infrequently found in the texts. It was thought desirable to mark them as such accentually by avoiding V01p156011.jpg. The poetical accentuation (the name will now be found appropriate), while primarily serving the requirements of sense, aims at the same time to do justice, as far as it can, to rhythm. It could safely be employed in books like Job, Proverbs, and Psalms, which were not read in public service, and for which therefore no established method of chanting existed (as is the case with Canticles and Lamentations); there was, of course, no room for it in the case of Ps. xviii. and cv. 1-15, which are repeated in II Sam. xxii. and I Chron. xvi. 8-22 in non-poetical surroundings. We subjoin here Ps. xviii. 16 = II Sam. xxii. 17, Heb. 16, which will illustrate the transposition of one system into the other:

V01p156012.jpg

A double accentuation is found in Gen. xxxv. 22 (one is intended for the verse ending at the Masoretic section; the other extends farther so as to slur over the uncomplimentary story concerning the misconduct of Reuben, V01p156013.jpg; or in order to imply the fanciful idea that, in spite of his misconduct, Reuben was still counted with the other sons of Jacob; see Rashi, ad locum, and sources) and in the Decalogue, Ex. xx. 3 et seq. and Deut. v. 7 et seq. (one divides the Decalogue into ordinary verses, neither too long nor too short; the other divides it into ten verses, one for each Commandment). According to the predominance of the lower (V01p156014.jpg) or upper (V01p156015.jpg) signs, one accentuation is spoken of as the "lower" V01p156016.jpg, and the other as the "upper" V01p156017.jpg.

With the superlinear vocalization goes a system of superlinear accentuation. The signs for the pausal Accents differ; some of them represent the actual or modified initial letters of their names; they are placed invariably above the line. The signs for the non-pausal Accents are the same as in the ordinary system, and are infralinear. The system also aims at simplicity. Ambiguous signs are avoided; V01p156018.jpg is used in the place of V01p156019.jpg and V01p156020.jpg which are wanting, also in the place of V01p156021.jpg repeated, and in other cases. There is no separate notation for the three books. Wickes ("Prose Accents," pp. 142 et seq.) proves conclusively that the superlinear system is derived from the ordinary one. Facsimiles may be found in Ginsburg's "XV. Facsimiles of Manuscripts of the Hebrew Bible," plate ii., London, 1897, and in Stade's "Hebräische Grammatik"; see also the reproduction in Baer's edition of Job. Compare also the literature quoted in article Vocalization.

Accentuation Supposed to be of Divine Origin.

5. The general belief of the Jews in the Middle Ages was that both the vocalization and accentuation originated with Ezra and the mythical Great Synagogue. Thus Ben Asher (V01p156022.jpg V01p156023.jpg, § 16 and elsewhere) speaks of the Accents as introduced by the prophets and princes of the diaspora (the exiled Jews in Babylon), to whom the interpretation of every word (Scriptural passage) was revealed; the accentuation which bears the seal of the prophets is therefore inspired. Some even maintained that the Torah Pentateuch which Moses received on Sinai and delivered to Israel was furnished with vowel-points and accent-signs, both of which were indeed as old as the alphabet and the language (communicated to Adam in paradise). The Sinaitic origin of the punctuation was emphatically denied by Mar Naṭronai II. (859-869), who accordingly prohibited its introduction into the Scrolls (see "Maḥzor Vitry," p. 91, Berlin, 1893, and Grätz, "Gesch. der Juden," 2d ed., v. 503).

Ben Asher's opinion of the sacredness of the Accents was shared by the contemporaries of Saadia (892-942). This gaon was accused by his detractors of ascribing to himself the gift of prophecy because he had written a treatise in Biblical style with vowel-points and Accents. In his defense Saadia pointed to extracanonical writings (such as Sirach, Scroll of the Hasmoneans, and others) which were pointed and accented. While Saadia evidently does not assign to the accentuation special sacredness, he is nevertheless far from suspecting its recent origin; for, speaking of Sirach's book, he says that he (Sirach) furnished it with points and Accents (wj'alahu musammanan mut'aman). See Saadia's V01p156024.jpg, ed. Harkavy, St. Petersburg, 1891, V01p156025.jpg; also V01p156026.jpg, note 2; V01p156027.jpg, note^*. The recently found fragments of Sirach have traces of points and Accents (see "Rev. Ét. Juives," xl. i. et seq.); on a text of the Scroll of the Hasmoneans with points and Accents (among the Cambridge manuscripts brought from Cairo), seeAbrahams, "Jewish Quarterly Review," 1899, xi. 291 et seq.

Post-Talmudic Origin.

6. The accentuation, like the vocalization, is certainly a post-Talmudic innovation. The treatise Soferim, in which for the first time reference is made to points marking the beginning (or, as it may be called, the end) of a verse (iii. 7), and possibly also to signs (points) by which the subdivisions of a verse are indicated, is post-Talmudic. V01p157001.jpg (Soferim, iii. 7) apparently means "to cut up a verse"; compare Meg. 22a: V01p157002.jpg, "I was not permitted to break up a verse"; in xiii. 1, reference is made to the stichic form of the texts of Psalms, Job, and Proverbs in which a verse (that is, a long verse) is said to be broken up into three parts by a blank left after the opening portion (V01p157003.jpg; corresponding to V01p157004.jpg's section), at the V01p157005.jpg (this is apparently the correct reading; see the edition of Müller, Leipsic, 1878) and at the end (V01p157006.jpg). Observe that the terminology is far from fixed. In the Talmud itself reference is made to the practise of reciting the text in a manner according with the logical pauses (Meg. 3a = Ned. 37b; Ḥag. 6b; in Ber. 62 mention is made of a system of hand movements used by teachers in training their pupils to pause in the proper places), and apparently also to the habit of chanting (Meg. 32a), but not to written signs by which pauses are marked. The beginnings of our system of accentuation may therefore safely be placed in the sixth century. The first to prove the post-Talmudic date of the points and Accents was Elias Levita (V01p157007.jpg 1538). See Vocalization.

7. One is led to the same conclusion by an examination of the Syriac system of accentuation introduced at the end of the fifth century by the grammarian Joseph Huzaya (Wright, "A Short History of Syriac Literature," pp. 115 et seq., London, 1894), to which the Hebrew system bears a striking resemblance and from which it is apparently derived. The Syrians, apt disciples of the Greeks, adopted from the latter their method of reading, and accordingly also their system of punctuation. The Greeks distinguished three kinds of reading (ἀνάγνωσις): oratorical or dramatic delivery implying declamation and gesticulation (καθ ὑπόκρισιν); reading in accordance with the tone, that is, word-accent (κατἁ προσῳδίαν), and reading in accordance with pauses required by the sense (κατἁ διαστολήν). A single point (στιγμή), placed above or below or in the middle of the line, indicated the pauses; the upper point (τελεία στιγμή) at the end of a period complete in itself (αὐτοτελής), the lower point (ὑποστιγμή) between protasis and apodosis, and the middle point (μέση στιγμή) in a long sentence in order to permit the reader to take breath. Upon this modest system, which is found in our oldest Syriac manuscripts, Huzaya founded a more elaborate one to mark the subordinate divisions in a more regular and careful manner. The following diagram will illustrate the system (A means protasis, and B apodosis):

V01p157008.jpg

Compare with this the Hebrew (prose) system in its essential parts:

V01p157009.jpg

The point employed at the end the Syrians call påsoḳå, that is, "sector"; V01p157010.jpg (corrupted into V01p157011.jpg) was apparently the name which in the Hebrew system belongs to the double point(:) marking the end of a verse. The Greeks also had a sign called ὑφέν (from which our "hyphen" is derived) to mark the coalescing of two syllables into one (synalepha). The Syrians employed the same sign to join together two Syriac words used in translation of one Greek word; hence the Hebrew hyphen (see § 3). In the Hebrew system the rhetorical Accents (they were the signs of interrogation, exclamation, etc.) are wanting. However, in distributing the pauses the Jewish accentuators frequently pay attention to the requirements of rhetorical declamation (see the quotation from the "Manuel du Lecteur," in Merx, p. 69, note 2; also Ḳalonymus ben David at the end of the Hebrew grammar of Abraham de Balmes, Venice, 1523). See Merx, "Historia Artis Grammaticæ apud Syros," pp. 62 et seq., Leipsic, 1889. On the origin (and function) of the minor pausal Accents see Büchler, "Untersuchungen zur Entstchung und Entwickelung der Hebräischen Accente," Vienna, 1891 (see also Grätz, "Monatsschrift," 1882, pp. 385-409).

8. It is doubtful whether the vocalization and accentuation were introduced simultaneously. Perhaps the latter followed the former. Both became an object of care to the Masoretes, who, in addition to the task of preserving the traditional consonantal text intact, undertook to watch over the traditional vowel-points and accent-signs. Compare, for example, the Masoretic note to Jer. i. 7: V01p157012.jpg V01p157013.jpg, that is, the words V01p157014.jpg occur four times (i. 7, iii. 11, xi. 6, xv. 1; contrast iii. 6 and xi. 9) in Jeremiah thus accented. On the accentual variations of the Orientals (V01p157015.jpg) and Occidentals (V01p157016.jpg) see Masorah. Even more minute are the differences between Ben Asher and Ben Naphtali. Our editions usually follow the former, who is the authority of the West.

Value in Bible Interpretation.

9. The accentuation offers an invaluable aid to the understanding of the Biblical text. One must, however, constantly bear in mind its limitations, which are of a twofold character. On the one hand, in attempting to accomplish too much, the system fails in important points. In short verses its pauses are unnecessary; in long verses there are not enough of them. Sense is not infrequently sacrificed to rhetorical effect. The imperfection of the system is particularly noticeable in the awkwardness with which a parenthesis is indicated (compare, for example, Jer. xx. 1). Nor is it always easy to tell just what the accentuators had in mind in choosing a certain mode of accentuation. While, for the finer points of Biblical exegesis, a knowledge of the Accents is indispensable, the beginner in the study of the Bible should not be burdened with learning more than a few of the important pausals, which are quite sufficient for ordinary purposes. On the other hand, the accentuation represents the interpretation current in the Jewish schools at a comparatively late period. While, on the whole, the accentuation endeavors to be true to the natural meaning (peshaṭ; which see) of the Biblical documents, it does not altogether keep itself free from dogmatic prejudices (see I Sam. iii. 3), which it indeed shares with the ancient versions. At best the accentuation is representative of traditional Jewish exegesis, which the student of the Bible is frequently forced to overrule. The rule laid down by Abraham ibn Ezra: V01p157017.jpg V01p157018.jpg "no interpretation of a Biblical passage which does not follow the accentuation should be accepted," was sinned against by every Jewish commentator of importance, including Ibn Ezrahimself. It should, of course, be remembered that the deviations from the accentual interpretation which are met in rabbinical commentaries were not always conscious transgressions. The minutiæ of the accentuation were not always present to the mind of the commentators. But there are cases where the Accents are avowedly disregarded (see Ḳimḥi on Hosea, xii. 12: V01p158001.jpg "in interpreting Scripture we are not always bound by the accents"; see also Luzzatto, "Prolegomeni," pp. 187 et seq.).

In Isa. xl. 3 there is a famous case where the accentuation (V01p158002.jpg) is unquestionably right. Accordingly the Revised Version (text) translates: "The voice of one that crieth, 'In the wilderness,'" etc. The quotation of the verse in Mark, i. 3 connects "in the wilderness" with "the voice of one crying" (implying the accentuation V01p158003.jpg). The New Testament accentuation (hardly invented for the occasion; the punctuation in the Septuagint is due to New Testament influence) is probably nothing more than a haggadic interpretation of the kind so often met with in midrashic works. A puzzling accentuation which goes with the rendering of the Septuagint and Vulgate may be found in Isa. vii. 3: V01p158004.jpg (et qui derelictus est, Iasub filius tuus; see Baer's edition, "Additamenta," p. 67).

The Accents in the ordinary editions of the Bible are frequently unreliable. Baer's and Ginsburg's Bible editions (where also important variants are noted) are indispensable to one interested in Biblical accentuation.

M. L. M.