SELAH (Hebrew, ):
Term of uncertain etymology and grammatical form and of doubtful meaning. It occurs seventy-one times in thirty-nine of the Psalms, and three times in Hab. iii. It is placed at the end of Ps. iii., ix., xxiv., xlvi., and in most other cases at the end of a verse, the exceptions being Ps. lv. 20, lvii. 4, and Hab. iii. 3, 9. Of the psalms in which it is found, twenty-three belong to the group in which "Elohim" is used to designate God; twenty-eight to that called by Briggs the "director's (="choir-leader"; See Psalms, Critical View) copies"; and twenty to the "Davidie" collection. Again, nine of the twelve Korahite and seven (LXX. eight, including lxxx. 8) of the twelve Asaph psalms have the term. Three psalms with "Selah" are headed "Miktam"; seven, "Maskil"; ten, "Shir"; twenty-six, "Mizmor"; while Habakkuk iii. is superscribed "Tefillah."Technical Term.
That the real significance of this curious term (or combination of letters) was not known even by the ancient versions is evidenced by the variety of renderings given to it. The Septuagint, Symmachus, and Theodotion translate διάψαλμα—a word as enigmatical in Greek as is "Selah" in Hebrew. The Hexapla simply transliterates σελ. Aquila, Jerome, and the Targum give it the value of "always" (Aquila, ἀεί; Jerome, "semper"; Targum, for the most part = "in secula" or = "semper"). Theodotion in Ps. ix. 17 has the translation ἀεί; the Quinta gives εἰς τοὺς αἰόνας (); and the Sexta, διαπντός, (in Ps. xx. 4, εἰς τέλος). Jacob of Edessa, quoted by Bar Hebræus (on Ps. x. 1), notices that instead of διάψαλμα some copies present αεί=; and he explains this as, referring to the custom of the people of reciting a doxology at the end of paragraphs of the liturgical psalms. In five passages (see Field, Hexapla on Ps. xxxviii. [Hebr, xxxix.] 12) Aquila offers, according to the Hexaplar Syriac, ="song," the ἀσμα by which, Origen reports Aquila to have replaced the διάψαλμα of the Septuagint. According to Hippolytus (De Lagarde, "Novæ Psalterii Græci Editionis Specimen" 10), the Greek term διάψαλμα signified a change in rhythm or melody at the places marked by the term, or a change in thought and theme. Against this explanation Baethgen ("Psalmen," p. xv., 1st ed. Göttingen, 1892) urges the circumstancethat the enigmatical expression occurs also at the end of psalms. The cogency of this objection would hold if the mark had been inserted by the original writer and not, as is most probable, by a later editor who may have expected the Psalms to be recited in succession without reference to the divisions in the Masoretic text; or if it were an indubitable fact that where in the Hebrew a psalm now ends it ended in the original. Augustin (on Ps. iv. 3) regards διάψαμα as indicating that what follows is not to be joined to the preceding. He suggests also the possibility that the Hebrew "Selah" meant "Fiat" = "Let there be [made]." The Masoretic accentuation always connects "Selah" with the preceding, as though it were part of the text or thought, most likely because it was held to mean "forever." In fact, the vowel-points in seem to indicate a "ḳere" (with "ḳameẓ" on account of ת)= "forever" (see B. Jacob in Stade's "Zeitschrift," xvi.  129 et seq.).Modern Views.
Nor is there greater unanimity among modern scholars than among the ancient versions. Only on one point there agreement, namely, that "Selah" has no grammatical connection with the text. It is either a liturgico-musical mark or a sign of another character with a bearing on the reading or the verbal form of the text. As thirty-one of the thirty-nine psalms with the caption "To the choir-master " present "Selah," the musical value of the mark has been regarded as well assured. In keeping with this it has been assigned to the root , as an imperative that should properly have been vocalized , "Sollah" (Ewald, "Kritische Grammatik der Hebräischen Sprache,"p. 554; König, "Historisch-Kritisches Lehrgebäude der Hebräischen Sprache," ii., part i., p. 539). The meaning of this imperative is given as "Lift up," equivalent to "loud" or "fortissimo," a direction to the accompanying musicians to break in at the place marked with crash of cymbals and blare of trumpets, the orchestra playing an interlude while the singers' voices were hushed. The effect, as far as the singer was concerned, was to mark a pause. This significance, too, has been read into the expression or sign, "Selah" being held to be a variant of "shelah" (="pause"). But as the interchange of "shin" and "samek" is not usual in Biblical Hebrew, and as the meaning "pause" is clearly inapplicable in the middle of a verse or where a pause would interrupt the sequence of thought, this proposition has met with little favor. Neither has that which proposes to treat it as a loan-word from the Greek ψάλλε = "strike the harp," etc.
Grätz ("Kritischer Commentar zu den Psalmen," i. 93 et seq.) argues that "Selah" introduces a new paragraph as it were, a transition in thought, and also in some instances a quotation (e.g., Ps. lvii. 8 et seq. from cviii. 2 et seq.). The fact that the term occurs four times at the end of a psalm would not weigh against this theory. As stated above, the Psalms were meant to be read in sequence, and, moreover, many of them are fragments; indeed, Ps. ix. is reckoned one with Ps. x. in the Septuagint, which omits διάψαλμα also at the end of Ps. iii., xxiv., and xlvi. B. Jacob (l.c.) concludes (1) that since no etymological explanation is possible, "Selah" signifies a pause in or for the Temple song; and (2) that its meaning was concealed lest the Temple privileges should be obtained by the synagogues or perhaps even by the churches.More Liturgical than Musical.
Another series of explanations is grounded on the assumption that its signification is liturgical rather than musical. It marks the place, and is an appeal, for the bystanders to join in with a eulogistic response. Briggs ("Jour. Bib. Lit." 1899, p. 142) accepts the etymology and grammatical explanation given above, i.e., that "Selah" is a cohortative imperative, meaning "Lift up [your benediction]," the eulogy with which psalms or sections of psalms were concluded. One would expect the imperative to be in the plural if the address was to more than one bystander. However, Briggs' explanation indicates the line along which the mystery connected with this term or combination of consonants is to be removed. It has been suspected that "Selah" is an artificial word formed from initials.Probably a Contracted Form.
That is probably the case, though the resolution of the initials usually suggested, (= "Return to the beginning, O singer"), has to be abandoned. The renderings in the versions, "'olmin," άεί, and the like (= "forever"), if they do not prove that is a corruption for —the word "'olam" standing for the first noun in the benediction—create a strong presumption that the initials of the verse in which "'olam" occurs are hidden in the puzzling word "Selah." Grätz (l.c.) shows that in Ps. lv. 20 is a corruption for (or even for ), meaning "destroy"; and a similar corruption of the first and third consonants throughout has contrived to make "Selah" the "crux interpretum." If in some instances or (= "destroy") be read and in others , the enigma disappears. "K l ḥ" represents the eulogy "Ki le'olam ḥasdo" (), hence the or ἀεὶ of the versions—a eulogy which is familiar and which is found as such in the Psalms (Ps. c. 5, cvi. 1, cvii. 1, cxviii. 1 et seq.; especially cxxxvi.; also I (Chron. xvi. 34, 41; II Chron. v. 13, xx. 21). This is confirmed by the fact that just such phrases as , and perhaps , actually do occur in passages where "Selah" might stand equally well and with as little bearing on the context (Ps. lii. 11, 12). In Ps. xxxiv. 11 at the end is certainly superfluous; but it stands where one would expect this very term ; and, therefore, it is not too bold a conjecture to read here in the sense of a technical abbreviation of the eulogy. In this connection the midrash on Ps. cxviii. is of importance; quoting Isaiah iii. 10, it commands that after the mention of the righteous the words should be added, but that after reference to an evil-doer a curse should be pronounced.
The latter injunction throws light on many passages in which "Selah" has another sense than that noted above, and in which it should be read or (= "Destroy them"), as one word. It is noticeable that the term occurs frequently after areference to evil-doers (Ps. iv. 3; vii. 6; ix. 21; xxxii. 5; xlix. 14 [xlix. 16 ?]; lii. 5; liv. 5; lvii. 4, 7; lix. 6; lxii. 5; lxvi. 7; lxxxii. 2; lxxxviii. 8; lxxxix. 46, 49; cxl. 6; Hab. iii. [A. V. ii.] 13); and at the mention of these the bystanders break forth into malediction, as they do into benediction at the mention of God's wonderful deeds. Their comment on the recital is "Destroy them," "Make an end of them," or "of the evils," i.e., "Forego" (as in Ps. lxxxviii. 8). "Selah" is thus identical with as twice repeated in Ps. lix. 14(Hebr.), "Destroy in anger; destroy that they be no more." This very verse ends with "Selah," which, as explained above, is a repetition (but in the mouths of the bystanders) of the passionate outcry (="Destroy").Sometimes Meaning "Delete."
Some few passages remain in which seems to fit in neither as a eulogy—i.e., as a corruption of or as an artificial combination of initials making —nor as an imprecation. But even in these the reading (="Destroy") suggests itself, not indeed as a liturgical response, but as a note to indicate that something in the text should be deleted. This seems to be the case in Ps. lv. 8 (R. V. 7), where verses 8 and 9 virtually conflict; for the desert is the place where storms blow. "Selah" here has the appearance of a sign that the verse, being a quotation from somewhere else and really not belonging to the psalm, should be omitted. The same holds good in Ps. lxxxi. 8, where the third member of the verse is clearly it marginal note explanatory of the preceding. "Selah" after , "at the waters of Meribah, "indicates this fact, and means (="Delete"). Another instance of this is Ps. lx. 6, where the words break the connection between verses 6a and 7, and really make no sense. In Hab. iii. (ii.) 3, 9, also, "Selah" points to some defect in the text.
Perhaps the latter use of the term will throw light on the origin of the Greek διάψαλμα. It may be connected with the verb διαψάω = "to rub away thoroughly," "to erase." At all events some of the versions point to a reading in which was visible, e.g., διαπαντός (Sexta), while the translation of Aquila according to the Hexaplar Syriac, , meaning "responsive, antiphonal song," corroborates the assumption that the benediction or malediction was marked as anticipated in the passage.
"Selah" occurs also in the text of the Shemoneh 'Esreh. This fact shows that at the time when the text of this prayer was finally fixed, the term had become a familiar one; and as the "Shemoneh 'Esreh" draws its vocabulary largely from the Psalms, the appearance of "Selah" in the prayer is not strange. In the Talmud that word is treated as a synonym of "neẓaḥ" and "wa'ed," all three signifying eternal continuance without interruption ('Er. 54a, ). Ḳimḥi connects the term with the verb (= "lift up"), and applies it to the voice, which should be lifted up, or become louder at the places marked by it (commentary on Ps. iii. 2). Ibn Ezra (on Ps. iii. 2) regards it as an equivalent of or , an affirmative corroborative expletive.