Manuscripts

All the reports certainly originate from different scribes, since the scribal hands are different in all of them. The four oracle collections, however, were all compiled by the same scribe. This can be established with certainty from an analysis of the sign forms and other scribal idiosyncracies occurring in these tablets, as contrasted with the other tablets of the corpus (see Table l ).^[[280]]

TABLE I. Sign forms ocurring in nios. 1, 2, 3, 7 and 9.

No. l (K 4310) is a three-column tablet measuring 28 mm (maximum thickness) x 118 mm (full width) x 155 mm (extant height). The upper part of the tablet is broken off so the beginnings of all the columns on the obverse and the ends of all the columns on the reverse have been lost. In addition, a small piece has broken away from the lower left corner of the tablet. The columns on the obverse (from left to right) measure 32, 42 and 35 mm in width, those on reverse 33, 42 and 35 mm. The space between columns is 2 and 2.5 mm on the obverse, 2-3 and 2 mm on the reverse. The portion of even thickness in the middle of the tablet measures 45 mm, the portion of decreasing thickness from there to the bottom of the tablet 65 mm. Assuming an identical curvature for the upper part of the tablet, this implies an original tablet height of 175 mm (= 2 x 65 mm + 45 mm) and a ratio of 2:3 between the horizontal and vertical axis of the tablet, as in nos. 2 and 3.

Vertical script density is 10 signs to 35 mm in cols. I and V, 10 signs to 38 mm in cols. II, III and VI, and 10 signs to 32 mm in col. IV. Allowing for an uninscribed space of 5 mm at the top of the obverse, as on the reverse, the amount of text lost on the obverse can hence be estimated as six lines at the beginning of cols. I and II, and as 14 lines at the beginning of col. III; in addition, another four lines have been lost due to surface damage at the beginning of col. I. Correspondingly, about 14 + 8 + 8 lines have been lost at the ends of cols. IV, V and VI. The lower portion of the last column is uninscribed, but the break at the end may have contained a short colophon and a date (see p. LIii above). The total number of lines lost in the breaks, excluding the colophon, is hence 42. Adding this to the total of extant lines (224), the original line total of the tablet can be established as 266, of which 84.2% are extant. Col. I originally contained 49 lines; col. II and III, 46 lines; col. IV, 49 lines; col. V, 44 lines; and col. VI, 32 lines of text ( + possibly a colophon and a date).

The individual oracles on the tablet and the authorship indications following them are separated from each other by horizontal rulings. The ruling before oracle 1.1 is a double one, implying that the text before it differed in nature from the rest of the tablet and hence probably did not contain a prophetic oracle. A similar double ruling occurs in no. 3, col. II, where it separates oracle 3.4 from a six-line section of ritual instructions attached to oracle 3.3.

The available data are not sufficient to determine the nature of the introductory passage preceding the double ruling in col. I with any certainty. All that can be said is that it evidently fell into two parts, a ten-line introductory section entirely lost at the beginning of col. I, and a three- line postscript separated from it by a horizontal ruling and closed by the double ruling just discussed. The text remaining from the postscript shows that it cannot have contained an authorship indication, and it is unlikely to have contained ritual instructions. It is possible that the introductory section was preceded by a short 1 -2 line heading, see below, p. LXIII with n. 285.

Note that the partially preserved ruling before col. I 1', which at first glance looks like a double one, almost certainly was a single ruling with an accidon- tally bifurcated tail. A similar single ruling with a 0.5 mm wide double tail occurs in col. III (between lines 6' and 7'). In the double rulings preceding oracles 1.1 and 3.4 the space between the rulings is much wider (1.5 mm in the former and 2 mm in the latter).

There are two horizontal impressions of the stylus ten lines apart in the space between cols. VI and V of the reverse, at the beginning of lines v 12 and v 21. The former is accompanied by a superscript winkelhaken ("ten" mark), which may indicate that these impressions were meant to plan the distribution of the text on the tablet. Similar marks are not found elsewhere in corpus.

Apart from the breaks, the tablet is in an excellent state of preservation. The script is clear and for the most part easily legible despite the three-column format, which occasionally caused the scribe problems of space and forced him to exceed the right-hand column margin, especially in col. I (see lines i 6', 7', 9', 1 0', 1 4', 28' and 32', and ii 13', 20' and 23'-26'). The handwriting is elegant and experienced, though it betrays traces of haste; erasures and scribal errors occur in i 17', 20', 28', and ii 6', 8', 34' and 39'. It is possible that some of the scribal mistakes derive from textual damage in the originals used. This is suggested by the unusually spelling na-kar^[ar]-ka "your enemy" in ii 8', which looks like a misreading for na-kar-u-ti-ka "your enemies"; note that ar cannot be interpreted as a phonetic complement here, since kar was not a polyphonic sign and thus did not need any reading specification. It is interesting that though the scribe evidently has applied his own conventions in the copying process, isolated traces of the orthography and ductus of the originals shine through here and there. ^[[281]] For evidence indicating that the scribe who inscribed the tablet also wrote Collections 2 and 3, see p. L V above with n. 280. On the probable date of the text (late 673), see p. LXIX.

No. 2 (K 1 2033 + 82-5-22,527) is a two-column tablet measuring 88 mm x 146 mm. One face of the tablet is flat, the other slightly convex; the flat face is the obverse, as in K 240 1 (no. 3). The entire left side of the tablet and the beginnings and ends of all the columns are missing. The right-hand column of the obverse measures 51 mm in width; the original width of the left-hand column can be determined to have been 54 mm on the basis of the restorations in lines l 1 '-22', which are certain. As the space between the columns is 3 mm, the original width of the tablet can be reconstructed as 108mm. On the reverse, the right-hand column (col. III) measures 54 mm, and as the space between the columns here is also 3 mm, the original width of the left-hand column must have been 51 mm.

Assuming that the ratio between the horizontal and vertical axes of the tablet was 2:3, as usual in three-column tablets, the original height of the tablet can be estimated to have been 162 mm.

Vertical script density is 10 signs to 38 mm in cols. 1-111 (in col. II partl 10 signs to 36 mm) and 10 signs to 40 mm in col. IV. Judging from the curvatures, very little text (about two lines only) has been lost at the bottom of the tablet. The amount of text lost at the top can be estimated as 4 to 6 lines depending on the column. Taking into consideration the lines lost in the breaks, cols. I and II originally contained 45 lines, col. II, 43 lines, and col. IV about 40 lines of text. The original line total of the tablet thus was about 173, of which 145 (= 83.8%) are extant.

As in nos. 1 and 3, the individual oracles are separated from each other by horizontal rulings, but in contradistinction to no. 1, the authorship indications are not correspondingly separated from the oracles. Instead, a blank space is left between the oracle and the authorship indication in 2.1. This space does not occur in other oracles of the tablet.

The scribe is the same as in nos. 1 and 3 (see p. LV). The script is clear but in several places (especially near the breaks) badly damaged and effaced, and therefore at times harder to read than in no. 1. In col. III, recent brushing and scratching has resulted in making the sixth sign in line 8' illegible beyond remedy. Scribal mistakes occur in lines ii 9' and 34', and text is occasionally continued over the column margin as in no. 1, see col. iv l ', 4', 1 3', 2 1' and 27'. On the probable date of the tablet (679 BC) see p. LXIX.

No. 3 (K 240 1) is a two-column tablet measuring 26 mm x 75 mm x 139 mm. As in no. 2, the obverse is flat, the reverse slightly convex. The left side and the lower edge of the tablet are missing; in addition, two small pieces have broken off from the middle of cols. I and III. The right-hand column of the obverse measures 42 mm in width; the space between the columns is 3.5 mm; the original width of the left-hand column can be determined to have been 46 mm on the basis of the restorations in lines 9-13 and 27-34. The original width of the tablet thus was 92 mm. On the reverse, the space between the columns is narrower (2 mm), and the right-hand column a little wider (44 mm) than on the obverse. The left-hand column probably had the same width as that on the obverse (46 mm).

Applied to the reconstructed tablet width (92 mm), the 2:3 axis ratio of two-column tablets yields 138 mm as the original height of the tablet. In actual fact, the tablet measures 139 mm in height, implying that the extant height is very close to the original one. This is confirmed by the curvatures which indicate that very little, possibly only the surface of the edge, is missing at the bottom of the tablet.

Vertical script density is 10 signs to 37 mm in cols. 1-111 and 10 signs to 35 mm in col. IV. The tablet originally contained a total of 1 45 lines, of which 139 (= 95.2%) are extant, many of them unfortunately only in part. Each column was originally inscribed with about 37 lines of text (cols. I and III = 37 lines; col. II = 36 lines; col. IV = 35 lines + blank space of two lines).

The individual oracles are separated from each other by horizontal rulings as in nos. 1 and 2; a double ruling separates the first three (coronation) oracles, accompanied by ritual instructions, from the rest of the tablet. In contradistinction to no. 1, but in keeping with no. 2, rulings are not used to separate the ritual instructions from the oracles, and a blank space is inserted before the authorship indication at the end of the tablet. This, as well as the two-column format of no. 2 and 3, indicates that no. 3 is temporally close to no. 2, while no. 1 was written at a different time, as is also implied by the analysis on p. LXVIIIff, which suggests that the tablets date from 680, 679 and 673 respectively.

The scribe is the same as in nos. l and 3 (see p. LV). The script is very clear and easily legible throughout.

No. 4 (83- 1-1 8,839) is a fragment from the surface of a clay tablet measuring 33 mm (width) by 39 mm (height). The surface of the fragment is entirely flat, which indicates that a multi-column tablet is in question, and the estimated column width (50 mm or slightly less) points to a two-column tablet. Compare the column widths in nos. 2 (51 to 54 mm) and 3 (42 to 46 mm), both two-column tablets, against those in no. 1 (32 to 42 mm), a three-column tablet. The vertical script density is the same as in nos. 2 and 3 (10 signs to 37-38 mm).

The available data are not sufficient to determine the original size of the tablet, but it is likely to have been close to that of no. 2 or 3 and hence probably contained between 170 and 150 lines of text. If the format of the tablet was the same as in nos. 2 and 3, the fragment belongs to the obverse, most probably to the beginning of column II. Since only one line seems to be missing from the beginning, it is unlikely to have contained the very first oracle of the tablet.

The script is clear and easily legible, and the scribal hand agrees with that of nos. 1-3. By its content, the fragment parallels nos. 1-2 and hence a date of composition ea. 680 BC is probable.

No. 5 (K 6259) is the left half of a horizontal tablet measuring 25 mm (thickness) x 60 mm (extant width) x 52 mm (full height). The original tablet width can be established as 104 mm on the basis of the 2: 1 ratio between the horizontal and vertical axes, which is constant in this type of tablet. Accord- ingly, more than 40% of text has been lost in each line.

The writing is big; signs measure 5 mm in height on the average, and vertical script density is 10 signs to 60 mm on the obverse and 10 signs to 73 mm on the reverse. Nevertheless, the tablet is in places very difficult to read; many signs, especially on the reverse, are badly obliterated or scratched beyond recognition. The scribal hand does not is not found on other tablets of the corpus.

No. 6 (Bu 91-5-9, 106+ 109) is a fragment from the left side of a horizontal tablet measuring 26 mm (thickness) x 43 mm (extant width) x 64 mm (height). The original width (twice the height) was 128 mm. Accordingly, more than two thirds of each line has been lost.

The script is clear and easy to read despite occasional surface damage. Vertical script density is 10 signs to 65 mm. The scribal hand is not the same as in no. 5 or other tablets of the corpus.

No. 7 (K 883) is an almost complete horizontal tablet measuring 22 mm (thickness) x 82 mm (width) x 41 mm (height). A small chip of 25 mm diameter has broken off the lower left-hand corner of the obverse, and there is minor surface damage in obv. 12 and rev. 1-2 and 11.

Vertical script density is l0 lines to 35 mm. The script is clear and beautiful, and in contrast to nos. 5-6, the text is elegantly distributed over the lines. It is almost certainly an archival copy of a more hastily prepared original, which it seems to have reproduced faithfully, judging from the distinctive features it shares with nos. 5-6 and 8, like the quotation particle ma-a (cf. above, p. LV). The scribal hand is not found on the other tablets of the corpus.

No. 8 (K 1545) is a horizontal tablet pieced together from two fragments; it measures 15 x 62 x 31 mm. A 1-2 cm wide triangular piece is missing from the middle, but otherwise the tablet is complete.

Vertical script density is 10 lines to 40 mm; script is clear and easily legible. Even though the available data are admittedly very limited, it is possible that the tablet was written by Assurbanipal's chief scribe, Ištar-šumu-ereš (cf. the ductus and sign forms, especially those of iq, ša, ti, in LAS 13 and CT 53 84, 177, 594 and 943).

No. 9 (K 1292 + OT 130) is a vertical single-column tablet measuring 20 mm x 57 mm (full width) x 104 mm (height). The upper left-hand corner and lower part of the tablet are missing. Vertical script density is 10 signs to 35 mm. Assuming that the ratio between the horizontal and vertical axes was 1 :2, as is normal in this type of tablet, the original height of the tablet was 1 14 mm. This means that about 4 lines have been lost at the end of the obverse and a corresponding number of lines at the beginning of the reverse. In addition, the tablet has a coating of very fine clay which has cracked off from the lower left-hand corner of the obverse and from the beginning of the reverse, causing additional loss of text. Altogether, at least 10 lines (= 24%) of the original prophecy have been totally lost. '

The tablet is beautifully inscribed and by all criteria represents a library copy rather than a report. The scribal hand closely resembles but is not identical with that of nos. 1-3. The script is very clear and easily legible on the obverse, but badly effaced and at times very hard to read on the reverse.

No. 10 (83-1-18,726) is a fragment from the left side of a clay tablet measuring 15 mm (thickness) x 12 mm (width) x 42 mm (height). The curvatures point to a vertical tablet originally measuring ea. 30 x 60 mm. Vertical script density is 10 lines to 41 mm. This implies (including the edges) that the tablet was originally inscribed with about 33 lines of text.

No. 11 (K 1974) is fragment of a vertical clay tablet from the upper left part of the reverse measuring 27 mm (width) by 49 mm (height). Curvatures suggest that the original width was ca. 40 mm, and hence the original height ca. 80 mm. Judging from the vertical script density (10 lines to 34 mm), the tablet was probably originally inscribed with about 50 lines of text.

The lines of the tablet are crowded with text, with signs packed tightly against each other, which combined with textual damage makes the reading and interpretation of some lines quite difficult. The scribal hand is not found on other tablets of the corpus.

280 ln no. 1, the determinative pronoun is written 29 times with the sign ša, 7 times with the sign šá. In nos. 2 and 3, sa is used 9 times each vs. 5 times each for sd; no data are available from no. 4. In sum, the scribe used both signs for writing the pronoun but preferred the sign ša, which appears in all authorship notes of these tablets except 2.3. No. 9 likewise uses ša in the authorship note and both ša and šá in the oracle itself. By contrast, the scribes of tablets 5, 6 and 8 used ša only for writing the pronoun, those of nos. 7 and 11 sá only. In no. 10, both ša and šá are used, but the latter is more frequent (2 attestations against one of fo). In nos. 1-4, the sign te occurs 17 times in final position (including forms with pronominal and enclitic suffixes like am-ma-te-ia), vs. 40 spellings with -ti. Note that virtually all the spellings with -te occur on the obverse of no. 1 (= oracles 1.1-6: 14 examples), which indicates that the scribe initially followed the orthography of the reports he was copying but later lapsed to his own conventions (i.e., the almost exclusive use of the sign ti in final position; cf. n. 281 ). In no. 7, there are 3 cases of final -te vs. 8 cases of -ti; in no. 8, one example of -te and -ti each; and in no. 9, 3 spellings with -te with no examples of -ti.

281 Note, e.g., the syllabic spelling ra-bi-tu in oracle 1.6 as against GAL-tu/tú in I .I and 1.3. The sign forms in nos. 1-4 are on the whole uniform, indicating a single scribal hand. However, oracle 1.6 surprisingly contains some sign variants deviating from the norm (see Table 1, p. LVII). This seems to indicate that the scribe, arriving at the middle of the tablet, had for a moment slipped to mechanically reproducing the sign forms of the original. See also n. 280 above.

Simo Parpola

Simo Parpola, 'Manuscripts', Assyrian Prophecies, SAA 9. Original publication: Helsinki, Helsinki University Press, 1997; online contents: SAAo/SAA09 Project, a sub-project of MOCCI, 2021 [http://oracc.org/saao/saa09/theprophecycorpus/manuscripts/]