Selected Textual Details
THE FEW STYLISTIC AND TEXTUAL characteristics of the Kai- feng Sifrei Torah that have been selected for presentation here are representative of the kind of problems with which the scribes of Kaifeng have unwittingly left us to grapple. They contain clues, as has previously been noted, from which we may hope to learn more than we now know about the history of the Jews of China.
1. LARGE AND SMALL LETTERS
In each of several verses of a Torah scroll one letter is customarily written either oversized or undersized. Thus, the first character in the first word of Genesis is written as a majuscule, while the last character in the first word of Leviticus appears as a minuscule. The extant Kaifeng Sifrei Torah (and the missing Scroll 9 as well) have absolutely no Large Letters or Small Letters. It seems unlikely, of course, that the scribes of Kaifeng could have completely overlooked such strikingly oversized or undersized characters if they were present in the Torah(s) from which the surviving Kaifeng exemplars were copied.
Question: Should the inference be drawn that the Kaifeng scrolls are patterned after an original (or originals) which came from an area (or from areas) in which the use of Large and Small Letters was not part of the soferic tradition? In this connection, it should be noted that these enlarged or reduced characters do not appear in either the two missing scrolls formerly held by the London Society or the American Bible Society scroll, both of which were written long before the flood of 1642.
2. THE COLUMNAR LETTERS
For many centuries it has been the common practice of most soferim to start all columns but five in a Torah scroll with a word beginning with the letter vav. The letters vav which stand at the heads of the columns are known as the vavei ha-amudim ("the columnar letters vav"), and are associated in the tradition with verse 27:10 of Exodus, in which the term vavei ha-amudim (translated as "the hooks of the pillars," or "the hooks of the columns") is used in connection with the instructions being provided for the building of the court of the Tabernacle.
Five columns of a Torah start with letters other than vav. These letters, when read in sequence and then terminated with a vav (so that they create a mnemonic phrase), spell out the two word b'yah sh'mo ("Whose name is the Lord," Ps. 68:5). The letters constituting the two words b'yah and sh'mo are bet-yod-he and shin-mem-vav respectively, and are written in most Torah scrolls emanating from western regions — i.e., in Ashkenazic scrolls — in this order: bet, from Bereishit (Gen. 1:1; and since Bereishit is the first word of the Pentateuch there is obviously no way of starting the first column of a Torah with any other letter than a bet); yod, from Yehuda (Gen. 49:8); he, from habaim (Ex. 14:28); shin, from sh'mor (Ex. 34:11); mem, from mah (Num. 24:5); and vav, from va-eidah (Deut 31: 28). The last of these columnar letters vav could in fact be incorporated into the words b'yah sh'mo from the initial letter of any other column following that in which Num. 24:5 appears, but tradition ascribes this particular vav to the word va-eidah.
All the Kaifeng scrolls I have seen observe the custom of using the letters of the words b'yah sh'mo at the beginning of six columns, but while the positioning of the b'yah sh'mo characters is the same throughout all these Kaifeng exemplars, there is a difference in the placement of three of these characters in the Kaifeng scrolls from that encountered in most western scrolls In the Kaifeng Torahs the yod is from yesakhar (Gen. 49:14), the shin is from shoftim (Deut. 16:18), and the mem is from motzei (Deut. 23:24). In the American Bible Society exemplar, which terminates at Lev. 18:19, we cannot of course tell whether the shin and mem of shoftim and motzei respectively stand at the heads of the columns of Deuteronomy in which we would expect to find them. However, the non-appearance in the American Bible Society scroll of a shin (from sh'mor) as a columnar heading at Ex. 34:1 adds support to the assumption that insofar as the positioning of the characters comprising the two words b'yah sh'mo is concerned the American Bible Society scroll falls within the same tradition as the other Kaifeng exemplars.
The differences in the placement of the yod, shin and mem of the b'yah sh'mo are of interest in that they provide a clue regarding the ecclesiastical sources by which the soferim of Kaifeng may have been guided — for the b'yah sh'mo pattern which characterizes the Kaifeng Torahs tends to indicate a Sefardic, Oriental or Yemenite influence rather than a European Ashkenazic influence. The Kai- feng scrolls, in short, exhibit b’yah sh’mo patterns which parallel those found in the scrolls written for Yemenite, Oriental and Sefardic communities.
3. THE INVERTED LETTERS NUN
Traditionally, the sofer writes an inverted letter nun (nun hafukha) immediately before Num. 10:35, and another immediately after Num. 10:36; but while an inverted nun is found at each of these two positions in the Bridwell, Oxford and Vienna scrolls, it does not appear at all in the exemplars at the Jewish Theological Seminary, British Museum and Cambridge. The American Bible Society scroll lacks the Book of Numbers, so that it is not possible to determine by inspection of this scroll whether the scribal tradition within which it falls required the inversion of the nun at these particular places in Numbers.
Questions: Does the lack of uniformity on the part of the Kaifeng scribes with respect to the use of the inverted letter nun suggest that they had more than one source from which to copy their Sifrei Torah? Alternatively, is it possible that a scribe who was copying from a scroll which contains these inverted characters erroneously omitted inserting them in the Torah he was preparing, and that the other two scrolls which do not have these nuns were then copied from his?
4. THE BROKEN LETTER VAV
Scribal tradition generally calls for the writing of one letter in the Torah defectively (the vav k’tiah), this being the vav which appears in the word shalom (shin-lamed-vav-mem) in Num. 25:12. This vav is deliberately "broken," or made defective, by leaving a small uninked gap in it. One suggested reason for the insertion of this defect is that because shalom means "peace and wholeness," the incompleteness of the one letter in this word is intended to symbolize the absence of an everlasting state of peace among the nations of the world.
None of the Kaifeng scrolls I have seen exhibits the broken vav. In the case of the American Bible Society exemplar, which does not extend as far as Numbers, it is of course not possible to determine whether the scribes who wrote it knew of the vav k'tiah tradition. I doubt, however, that the absence of a break in the vav of shalom should be taken to indicate that the texts of the Kaifeng Torahs were necessarily derived from scrolls which lacked such intentionally mutilated vavs. A scribe who was not aware of the tradition which calls for leaving a small portion of one vav in the Pentateuch uninked would probably not have noticed the defect in the letter as he copied Num. 25:12 from the Torah he was using as a model. Or, if he did notice it, he might have considered it to be no more than a scribal slip, and would therefore not have repeated the same "slip" in the exemplar he was preparing.
The name of Chedorlaomer, king of Elam, is found in Gen. 14:1 and is repeated several times in subsequent verses. The accepted Hebrew spelling of the name of the Elamite monarch is kaph-dalet-resh-lamed-ayin-mem-resh; and this, Hertz writes, is "a Hebraized form of Kudur, ‘servant of’, and Lagamar, the name of an Elamite deity." The explanation Hertz provides may thus explain the division of opinion that has persisted for many centuries as to whether Chederlaomer should be spelled as one word or as two (Chedor Laomer). The question has also arisen as to whether it is permissible (if the two-word spelling is accepted) to end one line of text with Chedor and start the next with Laomer.
Of interest here is the fact that the various Kaifeng scrolls exhibit all three readings at Gen. 14:1. The name of the Elamite king is spelled as one word in the American Bible Society exemplar. In the Jewish Theological Seminary, Bodleian, British Library, and Vienna scrolls it appears as two words, these being written on a single line. In both the Bridwell and Cambridge copies Chedor is written as the closing word of one line, and Laomer as the opening word of the next line.
It is unreasonable to suppose that it was solely by chance that in the seventeenth century the scribes of Kaifeng chose to divide in two a proper name in the Pentateuch over which there happens to be a dispute as to whether it should be so divided. I believe, however, that we should not make too much of the writing of the name on two lines, for the Kaifeng scribes who wrote Chedorlaomer as a two-part name probably did so without attaching any particular significance to what they were doing. If it was convenient to end one line with Chedor and start the next with Laomer, I should think that they simply did so. I doubt that any Kaifeng sofer's acquaintance with scribal tradition extended to the question of the spelling of the name of the Elamite monarch — but, of course, it may have.
Questions: If we assume that in the seventeenth century the scribes of Kaifeng were not aware of the unresolved problem vis-à-vis the writing of the name of Chedorlaomer, by what channels and in which centuries did texts written in the tradition that spelled the name as two words arrive in their city? Also: since the American Bible Society scroll, which was written well before the seventeenth century, spells Chedor- laomer as one word, are we justified in assuming that it derives from a source other than that upon which the surviving post-1642 Kaifeng Torahs are based?
The tractate Megillah of the Jerusalem Talmud states that in Num.10:23 the proper name Pedahzur (pe-dalet-he-tsade-vav-resh) should be written as one word. In the Bridwell, Cambridge, Bodleian, Jewish Theological Society, and British Library scrolls, however, this name appears as two words, Pedah and Zur (pe-dalet-he, followed by tsade-vav-resh). The word is missing from my microfilm copy of the Vienna scroll; and, of course, it does not occur in the incomplete American Bible Society scroll. (See note 3, above, for a roster of other biblical names which present problems to the scribe similar to those created by Pedahzur and Chedor- laomer.)
Question: Do we know of any particular group, or groups, of Jews who customarily wrote Pedahzur as it is written in at least four Kaifeng Torahs — that is, as two words?
7. THE PUNCTA EXTRAORDINARIA
The puncta extraordinaria (sing., nekuda; pl., nekudot) are dots which are found over certain letters and words in specific biblical passages, there being one punctum per letter of text. A punctum may be inserted over each letter of a word (or two words) in a particular verse, or over as few as one or two individual characters. Fifteen verses in the Hebrew Scriptures display pointing of this nature, and of these, ten occur within the text of the Five Books of Moses.
No agreement exists with regard to the origin and significance of the puncta, but a number of interesting hypotheses have been advanced to explain their existence. It has been suggested, for example, that certain early scribes may have copied specks of ink or dirt which looked like dots, and that the inked points thus inserted by these scribes were later recopied by still other scribes, thereby perpetuating the original errors and confounding generation after generation of biblical scholars. Another suggestion is that nekudot were introduced by Ezra as a means of drawing attention to the ten passages in the Pentateuch which in his opinion needed to be altered, but which he dared not alter for fear of violating the Deuteronomic injunction against making any additions or sub-tractions in the sacred text. Ezra, it is said, consequently decided to point out the questionable words by inserting dots over them, reasoning that the insertion of such small non-alphabetical markings would not constitute a violation of the Deuteronomic injunction. Still another theory suggests that puncta, like italics, are intended to add emphasis to the letters or words with which they are associated. When Elijah would eventually appear, thereby presaging the ar-rival of the Messiah, the argument went, he would decide whether the words in question actually needed alteration. If they did not, the dots would be removed; if they did, the dots would not only be removed, but the wording would be corrected. It has additionally been proposed that the presence of puncta over letters or words signifies that these are to be pronounced in a special way in all public readings of the scriptures — or, perhaps, that they are not to be pronounced at all. In the rabbinic literature the insertion of puncta in the Hebrew Scriptures is in certain instances interpreted as indicating that the reader would do well to go beyond the literal sense of the passages in which they occur and search for a deeper understanding of what is really meant.
The individual verses in the Pentateuch which call for pointing over specific letters or words are indicated below, together with the data I have been able to accumulate vis-à-vis the presence or absence of such pointing in the various Kaifeng scrolls. I have personally checked each of the seven extant Kaifeng Torahs for pointing, for the most part by examination at the repositories in which they are held. However, because of time limitations I found it more convenient in a number of cases to refrain from making thorough searches for pointing in the individual Kaifeng scrolls lying before me, believing that I could later undertake these procedures at leisure, by the use of microfilms which were to be forwarded to me. Unfortunately, when these microfilms eventually became available, it turned out that some of them were either incomplete or defective in spots. Accordingly, I have noted below three instances in which I am unable to certify the presence or absence of puncta in verses which ordinarily call for them.
GENESIS 16:5: (One punctum is required in this verse.) Present in the American Bible Society exemplar; absent from the British Library, Cambridge, Bodleian, Vienna, Jewish Theological Seminary, and Bridwell copies.
GENESIS 18:9: (Three puncta required.) Present in the American Bible Society exemplar; absent from the British Library, Cambridge, Bodleian, Vienna, Jewish Theological Seminary, and Bridwell copies.
GENESIS 19:33: (One punctum required.) Present in the American Bible Society exemplar; lacking in the British Library, Cambridge, Bodleian, Vienna, Jewish Theological Society, and Bridwell copies.
GENESIS 33:4: (Six puncta required.) Present in the American Bible Society, British Library, Cambridge, and Vienna exemplars; absent from the Bodleian, Jewish Theological Seminary and Bridwell copies. According to White, both the Hong Kong and the unnumbered London Society scrolls exhibited puncta at Deut. 33:4. White, however, is in error here, for Deut. 33:4 does not call for puncta. We must assume that he should have written Gen. 33:4 rather than Deut. 33:4.
GENESIS 37:12: (Two puncta required.) Present in the American Bible Society exemplar; absent from the British Library, Cambridge, Bodleian, Jewish Theological Society, and Bridwell copies. A defect in my microfilm of the Vienna exemplar makes it impossible for me to determine whether this copy exhibits pointing here.
NUMBERS 3:39: (Five puncta required.) Present in the British Library and Vienna exemplars; absent from the Cambridge, Bod- leian, Jewish Theological Society, and Bridwell copies. Since the American Bible Society Torah does not extend beyond Lev. 18:19, it is not possible to say whether this scroll (if, indeed, it ever was a complete scroll) contained puncta at Num. 3:39 and at the four subsequent passages in the Pentateuch which call for pointing. I would venture the guess that since the American Bible Society scroll is the only extant Kaifeng exemplar containing puncta at all five Genesis verses which call for them, it would also have exhibited such puncta in the verses in Numbers and Deuteronomy where they normally occur — if, of course, the ABS scroll ever existed as a complete Pentateuch.
NUMBERS 9:10: (One punctum required.) Absent from the Cambridge, Bodleian, Vienna, Jewish Theological Seminar, and Bridwell exemplars. My microfilm of the British Library copy is defective here, so that I cannot tell whether this scroll has a punctum in this verse.
NUMBERS 21:30: (One punctum required.) Absent from all the Kaifeng scrolls I have seen.
NUMBERS 29:15: (One punctum required.) Absent from six of the seven Kaifeng scrolls in which I have searched for it, but a defect in my microfilm of the British Library exemplar makes it impossiblefor me to determine whether this scroll does or does not contain a punctum at this position. .
DEUTERONOMY 29:28: (Ten puncta required.) Present in the Cambridge and Bodleian exemplars; absent from British Library, Vienna, Jewish Theological Seminary, and Bridwell copies.
The American Bible Society exemplar, the most ancient of the seven Kaifeng Sifrei Torah I have examined, displays puncta at all five of the positions in its surviving skins at which they are normally found. The Jewish Theological Seminary and Bridwell scrolls, on the other hand, contain no puncta whatsoever. Each of the other scrolls I have inspected displays the appropriate pointing at one or more of the verses in which one would expect to find such pointing, and the same, apparently, is (or was) true of the Hong Kong and London Society copies which are presently missing.
Question: Is it possible that the decreasing frequency of pointing that is observed in the individual surviving Kaifeng scrolls provides an indication of the sequence in which these scrolls were written? If so, the American Bible Society exemplar (and perhaps the London Society copy, if, that is, it displayed puncta at more positions than the solitary position referred to by White) would be the earliest to have been written, while the Jewish Theological Society and Bridwell copies would be the last.
8. DEUTERONOMY 32: 25
This verse, which occurs in the Song of Moses, reads: "Without shall the sword bereave, And in the chambers terror: Slaying both young men and virgin, The suckling with the man of gray hairs."
Father Jean Domenge reported that this verse was read publicly as part of the Shemini Atzeret Pentateuchal portion during his visit to the synagogue of Kaifeng on Saturday, 3 October 1722, and that the scroll from which it was read was the one which the con- gregation called the Scroll of Moses.
Domenge noted that in place of kfa, (t’shakel, tet-shin-kaph-lamed, meaning "bereave") the text contained the word kft, (tokhel, tet-aleph-kaph-lamed, meaning "eat," or "devour"). The sense of the verse, if not the accuracy of its spelling, he then commented, was thus preserved — with the sword devouring instead of bereaving.
All the scrolls I have examined (excluding, of course, the American Bible Society exemplar, which ends in the middle of Leviticus) spell t'shakel correctly; and, apparently, so does the missing Hong Kong scroll.
Oddly enough, these same scrolls all misspell the very next word, crj (hereb, or "sword"). The proper spelling of hereb is het-resh-bet, but in each of these scrolls it is rendered as ckj (het-lamed-bet) — a word which, depending on how it is vocalized, means either "milk" or "fat," and makes utter nonsense out of the phrase in which it occurs. I would venture the guess, however, that in the Scroll of Moses hereb was spelled correctly; for if it had not been spelled properly Domenge would assuredly have informed us of the error when he translated the three-word phrase as "from without shall the sword devour" and remarked that the correct phraseology was "from without shall the sword bereave."
It has been suggested that the substitution of a lamed for a resh reflects the difficulty that the pronunciation of the r sound (resh) presents to most Chinese, and that its replacement by an l sound (lamed) may be explained as the misspelling of a word because of an inability on the part of the scribe to sound it properly. This may be so, but if it is actually the case, why does not this resh/lamed interchange occur with even greater frequency throughout all the extant Kaifeng scrolls than it does?
Question: Do the spellings errors occurring in Deut. 32:25 suggest that the Scroll of Moses was not (as has been claimed in the literature) the exemplar from which the extant post-1642 scrolls were copied?
9. THE ABSENCE OF TAGIN
Tagin (sing.: tag) are flourishes that are placed by the scribe above certain letters of the Hebrew alphabet. These markings must be present in Pentateuchal scrolls used in Ashkenazic synagogues, but are rarely found in scrolls written for Sefardic houses of worship. Birnbaum, in a discussion of Plate 191 of his monumental work, The Hebrew Scripts, states that:
The kind of Square script used in synagogue scrolls is distinct from the Square used for other purposes. The most striking difference between the Scroll script and other Squares is the use of Tagin and the shape of the letter het [j]. The Tagin`crowns’ are small, hairfine ornamental strokes added to certain letters. Gimel [d], vav [z], tet [y], nun sofit [I], nun [b], ayin [g], tsade sofit [.], tsade [m], and shin [a] have three Tagin each. The scribes of certain groups also ornament bet [c], dalet [s], he [v], het [j], yod [h] and kuf [e], but with one stroke only. Het consists of two zayins [z] linked together by a bridge in the form of a tall circumflex called gimel-gimel [dd, ‘a roof’].
Not one of the post-1642 Kaifeng scrolls I have seen displays a tag of any kind, whether of the three-stroke or one-stroke genus. The same is true, moreover, insofar as the much older skins in the American Bible Society exemplar are concerned. Nor are any tagin found in the facsimile printed by White (Chinese Jews, ii, p. 157) of three columns of a skin which was part of the now-lost London Society Torah, a skin whose antiquity almost surely rivals that of the skins in the American Bible Society exemplar.
Birnbaum asserts that the letters zayin appearing in a facsimile he provides of the passage Gen. 7:16-8:4, as it appears in the British Library’s Kaifeng Torah scroll, display "the characteristic Parsic form." He does not elaborate, but later, when he turns his attention to Parsic calligraphy, he makes the following comment about its nature, origins and dispersion:
For centuries Babylonia formed part of Persia and Parthia, and it is therefore reasonable to assume that emigrants from that Jewish centre spread into Persia and carried their script with them....
While the Types [i.e., letter styles] hitherto dealt with are very similar to each other, this one differs markedly from them. That is true not only of the later stages, [for] our earliest specimens show a highly individual development.
In spite of the antiquity of the Parsic group, we have few documents older than the second half of the fifteenth century. They do not come from Persia proper but from farther afield in the east.
We cannot say with complete assurance that the absence of tagin in the Kaifeng Torahs indicates that they were necessarily copied from Sefardic originals, for it appears that in the era during which the Kaifeng kehillah first came into being Sefardic congregations used Torahs in which the presence of these markings was optional. Had this not been the case, Maimonides (1135-1204), himself a Sefardic Jew, would scarcely have made a point of declaring, as he did, that while the presence or absence of tagin in a Torah scroll is not of itself a critical factor in determining whether the scroll may be used in public worship services, it is preferable that it contain these em- bellishments. His reason, quite simply, was that the presence of tagin in a Torah reflected a laudable desire on the part of the writer to enhance the beauty of the text.
When one considers that the number of tagin one expects to find in any Torah scroll that contains them runs well into six figures, it strains the imagination to assume that a Kaifeng scribe who used a Torah displaying these ornamental serifs as his master copy could have failed to notice their presence. While it is possible that — for reasons unknown to us — it had become a deliberate practice of the Kaifeng scribes to omit these markings in the scrolls they produced, these omissions would appear to have been made unwittingly rather than the consequence of ignorance or carelessness, and it becomes reasonable to assume that their master copy or copies were altogether without tagin.
Questions: From which region in the world might the earliest Jewish emigrés to Kaifeng have brought the tagin-less Torahs which served as master copies for their descendants? Or, if the extant Kaifeng scrolls were copied from originals imported to China by these descendants, from whom did they get these? What, moreover, might the identification of the non-Chinese sources of these Torahs tell us about the nature and scope of the interactions between them and their Kaifeng coreligionists?
These are very numerous, running well into the hundreds, but only four are noted here as illustrative of the genre.
(A) At Gen. 19:12 the word het-tov-nun is spelled correctly in the American Bible Society, British Museum, Bodleian, Vienna, Jewish Theological Seminary, and Bridwell scrolls, but as he-tov-nun in the Cambridge copy. Seemingly, this error in the Cambridge exemplar may be dismissed as merely a scribal slip.
(B) At Gen. 30:26 the word tov-nun-he is spelled correctly in the American Bible Society scroll, but an aleph is im- properly substituted for the he in all the other scrolls I have seen.
(C) The word aleph-yod-bet-koph is spelled correctly at Deut. 20:1 in the Bodleian, Cambridge and Vienna scrolls, but in the British Museum, Jewish Theological Seminary and Bridwell exemplars a superfluous yod is written after the bet. (The American Bible Society scroll does not extend to Deuteronomy.)
(D) The very last letter of the Pentateuch, a lamed, is traditionally written higher than any other lamed in the Torah. None of the Kaifeng scrolls I have exarnined treats this final letter differently from any other lamed in the Pentateuch. It does not appear, of course, in the incomplete American Bible Society scroll.