The Shaphan Group

The Shaphan Group’s Prodigious Output

excerpted from The Shaphan Group by Preston Kavanagh

In the preceding chapter we learned that Shaphan, his family, and his colleagues wrote much of the Deuteronomistic History. Their coded fingerprints fill the books of Deuteronomy through Second Kings in the form of anagrams and hidden spellings. Shaphan, who was King Josiah’s secretary and whose career spanned the later monarchy and much of the Exile, probably led the group. Also among those who operated together were Shaphan’s three sons and his father; fellow-courtiers Asaiah, Micaiah, and Achbor; the prophets Huldah, Jeremiah, and Jacob; the brothers Jozadak and Ezra; the scribe Baruch; and probably Daniel.

In this chapter we shall establish that this same group of experts drafted a large amount of other Scripture, and we shall be specific about what the scripture was. Before we embark on that journey, however, here is a summary of discoveries about coded writing.

Shaphan Group Edits DH, Pentateuch

The Shaphan group used editing to do most of its work on the seven-book DH. Not infrequently, however, its members drafted entire chapters of new text. These are best identified by counting significant signature groups within each chapter and then applying a chi-square test. Joshua had seven chapters that passed this rigorous test (Josh 12, 13, and 15–19); Judges had two (Judg 5 and 12); and First Kings one (1 Kgs 6). The material in the seven Joshua chapters had sufficient spill-over to push the entire book into the beyond-coincidence zone.

Table 5.2 on the next page helps to demonstrate just how difficult it must have been to encode multiple spellings of names within a Hebrew text. The table shows the coding in Judg 12, which tells of the judge Jephthah, inter-tribal battles, and the Shibboleth password. We selected Judg 12 simply because it was among the shortest of the DH chapters that the group composed. By happenstance, it contains statistically significant groups from fourteen of the Shaphan group’s fifteen members. All but one of the chapter’s spellings employs athbash, which encoders used to create multiple versions of the same name. Although at 223 Hebrew text words the chapter is short, into it the authors crowded 266 spellings of coded words ranging from four to twelve letters each. The writers fashioned twenty-nine different groups of such spellings, each of which passed a stringent probability test (and in turn the twenty-nine together passed the same type of test for chapters). The Shaphan-group authors did all this while writing cogent accounts that, over two millennia later, we still find holy and marvelous.

The Shaphan group also played a major role in assembling the first three books of Scripture. According to the evidence from coded spellings, those writers accounted for fourteen chapters, including six Exodus chapters (25–27 and 36–38) and eight in Leviticus (1, 3, 4, 6, 9, 12, 15, and 19). Previously we identified Ezra as the P Source and said that he is also one of the fifteen in Shaphan’s group. Experts attribute to the P Source Exod 25–27 and 36–38, and we are here broadening authorship to others beside Ezra. In addition to him, authors of some or all of these Exodus P Source chapters include Achbor, Ahikam, Asaiah, Baruch, Daniel, Elasah, and Huldah. It seems that the Shaphan group also served as the P Source. If so, Shaphan-led writers also edited or composed substantial portions of Genesis and Exodus that our coded spellings approach is not picking up. Because chapters are the basic unit of this technique, it is not well suited for dealing with edited texts—such as those of the P Source—which have shorter passages inserted here and there. To address this, we could modify the design of coded spellings but, for this book at least, we shall stick with chapters as our basic unit. Please keep in mind, however, that results will be understated, and that Shaphan and his colleagues did more than we can immediately identify.

The opening chapters of Genesis offer a sharp lesson. In the preceding chapter we introduced Shaphan by detailing his extensive use of anagrams in Gen 1. There were sufficient anagrams, in fact, to produce a near-zero probability of coincidence. Yet the coded spellings approach found little in Genesis. The prior sketch of Shaphan anagrams in Genesis also rated chapters 10 (pre-history) and 35 (Jacob saga) as heavily salted with his anagrams, but coded spellings found nothing significant. Demonstrably, biblical authors had ways other than coded spellings to mark their work, and anagrams were among those ways. We now move to the book of Numbers.

In parsing the book of Numbers, we followed the analyses of experts. The second column of table 5.3 (see following), “Content,” shows this, with the shaded sections indicating high statistical significance. The table uses coded spelling groups per thousand text words as a simple way to display coding density. Scripture as a whole averages seventy-two groups per thousand. For Numbers, eight of the opening chapters (“Preparations” at 104) and the book’s final eleven (“Promised Land” at 92) are very high, while the “Camp Purity” and “Rebellions” chapters (Num 5−6, 11−14, and 16−17) seem to be holdovers without coding. The highly coded Preparations and Promised Land sections (Num 1−4, 7−10, and 26−36) could pertain to the 570s when the Cyrus-led rebellion was under way. Chapter 3 of this book describes what anagrams tell us about those years. Dividing Numbers into chapters featuring either laws or narratives shows that the Shaphan composers did more with legal chapters (91 per thousand) than with the narrative ones (68).

For scholarly purposes, however, this section-by-section approach to Numbers is short on details. As a remedy, table 5.4 gives a chapter-by-chapter listing of the signature groups per thousand words. Shaded chapters indicate chi-square coding that exceeds the .001 base line standard. For most chapters, a higher level of spellings produces a solid probability result. A notable exception is chapter 24, which carries a low 42 per thousand. It earns its shading because it contains six different high-probability groups of “Elasah,” the name of one of Shaphan’s sons. Almost half the chapter’s coded groups had Elasah’s signature. All in all, this proves the worth of chi-square testing in comparison to a different measure like spellings per thousand text words (S/1000 in table 5.4). The table shows that four other chapters (7, 10, 33, and 34) also conceal plenty of coded spellings. Including these, the Shaphan group, in part or whole, wrote chapters 1–3, 7, 10, 19, 24, 26, 28, 29, 33, 34, and 36. The table reveals that Num 1, with 588 signatures per thousand words, is in a class by itself. Certainly it is among the highest in all of Scripture. Ostensibly, the chapter is about Moses taking a census of Israel. In actuality, it must have served a more immediate exilic purpose, such as a count of the forces assembled in the 570s to take back the Promised Land.

To illustrate further, details follow about coding of the Shaphan group members. A dozen of the fifteen authors had a lot to do with the book of Numbers. Ranking serves only limited purpose since the coded spellings of each have essentially a zero percent probability of coincidence. Nevertheless, here they are. Micaiah, who was Shaphan’s contemporary in Josiah’s court, led all others with ninety-four coded groups. Shaphan and two of his sons, Gemariah and Elasah, followed Micaiah. Grouped in the high seventies were Daniel, Jeremiah, and the two priestly brothers Jozadak and Ezra, while Huldah, Jacob, Asaiah and three others were below that level.

At the bottom of the list of writers (or perhaps of subjects, since positive coding could signify either subject or author) is Shaphan’s father Azaliah. If we are correct about a date of around 570 BCE for Numbers, Azaliah would have been dead for many years. Also at the bottom of the ranking list is Micaiah’s son Achbor and Shaphan’s son Ahikam. Though both led in drafting the history books, they seem to have played only a minor role in Numbers. All of this is preliminary, since Scripture contains some thirteen hundred Hebrew personal names and we are working with only a handful. Eventually, persistent scholars with all possibilities before them will perform a more comprehensive examination.

Disclosure of First Importance

This writer now presents a disclosure of first importance, one that could rival any ever made in the study of the Hebrew Bible. The Shaphan group wrote or strongly influenced a large amount of the Writings and of the Prophets. We draw this conclusion based upon clear—we think irrefutable—evidence of coded writing within Scripture. So far we have found that Shaphan and his colleagues edited the books from Deuteronomy through Second Kings, and that these same experts shaped portions of the Torah—Genesis, Exodus, and Numbers. Now we add this: these fifteen Judahites also wrote or influenced to a statistically significant degree the books of Joshua, Psalms, Proverbs, Job, Songs, Lamentations, and Daniel. Moreover, the Shaphan group initiated or heavily edited much of the Prophets, including the books of Isaiah, Hosea, Joel, Amos, Micah, Nahum, and Habakkuk. Finally, the names of the members of the Shaphan group are also significantly coded in the books of Ezra and Nehemiah. (Appendix 2 gives details of all these books.)

We came to form these startling conclusions by applying probabilities to coded writing, of which the basic unit of measurement is the statistically significant group. With rare letter combinations it can take only a single spelling to form a significant group, but with common letters it can require one hundred spellings or even more. Scripture-wide, the average number of hidden spellings necessary to achieve statistical significance is eleven.

The Shaphan Group Writes Psalms

A real-life example from Psalms follows. Hidden within Ps 1 are three coded spellings of an athbash of “Huldah.” In the rest of Scripture, that spelling occurs only 422 other times. Psalm 1 contains 67 text words, while the remainder of Scripture has 305,429. The two chi-square proportions are 3 / 422 and 67 / 305,429, and they yield a probability of coincidence of 1.21 × 10-20. This decimal is well below our threshold of .001. The three Huldah spellings in Ps 1, then, qualify as a statistically significant group. We tested all of Scripture for coded spellings of “Huldah,” “Huldah the prophetess,” and athbash variants of the two, accumulating statistically significant groups as we went. We then repeated the process for the other members of Stephan’s group.

In total, the fifteen pairs of names produced 3,250 groups within Psalms, and 18,724 in the rest of Scripture. Knowing the exact numbers of text words in both, it was simple to test whether this could be coincidental. The calculation’s result was 0. That is, the probability that a text the size of Psalms would contain that many coded groups by chance alone was simply zero. Common sense confirms this statistical conclusion: the book of Psalms, with 6 percent of the Bible’s text words, has 15 percent of its Shaphan-related significant groups. Given that we are working with thousands of both text words and statistically significant groups, chance is out of the question. Only intent could have accomplished this.

There are numerous psalms. Does this zero probability of coincidence for the entire book mean that Shaphan-group members were involved in every psalm? No, but only nine psalms have a below-average number of significant groups; for all the others, the balance is well in favor of Shaphan-related coded spellings. Of the nine below-average psalms, two (Pss 118 and 134) have statistically significant numbers of “Baruch” biblical anagrams. A third, Ps 53, ends with the words “Jacob will rejoice; Israel will be glad”—a Jacob-Israel parallel that Second Isaiah used to sign his work. In summary, 144 psalms appear to be by or about the Shaphan group, and only Pss 28, 29, 33, 96, 100, and 115 do not, at present, qualify.

There is another, more stringent, way of deciding whether the Shaphan group authored any single psalm. Appendix 1 displays the results of this rigorous approach. Within the appendix are eighty-seven psalms that satisfy a probability standard of .001 (one chance in a thousand of being coincidental). Additionally, eleven psalms make the appendix by slightly relaxing that standard to .005 (one chance in five hundred). With the chi-square method, the total is either eighty-seven or ninety-eight psalms, depending upon where one draws the probability line. The larger number is about two-thirds of the 150 psalms. The above-average approach concluded that “144 psalms are by or about the Shaphan group,” while appendix 1 offered either eighty-seven or ninety-eight psalms. Readers may take their choice (or make no choice at all). Whatever the total actually is, Shaphan and his colleagues were in good part responsible for the book of Psalms.

Did each of Shaphan’s compatriots substantially contribute to the book of Psalms? The answer is a resounding yes. Each and every one of those fifteen Judahites had their names encoded within the psalm texts to a statistically lopsided extent. The proof is that not one of the fifteen chi-square results had fewer than fifty zeros to the right of the decimal point. (After four or five zeros, most would conclude that there is no real possibility of coincidence.) However, clearly all fifteen could not have written or been the subject of every psalm, so a psalm-by-psalm analysis remains to be made. As an aid to that, appendix 1 shows which Shaphan-group members have their names significantly encoded within ninety-eight of the psalms.

Other tests of the breadth of authorship within Psalms are easily done. Most psalms have headings, which range from one word to more than a dozen. Two examples are “To the leader. A Psalm of David” and “A Song. A Psalm of the Korahites.” Over the centuries, psalm headings have drawn intense scholarly interest. Reasonably, those who study Scripture have suspected that the headings contained clues about dating and authorship. So far, however, no conclusions have emerged—until now, with the discovery of coded signatures. Probabilities tell us that all fifteen members of the Shaphan group made strong contributions to the book of Psalms. To see whether this held true in smaller lots, I used Nahum Sarna’s categories to test psalms with Korah and Asaph headings, along with the so-called royal psalms.

Using pairs of the Shaphan-group names and their athbash variants, I ran over six hundred name versions against the batches of Korah, Asaph, and royal psalms. I expected to find that some of the fifteen had worked on one batch and others on another. Instead, the results were different. The name of almost every member of the Shaphan group was significantly coded in the Korah, Asaph, and royal batches of psalms. Of the forty-five trials (fifteen names times three psalm batches), forty-one had essentially no chance of coincidence. As to the four remaining, they recorded modest 4, 3, 3, and 2 percent chances of coincidence. Ezra, Shaphan, and Jeremiah had those lower scores, so they probably did not work on all of the Korah and Asaph psalms. (Chi-square results showed full participation in the royal psalms.)

Whoever assembled Psalms grouped them into five books. This, too, has been a subject of intense research for many a century. The thinking is that this five-part division is the product of “a long and complex history involving several small collections and their combination into larger units.” To see if the coded-writing discovery could shed any light on this aspect of psalm research, I ran the fifteen sets of Shaphan names against Book V of Psalms, which includes Pss 107–50. The outcome at least equaled those of the Korah, Asaph, and royal trials. All but one of the fifteen names from the Shaphan group had probability scores that ruled out coincidence. The sole exception was Jeremiah. He had a .0016 score, which barely fails our .001 cutoff. Whatever the source of the Book V collection, it shared the same composing group as the entire book and as the Korah, Asaph, and royal batches. We can infer, then, that neither the five divisions by book nor the three groupings by psalm-headings pertain to authorship. Perhaps, instead, they might refer to periods of composition, or to life situations such as wars, persecutions, or the reigns of Judah’s final kings.

It appears likely that the Shaphan group wrote most of the book of Psalms. Almost certainly there are additional authors we have not yet uncovered, but these fifteen are fifteen more than anyone else has ever discovered, and so we begin with them. In 621 BCE, workmen found “the book of the law” while repairing the temple. In Josiah’s court at the time were Shaphan and his son Ahikam, along with Micaiah and his son Achbor. These four are on the list of fifteen as is Azaliah, Shaphan’s father. This evidence indicates that the oldest members of the group that wrote Psalms are Azaliah, Micaiah, and Shaphan himself. In 621, Micaiah and Shaphan were fathers and Azaliah was a grandfather. Considering that King Josiah took the throne in 640, a date of about 625 seems as good a year as any to fix the earliest date of composition for any psalm. This would allow nearly forty years of the monarchy to provide a backdrop for the eighty-one psalms with a David heading. (An enthronement in Jerusalem during the Cyrus revolt might also have inspired royal psalms.)

The Cyrus possibility helps to set an ending date for the book. Scripture contains 999 biblical anagrams for “Cyrus,” and the book of Psalms, with thirty-two BAs, has more than its proportionate share. The probability of coincidentally finding so many Cyrus BAs within that book is essentially zero. Because of BAs, Pss 8, 110, and 129 stand by themselves as related to Cyrus, but even more important than these is an entire family of psalms that announce Cyrus as their subject. Psalms 120–34 begin with “A Song of Ascents,” an athbash anagram for “Cyrus.” (Psalm 121 is a possible exception. The Leningrad text lacks h, but several other versions include it and so complete the Cyrus BA.) These fifteen ascent psalms could even have been bulletins about the revolt’s progress and the rebuilding of Jerusalem’s defenses.

Psalm 110 has two Cyrus BAs and also features Melchizedek, the mythical king of Salem who blesses Abraham near what probably was Jerusalem. We believe that this book’s second chapter establishes the fact that Melchizedek and Cyrus are one and the same. The Persian first surfaces in Scripture in the late 570s, which helps us assign an ending date to the book of Psalms. Another late psalm is number 90, in which the eighty-year-old Shaphan bemoans God’s wrath. We have previously reasoned that Shaphan wrote Psalm 90 between 570 and 565. The span during which the Shaphan group composed the book of Psalms, then, would be about 625 to 565 BCE.

Writing One-Third of the Hebrew Bible

The same 625 to 565 BCE period would also be the span of years during which the Shaphan group wrote the flower of Old Testament literature. And how much does that include? Appendix 1 lays it out: up to 331chapters of the 929-chapter Bible—more than one-third of Scripture! Remember, too, that this is conservative. It includes only whole chapters with statistically significant groups of coded spellings. Partial chapters have not been considered, nor have texts with improbable concentrations of anagrams. For illustration, table 5.5, excerpted from appendix 1, includes a few chapters from Numbers, Deuteronomy, and Joshua. Three of the center columns list the number of coded signature groups, chapter text words, and chi-square values. At the far right are abbreviations of Shaphan-group names coded within those chapters. Most, though not all, chapters in the appendix feature long strings of name abbreviations. This means that, in many cases, groups rather than individuals composed Hebrew Scripture.

The Book of Isaiah

As the full appendix shows, the Shaphan group was deeply involved with the book of Isaiah. Its members wrote or were the subjects of half or more of the chapters of so-called First Isaiah (1–39), Second Isaiah (40–55), and Third Isaiah (56–66). There is a surprisingly even distribution among Isaiah’s three parts. This and the multiplicity of coded names within chapters indicate that the Isaiah corpus was in the group’s possession for an extended period, and that all fifteen writers applied their talents to it. The group generated about two thousand significant spelling groups within Isaiah’s sixty-six chapters. It is instructive to see which of the members led in which Isaiah section. Table 5.6 tells as much as has ever been known about the book’s authorship.

Jozadak, son of the high priest slain after Jerusalem’s fall, leads his fourteen Shaphan-group colleagues in coded signature groups within the whole Isaiah text. Jozadak ranked first in both the First and Second Isaiah chapters, and made an average contribution to Third Isaiah. Jozadak must have had lengthy experience in drafting Scripture before his exile in 586 BCE, and so it seems reasonable that Isa 1–39 was penned during the later monarchy, with many of the chapters coming from Shaphan’s father Azaliah. He was second, fifth, and second, respectively, among the Isaiah segments. Ahikam, one of Shaphan’s sons, was third in coded groups within the thirty-nine-chapter opening section of Isaiah, and Micaiah and Baruch also led. Both were known scribes.

Within Second Isaiah chapters, surprisingly Jozadak again placed first. Daniel and Asaiah ranked second and third, though this might have been because they were subjects rather than authors of the chapters. We can assume that Daniel’s career in Babylon would have been watched closely by the exilic community. As to Asaiah, an official under Josiah, there is reason to suspect that he was a military leader during the Cyrus revolt. Jacob, our favorite for Second Isaiah, is notable by his absence from the top five. (He ranked seventh of fifteen.) But given Jozadak’s preeminence, he must also become a candidate for the Second-Isaiah title.

Remarkably, one person is missing from our Isaiah table. That person is Shaphan, Secretary to King Josiah. Shaphan became the central figure in our search for the Dtr. Coded-spelling probabilities supported that he, his family, and several associates (Micaiah, Achbor, Asaiah) joined three prophets (Huldah, Jacob, Jeremiah) and several others (Ezra, Jozadak, Daniel, Baruch) to rework the Bible’s history books, and beyond that to give us far more Scripture than we imagined possible. Among the fifteen authors of Isaiah’s three sections, Shaphan ranked fifteenth, fifteenth, and fourteenth, respectively. Overall, he had less than half as many coding groups as Jozadak, the leader. A likely answer was that Shaphan oversaw and coordinated the work that others performed, or perhaps as the years passed his responsibilities changed. Our analysis of Ecclesiastes indicates that in the course of the Cyrus revolt Shaphan may even have become king in Jerusalem. While holding these possibilities in mind, we shall continue to employ the term “Shaphan group.”

Table 5.7 on the next page displays in detail the Shaphan group’s impact upon those chapters traditionally allotted to Second Isaiah. It shows plainly that members contributed heavily to six chapters, while only one or two writers are encoded within Isa 43 and 55. For eight of the chapters—44, 45, 47, 49–53—the appendix has no entries at all. This appendix 1 excerpt does not single out Jozadak, Daniel, Asaiah, Micaiah, and Azariah as leading contributors to the Second Isaiah corpus. Instead, the appendix works from a collective, more conservative, basis. While both approaches are useful, they make it clear that we need far more coding data than these tables present. But we have made a beginning.

Turning to Third Isaiah, two of Shaphan’s sons—Gemariah and Elasah—helped draft those eleven chapters. Jacob himself joined them, along with Jeremiah, his fellow prophet. Although Shaphan’s father Azaliah was probably the oldest of the fifteen, he had the second highest number of coded groups in Isa 56–66. This plants the thought that the Third Isaiah Scripture originated during the monarchy. Chronologically, then, it may belong with the First Isaiah chapters.

Overall, the three sections of Isaiah have a suspicious uniformity of composition—each has nearly the same percentage of chapters drafted by the Shaphan group, about 53 percent. It is enough to make one think that a second group of authors might be responsible for part or all of the balance of Isaiah. And if so, why not look to Babylonia to find them? For example, by tradition, Jews exiled with Jehoiachin established themselves in Nehardea on the Euphrates. (The town became famous as a center of Jewish scholarship.) Shaphan group members would have known those early settlers personally. One can envision groups of writers at Nehardea or Jerusalem composing new Isaiah chapters in response to previous chapters and then circulating them to others. We have accounted for only thirty-five Isaiah chapters thus far and someone wrote the other thirty-one. Perhaps further decoding work can identify them.

Within the book of Isaiah, the group’s members were responsible for oracles against foreign nations in Isaiah chapters 13, 17, 18, and 21–23. Appendix 1 shows that the Shaphan group also composed similar chapters in the books of Jeremiah (46 and 47) and Ezekiel (27 and 30).

Interestingly, the group contributed only two other chapters to Jeremiah’s text, one of which (Jer 10) was due solely to heavy encoding of that prophet’s own name. For his full book, we note a surprising dearth of Jeremiah coded spellings.

Ezekiel and Joshua

As for Shaphan-group coding, the book of Ezekiel is somewhat different. Eighteen of Ezekiel’s forty-eight chapters bear the signatures of the group’s members, and eight chapters (5, 6, 19, 22, 23, 26, 27, and 42) have coding from all or most of the entire membership. This extensive coding indicates that Shaphan members very likely composed those specific chapters. And by extensive coding we mean five to seven hundred intentionally crafted spellings per Ezekiel chapter. This confirms scholarly opinion that the book contains a considerable amount of redaction of the prophet’s own words. However, Shaphan-group coding does not go much beyond those eighteen chapters. Probability trials on the remainder of the book yield nothing significant. Still, one thing deserves special mention.

Six Ezekiel chapters featured spellings of “Asaiah” and “Asaiah-servant-of-the-king”—something that Ezekiel himself probably wove into his texts. For Asaiah-related coding, Ezek 1, 7, 16, 24, 45, and 47 are among the strongest in Scripture. This emphasis upon Asaiah points us to another area of Scripture—the book of Joshua.  “Joshua” conceals an anagram of “Asaiah.” According to appendix 1, Joshua 12, 13, and 15–19 brim with Shaphan-related coded groups, over 500 of them. The story line that I believe will prove true is that these Joshua chapters are about the revolt of the 570s, which was led by members of the Shaphan group, Asaiah among them. Joshua is Asaiah; indeed, Joshua may have been invented for that campaign. The revolt was a disaster and Joshua-Asaiah fell into Babylonian hands. They executed him as a substitute king—an event memorialized in the Twenty-third Psalm. Three “Asaiah” coded spelling groups are embedded within that psalm, something which has a zero probability of coincidence. A theory as novel as this certainly deserves fuller discussion, but this is as much as can be spared at this writing. The reason for introducing Asaiah’s nom de guerre is to explain his exceptional coded presence in the book of Ezekiel.

The Minor Prophets

The Shaphan group had as great an impact on the Minor Prophets as it did on the books of Isaiah and Ezekiel. The fifteen writers originated or heavily edited almost half the chapters attributed to the twelve prophets. Shaphan’s father Azaliah has the most significant coded groups, followed closely by Jozadak, heir to the chief priesthood. This suggests that a portion of the revision of the Minor Prophets took place during the later monarchy. Azaliah, as senior member of the Shaphan group, would then have been in his prime. Also, because Jozadak was exiled to Babylon after Jerusalem’s fall, we can assume that much of his contribution came before that. However, coding evidence also suggests an exilic date for the additional chapters. This is because of higher coding totals from Shaphan’s son Gemariah and from Jacob (Second Isaiah) and Daniel. Shaphan’s colleague Micaiah also belongs in that group. Probably the revisions came during both periods. Among all fifteen group members, Shaphan was dead last in coded signatures within the Minor Prophets—as he was in the Major Prophets. Appendix 1 and this note both list the prophetic chapters affected.


The book of Proverbs holds but 2.3 percent of Scripture’s text words, yet it contains 5.7 percent of its Shaphan-group coded spelling batches. This astounding imbalance is mirrored in appendix 1, which aligns coding results with Proverbs chapters. Four-fifths of them were signed and signed and signed yet again with the names and titles of most or all of the fifteen members of Shaphan’s group. Given this evidence, there can be no question as to who wrote virtually all of Proverbs or the period during which it was written. Shaphan group members wrote it between 625 and 565 BCE. Without the coding discovery, this book might never have been dated, due to the difficulty of connecting specific historical events to the generalities of wisdom literature. The leading contributors of coding groups to Proverbs were Jozadak, Ahikam, Gemariah, Jacob, and Micaiah.

This author has written elsewhere that an ongoing battle between the prophet Jacob and his unknown critics is spread across eight chapters of Proverbs (2, 4, 8, 10, 11, 20, 28, and 31). Two passages, generously coded with Jacob spellings, will illustrate. This first is an attack upon Jacob’s wife: You “who rejoice in doing evil and delight in the perverseness of evil; those whose paths are crooked, and who are devious in their ways. You will be saved from the loose woman, from the adulteress with her smooth words” (Prov 2:14–16). After a number of thrusts and counterthrusts undergirded with Jacob coding, in the final chapter the prophet concludes with his famous hymn to a woman of worth: “A capable wife who can find? She is far more precious than jewels. The heart of her husband trusts in her, and he will have no lack of gain. She does him good, and not harm, all the days of her life” (Prov 31:1–3). In the future, scholars may choose to make the Jacob controversy a starting point for their Proverbs examinations.

Job and Songs

The same people who wrote the book of Proverbs also did their work in Job. Ahikam, Gemariah, Jozadak, Micaiah, and Jacob occupy the first five places (out of fifteen) in Proverbs, just as they did in Job, though they rank in slightly different order. About 40 percent of Job’s chapters have significant numbers of coding groups by the fifteen Shaphan members, and another 25 percent conceal a significant variety of coded spellings for individual members. In all, close to two-thirds of the Job text is so marked. This author finds no distinctive patterns in the Shaphan coded chapters, though someone more expert in the book could perhaps do so. Speakers in chapters coded with the full fifteen Shaphan members run Eliphaz-Job-Job-Job-Eliphaz-Bildad-Job-Zophar-Job-Job-Job-Lord-Lord-Lord-Lord (chapters 4, 10, 12, 14, 15, 18–20, 23, 28, 29, and 38–41). Chapters with exceptional concentrations of a single member’s name seem almost random: Ezra, Daniel, Gemariah, Azariah, Jacob, Ahikam, Gemariah, Jozadak, Ezra, Asaiah, and Jozadak (chapters 2, 5, 9, 11, 16, 17, 21, 24, 26, 30, and 36). It is no small thing to establish the date of the book of Job. On the basis of linguistic evidence, scholars will be comfortable with our seventh-sixth century finding. What is most notable is that Job should be added to the already impressive bibliography of the Shaphan group.

Song of Songs is yet another book by the fifteen. Songs consists of eight chapters, six of which are by the collective membership of the Shaphan group. Chapters 1 and 8 are not by them, and presumably were added later. The text as a whole conceals 214 coded groups apportioned to from nine to fourteen of the individual members, depending on the chapter. This coding concentration cannot be coincidental. Writers who took the lead in composing are Gemariah and Micaiah, the same ones who worked in Proverbs and Job. But unlike those books, Songs also had Asaiah, Baruch, and Daniel as leaders in coding groups. Scholars have been unable to identify the authors of Song of Songs, and are split as to whether the book is pre- or post-exilic. We are glad to shed light on these questions.

As to the book of Lamentations, chapters 4 and 5 are by the Shaphan group, and there is no statistical reason to attribute any of that text solely to Jeremiah, who is sometimes thought to be the author.

Daniel and Chronicles

A comparison of the MT of Daniel and various Aramaic fragments found at Qumran shows that the Aramaic in Daniel is much earlier than that used during the second century BCE. It seems to follow that much of the book itself was written when the Bible says it was—in the time of Daniel, which was during the Exile. This is supported by coded spellings in Dan 2–5 and 7–12. Indeed, chapters 2–5, 8, and 11 were written with the participation of all, or all but one, of the fifteen Shaphan-group members. The entire book contains 761 statistically significant coded groups. The probability that this many groups could appear in Daniel by chance has fifty-two zeros to the right of the decimal.

Given such broad coding support, the matter of which individual had the most coding groups becomes less important. Nevertheless, Jacob, Ezra, Baruch, and Huldah ranked one through four with groups in the lower 60s. Jozadak, Elasah, Gemariah, Achbor, and Daniel himself are in the 50s, and Shaphan and Ahikam reach the high 40s.

The book of Daniel contains much apocalyptic and miraculous material. Moreover, since the nineteenth century, many have thought that the latter half of the book dates from the Maccabean period. Whichever way these discussions go, they must now take as proven that the bulk of the book of Daniel is awash in sixth-century coding. When assigning a date to Daniel, the powerful exilic coding should be considered alongside the early character of the Aramaic.

The books of Esther and Ruth each contain a single chapter with significant Shaphan-group coding in it. Esther clearly was written well after the Exile and Ruth may have been. The probabilities are such that neither chapter could have been coincidentally encoded, so the authors appear to be keeping alive the secrets about Shaphan and his colleagues. The sixty-five chapters in Chronicles are somewhat different. Appendix 1 shows that about a dozen of those chapters contain statistically supported stand-alone coding of the fifteen writers. Perhaps the chapters were written earlier than is generally assumed, though it is more likely that the authors were restating their own version of exilic events. When we get valid coding for those other Chronicles chapters we should have a better reading on when they were written. Taken as a whole, Chronicles stands 60 percent above the coincidence level set by nonsense words, so we need to take this Shaphan coding seriously.

About half the chapters in the Ezra-Nehemiah books are heavily encoded with Shaphan-group names, and without a doubt they were composed well after Shaphan and his team had finished work. The coding is real, not coincidental, so we must acknowledge this new fact: biblical authors used coding not just to convey present happenings but also to evoke past events.

This writer counted the number of coded groups in Hebrew Scriptures for the fifteen members of Shaphan’s group. The group totals show a notable equality of effort among those worthies. Shaphan’s colleague Micaiah had the most groups (1,677), while two of Shaphan’s sons—Ahikam and Gemariah—rounded out the top three. Jeremiah, Shaphan himself, and Micaiah’s son Achbor had the fewest, though their totals were still not far below the mean of 1,465 groups. Eleven of the fifteen members were within one standard deviation of that mean. This reinforces that the group which wrote so much of Hebrew Scripture worked closely together—and for a period of years.

This book has presented many lists and tables, as well as citing figures galore. It is a different approach to Scripture than the one normally taken. The writer hopes that readers agree the results justify this method of proceeding. Behind all the percentages, probabilities, and chi-square outcomes is one truth—Scripture was written by people, frequently by individuals in extreme circumstances. Here are two examples.

The Twenty-third Psalm, the most-beloved text in the Hebrew Bible, is signed by fourteen of the fifteen members of Shaphan’s group. Although short, the psalm still contains eighteen separate coded groups—a number that cannot be coincidental. Despite many hands in the composition, the psalmist speaks as a single individual. That person is surrounded by enemies, is in imminent danger, and faces probable death.

Psalm 23 is about a substitute king in Babylon. In Assyrian and Babylonian usage, a substitute was picked after certain types of eclipses to assume the evil omens that otherwise would attach to the reigning king. After forced drinking, the substitute was made to kneel before the sun god and swear to accept the true king’s fate. During a brief reign—one hundred days or fewer—he was clothed as a king, paraded about Babylon City, set upon a throne, had curse tablets sewn into his clothes, hosted nightly banquets, and was guarded by a retinue of several hundred men. At a convenient time, the authorities beheaded him and killed his family. After a royal funeral, they interred the victim alongside other former substitutes.

In the psalm itself, the Lord is the shepherd, not the king of Babylon (who was often portrayed as a shepherd). “Still waters” could have been the canal that ran through the city and “paths of righteousness” might have been Babylon’s Processional Way. “Prepare a table” described the special table made for each substitute and “in the presence of my enemies” told of his armed guards. The “valley of the shadow of death” probably meant the approaches to the Ishtar Gate, a death trap for assaulting forces. (The mock king’s inauguration march led through that gate.) The substitute was indeed anointed with oil, and then forced to down wine from an overflowing cup to ease the oath-taking. “Rod” can also be translated scepter. Finally, his desire to “dwell in the house of the Lord” was a wish that his body be sent to Jerusalem for burial—a wish not granted.

Who was the substitute? The strongest candidate is Asaiah, a member of the Shaphan group and the former “servant of the king” to Josiah. In the psalm, his coded spellings are the strongest among the fifteen and he also could well have been the Joshua of the Bible. Each “Joshua” is an anagram of “Asaiah,” and it may be that Asaiah led Israel’s invading army that tried to retake the Promised Land. The effort failed and the forces were slaughtered. Asaiah lived long enough to perish as a substitute for King Nebuchadnezzar. The purpose here is not to sell Asaiah as the subject of the Twenty-third Psalm. Although he is a strong candidate, subsequent coding work may turn up a better one. Instead, we want readers to understand how the Shaphan-group writers used Scripture.

For the mock king, the situation was dire. The Hebrew substitute faced an absolutely certain sentence of death—by decapitation. After suitable persuasion, he had been forced to do the most odious thing imaginable—to kneel in a pagan temple and swear oaths before the idol of the sun god. Assuming that Asaiah was this substitute, what success had he had in his life’s mission, which was to free the Promised Land and to restore worship at Jerusalem? After some initial gains, his venture had ended in disaster. The forces that Asaiah led were now dead or enslaved; he himself was soon to be the victim of a judicial execution. And in face of these realities, what did he say? “I fear no evil; for you are with me” (Ps 23:4). This summarizes the entire psalm, and indeed of much of the Scripture that Asaiah and his fourteen Shaphan-group associates created. How many believers have approached torment or death with this psalm in the eye of their minds? God alone knows. And how have they faced their plight? With the thought, “I fear no evil; for you are with me.” The gift that God gave the Shaphan writers and that they gave to us has been Scripture itself—Scripture written out of their human experience.

This book’s title contains the name “Shaphan.” Let us end with what Shaphan himself thought he had achieved. Like Asaiah, Shaphan had reason to think that his life had been a failure. He lived through the loss of Judah’s land, its temple, and its monarchy. He led an abortive revolt that proved catastrophic to Judah and to the exiles. The years of his long life were filled with stress and turmoil. The deuteronomistic view was that when Israel acted according to God’s will, the Lord made it prosper; when it did not, the Lord punished His people. Shaphan, when he was past eighty, wrote that, “By your wrath we are overwhelmed. You have set our iniquities before you, our secret sins in the light of your count-enance [Shaphan BA]” (Ps 90:7–8)—sin and be punished, and if punished one must have sinned. Yet the old man still could hope: “Satisfy us in the morning with your steadfast love [Shaphan BA], so we may rejoice and be glad all our days [Shaphan BA]” (Ps 90:14). Shaphan may have despaired, but in a way that he did not understand, his life had been a triumph. Shaphan and those he led gave us hundreds of chapters of Scripture. As God intended, they helped to shape the Hebrew Bible.