Reconstructing the Great Stemma


Introduction: Palaeographic clues can be discovered which reveal the original appearance of the Great Stemma document before it became disordered through repeated scribal copying. To my knowledge there has never been an attempt before now to reconstruct the document as it appeared during antiquity.

Although Zaluska's specialization was art history, she adopted a philological approach in her analysis of the document in the 1980s, noting that its text was static, but evolved:

...son histoire ultérieure consiste en grande partie en des corrections et des remaniements d'après la Vulgate ... Les efforts des recenseurs, pourtant, ne se sont pas bornés à la simple révision du texte reçu et ainsi différentes adjonctions et enrichissements se sont greffés autour de l'ossature commune.[*]Zaluska, Stemmata, 143-4.

The following note, which is necessarily technical and requires some familiarity with medieval, antiquity and biblical studies, takes up a puzzle left unsolved by Zaluska and not only proposes a solution, but also demonstrates that the Great Stemma began its existence as an integrated chart on a single roll and was only sectioned later into codex pages. If you are mainly interested in the date and form of the Great Stemma, you may wish to skip this page, but if you are interested in manuscript detective work, read on.

A further area of prime interest - precisely because the arrangement has become seriously corrupted - is the Horrite chiefs list , which Zaluska was more successful in analysing.

A section from Kings: The analysis which follows concentrates on a section that constitutes about 2 to 3 per cent of the Roda stemma (see the high-resolution image) and is illustrated by a detailed drawing which the reader should look at before taking in the argumentation below. The drawing is based on the Roda manuscript, but the text has been slightly emended using the Plutei manuscript wherever that seemed closer to what we read in the Liber Genealogus. This is not intended to be read as a canonical text: a future editor will have to settle the best text. I merely wish to set out a text that makes the arguments presented below comprehensible.

The section chosen for analysis spans the tops of plates 11 to 13. It displays 16 generations of the ancestry of Christ, running left to right, starting with "Asus genuit Iosafat" (English: Asa fathered Jehoshaphat), culminating in "Azor genuit Sathoc" (Azor fathered Zadok).

This section is typical of the peculiar wiring at the Great Stemma's upper margin: the connecting line does not proceed directly from father to son in series, with roundels arranged like beads on a string as one might expect. Instead the roundels are arranged with their connectors suspended from a single top line or bar, like garments hung to dry on a clothesline. This resembles the arrangement of modules in an electrical circuit diagram when they are shown as connected in parallel, via a bus bar, rather than in series.

Whether this pattern is the one devised by the Great Stemma's author or the result of an error in scribal transmission remains an open question. As far as I know, most versions of the Great Stemma share this clothesline-style arrangement, but a derivative 11th-century stemma, which is bound into the Codex Amiatinus III in Florence, corrects this, beading the roundels for the main lines along the same connector.

The father-son succession of Judaean kings in this section of the stemma is that stated in the Gospel of Matthew, which conflicts with the father-son sequence which had been set out in 2 Kings and 1 Chronicles, since it leaves out a total of four names. Those names are no longer present in the Great Stemma as we know it today, but I can demonstrate that the Great Stemma's author must in fact have written them on the page.

To make this discussion easier to understand, I have added these names in red text to the detailed graphic. I have given the four king names missed from the Gospel of Matthew in their Vulgate Latin spellings: Ahazias, Ioas, Amasias and Ioiachim. I have entered some further missing names in red text, using the spellings in the G recension of the Liber Genealogus, which is almost certainly a transcript from the lost ur-form of the Great Stemma. (The G recension of the Liber Genealogus is reproduced on this website.)

1: For each king of Judah, a wife is shown in a connected roundel, just below the king's own roundel, for example, "Cazoba uxor Asa" (Azubah, wife of Asa). (It might perhaps be noted that the concept of uxor is a Christian construct here: the bible merely identifies these persons as the mothers of the next ruler, implying that these were harem women - wives or concubines - of the previous ruler. But that does not affect the palaeographical issues.) The first anomaly to note is a wide blank space in the Roda version beneath the roundels for the first three kings ("Asus", Iosafat and Ioran) of plate 11. The correct wives' names of this trio have been inserted into roundels, but these are placed well down the page. Only one of the three, that for Cazoba (here "Zazo"), is connected by a line to its king. One suspects that the Roda editor faced some kind of problem interpreting the model and has left that problem unresolved. The later Alpha versions, as well as Epsilon, have re-aligned the roundels and closed up the gap, but the Roda scribe has preserved the evidence of this dilemma. As we proceed, we shall understand what this difficulty of interpretation was, and how it is linked to a much larger confusion.

2: All of the 13 remaining kings are shown with wives, but the oddity is that none of those 13 wives conform with the names stated in the Old Testament and the Liber Genealogus. This issue was first explored by Zaluska:

... nous remarquons que les rois de Juda (de Salomon à Jéchonias) et leurs cinq premiers successeurs (de Salathiel à Azor) sont accompagnés chacun de leur épouse. Celles-ci ne proviennent pas de Mathieu bien entendu. À première vue ces noms suscitent l'intérêt lorsqu'on sait que la Bible ne donne pas les noms des femmes de Jéchonias, ni de Salathiel, ni de Zorababel, ni à plus forte raison de celles de leurs descendants, ces derniers personnages n'étant pas attestés par l'Ancien Testament. On pourrait donc espérer que ces noms proviennent d'une source apocryphe peu connue, mais leur analyse permet d'écarter une telle supposition, de même qu'elle jette une nouvelle lumière sur les origines obscures de nos tables. On note en effet que ces noms, fondés pour la plupart sur des formes de la Vieille Latine, sont tous d'origine biblique et que seule leur distribution ne l'est pas. Étant donné que la généalogie selon Mathieu est plus courte que la liste biblique des rois et qu'en outre les fils de Josias étaient nés de deux mères différentes, cinq noms de femmes ont pu être en quelque sorte récupérés. Ceci a permis la "création" de cinq nouvelles épouses.

In the drawing, I have added in red print the female names as they are set out in the Liber Genealogus. Scripture and the Liber do not identify any wives for Iheconias and his five descendants. So where do the false names come from? Careful comparison of the names in red and the content of the roundels reveals that eight of the wife names have simply been transposed four positions to the right from their correct positions on plate 12. (At the moment I am only counting couples who are considered as ancestral to Jesus by Matthew.)

In my explanatory drawing, these eight roundels, and their original positions according to the textual sources, are marked by the rightmost, horizontally elongated grey box of eight roundels. The overlapping box containing seven roundels at left marks the places where they should have appeared. It is to be noted that two "wives" of Iosias are named in 2 Kings of whom Zebidah (here Zeccora) is likely to have been considered by a biblical genealogist as the source of the "royal line" since it was her grandson who ranked as the last king of Judah, eating in imperial palace in Babylon (2 Kings 25; 27-30).

One conjectures that both Amitaal and Zeccora were shown, appearing in either a column or overlapping, in the space beneath Iosias in the original diagram.

This error of the eight roundels is so wide as to suggest the blunder must have occurred while the Great Stemma was being copied from one large roll onto another roll, and not while it was being copied page by page to new codex folios. A scribe would surely have notice such an error when taking his bearings to begin a fresh page, and indeed it seems to me there is no convincing way to imagine such corruption ocurring after the point after the diagram had been sectioned into separate pages.

The error is thus our most convincing evidence that the diagram was originally authored in roll form, and this conclusion is supported by the observation that the diagram has an overall structure, with its fila and closure on the final page, which begs to be displayed in totality in a single view, not broken up into a series of constricted "windows" where its flow is withheld from view.

Zeccora has been displaced all the way into what is now Plate 13 of the Roda manuscript. In Epsilon, the same defect is spread across plates 12, 13 and 14 (with Ieiadada, Amitaal, Zecora, Nasta, uxor Sadoch and uxor Achim on the last), further confirming that the mix-up preceded the division of the stemma into the sections we see today.

3: We must now consider where the copyist found the four names (Coboda, Ioade, Gotolia and Iecelia in the Roda manuscript) to fill the gap of four positions under the names Ocias, Ioatam, Agaz and Ezezias. When we study the Vetus Latina and Septuagint texts and the Liber Genealogus, this becomes readily apparent. Three of these names, Chobodalla, Ioadae and Ieccillia, are stated in the LG to be the wives of three Judaean kings omitted from the genealogy in the Gospel of Matthew. If we accept that the Liber Genealogus is an ecprhasis of the Great Stemma, this provides aditional evidence that the original version of the Great Stemma must have somehow commented on the divergence between the Old Testament and the Gospel of Matthew. It almost certainly contained, somewhere on the page, roundels for these three absent kings.

The Septuagint forms of these names are Abia, Ioadin and Chalia. The differences in orthography are interesting and suggest that the Vetus Latina text from which the three names were obtained may have been modified in the light of a Hebrew source.

The following tabulation compares the known Latin spellings for these 14 wives, listing the dominant spellings in early recensions of the Great Stemma (GS), the attested Vetus Latina spellings (mainly from the Liber Genealogus, in one case from the Palimpsest in Naples), Septuagint forms and transliterations (taken from NETS) and the biblical references and Vulgate equivalents:

  Wife in Stemma attested in Vetus Latina LXX / NETS Vulgate King  
a Macca Maccha (LG) μααχα Maacha Maacha Roboam 1Ki 15:2
b --- An(n)a (LG, GM Gn15:13) ανα Ana Maacha Abia(m) 1Ki 15:10
c Cazoba Gaziba (LG) αζουβα Azouba Azuba Asa 1Ki 22:42
d --- Faceae (Fache) (LG) --- --- Iosaphat 2Ki 8:16
e Gotholiah Gotholia (LG, GM Gn10:36) γοθολια Gotholia Athalia Ioran 2Ki 8:26
f Cobo(c)da Chobodalla (LG) αβια Abia Sebia Ochozias (LG) 2Ki 12:1
g Ioade/Ioede Ioadae (LG) ιωαδιν Ioadin Ioaden Ioas (LG) 2Ki 14:2
h Iecelia Ieccillia (LG) χαλια Chalia Iecelia Amessias (LG) 2Ki 15:2
1 Darusa Darvia (LG, Palimpsest) ιερουσα Ierousa Hierusa Ozias 2Ki 15:33
2 --- Saba (LG) --- --- Ioatam 2Ki 16:2
3 Rabuti Rabutthi (LG) αβου Abou Abi Agaz 2Ki 18:2
4 Ebsiba Epsiba (LG) οψιβα Hopsiba Aphsiba Ezezias 2Ki 21:1
5 Musaella Mastelabitaba (LG) μεσολλαμ Mesollam Mesallemeth Manasse 2Ki 21:19
6 Iedadida Ieddadida (LG) ιεδιδα Iedida Idida Amon 2Ki 22:1
7 Amitaal Amithal (LG) αμιταλ Hamital Amithal Iosias 2Ki 23:31

Some of the above spellings may differ from those in my drawing. This is because this page was modified in 2014, four years after the graphic was made, and I have not had time to reconcile the two. The complete Vetus Latina text of 1-2 Kings is mostly lost, but one of the irregular names above, Daruia (to be read "Darvia"), can be confirmed using the fragmentary text of the Palimpsestus Vindobonensis (Naples, published 1983 by Fischer, PDF). The Great Stemma's version of this name, Darusa, differs from Daruia by only one letter. It is most likely to be a simple misreading by a scribe, but a case might be made out that it is a deliberate back-formation, where the I has been changed to an S in a timid attempt to approach the Vulgate form, Hierusa.[*]This table will not be complete until all the references have also been checked in Beuron Abbey's Vetus Latina card index: images of the cards, commercialized by Brepols, are accessible in a few research libraries.

Gotholia is an attested name from the Vetus Latina, as evidenced by the Glosas Marginales (GM in the table above) in the Spanish bibles. Ana is also preserved in one such gloss. The absence of Ana (row b, an indubitable element of the LXX/Vetus Latina tradition) from the diagram (although she is preserved in the Liber Genealogus) is one of the more notable mysteries of the Great Stemma as we see it in the manuscripts today. It seems implausible that Ana would have been omitted when so much else here conforms to the Vetus Latina. Ana's absence (and indeed that of Faceae and Saba) is likely to be the result of editorial intervention late in the transmission of the Great Stemma. [*]Moreno Hernández, Antonio. Las glosas marginales de Vetus Latina en las Biblias Vulgatas Españolas: 1-2 Reyes. Madrid: CSIC Press, 1992.

4: Reinforcing proof for the roll hypothesis comes from another corruption: the mysterious roundel inscribed "Iosobe filius Ioram". This is attached to a king Jehoram of the Northern Kingdom (Israel or Samaria). No such relationship is recorded in scripture or any early Christian writings. However the Second Book of Kings presents the history of a ruler in the neighbouring Southern Kingdom (Judah) bearing the same name, Jehoram. 2 Kings 11: 1-3 describes how his daughter, Jehosheba, managed to save and hide her royal nephew, Jehoash, while all his siblings were massacred. The name Iosobe in the stemma closely resembles the Vulgate Latin form of Jehosheba, Iosaba. The transformation from a filia to a filius is probably an error of negligence by a copyist mechanically writing the same word filius hundreds of times on the chart. If the original text was "Iosabe filia Ioram", it is easy to see why a scribe would have then been misled into drawing a connecting line to the Ioram of the Northern Kingdom (Samaria) and relocating the roundel slightly to be directly above that of the putative father.

To a biblical scholar, Jehosheba is an important character for the understanding of the Judaean political succession, because by her authority and courage she restored the old royal line. Her inclusion is not by whim, but makes it plain that her brother Ahazias and her nephew Joash (Vulgate Latin form, Ioas) were present on the chart too and that the Great Stemma was more than a genealogy. It was also a detailed graphic representation of the political history of the Jews.

All this evidence indicates that the Great Stemma once contained a diagram midway between the segments on the Southern and Northern kingdoms which detailing the exact political succession between Jehoram and Uzziah (Great Stemma: Ocias or Ozias). This analysis would have surely been one of the most heavily used parts of the Great Stemma in any Late Antique library, logically explaining to the bible scholar the succession within the two Jewish kingdoms which is so confusingly set out in the two books of Kings. The fact that this useful diagram has been lost may perhaps also be taken as a sign that it was heavily handled and more quickly destroyed by folding, smudging and the chemical corrosion from the secretions on a succession of hands.

The diagram must have been irretrievably lost when a copyist, lacking the authority, the time or perhaps the enthusiasm to restore the correct layout, bodged together the visible scraps into a kind of order to reproduce the stemma yet again in chart form. The sideways displacement of 13 roundels suggests either an extreme lapse in attention, or else that the pieces were only available as torn shreds which could scarcely be recomposed into any intelligent format. By the time that the Great Stemma was converted into a document divided into sections suitable for a codex, the original must have no longer existed and only the single, very defective reproduction was available. This heavily corrupted state and the difficulty of repairing the text must surely have been a motivation for scholars such as Petrus of Poitiers in the 13th century to discard the Great Stemma entirely and to commence a compilation in chart form starting out anew.

5: The Liber Genealogus also witnesses to several other curious emendations in the uxor names. Asa's mother is identified there as Anna, not the second Maaca of 1 Kings 15:19. Ioram's mother is identified as Faceae, not omitted as in 2 Kings 8:16. Ahaz's mother is identified as Saba, not omitted as in 2 Kings 16:2. None of these three corruptions appear in our Great Stemma manuscripts, but it is conceivable that they come from an early version of the diagram.

6: We now proceed to discuss clues to other lacunae in this area of the stemma. One of the names here is of major interest in the political history of Judah: Gotolia or Catholia. This is the Vetus Latina name, based on the Septuagint, of a character we know as Athalia. As portrayed in 2 Kings, she was an extraordinarily bloodthirsty ruler who seized power after her son died of war wounds (2 Kings 11:3). It was Athalia who ordered the massacre of her own grandchildren and the rest of the royal family to consolidate her own power, only to be foiled by Jehosheba (above) and be ousted and killed six years later. As might logically be expected, Athalia figures twice in the Great Stemma, firstly as the wife of Jehoram (Gotolia uxor Ioram) and later as a ruler in her own right in the timeline of kings. The latter mention has been partly erased, but has not been entirely lost, since this annotation has been accidentally shifted by a copyist and converted into an uxor roundel for king Ahaz (Roda: Gotolia uxor Agaz), a person who ruled Judah 100 years later. It is probable that the original annotation was at mid-page - it may have taken the form of an arch or a roundel - and was part of a timeline of some nature, possibly with a reign duration for the queen, as for example, *Gotolia annos VI.

7: The Gospel of Matthew again skips an ancestral king when it comes to Jehoiakim (Vulgate Latin: Ioiachim; Liber Genealogus: Ioachim). The Great Stemma author evidently included this Judaean king in an additional political succession diagram, which must have been placed below the genealogical succession as given in Matthew. To make the following discussion easier to follow, I have tabulated the kings' names below in the three language versions.

  Liber Genealogus Vulgate English
  LG Ruler His Parents Ruler His Parents Ruler His Parents
1 409 Iosias Amos & Ieddadida Iosias Amon & Idida Josiah (killed in battle against Egyptians) son of Amon and Jedidah
2 413 Ioas Iosias & Amital Ioahaz (half-brother to Eliachim below) Iosias & Amital Shallum (personal name) Jehoahaz (throne name) (taken captive to Egypt) son of Josiah and Hamutal
3 411 Ioachim Iosias & Zecchora/Ieccora Eliachim Ioiachim (half-brother of Ioahaz above) Iosias & Zebida Eliakim (personal name) Jehoiakim (throne name) son of Josiah and Zebidah
4 416 Iecchonias (ductus est in Babylonia) Ioachim & [omitted] Ioiachin alias Iechonias (taken to Babylon) Ioiachim & Naestha Jehoiachin alias Jechonias (taken to Babylon as vassal) son of Jehoiakim and Nehushta
5 413 Sedecia Iosias & Amital Matthanias Sedecia Iosias & Amital Mattaniah (personal name) Zedekiah (throne name) (the uncle of Jehoiachin, he was blinded and chained, taken to Babylon) son of Josiah and Hamital
6 419 Salatiel Iecchonias & ... Salathihel Iechonias & ... Shealtiel son of Jechoniah

This is a particularly difficult series to represent graphically using a stemma, given that three of these men are sons of Josiah by two different mothers and that the political leadership then jumps backwards a generation when Zedekiah succeeds his nephew. The generation skipped by Matthew would appear to be reign number 3 of the tabulation: Jehoiakim/Ioiachim.[*]This is the interpretation offered by Prat, Ferdinand. "2. Généalogie De Jésus-Christ." Dictionnaire De La Bible (1926) at columnn 1, page 167.

It is plain that reign 3 must have been somehow present in early versions of the Great Stemma, even if Matthew omits it, since the name of this king's mother, Zebidah (2 Kings, 23:36), can be found on the extant manuscripts in the corrupted form Zeccora as we have noted above, and in my expanded discussion of this issue on a separate page of this website.[*]One of the Liber Genealogus manuscripts offers the spelling Zecchora, so it is plausible to suggest that this was a current Vetus Latina form. The Septuagint's form of the same name is even more distant from Zebidah: Ieldaph.

It is inconceivable that the Great Stemma author would have introduced both wives of Iosias into the diagram while deliberately excluding their three sons. Zebidah plays no other role in the biblical account except as a giver of birth, so her son must have been recognized and mentioned as a separate person in the Great Stemma. The copyist has given Zebidah in marriage to Eliacim, a person attributed to the post-Babylon period.

8: Judaean reigns 2 and 5 are also present in the diagram, but have slipped away by scribal error to become appended to the list of kings of Samaria, the northern kingdom, with the legends Ioacas filius Iosie and Sedechias filius Iosie. In the Roda manuscript Jehoahaz (here: Ioacus) and Zedekiah (here: Sedecia) are on the same page, but in other manuscripts they move to the next page. The presence of their mother Hamital (here: Amitaal) offers additional confirmation that the reigns of both her sons were accounted for in the original chart. She has been given in marriage to a post-Babylon figure, Abiut.

If the Liber Genealogus is any guide, Ioacas may perhaps have been shortened to Ioas in one early version of the diagram, but the name should not to be confused with that of the king usually denominated Ioas in Vetus Latina texts: Joash/Jehoash of Israel at 2 Kgs 11:2 and 12:1- (nor for that matter with Jehoash of Samaria, 2 Kgs 13:10).

9: A gloss on the chart appears to contend that Ieconias and Zedekiah, who was blinded and dragged to Babylon in chains, are not two separate persons but merely one person under two names: Ieconias qui est Sedecius dedactus in Babiloniam or Hiechonias qui est Sedechius ductus in Babilonia.

It is not entirely clear who is meant by Ieconias here. Is it the Ioiachim (reign 3) who was unaccountably omitted from Matthew's list but is represented on the diagram by his mother, or the Ioiachin (reign 4) who was deported to Babylon as a vassal, figures in Matthew's list of Jesus's ancestry and has a roundel of his own? Only one consonant separates them in the Greek and early Latin text of 2 Kings and 2 Chronicles. This issue is discussed in more detail on my page dealing with exegetical issues.

10: Nehushta, mother of Jehoiachin (reign 4) is also present. She has been falsely attached to Azor as Nastha (beta) or Nasta (epsilon/sigma). This orthography is fairly close to the Vulgate Latin form Naestha and the Septuagint Nestha. Though we cannot appeal in this case to the Liber Genealogus (this mother is not listed in extant manuscripts of that book), the similarity is already so striking that we do not require any additional documentary support to make this connection.

11: Whether the proto-text ever would have suggested that the transport to Babylon marked the start of a new age of the world seems to me highly doubtful. Zaluska, in her analysis, felt that this detail had perhaps previously been there and had gone missing: she pointed out that the departure of Zedekiah marked the start of the Fifth Age. But I do not think such an argument can be sustained. The theory of the Seven Ages of the world derives from Augustine of Hippo. It does not seem to influence the Great Stemma at all, which should be hardly surprising if one adopts the view that the diagram was compiled before (or only shortly after) the publication of Augustine's City of God. The Stemma certainly takes an interest in "universal time" and synchronisms in history, but its theoretical underpinnings, both in its first version and in the Eusebian revision, are different from Augustine's. Our manuscripts contain calculations of Eusebian origin for the five first ages (to the flood 2,242; to Abraham 942; to Moses 505; to David 435; to the restoration of the temple 560) but these are additions drawn from the Ordo Annorum Mundi and they surely date only from the Eusebian revision of the diagram. These issues are discussed on my chronology pages.

12: What happened next: As we noted above, a 10th-century manuscript, the Codice of Roda, at least preserves a certain awkwardness of layout from which an alert reader might pick up a clue that all is not right. But no copyist was ever able to penetrate the mix-up and restore the correct order.

Later manuscripts only compound the corruption. The kings always remain connected to the wrong wives. Sigma repeats the same king-wife misconnections all the way to Eliachim/Zechora and Azor/Nasta. The Delta manuscripts (I have not seen the Calahorra bible, but presume it is no different) adopt their own unique layout, but the content is just as misconnected as in all the other Spanish manuscripts. The Beta manuscripts attempt to smooth over the difficulty by aligning all the wife roundels in a neat row on the same height, but only make the problem worse. Facundus for example, adds an extra roundel for the wife of Saddoc ("uxor Saddoc") so that there will be no untidy gap at the end of the row.

In the 20th century, this anomaly was noticed and described by Zaluska, but she was unable to offer any explanation for it. In her article on the stemma in 1984 she speculated, correctly as it turns out, that the author of the Great Stemma would surely have wanted to mention the four kings of Israel which are omitted from the Gospel of Matthew. She pointed out that non-scriptural wives for the period the Babylonian exile were named in the document. But her only suggestion towards a solution was the observation that the names did seem biblical in type and must have somehow been drawn into the document by some unexplained process. She did not venture any suggestion as to how this happened. Thanks to the analysis above, the mystery behind this muddle has been solved at last.

Back to History Table of Contents

Back to Macro-Typography on the Web