# Supplementary files: assessment of 9 September 2026

This addendum updates REPORT.md's statements about unavailable figures and supplementary discussion. It preserves the earlier review as a dated record. No website edits, commit or publication.

## Material now examined

Mike supplied three files on his Desktop:

- `media-1.pptx`: 13 slides containing Figures 1, 2a, 2b, and 3-12. Inspected all embedded figure images at source resolution; inspected the native text and structure of slide 3, which contains Figure 2b as editable objects.
- `media-2.pdf`: two pages of figure legends, read in full.
- `media-3.docx`: supplementary discussion of the eleven divisions of J-Z18271 and the other Cohen branches, read from its document XML in full.

The files do not supply the advertised separate SNP tables, individual sequence calls, a complete dated tree, executable dating configuration or calibration list. Figure 12 does supply the names of the 25 proposed CB-01 markers. File hashes and a local marker-name crosscheck are in supplement-marker-check.json.

## What becomes better supported

The figure and accompanying discussion substantiate the structure previously described in the main manuscript. Within CB-01, J-Z18271 divides into eleven branches. Eight have reported priestly affiliations: FT280381, FT303739, ZS2458, BY64521, Y146901, S17446, FT34605 and FT157560. Some contain multiple religious or geographical groups and do not consist exclusively of priests. ZS2434 has Jewish affiliation but lacks a demonstrated priestly affiliation in the description. ZS237 and Y191714 are the two other divisions. A tradition of being priests among some Pontic Greek matches to Y191714 must not be assumed to mean Aaronic priesthood.

This is a substantive reason to prefer an ancestral priestly tradition to an explanation consisting only of unrelated modern surname claims. It is compatible with descendants dispersing into different communities and sometimes losing priestly status. The eleven present divisions are not eleven independent experiments: status can be inherited from common ancestors, and communities can exchange migrants.

The J2 alternative is more precisely delimited. CB-03, FGC4992, splits into FGC4942 and FT384026. FGC4942 has widespread priestly affiliation. For FT384026, the supplement describes two Mountain Jewish and two Cochin Jewish men; one Mountain Jewish man has priestly tradition, while the Cochin men's status is unknown. The authors themselves entertain either loss of status or later acquisition, and offer FGC4942 as a narrower, higher-confidence priestly grouping. Unknown status must not be coded as known non-priestly ancestry.

CB-02, E-BY8508, remains an important competing old priestly cluster: both major divisions have priestly traditions. Its smaller size does not make it biologically incapable of descending from Aaron. The genealogical framework cannot make E and J1 both ordinary biological paternal lineages from the same recent Aaron, so multiple ancient priestly traditions do not by themselves resolve which is his.

Within Mike's framework, CB-01 remains the strongest qualitative research candidate for surviving Aaronic descent. The supplementary material clarifies the basis for that preference. It supplies neither historical posterior odds nor an unconditional current descendant minimum.

## What the growth charts actually show

Figure 3 plots counts of branching by century. Figure 4 has the corresponding cumulative pattern. These are two presentations of the same historical reconstruction, not independent support for population growth. The first post-root plotted count is three; it should not casually be described as three independently dated ancestral men.

The graphics display an early concentration and later resumption of branching, but show no uncertainty bands, alternative date assignments or sampling correction. The caption expressly limits the analysis to branches involving tested modern individuals. No event-level table or calculation is supplied in these three files, so the counts cannot be reconstructed independently from them. A real demographic interpretation would need survival, sampling and date uncertainty handled explicitly.

The plotted cumulative value is about 60 at the turn of the era, consistent with the manuscript's reported tally. It is not a census of 60 priests, and it cannot establish that every represented ancestor then held priestly status. Nor does it establish a current minimum of Aaronic descendants before the ancestral identification has been made.

## Figure 12 and the proposed stem

Figure 12 places CB-01 (ZS222), dated about 833 BCE, beneath ZS227, dated about 2150 BCE with a stated interval of 2478-1864 BCE. The other child is a basal ZS227 individual. It lists 25 markers on the CB-01 side, but it does not order their mutations chronologically or identify a biblical man anywhere along the segment.

**Correction after an all-occurrences recheck:** 17 of the 25 Figure 12 marker names match J-Y3088, not 16. FGC13874 occurs in its defining block `Y3088/FGC13874/Z18270`. The identical alias block also occurs at J-FT115146 under J2; that second occurrence does not cancel its occurrence at J-Y3088. The original JSON already recorded both, but my written summary omitted the J-Y3088 match and understated the agreement. This correction strengthens the marker reconciliation with ZS222; FGC13874 is not evidence of a misplaced Figure 12 marker.

The complete tally is **17 names at J-Y3088, two at its immediate parent J-S20075 (FGC13893 and ZS2405), and six names not found**. No marker is found exclusively in the unrelated J2 branch. These are counts of supplied names, not necessarily independent mutations. The six unmatched names are BY32807, BZ4063, FT2417, Y5408, ZS243 and ZS8006. All were rechecked across every node's `snps` field with word-boundary matching, including slash-separated aliases. None appears literally, including inside an alias block. Their names also have no entry in the locally saved Sahakyan supplementary table S2, so that table supplies no alternative name to bridge the gap. This does not establish that their underlying mutations are absent: a differently named mutation requires an independently supported alias or coordinate-and-allele mapping.

The reproducible [recheck-markers.py](recheck-markers.py) records every matching node, ancestral path and matching alias block in [supplement-marker-recheck.json](supplement-marker-recheck.json). The original JSON is retained as the earlier record. These are name comparisons, not allele checks.

The comparison means we cannot currently treat all 25 markers as a verified invariant block common to both trees. Neither the figure nor the local comparison justifies calling the segment the Adam-to-Noah stem. On the paper's own point dates it is substantially later than that interval in Mike's chronology. A genuine Aaronic descendant could carry mutations that arose after Noah, and Noah could lie farther upstream.

The later Sahakyan table reconciliation establishes B877 = S8223 = ZS222. B877 itself is absent from the dump's defining SNP fields; the bridge runs through S8223, which occurs at J-Y3088 in `S8223/ZS222`. This identifies the mutation across naming systems, not every historical branch boundary or ancient carrier. An ancient member of a broader branch remains distinct from an ancient priest identified as CB-01.

### Keep the dating estimates separate

For the requested comparison, use reference year **2025 CE** and explicitly subtract the age in years before present. This is the approximate subtraction convention used here, not a claim that the dump is calibrated to the radiocarbon reference year 1950. Rounded BC labels below do not introduce a one-year no-year-zero correction.

| Estimate | Explicit conversion or source | Approximate date and interval | Includes 1266 BC? |
| --- | --- | --- | --- |
| Local J-Y3088 TMRCA | 2025 − 2800 = −775; interval endpoints 2025 − 3200 = −1175 and 2025 − 2300 = −275 | 775 BC; 1175–275 BC | No: the older endpoint is 91 years later |
| Lipson Table 3 | Already reported in BCE, no BP subtraction | 834 BC; 1291–440 BC | Yes |

Reference-year qualification: clade-denominator/FROZEN-CRITERIA.md actually pairs 4076–4926 BP with 2050–2900 BC, implying a **2026** reference under direct subtraction. Using that convention instead gives 2026 − 2800 = −774 and interval endpoints −1174 and −274, approximately 774 BC and 1174–274 BC. Its older endpoint is 92 years after 1266 BC. The one-year convention difference does not change the conclusion; the table above explicitly uses 2025 rather than silently attributing it to that frozen record.

Lipson's prose instead gives 833 BC and 1289–440 BC; that small reporting difference also leaves 1266 BC within its interval. Do not merge either paper interval with the dump interval. The local interval does not reach the Exodus date. That does not exclude descent from Aaron: the common ancestor of a surviving subfamily can be younger than Aaron. Conversely, an interval overlapping his period does not identify its ancestor as Aaron.

## Specific corrections and ambiguities

- The Figure 1 caption says FT246134, whereas the figure and other descriptions say FT246135.
- The Figure 2b caption says eleven primary sub-branches. Its native slide objects show four: FT246365, Z11314, FT246135*, and FT266880. Four agrees with the main paper.
- Figure 7 dates FT395341 to 1515 BCE beneath a parent dated 1 BCE. This reverses temporal ancestry. The legend places the child in the fifteenth-sixteenth centuries CE, supporting a BCE/CE labelling error rather than an extraordinary biological result.
- In the CB-03 discussion, the sentence assigning 'CB-03 (c. 55 CE)' to the western group evidently means FGC4942, as Figure 6 shows; the broader CB-03 is dated 539 BCE.
- The CB-05 discussion switches to 'CB-04' in an upstream description. Context indicates CB-05.
- Dates vary slightly between tables, prose and figures. These variations are much smaller than the stated intervals, but exact-year historical matches should not be claimed.

These errors are reasons to check labels and calculations, not reasons to discard the positive association between the major J1 cluster and priestly tradition.

## Consequence for the wider inquiry

The new material makes an ancient priestly origin for CB-01 easier to assess and gives a clearer reason to investigate J1 before J2 in the Aaron route. It does not supply an independently identified Aaronic ancestor. Under the accepted biblical genealogy, establishing such an ancestor would connect his biological descendants to Levi, Noah and Adam. It would favour placement within the ordinary J1 tree, but would still not establish exact sequence copying from a particular man.

The productive outstanding checks are now specific: reconcile the 25 markers by allele and coordinate, obtain the full dated tree and calibration inputs, and obtain sample-level status and assignment provenance. We should not repeat 'the figures are missing'; they are now examined. The historical identification and reproducible dating inputs remain missing.