# C905 — Separate common topology from source-specific measurement

## Question

What does the full NT insertion response add to the earlier Cainan comparisons?

## Inputs

```json
{}
```

## Sources

```json
[
  "C864",
  "C831 insertion and BJ deletion results"
]
```

## Opened utc

2026-09-28T02:55:25.784204+00:00

## Results

```json
{
  "transfer": {
    "path": "model/topology_measure_transfer.json",
    "sha256": "13776a4ab8d0a07e54ce29312bdf2eda35338b9278b39a4f75b1eb5616ebd907",
    "bytes": 1698
  },
  "upstream_delta": -130,
  "downstream_delta": 0
}
```

## Finding

The complete NT/MT field changes by−130 at every common node through Arphaxad and by0 from Shelah onward. This makes the insertion support explicit. The130,460,70 and BJ57 measures retain their source-specific roles even when name bookkeeping aligns.

## Reassessment

Generalize the retained-biography distinction as a subdivision rule.

## Checks

```json
{
  "support": true,
  "four_roles": true
}
```

## Closed utc

2026-09-28T02:55:25.785485+00:00

## Record provenance

Exact record hash: `d61dc9a2affee12b43710c0da765738e1f7d31356015e375dfdab562fcd63337`

Source journal: `Research_Cycles/C0832_C0931/journal.json`