# Authorship and comparison controls

## The attribution question remains open

No institutional record or publication located in this investigation attributes the present sheet to Barent Fabritius, Carel Fabritius, Gerard ter Borch II, Moses ter Borch, Willem Drost, Rembrandt, or another named artist. A comparison can support a question or eliminate an overstatement; it does not transfer an attribution from one sheet to another.

## Barent Fabritius

**SOURCE-REPORTED.** A small institutional and scholarly comparison corpus exists for drawings currently associated with Barent, but its boundaries are unstable and some historical Barent attributions are now rejected or qualified. Identified Barent records show a wider range of mixed media than can be diagnosed from the present sheet's photograph.

**INFERRED.** The available present-sheet image does not presently show a sufficiently distinctive correspondence with the controlled Barent corpus to support an attribution. This does not exclude Barent: media and sequence in the present sheet remain unexamined, and the secure drawing corpus is limited.

## Carel Fabritius and the family boundary

Carel is a necessary near-neighbour control, not a secure negative corpus. The published and institutional boundaries of drawings associated with Carel are themselves qualified. A familiar name, a hat, a youth, paper, or a Rembrandt-school context cannot decide between Carel, Barent, another pupil, or an imitator.

## Rembrandt, pupils, and anonymous school

The current evidence remains compatible with several Rembrandt-school possibilities, including an anonymous school or circle hand and a later derivation. It does not support direct Rembrandt intervention. Historical workshop practice provides context for repeated models and copies, but does not supply an object-specific link to the present sheet.

## How comparisons should be weighed

| Comparison type | What it can contribute | What it cannot contribute alone |
|---|---|---|
| Securely catalogued drawing of a named artist | A controlled range of medium, line, date, and working method | Automatic attribution of the present sheet. |
| Qualified or disputed drawing | A boundary case and alternative hypotheses | A secure corpus anchor. |
| Same composition or close design | A reason to test derivation or common source | Same hand, direction, or date. |
| Shared medium, hat, or format | Contextual compatibility | A diagnostic authorship criterion. |
| Historic inscription or sale heading | Attribution history | Authographic signature or provenance without object verification. |

## Current conclusion

The defensible public description is **“unattributed drawing under comparative research.”** The strongest new question concerns the relation of the RKD design to the Ter Borch family/workshop context; this is a research priority about the comparator's design, not a new attribution of the present sheet.

**References:** Barent corpus report; Carel boundary report; British Museum and Rijksmuseum comparison records; Ter Borch audit; Drost assessment. Full citations are in [Sources and next tests](10-sources-and-next-tests.md).