Interpretive Saturation Log¶

Saturation monitor — instantiates Textual Close-Reading Mode

Tracks whether each further pass still changes a scoped claim, and calls the reading closed — with its open questions logged — once the passes stop paying off.

Close reading has no natural end — there is always another pass to make, another relation to notice. The Interpretive Saturation Log supplies the missing stop rule. After each pass over the artifact it records one thing: did this pass change any scoped claim, surface an unexamined material feature, or resolve a standing uncertainty? When several passes in a row return no on all three, the reading has saturated and the log calls it closed — not because everything has been said, but because further detail has stopped being consequential. Its defining output is a decision paired with a residue: the reading is declared done and its unresolved ambiguities and deliberately-excluded scope are written down, so closure is honest rather than a claim of completeness.

Example¶

A scholar is unpacking a single notoriously dense sentence from Kant's Critique of Pure Reason, and keeps a saturation log as she goes. Pass one settles the main clause and its subordinate structure — a real change, logged. Pass two notices that a qualifying phrase could attach to either of two nouns, changing what the sentence claims — a new material feature, logged. Pass three resolves the attachment by comparing a parallel construction earlier in the section — an uncertainty closed, logged. Passes four and five turn up further nuance of phrasing but nothing that alters the scoped claim she is actually making about the sentence.

At that point the log shows two blank passes in a row against all three questions, and she stops — recording, in the ledger, the one ambiguity that never resolved (a term that genuinely bears two senses) and the scope she is not claiming to have settled. The log's verdict is not the sentence is exhausted; it is further reading has stopped changing my claim, and here is what remains open.

How it works¶

Fix the scoped claims first. Saturation is always relative to what the reading is responsible for; without a scoped claim, "done" is undefinable and any pass can seem to matter.
Score each pass on three questions. Did it change a scoped claim, reveal an unexamined material feature, or resolve a recorded uncertainty? A pass that does none is marked inert.
Apply the stop rule. After a set number of consecutive inert passes, declare saturation — the marginal reading has stopped paying its way.
Log the residue at close. Record the ambiguities that never resolved and the scope deliberately left out, so closure carries its own caveats instead of implying completeness.

Tuning parameters¶

Saturation window — how many consecutive inert passes trigger the stop. A wide window is thorough but shades into endless formalism; a narrow one is efficient but risks calling a reading done while it is still moving.
Materiality threshold — how large a change has to be to count a pass as not inert. Set it high and small-but-real shifts get ignored; set it low and trivial rephrasings keep the reading open forever.
Scope tightness — how narrowly the scoped claim is drawn. A tight scope saturates sooner — legitimately, if the question is narrow; prematurely, if the scope was gerrymandered to reach closure.
Ambiguity tolerance — how much unresolved ambiguity the log will close over rather than chase. Some ambiguity is in the text and no pass will remove it; the dial sets how much of that you accept and record.

When it helps, and when it misleads¶

Its strength is that it turns "we've read enough" from a feeling into a recorded decision. It guards the two ends the archetype worries about at once: endless formalism (a reading that never closes because there is always one more relation to note) and premature closure (a reading that stops the moment a convenient answer appears). And because it closes with an ambiguity ledger, it never pretends a stopped reading is a finished one.

Its failure mode is gaming the stop rule — quietly tightening the scope or raising the materiality threshold until the current reading counts as saturated, which is premature closure wearing the costume of rigor. Its classic misuse is the mirror of that: running passes not until yield falls but until the wanted claim finally surfaces, then declaring saturation. The disciplines that keep it honest are fixing the scope and the saturation window before the passes, requiring the residue to be logged, and — where stakes justify it — having a second reader confirm that the inert passes were genuinely inert.^[1]

How it implements the components¶

interpretive_saturation_and_stop_rule — it is the stop rule: it scores each pass for marginal yield and declares the reading closed when successive passes go inert.
ambiguity_and_uncertainty_ledger — at close it records what stayed unresolved and what scope was excluded, so the stop carries its caveats.

It decides when reading ends, not what the reading is: the argument is the Micro-to-Macro Claim Matrix, its adversarial test the Counterreading Stress Review, and the context it must answer to the Contextual Reentry Memo.

Instantiates: Textual Close-Reading Mode — supplies the stop rule that lets a close reading end.
Sibling mechanisms: Counterreading Stress Review · Contextual Reentry Memo · Line-by-Line Annotation Protocol · Diction, Syntax, and Punctuation Trace · Close Reading Table · Motif Concordance and Recurrence Map · Sequence and Placement Map · Omission and Negative-Evidence Audit · Parallel-Passage Comparison · Paraphrase-Difference Test · Micro-to-Macro Claim Matrix · Edition or Variant Collation Table

Notes¶

Saturation is always relative to the scoped claim, never absolute. Widen the interpretive question and a reading the log closed can legitimately reopen — the same artifact is nowhere near exhausted, only the narrower question was. A saturation verdict should therefore always be read together with the scope it was made under; detached from that scope it means nothing.

References¶

[1] The idea is borrowed from theoretical saturation in grounded theory (Glaser & Strauss) — the point at which gathering more data stops yielding new categories. Close reading adapts it to passes over a fixed artifact: saturation is when another pass stops changing the scoped claim, not when the reader tires. ↩