Memory Source Probe¶

Source diagnostic — instantiates Use-Time Source Attribution Calibration

Interrogates one recalled item at the moment of recall for its source cues, then applies a rule to classify where it actually came from.

A memory arrives with a feeling of certainty but no receipt — it doesn't announce whether you saw it, were told it, inferred it, or imagined it. Memory Source Probe is the point query that recovers the missing receipt for one recalled item at the moment it is used: it inspects the item's source cues — how much perceptual and contextual detail it carries, how much it bears the fingerprints of reasoning or invention — and applies a rule that maps those cues to a source class. Its defining move is that it works on a single item at recall and returns a classification, which sets it apart from a full monitoring sweep and from a confidence rubric: the probe answers "what kind of source is this one?", not "how do I audit all of them?" or "how sure am I on a scale?"

Example¶

Halfway down the street, someone freezes: did I lock the front door, or am I just remembering that I always lock it? The probe interrogates the specific memory rather than the habit. Is there perceptual detail — the cold of the key, the resistance of the bolt, the particular sound of this morning's lock? Is it anchored to context — what they were carrying, what they'd just said? Or is it only a smooth, generic sense of of course I did? Sparse, schematic cues with no sensory grain point to a reconstruction from habit, not a perception of today; the probe classifies it "inferred, not observed," and the sensible action follows — go back and check. A therapist helping a client test a vivid early "memory" that may be assembled from an old family photo runs the same probe on the same cues.^[1]

How it works¶

The probe reads cues and applies a rule, on one item, at retrieval. The empirical basis it exploits is that perceived events tend to carry more sensory and contextual detail while generated ones carry more traces of the cognitive operations that produced them; the probe checks which cues are present and rich and maps that profile to a source class through an explicit decision rule. It is a point diagnostic, not a survey — deliberately narrow, run on the item in hand rather than across the whole store — which is what keeps it cheap enough to fire at the moment of use.

Tuning parameters¶

Cue set — which cues are inspected: sensory detail, contextual anchoring, traces of reasoning effort, emotional tone. More cues yield finer classification at more cognitive cost per probe.
Rule strictness — how much cue evidence is required before an item may be called "perceived." A strict rule guards against the false "I definitely saw it"; a lenient one avoids needlessly doubting genuine memories.
Probe trigger — whether it runs on every recall or only on surprising or consequential ones, trading vigilance against the friction of second-guessing everything.
Default on ambiguity — which class an under-cued item falls to, usually the safer "uncertain / not perceived" rather than a confident guess.

When it helps, and when it misleads¶

Its strength is catching the confident memory that has no perceptual basis, immediately and at almost no cost, right where the item is about to be relied on.

Its weakness is that cues are fallible in both directions: a vivid imagining can carry rich sensory detail, and a real but old memory can be sparse — and because source cues decay faster than content, the probe reads weakest exactly on the aged memories most likely to be wrong. Its classic misuse is to run it backwards — probing only the memories whose content you already want to discredit, and reading their thin cues as "confirmation" they were imagined. The discipline that keeps it honest is to apply the same rule regardless of whether you like what the memory says, and to treat a sparse-cue result as uncertain rather than false.

How it implements the components¶

source_cue_feature_set — it reads the specific qualitative cues a recalled item carries: sensory detail, context, and the marks of cognitive operations.
use_time_attribution_rule — it applies, at the moment of recall, the decision rule that maps that cue profile to a source class.

It does not score how confident the resulting attribution is — that's Source Attribution Confidence Rubric — nor decide what to do with a suspect verdict (Hallucination Intrusion Triage), nor resolve it by fresh observation (Observation Recheck or Replication).

Instantiates: Use-Time Source Attribution Calibration — it performs the core act of classifying one item's source at the moment of use.
Sibling mechanisms: Observation Recheck or Replication · Hallucination Intrusion Triage · Reality Monitoring Checklist · Source Attribution Confidence Rubric · Borrowed Idea Attribution Scan · Chain-of-Custody or Lineage Check · Generated Content Disclosure Gate · Provenance Lookup Before Publication · Source Attribution Training Set · Source Confusion Matrix Review · Source-Label Preserving Summary Template

Notes¶

Source cues decay faster than the content they attach to, so the probe steadily loses power as a memory ages — which is exactly why a load-bearing old memory is better handed to Observation Recheck or Replication, which produces fresh evidence rather than reading a faded trace.

References¶

[1] The source-monitoring framework (Marcia Johnson and colleagues) describes how people attribute a memory to its origin by weighing its qualitative cues — perceived events tending to carry more perceptual and contextual detail, imagined ones more traces of cognitive effort. The probe operationalises that framework for a single item at the moment of use. ↩