Hippocampus/medial temporal lobe, prefrontal cortex, aPFC/frontopolar cortex, striatum, prospective-memory/timing systems
In plain language
Something can be part of the past and still matter now in several different senses. An earlier event may still physically affect a present outcome; information about it may still be available; that information may have been retained or reconstructed from current clues; it may change what the correct decision should be; or it may actually influence what the system does.
The paper treats those as separate questions. Its historical-relevance audit asks whether the past still has a controlled causal effect, what historical information is available now and where it came from, whether that information should matter to the present decision, and whether it actually controls behavior.
Abstract
Something can have happened earlier without remaining present in every scientifically relevant sense. An earlier distinction may still make a controlled causal difference to a later target; information about it may be available now; that information may be retained from the earlier event or reconstructed from present cues; it may change what an agent should do; and it may or may not actually control behavior. Ordinary expressions such as "it still matters" and "the system remembers it" can collapse these relations. This article develops a historical-relevance audit that keeps them separate without proposing new primitives for causation, memory, or decision making. Two constructive cases motivate the synthesis. First, aggregate endpoint control need not identify which temporally indexed intervention remains causally consequential. Second, a cognitive process can possess an objective and recoverable history without thereby possessing an online representation of its position in that history, and available temporal information need not be authoritative for continuation or closure. A matched-present construction then shows how systems can produce the same ordinary observations while differing in external causal persistence, provenance of present historical information, decision authority, and realized use. The methodological consequence is simple: evidence identifying one present relation to history does not identify the others without additional assumptions or interventions. No participant data are reported; the proposed cognitive dissociations remain open empirical questions.
Why it matters
Ordinary phrases such as “it still matters,” “the system remembers,” or “history affected the result” can hide very different scientific claims. A correct report does not prove that the information was retained from the original event, and a historical fact being available does not prove that it should influence the current decision or that it actually does.
The same problem appears across causal inference, memory, temporal cognition, credit assignment, and control: evidence that identifies one present relation to history can be silently promoted into another. The audit makes the target of each claim explicit so the necessary comparison or intervention can be matched to the relation actually being asserted.
The framework does not propose a new memory system, causal primitive, or decision mechanism. Its value depends on whether separating these historical relations prevents invalid inference and produces clearer tests than treating “the past still matters” as one undifferentiated condition.