Based on the sources, the detection of “weak signals” or outliers is interpreted as a dynamic interaction between individual sensory/cognitive capability and the conceptual “net”—the boundaries, metaphors, and logic levels—employed by the observer. While individual capability sets the “inelastic limit” of perception, the “net” determines what is eventually classified as a signal versus irrelevant noise.

1. Individual Sensory and Cognitive Capability

The sources acknowledge that human perception is the starting point for all systems inquiry, but it is inherently limited:

The Inelastic Limit: Human capability is restricted by an “inelastic limit” of absolute judgment, often cited as the “magic number seven,” which forces individuals to “chunk” information into larger, less detailed units[1][2].

Tacit Knowing: Emery argues that individuals possess vast amounts of tacit knowledge extracted from their environments that may not be immediately conscious[3][4]. Detecting a “weak signal” may involve a sensory “extraction” from the environment that has not yet been “abstracted” into formal knowledge[4][5].

Interest-Driven Perception: What one person finds complex (a “signal”), another finds simple. For example, a neurophysiologist sees the brain as a complex web of fibers, while a butcher sees it as a simple “meat”[6][7]. The “signal” is thus dependent on the observer’s interests and purpose[8][9].

2. The ‘Net’ as an Emergent Artifact of Framing

The “net” refers to the specific choices an observer makes to structure a “messy” situation. The detection of a signal is often an emergent property of how this net is cast:

Boundary Judgments: Systems are not “given” by reality; they are constructs defined by boundary judgments[10][11]. These judgments determine what is in the “spotlight” (the system) and what remains in the “shadow” (the environment)[12][13]. A signal remains “weak” or “invisible” simply because it falls outside the chosen boundary of pertinent knowledge[10][14].

Station Points and Partiality: Flood emphasizes that all views are “temporary and partial”[15][16]. The observer’s “station point”—their unique location in time, space, and logic—acts as an organizing Gestalt that gives meaning to certain details while rendering a vast array of others “meaningless or invisible”[14][17].

Metaphors as Filters: Systemic metaphors (e.g., viewing an organization as a “machine,” “brain,” or “culture”) act as filters placed over the observer’s lens[18][19]. A “brain” filter might detect signals of “learning to learn,” while a “machine” filter would categorize those same signals as “noise” or “dysfunction”[20][21].

3. The Role of Task Constraints and Logic Levels

The “net” is also defined by the task and the level of logic used:

Filtering for Parsimony: In modeling, data must be filtered to achieve parsimony (simplicity for usefulness)[22]. This filtering—whether done via “expert consultation” (modeler as sole filter) or “process consultation” (including participants)—determines which outliers are retained for their “explanatory power” and which are discarded to improve “predictive ability”[23].

Logic Level Shifts: Signals that are unsolvable or “noise” at the object level (operational reality) can be re-perceived as meaningful metadecisions at a higher logic level (epistemology)[26][27]. Confusing these levels can lead to “disastrous” failures in detecting the true nature of a problem[26].

Algedonic Filters: In the Viable System Model (VSM), a specific “algedonic” filter is used to separate out “weak signals” of urgent development or “pain/pleasure” that require senior management’s immediate attention[28][29].

4. The Ethical and Power Dimension

Finally, the detection of signals is a political and ethical act:

The Ethics of Exclusion: Critical Systems Thinking (CST) argues that the choice of “the net” is ethical because it defines the beneficiaries of an inquiry[12][13]. “Weak signals” may be signals from marginalized stakeholders (the “affected but not involved”) that are intentionally or unintentionally filtered out by the “knowledge/power” held by those in control[30].

Sweeping-In: To detect important weak signals, CST recommends a process of “sweeping-in,” which involves intentionally expanding the boundary to include more conditions and contrasting mental models to become “as well informed as possible”[33][34].