Signifier¶
Core Idea¶
A signifier is the perceptible cue placed on or near an action possibility so that the possibility itself is recognized in time to be used. The structural distinction is sharp and load-bearing: an affordance is what the world makes possible, while a signifier is what makes the affordance perceivable. A door can be pushed — that is an affordance — while giving no evidence that it should be pushed rather than pulled; the flat plate, the horizontal bar at hand height, or the engraved word PUSH is the signifier that publishes the affordance to a user who has not been told. The two are paired but distinct contributions to action-readiness, and they can fail independently: an affordance the user cannot perceive does not function as an affordance for that user, and a cue that points at no real affordance misroutes effort toward an action that does not exist.
The pattern has three structural ingredients. There is a latent action possibility, the affordance. There is a perceptual cue indicating the possibility — a shape, marking, sound, texture, label, light, gesture, or convention — designed or selected to be encountered prior to the action. And there is a perceiving agent who reads the cue against learned, innate, or conventional decoding rules. The cue's value is precisely the delta in predictability of correct action between the world-with-cue and the world-without. The commitment that distinguishes a signifier from the broader pattern of signaling is that a signifier names an affordance: it tells a perceiver what they can or should do, here, now, rather than merely that some information exists. Its success metric is action-readiness — on encountering the cue, does the perceiver know which action is available, where it begins, and what its effect will be? Two failure modes follow directly from this metric: false signifiers, cues that suggest an affordance not present (the broken button, the flat plate on a pull door), and missing signifiers, real affordances with no perceivable cue (the hidden control, the unlabeled feature). The practical dictum that follows is to supply signifiers more than affordances, because the work of design is largely the work of publishing affordances to perceivers who have not been told they exist.
How would you explain it like I'm…
The Push-Here Hint
The Action Clue
Cue That Names an Action
Structural Signature¶
a latent action possibility (the affordance) — a perceptible cue placed on or near it — a perceiving agent who reads the cue against a decoding rule — encounter prior to the action, at the point where action commits — a gain in predictability of correct action as the success metric — two failure modes: false signifiers and missing signifiers
The pattern is present when each of the following holds:
- A latent affordance. Some action the world makes possible, which may exist whether or not any perceiver can detect it. The signifier is distinct from this — it makes the affordance perceivable, not possible.
- A perceptible cue. A shape, marking, sound, texture, label, light, gesture, or convention indicating the possibility, designed or selected to be encountered before the action.
- A perceiving agent with a decoding rule. A reader who interprets the cue against innate, learned, or conventional rules; without a shared decoding rule the cue carries no action-information.
- Locality of encounter. The cue is met at the spot where action commits — not earlier, where it is forgotten, nor later, after the action is past.
- An action-readiness metric. Success is measured by the delta in predictability of correct action between world-with-cue and world-without: does the perceiver know which action is available, where it begins, and what its effect will be?
- Two failure modes. A false signifier (a cue suggesting an affordance not present) and a missing signifier (a real affordance with no perceivable cue). A false signifier is worse than none, since it pre-allocates effort to a nonexistent action.
These compose into the affordance-publishing function: the cue advertises what can be done, here, now, by this perceiver — the narrowing of "signal" to "names an available action" that distinguishes a signifier from generic signaling.
What It Is Not¶
- Not an affordance. See
affordance. An affordance is what the world makes possible; a signifier is what makes the affordance perceivable. They are paired but distinct and can fail independently — a real affordance with no cue, or a confident cue pointing at no affordance. This is the prime's defining boundary. - Not the Saussurean signifier. See
signifier_signified_duality. There, "signifier" is the perceptible form of any sign coupled to a signified concept — a general semiotic relation. Here, a signifier specifically advertises an available action, narrowing the broad sign-form to the action-publishing function. The shared word names different objects. - Not signaling in general. See
signaling. Signaling advertises anything — quality, identity, status, intent. A signifier advertises specifically what can or should be done, here, now, by this perceiver. The narrowing to action-readiness is exactly what carves the prime out of the general semiotic field. - Not feedback. See
feedback. A signifier governs the moment before action — does the perceiver know what to do. Feedback governs the moment after — does the perceiver know it worked. They sit at opposite ends of the action and fail in different ways. - Not interpretation. See
interpretation. Interpretation is the perceiver's general act of assigning meaning. A signifier is the cue itself placed at the action point; it presupposes a decoding rule but is the published form, not the reading of it. - Common misclassification. Diagnosing every "hard to use" complaint as a signifier problem. The catch: locate the failure on the four-way map of affordance, signifier, action, and feedback. "I didn't know what to do" is a signifier problem; "I couldn't tell if it worked" is a feedback problem; "the system can't do that" is an affordance problem — and only one fix applies.
Broad Use¶
- Human-computer interaction and product design. Button styling, hover states, cursor changes, scrollbars, draggable handles, focus rings, label text — the signifier is the working construct of interaction design.
- Architecture and wayfinding. Signage, floor-pattern changes at decision points, lighting that draws toward an exit, paint stripes announcing a step, handrails appearing at the head of a stair.
- Safety and industrial design. Lockout-tagout colors, chevron hazard markings, reverse-beepers, and truncated-dome textures at platform edges — each cue advertising an affordance or its hazard complement at the spot where action commits.
- Animal communication. Aposematic coloring (the warning that advertises "I will harm you if pursued"), display behaviors, scent marks at territory boundaries, and courtship signals — the cue advertises an action-relevant state, giving genuine non-human carry.
- Traffic engineering. Lane markings, traffic signals, road-edge reflectors, and rumble strips — making affordances and hazards available to action-time perception.
- API design and software ergonomics. Method names, type signatures, auto-complete, tooltips, null-safety operators, and deprecation warnings — the signifier face of an API publishes which calls are available and how to compose them.
- Discourse markers. "However," "for example," "in conclusion" — verbal signifiers that publish the discourse move about to be made, helping the reader anticipate what follows.
Clarity¶
The signifier prime separates four things that intuition collapses into one: what the system can do (the affordance or capability), what the system tells you it can do (the signifier), what you do (the action), and what follows (the feedback). This four-way separation is diagnostically powerful, because most usability failures resolve cleanly onto exactly one of the four. An action-readiness failure — the user does not know what to do — traces to a missing or misleading signifier. An outcome failure — the user acted but cannot tell whether it worked — traces to missing feedback. And an action mistake — the user did the wrong thing — traces to a misread cue or a misclassified affordance. Without the separation, all of these present as the undifferentiated complaint "this is hard to use," and the fix is guessed at rather than diagnosed. The frame also separates the signifier from the broader notion of signaling. Signaling can advertise anything — quality, type, identity, status — whereas a signifier specifically advertises what can be done here, now, by you. This narrowing is the whole point: by carving the action-advertising function out of the general semiotic field, the prime makes it possible to reason about action-readiness as a distinct design target, with its own success metric and its own failure modes, rather than folding it into either the general theory of signs or the bare notion of capability.
Manages Complexity¶
A complex environment full of actions and hazards becomes walkable when its signifiers do their work. A subway map, a medication label, a power-tool warning sticker, a button label — each compresses what could have been a tutorial into a cue encountered at the action point, so that the perceiver need not carry a manual but can read the relevant action possibility off the world at the moment of acting. This is a genuine compression of the agent's cognitive burden: instead of learning and remembering the full set of available actions in advance, the agent reads each affordance as a cue when and where it becomes relevant. The signifier-affordance distinction also compresses the design problem. When an interface is hard to use, the diagnosis lands on which signifiers are absent, misleading, or out of place — and this diagnosis is independent of whether the underlying capability is present, which separates a question that is often easy to answer (is the cue there and correct?) from one that is often harder (is the capability there?). The designer can therefore work on the signifier layer without re-engineering the capability layer, and vice versa. This compression travels across substrates: the architect routing pedestrians through a museum and the API designer routing developers through a library face the same load-bearing problem — publish the available actions at the decision points — and differ only in medium, so a solution discipline developed in one transfers to the other.
Abstract Reasoning¶
The signifier prime supports a tight family of inferences. There is an action-readiness lower bound: the probability of correct action is bounded above by the probability that the relevant signifier is perceived in time, so supplying capability without a signifier is wasted capacity — the affordance simply does not function for a user who cannot find it. There is convention dependence: signifier reading depends on a decoding rule that may be innate, learned, or culturally conventional, so removing a convention — retiring skeuomorphic icons, say — without replacing it leaves orphan affordances that the capability still offers but no perceiver can locate. There is the cost of false signifiers: a misleading cue is worse than no cue, because it pre-allocates the perceiver's effort to the wrong action, so a door-shaped panel that is actually a wall, or a save icon that does not save, is a more expensive failure than a blank surface. There is the stacking inference: redundant cues — text plus icon plus color — raise action-readiness for users with different perception channels, which is the structural basis of the accessibility argument. And there is the locality principle: a signifier should be encountered at the spot where action commits, not earlier, where it is forgotten, nor later, after the action is past. Each of these is stated in terms of the cue, the perceiver, and the action-readiness gain, which is why they hold across substrates rather than only in software.
Knowledge Transfer¶
The signifier concept transfers across substrates because its structure — a cue that advertises an action possibility to a perceiver against a decoding rule — names no particular medium. It ports from HCI to architecture directly: the discoverability critique of a building (can a visitor find the exit, the stair, the entrance?) is the same critique as the discoverability critique of an interface, and the wayfinding designer and the interaction designer are solving one problem in two media. It ports from architecture to API design through the desire-path analogy: users will find their own route if the signifiers fail, which reappears in the idiomatic library-usage patterns and keyboard shortcuts that developers invent when the published interface does not advertise the action they need. It ports from ethology to safety design with the cost structure intact: the wasp's aposematic stripes are the biological case of the hazard-signifier pattern that engineering uses for high-voltage labels, and the cost of not signaling the affordance — being eaten, being electrocuted — is structurally identical. It ports from discourse markers to UI microcopy: the role of "however" in marking the next discourse move is the role of a prompt suggestion in marking the next user action. And it ports into novel substrates as a generator of design questions: how to signify model uncertainty so a user can act under it is the signifier problem in the AI-interface substrate, recognizable as such before any solution is in hand.
The role mappings that make these transfers reliable are direct. The latent affordance maps to the pushable door, the available API call, the safe-to-cross moment, the territory available to a rival, the upcoming discourse move. The perceptual cue maps to the flat plate, the method name, the green light, the aposematic stripe, the discourse marker. The perceiving agent maps to the hurried user, the developer reading an IDE, the driver, the predator, the reader. The action-readiness gain maps in every case to the increase in probability of correct action that the cue supplies. Because the structure and its two failure modes are shared, the same diagnostic transfers without substrate change: an API method named delete() that actually performs a soft-archive is a Norman door, a "Save" button that opens a "Save As" dialog is a Norman door, a museum corridor that ends in a janitor's closet is a Norman door, and a warning chevron on safe equipment is a Norman door in the other direction — a false signifier. What transfers is the discipline of distinguishing the capability from its advertisement, locating usability failures on the four-way map of affordance, signifier, action, and feedback, and supplying cues at the action point that publish, truly and locally, what can be done there.
Examples¶
Formal/abstract¶
The "Norman door" is the canonical worked instance, and it isolates the signifier from the affordance with diagnostic precision. Consider a glass door that opens by pushing. The latent affordance is the push: the world genuinely makes that action possible, whether or not anyone can tell. The perceptible cue should publish it — a flat horizontal plate at hand height affords nothing but pushing, so it is a true signifier that names the push action. The failure case swaps in a vertical pull-handle: the perceiving agent, reading the cue against the learned decoding rule "a graspable handle means pull," predicts the wrong action. This is a false signifier — a cue suggesting an affordance (pull) that is not present — and the frame's inference that a false signifier is worse than none is exactly borne out: a blank push-plate-less door would leave the user to try both, but the handle pre-allocates effort to pulling, producing the characteristic yank-then-stumble. The locality of encounter matters: the cue must be at the spot where action commits (on the door, at hand height), not on a sign ten feet back that is forgotten by the time the user arrives. The action-readiness metric is the delta in probability of correct action between handle-door and plate-door, and it is measurable. The frame's prescribed intervention follows directly and is independent of the capability: do not re-engineer the hinge (the affordance is fine); fix the signifier — replace the handle with a plate so the cue's decoding rule points at the action the world actually affords. The stacking inference adds the engraved word PUSH as a redundant cue for users whose handle-reading dominates their shape-reading.
Mapped back: The push is the latent affordance, the flat plate is the true signifier and the pull-handle the false one, the user's "handle-means-pull" rule is the decoding rule, and the yank-then-stumble is the cost of a false signifier — fixed at the signifier layer without touching the affordance.
Applied/industry¶
Aposematic coloration in animal communication is the signifier pattern operating in a non-human substrate, which is what gives the prime genuine cross-substrate reach beyond human design. Consider a wasp's bold black-and-yellow stripes. The latent affordance — strictly, the action-relevant state the cue advertises — is "pursuing me will harm you": the wasp possesses a sting, a real capability. The perceptible cue is the high-contrast warning pattern, selected by evolution to be conspicuous and encountered before the predator commits to an attack. The perceiving agent is the predator, reading the cue against an innate or learned decoding rule that associates the pattern with prior bad experience. The locality of encounter is the structural crux, exactly as with the Norman door: the cue must be perceived before the attack commits, because a warning read after the predator has already struck is worthless. The action-readiness metric is the gain in the predator's probability of correct action (avoid) that the stripes supply. The two failure modes both appear: a palatable mimic wearing the warning pattern without the sting is a false signifier (advertising an affordance it lacks), and a dangerous-but-drab insect is a missing signifier (a real hazard with no perceivable cue, hence eaten). The cost structure is identical to engineered hazard signifiers — the high-voltage chevron, the truncated-dome texture at a platform edge — where the cost of not signaling (electrocution, a fall) maps onto the biological cost (being eaten). A safety engineer placing a hazard label at the point of action and evolution placing warning coloration on a wasp are, structurally, solving the same problem: publish an action-relevant state to a perceiver, at the spot where action commits, against a shared decoding rule.
Mapped back: The sting is the latent capability, the warning stripes are the signifier, the predator's learned avoidance is the decoding rule, pre-attack perception is the locality principle, and the palatable mimic is the false signifier — the identical structure as a high-voltage chevron, in a substrate with no designer.
Structural Tensions¶
T1 — Signifier versus Affordance (the Defining Boundary). A signifier makes an action perceivable; an affordance makes it possible — and they fail independently. The prime's whole leverage comes from holding them apart, yet they collapse together under the loose word "usable." The failure mode is treating an action-readiness problem as a capability problem (re-engineering a hinge that works fine) or vice versa (adding a label to a button that does nothing). The diagnostic is to ask separately whether the capability exists and whether a correct cue advertises it: a real affordance with no perceivable cue is a missing-signifier problem fixed at the cue layer, while a confident cue pointing at no capability is a false-signifier problem fixed at the affordance layer — and conflating them aims the repair at the wrong layer.
T2 — False Signifier versus No Signifier (Sign of the Error). The two cue-failures are not symmetric: a false signifier is worse than a missing one, because it pre-allocates the perceiver's effort to an action that does not exist. Absence leaves the perceiver searching; misdirection commits them wrongly. The failure mode is treating any cue as better than none, adding signifiers for completeness and thereby planting confident pointers at nonexistent or wrong affordances — the door-handle on a push door, the delete() that soft-archives. The diagnostic is to ask, of each cue, whether it points at the action the world actually affords: an unsignified affordance costs a search, but a mis-signified one costs a committed wrong action plus the recovery, so a cue whose accuracy is uncertain is more dangerous than a blank surface.
T3 — Cue Present versus Decoding Rule Shared (Convention Dependence). A signifier carries action-information only against a decoding rule the perceiver holds; the cue is inert without the convention. This makes signifier-reading dependent on something outside the cue itself, and conventions vary across cultures, eras, and expertise. The failure mode is retiring or assuming a convention without checking it is shared — dropping skeuomorphic icons and orphaning the affordances they advertised, or designing cues legible only to the designer's own in-group. The diagnostic is to ask which decoding rule the cue presumes and whether the actual perceiver holds it: a cue perfectly clear under the author's convention signifies nothing to a perceiver who lacks it, so the convention, not just the cue, must be verified against the real audience.
T4 — Locality of Encounter versus Displacement (Temporal Placement). A signifier must be met at the spot where action commits — earlier it is forgotten, later the action is already past. Action-readiness is not just whether the cue exists but whether it arrives at the decision moment. The failure mode is a correct cue placed wrong in time or space: the warning on a sign ten feet back, the tooltip that appears after the click, the aposematic signal perceived only after the predator has struck. The diagnostic is to ask when and where the perceiver actually commits to the action, and whether the cue is present there: a signifier displaced from the commit point fails even when its content is perfectly correct, so placement at the decision locus is a separate requirement from accuracy.
T5 — Single Cue versus Stacked Cues (Channel Coverage). One signifier serves perceivers who read that channel; redundant cues — text plus icon plus color plus texture — raise action-readiness across perceivers with different channels, the structural basis of the accessibility argument. But stacking trades against clutter and can dilute the load-bearing cue. The failure mode runs both ways: a single cue that excludes anyone whose perception channel differs, or so many redundant cues that the environment becomes noise and the real signifier is lost. The diagnostic is to ask which perceiver channels must be covered and whether each added cue raises readiness for an uncovered channel or merely repeats a covered one: redundancy that reaches a new channel is accessibility, redundancy within one channel is clutter.
T6 — Action-Readiness versus Outcome-Feedback (Two Different Failures). The signifier governs the moment before action — does the perceiver know what to do — and is distinct from feedback, which governs the moment after — does the perceiver know it worked. Usability complaints collapse both into "hard to use," but they live at opposite ends of the action. The failure mode is misdiagnosing an outcome failure as an action-readiness failure, adding signifiers when the user already acted correctly but cannot tell whether the action succeeded. The diagnostic is to locate the complaint on the four-way map of affordance, signifier, action, and feedback: "I didn't know what to do" is a signifier problem, but "I did it and couldn't tell if it worked" is a feedback problem, and improving the cue does nothing for a failure that lives after the action commits.
Structural–Framed Character¶
Signifier sits just on the structural side of the middle of the structural–framed spectrum — a mixed-structural prime with a 0.4 aggregate. Its core is a bare relational object: a perceptible cue placed on or near a latent action possibility, read by a perceiving agent against decoding rules, whose value is the delta in predictability of correct action between the world-with-cue and the world-without. That three-part structure (latent affordance, perceptual cue, decoding perceiver) is structurally clean, but the prime carries a residue of its HCI home, which keeps it from reading as cleanly structural as percolation.
One diagnostic is fully clean: evaluative_weight is 0, since a signifier is neither good nor bad — false signifiers and missing signifiers are named as structural failure modes against the action-readiness metric, not as value verdicts. The other four sit at the partial 0.5 mark. Human_practice_bound is 0.5 because, while the canonical cases are human interaction design, the cue-that-advertises-an-action-possibility pattern reaches into genuinely non-human substrates — aposematic coloration in ethology, where a wasp's warning stripes signify "do not handle" to a predator with no designer and no human in the loop, is a real signifier — so the pattern is not bound to human practice. Vocab_travels is 0.5 because the Norman/HCI vocabulary (signifier, affordance, action-readiness) translates with mild friction into architecture, traffic engineering, API ergonomics, and discourse markers. Institutional_origin is 0.5 because the prime was sharpened within HCI and interaction design, a soft disciplinary origin. Import_vs_recognize is 0.5 because invoking the prime half-imports the affordance/signifier interpretive split and half-recognizes a cue-pointing-at-a-possibility structure already present. This profile — value-neutral with genuine non-human carry, but tinted by an HCI lexicon and origin and pairing structurally with affordance — is exactly what the mixed-structural label with its 0.4 aggregate records.
Substrate Independence¶
Signifier is a strongly substrate-independent prime — composite 4 / 5 on the substrate-independence scale. Its breadth is good: the cue-that-advertises-an-affordance pattern recurs across HCI and product design, architecture and wayfinding, safety and industrial design, animal communication, traffic engineering, API ergonomics, and discourse markers — and aposematic coloration in ethology gives it genuine non-human carry, a wasp's warning stripes signifying "do not handle" to a predator with no designer in the loop. The signature is highly relational — a latent affordance, a perceptible cue, a perceiving agent with a decoding rule, locality of encounter, an action-readiness metric, the false/missing failure modes — stated medium-neutrally, so the diagnostic (locate the failure on the four-way affordance/signifier/action/feedback map) and the interventions carry intact. Transfer is concrete and documented: the "Norman door" diagnostic ports identically to a mis-named API method, a museum corridor, and a hazard chevron, and the cost structure of a false signifier maps exactly from the palatable mimic to the high-voltage label. What holds it a notch below 5 is the thin residue the structural–framed analysis names: the Norman/HCI lexicon (signifier, affordance, action-readiness) translates with mild friction rather than dropping away, and the prime pairs structurally with affordance from a soft HCI origin. Recognized rather than translated across its range, with real non-human reach, it earns a composite 4.
- Composite substrate independence — 4 / 5
- Domain breadth — 4 / 5
- Structural abstraction — 4 / 5
- Transfer evidence — 4 / 5
Relationships to Other Primes¶
Parents (2) — more general patterns this builds on
-
Signifier is a kind of, typical Signaling
The file: signaling advertises ANYTHING (quality, identity, status); a signifier narrows this to advertising 'what can or should be done, here, now, by this perceiver' at the commit point. Genus-to-species (a different axis than the affordance partnership).
-
Signifier presupposes Affordance
The file's defining boundary: an affordance is what the world makes POSSIBLE; a signifier makes the affordance PERCEIVABLE. The signifier presupposes a latent affordance to advertise (they fail independently). Presupposes-parent — the 0.89 nearest is the structural PARTNER, not a duplicate.
Path to root: Signifier → Affordance
Neighborhood in Abstraction Space¶
Signifier sits in a sparse region of abstraction space (89th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely rather than landing on a neighbor.
Family — Emergence & The Irreducible Whole (6 primes)
Nearest neighbors
- Affordance — 0.78
- Absorbing State Under Restricted Modality — 0.72
- Invisible Affordance — 0.70
- Action Bias — 0.66
- Evolutionary Trap — 0.66
Computed from structural-signature embeddings · 2026-06-14
Not to Be Confused With¶
The defining confusion — and the embedding-nearest neighbor at similarity 0.89 — is with affordance, the prime's structural partner. They are so tightly paired that everyday usage fuses them into "usable," but the whole leverage of the signifier concept comes from holding them apart. An affordance is a relation between the world and an agent's capabilities — what the environment makes possible (a door that can be pushed, an API call that exists, a ledge a bird can perch on). A signifier is the perceptible cue that makes that possibility discoverable in time to act on it (the flat push-plate, the method name, the visual ledge-marking). The decisive fact is that the two fail independently: an affordance with no perceivable signifier does not function for a user who cannot find it (a hidden control, an undocumented feature), and a signifier pointing at no affordance misroutes effort (the pull-handle on a push door, the delete() that soft-archives). This independence is what makes the distinction a diagnostic: when something is "hard to use," the repair lands at different layers depending on which half failed — fix the cue (signifier) without re-engineering the capability (affordance), or vice versa. Collapsing the two leads to the classic mis-repair of redesigning a hinge that works perfectly because the real fault was the handle, or adding a label to a button that does nothing because the real fault was the missing capability. The prime exists precisely to keep "can it be done?" separate from "can the user tell it can be done?"
A second confusion, sharpened by the literal shared name, is with the Saussurean signifier_signified_duality. In semiotics, "signifier" denotes the perceptible form of any sign (the sound-image "tree") coupled to its signified concept, and the duality is the general claim that every sign pairs a form with a meaning. The HCI signifier is a specialization with a narrowed referent: its "signified," loosely, is always an available action — what can or should be done, here, now, by this perceiver — not an arbitrary concept, identity, or category. So the semiotic signifier is the genus (any form-bearing sign) and the action-signifier is a species defined by what it signifies (an affordance) and its success metric (action-readiness at the commit point). The shared word invites the error of importing the full semiotic apparatus — arbitrariness of the sign, paradigmatic and syntagmatic relations — into interaction design where only the action-publishing slice applies, or conversely of treating every UI cue as a deep linguistic sign when its job is the narrow one of telling a hurried user which action is available. Keeping the two apart prevents both over-theorizing the button and under-theorizing the word.
A third confusion is with signaling in the broad sense (and its biological cost-theory cousin). Signaling covers the advertisement of any action-relevant or state-relevant information — quality, type, status, intent, threat — to a receiver. The signifier narrows this to the advertisement of an available action to a perceiver at the point of action. The relationship is genus-to-species again, but along a different axis than the Saussurean case: signaling is broad in what is advertised (any state), while the signifier is specific in what is advertised (an affordance) and where (at the commit point). The wasp's aposematic stripes sit in the overlap — they are a signal (advertising a state, "I am dangerous") that functions as a signifier (publishing the action-relevant possibility "pursuing me will harm you," read before the predator commits). But a peacock's tail advertising mate quality is signaling that is not an action-signifier in the prime's sense — it publishes a state for assessment, not an available action at a commit point. Conflating them imports signaling's honesty/cost machinery into cases where the question is simply "is the cue at the right place and does it name the right action?", and misses the signifier's distinctive locality and action-readiness metrics.
For a practitioner, these distinctions route both the theory and the fix. First separate capability (affordance) from advertisement (signifier) — they break independently and demand layer-specific repairs. Reserve the semiotic signifier apparatus for genuine sign-systems and the signaling cost apparatus for state-advertisement among agents; for the everyday work of making actions discoverable at the moment of acting, the action-signifier — with its locality principle, its action-readiness metric, and its false/missing failure modes — is the precise tool, and borrowing the broader frameworks only blurs it.
Solution Archetypes¶
No catalogued solution archetypes reference this prime yet.