Reality is richer than any representation
Every enterprise acts through representations of reality. Some are explicit: financial models, architectures, forecasts, roadmaps, business cases, org charts, scenarios, dashboards, requirements, contracts, and portfolio views. Others arise almost immediately in perception.
Imagine a rock beside a tent stake. The useful relationship is obvious enough that no formal model is required. The rock is hard and heavy enough to drive the stake. That representation is sufficient for the task, even though it contains almost none of the total reality of the rock, the soil, the body, or the surrounding environment.
That incompleteness is not a defect. The representation does not need to contain the rock's chemical composition or the soil's microbial life because those facts do not materially bear on the immediate purpose.
The practical question is never whether a representation contains all of reality. It cannot. The question is whether it represents the right slice, with enough fidelity and enough richness, and whether other materially implicated slices also need to enter the picture.
Four questions govern a representation
What is it for?
A representation exists for a purpose. The purpose determines which slice of reality must become legible, for whom, at what scale and time horizon, and toward what judgment or action.
A map can be highly accurate and still be useless if it faithfully represents the wrong slice for the question being asked.
How faithful is it to that slice?
A useful representation preserves the distinctions, relationships, constraints, interactions, states, and dynamics that materially determine the answer for its intended purpose.
Factual correctness alone is not enough. A model can contain individually correct facts while drawing the wrong boundary around the economic object or omitting a relationship that changes the answer.
How rich does it need to be, and how rich is it actually?
The purpose establishes the required richness. Some judgments need only a few variables. Others require much more resolution, causal depth, temporal depth, uncertainty treatment, or interaction detail.
Too little richness creates erasure. The representation can be accurate as far as it goes while still omitting structure required for the judgment.
Too much richness can also make the representation worse. Excess detail can bury signal, increase cognitive and coordination burden, slow judgment, create false precision, and make the map harder to keep current.
The objective is not maximal detail. It is sufficient richness for the purpose.
Which other representations materially bear on this one?
A focal representation can be purpose-correct, faithful, and sufficiently rich on its own terms and still support a bad judgment because another materially interacting slice of reality has been left outside the picture.
The answer is not one master representation. In consequential enterprise movement, finance, architecture, operations, market strategy, organizational design, risk, regulation, procurement, portfolio management, and implementation may each require distinct maps. The important question is which relationships among those represented slices materially change the enterprise answer.
Several maps may therefore need to inform one another while remaining specialized.
Why Systemicity raises the representational burden
At very low Systemicity, the purpose-relevant relationship can be small, stable, and directly apprehensible. The rock and tent stake are the useful extreme.
As Systemicity rises, more economically material structure can lie outside any one representation. Causes and effects can sit at different scales. Value can depend on configuration with other elements. Sequence can change feasibility. Present choices can create or foreclose later possibilities. Knowledge can be distributed across functions. The territory itself can move while the enterprise is still deciding and acting.
Representation therefore becomes more necessary at the same time that any one representation becomes less capable of carrying the whole relevant answer.
The practical response is not complete representation. It is the right partial representations, at the right levels of richness, connected in the right ways for the purpose at hand.
Representation is one surface of Coherence
Representation is one way an enterprise can become more or less coherent with the reality it is navigating. The wrong map, an inaccurate map, an impoverished map, an overbuilt map, an isolated map, or a stale map can all degrade enterprise judgment.
But representation is not the whole of Coherence. An enterprise must still discover and construct possibilities, judge among them, commit, sequence, execute, learn, adapt, and revise.
Representations have useful lives
Feedback can change what must be represented, the boundary of the map, how much richness is required, which other maps need to be considered, or even the question the representation exists to answer.
The objective is not fidelity to the original map. It is continued adequacy to reality as the enterprise moves.
Bibliography
Adjacent intellectual territory includes map-territory distinctions, bounded rationality, modeling and abstraction, distributed knowledge, sensemaking, systems thinking, scientific models, information compression, decision theory, cognitive representation, and organizational information processing. Talbot West uses representation for the recurring enterprise operation of making selected slices of reality legible enough to reason and act.