Business Analysis Notes

Systems thinking

Twelve Properties of Systems

4 min read · Updated

In short Instead of a one-line definition, Tarasenko describes a system through 12 properties that every system has. Used as a checklist, they help you build three basic models (black box, composition, structure) and remind you that the results you care about usually come from how parts interact.

Based on: F. P. Tarasenko, Applied Systems Analysis (in Russian), ch. 2, in the systems tradition of R. L. Ackoff.

On this page

Key ideas#

The 12 properties fall into three groups of four.

Static properties: a snapshot#

  1. Wholeness. A system can be told apart from its environment; drawing that line splits the world into “system” and “everything else”.
  2. Openness. The system has inputs and outputs. Listing them gives a black-box model. Tarasenko names four ways to get it wrong: including an irrelevant link (wasted effort), missing an essential one (the goal becomes unreachable), not knowing a link exists, and mistaking an input for an output or the reverse.
  3. Internal heterogeneity. The system has distinguishable parts; listing them gives a composition model. There is no single “true” split: how you divide the whole, where you stop dividing, and even the outer boundary all depend on your purpose, because every system is part of a larger one.
  4. Structure. The essential links between parts form the structure model. A system’s behavior depends on how parts are connected, not only on what they are; an organization’s formal and informal structures behave differently.

Dynamic properties: a film#

  1. Functionality. Outputs are the system’s functions. There are many, and whether one is main, secondary, unwanted or useless depends on who judges.
  2. Stimulability. Inputs act on the system and change it.
  3. Variability. Tarasenko separates functioning (parameters change, structure does not), growth (more of the same) and development (better qualities). Growth consumes resources and always hits limits; development is learning. He quotes Ackoff’s view of development as a rise in the desire and ability to meet one’s own and others’ needs, and stresses that it can only be self-development: you can help an organization learn, not develop it from outside.
  4. Existence in a changing environment. In a turbulent environment a system survives only by matching outside change with inside change.

Synthetic properties: the whole#

  1. Emergence. A system has qualities that none of its parts has, produced by the links between them. The same parts wired differently behave differently. The lesson for managers: improve the interaction between units rather than each unit on its own, the way a conductor works on how musicians play together.
  2. Indivisibility. Remove a part and both change: the whole loses some property, and the part loses abilities it had only in context.
  3. Inherence. How well the system fits its environment for a given function. Fit can be added through an intermediary such as a translator or an adapter.
  4. Purposefulness. In a designed system, parts and structure serve a goal. Tarasenko separates the subjective goal (the desired future and the path to it) from the objective one (the future that actually arrives); a subjective goal is reachable only if it does not contradict natural laws. Ideals can be approached without limit but never reached.

How to apply it#

  1. Write down the purpose of your model first; the right boundary and split depend on it.
  2. Draw the black box. Check every input and output against the four error types.
  3. Build the composition model only as deep as your purpose needs.
  4. Draw the structure model: who hands what to whom. Include informal links.
  5. List the functions per stakeholder, including unwanted ones.
  6. Ask which results are emergent, produced by hand-offs rather than by any one unit. Improvement usually pays off there.
  7. Check fit with the environment and where an intermediary would help.

Example#

Model one shop of The Workshop for the online-booking project.

  • Black box. Inputs: bikes, bookings, parts deliveries, customer questions. Outputs: repaired bikes, invoices, status updates, worn parts for disposal. A first draft missed the supplier’s delivery schedule, an error of the second kind: turnaround cannot be planned without it.
  • Composition. Front desk, mechanics, parts store, workshop manager.
  • Structure. Front desk to mechanic (job card), mechanic to parts store (request), parts store to supplier (order), front desk to customer (status). Informally, mechanics phone customers about extra work.
  • Emergence. Turnaround time belongs to none of these parts. It comes from the hand-offs, so the project targets the job-card and parts-request links, not mechanic speed.

Pitfalls#

  • Searching for the “real” decomposition → choose one that serves your purpose and say so.
  • Decomposing and losing the links → pair every composition model with a structure model.
  • Optimizing each unit separately → find the interaction that produces the result.
  • Treating development as something you deliver to a client → create conditions for them to learn.

Sources and further reading#

  • F. P. Tarasenko. Applied Systems Analysis (in Russian). KnoRus, 2010, ch. 2.
  • R. L. Ackoff — the definition of development quoted in ch. 2; Tarasenko’s framework builds on Ackoff’s systems thinking.