Think twice / 03
The Watchmakers
The telephone will ring. How much of your work will still be there when you return?
How to play
You organise a watchmaking bench beside Tempus's. Both benches assemble identical watches and hear the same telephone calls, but your design can protect partly completed work. Your challenge is to finish more useful work without spending the afternoon protecting it. Find a design you can justify under both busy and quiet conditions. Tempus keeps all 24 parts loose until a watch is finished. At your bench, Hora can secure smaller groups called modules. A phone call scatters loose parts; secured parts survive. Securing takes time that could otherwise be spent assembling.
- Press Run afternoon with the starting settings. Each watch needs 24 parts. Watch both benches through the same 120-minute afternoon, or use See results now.
- Pause or step one minute to inspect what survives a call. Green markers are protected; amber markers are loose. Try Secure my loose parts now and notice that securing also takes time.
- Change Parts per module and compare completed afternoons. Then try A quiet shop. Use the results table to identify what your design gained in finished watches, what it lost to interruptions, and what it spent on securing.
What to notice
Tempus and Hora both build watches and both keep getting phone calls. Tempus works on one fragile watch at a time. When he puts it down to answer, his unfinished work falls apart. Hora builds small, stable modules first. A call costs him only the loose pieces he is holding. That difference, not greater speed or skill, is the heart of Herbert Simon's watchmaker story.
Run both benches through the same afternoon. Watch what is finished, what falls apart and how often Hora stops to secure a module. Then change the module size and run them again. Smaller modules protect progress sooner, but securing each one takes time. Larger modules save that overhead while leaving more work exposed to the next call.
Simon used the parable in 1962 to ask how complex systems can be built from stable intermediate parts. A saved draft, a tested software component or a recoverable checkpoint gives work somewhere solid to resume. Our watches use fewer parts and add an explicit securing cost, so they are an illustration rather than Simon's numerical example. How much progress is worth protecting before you continue?
Read more: Herbert Simon, The Architecture of Complexity.
An afternoon at the workbench
Minute 0 / 120Securing needs uninterrupted time. A call before it finishes still destroys loose parts.
Choose how to assemble, then run both benches. About 30 seconds.
What happened when the phone rang?
What survived?
| Afternoon / design | Tempus watches | Your watches | Parts lost: T / you | Your securing time |
|---|
Rules & model
Making and protecting a watch
Each watch needs 24 parts. Adding a part takes one minute. Tempus builds in one piece. Your bench automatically secures a module when it reaches your chosen size. Green parts are secured; amber parts are loose.
Securing a module uses the chosen number of working minutes. It protects the parts only when that process is complete. You may start securing a smaller batch early, but it takes the same time. A finished 24-part watch is complete immediately, without an extra final securing cost.
When the phone rings
A call takes one minute from both benches. No assembly or securing happens in that minute. All loose parts are lost, and unfinished securing is cancelled; already-secured modules and completed watches survive.
The optional three-minute warning gives you information, not free protection. If the phone rings before securing finishes, those parts are still loose. The afternoon ends after 120 minutes, even if a watch is unfinished.
Comparing fairly
Both benches face the same calls. Changing module size restarts that same afternoon. New afternoon changes the call schedule for both. Quiet, busy and clustered schedules are invented; the time used for securing is our addition to the original watchmaker idea.
The comparison shows a trade-off under these rules. It does not establish a best module size for every craft, project or organisation.
Where this began
Herbert Simon’s 1962 essay The Architecture of Complexity tells a parable about watchmakers Hora and Tempus. Stable subassemblies help Hora preserve progress through interruptions. Our smaller watches, securing cost, call schedules, manual intervention and finite shift are additions for this experiment.