6 Comments
User's avatar
Ducky McDuckface's avatar

Attractors? The system might have several states that are good enough, and it might bounce between them?

Or the cat decides to shred the elastic...

Expand full comment
Maxim Raginsky's avatar

That distinction is correct: homeostasis requires maintenance, and can typically be modeled as a result of some control process which optimizes (or at least satisfices) some cost over time. It can be often thought of as equilibrium, not on the system state space but on the space of system trajectories in time.

Expand full comment
Tim Wilkinson's avatar

Isn"t equilibrium a subtype of homoeostasis?

Expand full comment
Dan Davies's avatar

probably yes in a mathematical sense, but in my opinion it would be a bit weird to describe a bowl of water at room temperature as a homeostatic system.

Expand full comment
Tim Wilkinson's avatar

True, maybe because there's nothing very salient about 'being at room temp' (especially given that room temp is not fixed nor even independent). A bit like 'being a Pareto-optimal distribution'...

Expand full comment