Where it sits
Most real cases sit here — explicable but not foreseeable. Strong emergence sits further right and is disputed.
Tell it apart
| Aspect | Weak emergence | Strong emergence |
|---|---|---|
| Derivable in principle | Yes | No |
| Needs simulation | Yes | — |
| Breaks reductionism | No | Would |
| Evidence | Abundant | Contested |
Not opposites, despite the framing. Weak emergence is fully compatible with reductionism — the lower-level physics is sufficient, just computationally intractable. Only strong emergence would genuinely break reduction, and that case is unproven.
Opposite of Emergence
- ReductionismThe opposing view: a whole can be fully explained by its parts and their interactions, with nothing genuinely new appearing at the higher level.
What it doesn't mean
Most emergence is weak emergence, which is fully explicable from component interactions; it's simply not PREDICTABLE without running the simulation. Only strong emergence claims genuine irreducibility, and it's philosophically contested rather than an established finding.
It's at least two, and conflating them is the standard error. Weak emergence is derivable in principle from lower-level laws but only via computation. Strong emergence is not deducible even in principle. Popular writing routinely uses evidence for the first to argue for the second.
What it really means
Emergence is usually introduced with a phrase that does more harm than good: the whole is more than the sum of its parts. It sounds like a claim that science hits a wall somewhere above the atoms.
Mostly it isn't. The distinction that matters is between weak and strong emergence, and almost everything anyone points to is the weak kind.
Weak emergence means a macro-level property can be derived from the micro-level rules — but only by running the simulation. There's no shortcut, no closed-form equation that tells you the answer without stepping through every interaction. Conway's Game of Life is the cleanest example: the rules are four lines long, fully deterministic, and there is no way to know what a given pattern does except by watching it do it. Nothing is hidden. It's just not foreseeable.
Strong emergence is the much larger claim — properties not deducible even in principle from the lower level, potentially involving downward causation. That's philosophically contested and not an established scientific finding.
The standard error, and it's everywhere, is using evidence for the first to argue for the second. Ant colonies and murmurations are spectacular weak emergence, and they're fully compatible with reductionism — the physics underneath is sufficient, it's simply intractable to compute.
The Latin is nicely exact. Ē- + mergere: to rise out of, as from beneath water. Something surfacing from a level below — visible, but arriving from somewhere you weren't looking.
Where it comes from
Latin — out of
Latin — to dip, plunge
The literal image is surfacing from beneath water — something coming into view from a level below. An unusually apt etymology for a technical term.
- 1972
Philip W. Anderson publishes 'More is Different', arguing not against reduction but against the assumption that it implies construction — new laws and concepts appear at each level of complexity.
- 1997
Mark Bedau's criterion: a macrostate is weakly emergent if and only if it can be derived from the microdynamic and external conditions, but only by simulation.
In real life
- No starling directs a murmuration: each bird tracks a few neighbours, and the swirling shape of the whole flock emerges from that.
- In Conway's Game of Life, a grid with a handful of simple rules produces 'gliders' that crawl across the screen — something none of the rules mentions.
- A single molecule has no temperature; temperature appears only in the collective motion of very many of them.
Questions
Q01What is emergence?
Phenomena arising from interactions among a system's components that can't be readily inferred by examining those components in isolation. Ant colonies building bridges, starling murmurations and temperature as a collective property are standard examples.
Q02What's the difference between weak and strong emergence?
Weak emergence can be derived from the lower-level rules but only by simulation — you have to run it to find out. Strong emergence cannot be derived even in principle, and would involve genuine downward causation. The first is well evidenced; the second is philosophically contested.
Q03Does emergence disprove reductionism?
Weak emergence doesn't — it's fully compatible with the lower-level physics being sufficient, just computationally intractable. Only strong emergence would break reduction, and that case isn't established. Popular writing tends to blur the two.
Sources
- Conflicting Emergences: weak vs. strong emergence for the modelling of brain function — PMC.
- What is emergence, after all? — arXiv.
- emergence (etymology) — Etymonline.