Intro  Learning  Patterns  Real Life  Genetics  Classifiers  Biology  Neural Nets  Connectionism  Life  AI

Essays about Complex Systems, Adaptation, and Evolution, inspired by the book Complexity by M. Mitchell Waldrop

This sub-series of essays is a response to Complexity Chapter 8 Waiting For Carnot > The Growth of Complexity, starting on page 294.

1. Why does evolution seem to trend in the direction of greater complexity?
2. Does evolution have any direction? Where is it going? What will it lead to?
3. Complicated systems are not inherently better than simple ones.
4. What does "survival of the fittest" mean? What is evolutionary fitness?

What does "survival of the fittest" mean? What is evolutionary fitness?

Complexity, page 296, provides the following quote:

"With no objective definition of fitness, survival of the fittest becomes a tautology: survival of the survivors. ..."

The speaker seems to be complaining that the tautology has an unsatisfying feel to it, but it is nonetheless correct. The correct interpretation of "survival of the fittest" is "survival of the survivors". The reason lies in the correct definition of fittest, which is "those who survived".

Fitness and capabilities are two different things.

Capabilities of species are absolutely measurable

Capabilities are measurable using absolute scales. A cheetah has a physique that allows it potentially to run 70 miles an hour. A polar bear has a keen sense of smell that potentially allows it to detect a seal 1 mile away.

Capability measurements of two agents can be compared and ranked meaningfully somewhat independent of context or environment.

But specific capabilities in isolation are not the whole story and are not the same as fitness.

Evolutionary fitness is only a valid concept within an environmental context

Fitness is a balancing act, an optimization problem. The fitness of an agent or population is determined by how efficiently it utilizes resources to produce the right balance of all the capabilities it needs in the context of all the demands of the environment in which it exists.

Fitness describes the relationship between an agent and its environment, the degree of match. Without an environment, capabilities are just capabilities. Within an environment they acquire practical significance as the determinants of survival, the measure of fitness. It is the environment that determines which capabilities are important, or not. Without an environmental context, there is no such thing as fitness. 

If the environment changes, the equation changes, the whole balance changes. The fact that a cheetah can run 70 miles an hour is no use to one that is dropped into the open ocean. The fact that a polar bear can smell a seal 1 mile away is useless to one released into the Arizona desert. In those contexts, the capabilities are the same, but the fitness equation is much different. Capabilities do not determine survival. Fitness does.

For that reason, although humans are more intelligent than bacteria, it doesn't follow that we are more fit. What is important is: how well do the abilities of each match what they need to succeed in their environments? That's what fitness is.

Survival of the fittest does not mean survival of those who can run fastest, nor those who are smartest, nor those who are the most complicated, nor those who according to our calculations ought to survive.

It means "survival of those who do survive", those who have whatever characteristics are most needed when the survival challenge comes along that kills off all the others. That challenge could consist of almost anything, and the capabilities that might be relevant to surviving it are almost infinite in complexity and are completely dependent on what the challenge does consist of. The only measure of fitness that takes into account all possible factors is the actual outcome: who actually does survive. 

The reason "survival of the fittest" often leads to misunderstanding is that the word "fittest", as it is generally used, has connotations implying that something is better, in an absolute sense, than something else. In evolution, there is no such absolute context within which to judge something.

A better definition of "fittest" in the evolutionary sense is "those who fit best". It doesn't imply a value judgment. It just means those invididuals who fit best into one particular environment. 

The next section quotes a seemingly innocent statement from someone who misinterprets "fittest" in a slightly different way, but it still leads to trouble. 

Evolutionary fitness is as arbitrary and fickle as the environment

If the environment changes suddenly, those who were fittest (those who fit best) can suddenly be rendered totally unfit and go extinct. The fact that something is excellently suited to the current environment doesn't imply anything about its robustness, its flexibility, or its fitness relative to the next environment to come along. A change in the environment might destroy it.

Complexity, page 296 asks:

"Were our mammalian ancestors of 65 million years ago more advanced than Tyrannosaurus rex, or just luckier in surviving the impact of a marauding comet?"

Luckier would be the better choice, but neither characterization is satisfactory.

If you argue that dinosaurs were unlucky, it implies unfairness, as though what would otherwise have been some sort of "standard" evolutionary path was interrupted by an aberration. It tends to be dismissive of environmental change, a key evolutionary driving force and thus important and not aberrational.

The comet's impact and the post-comet world were a new environment for which "our mammalian ancestors" AND bacteria AND fish AND fungi AND myriad other species were better suited than dinosaurs.

If you argue that our mammalian ancestors survived the comet because they were more advanced than T. rex, you are forced to concede that all those other species were, too. 

A better route is to forget about the concept of "more advanced" with regard to evolution.

"Fittest" (in its common usage) and "more advanced" are concepts that depend on human definitions and prejudices.

In evolution, survival of the fittest is survival of those who "fit best" in a given environment at a given time. No value judgment is implied.  

To end with a sports analogy, you can analyze the capabilities of two teams by comparing statistics about the teams and their individual players, and you can decide based on their capabilities which team you think is most capable, or "best", by all the measures you think should be relevant. But the only way to determine the real winner, the survivor, on a particular day is to actually play the game.