[A]ll observation must be for or against some view if it is to be of any service!
—Charles Darwin76
What does it mean to say that we understand biological design? We must understand how particular causal forces have shaped the designs that we see in nature.
How do we match a causal force to the shaping of design? We must build on the idea that a change in force causes a change in state.
Why a change in force? Because many forces may initially be in balance. We identify a force and its consequences when the force changes.
How do we infer forces when studying biological design? Usually, we do not observe forces directly. Instead, we must infer what we cannot see. We suppose that an unseen force mediates between an observable partial cause and an observable effect, P → F → E.
What is an observable partial cause? A change in something measurable that we think may have an effect. Greater resource flow. Increased lifespan of resource patches. More competition between genotypes.
Why the partial qualifier for a cause? Because every effect has many causes. Each cause partially determines outcome. Many things alter growth rate. Increased resource competition may be one partial cause.
What is an observable effect? A change in any measurable attribute that we consider related to design. A trait, a tradeoff, the amount of variability in something, the dynamic tendency to fluctuate, the way an organism adjusts to its environment.
What is a mediating force? A process that mediates between cause and effect, defined inductively by the accumulation of reasonable interpretations. Kin selection mediates between a causal change in the genetic similarity of competitors and the effect on growth rate.
Why is cause used for both the observable change in the condition, P, and the mediating force, F? The observable change that drives the process is what we see as a cause. The unobservable force acts as the mediating causal process that shapes design.
What is a comparative prediction? A change in some condition, acting as a partial cause, that leads to a change in some effect, mediated by a force, P → F → E. Comparing different conditions predicts the direction of change in the designed effect.
Why must we use comparative predictions to study design? In practice, inferring cause requires associating the change in one observable with the change in another observable. Inferring cause depends on change, and change arises from comparison.
What is the relation between observable change and inferred cause? Observable change is the story. Inferred cause is a compelling plot to explain the story.
How can comparative predictions mislead? We can observe predicted associations, but our explanations may give the wrong reasons for the observed associations. The best we can do is to test for potential confounding factors and to test over many different conditions.
Are there alternative ways to study the causes of design? Not really. Consistency between an observed story and an explanatory plot is good but leaves open too many alternative consistent plots. Comparison restricts alternatives more strongly than does consistency.350
My emphasis on comparison is not new. Darwin often made comparative predictions about how a novel environmental challenge would likely cause an altered design for a matching trait.
Darwin also understood the importance of using phylogenetic history to develop meaningful comparisons.75 A changed environment might cause all species of a genus to share a novel design. All of the altered species together count as a single change because they share by common descent the same single change in design.
The number of separate comparative observations of change depends on the number of independent events when mapped onto the phylogenetic history.173
Darwin’s comparative method applies to historical, uncontrolled comparisons. By contrast, the experimental method provides well-controlled comparisons. With proper randomization, we increase the isolation of partial causes. Greater isolation improves the chance that an observed association between a putative cause and a consequent effect is mediated by the hypothesized force rather than by other correlated causes.
Given the highly developed understanding of comparison, why have I emphasized that aspect so strongly? Because most theory about biological design and most studies that discuss the causes of biological design do not focus sufficiently on comparison.
Three reasons may explain the lack of emphasis on comparison. First, biological data are often collected for reasons other than inferring the causes of design. Data often present a pattern rather than how pattern changes with some hypothesized cause. An observed pattern invites a consistent explanation. Consistency greatly dominates over comparison in the literature.
Second, evolutionary theory, which should drive the study of design, often fails to highlight comparative predictions. If a theory does not explicitly conclude and strongly emphasize that, as a condition A changes, the theory predicts that the design feature B changes in a specific way, then that analysis has failed to provide the proper impetus for study.
Third, there is often a mismatch in scale. Forces of design may change relatively quickly. Observed changes in traits may be measured over longer timescales. Microbes provide opportunities to match the scales of force and change. That potential match of scales in microbial studies motivated this book.
There are, of course, many examples of comparative predictions and empirical tests. But progress is slowed by the more numerous studies that emphasize consistency rather than comparison.
Part 1 summarized the forces that, in theory, shape biological design. Those forces lead to comparative hypotheses across many aspects of design and across all forms of life.
In Part 2, I apply the comparative analysis of design to microbial metabolism. To develop that topic, I synthesize the current understanding of microbial metabolism, critique the recent literature’s approach to the study of design, and present an improved way to study design by emphasizing comparative predictions and empirical tests.