Another example is the "camera eye".  It is an extraordinary organ.  How does it work?  Well, break it down.  There is a retina that responds to visible light by sending neural impulses to the brain.  There is a lens that focuses light on the retina.  There is an iris that regulates the amount of light entering the eye.  There is a cornea that bends the light initially, and there are two gelatinous "humors" that give the eye shape.  WOW!  It is a complex system, but by breaking it down into its component subsystems and learning what they do, we describe--in part--how an eye works.
 

        b. THE COMPARATIVE METHOD:

            When any complex system is considered in isolation, the observer is impressed with its complexity, integration of function, and internal causalities, and the very complexity of it seems to make figuring out it's origin nearly impossible. When we ask "why an eye?" or even the more mundane question of "How an eye?", there seems to be no place to start.

            Here's where the comparative methods finally comes into play, particularly within biology and the context provided by evolution. If we hypothesize that eyes evolved from simpler structures, maybe we can COMPARE the camera eye to the visual structures used by simpler, more primitive groups. We know they function, so by definition, these primitive systems are internally plausible and evidentiarily true - they exist, they aren't hypothetical constructs. So,let's compare the'eyes' in molluscs.... from limpets to snails to nautiloids to squid.

             Hmmm.... retina first, lens second, humours third, cornea and muscles last. Functionally efficient at each step, satisfying the limitations of a functional non-random sequential process. Good answer to an initally apparently intractible, unanswerable question using reductionism and the comparative method in an evolutionary context. We'll see this argument play out again tomorrow.

The comparative method is a very powerful tool in biology, because there is a REASON that organisms might be similar in structure or function. That reason is common ancestry - all species share a common ancestor with one another. The more recent that ancestor is, the more genes the species share and the more similar we might expect them to be. Even organisms that are very different - like flies and humans - can trace their separate lineages back to a common ancestor - probably a very primitive bilaterally symmetrical animal of the pre-cambrian, 600 million years ago. Although flies and humans are very different, they share the same basic stages of animal development from a zygote through to the establishment of a bilateral embryo. And indeed, they share the genes that co-ordinate this early development, the establishment of a head and tail end, and cues to specify how intervening body regions will develop. This common ancestry, and this similarity in structure and function, is the basis for our ability to use other species as models for humans - we learn about human genetics by studying flies, and we learn about human disease by studying mice.

        c. EXPERIMENTATION: 
                    Finally, the most direct way to tease apart causality from a nearly infinite set of coincident events is experimentation. For instance, if you want to know how eyes develop, well, there are an infinite set of events occuring whe an eye is developing, including genetic events and environmental events.

                    1) The first element is REPLICATED OBSERVATION - you have to observe something alot of times to get a feel for which events happen concurrently and might be putative causal agents. For instance, you might be watching the development of a fruit fly and you might notice that the eye begins development on a rainy day. Well, did rain cause eye development? NO WAY TO TELL. But, if you observe eye development in 100 flies over a period of time, you will probably notice that eye development occurs on rainy AND sunny days, so neither rain nor sun correlates with eye development and thus are probably not causal agents. Through careful observation and some knowledge of the system, a subset of factors possibly responsible can be determined. So, observation does not invovle looking at ONE thing - it involves looking at MANY THINGS and observing patterns of correlation among these things. Where to from here?

                    2) The second step is to construct a HYPOTHESIS; a statement of causal relationship.  What is actually causing the phenomenon that you observed?  For a hypothesis to be scientific, it must be falsifiable with evidence from the physical world.  You see, science is NOT the process of dreaming up ideas and then only seeking data that confirms this idea.  The fundamental process of science is testing falsifiable hypotheses.  What does this mean?  Well, a falsifiable hypothesis is one that could be proven false—it is a statement for which you can envision contradictory evidence.  For example, the statement that “humans evolved from other primates” is a falsifable statement.  We can envision collecting data that would disprove it.  If, for example, we found human fossils DEEPER in sedimentary deposits than any other primates, then this would suggest that humans lived before all other primates and thus could NOT be descended from other primates.  We TEST hypotheses by looking for BOTH contradictory and supporting evidence in an unbiased way.  So, we dig deeper into sedimentary strata in places we haven’t dug yet.  We don’t know what is there, so it is an unbiased search.  We could find human fossils (which would disprove the hypothesis), or we could find only other primates (which would support our hypothesis).  The KEY is that, in an experiment, both falsification and support is possible. In this way, questions about the past are testable, too; not just things that are happening today. Hypotheses are either supported or rejected by the experiment. When multiple results converge on a single explanation, that explanation, that explains all the existing patterns and makes PREDICTIONS about other patterns, is called a THEORY. So, the word "theory" in science does NOT mean "untested idea". We have a differnt word for that, remeber? (Hypothesis). Rather, "theory" means a broad, TESTED, explanatory model of how the physical universe works. They are supported by previous tests. They are used to make predictions of subsequent events in the physical world. But they are incomplete; they can be modified to become more precise. Consider these THEORIES in science:
              Physics: Atomic Theory  (until THIS YEAR, no one has seen an atom, but this theory explains how matter behaves)
              Astronomy: Heliocentric Theory (no one has stood outside the solar system, but this model predicts where the planets will be in relation to each other and the sun.)
              Chemistry: Bond Theory  (until 2009, no one has seen atoms bound together as molecules, but this theory predicts which the binding properties of chemicals)
              Biology: Evolutionary Theory (no one has seen a living dinosaur, but morphological, paleontological, geological points to a relationship with birds, and this predicts where subsequent fossils are found).