lpetrich

Some creationists argue that the concept is a tautology ("similar features indicate common descent because they indicate common descent"). This is much like the argument that natural selection is a tautology ("those that survive, survive").

But I think that a reasonable definition is structural similarity that coexists with functional dissimilarity.

There are numerous well-known examples, but I'll focus on my favorite sort of homology: serial homology. There are numerous relatively trivial examples, like lookalike or trivially different body parts (hair or feathers or scales with different colors, for example); I'll mention some less trivial examples.

Flower petals are homologous with leaves, because they have similar structure, though different functions (attracting pollinators, photosynthesis).

Also, arthropod antennae, mouthparts, and legs have similar structures, despite having different functions (sensing, getting food into the mouth, walking).

One may wonder how different body parts get distinguished, but "homeotic mutations" come to our rescue. They make one body part develop like another -- and which develops like which fits clear patterns of serial homology.

The famous "homeobox" or Hox genes were discovered by studying homeotic mutations in fruit flies, like one which makes their antennae develop like legs (Antennapedia) and one which makes their mouthparts develop like legs (Proboscipedia).

Likewise, mutations in certain "MADS box" genes cause flower parts to develop like other flower parts or like leaves. This includes fairly obvious cases like sepals and petals -- and less obvious cases like stamens and carpels (pistil parts).

Serial homology suggests a simple path for the evolution of complexity -- making several copies of some body part, and then making different copies develop differently. Among arthropods, there is evidence of that sort of evolution caught in the act; early arthropods, like trilobites, have limbs that resemble each other much more those of many later arthropods.

The Hox genes themselves are an example of serial homology; they are recognizably related, and their position of expression exactly corresponds to their order on the genome. They also specify identity in a rearward fashion; non-expression of a Hox gene causes affected parts to develop like forward ones. This is consistent with a common form of embryonic development: adding segments at the rear end. So some distant ancestors would have added duplicates of Hox genes that control new rear areas as they emerge from segment addition.

Hox genes are widespread in the animal kingdom; they are homologous between fruit flies and mice, two well-studied model systems. Though mice have not one but four sets of Hox genes(!), which suggests some Hox-gene duplications early in vertebrate history.

Hox genes are only one example of "deep homology" in the animal kingdom -- the common ancestor of insects and vertebrates turns out to be surprisingly complex, having a central nervous system, a heart, a gut that extends the length of the body, eyespots, some sort of appendages, and so forth.

One particularly interesting bit of deep homology is dorsoventral inversion; the arrangement of internal organs between arthropods and vertebrates is upside-down!

A:
dorsal side
heart
gut
CNS, muscle blocks
ventral side

V:
ventral side
heart
gut
CNS, muscle blocks
dorsal side

This homology had been proposed in the early 19th cy. by Geoffroy St. Hilaire, and it would be re-proposed every 20 years, only to be picked apart. There were difficulties, such as the CNS having different architectures (arthropod: a ladder; vertebrates: a tube). But this hypothesis had made a comeback when it was discovered that the genes that control dorsoventral patterning are homologous between insects and vertebrates -- and in exactly that arrangement!

So dorsoventral inversion has had a continental-drift-style comeback; the hypothesis had languished a long time before important discoveries were made that supported it.

Deep homology sometimes takes odd forms. The distal-less gene gets expressed at the ends of several kinds of limbs across the animal kingdom; this suggests that limbs without any obvious homology share the same growth mechanism, suggesting that this mechanism can be activated at different places in a body.

How do creationists handle homology? Often by denying its existence, such as denying that land-vertebrate limbs are homologous to fish side fins, a denial which appears in some creationists' ridicule of "Darwin Fish". A similar denial is that amniote-embryo gill-like structures are homolgous to fish gills.

And as I have mentioned, some creationists have taken homology denial far enough to call the concept a tautology.

Others accept the reality of homology, but maintain that it is due to the creator choosing a common plan; this was the usual pre-Darwinian view. However, there are many cases where homology is far from trivial; an entity that can create from scratch may prefer to use a totally different structure in such cases.

Another pre-Darwinian view is that homologous features are different manifestations of some ideal archetypical feature, which sounds rather mystical.

[ August 31, 2002: Message edited by: lpetrich ]</p>

scigirl

Hey, I was going to post something this weekend about certain anatomical structures as proof of descent with modification.

One way to show that homologies are not just tautologies is to study embryology. If two structures are homologous - like a whale's flipper and a human's arm - then they should be derived from the same embryological structures (like the 4th somite of mesoderm for example). Careful cell tracing studies make these types of analyses possible (although not in whales! ) From what I learned in embryology, it seems that the data did support homology inferences made long long ago by Darwin.

Just like the chimp chromosome fusion example I love to talk about - the embryology not only supports homology, but also gives us a hint as to how the homologies are occuring. Sure, a creationist would say, "well the embryos are similar because the same creator made everything." But then why the amazing coincidence that homologous structures with different functions have the same developmental pattern, whereas analogous structures (that have similar function but are not directly evolutionary related) have different patterns?

Studies of various embryos along the evolutionary tree, to me, clearly show not only homology & conservation of structures, but evidence that the developmental systems themselves are homologous among evolutionary lines.

I wonder what the creationist answer is to embryological features such as tails on fetuses.

scigirl

Edited to add. . . that's human fetuses!

[ August 31, 2002: Message edited by: scigirl ]</p>

pz

I've seen several creationists wrestle with this question. For an example, see this <a href="http://www.christiananswers.net/q-eden/edn-c024.html" target="_blank">
creationist site on human fetus development</a>.

The answer is simple: denial. It's not a tail, it's a "fatty tumor". Never mind that it contains all the classic embryological tissues of the chordate tail.

The other approach is to lie. Pharyngeal arches are complex and undeniably present in human embryos, so what they claim is that evolutionists say that human embryos have gills -- which they don't, obviously. Unfortunately for their argument, evolutionists and embryologists don't say that. Human embryos have pharyngeal structures which have colloquially been called gill arches because of their similarity to and homology with structures that develop into gill supports in fishes, but they never develop gills.

tgamble

Denial.

tgamble

This claim is addressed at talkorigins
<a href="http://www.talkorigins.org/faqs/wells/#Vertebrate-limb-homology" target="_blank">http://www.talkorigins.org/faqs/wells/#Vertebrate-limb-homology</a>