Darwin’s Dangerous Idea: Chapter 4

Dennett begins this chapter acknowledging that he used language that would need further explanation. "In order to sketch the big picture, I helped myself to several controversial claims, promising to defend them later." (85) So in this chapter, he is going to begin to lay out how this notion of 'Design Space' works in natural selection. As a reader, I'm looking for him to develop two lines of thought:

1. What physical evidence serves as the foundation for his model and historical account of how natural selection works?

2. What are the philosophical arguments for the necessity of naturalism as the ontology that must provide the foundation for natural selection?

Dennett begins the chapter by again reinforcing the claim about the nature of the evidence supporting the neo-Darwinian model: "There is no serious controversy about the fact that all the diversity of life that has ever existed on this planet is derived from this single fan-out; the controversies arise about how to discover and describe in general terms the various forces, principles, constraints, etc., that permit us to give a scientific explanation of the patterns in all this diversity." (86).

Instead of presenting or citing evidence, Dennett's description of the process appears to be built off of it. "The Earth is about 4.5 billion years old, and the first life forms appeared quite 'soon'…at least 3.5 billion years ago, and for probably another 2 billion years, that was all the life there was: bacteria, blue-green algae, and their equally simple kin. Then about 1.4 billion years ago, a major revolution happened: some of these simplest life forms literally joined forces…The second revolution--the emergence of the first multi-celled organisms--had to wait 700 million years or so."

The Tree of Life

Dennett's discussion of the structure of the tree of life consists of a mixture of what I assume to be summary information of physical evidence of some sort including current DNA evidence, and philosophy. He describes what the tree looks like and why it looks the way it does. His picture of the tree appears to be constructed from hard physical evidence and the rest filled in via the inductive method. But Dennett cites no evidence and offers no footnotes from which any evidence could be researched. His description of the tree of life includes statements like the following.

"What is the relation between time and Design, or what could it be? Could things that are more designed come first and gradually lose Design? Is there a possible world in which bacteria are the descendants of mammals and not vice versa? These questions about possibility will be easier to answer if we first look a bit more closely at what actually happened on our planet." (87-88)

"There must have been billions of failed design experiments, but perhaps none ever became very distant departures from a single unicellular norm." (89)

"Once sexual reproduction becomes established many millions of years later, up in the fronds of our Tree (and sex has apparently evolved [sic] many times, though there is disagreement on this score), if we zoom in and look closely at the trajectories of individual organisms, we find a different sort of juncture between individuals--matings--with starbursts of offspring resulting." (90)

Much of his description follows a similar pattern. This way of approaching history initially strikes me as either a summarized history or a just-so story and it’s tough to tell which it is. I recently read a history on the revolutionary war. It was written in narrative style in the form of a novel. The historian, of course, didn't reference all his sources every time he cited a fact yet I took the novel to be a fairly accurate historical account of what happened. The author regularly quoted from diaries, journals, newspapers, letters, documents and summarized the rest from other documents that existed at the time and put a "story" or "picture" together. However, other than a bibliography, he didn't cite a source for every "fact" he referenced in footnotes or endnotes nor would I have expected him to. There are histories that do this (Paul Johnson's A History of the American People for example endnotes fairly heavily). But a history-novel doesn't need to do this to constitute a good history.

Histories like this differ greatly from "histories" where there is little documentation and evidence to go on and the author interpolates from what little evidence there is to essentially create the story based on what might have happened given the evidence that does exist. So the bulk of the story, instead of being extrapolated from the evidence is created and is not factual in the sense of being directly derived from the evidence but largely comes from the mind of the author. Sometimes this yields fairly strong results other times the results are questionable. I recall when Oliver Stone released his movie JFK, controversy surrounded the movie because much of what Stone included in his version of who killed John F. Kennedy was not based on hard evidence. To this day, Stone's take is largely supposition and is not considered factual though there are many that would say his story is true. Many are saying the same thing about the recent, popular novel by Dan Brown, The Da Vinci Code.

I can't tell, from what Dennett gives in this chapter, whether his history is of the first or the second type. It seems from the very nature of the case that it would fit more in the latter category than the former simply because the type of evidence that one would need to create a factual history couldn't exist. When Dennett says, "There must have been billions of failed design experiments" or, when discussing the tree of life, "Perhaps it should be shown with various dead-end branches large enough to be visible…marking various hundred-million-year experiments in alternative unicellular design that eventually all ended in extinction." or Since the questions of what a species is, and how a species starts, continue to generate controversy, we can take advantage of the God's-eye perspective we have temporarily adopted to look closely at the whole Tree of Life and see what would happen if we tried to color-code a single species in it." I get the impression that he's doing some substantial editorializing. I can't tell from what he's written how many of his claims are based on fact and evidence and how much he's interpolating to create the picture he wants his reader to adopt.

Again, this seems reasonable. My only point of contention is around how the theory is presented. If one is going to claim that evolution is as factually verifiable as, say, the earth going around the sun (as Sagan and Lewontin famously asserted), then I expect the evidence to be commensurate with this claim.

Identifying Species

The general theme of the section on the Tree of Life is to explore how one would go about identifying a species. Since there are no natural kinds such that there are necessary properties or essences that a given species possesses, determining the boundaries of a given species becomes a bit difficult and the boundaries themselves become loose. Regarding the choices the scientist faces when attempting to determine when speciation has occurred, Dennett writes, "There are no secret facts that could tell us which choice is right, which choice carves nature at the joints, for we are looking right at the places where the joints would have to be, and there aren't any." (92), Further, he writes, "there are, again, no hidden facts no essences that could settle the issue." (93). Certainly this seems right on his model. Species are on a continuum not in silos according to Darwinian theory.

Dennett believes Darwin settled the issue. Though species are on a continuum, the only way we are able to settle the species boundary issue is because the intermediate forms have gone extinct and the surviving forms bracket species nicely! "As Darwin pointed out, if it weren't for the separations that time and the extinction of intermediate stepping-stones has created, although we could put the life forms into a 'natural arrangement' (of decent), we could not put them into a 'natural classification'--we need the biggish gaps between extant forms to form the 'boundaries' of such classes." (93-94). So we're able to determine species separation because biological life has nicely and naturally fallen into what appear to be natural kinds with enough genetic and morphological separation to make species determination easy. In realty though, this separation is only an appearance as all the confusing details that might make species separation difficult have been removed from the picture due to time and extinction.

As an aside, if there isn't physical evidence for the intermediates and speciation is derived from the extant forms, then the physical evidence--the raw data if you will--is entirely consistent with what essentialists claim. One does not, then, derive the existence of these intermediate form from the raw data alone but from the ontology that one brings to the table when examining the physical evidence just as one would derive the existence of essences from the raw data alone (though, it seems to me, the absence of hard evidence for intermediates seems to prima facie support essentialism). One is saying that the "story of the intermediates" has to be true even in the absence of physical data of these intermediates because essentialism is false. But that appears to be question begging of the worst kind.

Dennett ultimately concludes that there is no hard and fast way to define a speciation event on his (or Darwin's) model and that seems right. He does, however, want to wrestle with the fact that we do have hard and fast species--real species do exist. So how did they come about? Dennett has some suggestions on how to solve the problem.

Mitochondrial Eve, Isolationism, and Genetic Reproduction

Dennett begins his story by talking about the original female genetic ancestor of all human beings. "Mitochondrial Eve [ME hereafter], is the woman who is the most recent direct ancestor in the female line, of every human being alive today." (97) This woman, Dennett says, probably lived in Africa less than three hundred thousand years ago and maybe less than half that. (98). So there is a clean break between ME and forms prior to ME such that ME formed (or helped form), genetically, a new species, namely, homo sapiens. How did the break between the forms before ME and ME take place? One answer according to Dennett might be this: Geographical isolationism along with genetic staying power. "When has speciation occurred? In many cases (perhaps most, perhaps almost all--biologists disagree about how important the exceptions are), the speciation depends on a geographical split in which a small group--maybe single mating pair--wander off and start a lineage that becomes reproductively isolated." (99) So a genetic mutation occurs and a single mating pair becomes geographically isolated. This geographical isolation sets the genetically different pair off from the rest of the population and the genetically isolated pair is unique enough to create a new species. Well, not so fast.

"Has speciation occurred? Not yet, certainly. We won't know until many generations later whether or not these individuals should be crowned as species-initiators." (99) Why? Because, "there is not and could not be anything internal or intrinsic to the individuals --or even to the individuals-as-the-fit-into-their-environment--from which it followed that they were--as they later turn out to be--the founder of a new species." (99-100). Essences are a chimera. So "whether or not the individual who has that mutation counts as a species-founder or simply as a freak of nature depends on nothing in its individual makeup or biography, but on what happens to subsequent generations--if any--of its offspring." (100) So being able to pass on that positive mutation is also important for a new species to emerge.

So if I understand the picture here, Dennett is saying that when a new species emerges, three things must happen.

1. A positive genetic mutation must occur.

2. The individual in whom the genetic mutation occurred must pair up with a mate with whom it can interbreed and that pair must become geographically isolated from the rest of the population.

3. That mutation must be passed on to future generations and become a consistent part of the genetic makeup of that lineage.

This process must be repeated over and over for a new species to occur. In between step 1 and 2 above, all we might end up with is a variety of the same species. However over many generations and more mutations and isolations occur, the genetic difference becomes so great that we really can say we have a new species. So the individual in whom the initial mutation occurred can only be said to be the species initiator in that it caused the chain reaction to start.

What we have on this side of history then are these large gaps of extant forms with significant genetic differences so determining species is much easier. If we were able to witness the intermediates, we wouldn't be able to tell when the difference actually occurs at the time the mutation happens ("speciation is…hard to witness" as Dennett says on page 99 and acknowledges on page 100 that even Darwin never witness speciation even in his observation of human breeders).

From an evidential point of view then, if speciation is hard to witness as its happening what physical evidence is there that Dennett's speciation story is true from the historical record? What evidence is there that Eve and Adam "wandered off?" What evidence is there that this happened over and over and over? What evidence is there of these hundreds, thousands, perhaps millions of intermediates? Dennett provides no direct references or citations so I assume he's basing his story on evidence he doesn't present in this text. I assume that the historical evidence not only supports the current extant genetically distant forms but also the intermediate forms that existed between the gaps. Why is this evidence important?

Well, it seems to me that if the physical evidence for these intermediate forms doesn't exist (and based on the evidence given in Dennett’s book I don't know if it does or not), then the existence of these genetically distant forms only along with the physical evidence that only supports these distant forms is consistent with essentialism. Dennett's story then becomes a matter of interpretation of the evidence and a just so story rather than a "hard facts" story as Dennett seems to imply that it is. Why, then, interpret the hard physical evidence the way Dennett does rather than the way the essentialist does? One reason, possibly, is because one comes to the evidence with an ontology that can only support one interpretation of the physical evidence: it just has to be true. It then becomes a matter of the philosophical foundation one brings to the table rather than a matter of the hard physical evidence that demands a given interpretation. I wouldn't mind so much if it were presented this way but many times it’s not.

To put the matter a bit more concretely, suppose one comes across two computer programs that are similar in morphology (they have many of the same external features such as similar menus, similar colors, similar functionality), but the code underneath is substantially different. Two external observers attempting to figure out the progeny of these programs draw different conclusions. One concludes that the programs came from a single code base that developed bugs and anomalies over time (for whatever reason) and at some point in its history the code got copied and the two copies became separated (physically) from each other. The code of each of the "branches" continued to change until they became the codically different programs they are now.

The other observer says no, these programs never had a unified code base. They were developed separately at different times even though they have very similar external (phenotypic) characteristics. The reason for the different code is due to the fact that the programs were developed separately and the external similarities are due to the function for which the program was developed but not at all due to a unified source that existed sometime in the past.

How does one decide which story is true? Well, one way might be to look at what physical evidence exists for each story. Suppose that in doing the research, both observers find that there is a history for each of the programs but that history only includes versions of the code for each of the branches such that the versions in each branch are only slightly different from the extant version and there are no versions of the code of the "intermediate" versions that the first observer says existed in the history of the code base for both programs. So essentially, the only history that is uncovered turns up ancestors of each of the programs but the ancestors are very close to the extant versions and there are no intermediates with substantial differences code-wise.

We might also stipulate that there are other, more primitive programs that were found that have moderately similar morphology and some similar code but the morphological and codical difference between these rudimentary programs and the ones under consideration are substantial.

Could the first story still be true? Sure it could. But it doesn't seem one can decide whether or not the first or second story is true solely based on the physical evidence. There would have to be either other evidence or a philosophical reason or reasons to conclude one is true over the other. Certainly the lack of intermediates would seem to put the burden of proof on the first observer but neither story could be said to be a priori true without other considerations. At this point, Dennett has not supplied enough information for me to determine whether this evidence exists. It may but Dennett hasn't given his reader a reason to think it does yet.

Branch Shapes

Surprisingly, Dennett says that there's an even "more interesting" question than the one we just looked at. Questions about the shape of the branches of the Tree of Life and, even more interesting than that, "the shapes of the empty spaces between the branches" are really where the serious work needs to focus. He concludes that it may not be possible to answer such questions. "Eyes have evolved independently in dozens of lineages, but feathers probably only once. As John Maynard Smith observes, mammals go in for horns but birds do not. 'Why should the pattern of variation be limited in this way? The short answer is that we do not know.'"

Then in a candid statement, Dennett goes on to say, "We can't rewind the tape of life and replay it to see what happens next time, alas, so the only way to answer questions about such huge and experimentally inaccessible patterns is to leap boldly into the void with the risky tactic of deliberate oversimplification. This tactic has a long and distinguished history in science, but tends to provoke controversy, since scientists have different thresholds at which they get nervous about playing fast and loose with the recalcitrant details." (101)

I'm not sure what to make of this? Is Dennett saying that we don't have the history of the intermediates and can't trace the progeny of extant (and admittedly genetically and morphologically distant) forms? Is he saying the existing physical evidence doesn't provide us with that "tape" that we might use to trace a thing's history? Is he saying that all we have are the gaps and because of this we have to simplify the story because we can't dig down into the details? Are the details recalcitrant because we don't have physical evidence to figure out what those details are? It's hard to know what Dennett is saying here. If one takes his words at face value, it should certainly raise questions about the "factual" status of evolution and if not that, it should at least temper the sometimes harsh criticisms of essentialism.

He does give some hint in that his answer to this issue he raises is to look at the "large patterns" and try to figure out what the "right level" of explanation is appropriate and adequate to explain these patterns. (102) He gives an example: why do giraffes have long necks? "There is one answer that could in principle be "read off" the total Tree of Life, if we had to look at it: Each giraffe has a neck of the length it has because its parents had necks of the lengths they had, and so forth back through the generations. If you check them off one by one, you will see that the long neck of each living giraffe has been traced back through long-necked ancestors all the way back…to ancestors who didn't even have necks. So that's how come giraffes have long necks. End of explanation." (102-103)

This "explanation" could be considered adequate as a history of the long-necked giraffe I suppose if the physical evidence for this history is strong and there are good reasons to believe that the no-neck giraffe and the long-neck giraffe are from the same historical line because of incontrovertible physical evidence to link them. I'm not sure if this evidence exists and Dennett gives no good reason to think it does. But if it does, it may provide a history of the development of long necks of giraffes but I don't see how it’s really an explanation.

But even more importantly, it doesn't provide us with a speciation or genesis story which seems to be the more relevant problem. It doesn't help us understand how the common ancestor of giraffes became elephants or monkeys or humans and this is the kind of evidence that really becomes essential for the Darwinian story to work. Variation in a species seems consistent with just about any worldview. What physical evidence exists to support this more grand story?