Tuesday, April 16, 2013

On Truth and the Limits of Knowing

The Australian comedian Tim Minchin has a routine that consists of an 8 minute beat poem called Storm about a girl he met at a dinner party. Minchin, who has made a name for himself as a vocal Skeptic (I capitalize because there’s something of a cult of skepticism that has arisen in recent years, and because calling them a cult will piss them off) was deeply offended by Storm’s new age-y beliefs, which – to paraphrase – stemmed from a belief that life was a mystery that we cannot fathom. Yes, we can, says Minchin, and goes off on a diatribe about how what we discover about life through the application of the scientific method is just as beautiful and elegant and lyrical is anything that can be imagined through superstition. Or something like that – you can watch the routine here,

I was thinking about Storm the other day, because I began to wonder if I was inadvertently playing the role of hippie in a discussion about the limits of scientific understanding. I tried to discuss this with my wife - who had dragged me into the whole thing by enlisting me to review course materials on phylogenetic systematics to be supplied to students and teachers - and after several minutes of my trying to explain myself, she snapped. “If I actually though you cared about this, I would never have asked you to do it.” So that was the end of that. Or so she thinks. Ha! Because when I get annoyed about something, I go and blog about it - so now the whole world knows what I think. That’ll show her!

Anyway, the job in hand, as I said, involved course materials for high school students and teachers about phylogenetic trees. Or, as the writers put it, The Tree of Life. Note how I capitalized again, because this is The Tree of Life we’re talking about, like the One Ring to Rule Them All, or the Word of God. Hey kids, when you do phylogenetic systematics, you’re not just figuring out how things are related to each other. You’re helping to build the One True Tree of Life, the Tree that links you to wombats, priapulid worms, and the fungus that grows on your skin when you have poor hygiene. How Sick is that!

As you can tell, I’m a bit jaundiced about the Tree of Life.

Let’s be clear: the idea is a very important one. All living things on this planet are linked to each other by varying degrees of common ancestry in a pattern that is dictated by the process of evolution. By mapping the pattern of shared characters between different organisms, you can reconstruct that shared ancestry and hence the processes that generated the pattern. And you can visualize the pattern as a series of nodes and branches that looks kinda like a tree, which is a powerful metaphor to use when talking to people about evolution, because trees are a living, naturally generated structure, and because we also use them – in genealogy – to talk about our own immediate ancestry.

Metaphors are dangerous things, because when stretched they can take on a life of their own. “Genes” are a good example of this. When I say “gene” in this context, I don’t mean what a molecular biologist would mean – a stretch of nucleic acid, corresponding to a unit of inheritance, whose molecular sequence codes for a functional polypeptide chain or RNA. I mean it in the selfish, “Dawkins” sense, which is a phenotypic difference between two organisms that has a heritable basis, and which affects the fitness of the organism.  I have a bigger tail than you, females dig guys with big tails, I make more babies than you, and my babies have bigger tails too. Voila, natural selection in action. Or, to put it simply, I have the “big tail gene.”

Of course, there’s probably no such thing as “a” big tail gene. More likely there are lots of  different genes, whose combined affect is to give me a bigger tail. But when you’re trying to explain something complicated, like kin selection and the evolution of altruism, to a lay audience, simplicity is vital. Hence “The Selfish Gene,” Richard Dawkins’ 1976 masterpiece, one of the all-time greatest pieces of scientific writing and the book that made me want to be a zoologist. I try to remember that when I see the author today, a wild-eyed fundamentalist ranting away about God on the TV and calling otherscientists “Quislings.” What on earth happened to the guy that wrote that book? It’s hard to believe they’re one and the same.

Like most people, my education in science followed a well-trodden path. Each year you were taught something new. Then, at the beginning of the next year, you were taught that what you learnt in the previous year was not exactly true, because at some level you weren’t ready for the truth.  A vicar friend of mine (yes, I am a scientist who has friends who are vicars. I am a Quisling!) referred to this, in the context of bible study, as “milk and meat” – you start out on milk, but as you get older and wiser you’re ready for some scriptural “meat.” Why is it that everything evangelical Christians say comes out sounding inadvertently sexual? But the fact that science and bible study have such commonalities suggests that that maybe the Templeton Foundation is on to something (“Quisling!”)

Unfortunately we live in a time when it’s all too easy to call someone a liar for telling you less than the complete truth. This is the narrow path we have to tread when we develop pedagogic materials for science. Hence my concern over phylogenetic trees versus The Tree of Life. A phylogenetic tree is just a hypothesis about evolutionary relationships, based on available evidence. It cannot “be” the evolutionary history of the group of organisms in question; most of them are missing due to the incompleteness of the fossil record. It’s also essentially subjective, because you pick the characters that you analyze, and – if you’re looking at anything more than a handful of them – you run them through phylogenetic analysis software that makes a bunch of assumptions about how evolution works.

This is not to say that phylogenetics is in any way bad science. But neither is it infallible. Any tree can change with the emergence of new evidence. Sometimes it’s a newly discovered organism, recent or fossil. Sometimes it’s the inclusion of a known organism that had previously been excluded from the analysis. Sometimes its because of a new character set, a new analytical technique, a change in the assumptions underlying the software algorithms that are being used to reconstruct the phylogeny. Some relationships are more stable than others, being robustly supported by multiple lines of evidence, but nothing is completely certain.

When I put this to one of my colleagues, she expressed concern that if we didn’t give people the sense that there is real information in the trees, they might not care about this at all. I understand this perspective, but it’s a dangerous one. As scientists, we are comfortable with uncertainty. In fact, for many of us, uncertainty is what drew us into science in the first place. By our nature, we question everything, repeatedly, and this questioning drives our understanding forward. Unfortunately, most people outside of science want certainty. They expect us to be able to provide them with answers – if I eat this, will I shorten my life expectancy: yes or no? And they don’t like it when the answer is “maybe.”

But this does not mean we should give them certainty where none exists. To use an example provided by the same colleague, it is true that we are probably not going to change our minds next year and decide that gorillas are more closely related to us than chimps. But neither can we say with certainty that we won’t, if evidence emerges to support the alternate relationship. Twenty years ago, you wouldn't have found many scientists who thought whales and hippos were closely related, but now it's a widely accepted hypothesis. This may be frustrating to the lay reader, but it’s also science. We go where the evidence takes us, which is why we differ from creationists.

From uncertainty it’s a relatively short step to the unknowable, and this is where a lot of my colleagues get really uncomfortable, because we’re talking about the limits of science. Tim Minchin may think “Storm” is a dumbass for believing that there are aspects of the human body that are a mystery, but where paleontology is concerned there are all sorts of mysteries. We are stuck in what my graduate advisor used to call “evolution's epistemological gap” – we can see the end result of evolution, in terms of the diversity of species, and we can analyze and manipulate the processes that may have generated that diversity at a micro-evolutionary level, but our lifespans aren’t long enough for us to directly observe one leading to the other. We have a lot of powerful supporting evidence, but the most critical component – the fossil record – is hopelessly patchy.

In some cases, we can fill in the gaps with new fossil discoveries, and these often change the evolutionary narrative significantly. When I was a kid, I was told that we “knew” that our distant ancestors were fish with lungs that developed 5-toed limbs from their fins. They were able to survive out of water because of their lungs and the fins helped drag them back to the water, and so evolved into limbs. Very neat and also (probably) not true. We now know that things were more complicated than that; between 400 and 360 million years ago there were a bunch of different animals that were not quite fish and not quite land animal, that developed feet with far more than five toes, and that had gills, not lungs. The first feet may have had more to do with moving in an aquatic environment than the terrestrial one. And we will continue to find fossils that fill out the gaps in this story, in sometimes surprising ways.

Or maybe we won’t. As I said, the fossil record is patchy. The circumstances under which an animal is preserved are very specific and very rare. The likelihood that erosion exposes that animal at a time and place where someone can discover it is equally rare. And fossilization does not capture all of the evidence about the organism. I may have an exquisitely preserved oreodont skeleton, but I can’t tell you what color fur it had, what sounds it made, what it smelt like, or how it digested its food. And given the circumstances under which these animals are usually found, which do not allow for the sort of exceptional preservation seen in other fossil deposits, I may never be able to tell you. I don’t know about you, but I find that really exciting.

The “Tree of Life” is a bit like the Rudolph Zallinger murals that we have on the walls of the Peabody Museum. At first sight, the Age of Reptiles and the Age of Mammals look like panoramic views of the evolutionary history of these groups. But they are not. We now know – because of new evidence - that many of the animals pictured looked nothing like their reconstructions and would not have done the things that they are shown doing. Many of the species pictured together are actually separated by millions of years, and sometimes tens of millions of years. Even when they are contemporaries, the animals shown represent only a tiny fraction of the ones that would have been present in the environments pictured, let alone all those that have ever lived.

Does this mean the murals are useless? Not at all. Even now, they remain a powerful tool for teaching about evolution. As I’ve discussedelsewhere in this blog, The Age of Mammals is one of the best summaries of the effects of global climate change that one could ask for, while the Age of Reptiles shows us how our ideas about dinosaurs have changed as a result of the work of paleontologists, while still being an awe-inspiring work of art in its own right.

The Tree of Life is the same thing. It is a powerful metaphor for the interconnectedness of life, a good way to visualize our current understanding of how different organisms are related to each other, and a starting point for developing hypotheses of how evolution might have generated this pattern. But the idea that its neat pattern of nodes and bifurcations completely encapsulates the infinitely complex, intricate, and painstakingly slow process by which populations of millions of individuals change over time to give rise to new species is no more correct than saying that Zallinger’s mammal mural is a photographic record of what life actually looked like in the Oligocene.

When I first mentioned my concerns to my wife, she said (in a tone of barely suppressed outrage) “Are you telling me that you don’t believe in the Tree of Life?” (No, I’m not kidding – she really did say that). And yes, at some abstract level, I do. It’s even possible to believe that one day we might be able to build a complete Tree of Life – that we will describe every species of life on Earth (in a way that accounts for individual and population-level variability); that we will find all the fossils that we need to reconstruct the complete evolutionary history of every group that has ever lived; that we will develop unambiguous character sets that are devoid of homoplasy and immune to the effects of convergence; that we will build algorithms for phylogenetic reconstructionthat accurately reflect the true process of evolution. And that, having done so, we will stand back and admire our handiwork; a complete, entirely inclusive, 100% stable, and totally true Tree of Life. But to believe this will happen requires, dare I say it, quite a measure of faith.

I think I’ll stop there.

Monday, March 25, 2013

Nazi Cow

On the desktop of my computer there is a picture of a building. It’s an undistinguished piece of architecture – whitewashed walls, a red tiled roof speckled with lichen, and a wooden structure atop the ridgeline that aspires to be a belfry, but falls short. This is Coteford Infant School, in the Northwest London suburb of Eastcote, and I spent three happy years there from 1971 to 1974.

Or rather, this was Coteford Infant School. It’s been gone for more than thirty years now. A new school was built, just up the road, and the old one, surplus to requirements, was torn down. There’s a small cul-de-sac of brick town houses there now, called “Grooms Drive.”

When I look at the picture, with the haze of an early summer day and the lush green of horse chestnut trees forming a backdrop, I can remember the smell of new-mown grass from the playing fields and the excitement of knowing that the summer holidays were almost there. It brings a terrible, but short-lived pang of longing. I’ll never be that young again, and the school building, with all its memories, is gone for good.

Time marches on and the world changes. As a paleontologist, I am intimately aware of this. The collection that I look after, at the Peabody Museum, is a record of past lives in long-gone worlds, albeit a massively incomplete one. I may be intimately familiar with the bones of oreodonts, for example, but I will never see a living one.

Even if, by some miracle of science, I could conjure one up, it would be devoid of meaning in the modern world. Oreodonts are an evolutionary product of the world of the middle Cenozoic, a world without grasslands, where vast herds of oreodonts browsed on low-lying shrubs, in open woodlands that covered much of North America. This biome does not exist anywhere today and so neither do oreodonts.

This idea of a changing planet lies at the core of the work that my colleagues and I are doing on the Peabody’s new fossil halls. All too often, people imagine that the world of past geological eras was much like that of today, only with different animals. It wasn’t. There are no modern equivalents, for example, of the vast forests that existed at high latitudes 55 million years ago– dawn redwoods and cypresses growing in complete darkness for months each winter. The climate was different, the composition of the atmosphere itself was different. That world is gone. We can imagine it - incompletely and from a great distance - through the few fossils that remain, but we can never truly experience what it was like.

All this came to mind when I was reading a report by Carl Zimmer on the ongoing debate regarding “De-extinction” – the (mostly theoretical) process of reviving recently extinct species through modern genetic techniques. Carl’s article is, as ever, an excellent summation of the technical and ethical challenges of doing so and I’m not going to repeat it here – you can go read it in National Geographic. But it did get me thinking about how resistant some people are to the idea of change.

Trying to recreate extinct species is not a recent preoccupation. In the first half of the twentieth century, a pair of German brothers called Heck spent an enormous amount of time and effort trying to recreate the aurochs, the extinct wild ox of Eurasia from which today’s domestic cattle originated. The Heck brothers lacked the sophisticated genetic tools of today’s biologists, but they theorized that if all domestic cattle had the aurochs as a distant ancestor, it might be possible to recreate the species by selective breeding.

We know what the aurochs looked like because our Pleistocene ancestors painted its likeness on cave walls, because its bones can be found in museum collections, and – most persuasively – because it actually survived well into historic times; the last one died in Poland in1627. So armed with this information the Hecks set out to find the most “primitive” living breeds of cattle, pick the most aurochs-like examples of each breed, breed them with each other, and keep back-crossing until they ended up with something that looked like an aurochs.

I say “looked like an aurochs,” because the Heck cattle were not aurochs. They were still domestic cattle. They weren’t even all that aurochs-like; Heck cattle were barely half the weight of their wild ancestors and retained the chunky physique of animals bred for meat, rather than the athletic body of a wild animal that has to live by its wits and avoid predators. In our collections at Peabody, we have the skull of a male aurochs – the real thing, from the Pleistocene. Its horns are wider than the span of my arms. None of the recreated cattle come close to this.

The idea of back-breeding the aurochs, or at least something very like it, from domestic cattle has proved to be a remarkably persistent one – there is a modern program, called TaurOS that is still attempting to do so. One of the main arguments in favor of the effort is that recreating a large, native European ungulate would promote the restoration of the continent’s lost biodiversity. Large ungulates modify the environment through grazing, which impacts vegetation and hence the communities of animals supported by it.

Of course, humans modify the environment as well. We’ve spent the last 10,000 years turning the planet into a place that can support over 7 billion of us, and along the way we’ve eradicated tens, if not hundreds of thousands of species. The charismatic vertebrates that form the focus of the current de-extinction debate are only the tip of a very large iceberg.

So here’s my problem with de-extinction. I can understand throwing resources at the preservation of charismatic megafauna – such as tigers, giant pandas, or elephants - in the wild, because you are also preserving the habitat in which they live and the biodiversity that is associated with it. Whether any of these habitats is actually “natural” after centuries of tinkering around by our species, plus the management protocols that get put in place to preserve them today, is a question for another blog post by someone better informed than me.  But at least they are there and they are real.

Recreating a species that inhabited a world that’s long gone has none of these “added value” advantages. The vast broadleaved forests of the Eastern USA that were the home of the passenger pigeon are mostly gone. So are the Mascarene forests that were the home of the dodo, and the high latitude steppes that supported mammoths.

Could we recreate the world as it was before we became agricultural and started clearing the forests? The story of the Heck cattle provides a cautionary tale and, at the risk of providing a proof of Godwin’sLaw, it’s worth looking at. You see, the Heck brothers’ work attracted the attention of a rather sinister patron, in the form of Hermann Göring. He’s better known as the former head of the Luftwaffe and a relentless thief of artwork from all over Occupied Europe, but Göring was also Reich Master of the Hunt (yes, they really had a state position for a Master of the Hunt),  Master of the German Forests, and an advocate for the protection of endangered species.

Göring saw the recreation of the aurochs as the centerpiece of the Nazi’s plans to create a primal forest reserve that would cover vast swaths of Eastern Europe and be the home not just to the aurochs, but also bison, wolves, bears, wild boar and other large mammals. The human population of the area wouldn’t be a problem, because the Nazis would murder or deport them all.

This is an extreme example, but the question of where you put your newly “de-extinct” species is a real one. Even with the full resources of a genocidal state behind him, Göring’s forest preserve would not actually have been primary forest, any more than the Heck cattle were aurochs. It would have been an artificial environment, created by humans to fit a particular need – in this case, a hunting preserve for a small group of the Nazi elite. And when de-extinction advocates talk about recreating the “mammoth steppe” of Northern Eurasia, for example, they’re not actually restoring a natural environment. They’re just doing what our species has been doing for thousands of years – modifying the environment to suit our own needs.

To be fair, some “de-extincters” are quite open about this. The TaurOS project, for example, doesn’t claim that it’s re-creating the aurochs. Instead, it’s creating cattle that could fill the role formerly occupied by the aurochs and thus generate environments that are more biologically diverse through grazing. What I like about TaurOS is that it recognizes and acknowledges that farmland grazing also promotes greater biodiversity – the creation of the new cattle is intended to tackle the problem of farm abandonment rather than recreate some mythical “wildwood.” It recognizes that human landscapes have intrinsic biodiversity value as well. From my perspective, this is a much more honest position than trying to claim that you’ve resurrected something that was lost.

So maybe this is a problem of semantics. Would I feel differently if, rather than claiming that they were breeding wooly mammoths, the various groups of researchers working on this problem were to say “we believe there are important benefits for humanity in the creation of a diverse grassland ecosystem in Northern Asia. A key element required for the project, but missing from the modern fauna of the region, is a mega-herbivore, a role previously filled by the wooly mammoth. We propose to use genetic engineering to combine genetic material from mammoths and elephants to create a cold-tolerant proboscidian”? OK, it’s less sexy than saying “we’re going to de-extinct mammoths,” but it’s a more accurate description of what’s actually being done.

The problem is that I’m still not convinced that it’s worth doing. Suppose, for argument’s sake, that it was possible to recreate the passenger pigeon, as has been proposed, through a mixture of DNA sequencing and selective breeding. As Carl Zimmer asks, would the inhabitants of suburban America really want vast flocks of pigeons roosting in the cities and crapping on their cars? But you could fix this. You could, perhaps, create a modified passenger pigeon that lived in small flocks. Or that pooped less. Or that had poop that smelled of roses. In principal, you could genetically engineer a passenger pigeon for the 21st Century. But why would you? We already have lots of pigeons. The ones that are adapted for living in our highly modified, urbanized world are the ones that are still with us.

I’m a custodian of extinct species – that’s my job. Over the years I’ve had the privilege of preserving the mortal remains of dodos, moas, aurochs, passenger pigeons, Labrador ducks, wooly mammoths, thylacines, Christmas Island rats, giant ground sloths, great auks and a host of other creatures that are no longer with us. Perhaps that makes me a little jaundiced. But it seems to me that the de-extinction debate is a distraction from thinking about the challenges that we face in the world today.

I could, if I had a large amount of money, buy all the houses in Grooms Drive and pull them down; dig up the post-modern brick roadway and lay black tarmac, carefully source roofing tiles, window frames, and a battered wooden belfry; and plant horse chestnut trees. I could scour London looking for original 1960s vintage classroom furniture and hire a crack team of chefs to recreate the truly abysmal cuisine of school lunches from 40 years ago. I could recreate Coteford Infants School in every conceivable detail. But I would still be 47, not 7, and my pastiche would not be Coteford Infants ca. 1974. Time changes and I have to get over it and move on. As scientists, and as a species, we need to do the same.

[yes, I'm aware it's nearly a year since I last did a post, but this one got me going. With thanks to Carl, who persuaded me to write it down rather than just going for a cold shower and forgetting about it]