Nature Podcast 19 April 2007

This is a transcript of the 19 April edition of the weekly Nature Podcast. Audio files for the current show and archive episodes can be accessed from the Nature Podcast index page, which also contains details on how to subscribe to the Nature Podcast for FREE, and has troubleshooting top-tips. Send us your feedback to


The Nature Podcast from Nature, the international weekly journal of science brought to you in association with Bio-Rad. Multiplex your SPR results and increase your throughput with a ProteOn XPR36 protein interaction array system from Bio-Rad Laboratories. On the web at

End Advertisement

Ben Valsler: This week — is realism really real? Markus Aspelmeyer explains why it may not be.

Markus Aspelmeyer: People have thought before that as soon as you allow 'spooky' actions, these actions at a distance, that then you can explain everything.

Ben Valsler: Also, we will be discovering how the continents have grown....

Stephen W. Parman: The continental crust grew in pulses and these will be large, tons of extreme volcanic activity, probably more extreme than we have ever seen in the modern geologic record.

Ben Valsler: ....and when it was that they first went green.

William E. Stein: By the end of the Devonian we had representatives of all the major classes of vascular plants already in established ecosystems that we could pretty much view as being modern.

Ben Valsler: I am Ben Valsler and welcome to the Nature Podcast. First, to that age-old philosophical question of whether things are still there when you are not looking at them. The idea that matter has fixed properties, regardless of whether it is being observed or not, is known as 'realism' and it has long been a standard assumption in traditional physics. Now, Markus Aspelmeyer from the Austrian Academy of Sciences has been testing whether realism also hold true in quantum systems. Nature 446, 871–875 (19 April 2007) .

Markus Aspelmeyer: So, the question that is at stake is whether quantum theory is a complete theory. The point is that quantum theory only makes probabilistic predictions for outcomes of measurements in a physics experiment and the big question is, can one go beyond that? So, is there a more complete description, a sort of underlying structure, in nature that quantum theory still has not crossed, but that could explain as only seemingly probabilistic events? It seems rather a philosophical problem, but it turns out that you cannot have such a hidden structure in nature and have at the same time no action at a distance.

Ben Valsler: So, by "action at a distance", you mean the effect where two particles become entangled and then observing a property in one particle immediately affects the other particle, regardless of how far apart they are?

Markus Aspelmeyer: Yes, it is this affecting which would be the action at a distance.

Ben Valsler: And the main problem with this is that it would imply that the effect can travel faster than light?

Markus Aspelmeyer: Exactly, yes. That is an obvious conceptual problem, but even if you would allow that certain actions could travel faster than light and would therefore account for these correlations in entangled systems, you could not allow, in addition to that, for an arbitrary realistic view of nature. People have thought before that as soon as you allow spooky actions, as soon as you allow these non-local interactions, these actions at a distance, that then you can explain everything by assuming a hidden structure. This was the status quo of the field before our work.

Ben Valsler: So, when you say a hidden structure, what you mean is that when we are not observing something that something is still definitely there?

Markus Aspelmeyer: Yes, exactly. So that you have properties of particles, for example, that are definitely there, but that our physical theory that we have, namely quantum theory, just cannot describe exactly. This is what I mean with saying that the question is whether quantum theory is a complete theory or not.

Ben Valsler: So, a complete theory would have to account for everything that we can observe and also for everything that we can't observe?

Markus Aspelmeyer: Exactly. People often refer to this possibility of hidden structure simply as 'physical realism', as a philosophical viewpoint, because hidden structure means that simply things or properties exist independent of the observation. So, that is basically a philosophical viewpoint of realism.

Ben Valsler: How did you devise a test for this?

Markus Aspelmeyer: What we are doing in the experiment is we create pairs of photons that are entangled, in specific properties of polarization. So that means when you measure polarization of one photon in the experiment you will always find the polarization of the other photon to be connected to the measurements on your first photon over arbitrary distances. We performed measurements along very specific directions on one particle and along very specific directions on the other particle, and then we investigate the correlations between the results that we get. If quantum theory is correct, then the correlations that you observe cannot be accounted for by assuming that these particles have the property of polarization and just interact via spooky action.

Ben Valsler: Okay, so is it possible to tease the joint assumptions apart and test each assumption separately?

Markus Aspelmeyer: Right now, no; however, we already know from Bell's theorem that we cannot have a hidden structure together with no actions at a distance. Now, we know with this new result that, at least for this specific kind of realism, we cannot even have action at a distance. So, no action at a distance plus action at a distance means that basically you exclude fully for this specific class of realism the possibility that action at a distance is at work at all. So, these results together already tell you that we do have a serious problem with the existence of a hidden structure at all.

Ben Valsler: And what have you been able to conclude?

Markus Aspelmeyer: What we were able to conclude is that, for this specific kind of hidden structure that we assume, not even actions at a distance can account for the correlations that we observe in an experiment. So, that means that if we want to explain now experimentally observed quantum phenomena by a complete description that describes everything in terms of realistic entities or properties, we cannot any longer assume a very naive realism together with just non-local interaction. This is simply a too-naive picture. There are two ways out at the moment— basically, either that there is no realism that physical theory would refer to, or that if it is something like that it is definitely not this naive picture that we have in mind.

Ben Valsler: Markus Aspelmeyer showing that the quantum world is fundamentally random rather than real. Next, how did the land beneath our feet form? The debate is being raging for years over whether the continental crust is constantly forming or is punctuated by bursts of growth. Now, Durham University's Stephen Parman may have found the answer by looking at isotopes of the noble gas helium. As the ratio of helium-3 to helium-4 can accurately tell us when a piece of the Earth's mantle melted, we can use this time signature to see when any given part of the continental crust was formed. Researchers can now use this technique to probe continental evolution. Earlier this week, Dr Parman explained to Azi Khatiri how this process works. Nature 446, 900–903 (19 April 2007) .

Stephen W. Parman: The Earth's crust is very interesting because there are two types of crust. There is the crust that we live on which we call the continental crust and then there is the crust that is under the oceans, which is the oceanic crust, and we know a lot about the oceanic crust. We know fairly well how it is formed and its age is under 200 million years, which is young for the Earth. But the continents, some parts of them are very old, up to four billion years old and the age of the Earth is four-and-a-half billion, so most of history of the Earth is recorded in the continent. And the two primary questions about the continental crust are how it is formed and when did it form. In my paper the question I was trying to answer is when did the continental crust form.

Azi Khatiri: So, how have you tried to answer these questions?

Stephen W. Parman: For a long time people have gone around the continental crust and dated the surface of the Earth. An interesting observation is that in fact there are certain ages of crust that are over-representative, there are peaks in the distribution and those ages are 1.2, 1.9, 2.7 and 3.3 billion years old, and there has been a debate that has been going on about what this means. And there's essentially two opposing views: one that these peaks mean that the continental crust was formed in pulses at those times, and the other side of the debate is that in fact it has always been growing the same amount of crust and that these peaks are artificial and it is just a crust in between these peaks have been destroyed and removed from the record and essentially this debate you could not solve it by just looking at the crust, and what my paper does is actually does not look at the crust at all. I am actually looking at the mantle, which is the interior of the Earth, and we think the continental crust was extracted from the mantle by melting and that the crust is essentially those melt. Then, theoretically it should leave an isotopic imprint in the mantle that record the time at which it was extracted.

Azi Khatiri: So, how do you go about testing this?

Stephen W. Parman: We have many isotopic systems, strontium, neodymium, lead, and people have looked for this sort of pattern and have never found it and that is because these isotopes and elements are in the oceanic crust. These pieces of crust get put back into the mantle and they take all their strontium and lead with them and then that mixes back in and it kind of obscures the isotopic signature we are looking for, but helium is pretty much unique, and that when magmas erupt on the surface of the Earth, the helium gets out of the rock because it is a gas, and then it goes into space, and so when the oceanic crust breaks it essentially has no helium in it — and so it turns out helium has the clearest signature of these ancient melting events. What the paper shows is that there are exactly the same peaks in the helium isotopes that we see in the continental crust and that very much supports the idea that the continental crust grew in pulses and these would be large, tons of extreme volcanic activity, probably more extreme than we have ever seen in the modern geologic record.

Azi Khatiri: How does this helium tell you about how the crust was formed?

Stephen W. Parman: To understand that, you need to know the decay chain for helium and the decay scheme is that uranium and thorium are radioactive and they decay and their daughter product is helium-4 isotope. So, helium-4 is building up in the Earth all the time. There is another isotope of helium, which is helium-3, and it essentially is not being produced and so I am looking at the helium-4 to helium-3 ratio. So what happens is when you melt the mantle, all of the uranium and thorium goes into the melt and so helium-4, which had been building up in that piece of mantle that got melted, now that stops, because uranium and thorium are gone and so that helium-4 to -3 ratio gets frozen in and so when we see that helium-4 to- 3 ratio in a rock, we can say, aha, we know that that piece of mantle must have been melted at this age.

Azi Khatiri: So, where does that helium-3 come from? You said it is not produced, but where does it actually come from?

Stephen W. Parman: It was inherited when the Earth was made. Interestingly, we think we know the initial 4-to-3 ratio of the Earth, I smile when I think of how people noticed. They know it because they sent a satellite into Jupiter's atmosphere and measured the 4-to-3 ratio. Now, there is almost no uranium and thorium in Jupiter compared to how much helium it has. So, it has not built up any helium-4 more than it had when it started and has not broken up any helium-3. So, its 4-to-3 ratio has not changed since the beginning of the Solar System and we assume that that was Earth's initial value, and it turns out that when I make this correlation between helium and mantle and the continental crust age peaks, it projects back exactly to this value and that was kind of the last piece of the puzzle, the most convincing part of this study.

Azi Khatiri: So, given how much is already known about the way Earth's continents have formed, how are your results contributing to that knowledge?

Stephen W. Parman: Well, I think the main one is the debate about whether the crust formed continuously through time or in pulses and just looking at the continent itself you could not settle that debate. But now looking at the mantle, finding essentially the opposite side of the coin, that if the continents are formed by melting, then we should see the signature of melt removal from the mantle and that we do find that matching signal. I think that is the main contribution to understanding how the continental crust is formed.

Ben Valsler: Stephen Parman telling Azi Khatiri how isotopes of helium can be used to probe the origins of continents and land masses and support the idea that the continental crust was formed during periods of violent volcanic activity.


Nature's podcast, bringing the world of nature to life.

End Jingle

Ben Valsler: In just a moment, we will be off in search of the world's oldest trees and also finding out what meteorites can tell us about their past, but first, how can you tell just from a glance if your pancake is cooked or when a road is icy or oily? Isamu Motoyoshi has been finding out what it is about an image that enables a human brain to make such judgements. Nature advance online publication (18 April 2007) .

Isamu Motoyoshi: We were basically interested in how the human observer can perceive the lightness or glossiness of natural objects.

Ben Valsler: How did you try and study this?

Isamu Motoyoshi: We first made randomized surfaces, then we measure their images with our tabulated colours, then we measure human observers' responses to those surfaces.

Ben Valsler: So, by observing the basic statistics you could get from looking at an image, at the light reflections and luminosity, you try to find out which statistics humans themselves use to determine qualities of surfaces, is that right?

Isamu Motoyoshi: Yes.

Ben Valsler: What is it that you actually found?

Isamu Motoyoshi: From the analysis of images, we first found that the skewness in the luminance histogram of the two-dimensional image could be a really nice skew for distinguishing between a light and a dark object to a glossy and a matte object. Then, when we next tried to measure the human responses to those surfaces, we also found a really high correlation between the shape of luminosity histograms such as skewness and the human responses for lightness and glossiness of the surfaces.

Ben Valsler: Could you describe what a luminance histogram would look like?

Isamu Motoyoshi: Basically, the luminance histogram is a kind of a graph that represents the frequency of pixel in the image as a function of the luminance of its pixels. Dark and glossy objects have a luminance histogram that is skewed to positive direction.

Ben Valsler: How did you conclude that the human visual system uses this information?

Isamu Motoyoshi: There have been two kinds of evidence that supported the human visual system actually utilizes the low-level statistics. One is that the skewness of the luminance histogram can be extracted like low-level visual mechanisms, which has been already observed by a number of physiological researchers. And another reason is that we found a sort of illusion, or after-effect, in the perception of the lightness or glossiness. We found human observers check the subsequent surface, glossier or more matte, after adapting to a certain image which had positive skewness in the luminance histogram. On the basis of these findings of after-effect, we believe that we can conclude that human visual system actually computes and utilizes the skewness in the luminance histogram to judge the glossiness or lightness of the surface.

Ben Valsler: You found that by artificially changing the skewness of the luminance histogram, you would get a predictable response from the people who were then looking at the image that you had artificially changed?

Isamu Motoyoshi: Yes, exactly. We considered that this experiment can ensure that there must be a certain closer relationship between the perception and the skewness in the luminance histogram.

Ben Valsler: And so what are the implications for this, now that we know that this particular statistic is used in human visual systems. Are there any future applications we could see for it?

Isamu Motoyoshi: Yes, our finding I believe will enable us to develop devices or programs that can enable us to manipulate or control or simplify the image that gives light to a certain impression of macular property.

Ben Valsler: Isamu Motoyoshi of the NTT Communication Science Lab in Japan on how the human visual system uses some very basic image statistics to make judgment from surface qualities such as glossiness. By digitally altering the skewness of a luminance histogram for a given image, he was able to elicit a predictable response from human observers. This runs counter to the long-held assumption that the human visual system uses high-level statistics to make these decisions. Next up, a report from Jijin Yang at the University of Massachusetts describes how the microstructure of a meteorite can tell us a lot about its history and from that we can see that all the meteorites so far studied have come from the Iron-rich inner core of asteroids from the main asteroid belt between Mars and Jupiter. The asteroids in this belt are thought to have arisen from collisions between the protoplanets, which formed within the first one-and-a-half billion years of our Solar System's existence. They were thought to be about 300 kilometres across. Nature 446, 888–891 (19 April 2007) .

Jijin Yang: There is a controversy about how the asteroids formed in the past and the information we got from meteorites because meteorite is remnant from the asteroid.

: Could you clarify what these meteorites are actually made of?

Jijin Yang: A meteorite is made up of 90% iron, around 10% nickel. We have many elements like cobalt, gold, germanium. So, basically we call it iron–nickel meteorite.

: So, where would these iron meteorites have come from to start with?

Jijin Yang: Scientists believe they are from the asteroid belt between Mars and Jupiter. So, we have studied more than 10 meteorites. We found the cooling rate across these meteorites is not constant. So the trouble here is the traditional law about differentiating asteroids, that cooling rate across the cores to be constant, but what we found is not constant.

: Why is this cooling rate important?

Jijin Yang: The cooling rate gives us the history, basically the sum of history, and the size of the asteroid and basically also could give us collision history. For example, right now we have a meteorite on Earth which is mostly from the Kuiper belt. But we don't have any piece of mantle part of the asteroid. So, one most possible reason is the collision starts in separation of mantle and the core, and later on there is a lot of later impact, but the metallic body is more durable, more strong, but the mantle becomes very, very small because the silicates are meant to be brittle than metallic core, finally maybe it's too small for the mantle, we cannot find on Earth. So, our data show this is true because from the cooling rate of iron meteorites, we know the core and the mantle separate very, very early.

: So, the iron meteorites are actually originating from the pieces that would have been part of the core, but we don't really see anything of the mantle?

Jijin Yang: Yeah.

: How did you test this? How did you go about finding this result?

Jijin Yang: We can know the cooling rate by taking the microstructure and the microchemistry. Slow-cooled iron meteorite has several very beautiful microstructures which we can observe using optical microscopy and electron microscopy. The cooling rate of an iron meteorite can be determined by analysing the section with an electron microscope and at the same time we need to complete a model to reproduce this data.

: You have collected samples from these meteorites and you are looking at the chemical structure in the actual meteorite and you are also using computer modelling in order to reproduce what you are saying?

Jijin Yang: Absolutely, yeah. For the last 40 years, people tried to understand why we have non-constant cooling rate. They are several other models proposed that cannot explain the correlation between the cooling rate and the core composition. Based on the data, we put forth another model to say if there is no mantle surrounding the core, in that case we can have non-constant cooling rate, and subsequently we can get a correlation between the cooling rate and core composition. That is the beauty of our model.

Ben Valsler: Jijin Yang telling Anna Katiri how knowing the cooling rate of a meteorite can give away its origins and that only the strong Iron-rich core of asteroids survived to land as meteorites on Earth.JingleEnd Jingle

Ben Valsler: Lastly for this week, I spoke to William Stein at the State University of New York at Binghamton about the exciting find which has enabled him to identify the world's oldest trees. At Gilboa in New York State, 380-million-year-old remains of tree–fern-like cladoxylopsid trees have been found and these plants represent the first fossil evidence of forests and some of the ecosystems that we see today. Nature 446, 904–907 (19 April 2007) .

William E. Stein: It is a discovery. You never know when you are collecting specimens as a palaeontologist what is going to show up and this is an example of where nature provides us something that you would not have ever expected to have found and so you take advantage of that.

Ben Valsler: So, what have you found?

William E. Stein: We have actually finally figured out what Eospermatopteris was, these great trunks that we've known about for many years, so we now know what their identity was and pretty much what they looked like as trees, the earliest trees.

Ben Valsler: So prior to this discovery, what did we know about earliest trees?

William E. Stein: The ages of trees are roughly 360 or 380 million years. These stumps have been described and have been known for about a 100 years. They have been discovered throughout the Schoharie Valley. And the problem with them is that many of them are rooted in place and very large, but we had no idea what they represented, what major group they belong to or what they were. Yet, they were the oldest. So, it was very interesting to know pretty much what is going on.

Ben Valsler: And before these trees evolved, what plant life would be found?

William E. Stein: The Devonian period is often called the age of fishes, even more importantly can be considered the age of greeting of the Earth. Prior to the Devonian, in the Silurian period, we have pretty scant evidence of quite small plant-like organisms on the surfaces of the Earth. By the end of the Devonian, we have representatives of all the major classes of vascular plants, with the exception of flowering plants, already present and probably in already established ecosystems that we could pretty much view as being modern. So, the Devonian period is the time during which all of the basic terrestrial ecosystems become established and the basic forms of plants become established as well. So, these plants representing the first large evidence of forest in the fossil record are of considerable interest.

Ben Valsler: What happened to leave these particular samples so well preserved?

William E. Stein: The original Gilboa stumps were probably the result of periodic and multiple events of flooding preserving the lower parts of the plants, where they were embedded in the sandstorms or the floods themselves, and then they were truncated at the top as a result of actually being broken during the flood probably, or at least were rotting shortly thereafter and they were filled in with sand. They were sandstone casts. The specimens we have are actually compressions. So, the original organic matter is preserved in these things and they were preserved as part of a flood event where the plants were already uprooted and actually deposited in the bedding plains of the ancient delta where they were preserved.

Ben Valsler: So, what do we now know about these trees?

William E. Stein: Well, we know that the members of the cladoxylopsids have been of intense interest in this group currently throughout though, well quite a few people are studying these things. They represent some of the largest early plants and we now have a real clear idea of what they look like. They were very palm-like in their morphology and they were quite a bit larger, perhaps, than we had suspected previously.

Ben Valsler: So, how could the development of the tree shape have affected the environment and the ecosystem?

William E. Stein: They had a major main stem and then a crown of branches that were actually leafless. You can imagine perhaps a spotty, patchy kind of network of these trees and perhaps different sizes forming a canopy, much like a tree-fern forest, but not as dark as modern forests. At the bottom of these trees, you might recognize arthropods of various types, perhaps early tetrapods, at least by the end of the Devonian period.

Ben Valsler: So, would a plant of this shape create large amount of leaf littler for other species to take advantage of for shelter or for food?

William E. Stein: Well, as you are observing these things, as if they have a crown consisting of modular units that are acting very much like large fronds in at fern. This is a surprise. This is a new finding here for us. That would indicate that they actually dropped these large modules and that most of what represents the fossil record are these separate branches and they probably dropped a lot of it, indicating that there is a significant amount of leaf litter or branch litter, which again comes as a surprise.

Ben Valsler: And does this offer a greater understanding of the environment in which trees evolved?

William E. Stein: It turns out that this shape is very recognizable. We see it in many different groups of plants. Not all of them can be traced to the same common ancestor. This is a recurrently evolved life forms of plants, which suggests that there is some fundamental ecological shaping of these forms for reasons that are actually pretty unclear at this point.

Ben Valsler: So, what do you think the future of research in this area will be?

William E. Stein: Well, you always wish to find more of these things. There is a lot yet to be learned about the cladoxylopsids, especially there has a lot of work being done in the middle to late Devonian throughout the world, and we have just touched the surface about what these plants really represented and how they actually may have structured the first forest.

Ben Valsler: William Stein describing the miraculously preserved cladoxylopsid trees, which made up the world's first forest 380 million years ago, the first time a tree-shaped plant had evolved. That is it for this week's Podcast. I hope you would join us again next week, when we will be discussing the basis of addiction and the cappuccino. Until then, please send any comment or feedback to For more science in the meantime, this week's Naked Scientist's Podcast explores new directions in the science of cancer. That is available for free from the Naked Scientist's website at This week's show was produced and presented by me, Ben Valsler, and Anna Lacey, with additional production by Sabina Michnowicz. Until next time, good-bye.


The Nature Podcast is sponsored by Bio-Rad, at the centre of scientific discovery for over 50 years, and on the web at

End Advertisement