There is an important lesson here about what it means to “explain something.”

Rudyard Kipling wrote a children’s book of myths about the origins of animals titled Just-So Stories. In it he explains the origin of the elephant’s trunk, how the camel got his hump, and where the  leopard’s spots came from (they were drawn by an Ethiopian from the leftover black of his own dark skin, so that the leopard would better blend into the background when they hunted zebra together.) Clearly, making sense is not the criterion for truth. It’s very easy to forget this, when someone gives you a complex explanation and you get that “aha! I understand” feeling.  Human beings constantly confuse congruence with truth.

Sensible and false explanations are such a problem in science that the term “just-so story” has come to refer to any sort of explanation that fits the facts, but cannot be verified. Scientific theories are supposed to differ from literary criticism and other forms of creative writing by demanding explanations that are true. This means testing them against reality.

A crucial point here:  you can’t test a theory against the same facts that you used to come up with the theory to begin with. Of course a theory is going to fit the facts that inspired it! Instead, a theory — an explanation of something — needs to predict things that haven’t been observed yet. Prediction is the essence of science; it is the ability to say what will happen before it happens that makes it possible to “design” a bicycle rather than just gluing random objects together until they roll. If our aim is to come up with a true theory about evolution, we need to use the length of  the giraffe’s neck to make predictions about something else, something we can go check (repeatedly, if we are serious about testing the theory.)

This seemingly philosophical notion is incredibly useful for spotting subtle bullshit that sounds like science.

Consider, for example, the trial of a vitamin for preventing the common cold. Let’s say it’s even a controlled trial. One hundred volunteers are given Vitamin Z daily, while another hundred are (unknowingly) given a placebo. At the end of the study, the Vitamin Z group had the same number of colds. But, the researchers discover as they analyze the data, they had fewer headaches. Does this mean Vitamin Z prevents headaches? Not necessarily, because the theory “Vitamin Z prevents headaches” was formulated  by noticing a pattern, any pattern, then making up a story about how that pattern came to be. That doesn’t make the story true. And there will always be patterns. If the volunteers can suffer from hundreds of different ailments, then by sheer dumb chance the Vitamin Z group will be found to suffer from less of at least one of them. (Applied to controlled experiments, this notion can be made mathematically precise, by the way. See post-hoc analysis.)

Put another way, if you keep turning over rocks you will eventually find something. The whole point of a theory — an explanation, a model, a statement of the causal relationships of reality — is to say what you will find before the rock is turned over. Otherwise you only have a story that fits the facts, a just-so story.

I have found just-so stories to be most common in alternative medicine, economics, and evolutionary explanations of human behavior. If nothing testable has been predicted, then nothing has been “explained.”

]]> http://jonathanstray.com/we-were-wrong-about-giraffes/feed 0 http://jonathanstray.com/fight-global-warming-on-your-desktop http://jonathanstray.com/fight-global-warming-on-your-desktop#comments Thu, 19 Mar 2009 04:07:08 +0000 Jonathan Stray http://jonathanstray.com/?p=495

Or at least help us to understand it. Climateprediction.net is a large-scale scientific computing experiment, relying on individual computer users who donate their computer time for the simulation of tens of thousands of global warming scenarios. This is important because, lacking other Earths to experiment with, computer simulations are really the only way we can validate our existing models of climate change — and then predict the future with models we think are accurate.

The climateprediction.net project comprises three separate experiments – one to explore the model we are using, the second to see how well the models replicate past climate and the third to finally produce a forecast for 21st century climate. Each model that we distribute will be used for all three experiments.

Built upon the BOINC scientific computing framework oriignally developed for the groundbreaking SETI@Home project, Climateprediction.net relies upon hundreds of thousands of volunteer users who donate their spare computer time. All of these machines together are effectively one of the largest supercomputers in the world, and this allows previously impossible scientific studies. The Climateprediction.net scientific team can run not just one or a few climate prediction simulations, but hundreds of thousands. One study performed this way was the Seasonal Attribution Experiment:

We focus on extreme weather events that occur on a seasonal timescale, and in our current project we focus specifically on the United Kingdom floods of Autumn 2000 which occurred during the wettest autumn ever recorded, causing widespread damage and an estimated insured loss of £ 1.3 billion.

Half of the climate model simulations we run are of the Autumn 2000 period, specifically including within them the effects of human-induced climate change caused by the emission of greenhouse gases. The other half will simulate a representation of the the Autumn 2000 climate had there not been any human-induced emissions of greenhouse gases over the last century. By then comparing the results of these two simulated climates, and recording the occurrence of floods like that of Autumn 2000 in each of them, we can determine how the frequency of occurrence (or ‘risk’) of such a flood has changed, and therefore how much risk is attributable to human-induced emissions of greenhouse gases over the last century.

So far, so good. This sort of modeling has been attempted before. Here’s the fun part: because so much computing power is available, a statistical study becomes possible. Rather than looking for floods in a handful of simulated case studies, the CPDN researchers are running 10,000 simulations each of the climate with and without post-industrial greenhouse gas emissions. Each simulation is started with slightly different initial conditions, which are then amplified by the classic butterfly effect into completely different simulated histories. Not every simulation will result in floods in Britain in the Autumn of 2000; in fact, because the flood was a once-in-a-century event in any case, most of them won’t. But by counting the number of floods with and without humanity’s greenhouse gas emissions, we can determine how much our industrial activity has amplified the risk of large flood.

This is akin to an epidemiologist determining how much smoking increases the risk of lung cancer, and the massive number of Climateprediction.net simulations are like to having not just ten but 10,000 case histories. It’s a breakthrough, and it’s only one of several different Climateprediction.net experiments, which have now cumulatively run almost 400,000 different climate simulations covering 41,000,000 years of hypothetical history. It made me very happy to learn about this project; it gives me hope that climate scientists have unlimited access to one of the largest supercomputers that has ever existed, given to them freely by the citizens of (it says on the map here) 132 different countries. You too can help.

It is now possible to see what a person is looking at by scanning their brain. The technique, published last November by a team of Japanese neuroscientists, uses FMRI to reconstruct a digital image of the picture entering the eye, albeit at very low resolution and only after hundreds of training runs. Still, it’s an awesome development, and many articles covering this research have called it “mind reading” (1, 2, 3, 4, 5). But it really isn’t, and it’s fun to explore what real “mind reading” would imply.

When I hear “mind reading” I want psychic abilities. I want to be able to know what number you’re thinking of, where you were on the night of March 4th, and what you actually think of my souffle. This is the sort of technology that could be badly misused, as the comments on one blog note:

Am I the only one finding this DEEPLY disturbing? It opens the doors to some of the scariest 1984-style total-control future predictions. Imagine you can’t hide your f#&%!ng MIND!

Fortunately, we’re not there yet. Morover, if we did have the technology to read minds, we’d have much bigger societal issues than privacy to deal with. The existence of “mind reading machines” would imply that we possessed good formal models of the human mind, and that is a can of worms.

But back to today. The paper by Yoichi Miyawaki and colleagues describes a technique for exploiting retinotopy, the fact that certain areas of the visual cortex are direct “maps” of the retina. First, a series of 10×10 black and white test images are shown to a someone while their neural activation is recorded by FMRI. The responses to these test images are used to ascertain which areas of the visual cortex correspond to which areas of the subject’s field of vision. When the neural map is complete, it can be read “backwards,” going from neural scanner results to a low resolution representation of whatever the subject is currently looking at.

This is a long way from a tool for the thought-police. First, the algorithm requires training on each new person. Also, an MRI machine is a huge, expensive, complicated piece of machinery which requires the subject to stay very still over a period of minutes — widespread brain scanning is, for the moment, completely out of the question. But most fundamentally, the information recovered is nothing more than what the eye is currently looking at. You might as well just tape a digital camera to the subject’s head. The pictures would be a lot better.

What is it that we imagine for a mind reading machine? Perhaps a printout, in words, of every thought that goes through someone’s mind. But do people really think exclusively in words? What about their emotions, or their unconscious responses, or even the complete set of minor joint aches and temperature sensations all over their body? Or how about a video playback of the events of yesterday evening? Impossible, because that’s not how human memory works. When we think about it carefully, we realize that we have an extremely poor conception of what is actually “in someone’s head.”

Compounding this problem is the fact we can’t even say what’s in our own heads. We think we can, but we can’t. Decades of psychological experiments show that access to the contents of our own minds and the working of our own thought processes is very limited. Consequently, we cannot answer the question “what would a mind-reader read?” through introspection.

This is why, before we could build a mind-reading machine, we would first need formal models of a “mind.” We need the sort of mathematical models that one can manipulate with a computer, because computers will surely be intensely involved in any mind reading technology. If recent developments in linguistics and artificial intelligence research are any guide, these models will be huge, associative, and statistical in nature, nothing like the structured logic we think we possess. For example, Google translates web pages between different languages without using anything like formal grammar models.

In other words, we cannot “read minds” because we have very little idea of how minds might be stored on a computer. This problem is known in AI as “knowledge representation,” and we still know very little about it.

Good formal models of the mind, if possible, are the technological precursor to entire fields of information engineering, and this is why I’m not worried about mind-reading technology per se. We’ll get beneficial things like accurate machine translation and computers that respond to voice queries — no more fighting with software that just doesn’t understand what you want. (Think also of the possibilities for art and expression.) We’ll also get uncomfortable technologies like sickeningly effective advertisements that exploit behavioral quirks we didn’t know we had, and NSA-funded conversation snooping programs that make existing keyword scanners look like the toys that they are. Finally, it would be possible to use accurate human mind models for pure evil: imagine a computer virus that was designed to read your personal files and figure out how best to convince you that the Dictator was beneficent. All of this may sound very far-fetched, but we’re going to build these things if we possibly can: think of how much money Google makes from each percentage point of improvement in ad clickthroughs.

If the Japanese FMRI technique seems positively simplistic in this light, that’s because it is. They have read retinas, not minds. They are extracting a representation we already have abundant experience with: images. Saying that we’ve made a step towards reading minds is ridiculous; Thomas Edison might just as well have claimed to “record thoughts” when he announced the phonograph.

I bother with all of this both because I think science journalism is often done badly, and because I believe that it’s important to get hysterical about the right thiings. One comment posted to a video of the research reads, “this is the beginning of the end of free thought.” Perhaps the continuation of this type of FMRI research really will one day lead to the ability to determine what someone is thinking without invoking their consent, but torture already does that. To me, the ability to represent someone’s thoughts in electronic form has far greater implications than mind-reading per se, and this sort of FMRI research — as impressive as it is — contributes little to that enterprise.

The edit network for “telephone tapping” shows a bipartite structure, indicating that the topic is controversial (image from Brandes et al.)

An interesting new paper defines the “edit network” of a Wikipedia article by drawing edges to indicate that one person has deleted or restored text written by another. While it’s always fun to look at pictures, the surprise here is that we can verify that the resulting graph structure really does tell us something useful about the article. In this study, articles with a more “bipolar” edit network — meaning that the authors split into basically two camps who routinely undid each other’s edits — were also much more likely to appear on a manually-maintained list of controversial pages.

Although there has been previous work on network mapping of Wikipedia in particular (and of course volumes of work on social networks in general) I find this paper interesting because it tries very carefully to understand whether the pictures mean anything. Like all science, what you find depends on where you look, and the practitioner of network analysis has an absurd amount of freedom to define what a “node” is, what an “edge” is, and how the resulting graph is visually laid out (since the point of a map is a visual representation, it’s very important that graphical properties such as distance, size, color, etc. have the right sort of metaphorical relationships to the more abstract properties we are trying to understand.)  

For example, one might try to identify people who hold similar opinions by analyzing who “interacts” with whom and looking for clusters. In this case the nodes are people, and the edges are “interactions.” But what is an “interaction?” A hallway conversation? A co-authored paper?  An appearance on the same talk show? If we draw an edge between two people if they’ve ever stood in the same room, we won’t necessarily get a good map of who “agrees” with whom. It is therefore very important to make sure that your definition of an “edge” properly embodies the question you are trying to ask — in this case a question about “similar opinions.”  (The most cogent critique of my COIN Policy Author Graph made exactly this point.)

This is a real and serious methodological problem in network analysis, and it’s made worse by our preconceptions. Suppose we suspect that republicans and democrats read different books. We might look for a definition of “edge” that we can apply to sales or reading data, and choose the one that gives the cleanest separation of people into two distinct groups. This makes pretty pictures, but it’s not clear that we can learn anything from such an exercise: all we’ve done is thrown out all the evidence that didn’t prove the notion we had already decided upon.

Back to Wikipedia: the paper is titled “Network Analysis of Collaboration Structure in Wikipedia,” and was written by Ulrik Brandes, Patrick Kenis, Jürgen Lerner and Denise van Raaij, who wanted to know if the structure of the edit network for a particular article could tell us something about where that topic fits into a broader discourse. In particular, they decided to see if they could produce a numerical measure of an article’s “controversiality” by measuring how close the graph is to being bipartite, that is, whether the authors tended to split into two camps, each of whom routinely deleted words written by the other. It is important to note that both the text processing that mines the data and the selection criteria which define an edge are very complex — meaning that other criteria which may not give “good results” are effectively excluded from this study.  For this reason, it’s very important to have something to compare their graph metrics against, and they do: the manually-maintained list of controversial pages.

To test whether high values of the bipolarity indicator  point to controversy in authors’ opinions, we computed the bipolarity of articles linked from the page Wikipedia:List of controversial issues, with our hypothesis being that bipolarity is high on those controversial articles and lower on non-controversial ones. … To compare controversial articles with non-controversial ones that did receive enough attention, we have chosen so-called featured articles which are listed on the page Wikipedia:Featured articles. 

…

The bipolarity index of controversial articles is [statistically] significantly higher than the bipolarity of featured articles. Thus, the controversy of topics is indeed reflected in the edit behavior on the associated Wikipedia article. 

Okay, but how did I end up on this side and not that side?

I went through this in Russia last year, when I was hosted in Moscow by a global warming skeptic; apparently it’s politically popular there to deny global warming, which sounds like a slight to Russia except when you remember that it’s politically popular here, too. But anyway, I was plunged headfirst into the debate with an ambitious little snot of a web-startup wannabe millionaire (“You should see our new offices! The Mafia used to operate out of there! They still visit someimes.”) Running through the arguments in great detail (as I previously reported here) I was forced to ask the very pertinent question, why do I believe that global warming is real, and man-made, and a serious problem?

The quick answer is that I believe the Intergovernmental Panel on Climate Change reports, but that’s also just ink on paper. Why do I trust them?

It has to do with process. To begin with, I know what the IPCC process actually is. They have devoted almost as much dead tree to how they reached their conclusions as to the conclusions themselves. In short, they collected something like 600 climate scientists from 40 countries, locked them in a library with a complete and current set of all relevant academic and scientific publications, and threw raw meat at them through the bars until they reached consensus. Actually, that’s not quite how it happened. Some of them were vegetarians.

The 2007 Fourth Assessment Report was then further reviewed by another 600-odd people, corrected, argued over, politicized, and finally published. Although there is no way to guarantee against bias in the author and reviewer selection process, at least a very diverse range of viewpoints could be expected to be represented, and at least the people involved have some reason to know what they’re talking about, having spent significant chunks of their lives asking questions about the ecosystem. This is as global and as sincere an effort to answer a question as humanity has ever seen, and it was all meticulously open and transparent.

There is a moral here.

One of the great things that thorough education teaches you — any education — is just how deep the rabbit hole of knowledge goes. It’s a smart person who realizes how big and complex and subtle any real discipline is; and I am absolutely at a loss to answer the tricky questions of someone else’s field, be they about global warming, the effectiveness of acupuncture, or whether cutting taxes will really help with unemployment (or not.) The only truly universal approach, our only hope for living in a world too big for reason, is to learn to evaluate how any given body of knowledge decides what is true and what is not true. In painful depth and detail.

The method, philosophy, and process of coming to believe: that is everything. I can’t say I even understand this process in myself, let alone an entire civilization, but I can say with conviction that it’s my favorite field of study.

The North Sage said that he had learned through meditation that each person was connected to the cosmos. The South Sage said that his people had developed powerful tools that could penetrate the heart of the invisible. The North Sage insisted that all knowledge would come from within. The South Sage asked how that could be possible, and claimed that one could only truly learn from observing nature.

Neither was stupid enough to insist that the other was wrong. The South Sage understood that if he had been born in the North, he would have learned all that the North Sage had. And the North Sage could imagine forgetting everything he knew; he saw that only with a beginner’s mind would he ever be able to  comprehend the wisdom of the South.

They stayed many days at this crossroads, on the mountain top. Each day they sat in the shade of an enormous old tree which was neither the native Willow of the North nor the sweeping Banyan of the South. They talked all day, then each returned to their own camp at nightfall.  The North Sage watched the stars and meditated. He sought not understanding but clarity. The South Sage wrote by moonlight in a huge old book. He wrote not what he had learned, but questions he had discovered.

After many months, it came to this.

“You,” said the South Sage, “you believe that all answers come from within. In your life you have found this method a far more reliable guide than mere tabulation of nature. But you cannot convince me that this method is better, because all your proofs come from where I cannot see them.”

“And you, my friend,” said the North Sage, “everything you have learned shows you that reason and study can unravel any mystery. But reason alone cannot show me that reason is all-encompassing, and so I must look elsewhere for deeper truths.”

“We cannot ever convince the other of our truths,” nodded the South Sage, “because we each ask for a type of proof that the other does not believe in.”

“What shall we do?” moaned the North Sage. “How can we resolve this dilemma? How can we learn from each-other? Must we stay at this crossroads forever?”

“No,” said the South Sage. “We are free to leave. We may each return to our homes. Or, you may continue walking South, and I may continue walking North.”

“Alternatively,” said the North Sage, “we could each walk in any direction we pleased, and hope to discover new lands to the East and West.”

“That is a good idea,” said the South Sage. “We will never know all the places in between, but it would help us both to understand the boundaries of the map. Do you think we will ever meet again, my wise friend?”

The North Sage blinked. “Of course,” he said, “for neither of us can escape this world.”

Recent research on physiological differences between liberal and conservative voters seems like a dream come true if you’re in need of a catchy press release, like this one from the National Science Foundation. I read the actual paper, and it says that people who answer more conservatively on a questionnaire about their politics tend also to have more pronounced “fight-or-flight” reactions to disturbing or surprising stimuli, as measured by skin conductance and startle response.

The press release tells a different story, and I believe that the NSF science writer told the wrong story. I attribute this partially to the politics of publicity, but mostly to the fact that science is actually very subtle, and hard to summarize.

“Those with the strongest eye or skin reactions to unexpected noises or threatening pictures such as a spider on a person’s eyeball tended to endorse political positions that were interpreted as protective of social groups,” said John Hibbing, professor of political science at UNL.

Hibbing defined those “protective policies” as more defense spending, more government resources directed at fighting terrorism and tighter controls on immigration. “People in this group are more willing to sacrifice a little of their privacy to protect the social unit,” Hibbing said. “On the other hand, the subjects who reacted less strongly to the stimuli were more likely to favor policies that protect privacy and encourage gun control.”

The first group believes the greatest threat to them and their communities comes from other people; they want to arm themselves and their government to defend against those threats. The latter group sees less threat from people and more threat from technology and inanimate objects such as guns that can kill or harm innocent people. They want policies in place to protect their individual privacy and safety: They oppose the death penalty and favor strong gun control. The study controlled for subjects’ gender, age and income.

What the NSF science writer has given us in the last paragraph is an interpretation of the results. This interpretation may come more or less from the researchers themselves, but it is still a story about what the experiment means, and that meaning is only possible with reference to existing ideas about the world. Notably, the researchers had decided that people could be categorized into more and less “socially protective” groups before the experiment was performed, and defined “socially protective” people as “threatened”  by others. Based on this idea, they devised a survey to measure overall levels of “social protectiveness” by asking each subject 18 questions on topics such as military intervention, the death penalty, etc. The researchers then chose the well-studied physiological tests of skin conductance and startle response as proxies for feeling “threatened.”

With these choices, it seems possible to answer the headlining question “does being easily threatened lead to more conservative politics?” But this is not the question the experiment actually tests, which is “do people whose survey answers put them in the ‘more socially protective’ group have stronger skin conductance and startle responses?” This may seem like semantics, but it’s actually slight-of-hand, an expansion of the narrow results into much more interesting generalities.

I think the scientists who performed the reasearch understand these problems — indeed that is their job, and a close reading of the paper suggests that the authors fully appreciate the epistemological and methodological problems involved. Rather, the problem here is that the NSF press release completely fails to communicate these subtleties, leaving the reader with a false (but sensational) result.

In fact, no experiment of this type can be used to confirm or deny the hypothesis that “being easily threatened influences political position.” A test like this one can only demonstrate that certain kinds of startle responses and certain kinds of survey responses occur together. It is unreasonable for the press release to speculate on which is cause and which is effect; this is the classic correlation/causation error, as discussed in the paper:

Our data reveal a correlation between physiological responses to threat and political
attitudes but do not permit firm conclusions concerning the specific causal processes at work. Particular physiological responses to threat could cause the adoption of certain political attitudes, or the holding of particular political attitudes could cause people to respond in a certain physiological way to environmental threats, but neither of these seems probable. More likely is that physiological responses to generic threats and political attitudes on policies related to protecting the social order may both derive from a common source.

A deeper problem is the identification of “political attitudes” with survey results, and “environmental threats” with the disturbing pictures and startling sounds used in the experiment.

What, exactly, is a “political attitude?” We can devise a survey that categorizes people based on their beliefs, but we can always draw a line down the middle of any group of responses and put people on one side or the other; we might as well choose “wears jeans” versus “does not wear jeans” as our study variable. At issue here is whether the thing actually measured corresponds to our intuitive notions of political ideology, and whether those intuitive notions are even representative of reality. There are ways of answering such questions — such as looking at the variability of results when somewhat different questions are used, or doing cross-validation against other measures — but I do not see that the researchers have applied them.

Similarly, does skin conductance and startle response actually measure “physiological response to threat?” There seems to be good evidence for this, based on years of research into the “fight-or-flight” response of the sympathetic nervous system. However, we have no idea if this same biological response is in fact activated when, for example, a conservative voter thinks about terrorism. That is the sort of information we might learn by, say,  hooking people up to physiological recorders and asking them to make political decision.

None of this means that the research is bad, or even that the line of questioning is unreasonable — I for one would be very interested to know if politics are partially physiological, perhaps even heritable (as other recent research claims to show.) The experiment as designed is an interesting step towards answering these questions, sort of a proof-of-concept test of the idea. But nothing has yet been answered. All we now know is that (some aspects of) physiology and (some aspects of) politics do seem to be realted, but exactly what is related and exactly what is the relationship? What do “politics” and “physiology” mean in a measurable sense, how do our pre-existing concepts of what these things are influence what we choose to measure, and what is the causality here? The research done to date creates far more questions than answers, and perhaps that is what the NSF press release most failed to communicate.

The “beauty myth” holds that attractiveness is a cultural construct, a control mechanism of the patriarchy, a way to sell us cosmetics we do not need. Modern social theory says that the beauty of the face and body depends on when and whom you ask, and that we could learn to love anyone if only our iron-maiden social norms permitted it. This is a pleasing thought. It says that accidents of ancestry do not truly count. Unfortunately, when we actually do the experiments and test nature versus nurture in the case of facial attractiveness, genetics wins. Facial beauty is, quite emphatically, not culturally defined.

The best survey is a meta-analysis by Langlois and colleagues, published in the Psychological Bulletin in 2000. They did an exhaustive search of the experimental literature on facial “attractiveness” and ended up including 919 separate experiments on how attractiveness is perceived, how we judge ourselves and others based on attractiveness, and how attractive versus unattractive people actually behave. It turns out that facial beauty is “real” in the sense that it does not depend on who you ask; ratings of attractiveness are very stable across ethnicities, cultures, age, and gender:

Both our cross-cultural and cross-ethnic agreement effect sizes are more than double the size necessary to be considered large, suggesting a possibly universal standard by which attractiveness is judged. These analyses seriously question the common assumption that attractiveness ratings are culturally unique and merely represent media-induced standards. These findings are consistent with the fact that even young infants prefer the same faces as adults.

If beauty was a cultural construct, we would expect different cultures to rate the same people differently. However, we might imagine that global media has erased the diversity that used to exist. This is why the experiments with children and babies are so important: if beauty is learned, then children would have to learn it, presumably beginning in early childhood. Yet the age of the beholder seems not to matter at all in subjective rankings of the attractiveness of others, and even newborn babies will look longer at photographs of people that adults have judged “attractive.” This cannot be learned behavior.

When we check the evidence, (facial) beauty standards seem to be universal and innate (biological). Doubtless, this is going to upset some people.

Here is what I think happened: we saw beauty used as a weapon. We saw women especially destroying their self-image under the assumption that they were ugly. Capitalism grew up around selling to our insecurities, and yes, the patriarchy said that girls were only as valuable as their looks. In defense, we decided that beauty was a myth. We tried to escape from these traps by telling ourselves that the object of contention didn’t exist.

But it does. Some people really are more beautiful than others, says our universal physiological nature. Accepting this implies a new social project. Rather than trying to teach ourselves that beauty does not exist, perhaps we should be trying to decouple value from beauty. It’s not that everyone is equally attractive, it’s just that attractiveness, like intelligence, doesn’t actually make you a “better” person. The corollary is that it’s worth understanding what roles beauty actually plays in human life, both cultural and psychological. What emotions does a beautiful face trigger in others? What assumptions are made? What does beauty actually do? While beauty may be objective, our reactions to it may still be learned.

The situation is somewhat analogous to the violent impulses that everyone is capable of: rather than pretending that anger doesn’t exist, we learn to deal with it in some healthy way. Similarly, it seems to me to be far healthier in the long run to acknowledge that beauty, at least facial beauty, is real and universal. To the best of our knowledge so far, this also has the advantage of being true.

Most people don’t even know that the world’s academic journals exist, but this is the master record, the huge source that all those science blogs and mis-representative popular articles draw from. These research journals are the collective output of every professional researcher in the world, in all subjects — only you’re not allowed to read them.

If you don’t believe me, go to scholar.google.com and type the name of your favorite object of curiosity. Then click through. Chances are you’ll get an abstract, but not more, not for free. The international university community has access to the full text, but if you as a member of the public wish to read an article in the latest issue of Science, Nature, The Physical Review or The Journal of Conflict Resolution, or thousands of others, you have to actually physically go to the nearest major university library and hope that they have an access policy for the general public, then look through the paper stacks… wow.

The foremost discoveries of our age, in every field, are locked up by a small number of scientific publishers, latched fast by copyright.

Basically, this is because it is the business model of academic publishers to charge for access. This made sense, way back when. Handling the time-consuming, detail-oriented process of peer review, then printing up those thick journals on archival paper and distributing them world-wide is an expensive process. In the era of physical libraries, it seemed reasonable to ask interested parties to journey to their nearest university to find a copy. Because each university serves a large community, these subscriptions are expensive — tens of thousands of dollars per year, for each of thousands of journal titles. When academic publishers started moving their journals online, instead of changing their business model they chose to artificially limit access.

At a biology conference last week I attended a discussion mediated by a woman from the journal Nature. She wished to explore the concept of copyright in science. She wanted to have a discussion about whether it would make more sense for the paper author, the university, or the journal to retain copyright. She was looking for a new business model. Bravo.

Except she missed the point. This is not the MP3 war all over again. You can make a really good case for rights management for artists on the assumption that they have to get paid at some point during the creation and distribution of their work. Even if musicians choose distribute their music for free, it’s still reasonable to consider it theirs in the sense that no one can use their song in a car commercial without permission. Copyright on a song still makes sense in the modern era, but academic and scientific research is not at all like music production. To begin with, it would be really, really hard to fund your research through sales of the resulting technical papers! All this research is also presumably performed for the benefit of all humanity, hence it actually becomes more valuable as it is redistributed.

My argument is that copyright as we understand it simply does not apply to communications of academic research. Copyright is a construction designed to forward the broader interests of society by providing incentives for creation, but it assumes that a) the creators can make some sort of living from selling their product and b) no real harm comes to humanity if those without money are denied access. Given that these assumptions are completely wrong for academic and scientific research, copyright is a total lose for scholarship.

This is the argument I made to the editor from Nature.

“It costs money to run a journal,” she pointed out. She’s right. It’s a lot of work to edit a peer-reviewed publication.

“The Public Library of Science manages to fund itself on page fees and philanthropic grants,” I replied. PLOS is a young open-access journal publisher, already well established with several high-profile bio and medical titles. They’re my heroes, an existence proof of the viability of open publishing. (I was handing out home-made “I heart PLOS” stickers at the conference. Nature Girl politely declined to wear one.)

“Yes, and we’re moving toward that model,” she said, “But who should control access to the redistribution rights? For example, if someone takes a paper and republishes it in a book and sells the book, should the author get a cut?”

“What do you think would be best for science?” I asked.

“Well, I think that we at Nature have always been devoted to the scientific community…”

“No, really, what do you think is best?”

She seemed suddenly confused. “You mean me, as a person?”

“Yes. What do you think is actually best for humanity here?”

Pause.

“I don’t know.”

I encouraged her to think about it.

In the bright Nepali sunshine, sipping chai in a tourist cafe overlooking the lake, I found I could not agree, no matter how cute the dreadlocked girl sitting across from me. I see a lot of idealism and projection in her arguments. I also see an iPod in her bag. But neither could I come up with a concrete reason to insist that technology is fundamentally good, that the human race should invest as heavily in technology as it has. I admit that I really enjoy both the intellectual playground of technology and the fruits it brings, but that’s no way to form a moral imperative.

Until Ethiopia. I was working on a trachoma epidemiology study. This is an ancient, simple disease, and so fragile that the merest hint of civilization will destroy it — we’re not quite sure why yet. It could be antibiotics used for other things wipes it out, it could be that just washing your hands daily in clean water prevents its spread. But if left untreated long enough, this feeble disease will make you blind.

I had the cliché moment. I hiked out across the roadless wilderness to that idealized little village, that tiny traditional portion of the way we used to live. The simple folk gathered round us, gazing strangely at our white skin and synthetic fabrics. In turn we stared at their traditional cotton garments and coarse shiny jewelry, artifacts of a society that makes everything with its own hands. We stood a moment in that field, contemplating one another across vast distances of education and context. Then I looked into the scarred corneas of a blind young man and felt suddenly: this sucks. This man cannot see, for no reason at all. Extremely simple medicine could have prevented that.

It’s one of those moments when you realize that you’re not okay with the world as it is.

Medicine is good because health is good. I see no other way to draw this conclusion. And medicine is technological. Antibiotics are in no sense natural, x-rays and heart transplants less so. Medicine is the moral justification for continued technological development and dissemenation. It’s the killer app for technology, because it’s not just medical technology that must be known: modern medicine requires an entire technological infrastructure to design and manufacture its many, many inputs. Computers. Polymers. Superconducting magnets. Refrigerators to make the ice to keep cold our collected samples, and enzymes to do the PCR to detect the trachoma DNA, mathematics to do the statistical analysis to determine if our mass antibiotic distribution is actually denting the epidemic. It takes a world to raise a hospital.

That’s the moral reason for continued technological development. That blind man. Go tell his mother that we’d all be happier as hunter-gatherers.

Of course, that’s not why we actually will continue to develop our technology.

In the late afternoon sunlight I lounged against a tree, waiting for the last few villagers to show up so we could test them. They had fed us some (traditional, natural, idealized) beer, and I was sleepy and idle. I extracted my MP3 key from my kit and put the headphones in, leaned back to something relaxed. A kid came up to me, looking expectantly. He must have been about twelve.

“MP3 player?” he said.

“Yeah,” I replied.

“How many gigabytes?” he asked. Then: “I want one.”

I find it hard to disagree with him.